FOOD FOR A HEALTHY PLANET II

UNIB20014

SEMESTER 2, 2016

Coordinator: Professor Mohan Singh

Director Food Systems Program, Building 142
mohan@unimelb.edu.au

Lectures: Lectures will be held in the

Sidney Myer Asia Ctr-B02 (Carrillo Gantner Theatre)
Mondays 1 pm 2 pm
Wednesdays 1 pm 2 pm
There will be 18 lectures
First lecture is on Monday 25th July 2016

Forums: Forums will be held in the

Sidney Myer Asia Ctr-B02 (Carrillo Gantner Theatre)
Mondays 1 pm 2 pm
Wednesdays 1 pm 2 pm
There will be 5 Forums
First Forum is on 31st August 2016

Tutorials: Tutorials will begin in Week 2

Beginning the week Monday 1st August 2016
Tentative Timetable 2016
Topic 1: Food Health and Nutrition - Candy, Steenkamp, Singh, Stevens (lectures 1 - 8)

Topic 2: Food Safety - Tribe (lectures 9, 10)

Topic 3: Food Economy and Politics - Freebairn (lectures 11 - 14)

Topic 4: Food Biotechnology & Food Issues Ford and Salisbury (lectures 15 - 18)

Week 1 Lecture 1: 25th July Lecture 2: 27th July No tutes

Candy Candy
Week 2 Lecture 3: 1st August Lecture 4: 3rd August Tute 1
Steenkamp Steenkamp
Week 3 Lecture 5: 8th August Lecture 6: 10th August Tute 2
Stevens Stevens
Week 4 Lecture 7: 15th August Lecture 8: 17th August Tute 3
Singh Singh
Week 5 Lecture 9: 22nd August Lecture 10: 24th August Tute 4
Tribe Tribe
Week 6 TEST: 29th August FORUM 1: 31st August No tutes
Wilson Hall/Uni Hall Martin
Week 7 Lecture 11: 5th September Lecture 12: 7th September Tute 5
Freebairn Freebairn
Week 8 Lecture 13: 12th September Lecture 14: 14th Tute 6
Freebairn September Freebairn
Week 9 FORUM 2: 19th September FORUM 3: 21st September Tute 7
TBA Stockley
HOLIDAYS
Week 10 Lecture 15: 3rd October Lecture 16: 5th October No tutes
Ford Ford
Week 11 Lecture 17: 10th October Lecture 18: 12th October Tute 8
Salisbury Salisbury
Week 12 FORUM 4: 17th October FORUM 5: 19th October Tute 9
Katalaris Cole
Tutorial Groups
Day Start Location

Tutorial 1 Monday 14:15 PAR-Sidney Myer Asia Ctr-113

Tutorial 13 Monday 12:00 PAR-Sidney Myer Asia Ctr-116

Tutorial 2 Monday 15:15 PAR-Sidney Myer Asia Ctr-120

Tutorial 3 Tuesday 10:00 PAR-Alice Hoy-101

Tutorial 4 Tuesday 10:00 PAR-Redmond Barry-1004

Tutorial 14 Tuesday 16:15 PAR-Redmond Barry-1004

Tutorial 6 Tuesday 17:15 PAR-Redmond Barry-1004

Tutorial 5 Tuesday 09:00 PAR-Redmond Barry-516

Tutorial 12 Wednesday 14:15 PAR-Alice Hoy-101

Tutorial 11 Wednesday 14:15 PAR-Sidney Myer Asia Ctr-116

Tutorial 9 Thursday 09:00 PAR-Alice Hoy-101

Tutorial 8 Thursday 11:00 PAR-Alice Hoy-101

Tutorial 10 Thursday 09:00 PAR-Alice Hoy-332

Tutorial 15 Thursday 12:00 PAR-Alice Hoy-332

Tutorial 7 Thursday 11:00 PAR-Sidney Myer Asia Ctr-G03

What is expected of students?
Hurdle: Must attend at least 7 of the 9 tutorials. Your attendance will be
posted on LMS.
Lectures: All lectures will be recorded on Lectopia and students are expected to
attend all lectures.
Forums: Forums will be recorded on Lectopia however the recording will be
released in SWAT vac. Students are expected to attend all Forums.
Tutorials: There will be a total of 9 tutorials. No tutorials in weeks 1, 6 and 10
Assessments: Students must be aware of the requirements and the due dates for
each of the components of assessments.
Readings: You will be given weekly required tutorial readings these
readings will be examinable.
Assistance: Students should seek assistance if things go wrong. Contact your
tutor, lecturer or the coordinator of this subject.
Submission of work: Submission of the Essay will be via LMS through turnitin

The subject covers Four Topics:

Topic One: Food Health and Nutrition - and , Steenkamp, n and Stevens
Topic Two: Food Safety - Tribe
Topic Three: Food Economy and Politics - Freebairn
Topic Four: Food Biotechnology and Food Issues - d and Salisbury

Forums
There will be a total of FIVE Forums. Forums are presented by invited speakers.
Forums will be recorded, however recordings will only be released in SWAT vac.You are
strongly encouraged to attend all Forums. Forums will be examined in the final exam.
 Assessment?
1 Mid-Semester Test worth 2 % on 29th August 2016 to be held during the lecture time slot in
Wilson Hall or University Hall
One 1000-word essay worth 1 % due nd 2016 by 10 pm
Two hours Exam worth 60% during the examination period. You will be permitted to bring
into the final exam one A4 page of notes. You can have notes on both sides.

Mid-Semester Test (25%)

Mid-Semester Test on 29th August 2016 during the lecture time slot from 1:05-2:05 pm. The
examination venue will be d n Wilson Hall or University Hall.

The Mid-Semester test will be on Topic One: Health and Nutrition and Topic Two: Food Safety.
You will be examined on lectures 1 to 10, tutorials 1 to 4 and tutorial readings required for
tutorials 1 to 4.
The test will comprise of 20 multiple choice questions and four short-answer questions.
Past tests will be provided on LMS.

Essay 1000 words (15%)

A research essay involves students finding out information about an issue and organizing
information into a logical, coherent discussion. Questions and arguments to do with the issue are
raised, information presented and the questions, and arguments evaluated.
Your essay is to be 1000 words in length and is worth 1 % of your final marks. Your essay is to
contribute to public debate and help inform policy-makers, about issues covered in Topic 1. More
specific information will be made available during tutorials and your essay is to be submitted
through turn-it-in on LMS by the nd 2016 by 10pm.

You are allowed to see the turn-it-in report and you are allowed to submit to turn-it-in multiple
times before the due date. Please be aware of our late penalties and plagiarism policies.
Final Exam (60%)
The final exam will be of two hours duration plus an extra 15 minutes for reading. This subject is
a Melbourne University Breadth subject which means that students are from a wide range of
backgrounds with varying skills and interests. For examination purposes students will be
expected to understand the important principles but not necessarily have mastered in depth the
detailed and complex underlying disciplinary knowledge.
The final exam will consist of 3 parts and will cover lectures, tutorials, forums and essential
readings. Part A is worth 30%, Part B is worth 50%, Part C is worth 20%

Part A: Multiple Choice Questions 0 n

his will cover a lectures, a a and a n a ad n
Part B: Short-Answer Questions you will be asked to answer 5 from a choice of 7 questions.
This will cover all lectures, all tutorials and all essential readings
Part C: Forum Questions you will be asked to answer 2 from a choice of 3 Forum questions.
a

Plagiarism
The University takes plagiarism very seriously. It is recommended you visit the university
Academic Honest and Plagiarism website on http://academichonesty.unimelb.edu.au/index.html

Electronic Submission and Turn-it-in

Report will be submitted through LMS via the on-line plagiarism scanning programme Turn-it-
in. This program checks to see how closely the wording of your work compares with reference
material and other student assignments.

Excessive length of essay

The written work should be concise as well as adequately covering the scope of the assigned
task. Essays must be at least 1000 words. However, up to 100 words in excess of the published
word limit will be accepted. Note that a deduction of 10% of the marks will apply to
assignments which exceed the allowed 1,100 words. Abstract, figures and tables are excluded
from the word count. Appendices are accepted where they provide supporting information
essential to the text.
Late Penalties
Any work submitted after the deadline date:
1-3 days late: loss of 10% of available marks
4-7 days late: loss of 20% of available marks
8-12 days late: loss of 30% of available marks and no feedback provided to student
13+ days late: automatic zero mark for that assignment.

Extensions
If you dont think you can hand in your assessment on time, you can apply directly to your
lecturer or tutor for an extension (make sure you speak to them before the due date). In general
extensions are permitted for unforeseen factors affecting your ability to complete a task on time.
The lecturer or tutor may advise you to also apply for special consideration depending on the
circumstances. Special consideration is available when illness or other causes:
significantly affected your university studies at any time during the semester;
prevented you from preparing for all or part of a component of assessment;
affected you during the performance of a component of assessment.

See the Universitys submission guidelines for Special Consideration. You can follow links from
this page to print out the HCAP or Statutory Declaration form.

You can submit your special consideration application through your Student Portal (Admin Tab

> Exams and Assessment).

 Writing a good essay
Sentences
Concept: the argument, the concept, the hypothesis
Context: why is it important?
Content: examples to support the idea (statistics, quotes, other studies)
Counterargument: a critical analysis of your main argument and why it is significant
Conclusion: summarize each paragraph

Paragraphs
Introduction: introductory paragraph should include a statement something like: In this
essay it will be shown . and then clearly state your core question/hypothesis.
Major point in defense of your argument
Secondary point in defense of your argument
Third and/or tangential point in defense of your argument
Conclusion: concluding paragraph should include a statement something like; In this
essay it has been shown., and then clearly state how you have answered the core
question/hypothesis.
Finally, spell check, then edit, then read your essay aloud, then re-edit.

Sources of Information
Recognize that you have many sources of information available to you. Lectures and tutorials will
provide you with some general knowledge and some specific information. You are expected to
conduct research to find the specific information you need, using the library and electronic sources
of journals (use the internet with discretion) Library websites are a good place to start. The
Language and Learning Skills Unit has a number of web- links to information that will help you
with hints on general aspects of writing and research.
For more on Writing reports and essay go to http://unilearning.uow.edu.au/main.html For Rules
and advice from the University Library go to http://www.lib.unimelb.edu.au/elib/techniq.html
For Search on-line go to http://www.lib.unimelb.edu.au/find.html
For Research skills go to http://www.lib.unimelb.edu.au/sg/tools.html
Referencing
Use recent information from a range of sources to show the breadth of your research Always use
credible publications (e.g. journal papers, books, DAFTA approved publications) to strengthen
your academic argument
Keep web-based material/ references to a minimum. (Not allowed to use Wikipedia as a reference
source because it is a blog site).
The source of photos, figures and graphs must be included in your reference list.
Choose a reference style and be consistent Harvard and APA styles are recommended
Always acknowledge the source of your information. This is done by including an in-text citation,
which gives the basic publication details (see Table 1). Secondly, you must include this citation in
your reference list (see Table 2).

Use of Quotes
Direct quotes should be used sparingly (5% of an assignment) if at all. It is best to summarize
and/or paraphrase information obtained from references, with proper in-text citation and
referencing. Failure to acknowledge direct quotes is plagiarism even if a reference is cited. Short
quotes (less than 30 words) should be set within the text, italicized and enclosed by single
quotation marks. Large quotes should be indented, italicized and placed in single quotation
marks. In-text citations with direct quotes should always indicate the page number(s) from which
the quote was taken.

Personal communication is a form of citation that should be avoided. The abbreviation pers.
comm. can be used in the in-text citation. The entry in the reference list must include the full
name, year, position title if known, exact date (day and month), and contact address (mail, email,
phone or fax) of the person providing the information. It is usually unacceptable to have more than
a few of these references.

Academic Skills Unit

The Academic Skills Unit helps undergraduate and graduate students with academic writing,
time and task management, oral presentations, exam preparation, English language development
and much more! ASU conduct workshops and individual tutorials and provide the resources that
all students need to be successful at university. Visit the Academic Skills Unit's website for help
with your essay. http://www.services.unimelb.edu.au/asu/
Table 1: Citation in text
One author
Two or three authors
(or editors)
Four or more authors
More than one reference
Unread reference
(where Williams & Pascoe is
the source containing the
unread reference by Martin)
Personal communication
Direct Quote
Internet
Table 2: References
Hard copy books One Elder, BJ 1995, The magic of Australia, Beaut Books,
author
Hard copy books - Two
authors
Hard copy books - Three
authors and 2nd edition
Edited book
Chapter in edited book
Journal article
One or more authors
Conference proceedings
from a website
White clover is an important component of dairy pastures
in northern Victoria (Jason 1996).
Or
Jason (1996) surveyed dairy pastures of northern Victoria
and
The found....
response to molybdenum has been demonstrated
in similar soils (eds Gilbert, Sullivan & Menzies 1995). Or
Hanna and Dickins (1991) demonstrated that .....
Growth rate is reduced at temperatures above 35OC
(Emerson et al. 1987) ... Or
Emerson et al. (1987) studied growth ...
.... this response has been demonstrated in a range of
environments (Jackson 1994; Jones 1992; Jones
1994; Julian & Alston 1992).
... (Martin 1982, in Williams & Pascoe 1994, pp. 20-25)...
.... there is little published information on
reproductive behavior in this species (Butler, pers. comm.
1996).
As stated by Campbell (1994, p. 279): a definition
of sustainability remains elusive.
As for other citations - author(s) and year.
Sydney.
Yeric, J & Todd, J 1989, Public opinion: the visible politics,
Peacock Publishers, Chicago.
Yeric, J, Todd, J & Muller, P 1999, Political perspectives,
2nd edn, Peacock Publishers, Chicago.
Danaher, P (ed.) 1998, Beyond the Ferris wheel, CQU Press,
Rockhampton.
Byrne, J 1995, Disabilities in tertiary education, in L
Rowan & J McNamee (eds), Voices of a margin, CQU
Press, Rockhampton.
Peterson, J & Schmidt, A 1999, Widening the horizons for
secondary schools, Journal of Secondary Education, vol. 3,
no.8, pp.89-106
Callaghan, S & Martin, B 2004, Igniting concern about
refugee injustice, paper presented at the Education and
Social Action Conference, 6-8 December 2004,
http://www.uow.edu.au/arts/sts/bmartin/pubs/04esa.html
 Lecturers
Professor Mohan Singh is subject coordinator. Professor Singh teaches in Molecular
Biology, Breeding and Biotechnology, Advanced Plant Breeding, Bioinformatics project
in Genome Science. His research interests are: Molecular biology of plant reproduction;
Transcriptome of flowering plant male gametes; Transcriptional regulation of male
gamete expressed genes; Molecular determinants of stem cells in legume meristems;
Genetic engineering of male sterility in crop plants; Genetic modification of Grass pollen
allergens for developing allergy vaccines; DNA repair genes in plants; and Expression
systems for Biopharmaceuticals.

Seona Candy is a research fellow with the Victorian Eco-innovation Lab at the University
of Melbourne with a particular interest in the integration of food and urban systems to
improve the resilience of cities. Her recent research work used scenario modelling to link
land and resource use with food consumption at a national level on an ARC Linkage
project, and at a city level with the Foodprint Melbourne Project. Dr Candy has a
background in engineering, renewable energy and international development. Her
PhD research focused on the use of appropriate technology to improve food security and
nutrition in remote mountain regions of Nepal. She has also worked on food security and
energy related projects with communities in Australia, Cambodia and East Timor.

Julia Steenkamp is a dietitian and food scientist. Julia has worked as a clinical dietitian
in the hospital setting as well as in private practice. During this time Julia managed
patients with obesity, diabetes, eating disorders, allergies and food intolerance. Julia has
also worked as a nutritionist for one of the largest food companies in the world. Over the
last six years, Julia's interest has been on the methods of securing safe and sufficient food
for the world and she recently completed a Masters in the area of food security and
sustainable rural development.

Dianne Stevens is a Lecturer in Nutrition and Dietetics and Community Domain Leader
in the Masters of Science in Dietetics for the College of Health & Biomedicine. She has
worked in clinical and community dietetics in Australia. Her particular interest is in
maternal, newborn and child health and nutrition in resource poor settings. Dianne has
gained considerable experience in managing programs in nutrition, food security and
health in international settings both in development and in emergencies. She has
experience in working with non-government organizations, the United Nations and with
donors.
Associate Professor David Tribe, is a senior lecturer in Department of Microbiology
and Immunology, University of Melbourne. Dr Tribe has expertise in genetics,
biochemistry, biotechnology, microbiology and food science.

Professor John Freebairn holds the Ritchie Chair in the Department of Economics at the
University of Melbourne. He received his tertiary education in agricultural economics
from the University of New England and the University of California, Davis. He has held
appointments with the NSW Department of Agriculture, ANU, La Trobe University, the
Business Council of Australia and Monash University before joining Melbourne in 1997.
John is an applied micro economist and policy analyst with current research interests in
taxation reform and environmental economics.

Dr John Ford is a marine and fisheries ecologist with a research interest on the coastal
habitat restoration and the sustainability of Australias fisheries. He combines expertise in
experimental and observational field ecology with a passion for our sustainable use of
marine resources. Dr Ford has spent the last eight years at The University of Melbourne
researching fish ecology, conducting over 500 scientific dives and has become expert on
the underwater world around Melbourne and the Victorian coastline. John is known
locally and internationally as an expert on seafood sustainability and has worked as an
assessor for the Marine Stewardship Council and the Sustainable Australia Seafood
Assessment Program. Current research projects include working with commercial fishers
to address seagrass decline in Corner Inlet, and restoring the lost shellfish reefs of Port
Phillip Bay. Dr Ford is also a regular presenter on RRR radios Sunday marine show
Radio Marinara and is a contributor to the goodfishbadfish.com.au website.

Associate Professor Phil Salisbury is a Senior Lecturer at the University of Melbourne.

Dr Salisbury teaches Advanced Plant Breeding and Biotechnology within the Master of
Agricultural Science.
 SCIENTIFIC AMERICAN

S U STA I N A B I L I T Y

CAN WE FEED
T

THE WORLD
S

AND SUSTAIN
E

F
THE PLANET? A five-step global plan
F

O could double food production

by 2050 while greatly reducing
O

environmental damage
By Jonathan A. Foley

R
IGHT NOW ABOUT ONE BILLION PEOPLE SUFFER FROM CHRONIC HUNGER. THE WORLDS
farmers grow enough food to feed them, but it is not properly distributed and,
even if it were, many cannot afford it, because prices are escalating. But anoth-
er challenge looms. By 2050 the worlds population will increase by two billion
or three billion, which will likely double the demand for food, according to sev-
eral studies. Demand will also rise because many more people will have higher
incomes, which means they will eat more, especially meat. Increasing use of
cropland for biofuels will put additional demands on our farms. So even if we solve todays
problems of poverty and accessa daunting taskwe will also have to produce twice as much
to guarantee adequate supply worldwide. And thats not all.

84

2015 Scientific American

 SCIENTIFIC AMERICAN

By clearing tropical forests, farming
marginal lands, and intensifying industri-
al farming in sensitive landscapes and
watersheds, humankind has made agri-
culture the planets dominant environ-
mental threat. Agriculture already con-
sumes a large percentage of the earths
land surface and is destroying habitat,
using up freshwater, polluting rivers and
oceans, and emitting greenhouse gases
more extensively than almost any other
human activity. To guarantee the globes
long-term health, we must dramatically
reduce agricultures adverse impacts.
The worlds food system faces three
incredible, interwoven challenges. It must
T guarantee that all seven billion people
H alive today are adequately fed; it must
double food production in the next 40
E
years; and it must become truly environ-
S
mentally sustainableall at the same time.
Could these simultaneous goals pos-
C
sibly be met? An international team of
I experts, which I coordinated, has settled
E on five steps that, if pursued together,
could increase by more than 100 percent
N
the food available for human consump-
C
tion globally while substantially reducing
E greenhouse gas emissions, biodiversity
losses, water use and water pollution.
Tackling the triple challenge will be one of
the most important tests humanity has
O
ever faced. It is fair to say that our response
F will determine the fate of our civilization.
F BUMPING UP AGAINST BARRIERS
O AT FIRST BLUSH, the way to feed more peo-
ple seems clear: grow more food, by
O
expanding farmland and improving yield
D (crops harvested per hectare). Unfortu-
nately, the world is running into signifi-
cant barriers on both counts.
Society already uses about 37 percent
of the earths land surface, not counting
Greenland or Antarctica, for farms or pas-
tures. Agriculture is by far the biggest
human use of land on the planet; nothing
else comes close. And most of that 37 per-
cent covers the best farmland. Much of the
remainder is covered by deserts, moun-
tains, tundra, ice, cities, parks and other
unsuitable growing areas. The few re-
maining frontiers are mainly in tropical
forests and savannas, which are vital to
the stability of the globe, especially as
stores of carbon and biodiversity. Expand-
ing into those areas is not a good idea, yet
five million to 10 million hectares of crop-
land have been created in each of the past
20 years, with a significant portion of that
land conversion happening in the tropics.
These additions enlarged the net area of
cultivated land by only 3 percent, howev-
er, because of farmland losses caused by
urban development and other forces, par-
ticularly in temperate zones.
Improving yield also sounds enticing.
TO GUARANTEE
THE GLOBES
LONG-TERM HEALTH,
WE MUST
DRAMATICALLY
REDUCE
AGRICULTURES
ADVERSE IMPACTS.
Yet our research team found that average
global crop yield increased by just 20 per-
cent from 1985 to 2005far less than had
been reported. Cereals yield has been ris-
ing at less than 2 percent a year since 2000
and yields of pulses (beans, lentils) and
root crops by less than 1 percentrates
that are nowhere near enough to double
food production by midcentury.
Feeding more people would be easier
if all the food we grew went into human
hands. But only 60 percent of the worlds
crops are meant for people: mostly grains,
followed by pulses, oil plants, vegetables
and fruits. Another 35 percent is used for
animal feed, and the final 5 percent goes
to biofuels and other industrial products.
Meat is the biggest issue here. Even with
the most efficient meat and dairy systems,
feeding crops to animals reduces the
worlds potential food supply. Grain-fed
cattle operations typically use at least 100
kilograms of grain to make one kilogram
of edible, boneless beef protein [see The
Greenhouse Hamburger, on page 108].
Chicken and pork are more efficient, and
grass-fed beef converts nonfood material
into protein. Overall, grain-fed meat pro-
duction systems are a drain on the global
food supply.
Another deterrent to growing more
food is damage to the environment, which
is already extensive. Only our use of ener-
gy, with its profound impacts on climate
and ocean acidification, rivals the sheer
magnitude of agricultures environmental
footprint. Our research team has estimat-
ed that by 2010 agriculture had already
cleared or radically transformed 70 per-
cent of the worlds prehistoric grasslands,
50 percent of the savannas, 45 percent of
the temperate deciduous forests and 25
percent of the tropical forests. Since the
last ice age, nothing has disrupted ecosys-
tems more. Agricultures physical foot-
print is nearly 60 times that of the worlds
pavements and buildings.
Freshwater is another casualty. Hu-
mans use an astounding 4,000 cubic kilo-
meters of water per year, mostly with-
drawn from rivers and aquifers. Irrigation
accounts for 70 percent of the draw. If we
count only consumptive water usewater
that is used and not returned to the water-
shedirrigation climbs to 80 or 90 per-
cent of the total. As a result, many large
rivers, such as the Colorado, have dimin-
ished flows, some have dried up altogeth-
er, and many places have rapidly declining
water tables, including regions of the U.S.
and India.
Water is not only disappearing, it is be-
ing contaminated. Fertilizers, herbicides
and pesticides are being spread at incredi-
ble levels and are found in nearly every
ecosystem. The flows of phosphorus and
nitrogen through the environment have
more than doubled since 1960, causing
widespread water pollution and enor-

IN BRIEF

The world must solve three food prob-
lems simultaneously: end hunger, dou-
ble food production by 2050, and do
both while drastically reducing agricul-
tures damage to the environment.
Five solutions, pursued together, can
achieve these goals: stop agriculture
from consuming more tropical land,
boost the productivity of farms that
have the lowest yields, raise the effi-
ciency of water and fertilizer use world-
wide, reduce per capita meat con-
sumption, and reduce waste in food
production and distribution.
A system for certifying foods based on
how well each one delivers nutrition
and food security and limits environ-
mental and social costs would help the
public choose products that push agri-
culture in a more sustainable direction.

86

2015 Scientific American

 SCIENTIFIC AMERICAN

L AY O F T H E L A N D

Farming Hits the Wall, but Not the Ceiling

Humankind now farms 37 percent of the earths ice-free land. Crops take up roughly one third of that area; pastures and rangelands for
livestock cover the rest. Little room exists for expansion because deserts, mountains, tundra or cities cover most of the remaining land.
Still, farms in many existing areas (insets) could be more productive.

Type of None
S
Agriculture
C

100% cropland 100% pasture Better Breadbaskets I

The world could grow much more food if the productivity of the poorest E
farms were raised toward the maximum possible, given climate and soil
Yield of Maize Farmland conditions. For example, the yield for maize (shown) could rise signifi- N

cantly across parts of Mexico, West Africa and eastern Europe if seeds, C
Far from maximum Close to maximum irrigation, fertilizer and markets were improved.
E

mous hypoxic dead zones at the mouths
of many of the worlds major rivers. Ironi-
cally, fertilizer runoff from farmlandin
the name of growing more foodcompro-
mises another crucial source of nutrition:
coastal fishing grounds. Fertilizer certain-
ly has been a key ingredient of the green
revolution that has helped feed the world,
but when nearly half the fertilizer we ap-
ply runs off rather than nourishes crops,
we clearly can do better.
Farming also accounts for 10 to 12 per-
cent of the warming effects of greenhouse
gases released by human activitya con-
COURTESY OF JAMES GERBER Institute on the Environment, University of Minnesota (maps)
expand or the mounting environmental
harm it imposes. Previous approaches to
solving food issues were often at odds with
environmental imperatives. We could
boost food production by clearing more
land or using more water and chemicals
but only at a cost to forests, streams and
wetlands. Or we could restore ecosystems
by taking farmland out of cultivation but
only by reducing food production. This
either-or approach is no longer accept-
able. We need truly integrated solutions.
After months of research and deliber-
ationbased on analysis of newly gener-
Slowing deforestation would dramati-
O
cally reduce environmental damage while
imposing only minor constraints on glob- F
al food production. The resulting dip in
farm capacity could be offset by reducing F
the displacement of more productive O
croplands by urbanization, degradation
O
and abandonment.
D
Many proposals have been made to
reduce deforestation. One of the most
promising has been the Reducing Emis-
sions from Deforestation and Degrada-
tion (REDD) mechanism. Under REDD,
rich nations pay tropical nations to pro-

tribution equal to that of all the road vehi-
cles on the planet. Most of the direct emis-
sions from farming come from methane
produced by animals and rice paddies
and from nitrous oxide released by over-
fertilized soils. Add in the effects of trop-
ical deforestation and other land clear-
ing, and agricultures share of global
emissions rises to 24 percent of the total.
FIVE SOLUTIONS
MODERN AGRICULTURE has been an incredi-
bly positive force in the world, but we can
no longer ignore its dwindling ability to
ated global agricultural and environmen-
tal dataour international team settled
on a five-point plan that deals with food
and environmental challenges together.
Stop expanding agricultures foot-
print. Our first recommendation is to
slow and ultimately stop the expansion of
agriculture, particularly into tropical for-
ests and savannas. The demise of these
ecosystems has far-reaching impacts on
the environment, especially through lost
biodiversity and increased carbon dioxide
emissions (from clearing land).
tect their rain forests, in exchange for car-
bon credits. Other mechanisms include
developing certification standards for ag-
ricultural products so that supply chains
can be assured that crops were not grown
on lands created by deforestation. Also,
better biofuel policyone that relies on
nonfood crops such as switchgrass in-
stead of food cropscould make vital
farmland newly available.
Close the worlds yield gaps. To double
global food production without expanding
agricultures footprint, we must signifi-

87

2015 Scientific American

 SCIENTIFIC AMERICAN

cantly improve yields of existing farm- S T R AT E G Y

lands. Two options exist. We can boost the
productivity of our best farmsraising
their yield ceiling through improved
Output Expands, Food Production
By 2050 global population will be
crop genetics and management. Or we can
improve the yields of the worlds least pro-
Harm Contracts two billion to three billion greater,
and a larger proportion of people
will have higher incomes, so they
ductive farmsclosing the yield gap To feed the world without harming the planet, will consume more per person.
between a farms current yield and its agriculture will have to produce much more food Farmers will need to grow twice
higher potential yield. The second option (blue) and find better ways to distribute it (red), as much as they do today.
provides the largest and most immediate while significantly cutting the damage it does to
gain, especially in regions where hunger is the atmosphere, habitat and water (yellow).
most acute.
Our research group analyzed global
patterns of crop yields and found signifi-
cant yield gaps in many regions: most
notably parts of Africa, Central America
T and eastern Europe. In these regions, bet-
H ter seeds, more effective fertilizer appli-
E
cation and more efficient irrigation could
produce much more food without in-
S
creasing the amount of land under culti-
vation. Our analysis suggests that closing
C
the yield gap for the worlds top 16 crops
I could increase total food production by
E 50 to 60 percent, without causing much
N
additional environmental damage.
Reducing yield gaps in the least pro- Food Access
C More than one billion of the earths
ductive agricultural lands may often re-
seven billion people suffer from
E quire the use of additional chemicals and chronic hunger. Poverty and poor
water. Farmers will have to irrigate and distribution of food must be overcome
Environmental Damage
To reduce harm, agriculture must stop expanding
fertilize in responsible ways. They can to provide adequate calories for everyone.
into tropical forests, raise the productivity of under-
also make use of other yield-lifting tech- performing farmland (which could boost production
O
niques, such as reduced tillage, which 50 to 60 percent), use water and fertilizer far more
F disturbs less soil and thus minimizes efficiently, and prevent soil degradation.
erosion. Cover crops planted between
F food-crop seasons suppress weeds and
O add nutrients and nitrogen to the soil
when plowed under. Lessons from organ- er. Our analysis found that farms can sig- Europewhere farmers could substan-
O
ic and agroecological systems can also be nificantly curb water use without much tially reduce fertilizer use with little or no
D adopted, such as leaving crop residues reduction in food production, especially in impact on food production.
on fields so that they decompose into nu- dry climates. Primary strategies include Among the actions that can fix waste-
trients. To close the worlds yield gaps, drip irrigation (where water is applied ful overfertilization are policy and eco-
we also have to overcome serious eco- directly to the plants base and not waste- nomic incentives, such as payments to
nomic and social challenges, including fully sprayed into the air); mulching (cov- farmers for: promoting watershed stew-
better distribution of fertilizer and seed ering the soil with organic matter to retain ardship and protection, reducing exces-
varieties to farms in impoverished re- moisture); and reducing water lost from sive fertilizer use, improving manure
gions and improved access to global mar- irrigation systems (by lessening evapora- management (especially manure storage,
kets for many regions. tion from canals and reservoirs). so that less runs off into the watershed
With fertilizers, we face a kind of Gold- during a storm), capturing excess nutri-
Use resources much more efficiently. ilocks problem. Some places have too few ents through recycling, and instituting
To reduce the environmental impacts of nutrients and therefore poor crop pro- other conservation practices. In addition,
agriculture, low- and high-yield regions duction, whereas others have too much, wetlands could be restored to enhance
alike must practice agriculture in ways leading to pollution. Almost no one uses their capacity to act as natural sponges
that produce vastly greater output of fertilizers just right. Amazingly, only 10 that filter out nutrients in runoff.
crops per unit input of water, fertilizer percent of the worlds cropland generates Here again, reduced tillage can help
and energy. 30 to 40 percent of agricultures fertilizer nourish the soil, as can precision agricul-
On average, it takes about one liter of pollution. Our analysis identified hotspots ture (applying fertilizer and water only
irrigation water to grow one calorie of on the planetparticularly in the central when and where they are needed and most
food. Some places use much more, howev- U.S., China, northern India and western effective) and organic farming techniques.

88 Graphic by Jen Christiansen

2015 Scientific American


Shift diets away from meat. We can
dramatically increase global food avail-
ability and environmental sustainability
by using more of our crops to feed people
directly and less to fatten livestock.
Globally, humans could net up to
three quadrillion additional calories
every yeara 50 percent increase from
our current supplyby switching to all-
plant diets. Naturally, our current diets
and uses of crops have many economic
and social benefits, and our preferences
are unlikely to change completely. Still,
even small shifts in diet, say, from grain-
fed beef to poultry, pork or pasture-fed
beef, can pay off handsomely.
Reduce food waste. A final, obvious but
often neglected recommendation is to re-
duce waste in the food system. Roughly
30 percent of the food produced on the
planet is discarded, lost, spoiled or con-
sumed by pests.
In rich countries, much of the waste
takes place at the consumer end of the
system, in restaurants and trash cans.
Simple changes in our daily consumption
patternsreducing oversize portions, the
food thrown in the garbage, and the num-
ber of takeout and restaurant meals
could significantly trim losses, as well as
our expanding waistlines. In poorer coun-
tries, the losses are similar in size but
occur at the producer end, in the form of
failed crops, stockpiles ruined by pests or
spoilage, or food that is never delivered
because of bad infrastructure and mar-
kets. Improved storage, refrigeration and
distribution systems can cut waste appre-
ciably. Moreover, better market tools can
connect people who have crops to those
who need them, such as cell-phone sys-
tems in Africa that link suppliers, traders
and purchasers.
Although completely eliminating waste
from farm to fork is not realistic, even
small steps would be extremely benefi-
cial. Targeted effortsespecially reduc-
ing waste of the most resource-intensive
foods such as meat and dairycould
make a big difference.
MOVING TOWARD
A NETWORKED FOOD SYSTEM
IN PRINCIPLE, our five-point strategy can
address many food security and environ-
mental challenges. If they were done
together, the steps could increase the
worlds food availability by 100 to 180 per-
SCIENTIFIC AMERICAN
FEEDING NINE
BILLION PEOPLE
IN A TRULY
SUSTAINABLE WAY
WILL BE ONE
OF THE GREATEST
CHALLENGES
OUR CIVILIZATION
HAS HAD
TO CONFRONT.
cent while significantly lowering green-
house gas emissions, biodiversity losses,
water use and water pollution.
It is important to emphasize that all
five points (and perhaps more) must be
pursued simultaneously. No single strate-
gy is sufficient to solve all our problems.
Think silver buckshot, not a silver bullet.
We have tremendous successes from the
green revolution and industrial-scale ag-
riculture to build on, along with innova-
tions in organic farming and local food
systems. Lets take the best ideas and in-
corporate them into a new approacha
sustainable food system that focuses on
nutritional, social and environmental per-
formance, to bring responsible food pro-
duction to scale.
We can configure this next-generation
system as a network of local agricultural
systems that are sensitive to nearby cli-
mate, water resources, ecosystems and
culture and that are connected through
efficient means of global trade and trans-
port. Such a system could be resilient
and pay farmers a living wage.
One device that would help foster this
new food system would be the equivalent
of the Leadership in Energy and Envi-
ronmental Design program now in place
for constructing new commercial build-
ings sustainably. This LEED program
awards increasingly higher levels of cer-
tification based on points that are accu-
M O R E TO E X P L O R E
Global Consequences of Land Use. Jonathan A. Foley et al. in Science, Vol. 309, pages 570574; July 22, 2005.
Enough: Why the Worlds Poorest Starve in an Age of Plenty. Roger Thurow and Scott Kilman. Public Affairs, 2010.
Food Security: The Challenge of Feeding 9 Billion People. H. Charles J. Godfray et al. in Science, Vol. 327, pages 812
818; February 12, 2010.
Solutions for a Cultivated Planet. Jonathan A. Foley et al. in Nature, Vol. 478, pages 337342; October 20, 2011.
89
2015 Scientific American
mulated by incorporating any of a wide
range of green options, from solar power
and efficient lighting to recycled building
materials and low construction waste.
For sustainable agriculture, foods
would be awarded points based on how
well they deliver nutrition, food security
and other public benefits, minus their
environmental and social costs. This cer-
tification would help us get beyond cur-
rent food labels such as local and
organic, which really do not tell us
much about what we are eating. Instead
we can look at the whole performance of
our foodacross nutritional, social and
environmental dimensionsand weigh
the costs and benefits of different farm- T
ing approaches. H
Imagine the possibilities: sustainable E
citrus and coffee from the tropics, con-
nected to sustainable cereals from the S
temperate zone, supplemented by locally
C
grown greens and root vegetables, all
grown under transparent, performance- I
based standards. Use your smartphone E
and the latest sustainable food app, and N
you will learn where your food came
C
from, who grew it, how it was grown, and
how it ranks against various social, nutri- E
tional and environmental criteria. And
when you find food that works, you can
tweet about it to your social network of
O
farmers and foodies.
The principles and practices of our dif- F
ferent agricultural systemsfrom large-
scale commercial to local and organic F
provide the foundation for grappling with O
the worlds food security and environmen-
O
tal needs. Feeding nine billion people in a
D
truly sustainable way will be one of the
greatest challenges our civilization has
had to confront. It will require the imagi-
nation, determination and hard work of
countless people from all over the world.
There is no time to lose.
Jonathan A. Foley is executive director of the
California Academy of Sciences in San Francisco.
From 2008 to 2014, he led the Institute on the
Environment at the University of Minnesota.
s c i e n t i f i c a m e r i c a n . c o m /m a g a z i n e /s a
Protein production: planet, profit, plus people?14
Harry Aiking
ABSTRACT
Food sustainability and food security are increasingly in the spotlight
and increasingly intertwined. According to some projections we will
need to nearly double food production in the next 4 decades. This article
argues that protein production and consumption are pivotal to sustain-
ability, because anthropogenic contributions to the nitrogen cycle are
100200% compared with a contribution of 12% to the carbon cycle
by mineral fuel combustion, with biodiversity as the main casualty.
Because 1 kg animal protein requires w6 kg plant protein, its large-
scale production by means of factory farming is a major driver of
biodiversity loss, climate change, and freshwater depletion. Further-
more, intensive livestock production is associated with antibiotics re-
sistance and increasing incidence of emerging diseases. Therefore,
a reversed diet transition back to less animal protein could make
a difference. Some European countries, such as the United Kingdom,
Sweden, and The Netherlands, have published integrated policy reports
addressing food security, sustainability, and health combined. The food
industry is focusing on food safety and increasingly on sustainability.
An important issue is consumer communication, because consumer
framing is radically different from that of governmental and indus-
trial policy makers. There is no one size fits all. A huge range of
differences exists between countries and between distinct groups of
consumers within countries; getting consumers to change their diets
in a more sustainable direction is likely to require much more than
gentle nudging. National governments and the United Nations should
assume their responsibilities and initiate a global strategy integrating
sustainability, food security, nutrition, and equity. To date, the profit
pillar of sustainability has taken precedence over planet and people. It
is time to redress the balance. Am J Clin Nutr 2014;100
(suppl):483S9S.
INTRODUCTION
Nutrition takes a prominent position among our universal
needs, demanding considerable resources along with water,
shelter, and energy (1). However, access to food is far from
equitable, with the result that presently w1 billion people are
obese and w1 billion go hungry (2). In a world speeding toward
a population of 9 billion, it is increasingly acknowledged that
nutrition should not just promote health but sustainability as
well (37).
World population and average income are at unprecedented
levels and will continue to rise. However, increasing food pro-
duction generates its own feedback inhibition, thus compro-
mising food security, food safety, and food sustainability. From
the perspective of food security and nutrient adequacy, there is
a bias toward calories and vitamins. From the perspective of food
sustainability, there is a bias toward climate change, and so to-
Am J Clin Nutr 2014;100(suppl):483S9S. Printed in USA. ! 2014 American Society for Nutrition
ward carbon and calories. However, nitrogenor protein for that
matteris an important but often disregarded macronutrient
with a pivotal role in biodiversity loss, climate change, and
human health risks.
Therefore, the current article argues that food security, food
sustainability, and nutrition are increasingly intertwined and
should be addressed in an integrated way, by identifying trade-
offs and setting priorities. Subsequently, it is argued that a re-
duction in the consumption of animal products is a prominent
option toward more sustainable food production. The role of
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
major stakeholders and policy options will be discussed.
SUSTAINABILITY
Humanity has the ability to make development sustainableto en-
sure that it meets the needs of the present without compromising the
ability of future generations to meet their own needs. (8)
This statement by the World Commission on Environment and
Development, linking 1) global environmental deterioration, 2)
poverty, and 3) rapid population growth, is often used as a def-
inition of sustainability. It links the environments ability to
meet present and future human needs with theories of social
justiceboth within and between generationsas a basis for
ecologic, economic, and social aspects of sustainability (9).
In this respect, ecology, economy, and society are known as the
3 pillars of sustainability, alternatively referred to as people,
planet, profit. For a heterogeneous society it is often easier to
agree on the ills to be avoided (eg, poverty) than on the ideals to
be achieved (eg, the ideal income distribution) (10), so sus-
tainability may be a generally accepted goal with relative con-
sensus on its ills (such as production-related impacts) but
hardly ever on its ideals. For example, an industrial framing
of sustainability (11, 12) and that of the average consumer may
be worlds apart, because the latter primarily associates sus-
tainability with attributes such as natural and just (13). At
any rate, sustainability is not a static notion but a moving target,
1
From the Institute for Environmental Studies (IVM), VU University, De
Boelelaan, Amsterdam, Netherlands.
2
Presented at the symposium Sixth International Congress on Vegetarian
Nutrition held in Loma Linda, CA, 2426 February 2013.
3
The work reported here is based in part on results from the Protein Foods,
Environment, Technology and Society (PROFETAS) research program
funded by the Netherlands Organization for Scientific Research (NWO).
4
Address correspondence to H Aiking, Institute for Environmental
Studies (IVM), VU University, De Boelelaan 1087, 1081 HV Amsterdam,
The Netherlands. E-mail: harry.aiking@vu.nl.
First published online May 28, 2014; doi: 10.3945/ajcn.113.071209.
483S
484S
which should be understood as a challenge to preserve the
adaptability and resilience of the natural (biotic and abiotic)
systems that form the basis of economic and social development.
FOOD SUSTAINABILITY
Many definitions of sustainability are known and are generally
based on its 3 pillars of ecology, economy, and society. The same
holds with respect to food sustainability (14). However, distinct
definitions of food sustainability may, or may not, address issues
such as human health, equity, and animal welfare, indicating
that the issue is fraught with ethical considerations. In fact, it is
often unclear whether food sustainability definitions refer to
production or consumption. The gap between production and
consumption can be 30%, which may be in the form of waste (15)
or other uses of food such as pet food. As an illustration of the
latter, De Silva and Turchini (16) conservatively estimated that
2.48 million tons of wild-caught fish is used directly by the cat
food industry, raising the ethical point that the use of a limited
biological resource to feed pets competes with human food
purposes.
Because it appropriates major shares of all ice-free land (33%),
freshwater (70%), and energy production (20%) (17, 18), food
production is one of the main drivers of environmental degra-
dation and resource depletion (19). As global food production and
consumption continue to increase, so will the associated envi-
ronmental impacts. In fact, the environmental impacts of food
production include both resource depletion and pollution on all
scales from local to global. Prominent examples include impacts
on biodiversity (20, 21), climate change (22), and human health
(23), clearly showing the range and importance of anthropogenic
impacts on the environment via food production.
FOOD SECURITY
Food security exists when all people, at all times, have physical,
social and economic access to sufficient, safe and nutritious food to
meet their dietary needs and food preferences for an active and
healthy life. The 4 pillars of food security are availability, access,
utilization and stability. The nutritional dimension is integral to the
concept of food security. (24)
Defined in this way, food security is shown to have many
dimensions. Technology does not seem to be the limiting factor,
and producing enough food for 10 billion people seems feasible
(25, 26). Rather, socioeconomic aspects such as sex and poverty
are likely to constrain access to food (2729). Another economic
aspect is the market exclusion of smallholder producers by
current intensification trends (30, 31), which also tend to disrupt
beneficial functions of biodiversity (32).
The real challenge is still to come, however. Initially, the FAO
projected that by 2050 world food demand will be w70% higher
than in 2005/2007 (33) but later revised this figure to 60% (34).
According to other recent analyses, however, a 100110% in-
crease in global crop supply will be needed from 2005 to 2050
(35). As a consequence of the foreseen increase in demand,
world market price projections of the International Food Policy
Research Institute showed that world grain prices may increase
3050% before 2050 (36). However, more recent sources sug-
gested that Global food prices are predicted to rise by 7090
percent by 2030 (37).
AIKING
After the 2008 price increases, Lawrence Haddad, one of the
authors of the Foresight report (38), warned in an interview, The
last 3 to 4 years have seen alarming spikes in hunger. The price
rises in 20078 were actually quite modest in a historical context
but it led to 100 million more people going hungry. Bigger price
rises could wipe out the development gains of the last 20 years
and promote violent conflict and migration (39). In 2012, after
droughts that threatened US corn harvests, similar warnings
were voiced by the FAO (40).
In fact, the effects of the seemingly inevitable price increases
are not restricted to developing countries (41). Browning et al
(42) summarized the impacts in the United Kingdom as follows:
There are over 4 million people in the UK currently living in
food poverty and The lowest income households tend to be the
hardest hit by changes in food affordability. The relative af-
fordability of foodmeasured by the share of total consumer
spending that goes on food and nonalcoholic drink for house-
hold suppliesis therefore a key indicator of household food
security.
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
NUTRITION AND HEALTH
Embedded in the definition of food security is the requirement
that food should be nutritious, safe, and healthy. Such is generally
addressed at the level of individuals by dietary guidelines (43).
However, by their effects on public health and global sustain-
ability, current food production volumes and methods have led to
a new situation, as is evident from publications linking nutrition,
health, and sustainability. For example, McMichael et al (23)
summarized: Together with persistent widespread under-
nutrition, overnutrition (and sedentarism) is causing obesity
and associated serious health consequences. Worldwide, agri-
cultural activity, especially livestock production, accounts for
about a fifth of total greenhouse-gas emissions, thus contributing
to climate change and its adverse health consequences, in-
cluding the threat to food yields in many regions. Particular
policy attention should be paid to the health risks posed by the
rapid worldwide growth in meat consumption, both by exacer-
bating climate change and by directly contributing to certain
diseases.
By the same token, the Health Council of The Netherlands
concluded that dietary guidelines should take full account of the
ecologic perspective. In their 2011 report on this issue, and based
on an international workshop, they advocated a diet transition
toward fewer animal products, because such a diet will result in
a tremendous reduction in the pressure on land, freshwater, and
biodiversity resources, with added benefits for human health and
animal welfare (7). In Italy, the Barilla Center complemented the
existing food pyramid based on nutrition with an environmental
pyramid based on sustainability (44), addressing food security in
the process (45). Once more, livestock products were shown to
have important health and environmental impacts.
In addition, resistant bacteria (eg, methicillin-resistant Staphy-
lococcus aureus, extended spectrum b lactamase) result to a large
extent from antibiotics used in intensive livestock production (46,
47). When a prophylactic addition to feed was forbidden in the
European Union (EU) in 2006, therapeutic antibiotic use in live-
stock production increased. The amount of antibiotics used in live-
stock is considerably (w5-fold) higher than is used in human health
care. Research by the European Medicines Agency showed that
 PROTEIN PRODUCTION: PLANET, PROFIT, PLUS PEOPLE?
in the EU antibiotics resistance kills w25,000 people and costs
w1.5 billion/y (48). Finally, an endless string of food scares as-
sociated with emerging zoonotic diseases (including Bovine Spon-
giform Encephalopathy, avian influenza, Q fever, enterohemorrhagic
Escherichia coli) can be linked to livestock products (49). The
concept of One World, One Health becomes even more powerful
when considering that biodiversity was recently shown to have
a positive effect in natural ecosystems by reducing disease (50).
INTEGRATED APPROACH
Until a decade ago, addressing food security without com-
promising sustainability was considered an ever-growing chal-
lenge (51, 52), but more recently the message was summarized in
a more direct way: Continuing population and consumption
growth will mean that the global demand for food will increase
for at least another 40 years. Growing competition for land,
water, and energy, in addition to the overexploitation of fisher-
ies, will affect our ability to produce food, as will the urgent
requirement to reduce the impact of the food system on the
environment. The effects of climate change are a further threat.
But the world can produce more food and can ensure that it is
used more efficiently and equitably. A multifaceted and linked
global strategy is needed to ensure sustainable and equitable
food security (53). Other authors advocate an integrative ap-
proach to nutrition (54). In addition, it is held that the effects of
price increases may be cushioned by adopting a more holistic
approach toward resources (55). In fact, some European coun-
tries, such as the United Kingdom (56), Sweden (4), and The
Netherlands (5), have published policy reports addressing the
combined issues of food security, sustainability, and health. In
an increasingly interlinked world speeding toward 9 billion
people an integrated approach is clearly the way to proceed.
However, to tackle such a complex issue efficiently, at least
some priority setting is required to establish a beachhead.
SETTING PRIORITIES
To quantify sustainability in terms of the carrying capacity of
our planet, the groundbreaking article by Rockstrom et al (57)
defined and established boundary values that should not be
transgressed for the most important anthropogenic environ-
mental issues (Table 1). Subsequent analysis indicates the fol-
lowing: 1) food production is an important driver underlying all
of these impacts; 2) the top 3 environmental impacts, ie, bio-
TABLE 1
Ranking environmental impacts according to the transgression of planetary
boundaries established by Rockstrom et al1
Rank Environmental impact Current status2
1 Rate of biodiversity loss .10
2 Nitrogen cycle disruption 3.45
3 Climate change (carbon cycle disruption) 1.11.5
4 Phosphate cycle disruption 0.770.86
5 Ocean acidification 0.81
6 Land-use change 0.78
7 Freshwater use 0.65
8 Stratospheric ozone depletion 0.50
1
Data adapted from reference 57.
2
The boundary value for sustainability equals 1.
485S
diversity loss, nitrogen cycle disruption, and climate change, are
strongly interlinked rather than independent of one another; and
3) protein production is the pin linking these 3 impacts (19). The
underlying causes are discussed in detail below.
Dietary protein is nutritionally crucial (58), because it is the
primary way to acquire nitrogen, which is an essential element in
DNA, RNA, and cell protein. Smil (59) also calculated that before
the large-scale application of fertilizers, the world population was
capped at w3 billion people by nitrogen limitation, less than
half the present number. The proportion of animal protein in the
diet is primarily income dependent (60), but the actual protein
source also depends on cultural aspects (61).
Importantly, a large proportion of fertilizer nitrogen is lost to
the environment. In 2005, just 17% was consumed by humans in
crop, dairy, and meat products, and the global nitrogen use ef-
ficiency of crops keeps dropping (21). In parallel, ammonia
emissions from manure are increasing. Much of this reactive
nitrogen is deposited in nitrogen-limited ecosystems via the
atmosphere. There it leads to unintended fertilization of eco-
systems unable to cope with this nutrient inflow (eg, several
types of forest), which makes it one of the leading causes of
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
terrestrial biodiversity loss (62, 63). Pollution from livestock
enterprises affects both terrestrial and aquatic ecosystems (64).
In addition, fertilizer runoff may lead to algal blooms and dead
zones in sensitive coastal ecosystems, with inevitable re-
percussions on aquatic biodiversity (21). In sum, the nitrogen cycle
is strongly linked to both terrestrial and aquatic biodiversity loss.
The tremendous energy input in nitrogen fertilizer alone is
responsible for 2% of world energy consumption (63) and for
37% of all energy expenditure in US agriculture (3), thus causing
significant climate change (21, 59) and linking the anthropogenic
disruption of the nitrogen cycle to the disruption of the carbon
cycle (ie, to climate change). In short, nitrogen is crucial to
terrestrial and aquatic biodiversity loss, climate change, human
health, and many other issues (21, 23, 62, 65). Anthropogenic
contributions to the natural carbon cycle represent 12% by
mineral fuel combustion (18) but are 100200% to the natural
nitrogen cycle. In fact, the current production of fertilizer far
exceeds the formation of reactive nitrogen by natural processes,
such as lightning (21, 62, 66). Consequently, Rockstrom et al
(57) ranked the impacts of this nitrogen cycle disruption in
between those of biodiversity loss and carbon cycle disruption
(see Table 1). Because nitrogen cycle disruption has strong
impacts both on biodiversity and on the carbon cycle, protein
production was shown to be the pivotal link between the top 3
environmental issues in the Rockstrom et al ranking (19).
The environmental impacts of reactive nitrogen on biodiversity,
climate change, and human health are well established in the
United States (67) and in China (68). In Europe, the associated
costs have been estimated at 70320 billion annually, of which
w75% is related to health damage and air pollution (66). In
summary, reducing the losses of reactive nitrogen compounds in
the food chain will benefit biodiversity, human health, and a rap-
idly changing climate, thus killing several birds with one stone,
and making it stand out as a top priority to achieve sustainability.
REDUCING ANIMAL PROTEIN INTAKE
The conversion of plant protein into animal protein is a met-
abolic process optimized for animal survival. Turning protein
486S AIKING
from feed crops into animal protein for human consumption may
be economically feasible, but it is inherently resource-inefficient,
which makes intensive livestock production responsible for
a disproportionate share of environmental pressure (6971). On
average, 6 kg of plant protein is required to yield 1 kg of meat
protein (17, 72). Consequently, only 15% of protein and energy
provided by feed crops will be consumed by humans indirectly
and 85% of these crops are wasted (and, incidentally, 85% of
fertilizer inputs to grow them). In 2000, for example, 942 and
617 million tons of grain were used for food and feed, re-
spectively (36). Of the latter, .500 million tons constitute
a tremendous loss of resources. Maybe even more important,
this huge amount of potential food is turned into massive
emissions of reactive nitrogen (eg, ammonia from manure),
which pollutes the terrestrial and aquatic environment and re-
sults in biodiversity loss.
Both resource depletion and pollution by livestock production
were treated in a comprehensive FAO report (70). Moreover,
intensive livestock production was shown to play a crucial role in
all 3 of the planetary boundaries that have already been
overstepped by humanity (ie, biodiversity loss, nitrogen cycle
disruption, and carbon cycle disruption) (19). This conclusion
was fully confirmed by an exhaustive review of European pro-
tein impacts and options for their reduction (65). More recently,
Sutton and Dibb (73) estimated that 1) nearly one-third of global
biodiversity loss is attributable to livestock production, 2) meat
consumption is responsible for nearly half of the UK food
greenhouse gas emissions, and 3) the estimated cost to the Na-
tional Health Service in early deaths related to excessive meat
consumption is 1.28 billion.
From the above it is clear that plant protein production causes
far less pollution and requires far fewer resources than the
production of animal protein. The gain of a reversed diet
transition (74) back toward a diet with less animal protein (75)
is, in fact, amplified by substantial reductions in freshwater,
land, and fertilizer requirements (76), with beneficial effects on
all of Rockstrom et als (57) environmental impacts (Table 1),
without exception (19).
Total protein supply (= production + imports 2 exports) across
the 15 countries of the European Union ranged between 95.8
and 118.9 g/d (77). Correcting for household losses of 2530%
(15, 78), the average European consumption is therefore at least
150% of Dietary Reference Intakes, which are 5060 g protein/
d for adults (79). In the United States, Dietary Reference Intakes
are similar (4656 g/d) (80) to those in the EU. The supply of
plant protein is also similar; however, animal protein supply in
the United States is considerably higher (ie, 72.3 compared with
61.8 g/d in the EU) (81).
Consequently, on both continents there is ample room for a diet
that is less dependent on animal proteins and therefore attractive
from a sustainability perspective. In addition, reduced intakes of
animal products would benefit human health as well as the
ecology (7). European consumers should therefore consider to
change their diets by doing the following: 1) eating one-third less
protein (the average amount of overconsumption), 2) replacing
one-third with plant-derived protein, and 3) replacing the re-
maining animal protein (primarily meat) with protein from free-
ranging animals (82). In agreement with the results of the
multidisciplinary Protein Foods, Environment, Technology and
Society (PROFETAS) research program (83), these 3 steps
would benefit health, environment, and animal welfare. Similar
recommendations could be phrased for the United States.
Of course, a successful diet transition requires local differ-
ences to be addressed. For example, the global population has
been projected to increase by 50% between 2000 and 2050, and
both meat and dairy demand have been projected to double (70).
Most of the projected population increase will take place in
Africa and India, although meat demand will increase primarily
in China and South America and dairy demand in India. Such
geographic distinctions should be taken into careful consider-
ation.
STAKEHOLDERS AND POLICY
Reducing consumption in general or that of livestock products
in particular is easier said than done. In essence, there is no
problem owner. Although lip service is paid to less is more,
the bottom line is that to reduce consumption (or to be told to do
so) is in sharp conflict with the desires of the average consumer,
industrialist, and politician.
As indicated above, some countries have integrated policies
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
educating or advising consumers to reduce meat consump-
tion, placing the burden of responsibility on the consumers
shoulders, but they shy away from more substantive measures.
Options include promotion of meat substitutes and the taxation
of meat products, for example (84). According to the Royal
Society, stakeholder dialogue is a must (85), and a framework to
help consumers, producers, and policy makers out of deadlock
and into negotiation is available (86). However, in actual prac-
tice, little political effort is devoted to this issue. By the same
token, the food industry is focusing primarily on food safety and
increasingly on sustainability, but there is a lot of green
washing.
Despite the social status of meat (61), some consumers are
prepared to avoid meat (87), although health benefits are a much
stronger motivation than environmental concerns (88). Conse-
quently, consumer communication is crucial, because consumer
framing is fundamentally different from governmental and
industrial policy makers (89). There is no one size fits all. A
huge range of differences exists among countries (77) and
among distinct groups of consumers within a country (90, 91).
At any rate, getting consumers to change their diets in a more
sustainable direction is likely to require much more than gentle
nudging. Consumer awareness should be raised with regard to
protein overconsumption and the associated environmental and
health costs. More important, however, national governments
and the UN should shoulder their responsibilities and take the
lead, as proposed by the British Food Ethics Council (92).
CONCLUSIONS
Food demand has been projected to double by 2050. Simul-
taneously, the environmental impacts of food production will
have to be reduced strongly and urgently. Taken together, this
outlines the daunting task of almost quartering the impacts per
ton of food within a period of only 40 y. Because animal protein
production appropriates a huge and disproportionate share of
natural resources, it presents a perfect target as an option for
significant reduction. Reducing animal protein consumption will
therefore benefit both food security and food sustainability (89).
 PROTEIN PRODUCTION: PLANET, PROFIT, PLUS PEOPLE?
Food security, equity, health, climate, and biodiversity may all
benefit. In fact, several authors agreed that 1) a diet transition is
required and 2) better dietary health and better environmental
quality generally go hand in hand (7, 23, 63, 93).
Continuing on the present path is not an option, because, if
animal protein consumption is not reduced voluntarily, a transi-
tion toward less animal protein is likely to be brought about by
rising prices, which will hurt the poor and increase world hunger.
In the words of Roberts et al (92): So there is a problem:
business and government both look to consumers to lead the way
on sustainable consumption, but consumers do not want to as-
sume this responsibility. According to these authors, the food
business, government, and civil society should all take their
responsibility, which can and should be done in small steps. In
addition, they propose, We also need a global strategy, priori-
tized by the G20 and led by the UN, that takes a comprehensive
approach that puts sustainability, equity and hunger at the heart
of food security (92). Be that as it may, such global political
processes are renowned to be slower than the few decades that
remain to us to absorb the anticipated strain on ecology and
society.
So far, an important bottleneck has been the sluggishness of
World Trade Organization negotiations, or more in general, the
fact that stakeholders (including most governments, industry, and
consumers) give preference to economic issues over ecologic and
social issues. In this globalized world, the pressures of ethical
trade-offs (94) are mounting. Under the current conditions of an
unprecedented global population size it may be time to rethink
issues such as consumer freedom (diet choice) compared with
global food security, the use of 2.48 million tons of fish for cat
food, and free trade, not by abolishing these pillars of current
society but by adapting them to remain within Rockstrom et als
planetary boundaries. It is evident that all stakeholders should
take responsibility without exception (92); however, national
governments and the UN should take the lead and initiate
a global strategy integrating sustainability, food security, nutri-
tion, and equity. To date, the profit pillar of sustainability has
taken precedence over planet and people. It is time to redress the
balance.
The author is a member of the Scientific Advisory Committee of the Eu-
ropean Natural Soyfoods Association.
REFERENCES
1. Vellinga P, Herb N. Industrial transformation science plan. Bonn,
Germany: International Human Dimensions Programme, 1999. (IHDP
report no. 12.)
2. Beddington JR, Asaduzzaman M, Clark M, Fernandez A, Guillou M,
Jahn M, Erda L, Mamo T, Van Bo N, Nobre CA, et al. Achieving food
security in the face of climate change: final report from the Commis-
sion on Sustainable Agriculture and Climate Change. Copenhagen,
Denmark: CGIAR Research Program on Climate Change, Agriculture
and Food Security, 2012. Available from: http://ccafs.cgiar.org/
commission (cited 14 November 2012).
3. Lang T, Barling D, Caraher M. Food policy: integrating health, envi-
ronment & society. Oxford, United Kingdom: Oxford University Press,
2009.
4. Livsmedels Verket. The National Food Administrations environmen-
tally effective food choices. Uppsala, Sweden: Livsmedels Verket
(Swedish National Food Administration), 2009. Available from:
http://www.slv.se/upload/dokument/miljo/environmentally_effective_
food_choices_proposal_eu_2009.pdf (cited 2 November 2009).
5. Ministry of Agriculture, Nature and Food Quality (LNV). Policy
document on sustainable food: towards sustainable production and
487S
consumption of food. The Hague, Netherlands: Ministry of Agriculture,
Nature and Food Quality (LNV), 2009. Available from: http://www.
minlnv.nl/portal/page?_pageid=116,1640773&_dad=portal&_schema=
PORTAL&p_file_id=40703 (cited 9 October 2009).
6. Macdiarmid J, Kyle J, Horgan G, Loe J, Fyfe C, Johnstone A, McNeill
G. Livewell: a balance of healthy and sustainable food choices.
Aberdeen, United Kingdom: World Wildlife FundUK, Rowett
Institute of Nutrition and Health, 2011.
7. Health Council of the Netherlands. Guidelines for a healthy diet: the
ecological perspective. The Hague, Netherlands: Health Council of the
Netherlands/Gezondheidsraad, 2011. (Report no. 2011/08E.) Available
from: http://www.gr.nl (cited 13 May 2014).
8. Brundtland GH. Our common future. World Commission on Envi-
ronment and Development. Oxford, United Kingdom: Oxford Uni-
versity Press, 1987.
9. Langhelle O. Sustainable development and social justice: expanding
the Rawlsian framework of global justice. Environ Values 2000;9:295
323.
10. Lindblom CE. Inquiry and change: the troubled attempt to understand
and shape society. New Haven, CT: Yale University Press, 1990.
11. World Business Council for Sustainable Development. Homepage.
Available from: http://www.wbcsd.org (cited 8 March 2013).
12. Capper JL. The environmental impact of beef production in the United
States: 1977 compared with 2007. J Anim Sci 2011;89:424961.
13. Kloppenburg J, Lezberg S, De Master K, Stevenson GW, Hendrickson
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
J. Tasting food, tasting sustainability: defining the attributes of an al-
ternative food system with competent, ordinary people. Hum Organ
2000;59:17786.
14. Aiking H, de Boer J. Food sustainability: diverging interpretations. Br
Food J 2004;106:35965.
15. Ventour L. The food we waste. Banbury, Oxon, United Kingdom:
Waste & Resources Action Programme, 2008. Available from: http://
www.wrap.org.uk/retail/case_studies_research/report_the_food_we.
html (cited 29 September 2009).
16. De Silva SS, Turchini GM. Towards understanding the impacts of the
pet food industry on world fish and seafood supplies. J Agric Environ
Ethics 2008;21:45967.
17. Smil V. Feeding the world: a challenge for the twenty-first century.
Cambridge, MA: MIT Press, 2000.
18. Smil V. The Earths biosphere: evolution, dynamics, and change.
Cambridge, MA: MIT Press, 2002.
19. Aiking H. Future protein supply. Trends Food Sci Technol 2011;22:
11220.
20. Nierenberg D. Rethinking the global meat industry. In: Nierenberg D,
ed. State of the world 2006a Worldwatch Institute report on progress
towards a sustainable society. London, United Kingdom: Earthscan,
2006:2440.
21. Erisman JW, Sutton MA, Galloway JN, Klimont Z, Winiwarter W.
How a century of ammonia synthesis changed the world. Nat Geosci
2008;1:6369.
22. Carlsson-Kanyama A, Gonzalez AD. Potential contributions of food
consumption patterns to climate change. Am J Clin Nutr 2009;89
(suppl):1704S9S.
23. McMichael AJ, Powles JW, Butler CD, Uauy R. Food, livestock pro-
duction, energy, climate change, and health. Lancet 2007;370:125363.
24. World Summit on Food Security. Declaration of the World Summit on
Food Security. Rome, Italy: FAO, 2009. World Summit on Food Se-
curity; 1618 November 2009. Available from: ftp://ftp.fao.org/docrep/
fao/Meeting/018/k6050e.pdf (cited 8 March 2013).
25. Netherlands Scientific Council for Government Policy. World food supply.
In: Sustained risks: a lasting phenomenon. The Hague, Netherlands:
Netherlands Scientific Council for Government Policy, 1995:5171.
Available from: http://www.wrr.nl/fileadmin/en/publicaties/PDF-Rapporten/
R44_Sustained_Risks_a_Lasting_Phenomenon__r44_1995_.pdf (cited 17
September 2012).
26. Evans LT. Feeding the ten billion: plants and population growth.
Cambridge, United Kingdom: Cambridge University Press, 1998.
27. Saravia-Matus S, Gomez y Paloma S, Mary S. Economics of food
security: selected issues. Bio-based Appl Econ J 2012;1:6580.
28. Ivers LC, Cullen KA. Food insecurity: special considerations for
women. Am J Clin Nutr 2011;94(suppl):1740S4S.
29. Ndiku M, Jaceldo-Siegl K, Singh P, Sabate J. Gender inequality in food
intake and nutritional status of children under 5 years old in rural
Eastern Kenya. Eur J Clin Nutr 2011;65:2631.
Melbourne wastes 200 kg of food per person a year: it's time to get serious

Melbourne wastes 200 kg of food per

person a year: its time to get serious
June 3, 2016 6.24am AEST
Wasting food, wasting the earth. Matt Carey, Author provided

Seona Candy
Research Fellow: Sustainable Food Systems, University of Melbourne
Jennifer Sheridan
Researcher in sustainable food systems, University of Melbourne
Rachel Carey
Research Fellow, University of Melbourne

The Conversations partners

View partners of The Conversation

You know that feeling when you open the fridge and are met with something on the nose.
We all know what food waste looks and smells like.

But food waste stinks in more ways than one. It is expensive, costing the average household
over A$2,200 a year, and it undermines the resilience and sustainability of our food supply.

A new report from our Foodprint Melbourne Project has estimated the amount of food that is
wasted in feeding Melbourne. We found that feeding Melbourne generates more than 900,000
tonnes of edible food waste every year, or over 200 kg per person.

This is enough to feed more than 2 million people for a year*.

http://theconversation.com/melbourne-wastes-200-kg-of-food-per-person-a-year-its-tim... 8/07/2016
Melbourne wastes 200 kg of food per person a year: it's time to get serious Page 2 of 3

Food waste occurs at different stages for each food type.

Foodprint Melbourne

Undermining sustainability

Growing this wasted food uses 180 gigalitres of water each year, or 113 litres per person per
day. This is equivalent to running your shower for an extra 10 minutes a day.

This wasted food also uses around 3.6 million hectares of land around 41 ha per person, or
more than 20 times the area of the Melbourne Cricket Ground.

And this wasted food is responsible for around 2.5 million tonnes of greenhouse gas
emissions, 60% of which is generated by food waste rotting in landfill, and the rest in
producing the wasted food.

This uneaten food is not only a source of unnecessary greenhouse gas emissions. It
represents a waste of natural resources that are in increasingly limited supply.

Australia is a water-scarce region that is likely to become drier due to climate change, while
only 6% of Australias land is suitable for growing crops.

With the associated waste of natural resources, high levels of food waste add to the challenge
of producing sufficient food to feed a growing population.

Reducing food waste

There are many ways to reduce food waste at home. These include making meal plans,
sharing leftover food with friends or neighbours, checking the fridge before going shopping
and storing food correctly.

The Cloud-Freezer app can help you to keep track of whats in your freezer and fridge. Worm
farms, bokashi bins and other forms of composting are also great ways to divert food waste
from landfill.

While we can all take steps to reduce food waste at home, we need to look at the bigger
picture. Our research shows that more than 60% of food waste is generated before food
reaches your fridge or freezer.

Strict standards defining the shape, size and colour of fresh fruit and vegetables in
supermarkets can mean that a significant proportion of a crop never leaves the farm.

Low prices for second-grade produce can make it financially unviable for farmers to pick, pack
and ship imperfect produce. Pressure to keep supermarket shelves full for appearances sake,
losses during food processing and storage problems also lead to food being wasted.

http://theconversation.com/melbourne-wastes-200-kg-of-food-per-person-a-year-its-tim... 8/07/2016
Melbourne wastes 200 kg of food per person a year: it's time to get serious Page 3 of 3

Initiatives that aim to make more imperfect fruit and vegetables available, such as
Woolworths Odd Bunch campaign, go some way to reducing this problem, but more needs to
be done.

Our research estimates that if food waste was halved across the food supply chain,
Melbourne could save 1.8 million hectares of land, 90 million litres of water and avoid 1.3
million tonnes of greenhouse gases each year.

We need to halve food waste

In recognition of the significant challenge that food waste represents to sustainable food
systems, the new Sustainable Development Goals set a target to halve the global food waste
per person that is generated by retailers and consumers by 2030.

The United States government has also set a national target to reduce food waste by 50% by
2030. It has established a cross-sector partnership of stakeholders across the food system to
tackle the problem.

The UK government has been an early mover in taking action to tackle food waste. In 2007, it
launched the WRAP Love Food Hate Waste program aimed at reducing food waste. An
evaluation in 2012 showed that avoidable waste of food and drink (that could have been
eaten) had fallen by 21% in five years following the launch of the program.

Most of this reduction has been in household food waste. The WRAP program is now working
with the food industry to reduce waste in other sectors. The successful UK Love Food Hate
Waste program aimed at reducing household food waste has been taken up by state
governments in Victoria and New South Wales.

Australia is developing a national food waste strategy the Food Waste 2025 Strategy and
stakeholders from across the food supply chain meet this month to discuss how to reduce
food waste.

Australia should follow suit in setting a target to halve food waste across the food supply chain
to put Australias food system on a more sustainable footing.

*Correction: This figure has been updated. It previously incorrectly stated that Melbournes
food waste is enough to feed 2,000 people per year.

Food
Waste
Farming
Greenhouse gases
Melbourne
Carbon emissions
food waste
Greenhouse Gas
ugly food
eating
Tweet163
Share334
Get newsletter

http://theconversation.com/melbourne-wastes-200-kg-of-food-per-person-a-year-its-tim... 8/07/2016
Running on empty: Australia's risky approach to oil supplies

Running on empty: Australias risky

approach to oil supplies
February 26, 2014 1.58pm AEDT

Vlado Vivoda
Research Fellow, Griffith Asia Institute, Griffith University

The Conversations partners

View partners of The Conversation

Australia only keeps an estimated 23 days' worth of fuel in the country. Bidgee/Wikimedia Commons

It might sound unlikely, but Australias fuel gauge is worryingly low. Were one of the worlds
top energy exporters, but our stocks of liquid fuels such as the oil on which almost the whole
transport sector depends are far from healthy.

Australias worsening liquid fuel security problem was detailed this week in the latest
instalment of a report prepared for the National Roads and Motorists' Association.

A year after the issue was first raised, there has been no action to address the growing risks,
both logistical and financial, of keeping so little fuel in the tank.

http://theconversation.com/running-on-empty-australias-risky-approach-to-oil-supplies-... 8/07/2016
Running on empty: Australia's risky approach to oil supplies

Lack of refinement

According to the report, refinery closures coupled with growing demand for petroleum have
increased Australias dependency on imported refined petroleum products. As a result, oil-
dependent sectors such as transport, mining and agriculture are vulnerable to supply-chain
interruptions or future oil-price spikes.

The situation is made worse by the fact that Australia is consistently the only International
Energy Agency (IEA) member state that fails to maintain the mandated stockpile of 90 days'
worth of net oil imports. Australias stockpiles stood at 57 days as of November 2013. The
NRMA report estimates Australias in-country stockpile, which excludes shipments en route to
Australia, at only 23 days.

With growing demand for liquid fuels and a continued hands-off approach from government,
by 2030 Australia is on track to find itself with no refining capacity, less than 20 days' worth of
liquid fuel stocks, and entirely at the mercy of the international oil market.

The reports author, retired Air Vice-Marshall John Blackburn, has called on the government to
step in. This years National Energy Security Assessment and Energy White Paper both offer
opportunities for public debate. The report also suggests making fuel stocks a matter of
national security risk analysis.

Hurting the hip pocket

Relying on imported oil isnt just risky its expensive too. In 2012, Australias net oil imports
were 561,000 barrels per day. At an average of US$109.08 a barrel for that year, the bill was
US$22.3 billion, or 1.7% of Australias gross domestic product.

At the same time, there are fears of a slowdown in Australias mineral export revenues as
Chinas economy loses pace.

The government seems to be refusing to consider the

potential threats to Australias coal, liquid natural gas (LNG)
and other mineral exports. Australian LNG, for example, costs
more than its competitors in North America and Africa and
Australias customers in Asia are increasingly critical of the
prices they are being charged. Without giving any
consideration to potential threats to Australias market share
in Asia, industry minister Ian MacFarlane has hailed forecasts Are Australias LNG exports
that assume a continued increase in energy export revenues. masking a problem with its oil
imports? Ken

The back story Hodge/Wikimedia Commons,

CC BY
Australias complacency over oil imports can be traced back
to the discovery in the mid-1960s of oil in Bass Strait. While not large by world standards,
these reserves increased Australias oil self-sufficiency from 10% to 70% in the years leading
up to the 1973 oil crisis, when world prices rose sharply.

Four decades later the picture is very different, although people dont necessarily know it.
Blinded by positive headlines about energy exports, only 1% of Australians consider the
energy crisis, petrol prices and fossil-fuel supplies to be our most pressing issue.

http://theconversation.com/running-on-empty-australias-risky-approach-to-oil-supplies-... 8/07/2016
Running on empty: Australia's risky approach to oil supplies

Perhaps Australias good fortune during the 1970s has helped to entrench a free-market
approach to the oil market, while other nations that were more badly burned by OPECs price-
gouging have taken stronger steps towards oil independence.

The Hawke governments free-market attitude continued through the Howard years, as
evidenced in Australias first ever Energy White Paper. Released in 2004, it openly
acknowledged Australias oil deficit. But instead of raising the issue as a significant energy
security threat, it sought to reassure by pointing to surpluses in gas, uranium and coal. The
policy implication was clear: as long as Australia remained a net energy exporter, its liquid fuel
balance was nothing to worry about.

None of this changed with Labors return to power. The 2012 Energy White Paper also struck
a positive tone, based on Australias position as an energy producer and net exporter, which
implied overall energy security. And now the Abbott government, in preparing this years
Energy White Paper, seems to consider the international oil market so resilient that the supply
of crude oil and refined petroleum to Australia is guaranteed. However, it is worth noting that
its recent issues paper includes a reference to asking whether Australia needs to increase its
stockpiles to meet international obligations.

Real impacts

Energy security issues may impact ordinary Australians directly through high consumer
prices, fuel shortages and the environmental impacts associated with energy consumption.
Given the importance of energy to the Australian economy, it is vital that the government
reevaluates its approach to energy policy and engages in a systematic risk analysis regarding
both Australias growing liquid fuel import dependency and potential threats to demand for its
energy exports.

While Australia remains faithful to the basic principles of free-market economics, if they are
wrong the consequences could be dire. A bipartisan hands-off approach and general public
complacency regarding Australias energy future may not serve the nations best interest in
the long term.

In terms of practical solutions to our growing liquid fuels import dependency, meeting the
IEAs mandated stockpile levels would be a first step in the right direction. The government
should also consider measures aimed at curbing the demand for liquid fuels in Australia.
These might include higher fuel-efficiency standards, improving public transport infrastructure,
promoting carpooling, or introducing congestion charges in the CBDs of major cities.

Energy security
Transport
OPEC
Exports
Petroleum
Oil refining
International Energy Agency
Tweet32
Share88
Get newsletter

http://theconversation.com/running-on-empty-australias-risky-approach-to-oil-supplies-... 8/07/2016

Environmental Nutrition: A New Frontier
for Public Health
Food systems must operate
within environmental constraints
to avoid disastrous consequences
for the biosphere. Such con-
straints must also take into ac-
count nutritional quality and
health outcomes.
Given the intrinsic relation-
ships between the environ-
mental sciences and nutritional
sciences, it is imperative that
public health embraces envi-
ronmental nutrition as the new
frontier of research and prac-
tice and begins a concerted
focus on the new discipline of
environmental nutrition, which
seeks to comprehensively ad-
dress the sustainability of food
systems.
We provide an overview to
justify our proposition, outline
a research and practice agenda
for environmental nutrition, and
explore how the complex re-
lationships within food systems
that affect public health could
be better understood through
the environmental nutrition
model. (Am J Public Health.
Published online ahead of print
March 17, 2016: e1e7. doi:10.
2105/AJPH.2016.303046)
Published online ahead of print March 17, 2016
Joan Sabat, MD, DrPH, Helen Harwatt, PhD, and Samuel Soret, PhD
B alancing the requirements of
food supply, health, and the
physical environmentthe
healthenvironmentdiet
trilemma1is of great environ-
mental and public health im-
portance and will require new
priorities for public health. We
are now confronted with the
challenge of adequately feeding
a growing world population,
which is expected to increase by
around 2 billion people by 2050,2
while maintaining environmen-
tal conditions that can sustain all
life forms.3 The signicant det-
rimental impacts of human ac-
tivities on Earths ecosystems in
the Anthropocene era have
substantial implications for public
health.4,5 Prior to the Anthro-
pocene era, relatively stable
environmental conditions,
including temperatures, fresh-
water availability, and bio-
chemical ows, enabled the
development of many and varied
cultures for millennia.6 How-
ever, since the 19th century,
when the Industrial Revolution
was extended, human actions
have had a signicant impact on
Earths environment on a global
scale. Among other factors,
a growing reliance on fossil fuels
together with changes in food
consumption habits and agricul-
tural norms have brought us to
the point of almost exceeding
Earths biophysical capacity
that is, what the planet can sustain
in terms of resource provision and
absorption of wastes. A multi-
national team of scientists
concluded that more than 1
boundary needed to maintain
AJPH
equilibrium of the planets bio-
physical processes has been
transgressed.6 Of the 9 bound-
aries dened, the safe operating
space has been exceeded for
biodiversity, land system change,
biochemical ows (the phos-
phorus and nitrogen cycles), and
climate change.7 A consensus
exists that rapid climate change is
not only real but is occurring
faster than previously thought.8
Global food provisions are ex-
pected to be widely affected by
climate change,9 further threat-
ening food security, which is
already an escalating global
concern.10 There is a high level of
condence among leading sci-
entists that climate change is al-
ready adversely affecting crop
production in several regions of
the world. For example, since
2007 there have been several
periods of rapid price increase for
cereals and other foods, following
extreme climate conditions in
key producing regions.11
The link between diet, health,
and environment is well estab-
lished.12,13 An integration of the
earth and life sciences with nu-
trition science must begin. It is
essential that the protection of
Earths resources and their in-
terdependencies become a factor
in the advancement of nutri-
tion.14 We explore why, as
ABOUT THE AUTHORS
Joan Sabate, Helen Harwatt, and Samuel Soret are with the School of Public Health, Loma
Linda University, Loma Linda, CA.
Correspondence should be sent to Helen Harwatt, School of Public Health, Loma Linda
University, Loma Linda, CA 92350 (e-mail: hharwatt@llu.edu/hharwatt@gmail.com).
Reprints can be ordered at http://www.ajph.org by clicking the Reprints link.
This essay was accepted December 17, 2015.
doi: 10.2105/AJPH.2016.303046
Sabat et al.
ANALYTIC ESSAYS
a priority, the public health remit
should expand to encompass the
(physical) environmental and
ethical dimensions of food sys-
tems.10,1517 The proposition to
formally include environmental
nutrition within the public health
remit goes beyond the scope of
current discussions on sustainable
diets14,1820 to systematically
consider the complex relation-
ships within food systems that
affect public health at a local and
global scale, such as toxic con-
taminants, biological agents, and
climate change. We also explore
how the inclusion of such pa-
rameters could be executed to
necessarily incorporate a com-
plete understanding of food sys-
tems, through the environmental
nutrition model.
THE EVOLUTION OF
FOOD SYSTEMS
From a rudimentary perspec-
tive, food systems comprise the
interaction of 4 main facets
resource inputs, drivers of de-
mand, food outputs, and waste
emissions, as depicted in Figure 1.
Historically, the production of
food was based on inputs of solar
energy and of rain and surface
water, and it featured multicrop
enterprises that simultaneously
Peer Reviewed Analytic Essays e1
ANALYTIC ESSAYS
Natural
Resources
IN
PU
T PU
Food System
UT
TP
OU
Waste
Emissions
FIGURE 1Basic Model of a Food System
raised plants and animals and
utilized animal waste as fertilizer.
It had relatively few detrimental
impacts on the physical envi-
ronment, partly because of pop-
ulation size and production
techniques. Plants obtained nu-
trients from the soil and energy
from the sun. Animals obtained
nutrients and energy by eating
plants, other animals, or both.
Manure, crop rotation, poly-
cultures, and routinely aban-
doning elds and allowing them
to be taken over gradually by
natural vegetation were the main
sources of maintaining soil vital-
ity. Agriculture evolved in very
specic ecological settings that
facilitated the domestication and
selection of 4 main crops: wheat,
rice, corn, and potatoes; these
became the key foods to support
the expansion of human
populations.15
Although instances of in-
tensive agriculture have existed
for thousands of years, agriculture
largely became industrialized
over the past 2 centuries, which
was inevitable under the domi-
nant ideologies of political power
and economic growth that foun-
ded the Industrial Revolution.21
e2 Analytic Essays Peer Reviewed
Societal
Demands
T and waste emissions (Figure A,
IN
OU
TP
UT
Food
In addition to the manufacture
of synthetic fertilizers, the use
of vitamins and antibiotics allowed
more intensive types of agricul-
ture, including animal farming,
to become dominant. Globally,
agricultural production doubled
4 times between 1820 and
1975.22 Monocultures replaced
polycultures as the main crop
production model. The
Green Revolution saw the de-
velopment of disease-resistant
high-yield grain varieties and
the implementation of irriga-
tion, further increasing the in-
tensity of food production.
Despite a doubling of population
over the past 4 decades, the
Green Revolution has enabled
agriculture to meet the worlds
food needs.23 However, in-
tensifying the production of
food has adversely affected
biodiversity3,2426which is de-
clining at a global level and under
continued threat6,27and human
health through exposure to anti-
microbial agents such as antibi-
otics,28 biological agents such as
viruses, and toxic chemicals such
as some pesticides.29
The global food system now
has enormous adverse impacts on
Sabat et al.
the physical environment. Since
the Industrial Revolution, the
food system has become much
more complex, with an increased
range of inputs, drivers, outputs,
available as a supplement to the
online version of this article at
http://www.ajph.org). Prevail-
ing political, economic, and
cultural governing structures in-
uence the characteristics of in-
puts, drivers, outputs, and waste
emissions, and also how these key
aspects relate to each other. The
life cycle of food types has ex-
panded to include new stages
such as manufacturing, packag-
ing, long-distance transportation,
storage, and waste. Consequently,
the environmental costs of food
production are now substantial,
including the degradation of in-
land waterways, nitrogen and
phosphorus pollution of coastal
marine ecosystems leading to
dead zones, the development of
photochemical smog, and rising
global concentrations of nitrous
oxide and methane, which are
both greenhouse gases (GHGs)
contributing to the process
of climate change.30,31 Vast
amounts of oil and gas are used to
provide raw materials and energy
for the manufacture of fertilizers
and pesticides and throughout all
stages of food production, from
planting, irrigation, feeding, and
harvesting to processing, cold
storage, distribution, and pack-
aging. In addition, fossil fuels are
essential in the construction and
repair of equipment and in-
frastructure needed to facilitate
this industry, including farm ma-
chinery, processing facilities,
storage, transportation, and
roads.32 Hence, the current food
system is one of the biggest
consumers of fossil fuels and
one of the largest emitters of
GHGs.3,9,33
Wastage, including food loss,
is also a major issue in current
AJPH Published online ahead of print March 17, 2016
food systems. Analyses of the US
food system have shown that
avoidable food waste accounts for
between 29% and 40% of annual
production.34,35 At a global level,
it is estimated that one third of all
food produced for human con-
sumption is lost or wasted,36 with
much more of this occurring in
the industrialized world than in
developing countries.37
Demand for food is largely
dened by the technology, social
structure (including cultural
norms), availability of certain
foods, and food policy in a given
society. Societal demand and
dietary choices are of funda-
mental importance in terms of
the severity of environmental
impacts.1820,38 Certain food
groups have a much larger re-
source requirement and hence
environmental footprint, with
animal-based products generally
having the greatest impacts and
plant foods having the least.3943
This is important not only from
a current perspective but also in
consideration of the predicted
increase in demand for animal
products from emerging econo-
mies.44 The environmentally
hazardous aspects of animal
production include the use of
growth-promoting antibiotics
and manure as a waste product.45
Furthermore, it has been esti-
mated that livestock production
alone accounts for around 15% of
global GHG emissions,46 thus
contributing signicantly to cli-
mate change. The environmental
impacts of such dietary choices
in turn have signicant public
health implicationsfor exam-
ple, air and water contamination
from hazardous chemicals, ani-
mal waste, and zoonotic dis-
ease.5,45 In addition, climate
change is already having adverse
human health consequences and
is expected to have more,47 in-
cluding more deaths through
temperature extremes, increased

spread of infectious disease,4 re-
duced food yield in some regions,
and increased threat to food
security.11,39
THREE DIMENSIONS
OF NUTRITION
SCIENCES
Currently, human nutrition
covers 2 main aspects: individuals
and communities. Although
public health and commu-
nity are not necessarily in-
terchangeable, the use of
community here purposefully
allows differentiation of the
existing and proposed aspects of
public health nutrition. At its core,
the discipline of nutrition is, and
will remain, a biological science
centered on biochemistry, physi-
ology, and medicine. It is con-
cerned with the interaction of
food and other sources of nutri-
ents and with the physiological,
metabolic, and genomic systems
of the human body. By extension,
it addresses the effect of these as-
pects on health and disease48
(see the box on page e4). Public
health nutrition emerged in
recognition of the importance
of community welfare, including
social, economic, political, and
human aspects. Thus, nutrition
extended its disciplinary scope to
incorporate such concerns. This
research expansion has had a tre-
mendous impact on public policy,
both nationally and internationally.
Specically, the public health
nutrition agenda incorporates
a range of social and multidisci-
plinary sciencesincluding ep-
idemiology, anthropology, and
sociologyto address nutri-
tional deciencies, disease epi-
demics, disease prevention, and
food security (see the box on
page e4).
As public health nutrition
emerged from the science of
nutrition for individuals, we are
Published online ahead of print March 17, 2016
now on the forefront of a new physical and social dimensions
expansion to incorporate envi- of the food system and its un-
ronmental considerations (Figure B, intended and undesired conse-
available as a supplement to the quences on the biosphere,
online version of this article at including public health impacts.
http://www.ajph.org). The nu- Although this is relevant to all
trition of individuals and com- types of food systems, Western-
munities can only be maintained ized industrial food systems
within an environmentally sus- centrally controlled by govern-
tainable context, which is cur- ments or multinational con-
rently under serious threat.6 glomerates are the selected focus,
Environmental nutrition seeks to given their growing dominance
address the sustainability of food and inuence on other food
systems by integrating the envi- systems.
ronmental sciences with the nu- Through the ENM, the inputs
tritional sciences, addressing and outputs related to the whole
a range of issues from production life cycle are considered and
practices to societal demands on quantied for individual food
a biospheric scale (see the box on items, which can then be used to
page e4). The existing multidis- assess and compare whole diets.
ciplinary nature of public health The life cycle of food includes the
nutrition lends itself to in- production, processing, trans-
corporating further disciplines. portation, storage, retail, and
Given the intrinsic interrelation- consumption and disposal prac-
ships between the environment tices employed. For each food
and nutrition,26 the current public item, the ENM can produce
health nutrition agenda must ex- a comparable list of inputs such as
pand to embrace environmental water, energy, land, and chem-
nutrition as a new frontier. icals (including fertilizer and
pesticide), and of outputs, in-
cluding GHG emissions, manure,
and waste throughout the entire
life cycle. One of the aims of the
THE ENVIRONMENTAL ENM is to calculate a compre-
NUTRITION MODEL hensive range of environmental
Incorporating environmental outputs and impacts; therefore,
nutrition into public health will for certain food items, additional
necessitate a comprehensive un- aspects must be considered. For
derstanding of food systems and example, for livestock products,
their impact on the biosphere. In respiration and manure can be
an effort to clarify the interaction factored into the life cycle as-
between current food systems, sessment. Another important as-
the environment, and public pect to take into account is the
health, we propose an environ- missed opportunity for carbon
mental nutrition model (ENM). sequestration in relation to land
This model, portrayed in use and crop type.42 An assess-
Figure 2, is intended to provide ment of the biodiversity impacts
a useful didactic tool to explain, should also be included.25
understand, and ultimately con- Societal demands would be
tribute to the necessary modi- considered in the ENM, in-
cations and changes to the cluding consumer preferences,
current food system to achieve policy, marketing, and social
sustainability. The ENM is dynamics. The nutritional ade-
a conceptual framework that quacy of each food item or di-
encompasses the multifaceted etary pattern would also be
AJPH Sabat et al.
ANALYTIC ESSAYS
assessed, resulting in a compre-
hensive assessment of each food
item or dietary pattern in terms of
the environmental and public
health impacts.
As depicted in Figure 2, there
is a direct feedback between
natural resource use and envi-
ronmental pollution as the latter
inuences the former and vice
versa. The types and quantities of
resources used inuence the type
and amount of pollution created,
which in turn compromises the
quality of natural resources. The
required output of the food sys-
tem (i.e., food) serves and in turn
inuences societal demands and
composes the dietary patterns of
diverse populations. Diet-related
health issuesincluding obe-
sity, chronic disease, and
malnourishmentare a major
consequence of the food system.
Food consumption patterns or
consumer preferences can inu-
ence the extent of environmental
degradation (as outlined in the
section The Evolution of Food
Systems). Policy can also sig-
nicantly inuence consumer
demandsfor example, by sub-
sidizing the production of certain
food types more than others,
which in turn articially lowers
retail price and increases con-
sumption. In addition, policy and
governance are crucial factors for
food sovereigntyfor example,
food systems can become largely
controlled by a small number
of corporations rather than by
everyone involved, including
consumers.49
The environmental impacts of
the food system can in turn in-
uence nutritional outcomes.
For example, GHGs are an out-
put of the food system that
contribute to climate change,
which in turn affects food pro-
duction through reduced yields
and nutritional content.11 These
impacts subsequently increase
malnutrition and food insecurity.
Peer Reviewed Analytic Essays e3
ANALYTIC ESSAYS

FOCUS OF NUTRITION DIMENSIONS

Nutrition Dimensions
Human Nutrition Community Nutrition Environmental Nutritiona

Scientic disciplines covered by each dimension Biological sciences Social and multidisciplinary sciences Environmental sciences
d Biochemical d Epidemiology d Physical

d Physiological d Anthropology d Atmospheric

d Medical d Political science d Ecology

d Economics d Geography

d Sociology

Scope of Dimensions

Individuals Communities Biosphere

Issues addressed Growth Epidemics of chronic disease Agricultural practices

Adequate diets Nutritional deciencies Sustainable food systems
Nutrient requirements Disease prevention Societal demands and marketing
Disease management Food security Benign food technologies

Public policy
a
Environmental nutrition also covers disciplines covered by human and community nutrition.

Coupled with the global impacts
of climate change which the food
system signicantly contributes
to,9 malnutrition and food in-
security in turn raise issues related
to environmental justice and
intra- and intergenerational dis-
tributive justice.10,16
As a specic example, to fully
demonstrate the complex range
of inputs and outputs of the food
system, we provide a case study of
beef. In the United States, the
production of 1 kilogram of beef
requires the input of 10 kilograms
of grain feed,50 more than 8000
liters of water, almost 8000
kilojoules of energy, 150 grams of
fertilizer, 7 grams of pesticides,38
and 21 square meters of land.51
Inputs on the demand side in-
clude marketing, cultural norms,
technology, and such policies as
government subsidies, either as
direct payments to farmers or
through public nutritional assis-
tance programs, both of which
increase availability to con-
sumers. Governance issues can
also be an important consider-
ation; in the United States, for
example, beef processing is
largely controlled by 4 corpora-
tions. The environmental out-
puts of concern include solid,
liquid, and air pollution (from
animal manure, urine, blood, and
hair; antibiotic residues; fossil fuel
combustion; nitrogen and phos-
phorus), and GHGs (primarily
methane, nitrous oxide, and
carbon dioxide). Other outputs
include nutrients for human
consumption and contributions
to chronic disease52 and in-
fectious disease.5 Equity and
food security issues are relevant,
as producing livestock for
human consumption is a very
resource-inefcient process, with
about an 89% loss of biomass
through the life cycle.50 Com-
pared with producing 1 kilogram
of protein from kidney beans, the
production of 1 kilogram of
protein from beef requires ap-
proximately 18 times more land,
10 times more water, 9 times
more fuel, 12 times more fertil-
izer, and 10 times more pesti-
cide.51 In addition, the United
States uses 67% of total calorie
production for livestock feed and
could feed almost 3 times as many
people per cropland hectare by
shifting crop calories to direct
human consumption.53
ENVIRONMENTAL
NUTRITION AGENDA
Expanding public health to
include environmental nutrition
will require dialogue and research
in 4 overarching areas of the
global food system: food pro-
duction, food consumption,
nutrition policy, and the in-
tegrated assessment of these areas.
So far, food production has
attracted the most attention,
much of which is focused on
reducing the inputs required
through technological means,
rather than replacing less efcient
food crops with more efcient
ones.13 Hence, it is essential to
identify resource-intensive
commodities and those that
could replace them, and also to
assess the relative health impli-
cations. Although biodiversity
loss ranks among the major
drivers of ecosystem change, it
remains a relatively neglected
area of investigation.54 Loss of
biodiversity also has implications
for public health, including an
exacerbation of existing health
inequities through exposure
to environmental hazards or
through loss of livelihoods.49,55
Hence, biodiversity impacts must
also be a strong consideration in
food production.24
To successfully feed a growing
world population while pro-
tecting and remediating the en-
vironment, food consumption
patterns must be considered and
altered.14,56 The potential level
of control that consumers have
over their purchases, particularly
in developed nations, implies
a level of elasticity in demand and
hence a signicant potential to
modify the environmental im-
pacts incurred within the food
system and to alter personal
health outcomes. The level of
consumer control depends on
many factors, including the
amount of money they have
available and their access to al-
ternative foods. The key ques-
tions in this area are how, where,
and when can the sustainability
and nutritional quality of dietary

e4 Analytic Essays Peer Reviewed Sabat et al. AJPH Published online ahead of print March 17, 2016

choices be simultaneously opti-
mized? So far, there is a lack
a comprehensive analysis to ad-
dress these issues.13
The main area requiring at-
tention within policy research
relates to the integration of policy
options to address problems from
a holistic perspective. For ex-
ample, there is a need to assess
how food production can be
adjusted on a large scale to reduce
environmental impacts and pro-
vide health-promoting foods
while also accounting for the
required changes in food con-
sumption on the demand side.
Essentially, policies need to be
developed in a way that ensures
that all of the stages from food
production to postconsumption
are coherent and cooperative
with a global-health-promoting,
environmental-conservancy
agenda. This could involve
a mixture of regulatory and scal
measures.13 The latter, which
Natural World
Resources
Sun
Water
Land
Energy
Chemicals
Pollution
Solid and fluid waste
Biological contaminants
Greenhouse gas emissions
Climate
Change
FIGURE 2The Environmental Nutrition Model, Including an Example of the Relationships Between
Climate Change and Food Security
Published online ahead of print March 17, 2016
have been explored to some
extent, have been shown to re-
duce dietary GHG emissions.57
However, there remains much to
explore, particularly in relation to
how societal demands can be
adjusted to embrace healthy and
sustainable foods, and how
measures such as a restructuring
of food subsidies could support
such change.13
Given the interrelationships, it
is essential to integrate food
production, consumption, and
other relevant aspects in a way
that allows a comprehensive
analysis of food systems. It is
important to identify foods that
are both sustainable and healthy
and to develop a food-labeling
scheme to communicate such
information to consumers. De-
veloping a set of sustainability and
social indicators is likely to be
important in assessing the scale
and speed of the transition.
Measures that increase the
Human Societies
Food System
Life Cycle
Agriculture Processing
Livestock Packaging
Fisheries Transportation
Storage
Retail
Waste
Consumption Patterns of
Undernourished
Low Regional Food
Food Yields and/or Insecurity
Reduced
Nutritional Content
AJPH
sustainability of food systems
should be evaluated in terms of
whether they require techno-
logical change for the producer or
behavioral change for the con-
sumer, and the type and level of
policy intervention required to
implement such measures.
To conduct comprehensive
assessments of food systems, the
development of an analytical tool
is required. Such a tool would
take into account the health,
nutritional, and environmental
aspects of food choices and allow
for the exploration of additional
factors such as policy scenarios,
monetary costs for the consumer,
and equity and cultural consid-
erations.12,58 One of the key
components within such a tool is
an environmental life cycle as-
sessment of foods across the entire
food system. The growing body
of research in the life cycle as-
sessment of food offers a useful
insight into the current system
Demands
Consumer
preferences
Policy
Technology
Marketing
Social dynamic
Cultural norms
Food
Diverse Populations
Overfed
Malnutrition Obesity
Deficiency Chronic
diseases diseases
Sabat et al.
ANALYTIC ESSAYS
of food delivery, which can be
simple or complex. For example,
the life cycle of a potato involves
growing, harvesting, washing,
packing, transporting to a dis-
tributor and then to a retailer,
purchasing at a store, transporting
to a home, cooking, and dis-
posing of any waste. This is
a fairly straightforward route
compared with that of a tomato
grown to produce ketchup. The
process is more complex, and
even more so if the ketchup
becomes one of several food
items in a ready mealthe route
becomes staggering.
Currently, the focus of life
cycle assessments is on the GHGs
from dietary patterns. This must
be expanded to incorporate
a broad range of environmental
aspects, including water use,
energy use, chemical use (pesti-
cide and fertilizer), land use, and
waste products (including solid,
gaseous, and efuent) from each
aspect.12 The focus on GHGs has
partly arisen through a lack of
sufcient information on other
environmental aspectsfor ex-
ample, there is currently very
little analysis of the impacts of
global food wastage from an
environmental perspective.36
Hence, research efforts are also
required to develop the data in-
puts that would facilitate the in-
clusion of such environmental
impacts. In addition, regionally
specic analyses must be de-
veloped to increase accuracy and
applicability.12 Most life cycle
assessments currently focus on
food production, so there is
a need to expand the analysis to
include other stages of the food
system.59
CONCLUSIONS
Paradoxically, although the
current food system is reliant on
the natural world, it is also a major
Peer Reviewed Analytic Essays e5
ANALYTIC ESSAYS
contributor to its degradation.
Food production, processing,
transportation, storage, and
packaging, coupled with an
ever-increasing population and
consumer demand, is using nat-
ural resources and producing
waste at unsustainable rates. In
turn, this results in serious im-
plications for public health, in-
cluding climate change, food
insecurity, and changing food
consumption patterns. The
food system must operate within
environmental constraints to
avoid disastrous consequences for
the biosphere. Given the intrinsic
relationships between the envi-
ronmental sciences and the nu-
tritional sciences, it is imperative
that public health research and
practice begin a concerted focus
on the new discipline of envi-
ronmental nutrition, which seeks
to comprehensively address
the sustainability of food
systems.
CONTRIBUTORS
J. Sabat obtained funding, conceptual-
ized and designed the research, and edited
the article. H. Harwatt conceptualized the
research and wrote and edited the article.
S. Soret conceptualized the research and
edited the article.
ACKNOWLEDGMENTS
This research was funded by the McLean
Fund for Nutrition Research.
The concepts described in this essay
were presented at the 2nd World Congress
of Public Health Nutrition; September
2325, 2010; Porto, Portugal.
Note. The funder was not involved in
the content or production of this essay.
HUMAN PARTICIPANT
PROTECTION
Institutional review board approval was
not needed for this article because no
human participants were involved.
REFERENCES
1. Tilman D, Clark M. Global diets link
environmental sustainability and human
health. Nature. 2014;515(7528):518522.
2. Food and Agriculture Organization of
the United Nations. How to feed the
world in 2050. Rome, October 1213,
2009. Available at: www.fao.org/wsfs/
forum2050/wsfs-background-
documents/wsfs-expert-papers/en.
Accessed December 1, 2014.
e6 Analytic Essays Peer Reviewed
3. Foley JA, Ramankutty N, Brauman
KA, et al. Solutions for a cultivated planet.
Nature. 2011;478(7369):337342.
4. McMichael AJ. Globalization, climate
change, and human health. N Engl J Med.
2013;368(14):13351343.
5. Akhtar AZ, Greger M, Ferdowsian H,
Frank E. Health professionals roles in
animal agriculture, climate change, and
human health. Am J Prev Med. 2009;36(2):
182187.
6. Rockstro m J, Steffen W, Noone K,
et al. A safe operating space for humanity.
Nature. 2009;461(7263):472475.
7. Steffen W, Richardson K, Rockstro m J,
et al. Planetary boundaries: guiding hu-
man development on a changing planet.
Science. 2015;347(6223):736746.
8. Smith JB, Schneider SH, Oppenheimer
M, et al. Assessing dangerous climate
change through an update of the In-
tergovernmental Panel on Climate
Change (IPCC) reasons for concern.
Proc Natl Acad Sci USA. 2009;106(11):
41334137.
9. Vermeulen SJ, Campbell BM, Ingram
JSI. Climate change and food systems. Annu
Rev Environ Resour. 2012;37:195222.
10. Loos J, Abson DJ, Chappell MJ, et al.
Putting meaning back into sustainable
intensication. Front Ecol Environ. 2014;
12(6):356361.
11. Porter J, Xie L, Challinor A, et al. Food
security and food production systems. In:
Field CB, Barros VR, Dokken DJ, et al.,
eds. Climate Change 2014: Impacts, Adap-
tation, and Vulnerability. Part A: Global and
Sectoral Aspects. Contribution of Working
Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change.
Cambridge, UK: Cambridge University
Press; 2014:485533.
12. Heller MC, Keoleian GA, Willett
WC. Toward a life cycle-based, diet-level
framework for food environmental im-
pact and nutritional quality assessment.
Environ Sci Technol. 2013;47(22):
1263212647.
13. Garnett T. Food sustainability: prob-
lems, perspectives and solutions. Proc Nutr
Soc. 2013;72(1):2939.
14. Sabat J, Soret S. Sustainability of
plant-based diets: back to the future. Am J
Clin Nutr. 2014;100(suppl 1):476S482S.
15. Uauy R. Dening and addressing the
nutritional needs of populations. Public
Health Nutr. 2005;8(6A):773780.
16. McAlpine CA, Seabrook LM, Ryan
JG, et al. Transformational change: cre-
ating a safe operating space for humanity.
Ecol Soc. 2015;20(1). Available at: http://
www.ecologyandsociety.org/vol20/iss1/
art56. Accessed May 1, 2015.
17. Buttriss J, Riley H. Sustainable diets:
harnessing the nutrition agenda. Food
Chem. 2013;140(3):402407.
Sabat et al.
18. Scarborough P, Appleby P, Mizdrak
A, et al. Dietary greenhouse gas emissions
of meat-eaters, sh-eaters, vegetarians and
vegans in the UK. Clim Change. 2014;125
(2):179192.
19. Hallstro m E, Carlsson-Kanyama A,
Bo rjesson P. Environmental impact of
dietary change: a systematic review. J
Clean Prod. 2015;91:111.
20. Soret S, Mejia A, Batech M, Jaceldo-
Siegl K, Harwatt H, Sabat J. Climate
change mitigation and health effects of
varied dietary patterns in real-life settings
throughout North America. Am J Clin
Nutr. 2014;100(suppl 1):490S495S.
21. Cannon G. The rise and fall of dietetics
and of nutrition science, 4000 BCE2000
CE. Public Health Nutr. 2005;8(6A):701705.
22. Scully M. Dominion: The Power of Man,
the Suffering of Animals, and the Call to
Mercy. New York, NY: St. Martins
Grifn; 2003.
23. Tilman D. The greening of the green
revolution. Nature. 1998;396(6708):
211212.
24. Tscharntke T, Clough Y, Wanger TC,
et al. Global food security, biodiversity con-
servation and the future of agricultural in-
tensication. Biol Conserv. 2012;151(1):5359.
25. Burlingame B, Dernini S, eds. Sus-
tainable Diets and Biodiversity: Directions and
Solutions for Policy, Research and Action.
Rome, Italy: Food and Agriculture Or-
ganization of the United Nations; 2012.
26. Sutton MA, Bleeker A, Howard CM,
et al. Our Nutrient World: The Challenge to
Produce More Food and Energy With Less
Pollution. Edinburgh, UK: Centre for
Ecology and Hydrology; 2013.
27. Ripple WJ, Newsome TM, Wolf C,
et al. Collapse of the worlds largest her-
bivores. Sci Adv. 2015;1(4). Available
at: http://advances.sciencemag.org/
content/1/4/e1400103.full-text.pdf
+html. Accessed May 1, 2015.
28. Antimicrobial Resistance: Global Report
on Surveillance. Paris, France: World
Health Organization; June 2014.
29. Aktar MW, Sengupta D, Chowdhury
A. Impact of pesticides use in agriculture:
their benets and hazards. Interdiscip
Toxicol. 2009;2(1):112.
30. Eshel G, Martin PA. Geophysics and
nutritional science: toward a novel, uni-
ed paradigm. Am J Clin Nutr. 2009;89(5):
1710S1716S.
31. Vitousek PM, Naylor R, Crews T,
et al. Nutrient imbalances in agricultural
development. Science. 2009;324(5934):
15191520.
32. Pew Commission on Industrial Farm
Animal Production. Putting meat on
the table: industrial farm animal pro-
duction in America. 2008. Available at:
http://www.livablefutureblog.com/
pdf/Putting_Meat_on_Table_FULL.
pdf. Accessed December 22, 2010.
AJPH Published online ahead of print March 17, 2016
33. Church N. Why our food is so de-
pendent on oil. Energy Bulletin, 2005.
Available at: http://www.energybulletin.
net/node/5045. Accessed December 1,
2014.
34. Gunders D. Wasted: how America is
losing up to 40 percent of its food from
farm to fork to landll. Natural Resources
Defense Council issue paper, August
2012. Available at: https://www.nrdc.
org/food/les/wasted-food-ip.pdf.
Accessed December 1, 2014.
35. Venkat K. The climate change and
economic impacts of food waste in the
United States. Int J Food System Dynamics.
2011;2(4):431446.
36. Food Wastage Footprint: Impacts on
Natural ResourcesSummary Report.
Rome, Italy: Food and Agriculture Or-
ganization of the United Nations; 2013.
37. Gustavsson J, Cederberg C, Sonesson
U, van Otterdijk R, Meybeck A. Global
Food Losses and Food Waste. Rome, Italy:
Food and Agriculture Organization of the
United Nations; 2011.
38. Marlow HJ, Harwatt H, Soret S,
Sabat J. Comparing the water, energy,
pesticide and fertilizer usage for the pro-
duction of foods consumed by different
dietary types in California. Public Health
Nutr. 2015;18(13):24252432.
39. McMichael AJ, Powles JW, Butler
CD, Uauy R. Food, livestock production,
energy, climate change, and health. Lancet.
2007;370(9594):12531263.
40. Steinfeld HGP. Livestocks Long
Shadow: Environmental Issues and Options.
Rome, Italy: Food and Agriculture Or-
ganization of the United Nations; 2006.
41. Garnett T. Where are the best op-
portunities for reducing greenhouse gas
emissions in the food system (including the
food chain)? Food Policy. 2011;36(suppl 1):
S23S32.
42. Schmidinger K, Stehfest E. Including
CO2 implications of land occupation in
LCAsmethod and example for livestock
products. Int J Life Cycle Assess. 2012;17(8):
962972.
43. Stehfest E, Bouwman L, van Vuuren
DP, den Elzen MGJ, Eickhout B, Kabat P.
Climate benets of changing diet. Clim
Change. 2009;95(12):83102.
44. Food and Agriculture Organization.
Food outlook. Global market analysis.
November 2012. Available at: http://
www.fao.org/docrep/016/al993e/
al993e00.pdf. Accessed December 1, 2014.
45. Horrigan L, Lawrence RS, Walker P.
How sustainable agriculture can address
the environmental and human health
harms of industrial agriculture. Environ
Health Perspect. 2002;110(5):445456.
46. Gerber P, Steinfeld H, Henderson B,
et al. Tackling Climate Change Through
LivestockA Global Assessment of Emissions
and Mitigation Opportunities. Rome, Italy:
Food and Agriculture Organization; 2013.
REVIEW

CURRENT
OPINION What an anticardiovascular diet should be in 2015
David R. Jacobs Jr a and Linda C. Tapsell b

Purpose of review
Given scientific and public debate about optimal diet to prevent cardiovascular disease, and interest in diet
and other chronic diseases, we propose that following a few simple dietary principles would reduce
chronic disease incidence.
Recent findings
Nutrition research has been criticized for focusing on individual nutrients and foods, treated like drug
therapy. With a few important exceptions, clinical trials of supplemental nutrients have not shown benefit.
Although highly specific nutrition information is elusive, diet patterns have provided consistent answers,
important for public health. Observational cohort studies have found that some dietary patterns are
reported with high reliability over long periods and predict future cardiovascular and other inflammatory-
related diseases. Two randomized clinical trials confirmed this finding. There are many common features
of Mediterranean and prudent diets, particularly the plant-centered aspect, coupled with variety of foods
eaten. A dietary pattern characterized by high fruit, vegetable, legume, whole grain, nut, berry, seed, and
fish intakes, and possibly by intakes of dairy, coffee, tea, chocolate, and alcohol (not in excess), but low
meat and detrimentally processed foods is associated with reduced incidence of cardiovascular disease
and rates of noncardiovascular, noncancer chronic inflammatory-related mortality.
Summary
A plant-centered diet may be broadly recommended.
Keywords
cardiovascular disease, chronic disease, diet patterns, plant-centered diet

INTRODUCTION methods as biochemistry and molecular genetics)

There is substantial scientific and public debate
about optimal diet to prevent cardiovascular and
other chronic diseases. This diet can be represented
as a set of menu plans, a list of foods to choose from,
or the general characteristics of overall daily food
intake. US News and World Report captures this
debate annually in Best Diets, which in 2015 sum-
marized the rankings by a panel of nutrition pro-
fessionals of 35 different diet plans [1]. The US
Dietary Guidelines have made nutrition recom-
mendations, which for many years emphasized
reduction in total and saturated fat as one corner-
stone, often tending to avoid reference to specific
foods [2,3]. Many people have recommended vari-
ous forms of low carbohydrate diets (for example,
the Atkins diet). Many ways to view what to eat
have been formalized.
Concurrently, epidemiology has been severely
criticized [4], particularly observational nutritional
epidemiology [5], as well as the overwhelming body
of detailed studies of nutrition mechanisms [6].
Temple [6] classified the research as complex (the
detailed study of disease mechanisms using such
or simple (epidemiology, intervention trials, and
analogous studies on animals) and commented that
complex research has received most resources, but
more practical value has come from simple research.
Archer et al. [7] found that nutritional surveillance
in the US is seriously flawed because energy intake is
clearly underestimated. They concluded that the
ability to estimate population trends in caloric
intake and generate empirically supported public
policy relevant to diet-health relationships from
US nutritional surveillance is extremely limited,
thereby eliminating all nutritional epidemiology
that might be generated from this source. Ioannidis
a
Division of Epidemiology and Community Health, School of Public
Health, University of Minnesota, Minneapolis, MN, USA and bSchool
of Health Sciences, University of Wollongong, Wollongong, NSW,
Australia
Correspondence to David R. Jacobs, Division of Epidemiology and
Community Health, School of Public Health, University of Minnesota,
Minneapolis, MN 55454, USA. E-mail: jacob004@umn.edu
Curr Opin Lipidol 2015, 26:270275
DOI:10.1097/MOL.0000000000000184

www.co-lipidology.com Volume 26 ! Number 4 ! August 2015

Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.


KEY POINTS
! Much of the dietary advice over the past decades to
prevent cardiovascular disease has focused on the
reduction of saturated fat intake, associated with lower
coronary heart disease mortality and altered meat and
dairy intake.
! Nevertheless, this message appears not to be correct in
all respects and may have been too restrictive. It missed
important focuses on nutritionally rich plant food intake.
! Reductionist science operates through possibly
misleading syllogisms; conclusions about saturated fat
fostered a view of nutrition that was unnecessarily
restrictive and fostered growth of the dietary
supplement industry, despite contrary evidence from
long-term randomized clinical trials.
! There are many excellent approaches to a diet that
prevents chronic disease, but such approaches do not
depend on total fat or total carbohydrate intake.
! Diets that are associated with reduced chronic disease
risk and have been successful in long-term randomized
clinical trials are characterized by high fruit, vegetable,
legume, whole grain, nut, berry, seed, unrefined
unsaturated oils, and fish intakes, and more
speculatively intakes of dairy, coffee, tea, chocolate,
and alcohol (not in excess), but are low in meat and
detrimentally processed foods.
[5] stated, Definitive solutions will not come from
another million observational papers or a few small
randomized trials. He found a poor track record of
observational claims, but lauded two long-term
randomized clinical trials of dietary patterns [8,9 ].
&&
Some commentators have focused on nutrition-
ism [10 ,11], a formulation in which nutrition de-
&
cisions and policies are focused largely on nutrients,
with little or no consideration of the complex nature
of foods. Nutritionism at least implicitly supports
the financially thriving supplement industry. Sev-
eral isolated food chemicals have been tested as
drugs in randomized clinical trials, with no effect
or a harmful effect on long-term health [12]. A
recent clinical trial of niacin and the antiflushing
agent laropiprant found adverse effects for infec-
tion, bleeding, and diabetes, among other effects
[13]. Exceptions related to age-related macular
degeneration [14] and maternal folate supplement-
ation for prevention of fetal neural tube defects [15]
showed benefit. Similarly, viewing diet through the
lens of a single nutrient (for example, low fat, low
saturated fat, or low carbohydrate messages) is likely
to discard good foods as well as less desirable ones.
For example, among highly regarded foods, nuts are
high in fat and whole grain foods and all fruits and
0957-9672 Copyright ! 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Anticardiovascular diet in 2015 Jacobs and Tapsell
vegetables are high in carbohydrates. Dairy products
are high in saturated fat, but findings related to dairy
foods and various outcomes suggest benefit [16].
Therefore, concepts about the best diet for long-
term cardiovascular health have been changing.
This review reconsiders diet, focusing on dietary
patterns and long-term chronic disease outcomes.
A MULTIDIMENSIONAL VIEW OF FOOD
Food is complex [17]. There are major differences
between drug and food research [17,18 ]. Most
&&
salient difference is the nature of a drug versus food.
Drugs are well defined, purified compounds which
can be unobtrusively consumed and compared with
another well-defined regimen, such as a placebo or
an accepted therapy. Foods are the opposite of well
defined and constant: we eat formerly living organ-
isms that have all the complexity and variety of
life forms, nurtured in various ways by farmers,
breeders, and genetic modifiers, consumed several
times per day, every day, one food balanced against
another to maintain energy balance, and affecting
chronic disease as some sort of average across foods
and time. Many authorities require the internal
validity of long-term, randomized, clinical trials
characteristically used in inferences about drugs
before they consider evidence of the most persuasive
nature. Such trials are difficult at best for study of
food, yet it is critical to appraise health value of
foods. To this end, several authors have favored
reappraisal of the body of nutrition studies with a
view toward valid inference; and a more holistic
view of nutrition [17,18 ,1921], in which dietary
&&
policy and even understanding of nutrition itself
does not need a meticulous and reductionist break-
down into separate constituent parts [19].
Reductionist science employs syllogisms. An
example is the logic underlying avoidance of satu-
rated fat. A syllogism considers three propositions,
A, B, and C. It asserts that if A being true implies that
B is true, and B being true implies that C is true, then
it follows without further testing that A being true
implies that C is true. The problem of judging the
validity of C is reduced to solutions of easier prob-
lems. In the particular case, the proposition that
intake of saturated fat raises serum cholesterol is
strongly supported (A implies B) [2224]. The prop-
osition that cholesterol in the serum leads to athe-
rosclerotic disease (B implies C) is also strongly
supported by many kinds of studies, for example,
by presence of cholesterol in atherosclerotic plaque
[25,26], observational epidemiology [27,28], and
the success of statin drugs (which dramatically
reduce LDL cholesterol, among other things) in
reducing coronary heart disease and total mortality
www.co-lipidology.com 271
Hyperlipidaemia and cardiovascular disease
[29]. That coronary heart disease reduction also
follows serum cholesterol reduction even when total
mortality is not reduced [30] strengthens the prop-
osition that cholesterol in the serum plays a role
in atherosclerotic disease, because it suggests that
cholesterol reduction is specific to coronary heart
disease, whatever else might eventuate from serum
cholesterol lowering. Yet, this efficient logic is not
necessarily correct. Saturated fat is not the only
component of saturated fat-containing food; other
food components might be important. Many bio-
logical changes occur with cholesterol changes,
particularly considering the pleiotropic nature of
statins, and these also could be important. Direct
observation of the association between saturated fat
and coronary heart disease did not find the associ-
ation of saturated fat intake with coronary heart
which as expected by the syllogism [31]. Such syllo-
gisms are standard in science, for linking and inter-
preting the chains of evidence, but they should be
used cautiously because they are not always right.
In contrast, direct observation of certain dietary
patterns with various cardiovascular and other
chronic diseases has consistently shown reduced
risk [3235,36 ], although meta-analysis of specific
&&
foods and nutrients was found to be confusing [37].
This difference in findings possibly relates to the
stability of diet patterns over time and reliability of
reporting. Jacobs and Orlich [19] pointed out that in
most cases, diet patterns derived from food fre-
quency questionnaires had high correlation within
person over periods of years; two studies reported
tracking correlation over 0.5 over 18 and 20 years.
Correlations over time of most foods and nutrients
are much lower, on the order of 0.2 or less, although
a few well-defined foods that are habitually con-
sumed by some people, such as coffee, tea, or alco-
hol, have high-tracking correlations. In addition,
diet patterns integrate over foods eaten and over
long periods, whereas individual foods and
nutrients do not capture the sum of complex nutri-
tion effects.
MEDITERRANEAN AND PRUDENT DIET
PATTERNS THAT PROTECT AGAINST
CHRONIC DISEASE
There are many ways of describing diet patterns that
attempt to capture key characteristics, and to a
certain extent, culture, such as Mediterranean (char-
acterizing foods from that region), prudent (imply-
ing wise, judicious food choices), and Western
(reflecting certain aspects of food industrialization
in Western culture). The central focus of diet pattern
research has been Mediterranean or prudent diets
with various definitions [32]. In an early article
272 www.co-lipidology.com
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
studying middle-aged male health professionals,
Hu et al. [33] performed principal components
analysis and found two patterns: prudent, with
high weightings on green leafy vegetables, dark-
yellow vegetables, cruciferous vegetables, other
vegetables, legumes, fruit, tomatoes, fish, garlic,
poultry, and whole grains; and Western, with high
weightings on red meat, processed meat, refined
grains, sweets and desserts, french fries, high-fat
dairy products, eggs, high-sugar drinks, snacks,
condiments, margarine, potatoes, and butter. The
prudent diet was inversely related to incident cor-
onary heart disease, whereas the Western diet was
adversely related. This list of foods closely resembles
the Dietary Approaches to Stop Hypertension diet,
in which vegetables, fruit, fruit juice, whole grains,
nuts, seeds, legumes, and dairy products other than
whole milk were increased, and meat, poultry, fish,
refined grain, whole milk, sweets, snack foods, fats,
oils, and salad dressings were decreased [38].
The Prevencion con Dieta Mediterranea (PRE-
DIMED) study [9 ] instituted a 14-point Mediterra-
&&
nean diet compliance score. Foods favored are olive
oil, vegetables, fruit (including natural juices), wine,
legumes, fish or shellfish, nuts (including peanuts),
chicken, turkey, or rabbit meat instead of veal, pork,
hamburger, or sausage, and sofrito (a sauce made
with tomato and onion, leek, or garlic, simmered
with olive oil). Foods not favored are red meat,
hamburger, or meat products (ham, sausage, and
so on), butter, margarine, or cream, sweet/carbo-
nated beverages, and commercial sweets or pastries
(not homemade), such as cakes, cookies, biscuits, or
custard. Participants randomized to Mediterranean
diet groups had about 2 out of 14 points higher score
than the randomized control group, starting soon
after baseline and throughout the study (Table S1
and Figure S3 in [9 ]). Higher baseline scores pre-
&&
dicted cardiovascular disease outcomes [39].
Mursu et al. [36 ] studied various mortality out-
&&
comes in the Iowa Womens Health Study (IWHS)
according to two diet scores, the A Priori Diet Qual-
ity Score (APDQS) [4045] and the Alternative
Healthy Eating Index 2010 (AHEI) [46].
The APDQS originated in a case"control study
to differentiate myocardial infarction from healthy
controls according to combinations of foods [40]. It
was applied in several studies subsequently, show-
ing relationships with demographics [41], several
biomarkers [4245], cognitive function [47], and
fitness [48]. It has been reformulated for different
studies, depending on the questionnaire data avail-
able. Approximately 40 food groups, covering many
aspects of diet, are rated as favorable, adverse or
neutral and placed in quantiles according to intake.
For example, in the IWHS [36 ], 34 food groups
&&
Volume 26 ! Number 4 ! August 2015

were rated favorable (n 17), adverse (n 10), or
neutral (n 7) relative to future chronic disease,
based on the literature and expert judgment, and
then placed in quartiles (or a large nonconsumer
group and tertiles among consumers). Mursu et al.
[36 ] stated: The positively rated food groups were
&&
beans and legumes, beer, coffee, fish, fruit, green
vegetables, low-fat dairy, liquor, oil, other veg-
etables, poultry, seeds, and nuts, soy products, tea,
tomato, whole grains, and wine. The negatively
rated food groups were butter, fried foods, fried
potatoes, red meat, liver, processed meat, salty
snacks, soft drinks, sweets, and whole-fat dairy.
The remaining neutrally rated food groups were
chocolate, diet soft drinks, eggs, fruit juice, margar-
ine, potatoes, and refined grains.
Groups were given scores of 0 for lowest quartile
of consumption or nonconsumers to 3 for highest
consumption grouping. The APDQS was formed by
adding these values in a positive direction for bene-
ficially rated food groups, in a negative direction for
adversely rated food groups, and 0 for neutrally
rated food groups, for a theoretical maximal score
of 81. Thus, 1 point in the APQDS corresponds to a
difference of one higher intake category in benefi-
cially rated food groups or one lower intake category
in adversely rated food groups. Mean $ standard
deviation of the APDQS was 38.4 $ 8.2 in the
29 634 women aged 5569 years and free of self-
reported heart disease, diabetes, and cancer at base-
line in 1986 and followed for 22 years.
The AHEI-2010 [46] was formulated as an
improvement to an earlier similar score [49] that
provided dietary guidance related to the US Dietary
Guidelines. AHEI-2010 contains 11 components,
each of which receives 010 points (theoretical
maximum 110). Components included vegetables,
whole fruit, whole grains, nuts and legumes, red and
processed meat, long-chain (n-3) fatty acids eicosa-
pentaenoic acid docosahexaeneoic acid, polyun-
saturated fatty acids, and moderate alcohol intake
(all rated favorably) and trans fatty acids, sugar-
sweetened beverages and fruit juices, and sodium
(all rated adversely). In the IWHS, the mean AHEI-
2010 was 40.0 $ 10.4, with correlation of 0.65
between the AHEI-2010 and the APDQS.
Both the APDQS and AHEI-2010 predicted total,
cardiovascular, cancer, and noncardiovascular, non-
cancer inflammatory-related mortality, but not
other external causes of mortality. Prediction was
numerically strongest for inflammatory-related
mortality and weakest for cancer mortality, but
statistically significant for all. The risk factor
adjusted total mortality rates were 36.9, 34.6,
32.5, and 30.2 per 100 women followed for 22 years
across the increasing APDQS quartiles, and similarly
0957-9672 Copyright ! 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Anticardiovascular diet in 2015 Jacobs and Tapsell
for the AHEI-2010. We regard a difference of 36.9
versus 30.2 as of considerable clinical importance in
women initially average age 62 years, followed into
old age. Furthermore, the women were resurveyed
in 2004 (n 15 076) at average age 79 years, and
findings were similar or numerically stronger for
diet score prediction of all outcomes.
One of the goals of Mursu et al. [36 ] was to
&&
identify information in each score beyond what was
found in the other score. Each score did add some
prediction when the other was in the model, but it
was difficult to sort out what that information was.
More to the point for this review is the commonality
between these scores and of these scores with other
prudent and Mediterranean diet patterns.
CONCLUSION
The many diet pattern analyses cited here, although
consistent across many settings and having substan-
tial commonality across patterns studied, are obser-
vational and therefore considered insufficient in
themselves to make causal statements about diet
protecting against future chronic disease. However,
coupled with the randomized findings of the Lyon
Diet-Heart Study [8] and PREDIMED [9 ], we assert
&&
that certain diet patterns can prevent or delay a
substantial amount of currently existing chronic
disease, particularly cardiovascular and noncardio-
vascular, noncancer inflammatory-related diseases.
Such diet patterns are plant-centered, at the same
time minimizing intake of nutritionally poor plant
foods such as sugar, refined grains, and highly proc-
essed and very salty foods. Intake of plant seeds,
including whole grains, nuts, berries, coffee, and
chocolate, is promising [50 ]. The recommended
&&
prudent/Mediterranean diet patterns are highly var-
ied, as can be gleaned from the small increments to
the APDQS from 34 widely varying food groups.
Apparently, there are many ways to achieve a pro-
tective diet pattern.
The prudent/Mediterranean diet pattern
solution appears to be robust, based on its consist-
ency across many cohort studies, the Lyon Diet-
Heart Study and PREDIMED. However, some caveats
are in order. It seems to us that vegetarian diet
patterns work because of the presence of nutrition-
ally rich plant foods, not because of the absence of
meat. Thus, prudent/Mediterranean diet patterns
are not necessarily vegetarian, although much of the
meat in the diet is rated poorly in all patterns.
However, meat is less sustainable than other foods,
given the energy cost of growing meat products of
all types, and sustainability is a major issue for the
future [51]. Another caveat concerns fish, which
is highly recommended across all diet patterns. A
www.co-lipidology.com 273
Hyperlipidaemia and cardiovascular disease
problem is that the fish stock is substantially
polluted with heavy metals and other industrial
chemicals; this is an ongoing problem [52,53].
We may not ever know the more specific details
that have been so earnestly sought in many, many
nutrition studies. Indeed, the diet pattern approach
to nutrition recommendations taken here leaves
many questions unanswered. For example, for a
beverage with known toxicity at high intake levels,
alcohol intake may be overendorsed in the APDQS,
as points are given for each of wine, beer, and liquor
separately and with no upper limit for excessive
intake, as opposed to the moderate amount of alco-
hol recommended in the AHEI-2010 or the wine
with meals recommended in PREDIMED. There is
evidence that would support rating chocolate as
favorable rather than neutral [54], but food appli-
cations are highly variable. Furthermore, it is not
easy to know about separate effects of food groups
(as in APDQS) or dietary principles (as in AHEI-
2010). For example, prediction of cardiovascular
disease events in PREDIMED from the 14 individual
score components is not always consistent with
findings for prediction from the whole score, but
individual components do not take the whole diet
into account and are subject to confounding with
the rest of the diet [39]. The treatment groups
themselves in PREDIMED [9 ] and the Lyon Diet
&&
Heart Study [8] are not specific, but rather are highly
multifaceted. In PREDIMED, it is not possible to
fully sort out whether the benefit was from substi-
tuting extra virgin olive oil for the customarily
consumed and less expensive refined olive oil, from
the nut mixture, or from the overall Mediterranean
diet. Whether findings would have been different if
a different control regimen had been used is not
discernible from the PREDIMED trial.
The public and public health researchers do not
necessarily need to know details in order to identify
diet patterns that improve public health. The small
contribution of any one food to the total score
enhances robustness of the scores and supports
enjoyment of a varied diet as a major aspect of
eating. Thus, Pollans 7 word aphorism [2] and
our doubling of it to 14 words [19], eat food, mostly
plants, not too much, in colorful variety, maximiz-
ing nutrients per bite is a simple approach to
healthy and enjoyable eating.
Acknowledgements
D.R.J. wrote the first draft of this article. L.C.T.
performed the critical review and editing.
Financial support and sponsorship
Supported in part by a grant from the National Institutes
of Health, R01 HL 53560.
274 www.co-lipidology.com
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Conflicts of interest
D.R.J. is supported by diverse grants from the National
Institutes of Health and the Environmental Protection
Agency in the USA and is a member of the Scientific
Advisory Board of the California Walnut Commission.
L.C.T. is supported by grants from the Australian
Research Council and the Illawarra Health and Medical
Research Institute and is an unpaid member of the
Scientific Advisory Board of the California Walnut
Commission, and a member of the Science Advisory
Committee of the McCormick Science Institute. There
was no specific funding received for this paper.
REFERENCES AND RECOMMENDED
READING
Papers of particular interest, published within the annual period of review, have
been highlighted as:
& of special interest
&& of outstanding interest
1. US News and World Report. Best diets. 2015; http://health.usnews.com/
best-diet. [Accessed 26 January 2015]
2. US Department of Agriculture National Agricultural Library. Historical dietary
guidelines. 2015; http://fnic.nal.usda.gov/dietary-guidance/dietary-guide-
lines/historical-dietary-guidance. [Accessed 26 January 2015]
3. US Department of Health and Human Services. History of Dietary Guidelines
for Americans. 2015; http://www.health.gov/dietaryguidelines/history.htm.
[Accessed 26 January 2015]
4. Taubes G. Epidemiology faces its limits. Science 1995; 269:164169.
5. Ioannidis JP. Implausible results in human nutrition research. BMJ 2013;
347:f6698.
6. Temple NJ. Nutrition and disease: challenges of research design. Nutrition
2002; 18:343347.
7. Archer E, Hand GA, Blair SN. Validity of U.S. Nutritional Surveillance: National
Health and Nutrition Examination Survey Caloric Energy Intake Data, 1971
2010. PLoS One 2013; 8:e76632.
8. De Lorgeril M, Salen P, Martin JL, et al. Mediterranean diet, traditional risk
factors, and the rate of cardiovascular complications after myocardial infarc-
tion: final report of the Lyon Diet Heart Study. Circulation 1999; 99:779785.
9. Estruch R, Ros E, Salas-Salvado J, et al. Primary prevention of cardiovascular
&& disease with a Mediterranean diet. N Engl J Med 2013; 368:12791290.
This is a landmark report of a long-term, randomized clinical trial of dietary patterns,
one of few such studies of dietary patterns, and probably as well done and as close
to unambiguous as such a study can be.
10. Scrinis G. Nutritionism: the science and politics of dietary advice. New York,
& NY: Columbia University Press; 2013.
This book presents a useful integration of ideas about how to talk about nutrition
and the specific and unanticipated political and scientific effects of reducing most
discussion of nutrition to nutrients.
11. Pollan M. In defense of food: an eaters manifesto. New York, NY: The Penguin
Press; 2008.
12. Bjelakovic G, Nikolova D, Gluud LL, et al. Mortality in randomized trials of
antioxidant supplements for primary and secondary prevention: systematic
review and meta-analysis. JAMA 2007; 297:842857.
13. HPS2-THRIVE Collaborative Group. Landray MJ, Haynes R, et al. Effects of
extended-release niacin with laropiprant in high-risk patients. N Engl J Med
2014; 371:203212.
14. Age-Related Eye Disease Study Research Group. A randomized, placebo-
controlled, clinical trial of highdose supplementationwith vitamins C and E,
beta carotene, and zinc for age-related macular degeneration and vision loss:
AREDS report no. 8. Arch Ophthalmol 2001; 119:14171436.
15. Kim YI. Folic acid fortification and supplementation: good for some but not so
good for others. Nutr Rev 2007; 65:504511.
16. Weng LC, Steffen LM, Szklo M, et al. A diet pattern with more dairy and nuts,
but less meat is related to lower risk of developing hypertension in middle-
aged adults: the Atherosclerosis Risk in Communities (ARIC) study. Nutrients
2013; 5:17191733.
17. Jacobs DR Jr, Tapsell LC, Temple NJ. Food synergy: the key to balancing the
nutrition research effort. Public Health Rev 2012; 33:507529.
18. Satija A, Yu E, Willett WC, Hu FB. Understanding nutritional epidemiology
&& and its role in policy. Adv Nutr 2015; 6:114.
This article, written independently from Jacobs, Tapsell, and Temple, gives an
alternate but confirmatory view of issues in inferences in nutrition, with comparison
to drug research.
Volume 26 ! Number 4 ! August 2015

19. Jacobs DR Jr, Orlich MJ. Diet pattern and longevity: do simple rules suffice? A
commentary. Am J Clin Nutr 2014; 100 (Suppl 1):313S319S.
20. Fardet A, Rock E. Toward a new philosophy of preventive nutrition: from a
reductionist to a holistic paradigm to improve nutritional recommendations.
Adv Nutr 2014; 5:430446.
21. Mozaffarian D, Ludwig DS. Dietary guidelines in the 21st century: a time for
food. JAMA 2010; 304:681682.
22. Keys A, Anderson JT, Grande F. Serum cholesterol response to changes in
the diet, IV: particular saturated fatty acids in the diet. Metabolism 1965;
14:776784.
23. Mensink RP, Katan MB. Effect of dietary trans fatty acids on high-density and
low-density lipoprotein cholesterol levels in healthy subjects. N Engl J Med
1990; 323:439445.
24. Hegsted DM. A look back at lessons learned and not learned. J Nutr 1994;
124 (Suppl):1867S1870S.
25. Sabine J. Cholesterol. New York: Marcel Dekker, Inc; 1977.
26. Wissler RW. New insights into the pathogenesis of atherosclerosis as
revealed by PDAY. Pathobiological Determinants of Atherosclerosis in Youth.
Atherosclerosis 1994; 108 (Suppl):S3S20.
27. Jacobs D, Blackburn H, Higgins M, et al. Report of the conference on low
blood cholesterol: Mortality Associations. Circulation 1992; 86:10461060.
28. Prospective Studies Collaboration. Lewington S, Whitlock G, et al. Blood
cholesterol and vascular mortality by age, sex, and blood pressure: a meta-
analysis of individual data from 61 prospective studies with 55 000 vascular
deaths. Lancet 2007; 370:18291839.
29. Ebrahim S, Taylor FC, Brindle P. Statins for the primary prevention of
cardiovascular disease. BMJ 2014; 348:g280.
30. Gould AL, Rossouw JE, Santanello NC, et al. Cholesterol reduction yields
clinical benefit. A new look at old data. Circulation 1995; 91:22742282.
31. Hu FB, Stampfer MJ, Manson JE, et al. Dietary fat intake and the risk of
coronary heart disease in women. N Engl J Med 1997; 337:14911499.
32. Trichopoulou A, Martnez-Gonzalez MA, Tong TY, et al. Definitions and
potential health benefits of the Mediterranean diet: views from experts around
the world. BMC Med 2014; 12:112.
33. Hu FB, Rimm EB, Stampfer MJ, et al. Prospective study of major dietary
patterns and risk of coronary heart disease in men. Am J Clin Nutr 2000;
72:912921.
34. Stradling C, Hamid M, Taheri S, Thomas GN. A review of dietary influences on
cardiovascular health: part 2 " dietary patterns. Cardiovasc Hematol Disord
Drug Targets 2014; 14:5063.
35. Martinez-Gonzalez MA, Bes-Rastrollo M. Dietary patterns, Mediterranean diet,
and cardiovascular disease. Curr Opin Lipidol 2014; 25:2026.
36. Mursu J, Steffen LM, Meyer KA, et al. Diet quality indexes and mortality in
&& postmenopausal women: the Iowa Womens Health Study. Am J Clin Nutr
2013; 98:444453.
Among many articles reporting disease outcomes according to diet patterns in
prospective observational studies, this one stands out because it compares and
contrasts results for two highly predictive diet pattern formulations. Although the
authors focused on differences between the diet patterns, in the sense of the
current review the commonality of the two diet patterns is worthy of note.
37. Berciano S, Ordovas JM. Nutrition and cardiovascular health. Rev Esp Cardiol
(Engl Ed) 2014; 67:738747.
0957-9672 Copyright ! 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Anticardiovascular diet in 2015 Jacobs and Tapsell
38. Jacobs DR Jr, Gross MD, Steffen L, et al. The effects of dietary patterns on
urinary albumin excretion: results of the Dietary Approaches to Stop Hyper-
tension (DASH) Trial. Am J Kidney Dis 2009; 53:638646.
39. Schroder H, Salas-Salvado J, Martnez-Gonzalez MA, et al. Baseline adher-
ence to the Mediterranean diet and major cardiovascular events: Prevencion
con Dieta Mediterranea trial. JAMA Intern Med 2014; 174:16901692.
40. Lockheart MS, Steffen LM, Rebnord HM, et al. Dietary patterns, food groups
and myocardial infarction: a case-control study. Br J Nutr 2007; 98:380387.
41. Sijtsma FP, Meyer KA, Steffen LM, et al. Longitudinal trends in diet and effects
of sex, race, and education on dietary quality score change: the Coronary
Artery Risk Development in Young Adults study. Am J Clin Nutr 2012;
95:580586.
42. Jacobs DR Jr, Sluik D, Rokling-Andersen MH, et al. Association of 1-y changes
in diet pattern with cardiovascular disease risk factors and adipokines: results
from the 1-y randomized Oslo Diet and Exercise Study. Am J Clin Nutr 2009;
89:509517.
43. Meyer KA, Sijtsma FP, Nettleton JA, et al. Dietary patterns are associated with
plasma F2-isoprostanes in an observational cohort study of adults. Free Radic
Biol Med 2013; 57:201209.
44. Sijtsma FP, Meyer KA, Steffen LM, et al. Diet quality and markers of endothelial
function: the CARDIA study. Nutr Metab Cardiovasc Dis 2014; 24:632638.
45. Nettleton JA, Schulze MB, Jiang R, et al. A priori-defined dietary patterns and
markers of cardiovascular disease risk in the Multi-Ethnic Study of Athero-
sclerosis (MESA). Am J Clin Nutr 2008; 88:185194.
46. Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly
predict risk of chronic disease. J Nutr 2012; 142:10091018.
47. Zhu N, Jacobs DR, Meyer KA, et al. Cognitive function in a middle aged cohort
is related to higher quality dietary pattern 5 and 25 years earlier: The CARDIA
Study. J Nutr Health Aging 2015; 19:3338.
48. Shikany JM, Jacobs DR Jr, Lewis CE, et al. Associations between food
groups, dietary patterns, and cardiorespiratory fitness in the Coronary Artery
Risk Development in Young Adults study. Am J Clin Nutr 2013; 98:1402
1409.
49. McCullough ML, Feskanich D, Stampfer MJ, et al. Diet quality and major
chronic disease risk in men and women: moving toward improved dietary
guidance. Am J Clin Nutr 2002; 76:12611271.
50. Ros E, Hu FB. Consumption of plant seeds and cardiovascular health:
&& epidemiological and clinical trial evidence. Circulation 2013; 128:553565.
This review and thought piece is a very clever amalgamation of a series of thought
threads that combines and expands the important work on whole cereal grains and
nuts.
51. Sabate J, Soret S. Sustainability of plant-based diets: back to the future. Am J
Clin Nutr 2014; 100 (Suppl 1):476S482S.
52. Nstbakken OJ, Hove HT, Duinker A, et al. Contaminant levels in Norwegian
farmed Atlantic salmon (Salmo salar) in the 13-year period from 1999 to
2011. Environ Int 2015; 74:274280.
53. Ruzzin J, Bethune C, Goksyr A, et al. Comment on Contaminant levels in
Norwegian farmed Atlantic salmon (Salmo salar) in the 13-year period from
1999 to 2011 by Nstbakken et al. Environ Int 2015. [Epub ahead of print].
doi: 10.1016/j.envint.2015.01.003.
54. Latham LS, Hensen ZK, Minor DS. Chocolate: guilty pleasure or healthy
supplement? J Clin Hypertens (Greenwich) 2014; 16:101106.
www.co-lipidology.com 275
Diet and the environment: does what you eat matter?14
Harold J Marlow, William K Hayes, Samuel Soret, Ronald L Carter, Ernest R Schwab, and Joan Sabate
ABSTRACT
Food demand influences agricultural production. Modern agricul-
tural practices have resulted in polluted soil, air, and water; eroded
soil; dependence on imported oil; and loss of biodiversity. The goal
of this research was to compare the environmental effect of a vege-
tarian and nonvegetarian diet in California in terms of agricultural
production inputs, including pesticides and fertilizers, water, and en-
ergy used to produce commodities. The working assumption was that
a greater number and amount of inputs were associated with a greater
environmental effect. The literature supported this notion. To accom-
plish this goal, dietary preferences were quantified with the Adventist
Health Study, and California state agricultural data were collected
and applied to state commodity production statistics. These data were
used to calculate different dietary consumption patterns and indexes
to compare the environmental effect associated with dietary prefer-
ence. Results show that, for the combined differential production of
11 food items for which consumption differs among vegetarians and
nonvegetarians, the nonvegetarian diet required 2.9 times more wa-
ter, 2.5 times more primary energy, 13 times more fertilizer, and 1.4
times more pesticides than did the vegetarian diet. The greatest con-
tribution to the differences came from the consumption of beef in the
diet. We found that a nonvegetarian diet exacts a higher cost on the
environment relative to a vegetarian diet. From an environmental per-
spective, what a person chooses to eat makes a difference. Am J
Clin Nutr 2009;89(suppl):1699S703S.
INTRODUCTION
In developed countries, and throughout the world, there exists
a link between agricultural production and environmental deg-
radation (14). Public awareness of diverse global environ-
mental issues, such as climate change (57), toxic residues in
food (8), soil erosion (9, 10), and species endangerment (9, 11),
has brought about a call for sustainable food production prac-
tices (12) and responsible stewardship of our finite resources
(13). Particular skepticism has been aimed at supporting the
increased demand for animal products in the diet of the eco-
nomically advantaged persons of the world (4, 14).
To address concerns about the increased demand for animal
consumption, we can begin by asking a series of pertinent
questions. Do dietary choices really have an effect on the en-
vironment? More specifically, does animal consumption create
a heavier footprint than a vegetarian diet? If so, what are some of
the major environmental effects of an animal-based diet, and how
might these be measured?
In this article, we first identify and briefly review 6 major
effects that dietary choices may have on the environment and then
Am J Clin Nutr 2009;89(suppl):1699S703S. Printed in USA. ! 2009 American Society for Nutrition
describe a research program undertaken at Loma Linda Uni-
versity to quantify the environmental effects of vegetarian and
nonvegetarian diets.
MODERN AGRICULTURES EFFECT ON THE
ENVIRONMENT
The environmental effect of modern agriculture has increased
with the implementation of technologies designed to increase
crop yield and commodity production (15). Technologic ad-
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
vances in mechanization, irrigation, fertilization, and chemical
control of pests have facilitated substantial increases in agri-
cultural output since the 1940s (16). Simultaneously, there has
been an increase in total energy expenditure (17), depletion of
natural resources (12, 18), and generation of waste products (16)
associated with increased agricultural output. In fact, the point
has been reached where scientists and policymakers have begun
to seriously doubt the sustainability of these trends (19). In the
remainder of this section, we identify and briefly review 6 major
effects that dietary choices may have on the environment: water
resources, energy consumption, chemical fertilizer application,
pesticide application, waste generation, and land degradation.
Water resources
Most cropland in the United States is rain fed (20). Despite this
fact, agricultural production requires !80% of the water con-
sumed in the United States (15, 21) to irrigate 1015% of
cropland (15, 20, 22) and to water livestock (20). Critical water
issues exacerbated by agricultural practices include the pollution
of surface and groundwater sources (23), overdrafting of aqui-
fers (20), waterlogging and salinization of soils (12, 13), wet-
lands loss (24), and runoff, evaporation, and leakage from
1
From the Departments of Environmental Health (HJM and SS) and
Nutrition (JS), School of Public Health, Department of Earth and Biological
Sciences, School of Science and Technology (WKH and RLC), Department
of Allied Health Studies, School of Allied Health Professions (ERS), Loma
Linda University, Loma Linda, CA.
2
Presented at the symposium, Fifth International Congress on Vegetar-
ian Nutrition, held in Loma Linda, CA, March 46, 2008.
3
Supported by the Loma Linda University, School of Public Health, De-
partment of Nutrition, McLean Fund for Vegetarian Nutrition Research.
4
Reprints not available. Address correspondence to HJ Marlow, Depart-
ment of Environmental Health, School of Public Health, Loma Linda Uni-
versity, Loma Linda, CA 92350. E-mail: hmarlow@llu.edu.
First published online April 1, 2009; doi: 10.3945/ajcn.2009.26736Z.
1699S
1700S MARLOW ET AL
irrigation systems (25). These effects may have greater signifi-
cance during times of seasonal or extended drought (21).
Energy consumption
Increased use of fossil fuels and concurrent technologic
advances have allowed humans to increase the productivity of
natural systems by manipulating the environment (17, 23). The
energy intensiveness of agricultural production varies with type
of crop produced, amount of chemical inputs, and geographic
location (18). With market globalization and convenient trans-
portation choices, food has become available during seasons
when they were typically absent, and the increased energy re-
quirements are largely borne by consumers and driven by market
demand. In the United States, fossil fuel consumption doubled
during a 20-y span while the caloric return per calorie of input
on most crops diminished (15). Cheap sources of fossil fuels will
continue to allow for massive energy inputs to agricultural
systems, but, as prices increase and supplies dwindle, this
practice is likely to change (14). Conservation and optimization
of energy use will certainly be in the future of agriculture.
A positive return of 23 nutrient calories per calorie of primary
energy input is characteristic for most cereal grains and legumes
(26). Most fruit and vegetables return 0.5 calories, and animal
products return 0.010.05 calories (15). The energy inputs for
animal products may be 2.55.0 times greater than for plant
products (27).
Chemical fertilizer application
The natural fertility of soil in the United States has been de-
pleted and has been replaced by application of chemical fertilizers
at rates that, for a time, increased 10%/y since the 1950s (15).
Potassium and phosphate are produced from nonrenewable re-
sources, and the production of nitrogen fertilizer relies directly on
petroleum (28). The use of fertilizers represents the single
greatest energy input for many crops (15), and the overuse of
fertilizers has resulted in surface and groundwater contamination
(18, 19, 29), air pollution (30), and a decrease in biodiversity (31).
Pesticide application
Pesticide use has increased as much as 33-fold in the United
States since the 1940s (16, 32). At the turn of the century, 2.5
million tons of pesticides were applied annually to crops world-
wide (33). Despite this increase in the use of pesticides, an esti-
mated 37% of all crop production is lost annually to pests (3234).
Increased monoculture cultivation, positive cost-benefit ratios (8,
33), and neglect of the environmental or social cost of application
(35) have lead to unrestricted increases in pesticide usage.
Concerns over the environmental consequences of pesticide use
include: residues on food (34), ground and surface water con-
tamination (36), persistence in the environment and bioamplifi-
cation (35), damage to nontargeted species (36), increased chemical
resistance in pests (34, 35), and worker safety (37). Many of the
environmental effects are difficult to measure or assess accurately
(38).
In addition to direct and indirect environmental effects, a host
of acute and chronic human health effects have also been as-
sociated with pesticides. These include endocrine disruption,
immune dysfunction, neurological disorders, and cancer (39).
Waste generation
In addition to the previously identified pollution problems,
wastes generated by intensified animal production often result in
significant water, soil, and air pollution (40). In the United States,
7 billion livestock generate 130 times more waste than produced
by 300 million humans (41). These wastes, most of which go
untreated, contain high concentrations of nitrogen, phosphorous,
and potassium compounds and traces of metals and antibiotics;
these represent a serious public health problem according to the
World Health Organization and US Department of Agriculture
(4, 23, 40, 41). Concentrated livestock operations and livestock
waste also produce gases. Some, such as ammonia, have a more
local effect and are generally regarded as nuisance odors (40).
Others such as carbon dioxide, methane, and nitrous oxide exert
a global effect and have been implicated in climate change (4,
26, 40, 42).
Land degradation
Livestock production exacts a significant toll on natural
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
habitats. According to a recent report from the Food and Agri-
culture Organization of the United Nations, the livestock sector is
by far the single largest anthropogenic user of land, accounting
for 70% of all agricultural land and 30% of the land surface of the
planet (4). Livestock production, and its continuing expansion
and intensification, is a key driver of many destructive ecosys-
tem changes, including deforestation; replacement of herba-
ceous plants by woody plant cover; desertification; and soil
compaction, erosion, and subsequent sedimentation of water-
ways, wetlands, and coastal areas (4, 43, 44). Animal production
also facilitates the establishment and spread of invasive plants
and animals, as well as zoonotic diseases. The poultry industry,
for example, has been linked to the transmission of highly
pathogenic avian influenza (45).
MEASURING ENVIRONMENTAL EFFECTS OF DIETARY
PREFERENCE: VEGETARIAN COMPARED WITH
NONVEGETARIAN
Human health and the health of the environment are in-
extricably linked. The link is so clear for Fowler and Hobbs (46)
that they concluded that humanity is not sustainable. There
have been attempts to identify and quantify the ecologic con-
sequences associated with modern agricultural practices (15, 19,
23). At the turn of the century, Gussow (47) issued a call for
research permitting a direct comparison of the ecologic con-
sequences of different diets. Although several associations have
been suggested (3, 23), what mostly appear in the literature
are comparisons of discretely selected food items, not direct,
quantifiable comparisons of whole diets. One notable exception
is a recently published study involving the evaluation of ideal-
ized diets with the use of Life Cycle Assessment and computer
modeling (19).
Research location
At Loma Linda University we explored the relation between
dietary preference and environmental effects. An approach was
developed with the use of the state of California as a model to
quantify the environmental effect of agricultural practices used to
 VEGETARIAN DIET AND THE ENVIRONMENT 1701S
produce commodities for representative vegetarian and non- TABLE 1
vegetarian diets. California historically has been the largest Ratios of consumption of the 11 food items or groups that were significantly
producer of agricultural and food products in the United States, different between the diets of vegetarians and nonvegetarians1
hosting a wide range of operations (48). The goal of our research Ratio2
was to compare these 2 diets in terms of the water, energy, in-
Plant foods
organic fertilizers, and pesticides (ie, inputs) used to produce Dry fruit 1.1
the commodities for each. Our working assumption was that Canned fruit 1.8
a greater number and amount of inputs are associated with Winter fruit3 1.4
greater environmental effect. A complete description of the Seasonal fruit3 1.6
methods and results are available in Marlow (49). Citrus fruit 1.6
Fruit juice 1.1
Nuts 2.4
Beans 1.5
Quantifying vegetarian and nonvegetarian diets Animal foods4
There are many vegetarian diets, many unique to the individual Eggs 0.43
consumer. To make this project relevant, we selected the largest Poultry 0.04
Beef 0.03
vegetarian group in California, the Seventh-day Adventists
1
(Adventists), for whom ample data are available, to specify Food composition of vegetarian and nonvegetarian diets was calcu-
the composition of representative vegetarian and nonvegetarian lated from the Adventist Health Study (50).
2
diets. Among the 34,000 California Adventists participating in Expressed as ratio of vegetarian to nonvegetarian diet.
3
Winter fruit was referred to in the food-frequency questionnaire as
the Adventist Health Study I (AHS) cohort, 50% are vegeta-

Downloaded from ajcn.nutrition.org by guest on June 18, 2016

rians and 50% nonvegetarians by dietary preference. The AHS
was designed to investigate the relation between lifestyle, in
particular dietary choice, and health outcomes (50). Our in-
vestigation has extended the utility of the AHS into the field of
environmental health. This data set provided a means for
quantifying practically relevant consumption pattern differences
for specific food items or food groups in the 2 diets. Among 31
food items or food groups in the AHS questionnaire, 11 were
consumed at substantially different rates by vegetarians and
nonvegetarians, whereas the remainder of 20 food items or
groups was similarly consumed in both diet patterns. The food
items or food groups used in this research and their relative
contribution to the vegetarian or nonvegetarian diet are shown in
Table 1. The results show that vegetarians ate slightly more
plant foods and that nonvegetarians ate substantially more animal
foods in their diets.
Environmental effect analyses
Commodity production quantities and input statistics were
gathered from a variety of federal, state, and county agencies in
addition to industry associations. These organizations provided
information mostly in the form of published reports, databases,
technical assistance, and professional advice. Production and
input statistics were used to calculate overall variables generally
referred to as use efficiencies (49). There were corresponding Comparisons of environmentally relevant inputs for the combined
use efficiencies calculated for water consumption (water use
efficiency), energy used (energy use efficiency), pesticides ap-
plied (pesticide use efficiency), and fertilizers applied (fertilizer
use efficiency).
The outcome of our studies provided evidence for the much
higher ecologic cost of an animal-based diet. The approximated
effect ratios for water use efficiency, energy use efficiency,
pesticide use efficiency, and fertilizer use efficiency are presented
in Table 2. Our analyses further showed that these differences
resulted primarily from the inclusion of beef in the diet of the
nonvegetarian. This finding is similar to those published by
groups in Europe (4, 19), Japan (51), the United States (27, 52),
and Australia (6, 53).
fruit, such as an apple, for which availability was not seasonally limited.
Seasonal fruit, such as watermelon, was fruit that was limited to seasonal
availability.
4
Corresponding figures for the inverse ratio (nonvegetarian/vegetarian)
are 2.3 for eggs, 25 for poultry, and 32 for beef.
DISCUSSION AND CONCLUSIONS
It is important to remember that these efficiency ratios are
based on the differences between the diets that we chose to
analyze, each of which had a limited number of food items. If, for
example, the inputs from the remainder of the diet were added,
the ratios would be reduced, but the absolute differences would
remain unchanged.
For purposes of comparison, the absolute data are highly il-
lustrative. When comparing water, for instance, the difference in
water use for the vegetarian and nonvegetarian diet was 1000 L
(264 gallons)/wk. These results are consistent with those reported
by others (18, 22, 26, 52, 5456). According to the American
Water Works Association (57), the average weekly per capita
indoor water consumption for a home with no water-conserving
appliances is 1835 L (485 gallons), although this may be a
conservative estimate (58). With the use of this figure, the
TABLE 2
production of the 11 food items or groups in which California Adventist
vegetarian and nonvegetarian diets differ1
Input2 Ratio3
Water (L) 2.9
Primary energy (kJ) 2.5
Fertilizer (g) 13
Pesticides (g) 1.4
1
Food composition of vegetarian and nonvegetarian diets was calcu-
lated from the Adventist Health Study (50); inputs were estimated from
a variety of federal, state, and county agencies in addition to industry asso-
ciations (49).
2
Expressed as cumulative requirements.
3
Expressed as ratio of nonvegetarian to vegetarian input quantities.
1702S MARLOW ET AL
Adventist vegetarian diet conserves the equivalent of 54% of the
average weekly per capita indoor water consumption. This can
be compared with a savings of 35%, estimated by the American
Water Works Association, by installing more-efficient water
fixtures and regularly checking for leaks. From this comparison
it is apparent that a plant-based diet provides a significant water
conservation benefit. A similar ecologic cost effectiveness can
be determined for each of the other inputs in the study.
Considering the surmounting ecologic pressures that a bur-
geoning human civilization exerts on our planet, there is a need to
make hard decisions. Among these hard decisions, many societies,
and governments in particular, will have to reconsider the increasing
demand for an animal-based diet. Many governments, including
both the European Union and the US government, may need to
reassess agricultural subsidies (59, 60) and divert some of the
funding to support additional research, development, and applica-
tion of sustainable methods of food production. Outreach programs
may be necessary to educate and inform people about the health
and environmental benefits of a vegetarian diet. (Other articles in
this supplement to the Journal include references 6187.)
The authors responsibilities were as followsHJM: was the principal au-
thor of the manuscript and was primarily responsible for the design of the
experiment, collection and analysis of data, and writing of the manuscript;
WKH, SS, and JS: contributed to the writing and editing of the manuscript;
And RLC and ERS: provided significant advice and consultation during the
data analysis and manuscript preparation. None of the authors declared a con-
flict of interest.
REFERENCES
1. Gussow JD, Clancy KL. Dietary guidelines for sustainability. J Nutr
Educ 1986;18:15.
2. Goodland R. Environmental sustainability in agriculture: diet matters.
Ecol Econ 1997;23:189200.
3. Carlsson-Kanyama A. Climate change and dietary choices: how can
emissions of greenhouse gases from food consumption be reduced?
Food Policy 1998;34:27798.
4. Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, de Haan C.
Livestocks long shadow: environmental issues and options. Rome,
Italy: Food and Agriculture Organization of the United Nations, 2006.
5. Johnson JM, Franzluebbers AJ, Weyers SL, Reicosky DC. Agricultural
opportunities to mitigate greenhouse gas emissions. Environ Pollut
2007;150:10724.
6. McMichael AJ, Powles JW, Butler CD, Vauy R. Food, livestock pro-
duction, energy, climate change, and health. Lancet 2007;370:125363.
7. Noble BF, Christmas LM. Strategic environmental assessment of
greenhouse gas mitigation options in the Canadian agricultural sector.
Environ Manage 2008;41:6478.
8. Falconer KE. Managing diffuse environmental contamination from ag-
ricultural pesticidesan economic perspective on issues and policy
options, with particular reference to Europe. Agric Ecosyst Environ
1998;69:3754.
9. Huston M. Biological diversity, soils, and economics. Science 1993;262:
167680.
10. Trimble SW, Crosson P. Land useUS soil erosion ratesmyth and
reality. Science 2000;289:24850.
11. Tisdell C. Agricultural sustainability and conservation of biodiversity:
competing policies and paradigms. In: Dragun AK, Jakobsson KM, eds.
Sustainability and global environmental policy. Cheltenham, United
Kingdom: Edward Elgar Publishing Ltd, 1998:97129.
12. Pinstrup-Andersen P, Pandya-Lorch R. Food security and sustainable use
of natural resources: a 2020 vision. Ecol Econ 1998;26:110.
13. Worrell R, Appleby MC. Stewardship of natural resources: definition,
ethical and practical aspects. J Agric Environ Ethics 2000;12:26377.
14. Heitschmidt RK, Short RE, Grings EE. Ecosystems, sustainability, and
animal agriculture. J Anim Sci 1996;74:1395405.
15. Pimentel D, Pimentel M. The future of American agriculture. In: Knorr D,
ed. Sustainable food systems. Westport, CT: Avi Publishers, 1983:327.
16. Brown LR. Human food production as a process in the biosphere. Sci
Am 1970;223:16170.
17. Cleveland CJ. The direct and indirect use of fossil fuels and electricity in
USA agriculture, 19101990. Agric Ecosyst Environ 1995;55:11121.
18. Pimentel D, Pimentel M. US food production threatened by rapid
population growth. Washington, DC: Prepared for Carrying Capacity
Network, 1997.
19. Baroni L, Cenci L, Tettamanti M, Berati M. Evaluating the environ-
mental impact of various dietary patterns combined with different food
production systems. Eur J Clin Nutr 2007;61:27986.
20. Tanji KK, Enos CA. Global water resources and agricultural use. In:
Tanji KK, Yaron B, eds. Management of water use in agriculture. Berlin,
Germany: Springer-Verlag, 1994:324.
21. Zilberman D, Dunarm A, MacDougal N, Khanna M, Brown C, Castillo
F. Individual and institutional responses to the drought: the case of
California agriculture. J Contemp Water Res Educ 2002;121:1723.
22. Pimentel D, Berger B, Filiberto D, et al. Water resources: agricultural
and environmental issues. Bioscience 2004;54:90918.
23. Pimentel D. Environmental and economic benefits of sustainable agri-
culture. In: Kihn J, Gowdy J, Hinterberger F, van der Straaten J, eds.
Sustainability in question. Cheltenham, United Kingdom: Edward Elgar
Publishing Ltd, 1999:15370.
24. Lemly AD, Kingsford RT, Thompson JR. Irrigated agriculture and
wildlife conservation: conflict on a global scale. Environ Manage 2000;
25:485512.
Downloaded from ajcn.nutrition.org by guest on June 18, 2016
25. Wallace JS. Increasing agricultural water use efficiency to meet future
food production. Agric Ecosyst Environ 2000;82:10519.
26. Horrigan L, Lawrence RS, Walker P. How sustainable agriculture can
address the environmental and human health harms of industrial agri-
culture. Environ Health Perspect 2002;110:44556.
27. Reijnders L, Soret S. Quantification of the environmental impact of dif-
ferent dietary protein choices. Am J Clin Nutr 2003;78(suppl):664S8S.
28. Viglizzo EF, Pordomingo AJ, Castro MG, Lertora FA. Environmental
assessment of agriculture at a regional scale in the Pampas of Argentina.
Environ Monit Assess 2003;87:16995.
29. Cavero J, Beltran A, Aragues R. Nitrate exported in drainage waters of
two sprinkler-irrigated watersheds. J Environ Qual 2003;32:91626.
30. Cowling E, Galloway J, Furiness C, et al. Optimizing nitrogen man-
agement in food and energy production and environmental protection:
summary statement from the Second International Nitrogen Conference.
ScientificWorldJournal 2001;1:19.
31. Mineau P, McLaughlin A. Conservation of biodiversity within Canadian
agricultural landscapes: integrating habitat for wildlife. J Agric Environ
Ethics 1996;9:93113.
32. Pimentel D, Acquay H, Biltonen M, et al. Environmental and economic
costs of pesticide use. An assessment based on currently available U.S.
data, although incomplete, tallies $8 billion in annual costs. Bioscience
1992;42:75060.
33. Paoletti MG, Pimentel D. Environmental risks of pesticides versus ge-
netic engineering for agricultural pest control. J Agric Environ Ethics
2000;12:279303.
34. Pimentel D, McLaughlin L, Zepp A, et al. Environmental and economic
effects of reducing pesticide use in agriculture. Agric Ecosyst Environ
1993;46:27388.
35. Foster V, Mourato S, Tinch R, Ozdemiroglu E, Pearce D. Incorporation
external impacts in pest management choices. In: Vorley W, Keeney D,
eds. Bugs in the system. London, United Kingdom: Earthscan Pub-
lications Limited, 1998:94106.
36. Levitan L, Merwin I, Kovach J. Assessing the relative environmental
impacts of agricultural pesticides: the quest for a holistic method. Agric
Ecosyst Environ 1995;55:15368.
37. Kishi M, Ladou J. International pesticide use. Introduction. Int J Occup
Environ Health 2001;7:25965.
38. Burn A. Pesticide buffer zones for the protection of wildlife. Pest Manag
Sci 2003;59:58390.
39. Pretty J. Agricultural sustainability: concepts, principles and evidence.
Philos Trans R Soc Lond B Biol Sci 2008;363:44765.
40. Delgado C, Rosegrant M, Steinfeld H, Ehui S, Courbois C. Livestock to
2020: the next food revolution. Washington, DC: International Food
Policy Research Institute, 2001:279.
41. US General Accounting Office, Committee on Agriculture Nutrition and
Forestry. Animal agriculture: waste management practices: report to the
Honorable Tom Harkin, Ranking Minority Member, Committee on
Agriculture, Nutrition, and Forestry, U.S. Senate United States General
 AJCN. First published ahead of print June 4, 2014 as doi: 10.3945/ajcn.113.071522.
Sustainability of plant-based diets: back to the future13
Joan Sabate and Sam Soret
ABSTRACT
Plant-based diets in comparison to diets rich in animal products are
more sustainable because they use many fewer natural resources
and are less taxing on the environment. Given the global population
explosion and increase in wealth, there is an increased demand for
foods of animal origin. Environmental data are rapidly accumulating
on the unsustainability of current worldwide food consumption prac-
tices that are high in meat and dairy products. Natural nonrenewable
resources are becoming scarce, and environmental degradation is
rapidly increasing. At the current trends of food consumption and en-
vironmental changes, food security and food sustainability are on a
collision course. Changing course (to avoid the collision) will require
extreme downward shifts in meat and dairy consumption by large seg-
ments of the worlds population. Other approaches such as food waste
reduction and precision agriculture and/or other technological ad-
vances have to be simultaneously pursued; however, they are insuf-
ficient to make the global food system sustainable. For millennia,
meatless diets have been advocated on the basis of values, and large
segments of the world population have thrived on plant-based diets.
Going back to plant-based diets worldwide seems to be a reasonable
alternative for a sustainable future. Policies in favor of the global
adoption of plant-based diets will simultaneously optimize the food supply,
health, environmental, and social justice outcomes for the worlds
population. Implementing such nutrition policy is perhaps one of the
most rational and moral paths for a sustainable future of the human race
and other living creatures of the biosphere that we share. Am J
Clin Nutr doi: 10.3945/ajcn.113.071522.
WHAT ARE SUSTAINABLE DIETS?
Definitions of sustainability generally address aspects of ecology,
economy, and society and have different meanings depending on
the context. A sustainable diet will not necessarily be defined the
same way for consumers as for farmers or food manufacturers. In
2010 the FAO defined sustainable diets as those diets with low
environmental impacts which contribute to food and nutrition
security and to healthy life for present and future generations.
Sustainable diets are protective and respectful of biodiversity
and ecosystems, culturally acceptable, accessible, economically
fair and affordable; nutritionally adequate, safe and healthy; while
optimizing natural and human resources (1). Furthermore, ac-
cording to Fanzo et al (2), the determinants of a sustainable diet
are as follows: nutritional adequacy, environmental sustainabil-
ity, cultural acceptability, and low-cost accessibility.
In the arena of environmental sustainability, the focus of this
article, 2 important dimensions are considered: efficiency and
environmental protection. Efficiency is a measure of how natural
resources are used to obtain the foods of a given diet and is
Am J Clin Nutr doi: 10.3945/ajcn.113.071522. Printed in USA. ! 2014 American Society for Nutrition
Copyright (C) 2014 by the American Society for Nutrition
quantified by the ratio of inputs to outputs. Environmental protection
addresses the preservation of ecological systems that allow life on
earth: the biosphere. It is measured by environmental indicators such
as global warming potential, biodiversity, and eutrophication. Thus,
both key dimensions of environmental sustainability, the efficient
use of natural resources, and the avoidance of environmental deg-
radation in the production, preparation, and disposing of the food
consumed are to be considered in assessing the sustainability of
a diet.
ARE CURRENT FOOD PRODUCTION AND
CONSUMPTION PATTERNS SUSTAINABLE?
Evaluating the food systems in light of these 2 dimensions
provides a framework for assessing the environmental sustain-
ability of current practices. The basic inputs and outputs of the
food system as a whole are shown in Figure 1. The food system
takes inputs from the natural world in the form of natural
resourcesthat is, land, sun radiation, water, fossil energy, and
chemicals. Working together, these aforementioned inputs pro-
duce food for human societies. Food is the desired output of
the system; however, the food system also produces undesirable
outcomes in the form of solid, liquid, and gas waste. Societal
demand, which includes consumer preferences, is a major driver
of the food system (3). The life cycle of foods is determined by
the production, processing, transportation, storage, retail, and dis-
posal practices used; and consumer demands in a given society
define these interactions within the food system.
Agriculture is the practice of producing crops and raising
livestock. From an ecological perspective, agriculture involves
managing resources to capture solar energy and the transferring
of it to people for their use. For millennia, agriculture was a
spatially complex system of polycultures, and a variety of crops
and animals inhabited the same farm lands. Compared with
output (food produced), inputs were low and consisted of solar
energy, rain water, and animal waste for fertilizer (4). By effi-
ciency standards, the system was sustainable. With the advent of
industrial agriculture, farms became a monoculture enterprise,
1
From the Departments of Nutrition (JS) and Environmental Health and
Geoinformatic Sciences (SS), School of Public Health, Loma Linda University,
Loma Linda, CA.
2
Presented at the symposium Sixth International Congress on Vegetarian
Nutrition held in Loma Linda, CA, 2426 February 2013.
3
Address correspondence to J Sabate, Loma Linda University, Department
of Nutrition, Nichol Hall 1102, Loma Linda, CA 92350. E-mail: jsabate@
llu.edu.
doi: 10.3945/ajcn.113.071522.
1S of 7S
2S of 7S
FIGURE 1. Major drivers and outcomes of the food system.
with a single farm generally producing a single food item (5). The
main inputs are nonrenewable energy from fossil fuel and high
amounts of chemicals, and oil is also used to produce nitrogenous
fertilizers and irrigation water (5). These enormous inputs of energy
in modern agricultural practices have greatly increased food
production but have resulted in an energy imbalance.
FIGURE 2. Graphic summary of the various types of the food production: ratio of energy required to food energy delivery. Reprinted with permission from reference 6.
SABATE AND SORET
The increase in energy usage for food production from tra-
ditional to current practices is depicted in Figure 2. Originally,
agricultural activity resulted in a net gain in energy as more
energy was obtained from food than expended on its production.
One farmer could feed a family by using only the energy of his
labor and that provided by nature. As food production intensified
with the use of fossil fuel energy, the ratio increased for the
energy input to energy output from food (6). The imbalance
between the total energy required by the US food system and
the total food energy produced by the effort was reported by
the Center for Sustainable Systems (7). On-farm production
amounts to 21% of the total system energy usage, and 40% of
agriculture production energy go into making chemical fertil-
izers and pesticides. Large amounts of energy go into process-
ing, transporting, storing, and serving food. For every 10.3 quads
of the total energy used to produce food, only 1.4 quads of food
energy is created, yielding an overall energy efficient ratio
of .7:1 (7). From the energy perspective, the industrial food
system is very inefficient and because most of the energy inputs
are from nonrenewable sources such as fossil fuels, the current
system is unsustainable (5).
 SUSTAINABILITY OF PLANT-BASED DIETS
EFFICIENCY: ANIMAL COMPARED WITH PLANT FOODS
Raising animals for human food is an intrinsically inefficient
process. As we move up in the trophic chain there is a progressive
loss of energy. Grass-fed livestock subsists, but this is not the
main source of meat for human consumption in developed na-
tions. Modern husbandry (animal farms) is based on intensive
feeding of grain crops to animals (5). This grain could be a source
of food for humans. The same standards apply to the production
of other animal products such as eggs and dairy. Several authors
have computed the efficiency ratios of animal compared with
plant foods for human consumption. The amount of grain needed
to produce the same amount of meat varies from a ratio of 2.3 for
chicken to 13 for beef (Table 1). Pimentel and Pimentel (8)
established that, on average, 11 times greater fossil energy is
required to produce animal protein than plant protein for human
consumption. However, the energy-to-protein efficiency ratio
varies greatly by type of meat. More specifically, it is only 4
times greater for chicken protein compared with grain protein
but 40 times greater for beef protein compared with grain pro-
tein. We have previously reported that the ratio for water used in
the production of soy protein compared with the same quantity
of animal protein is from 4 to 26 and showed that the ratio
between soy protein and the different types of animal proteins
varies from 6 to 20 for fossil fuel usage (9). The land required to
raise the feed to produce animal protein is 617 times greater than
for soy protein (9).Thus, the conversion of plant foods to foods of
animal origin is an intrinsically inefficient process (w10:1).
The ratio of energy inputs to protein delivery is also qualitatively
different for animal compared with plant foods. As the concen-
tration of protein increases in plant foods, so does the efficiency. It
does not change or may even decrease in animal protein sources
(Figure 3) (10). High-protein plant foods such as soy beans and
other legumes have greater protein delivery energy efficiency
than cereals, which have a lower protein concentration. Therefore,
less energy is needed to produce the same amount of protein from
soy than from corn. However, very similar amounts of energy are
used to produce equivalent amounts of protein from different
sources of animal protein. In animal foods, the degree of protein
concentration seems to decrease the efficiency ratio of energy
inputs compared with protein outputs.
ENVIRONMENTAL INDICATORS: ANIMAL COMPARED
WITH PLANT FOODS
Research has been conducted on individual food items to determine
the impact of production and consumption on the environment.
The production of food for human consumption, particularly by
TABLE 1
Ratio of different inputs to animal outputs in US husbandry practices1
Grain fed:meat produced Fossil fuel energy:protein energy
kg:kg kcal:kcal
Beef 13:1 40:1
Eggs 11:1 39:1
Pork 5.9:1 14:1
Milk 14:1
Turkeys 3.8:1 10:1
Chickens 2.3:1 4:1
1
Adapted from reference 8.
3S of 7S
industrialized agricultural practices, causes significant emissions
of greenhouse gases (GHGs). These may occur directly from
carbon dioxide emissions from fossil fuel use on the farm or in the
supply chain, nitrous oxide emissions resulting from fertilizer
application, and methane emissions from animals or indirectly as
a result of land use change. In addition to agricultural production
process, the transport, processing, packaging, marketing, sales,
purchasing, and cooking of food also contribute to GHG emis-
sions. Berners-Lee et al (11) computed the GHG emissions from
61 food categories at point of purchase. With one exception, each
food with GHG emissions .10 kg CO2 equivalents/kg of product
is a meat or dairy food.
Approximately half of all food-related GHG emissions are
generated during farming. Farm-stage emissions include nitrous
oxide and methane from livestock and carbon dioxide from ag-
riculturally induced change in land use, especially deforestation.
Nitrous oxide from pasture land and arable land used to grow
food crops and methane from the digestive processes of ruminant
animals (eg, cows and sheep) account for 80% of all agricultural
GHG emissions (12). The emissions per unit of livestock product
vary by animal type and are much higher in ruminant animals
such as in cattle, sheep, and dairy farming than in pig and poultry
farming (13).
Beyond contributing disproportionately to GHG emissions
to the atmosphere, industrial husbandry operations damage the
environment by chemical runoff to water and land and animal waste.
This chemical pollution may cause acidification, algal blooms, and
dead zones in lakes and coastal areas; soil quality degradation;
habitat change; and biodiversity loss (5, 12, 14). We previously
computed 3 environmental degradation ratios of meat compared
with soy protein production and reported that the emissions of
acidifying substances, pesticides, and metals are respectively 7, 6,
and 100 times greater for meat protein compared with soy (9).
Thus, the intensive production of meat is considerably more taxing
to the environment than nutritionally equivalent plant protein foods.
COMPARING ENVIRONMENTAL INDICATORS OF
PLANT-BASED AND MEAT-BASED DIETS
Compared with plant foods, meat and dairy products are clearly
responsible for a hefty share of the natural resource utilization
and environmental burden of food production. However, looking
at dietary patterns, instead of single foods, is a more integrated
and realistic approach in the assessment of the environmental
impacts of producing foods for human consumption. We pre-
viously compared the environmental impacts of producing the
foods differentially consumed by vegetarians and nonvegetarians
from California and showed that the agricultural inputs required
for producing the nonvegetarian diet were 2.9 times more water,
2.5 times more primary energy, 13 times more fertilizer, and 1.4
times more pesticides than for the vegetarian diet (15).
Earlier studies reported several-fold increases in GHG emis-
sions and/or in environmental degradation scores for conventional
meat diets compared with vegetarian diets (16, 17). However,
these notable differences between dietary patterns also encom-
passed factors other than the diet composition, such as agricultural
production methods, food processing, and transportation. Such
factors, albeit important contributors to the overall environmental
footprint, are confounders in the assessment of the strict impact of
the composition of the diet.
4S of 7S SABATE AND SORET

FIGURE 3. Protein delivery efficiency in terms of energy use as a function of food protein content. Squares, plant-based foods; circles, animal-based foods.
The correlation coefficient r is shown in linear interpolations. Reprinted with permission from reference 10. GH, greenhouse; Vegs, vegetables.

Most of the research on the environmental impact of dietary
patterns has focused on the quantification of GHG emissions.
Berners-Lee et al (11) contrasted 6 meatless daily meal scenarios
with the average UK diet. Under the different dietary scenarios,
meat was replaced with either dairy products or plant-based
alternatives. Compared with the benchmark UK diet, reductions
in GHG emissions of 22% for vegetarian and 26% for vegan
scenarios were estimated (Figure 4). With the use of data from
Finland, Risku-Norja et al (18) estimated that a vegan diet would
reduce 48% of the agricultural and 34% of the overall food
system GHG emissions compared with the average Finnish diet.
Reductions in agricultural and overall food system GHG emis-
sions of only 33% and 23%, respectively, result when dairy is
removed and beef and mutton are replaced in the diet with pork
and poultry. Eshel and Martin (14, 19) estimated the decrease in
GHG emissions for a vegan diet compared with the mean US
diet on the basis of national food disappearance data. They cal-
culated a per capita annual decrease of 1.5 tons of CO2 equiv-
alents, which is equivalent to 33% fewer emissions from the
vegan food pattern than from the average US diet. This repre-
sents an 8% reduction in the per capita total GHG emissions in
the United States, which is similar to the reduction found for the
Finnish consumer (18).
Modeled reductions in GHG emissions ranging from 19% to
30% were estimated for several dietary scenarios involving the
partial replacement of animal products, meat and dairy, in the
average UK diet (20, 21). Furthermore, Stehfest et al (22) esti-
mated reductions in GHG emissions of 17% for CO2, 24% for
CH4, and 21% N2O at the global level from a complete switch to
meatless diets, in which all protein is derived from plants.

FIGURE 4. Reduction in greenhouse gas emissions of different prototype vegetarian and vegan diets compared with an average UK diet. Adapted by using
data from reference 11. CO2e, CO2 equivalents; p, person.
 SUSTAINABILITY OF PLANT-BASED DIETS
The aforementioned studies provide useful estimates of en-
vironmental impacts, but researchers relied on prototype vege-
tarian diets for the comparative estimates, an approach that has
inherent limitations. Prototype vegetarian diets may reflect the
preferences and potential bias of the researcher, and they do not
capture the variability and complexity of the foods consumed on
a daily basis by free-living individuals. Thus, impact assessments
based on ideal or designed diets may produce only theoretical
results and are not grounded in reality. We recently conducted
an assessment of the carbon footprint of dietary patterns using
a novel approach (23), and our methodology was based not on
researcher-designed, but on subject-defined food consumption
for a large and geographically representative study population
across North America. Furthermore, our study population exhibits
a great dietary variability with respect to the intake of animal and
plant foods, and we determined the emissions with global warming
potential (CO2, CH4, and N2O) of 210 foods included in a vali-
dated food-frequency questionnaire by implementing a life-cycle
impact assessment. We estimated the GHG emissions for each of
the .73,000 participants in the Adventist Health Study 2 according
to self-reported dietary preference. A reduction in emissions of
31% and 22% was estimated for vegetarians and semivegetarians,
respectively, compared with nonvegetarians (23). These esti-
mates should be regarded as conservative because the average
meat consumption of 64 g/d for nonvegetarian Adventists con-
trasts with the much higher 220 g/d in a typical Western diet.
THE CASE FOR PLANT-BASED DIETS AT THE GLOBAL
LEVEL
Is food security in the context of the current and projected
dramatic increase in global demand for animal products com-
patible with environmental sustainability? The answer to this
question must be considered against the backdrop of the worlds
demographic explosion and the increase in wealth among large
segments of the population in transitional and developing na-
tions (24, 25). These trends result in a dramatic increase in the
global demand for foods of animal origin, particularly meats and
dairy foods. Industrial livestock production is intrinsically re-
source-inefficient and highly taxing on the environment, ren-
dering the current food system environmentally and societally
unsustainable. We argue in this section that for a sustainable
future, a drastic reduction in the consumption of meat and dairy
foods by large segments of the worlds population is unavoidable.
Thus, plant-based diets at the global level are imperative (26).
The overall size and economic activity of humankind are exceeding
the biocapacity of the world. Until recently, it was assumed that
the worlds living and physical resources were inexhaustible. How-
ever, this is a false assumption attributable to a rapidly growing
reliance on nonrenewable natural resources, ie, fossil fuels. Many
human activities, including industrial agriculture, have reached
a level that could damage the systems that keep Earth in a desirable
state of ecological balance. The outcome could be irreversible
and, in some cases, lead to abrupt environmental change. Rockstrom
et al (3) identified 9 Earth-system self-regulatory processes and
associated thresholds that, if crossed, could generate unaccept-
able environmental change. Their analysis suggests that 3 of the
boundariesclimate change, rate of biodiversity loss, and inter-
ference with the nitrogen cyclehave already been trespassed.
Although the planetary boundaries are described in terms of
5S of 7S
individual quantities and separate processes, the boundaries are
tightly coupled. Humankind does not have the luxury of con-
centrating its efforts on any one of them in isolation from the
others (27). If one boundary is transgressed, then other bound-
aries are also under serious risk.
The world population is growing exponentially and is expected
to reach 9 billion by the year 2050 (25). The parallel increase in
wealth in large segments of the population of transitional and
developing countries has resulted in a markedly increased de-
mand for foods of animal origin, particularly meat and dairy
foods. Since 1963, there has been a 62% increase worldwide in
meat consumption, but a much greater increase of w300% has
occurred in developing nations. China has experienced a 9-fold
increase in per capita meat consumption of meat since the 1980s
(Figure 5) (28). In addition to wealth, other factors driving the
worldwide consumption of meat include urbanization, trade
liberalization, transnational food corporations, retailing growth,
food industry marketing, and consumer attitudes and behaviors
(29).
The projected increased size of the worlds population and the
increase in the appetite for meat are pushing our food systems to
unsustainable levels. Reports from authoritative agencies have
accumulated in the past few years that document the negative
environmental impacts of industrial meat production (12, 3032).
Industrial meat production is intrinsically resource-inefficient and
requires high inputs of nonrenewable natural resources and growth-
promoting antibiotics. It damages the environment through gaseous
emissions and chemical runoff into water and soil. Meat and dairy
production accounts for .80% of all GHG emissions from the
food sector (30) and for 24% of global GHG emissions (12). Thus,
the livestock industry is a major contributor to anthropogenic
GHG emissions liable for global climate change, one of the
planetary boundaries already transgressed by humanity and directly
threatening its sustainability.
Animal waste has become a public health problem and en-
vironmental hazard. Annually, 7 billion livestock in the US meat
industry generate 1.4 billion tons of wasteor w5 tons of waste
for every US citizen (5). These wastes, most of which go un-
treated, not only contain high concentrations of nitrogen, phos-
phorous, and potassium compounds and traces of metals but also
antibiotics, and are the source of .100 zoonotic pathogens that
may contaminate food and water supplies (33), thus representing
a direct threat to human health.
Food security and food sustainability are on a collision course.
The past half-century has seen a marked growth in food pro-
duction, allowing for a dramatic decrease in the proportion of the
worlds people who are hungry despite a doubling of the total
FIGURE 5. Trends in per capita meat consumption. Adapted by using
data from reference 28.
6S of 7S SABATE AND SORET
population. Nevertheless, 1 in 7 persons today consume diets
insufficient in protein and energy, and 2 in 7 persons have some
micronutrient deficiency. With a larger-than-ever world population
that has a greater appetite for meat and with the threat of lower
food yields because of substantial climate change (34), the world
food system is facing a set of intersecting challenges. These
challenges require radical changes not only in the way food is
produced, processed, stored, distributed, and accessed but also in
the types of food consumed.
Previously, the main solution to food shortages was to bring
more land into agriculture and to exploit fish stocks. These are not
sustainable options. Neither is increasing the intensification of
existing underperforming agricultural landscapes. Closing the
food yield gap, while considering sustainability goals, will re-
quire new approaches, including the adaptation of conventional
agriculture to more low-input and precision practices (3537).
Waste and losses along the supply chain are estimated to involve
more than one-third of the food produced in the world (36).
Reducing food waste could substantially improve food security
and simultaneously decrease environmental degradation. Improv-
ing food yields, reducing agricultures environmental impacts,
and reducing waste are necessary but are not sufficient strategies
to ensure global food security. More food can be delivered by
realigning agricultural and dietary preferences.
Shifting diets from animal-based to plant-based at the global
level is of paramount importance in achieving food security and
sustainability goals. Decreasing consumption of meat and other
animal products will free up large amounts of food that could be
consumed directly by humansfor example, soy and grains. Foley
et al (36) recently estimated the potential to increase the global
food supply by shifting 16 major crops from the current mix of
uses (eg, human consumption, livestock feed, and biofuels) to
human food consumption only: they estimated a 28% increase in
food availability, or the equivalent of a 49% increase in dietary
energy for human consumption. These staggering figures, re-
sulting from the implementation of a single strategy, have pro-
found implications for both food security and environmental
sustainability. By 2050, the worlds population is expected to
plateau at 9 billion, a 28% increase from the current 7 billion
(25). Shifting the portion of current crops directed to livestock
feed and nonfood uses to human consumption could fill the
future food gap and exceed dietary energy needs. In addition, the
proposed dietary shift would result in major environmental
benefits because it does not require an overall increase in plant
crops at the global level, and it would eliminate the environ-
mental impacts related to livestock. As presented above, in-
dustrial livestock production contributes disproportionally to
numerous forms of environmental degradation. For example,
livestock-related GHG emissions are .80% of those generated
by the food sector and 24% of global GHG emissions.
ISSUES AND CHALLENGES
The proposal to drastically reduce meat consumption at the global
level is ground-shaking. Some have even branded it a revolu-
tionary approach and have argued that wholesale dietary shifts
may not be realistic (36). However, the proposed transition does not
need to be an all or nothing process because even only incremental
steps could be extremely helpful in solving food availability and
sustainability challenges. We are fully aware that such a drastic
dietary shift is complex and implicates behavioral and policy
challenges at many levels.
The adequacy of meatless diets has been a recurrent theme
in the nutrition literature. On the basis of a higher concentration
of essential nutrients in animal products, meat and dairy were
considered essential in large proportions for adequate nutrition in
the daily diet, and consumption of plant-based diets was con-
sidered inadequate. This nutritional paradigm has changed in the
past few decades (38) as data now support that most plant-based
diets are healthier than meat-based diets and yield greater lon-
gevity and lower chronic diseases among those who consume
vegetarian diets (3942). Furthermore, there is growing evidence
linking meat consumption, in particular red meat and processed
meat, with detrimental health outcomes (4345). From a strict
health perspective, there is no need to consume meat.
For millennia, large segments of the worlds population thrived
on diets with little or no meat. In the past century, however, the
concept of eating meat as the paramount source of protein has
become deeply engrained in the psyche and culture of Western
countries and now pervades many other cultures and nations.
Undertaking a drastic downshift in meat consumption will face
serious obstacles and opposition at many levels: the consumers
taste preferences; some culinary traditions; established social norms;
economic forces, such as the livestock industry; and current national
and international food policies. Several proposals have been
advanced to accomplish the transition from animal protein to
plant-based protein consumption. Some of them include con-
sumer education focused on the environmental and health merits
of plant-based diets, the promotion of food guidelines based on
health and sustainability criteria, developing attractive and cul-
turally acceptable plant-based meat-alternative foods, and re-
aligning current fiscal policy (food subsidies and taxation) with
efficiency and environmental criteria (4650).
CONCLUSIONS
Plant-based diets in comparison to meat-based diets are more
sustainable because they use substantially less natural resources
and are less taxing on the environment. The worlds demographic
explosion and the increase in the appetite for animal foods render
the food system unsustainable. Food security and food sustain-
ability are on a collision course. Changing course (to avoid the
collision) will require extreme downward shifts in meat and
dairy consumption by large segments of the world population.
Although other approaches should be pursued, they are insufficient
to make the global food system sustainable, and therefore the
dietary shift is an inevitable strategy.
Throughout history, forced either by necessity or by choice,
large segments of the worlds population have thrived on plant-
based diets. In the past, meatless diets have been advocated on
the basis of religious, ethical, or philosophical values, not science.
It is only in the past 150 y that empirical evidence has yielded
dietary recommendations.
Will going back to plant-based diets be the best way forward
for a sustainable future? Agricultural and nutritional policies
that lead to the adoption of plant-based diets at the global level
will simultaneously optimize the food supply, health, environ-
mental, and social justice outcomes for the worlds population.
Implementing such policies is not free of political challenges
but is perhaps the most rational, scientific, and moral path for
 SUSTAINABILITY OF PLANT-BASED DIETS
a sustainable future of the human race and other living creatures
of the biosphere that we share.
The authors responsibilities were as followsJS: wrote the first draft of
the manuscript; and SS: contributed to the content and editing of the man-
uscript. The authors had no conflicts of interest to declare.
REFERENCES
1. Burlingame B, Dernini S, eds. Sustainable diets and biodiversity: di-
rections and solutions for policy, research and action. Proceedings of
the International Scientific Symposium. Biodiversity and sustainable
diets united against hunger; 35 Nov 2010; Rome, Italy. Rome, Italy:
Food and Agriculture Organization, 2012.
2. Fanzo J, Cogill B, Mattei F. Technical brief: metrics of sustainable diets
and food systems. Maccarese, Italy: Bioversity International, 2012:18.
3. Rockstrom J, Steffen W, Noone K, Persson A, Chapin FS III, Lambin
EF, Lenton TM, Scheffer M, Folke C, Schellnhuber HJ, et al. A safe
operating space for humanity. Nature 2009;461:4725.
4. Altieri MA. Linking ecologists and traditional farmers in the search for
sustainable agriculture. Front Ecol Environ 2004;2:3542.
5. Horrigan L, Lawrence RS, Walker P. How sustainable agriculture can
address the environmental and human health harms of industrial agri-
culture. Environ Health Perspect 2002;110:44556.
6. Steinhart CE, Steinhart JS. Energy: sources, use and role in human
affairs. North Scituate, MA: Duxbury Press, 1974.
7. Center for Sustainable Systems, University of Michigan. U.S. Food
System Factsheet. Publication CSS01-06. Available from: http://css.
snre.umich.edu/css_doc/CSS01-06.pdf (cited 23 April 2013).
8. Pimentel D, Pimentel M. Sustainability of meat-based and plant-based
diets and the environment. Am J Clin Nutr 2003;78:660S3S.
9. Reijnders L, Soret S. Quantification of the environmental impact of
different dietary protein choices. Am J Clin Nutr 2003;78:664S8S.
10. Gonzalez AD, Frostell B, Carlsson-Kanyama A. Protein efficiency per
unit energy and per unit greenhouse gas emissions: potential contribution
of diet choices to climate change mitigation. Food Policy 2011;36:56270.
11. Berners-Lee M, Hoolohan C, Cammack H, Hewitt CN. The relative
greenhouse gas impacts of realistic dietary choices. Energy Policy
2012;43:18490.
12. Steinfeld HGP. Livestocks long shadow: environmental issues and
options. Rome, Italy: Food and Agriculture Organization of the United
Nations, 2006.
13. Friel S, Danogour AD, Garnett T, Lock K, Chalabi Z, Roberts I, Butler
A, Butler CD, Waage J, McMichael AJ, et al. Public health benefits of
strategies to reduce greenhouse-gas emissions: food and agriculture.
Lancet 2009;374:201625.
14. Eshel G, Martin PA. Geophysics and nutritional science: toward
a novel, unified paradigm. Am J Clin Nutr 2009;89:1710S6S.
15. Marlow HJ, Hayes WK, Soret S, Carter RL, Schwab ER, Sabate J. Diet
and the environment: does what you eat matter? Am J Clin Nutr 2009;
89(suppl):1966703S.
16. Carlsson-Kanyama A, Gonzalez AD. Potential contributions of food
consumption patterns to climate change. Am J Clin Nutr 2009;89
(suppl):1704S9S.
17. Baroni L, Cenci L, Tettamanti M, Berati M. Evaluating the environ-
mental impact of various dietary patterns combined with different food
production systems. Eur J Clin Nutr 2007;61:27986.
18. Risku-Norja H, Kurppa S, Helenius J. Dietary choices and greenhouse
gas emissions assessment of impact of vegetarian and organic options
at national scale. Progr Ind Ecol 2009;6:34054.
19. Eshel G, Martin PA. Diet, energy and global warming. Earth Interact
2006;10:117.
20. Macdiarmid JI, Kyle J, Gorgan GW, Loe J, Fyfe C, Johnstone A, McNeill G.
Sustainable diets for the future: can we contribute to reducing greenhouse
gas emissions by eating a healthy diet? Am J Clin Nutr 2012;96:6329.
21. Scarborough P, Allender S, Clarke D, Wickramasinghe K, Rayner M.
Modelling the health impact of environmentally sustainable dietary
scenarios in the UK. Eur J Clin Nutr 2012;66:7105.
22. Stehfest E, Bouwman L, van Vuuren DP. Climate benefits of changing
diet. Clim Change 2009;96:83102.
23. Soret S, Mejia A, Batech M, Jaceldo-Siegl K, Harwatt H, Sabate J.
Climate change mitigation and health effects of varied dietary patterns
7S of 7S
in real-life settings throughout North America. Am J Clin Nutr 2014;
100(suppl):490S5S.
24. Credit Suisse Research Institute. Global Wealth Report 2013. October
2013. Available from: https://publications.credit-suisse.com/tasks/render/
file/?fileID=BCDB1364-A105-0560-1332EC9100FF5C83 (cited 25
November 2013).
25. US Census Bureau. Available from: http://www.census.gov/population/
international/data/idb/worldpoptotal.php (cited 5 October 2013).
26. Saxena AM. The vegetarian imperative. Baltimore, MD: The Johns
Hopkins University Press, 2011.
27. Aiking H. Protein production: planet, profit, plus people? Am J Clin
Nutr 2014;100(suppl):483S9S.
28. FAO. Food balance sheets. 2009. Available from: http://faostat3.fao.
org/home/index.html#DOWNLOAD (cited 5 October 2013).
29. Kearney J. Food consumption trends and drivers. Phil Trans R Soc B
2010;365:2793807.
30. McMichael AJ, Powles JW, Butler CD, Uauy R. Food, livestock pro-
duction, energy, climate change, and health. Lancet 2007;370:125363.
31. Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor
M, Miller HL, eds. Climate Change 2007: The Physical Science Basis.
Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. New York, NY and Cam-
bridge, United Kingdom: Cambridge University Press, 2007 [996 pp].
32. Pew Commission on Industrial Farm Animal Production. Putting meat
on the table: industrial farm animal production in America. Available
from: http://www.pewtrusts.org/our_work_report_detail.aspx?id=38442
(cited 18 February 2013).
33. Akhtar AZ, Greger M, Ferdowsian H, Frank E. Health professionals
roles in animal agriculture, climate change, and human health. Am J
Prev Med 2009;36:1827.
34. Schmidhuber J, Tubiello FN. Global food security under climate
change. Proc Natl Acad Sci USA 2007;104:197038.
35. Mueller ND, Gerber JS, Johnston M, Ray DK, Foley JA. Closing the yield
gap through nutrient and water management. Nature 2012;490:2547.
36. Foley JA, Ramankutty N, Brauman KA, Cassidy ES, Gerber JS,
Johnston M, Mueller ND, OConnell C, Ray DK, West PC, et al. So-
lutions for a cultivated planet. Nature 2011;478:33742.
37. Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, Muir
JF, Pretty J, Robinson S, Thomas SM, Toulmin C. Food security: the
challenge of feeding 9 billion people. Science 2010;327:8128.
38. Sabate J. The contribution of vegetarian diets to health and disease:
a paradigm shift? Am J Clin Nutr 2003;78(suppl):502S7S.
39. Rizzo NS, Sabate J, Jaceldo-Siegl K, Fraser GE. Vegetarian dietary
patterns are associated with a lower risk of metabolic syndrome: the
Adventist Health Study 2. Diabetes Care 2011;34:12257.
40. Orlich MJ, Singh PN, Sabate J, Jaceldo-Siegl K, Fan J, Knutsen S,
Beeson WL, Gary E, Fraser GE. Vegetarian dietary patterns and mortality
in Adventist Health Study 2. JAMA Intern Med 2013;173(13):12308.
41. Tonstad S, Stewart K, Oda K, Batech M, Herring RP, Fraser GE.
Vegetarian diets and incidence of diabetes in the Adventist Health
Study-2. Nutr Metab Cardiovasc Dis 2013;23:2929.
42. Tantamango-Bartley Y, Jaceldo-Siegl K, Fan J, Fraser G. Vegetarian
diets and the incidence of cancer in a low-risk population. Cancer
Epidemiol Biomarkers Prev 2013;22:28694.
43. American Institute for Cancer Research; World Cancer Research Fund.
Food, nutrition, physical activity and the prevention of cancer: a global
perspective. Washington, DC: AICR, 2007.
44. Sinha R, Cross A, Graubard B, Leitzmann M, Schatzkin A. Meat intake
and mortality. Arch Intern Med 2009;169:56271.
45. Mozaffarian D, Hao T, Rimm EB, Willett WC, Hu FB. Changes in diet
and lifestyle and long-term weight gain in women and men. N Engl J
Med 2011;364:2392404.
46. deBoer J, Aiking H. On the merits of plant-based proteins for global
food security: marrying macro and micro perspectives. Ecol Econ
2011;70:125965.
47. Vinnari M, Tapio P. Sustainability of diets: from concepts to gover-
nance. Ecol Econ 2012;74:4654.
48. Clonan A, Holdsworth M. The challenges of eating a healthy and
sustainable diet. Am J Clin Nutr 2012;96:45960.
49. Goodland R. Environmental sustainability in agriculture: diet matters.
Ecol Econ 1997;23:189200.
50. Lang T, Barling D. Nutrition and sustainability: an emerging food
policy discourse. Proc Nutr Soc 2013;72:112.
Dietary Fatty Acids
BRETT WHITE, MD, University of Southern California, Keck School of Medicine, Los Angeles, California

Fatty acids can be divided into four general categories: saturated, monounsaturated, polyun-
saturated, and trans fats. Saturated fatty acids and trans fats are associated with an increased
risk of coronary heart disease. Monounsaturated fatty acids and polyunsaturated fatty acids
are associated with a decreased risk of coronary heart disease, although these associations are
not uniformly supported in the literature. Omega-3 fatty acids, which are a type of polyun-
saturated fatty acid, have been studied as potential therapy for a variety of medical conditions
because of their suspected anti-inflammatory properties. Omega-3 fatty acids have also been
shown to provide some benefit to patients with cystic fibrosis, and may have a protective effect
against dementia. Physicians should counsel patients about the importance of avoiding hydro-
genated oils and foods containing trans fats because of their association with coronary heart
disease in observational studies. (Am Fam Physician. 2009;80(4):345-350, 372. Copyright
2009 American Academy of Family Physicians.)

F
See related editorial atty acids are long-chain hydro- hydrogen atoms is missing because of a double
T

on page 330. carbons that can be separated into bond between two carbon atoms, it is called
Patient information: four categories: saturated, mono- an unsaturated fatty acid. A fatty acid with
T

A handout on trans fats,
written by the authors of
this article, is provided on
page 372.
unsaturated, polyunsaturated, and
trans fats. More than 20 types of fatty acids
are found in foods; some of these are listed in
Table 1. Sources of fatty acids include fruits,
vegetable oils, seeds, nuts, animal fats, and
fish oils. Essential fatty acids, such as omega-3
fatty acids, serve important cellular functions.
They are a necessary part of the human diet
because the body has no biochemical pathway
to produce these molecules on its own.
Biochemical Structure
In saturated fatty acids, the carbon chain has
the maximum number of hydrogen atoms
attached to every carbon atom. If a pair of
a single double bond is monounsaturated,
whereas a fatty acid with more than one dou-
ble bond is polyunsaturated (Figure 1). The
carbon-carbon double bond found in mono-
unsaturated or polyunsaturated fatty acids
can exist in the cis or trans configuration.
When the two hydrogen atoms are on oppo-
site sides of the double bond, the configura-
tion is called trans. When the hydrogen atoms
are on the same side of the double bond, the
configuration is called cis (Figure 2).
Saturated Fats
Studies have shown that consuming satu-
rated fatty acids has a detrimental effect on

Table 1. Well-Known Fatty Acids

Number of Essential
Name carbon atoms Type of fatty acid fatty acid Common sources

Palmitic acid 16 Saturated No Palm oil

Stearic acid 18 Saturated No Animal fat
Oleic acid 18 Monounsaturated No Olive oil
Linoleic acid 18 Polyunsaturated Yes Safflower oil
Linolenic acid 18 Polyunsaturated Yes Soybean oil
Arachidonic acid 20 Polyunsaturated Yes Meat, dairy
Eicosapentaenoic acid 20 Polyunsaturated Yes Fish oil
Docosahexaenoic acid 22 Polyunsaturated Yes Fish oil

Downloaded from the American Family Physician Web site at www.aafp.org/afp. Copyright 2009 American Academy of Family Physicians. For the private, noncommer-
August 15,
cial 2009
use of oneVolume 80, Number 4 site. All other rights reserved.
www.aafp.org/afp American requests. 345
Family Physician
V
individual user of the Web Contact copyrights@aafp.org for copyright questions and/or permission
 SORT: KEY RECOMMENDATIONS FOR PRACTICE

Evidence
Clinical recommendation rating References

Increased intake of saturated fats and trans B 4-7, 9, 40, 41 serum lipids by increasing low-density lipo-
fats may increase the risk of coronary heart
disease.
protein (LDL) cholesterol levels.1 There is
Increased intake of monounsaturated and B 4, 7-13
some evidence that short-chain fatty acids
polyunsaturated fats may decrease the risk (fewer than 10 carbon atoms) are less likely
of coronary heart disease. to affect serum cholesterol levels, whereas
Regular omega-3 fatty acid supplementation B 20, 21 longer-chain fatty acids (12, 14, or 16 carbon
may provide some benefits for persons with
atoms) are more likely to raise LDL levels.2
cystic fibrosis and dementia.
One exception to this is stearic acid (18 car-
There is no clear effect of omega-3 fatty B 18
acid consumption on cardiovascular bon atoms), which does not appear to raise
events in persons with, or at high risk of, serum cholesterol levels.3 Increased con-
cardiovascular disease. sumption of saturated fatty acids has also
been associated with an increased risk of
A = consistent, good-quality patient-oriented evidence; B = inconsistent or limited-
quality patient-oriented evidence; C = consensus, disease-oriented evidence, usual coronary heart disease (CHD; Table 2).4-13
practice, expert opinion, or case series. For information about the SORT evidence
rating system, go to http://www.aafp.org/afpsort.xml. Monounsaturated Fats
Several large observational studies have
found an association between an increased
Saturated fatty acid (stearic acid) intake of monounsaturated fatty acids and a decreased
risk of CHD (Table 2).4-13 One large study failed to
H H H H H H H H H H H H H H H H H
| | | | | | |
O
| | | | | | | | | | find a similar association, although it was limited to a
HCCCCCCCCCCCCCCCCCC=O
| | | | | | | | | | | | | | | | | H trial of beta-carotene and alpha-tocopherol in persons
H H H H H H H H H H H H H H H H H who smoke.9 Evidence from controlled clinical stud-
ies has shown that monounsaturated fatty acids favor-
Monounsaturated fatty acid (oleic acid) ably affect a number of risk factors for CHD, including
H H H H H H H H H H H H lowering total and LDL cholesterol levels, protecting
| | | | | | | | | | | | H H H against thrombogenesis, reducing LDL susceptibility
HCCCCCCCCC=C C C | | | H H
| | | | | | | |
| | C C C
| |
to oxidation, and producing a more favorable glycemic
| | C C
H H H H H H H H H H C = O
H H
| |
| profile.14
H H
O

H
H

Polyunsaturated Fats
Polyunsaturated fatty acid Polyunsaturated fatty acids, which include omega-3 fatty
(linolenic acidan omega-3 fatty acid)
acids, have been studied extensively for their effect on
H H H H H H H several diseases (Table 3).15-25 Omega-3 fatty acids have
H H| | | | | | | | H H
| | H H H
|
C= CCC=CC C= C C | | | H H H
been shown to provide some benefit to patients with
C
H C| | | | | C
| | C C C
| |
| cystic fibrosis, and may have a protective effect against
H H H H
H H
| | C C
C = O
H H H
| |
| dementia.20,21 Omega-3 fatty acids are thought to be ben-
H H
O

H eficial in some inflammatory-related diseases because

H

they displace omega-6 fatty acids, including arachidonic

Figure 1. Molecular structure of fatty acids. acid, in the cell membrane. This reduces the creation
of metabolic end products, including prostaglandins,
thromboxanes, and leukotrienes.26
H
H

Additionally, maternal omega-3 fatty acid supple-

|

C=C C=C
|

mentation during pregnancy and lactation may pro-

H

cis trans vide a beneficial effect on the cognitive development

of infants and children, but evidence is inconclusive
Figure 2. The carbon-carbon double bond found in mono- about the benefits of omega-3 supplementation in pre-
unsaturated or polyunsaturated fatty acids can exist in term and full-term infants.27 Several review articles have
the cis or trans configuration. When the two hydrogen examined the relationship between omega-3 fatty acid
atoms are on opposite sides of the double bond, the con-
figuration is called trans. When the hydrogen atoms are intake and cancer incidence, including prostate cancer,
on the same side of the double bond, the configuration colon cancer, and skin cancer.28-30 A recent systematic
is called cis. review concluded that the literature does not support an

346 American Family Physician www.aafp.org/afp Volume 80, Number 4 V August 15, 2009
 Dietary Fatty Acids
Table 2. Fatty Acids and Coronary Heart Disease

Years of
Study Number of patients Study type follow-up Findings

Saturated fatty acids

Seven Countries Study 4 12,770 men Cross-population 5, 10, 15 Strong correlation between total cholesterol and
the percent of energy intake from saturated
fatty acids

Japan-Honolulu- 11,900 men Cross-population NA Correlation between increased consumption of

San Francisco Study 5 saturated fatty acids, and increased serum choles-
terol levels and increased rate of CHD mortality

Ireland-Boston Diet- 1,001 men Prospective 20 Patients who died of CHD had a higher intake of
Heart Study 6 cohort saturated fatty acids and cholesterol

Nurses Health Study 7 80,082 women Prospective 14 Positive association between percent of energy
cohort intake from saturated fatty acids and increased
risk of CHD

Monounsaturated fatty acids

Seven Countries Study 4 12,770 men Cross-population 5, 10, 15 Low death rate from heart disease in men
consuming large quantities of olive oils (primarily
monounsaturated fatty acids)

Nurses Health Study 7 80,082 women Prospective 14 Association between increased energy intake from
cohort monounsaturated fatty acids and a relative risk
reduction in coronary disease

Coronary Mortality in NA Epidemiologic NA Association between increased consumption of

France and Finland 8 monounsaturated fatty acids and low rates of
CHD mortality

Alpha-Tocopherol, 21,930 men who Prospective 6.1 No association between monounsaturated fatty
Beta-Carotene Cancer smoke cohort acids and the risk of CHD
Prevention Study 9

Polyunsaturated fatty acids

Finnish Mental Hospital 676 men RCT 6 Increased consumption of polyunsaturated fatty
Study 10 acids associated with reduction of CHD

Los Angeles Veteran 424 men RCT 8 Increased consumption of polyunsaturated fatty
Hospital 11 acids associated with reduction of CHD

Oslo Diet-Heart Study 12 206 men with history RCT 5 Increased energy from polyunsaturated fatty
of myocardial acids associated with significant reduction in
infarction cholesterol and coronary events

Minnesota Coronary 4,393 men, RCT 4.5 Increased consumption of polyunsaturated

Survey 13 4,664 women fatty acids associated with reduction in serum
cholesterol but no change in CHD

CHD = coronary heart disease; NA = not applicable; RCT = randomized controlled trial.
Information from references 4 through 13.

association between increased omega-3 fatty acid intake
and reduced cancer incidence.31
Various studies 4-13 and a report from the Institute of
Medicine32 support the increased consumption of polyun-
saturated fatty acid for the prevention of CHD (Table 2 4-13).
However, a Cochrane review of 48 randomized controlled
trials (RCTs) and 41 cohort analyses suggests that fur-
ther high-quality trials are needed to confirm the protec-
tive effect of omega-3 fatty acids in persons at increased
risk of cardiovascular disease.18 The authors concluded
that increased consumption of omega-3 fatty acids did
not significantly alter total mortality or combined car-
diovascular events in persons with cardiovascular dis-
ease, in persons at high risk of cardiovascular disease,
or in the general population.18 Excessive consumption of
polyunsaturated fats is not recommended because of the
increased risk of excessive weight gain and the increased
risk of gallstone formation in some persons.33

August 15, 2009 V Volume 80, Number 4 www.aafp.org/afp American Family Physician 347
 Table 3. Cochrane Reviews of Omega-3 Fatty Acids

Disease process Cochrane review conclusions

Asthma Little evidence to recommend persons with asthma modify intake of omega-3 fatty acids15
Cancer cachexia Insufficient data16
Claudication Omega-3 has limited benefit in persons with intermittent claudication, but no evidence of consistent
improved clinical outcomes17
Coronary heart disease No clear effect of omega-3 fatty acid consumption on cardiovascular events in persons with, or at high
risk of, cardiovascular disease18
Crohn disease Insufficient data to recommend routine use of omega-3 supplements for maintenance of remission19
Cystic fibrosis Regular omega-3 supplements may provide some benefits for persons with cystic fibrosis 20
Dementia May be a protective effect of omega-3 fatty acids against dementia 21
Diabetes (type 2) Omega-3 supplementation lowers triglyceride and very-low-density lipoprotein levels in patients with
diabetes, but may raise low-density lipoprotein levels; no effect on glycemic control or fasting insulin 22
Kidney transplant recipients Insufficient evidence to recommend fish oil therapy 23
Schizophrenia Inconclusive results; use of omega-3 remains experimental 24
Ulcerative colitis No evidence to support omega-3 fatty acids in maintenance of remission 25

Information from references 15 through 25.

Table 4. Fat and Cholesterol Content of Foods Containing Fatty Acids*

Percent daily Combined Percent daily

Common Total Saturated value for Trans saturated and Cholesterol value for
Product serving size fat (g) fat (g) saturated fat fat (g) trans fat (g) (mg) cholesterol

Butter 1 tablespoon 11 7 35 0 7 30 10
Cake (pound cake) 1 slice (80 g) 16 3.5 18 4.5 8 0 0
Candy bar 1 (40 g) 10 4 20 3 7 <5 1
Cookies (cream filled) 3 (30 g) 6 1 5 2 3 0 0
Doughnut 1 18 4.5 23 5 9.5 25 8
French fried potatoes Medium size 27 7 35 8 15 0 0
(fast food) (147 g)
Margarine, stick 1 tablespoon 11 2 10 3 5 0 0
Margarine, tub 1 tablespoon 7 1 5 0.5 1.5 0 0
Mayonnaise (soybean 1 tablespoon 11 1.5 8 0 1.5 5 2
oil)||
Milk, skim 1 cup 0 0 0 0 0 5 2
Milk, whole 1 cup 7 4.5 23 0 4.5 35 12
Potato chips Small bag 11 2 10 3 5 0 0
(42.5 g)
Shortening 1 tablespoon 13 3.5 18 4 7.5 0 0

*Nutrient values rounded based on the U.S. Food and Drug Administration (FDA) nutrition labeling regulations.
From the FDA Table of Trans Values, January 20, 1995.
1995 U.S. Department of Agriculture (USDA) composition data.
From the 2002 USDA National Nutrient Database for Standard Reference, Release 15.
||Prerelease values derived from the 2003 USDA National Nutrient Database for Standard Reference, Release 16.
Adapted from the U.S. Food and Drug Administration. Revealing trans fats. FDA Consumer Magazine. SeptemberOctober 2003. Publication no.
FDA051329C. http://www.fda.gov/FDAC/features/2003/503_fats.html. Accessed March 11, 2009.

Trans Fats
Most trans fats are manufactured through a process
called hydrogenation, which is the artificial addition of
hydrogen atoms to unsaturated oils. Hydrogenation con-
verts liquid vegetable oils to solid or semi-solid fats that
remain stable at room temperature. These fats can then
be incorporated into certain food products (e.g., cook-
ies, chips) to increase shelf life (Table 434). Until recently,
trans fats have been the predominant fat used in most
types of commercial baked goods.

348 American Family Physician www.aafp.org/afp Volume 80, Number 4 V August 15, 2009

Table 5. Trans Fat Consumption and CHD
Number Years of
Study of patients follow-up
Alpha-Tocopherol, 21,930 men 6.1
Beta-Carotene Cancer who smoke
Prevention Study 9
Zutphen Elderly Study 39 667 men 10
Health Professionals 43,757 men 6 1.36 (95% CI,
Follow-up Study 40
Nurses Health Study 7 80,082 women 14
CHD = coronary heart disease; CI = confidence interval.
*For each isocaloric substitution of 2 percent of total energy intake with trans-fatty
acids.
Information from references 7, 9, 40, and 41.
The consumption of trans fats has been directly linked
to an increase in CHD. Trans fats appear to have a detri-
mental effect on serum lipids by increasing LDL choles-
terol and triglyceride levels, and reducing high-density
lipoprotein cholesterol levels.35-39 Studies have shown that
substituting 2 percent of total energy intake with trans-
fatty acids is associated with a 14 to 36 percent increase in
the incidence of CHD (Table 5).7,9,40,41 Observational data
from the Nurses Health Study suggests that replacing
5 percent of energy from saturated fat with energy from
unsaturated fats could reduce the risk of CHD by 42 per-
cent; however, replacing only 2 percent of energy from
trans fats with energy from nonhydrogenated, unsatu-
rated fats could reduce the risk of CHD by 53 percent.7
There are no prospective RCTs that have demonstrated a
reduction in morbidity and mortality through selective
reduction of trans-fat consumption.
Dietary Recommendations
The Nutrition Committee of the American Heart Asso-
ciation has recommended that no more than 30 percent
of a persons daily calories come from fat.42 Of that, less
than 7 percent of total calories should be from satu-
rated fatty acids, and less than 1 percent should be from
trans-fatty acids.42 Certain fatty acids, such omega-3
fatty acids, are preferable to saturated fats and should be
substituted for saturated fats when possible. Physicians
should emphasize to patients that consuming too much
of any fat contributes to caloric intake and weight gain.
Physicians should also stress the importance of minimiz-
ing or avoiding trans fats from hydrogenated oils because
August 15, 2009 V Volume 80, Number 4
Dietary Fatty Acids
of their harmful effects on cholesterol levels
and their link to heart disease.
Relative risk of The Author
CHD with trans-
fat consumption* BRETT WHITE, MD, is medical director of the Family Medi-
cine Center at the Keck School of Medicine at the Univer-
1.14 (95% CI, sity of Southern California (USC) in Los Angeles. He is also
0.96 to 1.35) an assistant professor in the Department of Family Medi-
cine at USC.
1.28 (95% CI, Address correspondence to Brett White, MD, University
1.01 to 1.61) of Southern California, Keck School of Medicine, 1510
San Pablo St., Suite 104, Los Angeles, CA 90033 (e-mail:
1.03 to 1.81) brettwhitemd@gmail.com). Reprints are not available
1.33 (95% CI, from the author.
1.07 to 1.66) Author disclosure: Nothing to disclose.
REFERENCES
1. Mensink RP, Katan MB. Effect of dietary fatty acids on
serum lipids and lipoproteins. A meta-analysis of 27 tri-
als. Arterioslcer Thromb. 1992;12(8):911-919.
2. Chait A, Brunzell JD, Denke MA, et al. Rationale of the
diet-heart statement of the American Heart Association.
Report of the Nutrition Committee. Circulation. 1993;88(6):3008-3029.
3. Hu FB, Manson JE, Willett WC. Types of dietary fat and risk of coronary
heart disease: a critical review. J Am Coll Nutr. 2001;20(1):5-19.
4. Keys A. Coronary heart disease in seven countries. Circulation. 1970;41
(4 suppl):1-198.
5. Kagan A, Harris BR, Winkelstein W Jr, et al. Epidemiologic studies of
coronary heart disease and stroke in Japanese men living in Japan,
Hawaii and California: demographic, physical, dietary and biochemical
characteristics. J Chronic Dis. 1974;27(7):345-364.
6. Kushi LH, Lew RA, Stare FJ, et al. Diet and 20-year mortality from coro-
nary heart disease. The Ireland-Boston Diet-Heart Study. N Engl J Med.
1985;312(13):811-818.
7. Hu FB, Stampfer MJ, Manson JE, et al. Dietary fat intake and the risk of cor-
onary heart disease in women. N Engl J Med. 1997;337(21):1491-1499.
8. Artaud-Wild SM, Connor SL, Sexton G, Connor WE. Differences in
coronary mortality can be explained by differences in cholesterol and
saturated fat intakes in 40 countries but not in France and Finland. A
paradox. Circulation. 1993;88(6):2771-2779.
9. Pietinen P, Ascherio A, Korhonen P, et al. Intake of fatty acids and
risk of coronary heart disease in a cohort of Finnish men. The Alpha-
Tocopherol, Beta-Carotene Cancer Prevention Study. Am J Epidemiol.
1997;145(10):876-887.
10. Turpeinen O, Karvonen MJ, Pekkarinen M, Miettinen M, Elosuo R,
Paavilainen E. Dietary prevention of coronary heart disease: the Finnish
Mental Hospital Study. Int J Epidemiol. 1979;8(2):99-118.
11. Dayton S, Pearce ML, Hashimoto S, Dixon WJ, Tomiyasu U. A controlled
clinical trial of a diet high in unsaturated fat in preventing complications
of atherosclerosis. Circulation. 1969;40(1 suppl 2):1-63.
12. Leren P. The Oslo diet-heart study. Eleven-year report. Circulation. 1970;
42(5):935-942.
13. Frantz ID Jr, Dawson EA, Ashman PL, et al. Test of effect of lipid low-
ering by diet on cardiovascular risk. The Minnesota Coronary Survey.
Arteriosclerosis. 1989;9(1):129-135.
14. Kris-Etherton PM. AHA science advisory: Monounsaturated fatty acids
and risk of cardiovascular disease. J Nutr. 1999;129(12):2280-2284.
15. Woods RK, Thien FC, De Luca S, Abramson MJ. Dietary marine fatty
acids (fish oil) for asthma in adults and children. Cochrane Database Syst
Rev. 2002;(3):CD001283.
www.aafp.org/afp American Family Physician 349
Annals of Internal Medicine
Cholesterol Control Beyond the Clinic: New York Citys Trans Fat
Restriction
Sonia Y. Angell, MD, MPH; Lynn Dee Silver, MD, MPH; Gail P. Goldstein, MPH; Christine M. Johnson, MBA; Deborah R. Deitcher, MPH;
Thomas R. Frieden, MD, MPH; and Mary T. Bassett, MD, MPH
Decades after key modifiable risk factors were identified, cardiovas-
cular disease remains the leading cause of preventable death, and
only one quarter of persons with high cholesterol levels have at-
tained recommended levels of control. Cholesterol control efforts
have focused on consumer education and medical treatment. A
powerful, complementary approach is to change the makeup of
food, a route the New York City Department of Health and Mental
Hygiene took when it restricted artificial trans fata contributor to
coronary heart diseasein restaurants. The Department first under-
took a voluntary campaign, but this effort did not decrease the
proportion of restaurants that used artificial trans fat. In December
2006, the Board of Health required that artificial trans fat be
phased out of restaurant food. To support implementation, the
Department provided technical assistance to restaurants. By No-
A lthough blood cholesterol levels in the United States
have decreased slightly since the late 1980s (1), con-
sumption of saturated and trans fats remains above recom-
mended maximum levels (2, 3), and only one quarter of
U.S. adults with high cholesterol have it under control (4).
Some 24 million U.S. adults report taking cholesterol-
lowering medications (5). Changing a persons nutrition-
related behavior to improve cholesterol control is challeng-
ing (6) and a source of frustration for clinicians (7). Public
health action can complement health care provider efforts.
The New York City Department of Health and Men-
tal Hygiene adopted such a strategy when it restricted the
use of artificial trans fat in restaurant food. We describe the
rationale and process that led to this New York City
Health Code amendment and the Departments experience
implementing the rule.
Artificial trans fat is an unnecessary ingredient that
poses a substantial risk to heart health. Although no ran-
domized, controlled trials have assessed the effect of trans
fat consumption on cardiovascular events, evidence from
experimental studies, dietary trials, and prospective obser-
vational studies (3, 8) demonstrates that consumption of
trans fatty acids provides no apparent nutritional benefit
and has considerable potential for harm. The Institute of
Medicine (3) concluded that a positive linear trend exists
between trans fatty acid intake and low-density lipoprotein
cholesterol level, which increases the risk for coronary heart
disease. Unlike saturated fat, trans fat also decreases high-
density lipoprotein cholesterol levels (3, 8, 9). Trans fat
intake has a large effect on cardiovascular health: A mere
2% increase in energy intake from trans fat may increase
the risk for a coronary event by up to 23% (8). Other
potential adverse effects have been noted, such as affecting
insulin sensitivity and increasing systemic inflammation,
Perspective
vember 2008, the restriction was in full effect in all New York City
restaurants and estimated restaurant use of artificial trans fat for
frying, baking, or cooking or in spreads had decreased from 50%
to less than 2%. Preliminary analyses suggest that replacement of
artificial trans fat has resulted in products with more healthful fatty
acid profiles. For example, in major restaurant chains, total satu-
rated fat plus trans fat in French fries decreased by more than
50%. At 2 years, dozens of national chains had removed artificial
trans fat, and 13 jurisdictions, including California, had adopted
similar laws. Public health efforts that change food content to make
default choices healthier enable consumers to more successfully
meet dietary recommendations and reduce their cardiovascular risk.
Ann Intern Med. 2009;151:129-134. www.annals.org
For author affiliations, see end of text.
which raises additional public health concerns (8, 10).
Each year, 6% to 19% of coronary heart disease events and
30 000 or more premature deaths are estimated to occur in
the United States because of trans fat consumption (8, 11).
The Institute of Medicine, the American Heart Associa-
tion, the U.S. Department of Agriculture, and other lead-
ing health organizations recommend minimizing trans fat
intake (3, 1215).
To achieve this, health care providers are asked to
counsel patients (12). However, such advice is unlikely to
be broadly effectivemillions of Americans would need to
routinely study package labels on groceries and actively
seek information on trans fat in restaurant foods, where a
growing proportion of calories are now consumed. Health
would be better served by making the use of safer fats the
default in food preparation.
THE EMERGENCE OF ARTIFICIAL TRANS FAT AND
RECOGNITION OF ITS RISKS
Trans fats occur naturally in small amounts in some
meat and dairy products (3); however, these sources make
up only 21% of total trans fat intake (16). The remaining
trans fat in the U.S. diet is artificial, produced by partial
See also:
Print
Editorial comment. . . . . . . . . . . . . . . . . . . . . . . . . . 137
Web-Only
CME quiz
Conversion of graphics into slides
2009 American College of Physicians 129
Perspective Cholesterol Control Beyond the Clinic
hydrogenation of vegetable oils and found mostly in pro-
cessed foods especially baked and fried goods and
spreads. Partial hydrogenation was patented in 1903. Eight
years later, Crisco (J.M. Smucker, Orrville, Ohio) became
the first such product widely marketed in the United States
(17). These chemically modified oils, like several food in-
gredients in the 20th century, entered the food supply
without a full evaluation of human health effects and were
generally recognized as safe on the basis of experience
(18). Industrial advantages, such as extended product shelf
life and fry oil stability, led to increased use in the 1950s.
In the 1970s, margarine made from partially hydrogenated
oil was promoted as a healthier alternative to butter (19).
However, by the 1990s, studies began to identify trans fat
as a health hazard (9).
In 2003, the U.S. Food and Drug Administration
(FDA) required that all packaged goods display informa-
tion on trans fat content by 2006 (20). The regulation
allows products that contain less than 0.5 grams of trans fat
per serving to be labeled zero grams trans fat. This
prompted a widespread industry response of reformulating
products to declare zero grams trans fat on front-of-
package labeling (2123), although such products may still
contain trans fat in small amounts, which can add up.
The nutrition facts labeling allowed attentive consum-
ers to avoid trans fat in packaged foods; however,
restaurantswhich provide one third of daily caloric in-
take (24) generally do not disclose nutritional informa-
tion, leaving consumers no practical way to avoid trans fat.
At the time of the FDA action, only Denmark limited
trans fat in its entire food supply (25).
Although local health departments do not commonly
regulate packaged goods, inspecting, licensing, and regulat-
ing restaurants and retail food outlets are core local public
health functions (26). Building on its food safety infra-
structure, the Department trained restaurant inspectors in
May 2005 to assess the presence of artificial trans fat in
products used for frying, baking, or cooking or in spreads
by means of food label inspection. Because inspection visits
are scheduled each year on the basis of original restaurant
licensing date and compliance records, a survey fielded
during regularly scheduled inspections approximates a ran-
dom sample of these establishments. In May and June
2005, inspectors assessed the use of trans fat in oils, short-
enings, and spreads used for frying, baking, or cooking or
in spreads in 529 restaurants. Of the 478 restaurants that
used fats for any of these purposes and for which complete
data were collected, trans fat presence could not be deter-
mined in one third. Where it could be assessed, 50% of
restaurants used artificial trans fat to prepare food.
EDUCATIONAL CAMPAIGN
In August 2005, the Department launched an educa-
tional campaign to reduce restaurant artificial trans fat use.
Information was sent to the 30 000 licensed food outlets in
130 21 July 2009 Annals of Internal Medicine Volume 151 Number 2
the city, 15 000 suppliers and supermarkets, and hundreds
of thousands of consumers. Mailings urged restaurants to
remove artificial trans fat from food, suppliers to promote
0-grams trans fat products, and patrons to inquire about
oils used. Press coverage was extensive (2729). More than
7800 restaurant operators received training on the issue
during the mandatory food protection course.
Nearly a year later, inspectors surveyed 1021 restau-
rants in April and May 2006 by following the same meth-
odology as the previous year. Despite the educational cam-
paign, prevalence of artificial trans fat use remained
virtually unchanged at 51% in places where it could be
assessed. Given survey findings of a wide range of restau-
rants that cooked without artificial trans fat, successful re-
formulation of packaged foods, and the Danish experience,
the Department concluded that food choice and quality
would be unaffected by replacing artificial trans fat and
proposed using its regulatory authority to reduce artificial
trans fat use (26).
CRAFTING REGULATION
Department clinical staff, food safety experts, and law-
yers considered a range of regulatory options, from
disclosure-only requirements to content restrictions. Label-
ing trans fat would benefit only consumers who were aware
of and would act on the dangers of trans fat and would be
difficult to enforce. A restriction, however, would benefit
all New Yorkers, regardless of language, literacy, level of
health awareness, or age. Protecting children, who con-
sume French fries and other quick-service foods that are
common sources of artificial trans fat (30), was particularly
important. Because artificial trans fat is both harmful and
fully replaceable, allowing continued use, even with disclo-
sure, could not be justified. Similarly, restricting trans fat
in cooking oils and spreads onlyalthough easier to
achievewould be insufficient because baked goods are
the largest dietary source of trans fat (16). Given that an
estimated one third of calories comes from away-from-
home sources, removing most trans fat from restaurants
could be expected to reduce trans fatassociated deaths
proportionately. The Department drafted a broad proposal
to restrict artificial trans fat use in fry oils; spreads; and all
other ingredients and products, including baked goods.
For the regulation to be clear and enforceable, the
Department adopted the FDA threshold used for labeling
of up to 0.5 grams of trans fat per serving for products
containing any trans fat (20). The regulation would apply
to all licensed food establishments, including restaurants,
school cafeterias, caterers, senior centers, and street-food
vendors. It intentionally did not affect products that con-
tained only natural trans fat.
The Department designed enforcement to rely on ex-
isting infrastructure. Inspectors check product labels,
which restaurants now must maintain, during their regular
annual inspections. The rule included a 6-month phase-
www.annals.org
 Cholesterol Control Beyond the Clinic Perspective

Table 1. Resources Created to Help Restaurants Comply With the Artificial Trans Fat Restriction

Resources Use Period

Training module in Health Department food protection courses 48 144 food service employees trained January 2006December 2008
Health Department informational brochures 87 800 brochures distributed March 2007December 2008
Trans Fat Help Center
Telephone helpline 502 calls received March 2007December 2008
Brochures on frying, baking, and purchasing prepared foods 130 000 brochures mailed March 2007December 2008
Trans Fat 101 classes in English, Spanish, and Chinese 743 food service employees trained July 2007December 2008
Web site content (www.notransfatnyc.org)* 52 218 Web hits July 2007December 2008

* Includes lists of replacement products classified by saturated fat content; general price lists for 0-grams trans fat products; information on how to comply with the citys
regulation; pages on frying, baking, and purchasing prepared products with 0-grams trans fat; and a baking troubleshooter guide for working with 0-grams trans fat products.
The site is offered in English, Spanish, and Chinese.

out period for artificial trans fat in fry oils and spreads and
18 months for all other items, to give restaurants time to
identify alternatives and reformulate (31). Fines for viola-
tions range from $200 to $2000.
FROM CONCEPT TO ACTION
The regulation was proposed as an amendment to the
citys Health Code, approvable by the Board of Health.
Nearly 2300 public comments were received, with sup-
porters outnumbering opponents 31 to 1. Local, regional,
and national organizations and professional medical asso-
ciations made statements of support.
Opposition came primarily from industry associations,
including the National Restaurant Association and its state
affiliate and the Grocery Manufacturers Association (32).
The restaurant industry, which often opposes regulation on
principle, alleged that the elimination of artificial trans fat
would result in more expensive and less flavorful food and
also claimed that the supply of alternative oils was insuffi-
cient and the regulatory timeline unreasonable. Some
media derided the proposal as nanny state meddling (33,
34).
In response to public comment, the Department mod-
ified the rule to allow a longer phase-out period for baked
RESTAURANT RESPONSE
By June 2008, on the basis of inspection reports, 99%
of all restaurants had successfully adapted to the fry oil and
spread restriction. The second phase, which covered all
other products, became effective on 1 July 2008. Nearly 6
months later, adherence was 92%, with most violations
due to missing documentation. When November 2008 ad-
herence figures are adjusted to provide approximate com-
parison with preregulation survey denominators, an esti-
mated 98% of restaurants were not using artificial trans fat
in oils, shortenings, and spreads, compared with 50% in
2005 (Figure).
As implementation proceeded, the Department fol-
lowed the use of saturated fat in replacement oils. A num-
ber of chains announced early that replacement of trans
fat containing fry oils would also reduce saturated fat by
20% to 35% (35) (Colon R. Personal communication, 14
June 2007). Leading fast food chains decreased the satu-
Figure. New York City restriction of artificial trans fat:
reduction in use in frying, baking, or cooking or in spreads.
Containing Oils, Shortenings, or Spreads, %

goods and fried-dough items, such as doughnuts, to ad-

Restaurants That Used Artificial Trans Fat

dress technical challenges. Although some raised concerns 100

90 Health Code
about increased saturated fat use, the Department noted 80 Restriction
that, gram for gram, trans fat had a more adverse effect on 70 Education Passed Phase 1
Campaign Effective*
the lipid profile than saturated fat (8, 9); it was unlikely 60
50 Phase 2
that most trans fat would be replaced with saturated fat 40 50% 51% Effective*
43%
(32, 35, 36); and the Department would encourage substi- 30
Survey 1 Survey 2
tution with unsaturated fats. The Board of Health ap- 20 Survey 3

proved the modified Health Code amendment in Decem- 10

Inspection Compliance
1.9% 1.6%
0 . .
ber 2006, thereby instituting the first legal restriction of June December June December June December June December
artificial trans fat in the United States. A third survey 2005 2005 2006 2006 2007 2007 2008 2008

among 996 establishments was performed in May to June

2007, after passage but before the laws effective date. In
establishments where trans fat content could be deter-
mined, use had decreased to 43%.
To help restaurants comply with the regulation and
switch to the healthiest alternatives, the Department created a
range of resources (Table 1). Although these were well-
received, most establishments found their own solutions.
www.annals.org
Data limited to use of artificial trans fat in oils, shortenings, and spreads
(for which trans fat content could be determined) in New York City
licensed food service establishments by users of these products. Data
from 2005 to 2007 are from surveys. July and November 2008 data are
based on restaurant compliance data collected during regularly scheduled
inspections. Compliance data are further adjusted to be consistent with
survey denominator.
* Phase 1 of the regulation only covered fats used for frying or as a
spread. Phase 2 covered all other foods and ingredients.
21 July 2009 Annals of Internal Medicine Volume 151 Number 2 131
Perspective Cholesterol Control Beyond the Clinic

Table 2. Trans and Saturated Fat Content of French Fries Sold in Major U.S. Fast Food Chains*

French Fries Fatty Acid Content

Before Effective Date After Effective Date Change

Saturated Trans Saturated ! Saturated Trans Saturated ! Saturated Trans Saturated !

Fat, g Fat, g Trans Fat, g Fat, g Fat, g Trans Fat, g Fat, g Fat, g Trans Fat, g
Wendys
Kids Meal (small) 1.5 0.0 1.5 1.5 0.0 1.5 0.0 0.0 0.0
Medium (small) 2.5 1.0 3.5 2.5 0.0 2.5 0.0 !1.0 !1.0
Biggie (medium) 3.0 1.0 4.0 3.0 0.0 3.0 0.0 !1.0 !1.0
Great Biggie (large) 4.0 1.0 5.0 4.0 0.0 4.0 0.0 !1.0 !1.0

McDonalds
Small 2.5 3.5 6.0 1.5 0.0 1.5 !1.0 !3.5 !4.5
Medium 4.0 5.0 9.0 2.5 0.0 2.5 !1.5 !5.0 !6.5
Large 6.0 8.0 14.0 3.5 0.0 3.5 !2.5 !8.0 !10.5

Arbys
Small 4.0 3.0 7.0 4.0 0.0 4.0 0.0 !3.0 !3.0
Medium 4.0 4.0 8.0 4.0 0.0 4.0 0.0 !4.0 !4.0
Large 7.0 5.0 13.0 7.0 1.0 8.0 0.0 !4.0 !4.0

White Castle
Regular, 4.12 oz 3.0 5.0 8.0 3.0 0.0 3.0 0.0 !5.0 !5.0
Sack, 8.6 oz 6.0 11.0 17.0 6.0 0.0 6.0 0.0 !11.0 !11.0

Mean grams 4.0 4.0 7.9 3.5 0.1 3.6 !0.4 !3.9 !4.3
Mean gram change, % !10.5 !97.9 !54.2

* Analysis restricted to French fries to illustrate the effect of the first phase of the trans fat regulation, which covered fry oils and spreads. The second phase covered all other
products and went into effect on 1 July 2008.
Values are based on nutrition information published online at each chains Web site. Access dates: Wendys, 1 December 2006 (before) and 1 April 2008 (after);
McDonalds, 25 January 2007 (before) and 22 May 2008 (after); Arbys, 6 February 2007 (before) and 23 May 2008 (after); and White Castle, 20 February 2007 (before)
and 25 August 2008 (after).

rated fat in their French fries by 10.5% and the total trans
fat plus saturated fat by 54%, on average (Table 2). Recent
experience in Canada (21) also suggests that saturated fat
content can remain the same or even be reduced when
trans fat is reduced. In New York City, the evaluation of
replacement products and changes in fatty acid composi-
tion are under way, and surveillance of cardiovascular dis-
ease is ongoing. However, as with other population inter-
ventions, isolating the specific effect of this regulation on
low-density lipoprotein cholesterol levels and cardiovascu-
lar health is challenging.
Two years after New York Citys action, 12 U.S. local
governments and 1 state had adopted similar measures.
Following the approval of the Health Code amendment,
more than 50 restaurant chains announced or reiterated
their intention to discontinue use of artificial trans fat
containing products nationally, and most have already
done so (37). Dozens of hotel groups, food manufacturers,
grocery retailers, and other companiesfrom Disney and
Marriott to American Airlines and Kraft Foods have also
announced elimination of artificial trans fat.
REFLECTIONS ON THE EXPERIENCE
To develop and implement the trans fat regulation,
Department staff needed to communicate effectively about
132 21 July 2009 Annals of Internal Medicine Volume 151 Number 2
this complicated topic, harness political and community
support for a regulatory approach, and navigate industry
opposition while addressing legitimate industry concerns.
Terms with which the public is unfamiliar, such as partially
hydrogenated vegetable oil and trans fatty acids, were aban-
doned in favor of the more accessible term artificial trans
fat. Although the voluntary initiative did not result in mea-
surable changes, it generated public and industry awareness
and demonstrated that a voluntary effort was insufficient to
eliminate the health risk. The restaurant industry was sup-
portive of the voluntary initiative; however, it organized in
opposition when the regulation was proposed (32) even
though the Health Code amendment was clearly within
the legal purview of the Board of Health and went unchal-
lenged in court. To address replacement product availabil-
ity and cost concerns and to reduce the risk for price goug-
ing at the retail level, the Department identified food
suppliers and manufacturers and tracked availability and
price of these products. Restaurants that switched to newer
fry oils with 0 grams of trans fat reported that the change
was cost neutral (38). Supply concerns proved unfounded
because major oil producers expanded production of oils
with 0 grams of trans fat (22, 39, 40) and ingredients for
baked goods with reduced levels of saturated fat (41). Al-
though evaluation continues, no negative economic effects
have been observed to date.
www.annals.org

PUBLIC HEALTH ACTION CHANGES THE DEFAULT
Reducing risk factors for cardiovascular disease is dif-
ficult for patients and physicians. Artificial trans fata
recognized contributor to coronary heart disease riskis
widespread in our food supply and, at the time of the
Departments action, was virtually impossible for the aver-
age consumer to avoid. An effective way to reduce this risk
was to change the food supply through regulation. None-
theless, proposing regulation as a tool to protect health
prompts industry opposition and complaints about govern-
ment intrusion. This opposition can often delay important
policy actions and lead to preventable disease. For example,
it took decades to achieve a national prohibition of lead
paint, despite unequivocal evidence for harm. New York
Citys Board of Health banned the product in 1960, 18
years before federal action.
Throughout the Health Code amendment process, de-
tractors likened the city to a nanny state (33, 34), as they
had with such previously controversial measures as seat belt
requirements, water fluoridation, and tobacco control. Just
as the debates surrounding those historic public health in-
terventions calmed after they were implemented, trans fat
restriction is now a largely unnoticed part of New York
City life. By the time the second phase went into effect,
press coverage was light-heartednoting that popular
foods, from cannoli to knishes, were still available (42).
Seventy-nine percent of trans fat in the diet is artifi-
cially produced (16). It is present only because we fail to
prevent industry from using it, even as we recommend that
consumers avoid it. Removing trans fat from the food sup-
ply will improve the lipid profiles of millions of persons
without requiring complex behavioral efforts and may re-
duce the need for medication. Although regulatory ap-
proaches are not appropriate for all health risks, our society
could improve health if available food choices more closely
resembled those recommended for heart health (43).
From the New York City Department of Health and Mental Hygiene,
New York, New York.
Acknowledgment: The authors thank Elliott Marcus, MSSW, and Rob-
ert Edman, BSc, of the Departments Bureau of Food Safety and Com-
munity Sanitation, as well as all the Bureaus directors and sanitarians for
implementation of the work described here. They also thank Wilfredo
Lopez, JD; Thomas Merrill, JD; Martha Robinson, JD; and Anna Caf-
farelli, MHS; the members of the New York City Board of Health;
Walter Willett, MD, MPH, DrPH; Colin McCord, MD; and the other
staff and interns of the Department whose work was critical to the suc-
cess of this initiative.
Grant Support: By the City of New York. The American Heart As-
sociation provided an educational grant to support the Trans Fat
Help Center.
Potential Financial Conflicts of Interest: None disclosed.
Requests for Single Reprints: Lynn Dee Silver, MD, MPH, New York
City Health Department, 2 Lafayette Street, 20th Floor, CN-46, New
York, NY 10007; e-mail, lsilver@health.nyc.gov.
www.annals.org
Cholesterol Control Beyond the Clinic Perspective
Current author addresses are available at www.annals.org.
References
1. Carroll MD, Lacher DA, Sorlie PD, Cleeman JI, Gordon DJ, Wolz M, et al.
Trends in serum lipids and lipoproteins of adults, 1960-2002. JAMA. 2005;294:
1773-81. [PMID: 16219880]
2. U.S. Department of Health and Human Services. Nutrition and overweight.
In: Healthy People 2010 Midcourse Review. Washington, DC: U.S. Department
of Health and Human Services; 2005:1-25.
3. Panel on Macronutrients. Letter report on dietary reference intakes for trans
fatty acids. In: Institute of Medicine. Dietary Reference Intakes for Energy, Car-
bohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Wash-
ington, DC: Institute of Medicine; 2002.
4. Hyre AD, Muntner P, Menke A, Raggi P, He J. Trends in ATP-III-defined
high blood cholesterol prevalence, awareness, treatment and control among U.S.
adults. Ann Epidemiol. 2007;17:548-55. [PMID: 17395483]
5. Mann D, Reynolds K, Smith D, Muntner P. Trends in statin use and
low-density lipoprotein cholesterol levels among U.S. adults: impact of the 2001
National Cholesterol Education Program guidelines. Ann Pharmacother. 2008;
42:1208-15. [PMID: 18648016]
6. Brunner EJ, Rees K, Ward K, Burke M, Thorogood M. Dietary advice for
reducing cardiovascular risk. Cochrane Database Syst Rev. 2007:CD002128.
[PMID: 17943768]
7. Beran MS, Fowles JB, Kind EA, Craft CE. State of the art reviews: patient
and physician communication about weight management: can we close the gap?
American Journal of Lifestyle Medicine 2008;2:75-83.
8. Mozaffarian D, Katan MB, Ascherio A, Stampfer MJ, Willett WC. Trans
fatty acids and cardiovascular disease. N Engl J Med. 2006;354:1601-13. [PMID:
16611951]
9. Ascherio A, Katan MB, Zock PL, Stampfer MJ, Willett WC. Trans fatty
acids and coronary heart disease. N Engl J Med. 1999;340:1994-8. [PMID:
10379026]
10. Lopez-Garcia E, Schulze MB, Meigs JB, Manson JE, Rifai N, Stampfer MJ,
et al. Consumption of trans fatty acids is related to plasma biomarkers of inflam-
mation and endothelial dysfunction. J Nutr. 2005;135:562-6. [PMID:
15735094]
11. Willett WC, Skerrett PJ, Giovannucci EL, Callahan M. Eat, Drink, and Be
Healthy: The Harvard Medical School Guide to Healthy Eating. New York:
Simon and Schuster; 2001.
12. Lichtenstein AH, Appel LJ, Brands M, Carnethon M, Daniels S, Franch
HA, et al; American Heart Association Nutrition Committee. Diet and lifestyle
recommendations revision 2006: a scientific statement from the American Heart
Association Nutrition Committee. Circulation. 2006;114:82-96. [PMID:
16785338]
13. U.S. Department of Health and Human Services, U.S. Department of
Agriculture. Dietary Guidelines for Americans. Washington, DC: U.S. Depart-
ment of Health and Human Services; 2005.
14. American Medical Association House of Delegates. Reducing Trans Fat
Content of Americans Diet. Chicago: American Med Assoc; 2006.
15. World Health Organization, Food and Agriculture Organization of the
United Nations. Joint WHO/FAO Expert Consultation on diet, nutrition and
the prevention of chronic diseases. Section Report of a joint WHO/FAO expert
consultation. Geneva, Switzerland: World Health Organization; 2003.
16. Revealing trans fats. FDA Consum. 2003;37:20-6. [PMID: 14666902]
17. Allen RR. Hydrogenation. In: Hui YH, ed. Baileys Industrial Oil Fat Prod-
ucts, Edible Oil and Fat Products: Processing Technology. 5th ed. New York: J
Wiley; 1996:213-6.
18. U.S. Food and Drug Administration. History of the GRAS list and SCOGS
Review. Silver Spring, MD: U.S. Food and Drug Administration; 2006. Accessed at
www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafeGRAS
/GRASSubstancesSCOGSDatabase/ucm084142.htm on 2 June 2009.
19. Chrysam MM. Margarine and spreads. In: Hui YH, ed. Baileys Industrial
Oil and Fat Products, Edible Oil and Fat Products: Products and Application
Technology. 5th ed. New York: J Wiley; 1996:65-114.
20. U.S. Food and Drug Administration. Food Labeling: Trans Fatty Acids in
Nutrition Labeling, Nutrient Content Claims, and Health Claims and Possible
Footnote or Disclosure Statements. Federal Register. Washington, DC: U.S. De-
partment of Health and Human Services; 11 July 2003:41434-506.
21 July 2009 Annals of Internal Medicine Volume 151 Number 2 133
 Nutrition & Dietetics 2012; 69: 256259 DOI: 10.1111/j.1747-0080.2012.01600.x

INSIGHT

Should trans fats be regulated? ndi_1600 256..259

Bill SHRAPNEL
Shrapnel Nutrition Consulting Pty Ltd, Sydney, New South Wales, Australia

Abstract
Aim: The aim of this paper is to critically assess recent calls for increased regulation to lower the level of trans fats
in the Australian diet.
Methods: Key milestones in the elucidation of the effects on trans fats on health were identified and reviewed.
Trends in intakes of trans fats in Australia and factors affecting those trends were described and compared with
those in Denmark, which has regulated to lower population intake of trans fats.
Results: The scientific evidence demonstrating adverse effects of trans fats on human health is consistent and
strong. Australian health authorities were quick to identify the potential risk of trans fats and communicate it to
health professionals and the food industry. The response from the margarine industry resulted in large falls in the
trans fat content of the Australian diet in the mid-1990s. A second wave of trans fat reduction across many foods
categories has occurred subsequently. Total intake of trans fats in Australia is now low, half the upper limit
recommended by the World Health Organization and lower than the intake in Denmark. Trans fats of industrial origin
comprise just one-eighth of 1% of dietary energy. These falls in trans fat intake mirror the large falls that occurred
in Denmark prior to regulation.
Conclusions: The case for increased regulation to lower intake of trans fats in Australia cannot be sustained. The
trans fat issue stands as a good example of self-regulation through collaboration between Australian health
agencies, the food industry and the government.

Key words: food industry, health, regulation, trans-fatty acid.

Introduction
In recent years, there have been calls from Australian
politicians and former politicians for a ban on trans-fatty
acids (trans fats),1,2 echoing similar calls overseas that have
resulted in changes to food regulations in Denmark3 and the
USA.4,5 Is there a sound case for regulatory intervention to
drive down the level of trans fats in the Australian diet or
does current progress obviate the need for regulation?
Unsaturated fatty acids comprise approximately half of
the fatty acids in the typical adult Australian diet6 and their
carboncarbon double bonds are generally in the cis con-
figuration. This conformation ensures that unsaturated fats
have lower melting points than fats dominated by saturated
fatty acids. In a small proportion of unsaturated fatty acids,
the carboncarbon double bonds are in the trans configura-
tion and the physical properties of these trans fats are similar
to those of saturated fats.7
B. Shrapnel, MHP, APD, Director
Correspondence: B. Shrapnel, Shrapnel Nutrition Consulting Pty
Ltd, 10 Blackwood Close, Beecroft, Sydney, NSW 2119, Australia.
Email: shrapnelnc@bigpond.com
Industry links: Bill Shrapnel consults to the food industry and current
clients include Goodman Fielder.
Accepted September 2011
There are two sources of trans fats in the human diet,
industrial and ruminant, and both are the result of a process
called partial hydrogenation.7 This process can be applied
industrially to harden a vegetable oil, that is, raise the
melting point of an unsaturated oil by generating increased
proportions of saturated and trans fats. Until the mid-1990s,
partial hydrogenation was used by the margarine industry in
Australia to generate a hard fraction, which when mixed
with liquid vegetable oils yielded a solid product suitable
for spreading. Hardened fats were also used in the baking
industry where fats with comparable physical properties to
butter were required. Trans fats are also generated in the
gastrointestinal tracts of ruminant animals, such as cattle,
sheep and goats.8 Once ingested, unsaturated fats from grass
and seeds are retained in the rumen of the animal where they
undergo partial hydrogenation, increasing the proportion of
saturated and trans fats. Hence, the meat and milk of rumi-
nant animals contain trans fats, which comprise about 35%
of total fatty acids in these foods.
In the popular media, it has become common to refer to
industrially generated trans fats as artificial and, less com-
monly, for ruminant trans fats to be described as natural.
This nomenclature is misleading as many 18:1 trans isomers
are common to both industrial and ruminant trans fats.7
For example, elaidic acid, with the trans bond in the 9 posi-
tion, which is prevalent in industrially hardened fats and

256 2012 The Author

Nutrition & Dietetics 2012 Dietitians Association of Australia
bs_bs_banner

sometimes singled out as a focus of concern, is also present
in ruminant fats. Vaccenic acid, with the trans bond in the 11
position, which is the most common ruminant trans isomer,
is also present in industrially hardened fats.7 Industrial and
ruminant trans fats largely represent different combinations
of the same molecules.
Methods
In relation to the health effects of trans fats, a computerised
Medline search was conducted of English language papers
published after 1985, using the following search terms: trans
fatty acids, serum lipids, coronary heart disease and cardio-
vascular disease. Key milestones in the elucidation of the
effects on trans fats on health were identified and briefly
reviewed. Data on trans fat levels in the Australian diet
were accessed through the Food Standards Australia New
Zealand (FSANZ) website and through manual literature
searches. Information about the chronology of changes to
the criteria for the Tick programme was gathered from the
Heart Foundation via personal communications. Details
about the extent and timing of industry initiatives to lower
the trans-fatty acid content of products were gathered from
food companies through personal communications. Data on
trans fat levels in the Danish diet were gathered via comput-
erised literature searches using the names of known Danish
researchers into trans fats.
Results and Discussion
Investigations of the health effects of trans fats in the 1980s
found that their effect on total blood cholesterol was almost
neutral, similar to that of monounsaturated fats, and there
appeared to be little cause for concern.9 As understanding of
blood lipid-related coronary heart disease risk became more
sophisticated, the focus shifted from total blood cholesterol
to the different lipoprotein fractions, in particular the detri-
mental LDL-cholesterol (LDL-C) and the protective HDL-
cholesterol (HDL-C). When the health effects of trans fats
were assessed through this prism, their unique double-
negative effect on blood lipids was revealedtrans fats
raised the LDL-C concentration and lowered HDL-C.10 In
their 2003 meta-analysis, Mensink and colleagues found that
trans fats raised the total cholesterol to HDL-C ratio signifi-
cantly more than saturated fat, suggesting trans fats con-
ferred higher coronary heart disease risk.11 Adverse effects on
this ratio have been observed with both industrial and rumi-
nant trans fats, as well as conjugated linoleic acida poly-
unsaturated trans-fatty acid found predominantly in dairy
fat.12 The possibility that dietary trans fats may pose greater
coronary heart disease risk than saturated fat was supported
by the findings of prospective cohort studies.13,14 The mag-
nitude of coronary heart disease risk associated with dietary
trans fats in these studies suggests that additional mecha-
nisms may also be at play, for example, adverse effects on
inflammation, insulin resistance and blood vessel function.15
Nutrition scientists now generally agree that dietary trans fats
pose a greater coronary heart disease risk than saturated fat
and other macronutrients.
2012 The Author
Nutrition & Dietetics 2012 Dietitians Association of Australia
Should trans fats be regulated?
There have been three estimations of the content of trans
fats in the Australian diet. In 1994, Noakes and Nestel
estimated that trans fats contributed 2.02.5% of the dietary
energy of Australian adults.16 Two estimates of trans fats in
the US diet in the early 1990s indicated that trans fat intake
in that country was twice17 or three times18 higher than that
in Australia. The high intakes of trans fats in the USA at the
time could be attributed to the widespread use of partially
hydrogenated soybean oil as a frying medium. In contrast,
the major frying fats used in Australia were tallow (beef and
lamb fat), which is low in trans fats, and palm oil, which is
virtually free of trans fats. Noakes and Nestel estimated that
approximately 60% of trans fats in the Australian diet were of
industrial origin, with domestic margarines being by far the
major source.16 FSANZ estimated the trans fat content of
the Australian and New Zealand diets in 2007 and 2009 in
response to a request from The Australia and New Zealand
Food Regulation Ministerial Council.19,20 In the more recent
study,20 mean trans fat intake in Australia was estimated to be
0.5% of dietary energy, half the upper limit recommended by
the World Health Organization.21 The fall in trans fat intake
in Australia over the preceding 15 years appears to have
been significant and, based on the differences between the
two FSANZ studies, was still underway in the late 2000s.
Falls in intakes of both industrial and ruminant trans fats
have occurred since the early 1990s, the fall in ruminant
trans fats most likely being due to lower use of tallow and
increased use of palm oil as a frying medium and in baked
goods. The fall in intake of industrial trans fats has been
significantly greater as the use of partially hydrogenated fats
has been phased down or out in many product categories.
Industrial trans fats now comprise just one-eighth of 1% of
dietary energy. Ruminant sources contribute 73% of total
trans fat in the current Australian diet.20
Two factors appear to have been instrumental in the
appreciable falls in the trans fat content of the Australian
dietscientific leadership and the response from the food
industry. Scientific leadership on the issue of trans fats has
come from two quarters. In 1992, just two years after the
adverse effects of trans fats on serum lipoproteins were
first observed,10 the National Health and Medical Research
Council released a report recommending that trans fats be
considered as equivalent to saturated fats in relation to their
effect on blood cholesterol.22 Two years later, the National
Heart Foundation of Australia argued a similar case in a
review of dietary fats and cholesterol and forecast likely
changes to relevant criteria for the Foundations Tick
programme.23
After extensive consultation with the margarine industry,
the Heart Foundation introduced a new Tick programme
criterion for table margarines in early 1996, the previous
limit on saturated fats being replaced by a limit on the sum
of saturated and trans fats. In that year, one of Australias
leading margarine manufacturers removed trans fats from all
its table spreads as part of a global initiative by the parent
company24 and the other leading margarine company largely
followed in 1997, resulting in a considerable fall in the
amount of trans fat present in the Australian food supply.
257
B. Shrapnel
Insight into the effects of this change on individual intakes of
trans fats was gleaned from a serendipitous study conducted
at the Commonwealth Scientific and Industrial Organisation
in Adelaide.25 Adipose tissue biopsies of fatty acids were
collected from two groups of subjects between 1995 and
1997, which coincided with the first margarine companys
removal of trans fats from its products. In both groups of
subjects, adipose tissue levels of trans fats fell by 23% during
the study, suggesting a substantial fall in dietary intake of
trans fats. In 2005, the Heart Foundation again amended
its Tick programme criteria for margarines, stipulating that
trans fats should comprise 1% or less of total fatty acids. This
encouraged smaller players in the margarine industry to
adopt low trans fat formulations. There are now 37 catego-
ries of supermarket foods in the Heart Foundations Tick
programme that include criteria for trans fats (personal
communication).
Although the major fall in the trans fat content of the
Australian diet is likely to have occurred during the 1990s,
there have been many initiatives undertaken in various
sectors of the food industry during the last decade that have
served to drive trans fat intake down to its current low level.
At least two multinational companies operating in Australia
have participated in global company programmes to lower
trans fats in their product ranges;26,27 the leading chain of
quick service restaurants has adopted a frying oil low in both
saturated and trans fats;28 the leading supermarket chain in
Australia has included trans fats in the nutrition information
panel of its house brand products thereby highlighting
problem foods in its portfolio (personal communication);
and a major donut manufacturer that had been the focus of
criticism for the high trans fat content of its frying oil has
adopted a low trans fat formulation.29 The Commonwealth
Government played a role in facilitating the removal of
industrial trans fats from products available through quick
service restaurants, convening a roundtable in March 2007
and reporting good progress in September of that year30 and
more recently.31
Advocates of increased regulation of trans fats in Australia2
have highlighted the situation in Denmark, which intro-
duced a regulation imposing a limit on the content of trans
fats in food ingredients in 2003. Although there has been a
remarkable (85%) fall in the amount of trans fats in the
Danish diet since the late 1970s, it should be noted that by
far, the majority of this fall occurred before the year 2000,
that is, before the introduction of regulation.32 The factors
underpinning the Danish trend are very similar to those at
play in Australiascientific leadership from the Danish
Nutrition Council33 and a rapid response from industry,
including the removal of trans fats from table margarines,
which corresponded with the period of the steepest fall in
the trans fat content of the Danish diet between 1990 and
1995.32 The current trans fat content of the Danish diet is at
least 0.7% of dietary energy,32,34 which is slightly higher than
that in the Australian diet.19,20 By far, the majority of trans fats
in the Danish diet is ruminant in origin, as is the case in the
Australian diet. Thus, Denmark and Australia have arrived at
near identical positions in relation to trans fats, one employ-
258
Nutrition & Dietetics 2012 Dietitians Association of Australia
ing regulation and the other adopting non-regulatory mea-
sures. It can be concluded that regulation has played a very
minor role in the Danish efforts to lower trans fat intakes and
would have a negligible effect if enacted in Australia. Given
the successful strategies employed by the scientific commu-
nity, the food industry and government to lower trans fats in
the Australian diet to their current low level, the case for
regulatory intervention in relation to trans fats is lacking.
In summary, although there is consistent evidence that
dietary trans fats are associated with adverse effects on blood
lipids and increased risk for coronary heart disease, the case
for increased regulation in this area is weak. Australian
health authorities were quick to identify the potential risk
associated with trans fats and communicate it to health-care
professionals and the food industry. The timely response
from major margarine companies followed by a wider indus-
try response has seen levels of industrial trans fats in the
Australian diet fall to just one-eighth of 1% of dietary energy.
Total trans fat intakes in Australia are already lower than
in Denmark, which has regulated to lower the trans fat
content of its food supply. Rather than there being a good
case for increased regulation, the trans fat issue stands as
a good example of self-regulation though collaboration
between Australian health agencies, the food industry and
the government.
Acknowledgements
Preparation of this paper was partly supported by Goodman
Fielder.
References
1 ABC television, Stateline (NSW). Carrs Crusade. 28 August
2009.
2 Kaye J. The Greens. Greens call for NSW govt support for bill
to terminate trans fats. Media release, 27 July 2008.
3 Danish Veterinary and Food Administration. Executive Order
No. 160 pursuant to section 13, section 55(2) and section 78(3)
of Act No. 471 (the Danish Food Act). 11 March 2003.
4 Board of Health, New York City. The regulation to phase out
artificial trans fats in New York City food service establishments.
Section 81.08, New York City Health Code 2007.
5 State of California. Health and Safety Code. Part 7, Division
104, Chapter 12.6, Trans fats 2008.
6 Australian Bureau of Statistics. National Nutrition Survey.
Nutrient Intakes and Physical Measurements, Australia, 1995.
1998 Commonwealth of Australia. ABS Catalogue No. 4805.0.
7 Expert Panel on Trans Fatty Acids and Coronary Heart Disease
Risk, Kris-Etherton PM, ed. Physicochemical properties, intake,
and metabolism. Am J Clin Nutr 1995; 62: 659S69S.
8 Sommerfeld M. Trans unsaturated fatty acids in natural pro-
ducts and processed foods. Prog Lipid Res 1983; 22: 22133.
9 British Nutrition Foundation. Trans Fatty Acids. London: British
Nutrition Foundation, 1987.
10 Mensink RP, Katan MB. Effect of dietary trans fatty acids on
high-density and low-density lipoprotein cholesterol levels in
healthy subjects. N Engl J Med 1990; 323: 43945.
11 Mensink RP, Zock PL, Kester AD, Katan MB. Effects of dietary
fatty acids and carbohydrates on the ratio of serum total to HDL
2012 The Author



 ScW W ^g Wg ^3g_c
W[ g Wg q W v 7
h ^gd [ g 7_
-<W DgW 7^' [lgi
/&&1Wc/&&2Wgd^g
gc}gq Wc[gWWgc
^q W v 7}
WWjg qWqW^W_dWd
orggd[lWbg
gg gg } B
gWg /&&1(2g^gs7g
^W^c Wg gg {G 4_
q W v5 /&&2  x}g
^gWgc /) g^g% EWdd*.
/&&2 02g^g _c[}
d^ W
g W w
W^W}g W[g
^Wkc
0/g^g/&&1
 

 


 




 ;g^WvW^dWl^+y_
[Wgc g g} t g ^W[
:V ?T8LC?I M? VTNLL? R8KOJ?R KR P> I> ^W Wd g cg}gg q WW
[gq[dmW[
^W @W W^c W [g _ ^\
W 9 _ q v W=c
WgW }gd[g[c Eg_
W^,c[g[d!d}gW
Wg g Wg dg t g [c
^W}W[gcgvW^cW
EX vc}g WW_
g cg q g c[l [c"
n~p~ p uZ X`e ^
Fg ^W gg q W w
}ggWgg[gkt
WWgd
v QWWgc v W^] |d
WWv^W[gWc[W^
}gWg W_ c W glW
g 9 ^W W w }

WWgc sY W^c# Wg c W



ggWg$

U d ^g [g ^
^ggd W[ l Wg q W
v6AWvW^dcg
gW[ggz EWcd#
gdgfd}W/&&/d
[HdcWc^gW}g^acgW
WgqWv^gWg^^g
 && 9V




 

  \p
 


 


 g 

_d 
&=^



 
)

 "
)

K
 




l 


   
i

 iR
`











 

1
 



&2&( "  
 

z

%8
!88 !%*l%e*8*8e!!%* !hU%
! !%!%*%!! *h*






  d
 9
 -Wk

 

1

E 

;a



"2&  

 

  




9V E


 




  Q- )


  22




 "=^


  
 

 





 
&=&"2(


 

w

 









1

  
  
 
   


 

 


  Qu
a  

   XW  
z

:

R 0




    
 "
0
:    

  
 
(\

 
P
 
 

 
D
 
    

 
    

A 0


  
Kt

 

 




  
  T@



 
1 "
 




 

  OG


T@

 3b>3H>3 b>>3 IH3I: I: ]
  

L  
  
 
 
  H].#MFF.B7jJ4.'f#5.B#Mm#f$jCo+B$6+//





 


 B$ ,'5 $ {#C.5 5#+' J6#C?' ~rC

A 

O
}  

 

X +$+4/m'4,/ 5?n,'6#FoMc${74$' ?$7'



( [
 

  ,J6?,/s~s.,YZy7YZ4r/n$7y6#S|S



 `
 


NqNN q|Sc+

 
 






1

  
A


 

 

g 

( 
\  
 q
 ,VKbeK'KEe,@U:Pe

eSPC\e*`,e7 @cK-@WeK)`-K"e .+*fa +


 }
  
 
,
> k

!+
Mg[!

U
~e

)
v
"
)
 




]
 
V).

"]` .n


U 
  
 




  9

EP R* K de
 7`K@5e H'e Y+e>N-@e 77Q
 
Ex


 G

<<<


 :
th 
r
#
V
  9LB6e H'e 4e
 

AD
} p  O
46$ed Y' G
G '
-
 "
0
 
G= )  1  5
5v#

;
Q2 
 ;
p

# (1"(n 
9K@=e Ie Ze K1@e
 Q-





   B>!
$
C
 o 
K @

<; 

D   9 \ 5 WJ<   ` !9M8Ng2  P

 



D   ,,!
R_J,Qe 'EC
B 


  <

*
  "-,  u*+ m  -


       -   
 
bSL  @aeF;@e

  




 8M@;eH'e>,/@e & &
Ay p&

)


   




 //   m  4  ;1e-
Z
" @!aeGe7`KD<eH(e?-@e H  .Re
AD


0

u

(_ 0
  v 

  
t 
^ 
^ saTh i

l:




<
:i 
(
bKW







   

)
9K@=eH(e[eN2Ae%ePT0]e X 
T@

L ([


  
F&z %
' I

 

 
 O     
  982 /NT
O
)
6 k
[eK1Ae,&ePT0^e % %
$F D 
e
w

L 
 


    
Z# / f
3)
X  3P 

;   P-




  e J L8[  7`LB=e
 [eO1Ae  X#3e




 
0
 k

d46 ? I
 = Ho00
j-(w


 {

%
U
x Ux

R S $



%|


  ; je .l$
"

 ?=

@@ Y_3 #_cc
 Image description. Better Health Channel logo End of image description.

Fats and oils

Ima
ge
des
crip

Fatisimportantformanybodyprocesses.Youneedtoeatsomefatinyourdiet.Fatprotectsyour
organs,keepsyouwarmandhelpsyourbodyabsorbandmovenutrientsaround.Italsohelps
hormoneproduction.However,somefatsarebetterthanothersandhavingtoomuchofanytype
isnotagoodidea.

Dietaryfatsareclassifiedbytheirstructure.Differenttypesoffatsreactdifferentlyinsidethebody.
Saturatedfats(foundmostlyinanimalproducts)increasebloodcholesterol,whichisariskfactor
incoronaryheartdisease.Mono-unsaturatedandpolyunsaturatedfatstendtolowerblood
cholesterol.

Dietary fats and blood cholesterol

Thetwotypesofbloodcholesterolarelowdensitylipoprotein(LDL)cholesterolandhighdensity
lipoprotein(HDL)cholesterol.

LDLisconsideredthebadcholesterolbecauseitcontributestothenarrowingofthearteries,
whichcanleadtoheartdiseaseandstroke.HDLcholesterolisconsideredtobethegood
cholesterolbecauseitactuallycarriescholesterolfromthebloodbacktotheliver,whereitis
brokendown,reducingtheriskofcardiovasculardisease.

Groups of fats
Eachgroupoffatsbehavesdifferentlyinsidethebody.Dietaryfatcanbeclassifiedintofour
groups.Theseare:

saturated
mono-unsaturated
polyunsaturated
transfats.

Saturated, mono-unsaturated and polyunsaturated fats

Saturatedfatscontributetotheriskofheartdiseasebyraisingbloodcholesterollevels.Thesefats
arecommonlyfoundinmanytakeaway(fast)foods,incommercialproductssuchasbiscuitsand
pastries,andindairyproducts.

Mono-unsaturatedandpolyunsaturatedfatsbothtendtolowerbloodcholesterolwhentheyreplace
saturatedfatsinthediet.Polyunsaturatedfatshaveaslightlygreaterimpactthanmono-
unsaturatedfats.

Replacesaturatedfatsinyourdietwitheithermono-unsaturatedorpolyunsaturatedfatswhenever
possible.Forexample,replacebutterinsomecookeddisheswitholiveoilormargarine.

Trans fatty acids

Transfattyacidsarerareinnature.Theyareonlycreatedintherumenofcowsandsheep,and
arenaturallyfoundinsmallamountsinmilk,cheese,beefandlamb.Transfattyacidsarealso
createdduringthemanufactureofsometablemargarinesandinsolidspreadsusedinthefood
industrytomakebakedproductssuchaspies,pastries,cakes,biscuitsandbuns.

Transfattyacidstendtobehavelikesaturatedfatsinthebody,astheyraiseLDLlevelsand
increasetheriskofheartdisease.Unlikesaturatedfats,theytendtolowerHDLcholesterol,soare
potentiallyevenmoredamaging.

Fatsandoils Page1of4
 Itisthetransfatsthatareproducedduringfoodmanufacturingthatyoushouldbemostconcerned
about,notthetransfatspresentnaturallyincertainfoods.Lookformargarinesthathavelessthan
onepercenttransfatsonthelabelorchoosefoodswiththeHeartFoundationTick.Limithow
muchtakeawayfoodandpackagedsnackfoodsyoueat.

Common fat-containing foods

Differentfoodscontaindifferentratiosoffattyacids.Forexample:

Saturated fatssourcesincludefattycutsofmeat,full-fatmilk,cheese,butter,cream,
mostcommerciallybakedproductssuchasbiscuitsandpastries,mostdeep-friedfast
foods,coconutandpalmoil.
Mono-unsaturated fatssourcesincludemargarinespreadssuchascanolaoroliveoil
basedchoices,oilssuchasolive,canolaandpeanutoils,avocado,andnutssuchas
peanuts,hazelnuts,cashewsandalmonds.
Polyunsaturated fatssourcesincludefish,seafood,polyunsaturatedmargarines,
vegetableoilssuchassafflower,sunflower,cornorsoyoils,nutssuchaswalnutsand
Brazilnuts,andseeds.

Sources of omega-6 and omega-3 fats

Polyunsaturatedfatscanbedividedintotwocategories,whichare:

Omega-3 fatsarefoundinbothplantandmarinefoods,althoughitistheomega-3fats
frommarinesourcesthathavethestrongestevidenceforhealthbenefits(including
reducingtheriskofheartdisease).Plantfoodsourcesincludecanolaandsoyoils,and
canola-basedmargarines.Marinesourcesincludefish,especiallyoilyfishsuchasAtlantic
salmon,mackerel,Southernbluefintuna,trevallyandsardines.
Omega-6fatsarefoundprimarilyinnuts,seedsandplantoils,suchascorn,soyand
safflower.

Benefits of omega-3 fats

Researchisongoing,butthebenefitsofomega-3fatsinthedietseemtoincludethatthey:

lowertriglyceridelevels,whichareimportantriskfactorsincoronaryheartdisease
improvebloodvesselelasticity
keeptheheartrhythmbeatingnormally
thintheblood,whichmakesitlessstickyandlesslikelytoclot
reduceinflammationandsupporttheimmunesystem
reducebloodpressure
mayplayaroleinpreventingandtreatingdepression
contributetothenormaldevelopmentofthefetalbrain.

Plant sterols
Plantsterolsarepresentinallplants.Intakesof23gofplantsterolsperdayhavebeenshownto
reducebloodcholesterollevelsbyanaverageof10percent.Thisisbecausetheyblockthebodys
abilitytoabsorbcholesterol,whichleadstoareducedlevelofcholesterolintheblood.

However,itishardtoeatthisamountofplantsterolsfromnaturalsources,sotherearenowplant
sterol-enrichedmargarinesanddairyproductsonthemarket.Eating11.5tablespoonsofsterol-
enrichedmargarineeachdaycanhelptolowerbloodcholesterollevels.

Energy density of dietary fat

Dietaryfathasmorethandoubletheamountofkilojoulespergram(37kJ/g)thancarbohydrateor
protein(17kJ/g),makingitveryenergydense.Foodshighinfatareusuallyhighinkilojoules.
Someresearchsuggeststhatsaturatedfatsaremorelikelytocontributetoweightgain(especially
aroundthewaistline)thanpolyunsaturatedfatandmono-unsaturatedfats,eventhoughtheyhave
thesamekilojoulecontent.

Fatsandoils Page2of4
 Carryingtoomuchbodyfatisariskfactorinmanydiseases,includingcoronaryheartdisease,type
2diabetesandmanycancers.

Cholesterol in food
Peoplewithhighbloodcholesterolorwhoareatriskofheartdiseaseshouldalsotrytolimittheir
intakeofcholesterol-richfoods.However,whilecholesterolinfoodcanraisebloodcholesterol
levels,theeffectissmallcomparedtotheeffectsofsaturatedfat.

Dietarycholesterolisonlyfoundinanimalproductssuchas:

full-fatdairyproducts
fattymeats
eggyolks
shellfish
offalforexample,liver,kidneyandbrains.

The Mediterranean diet

Researchersareinvestigatingthepossibilitythatadietrichinmono-unsaturatedfats,suchasolive
oil,maybeprotectiveagainstthedevelopmentofcoronaryheartdisease.Peoplewhohaveahigh
consumptionofmono-unsaturatedfatsfromoliveoil(forexample,inGreeceandItaly)tendto
havelowratesofcoronaryheartdisease,regardlessoftheirbodyweight.

Studieshaveshownthatoliveoilconsumptionmayhaveaprotectiveroleonbreast,colon,lung,
ovarianandskincancerdevelopment.Compoundsspecifictooliveoil,knownasphenolics,seemto
possessfreeradical-scavengingpropertiesandsomaybeabletoreduceoxidativedamagetoDNA.

Anumberofstudieshavealsoshownthatoliveoilmayhaveadditionalbeneficialeffectsonblood
pressure,obesity,rheumatoidarthritisandimmunefunction.

Wemustremember,though,thattheMediterraneandietcontainsmuchmorethanoliveoil.Its
possiblethatthelowrateofcoronaryheartdiseaseinthesecountriesrelatestoahighintakeof
vegetables,legumes,fruitsandcereals,whicharerichinantioxidants.

Current recommendations on fats in your diet

Nutritionistsrecommendthatwelimittheamountoffatsinthedailydiet,particularlysaturated
andtransfats.Simplesuggestionsinclude:

Usemargarinespreadsinsteadofbutterordairyblends.
Usesaladdressingsandmayonnaisemadefromoilssuchascanola,sunflower,soyand
oliveoils.
Useloworreduced-fatmilkandyoghurtoraddedcalciumsoydrinks.
Trytolimitcheesetotwiceaweek.
Havefish(freshorcanned)atleasttwiceaweeklimitfishthatishighinmercurytoonce
aweek.Veryyoungchildrenandwomenwhoarepregnant(orintendingtobecome
pregnantwithinthenextsixmonths)areadvisedtoeatfishthatislowinmercury
(shellfish,salmonandcannedtuna).
Selectleanmeat(meattrimmedoffatandchickenwithoutskin).Trytolimitfattymeats,
includingsausagesanddelicatessenmeatssuchassalami.
Snackonplain,unsaltednutsandfreshfruit.
Incorporatedriedpeas(forexample,splitpeas),beans(forexample,haricotbeans,kidney
beans,threebeanmix)orlentilsintotwomealsaweek.
Makevegetablesandgrain-basedfoods,suchasbreakfastcereals,bread,pasta,noodles
andrice,themajorpartofeachmeal.
Trytolimittakeawaytoonceaweekorless.
Trytolimitsnackfoodssuchaspotatocrispsandcorncrispstoonceaweekorless.
Trytolimitcakes,pastriesandchocolateorcreamybiscuitstoonceaweekorless.
Peoplewithhighcholesterolshouldtrytolimitcholesterol-richfoodssuchaseggyolks,
andoffallikeliver,kidneyandbrains.

Where to get help

Fatsandoils Page3of4
EBIOM-00561; No of Pages 10
EBioMedicine xxx (2016) xxxxxx

Contents lists available at ScienceDirect

EBioMedicine

journal homepage: www.ebiomedicine.com

Research Paper

Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and

Brain Disorders
Qingying Meng a, Zhe Ying a, Emily Noble a, Yuqi Zhao a, Rahul Agrawal a, Andrew Mikhail a, Yumei Zhuang a,
Ethika Tyagi a, Qing Zhang a, Jae-Hyung Lee a,f, Marco Morselli b, Luz Orozco b, Weilong Guo b,g, Tina M. Kilts c,
Jun Zhu d, Bin Zhang d, Matteo Pellegrini b, Xinshu Xiao a, Marian F. Young c,
Fernando Gomez-Pinilla a,e,, Xia Yang a,
a
Department of Integrative Biology and Physiology, University of California, Los Angeles, Los Angeles, CA 90095, USA
b
Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, Los Angeles, CA 90095, USA
c
Craniofacial and Skeletal Diseases Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD 20892, USA
d
Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York 10029, USA
e
Department of Neurosurgery, University of California, Los Angeles, Los Angeles, CA 90095, USA
f
Department of Maxillofacial Biomedical Engineering, School of Dentistry, Kyung Hee University, Seoul 130-701, Korea
g
Center for Synthetic & Systems Biology, TNLIST, Tsinghua University, Beijing 100084, China

a r t i c l e i n f o a b s t r a c t

Article history: Nutrition plays a signicant role in the increasing prevalence of metabolic and brain disorders. Here we employ
Received 1 February 2016 systems nutrigenomics to scrutinize the genomic bases of nutrienthost interaction underlying disease predispo-
Received in revised form 5 April 2016 sition or therapeutic potential. We conducted transcriptome and epigenome sequencing of hypothalamus
Accepted 7 April 2016
(metabolic control) and hippocampus (cognitive processing) from a rodent model of fructose consumption,
Available online xxxx
and identied signicant reprogramming of DNA methylation, transcript abundance, alternative splicing, and
Keywords:
gene networks governing cell metabolism, cell communication, inammation, and neuronal signaling. These sig-
Systems nutrigenomics nals converged with genetic causal risks of metabolic, neurological, and psychiatric disorders revealed in humans.
Fructose Gene network modeling uncovered the extracellular matrix genes Bgn and Fmod as main orchestrators of the ef-
Omega-3 fatty acid fects of fructose, as validated using two knockout mouse models. We further demonstrate that an omega-3 fatty
DHA acid, DHA, reverses the genomic and network perturbations elicited by fructose, providing molecular support for
Epigenome nutritional interventions to counteract diet-induced metabolic and brain disorders. Our integrative approach
Transcriptome complementing rodent and human studies supports the applicability of nutrigenomics principles to predict dis-
Brain networks
ease susceptibility and to guide personalized medicine.
Metabolic diseases
2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
Brain disorders
Extracellular matrix (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction dietary components introduced through industrialization (Chassaing

Metabolic disorders (MetDs) such as metabolic syndrome, obesity,
and type 2 diabetes (T2D) have become a pressing health apprehension
worldwide due to their increasing prevalence and high mortality rate,
and even more recently to their ability to escalate the pathology of neu-
rological and psychiatric disorders (Bomm et al., 2012, Newcomer,
2007; Farooqui et al., 2012, Lowette et al., 2015). Among the potential
culprits for the rising epidemic of metabolic and brain disorders are
Correspondence to: F. Gomez-Pinilla, Department of Neurosurgery and Department of
Integrative Biology and Physiology, University of California, Los Angeles, Los Angeles, CA
90095, USA.
Correspondence to: X. Yang, Department of Integrative Biology and Physiology,
University of California, Los Angeles, Los Angeles, CA 90095, USA.
E-mail addresses: fgomezpi@mednet.ucla.edu (F. Gomez-Pinilla), xyang123@ucla.edu
(X. Yang).
et al., 2015, Suez et al., 2014). In particular, fructose, which has been
widely used as a safe and healthy sweetener in soft drinks and proc-
essed foods in the past decades, is emerging as a signicant contributor
to MetDs in humans (Lyssiotis and Cantley, 2013, Lustig et al., 2012).
Fructose-induced MeDs has been shown to reduce hippocampal-
dependent memory (Agrawal and Gomez-Pinilla, 2012) and to worsen
the pathology of brain disorders in rodents (Agrawal et al., 2015).
Conversely, the omega-3 fatty acid docosahexaenoic acid (DHA) has
been shown to attenuate MetDs (Steffen et al., 2015, Virtanen et al.,
2014, De caterina, 2011), and to counteract the deleterious effects of
fructose on brain function and plasticity (Bremer et al., 2014, Agrawal
and Gomez-Pinilla, 2012). Our understanding of the molecular mecha-
nisms underlying the actions of fructose and DHA on MetDs and brain
disorders has been limited by conventional approaches focusing on
isolated molecular events. This limitation has delayed major advances

http://dx.doi.org/10.1016/j.ebiom.2016.04.008
2352-3964/ 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
2 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx
in the utilization of nutrient-based strategies for the prevention and
treatment of common complex disorders.
As fundamental aspects of gene regulation, disruptions in
epigenomic reprogramming, transcript abundance, alternative splicing,
and genegene interactions are increasingly recognized as core aspects
of wide-ranging pathogenesis (Chen et al., 2008, Zhang et al., 2013,
Rhinn et al., 2013, Narayanan et al., 2014, Makinen et al., 2014, Yang
et al., 2009). Systems nutrigenomics is emerging as a powerful ap-
proach to reveal the hidden aspects of pathogenesis under dietary mod-
ulation (Zhao et al., 2015, De Graaf et al., 2009, Panagiotou and Nielsen,
2009). Here we apply systems nutrigenomics to unveil the multidimen-
sional molecular interactions driven by fructose and DHA that regulate
pathogenesis and recovery, and to provide proof-of-principle on the po-
tential of systems nutrigenomics to guide personalized medicine. The
comparative account of nutrigenomics signals between fructose and
DHA is crucial to understand how select diets impact the molecular
substrates governing the balance between normal brain function and
disease, and holds potential for guiding effective preventative and
therapeutic strategies to mitigate common human diseases.
2. Materials and Methods
We describe essential methods in the main text and detailed exper-
imental procedures are available in the Supplemental Materials.
2.1. Overall Study Design
As depicted in the analysis ow in Fig. 1, we focus our study on two
key regions in the rodent brain that are important for the regulation of
metabolism (hypothalamus) and cognition (hippocampus), and there-
fore can play a major role in fructose-induced metabolic and brain
dysregulation as well as DHA-mediated recovery. We analyzed the tran-
scriptome and DNA methylome using next generation sequencing,
followed by investigation of the regulatory relationship between the
methylome and transcriptome. Network approaches were then applied
to model genegene interactions and to predict essential perturbation
or regulatory points. Knockout mouse models were subsequently used
to validate the predicted regulatory genes with regard to their ability
to modulate metabolic and behavior phenotypes. To infer translatability
to human pathophysiology, we assessed the intersection of the
Fig. 1. Overall study design and analysis ow.
Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
molecular signals from our rodent models with human genome-wide
association studies (GWAS) of metabolic and brain disorders.
2.2. Rat model of fructose consumption and DHA supplementation
Male SpragueDawley rats (Charles River Laboratories, Inc., MA,
USA) of 2 months old weighing 200-220 g were randomly assigned to
15% fructose treatment (n = 8, 15% w/v fructose in the drinking
water), 15% fructose plus an omega-3 fatty acid diet rich in DHA (n =
8; 0.5% of axseed oil supplying ALA and 1.2% of DHA capsule oil, Nordic
Naturals, Inc., CA, USA), or a control group (n = 8, without fructose in
drinking water or DHA supplement) for six weeks. Sample size was cho-
sen to yield N 80% statistical power to detect 30% between-group differ-
ence with 10% within-group difference in a phenotype using two-sided
Student's t-test. The rats were singly housed in polyacrylic cages with
free access to water and their respective diets, and maintained under
standard housing conditions (room temperature 2224 C) with 12 h
light/dark cycle. Daily food and drink intake were monitored. The
fructose intake level is approximately equivalent to long-term
daily consumption of 130 g sugar in 12 l soda drinks in a 60 kg
human. The total fat content in the control and DHA diets was 10 g
per 100 g of diet. The rats were then examined for changes in MetD-
related phenotypes (serum levels of insulin, glucose, and triglycerides,
and insulin resistance index (fasting glucose [mg/dl] fasting insulin
[ng/ml] / 16.31)). Rats were trained in the Barnes maze test for 5 days
to learn the task prior to diet treatment and then tested for memory
retention in the Barnes Maze after 6 weeks of treatment as previously
described (Agrawal and Gomez-Pinilla, 2012). Mice were sacriced,
and hypothalamus and hippocampus were dissected out, ash frozen,
and stored at 70 C for transcriptome and DNA methylome sequenc-
ing experiments.
2.3. RNA Sequencing (RNA-Seq) and Data Analyses
RNA-Seq libraries were prepared for 24 RNA samples (n = 4 per
treatment group per brain region) and sequenced in paired-end mode
by HiSeq 2000 (Illumina Inc., CA, USA) as detailed in Supplemental Ma-
terials. The short reads data were analyzed for different transcription
between treatment and control groups using the Tuxedo tool package
(Trapnell et al., 2012) and false discovery rate (FDR) was estimated
using the q-value approach. One hypothalamus sample from the
fructose group failed standard quality control and was removed from
analysis. Genes and transcripts showing differential expression or alter-
native splicing at p b 0.01 in each brain region were dened as a gene
signature for further integrative analyses. The reliability of tran-
scriptome signals at p b 0.01 from RNA-Se1 was conrmed using quan-
titative real-time PCR (qPCR) for 12 selected genes (Supplemental
Materials). Relative gene expression was represented by delta Ct =
Ctgene CtGapdh. The RNA-Seq data was deposited to Gene Expression
Ominbus (GEO) under accession numbers GSE59918 (control and
fructose groups) and GSE 64815 (DHA).
2.4. Reduced Representation Bisulte Sequencing (RRBS) of DNA
Methylome
RRBS libraries were constructed for 24 DNA samples (n = 4 per
treatment group per brain region) and sequenced using an Illumina
HiSeq 2500 System (Illumina Inc., CA, USA) as described in Supplemen-
tal Materials. Bisulte-converted reads were processed using the bisul-
te aligner BS Seeker2 (Guo et al., 2013). One hypothalamus sample
from the DHA group failed standard quality control and was removed
from analysis. Differential methylation between treatment and control
groups was calculated using the R package methylKit (Akalin et al.,
2012) and FDR was estimated using the q-value approach. Loci
with methylation levels N 25% between groups and FDR b 0.05 were de-
ned as differentially methylated loci (DMLs). The potential cis- and
 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx
trans-regulatory effects of DNA methylome on transcriptome were fur-
ther examined (Supplemental Materials). RRBS data was deposited to
GEO under accession numbers GSE59893 (control and fructose groups)
and GSE64816 (DHA).
2.5. Assessing Overlap Between Fructose Signatures and Candidate Causal
Genes From Human GWAS of MetDs and Brain Disorders
We tested the consistency between the genes identied from fruc-
tose treatment in our study and those from human GWAS related to
MetDs and brain functions in two ways. First, we cross-checked genes
identied in our study with top GWAS candidate genes reported in
the GWAS catalog (http://www.genome.gov/gwastudies/) (Hindorff
et al., 2009) for direct overlap. Second, we used a SNP Set Enrichment
Analysis (SSEA) (Makinen et al., 2014), which goes beyond the top
GWAS hits and tests the overall enrichment of the genes affected in
our fructose animal model for SNPs that demonstrated disease associa-
tion in humans using KolmogorovSmirnov (KS) test and Fisher's exact
test. Bonferroni-corrected p b 0.05 by either KS test or Fisher's exact test
was considered signicant. The detailed methods and GWAS datasets
involved are described in Supplemental Materials.
2.6. Network Modeling and Identication of Key Drivers (KD) of Fructose
and DHA Signatures
We obtained both proteinprotein interaction (PPI) networks from
the HPRD database (http://www.hprd.org) (Keshava Prasad et al.,
2009) and tissue-specic Bayesian networks (BNs) (Zhu et al., 2004,
2008) for hypothalamus and hippocampus constructed based on genet-
ic and transcriptomic data from multiple large-scale mouse studies
(Supplemental Materials). We used the fructose or DHA signature
genes identied from the RNA-Seq analysis as seeds to extract the top
most connected subnetworks as described previously (Yang et al.,
2009). The network components including nodes and edges as well as
topological structures were visualized using Cytoscape (Smoot et al.,
2011). A Key Driver Analysis (KDA; detailed in Supplemental Materials)
(Yang et al., 2010, Zhang et al., 2013) was used to identify potential key
regulators for the fructose and DHA signatures based on the topology of
BNs and PPI networks. Briey, the neighboring subnetwork of each gene
in a given network was rst extracted and then compared with the
tissue-matched fructose or DHA signature to assess the enrichment of
genes in the latter using Fisher's exact test. Network genes that reach
Bonferroni-adjusted p b 0.05 were reported as KDs. KDs identied
from multiple networks were ranked according to a composite score
that favors KDs that are highly signicant in each KDA and are consis-
tent across network models used for KDA (Supplemental Materials).
2.7. KD Validation Using Bgn and Fmod Knockout (KO) Models
Bgn KO mice (Xu et al., 1998) and Fmod KO mice (Svensson et al.,
1999) were given as gifts from National Institute of Dental and Cranio-
facial Research and re-derived at the University of California, Los
Angeles. Mice were housed in standard polyethylene cages (34 mice
per cage) in an environmentally controlled room (2224 C) with a
12 h light/dark cycle and free access to food and water. Homozygous
male KO mice (Bgn: n = 16, Fmod: n = 11) of 9-week age were tested
against male wild type (WT) controls (n = 8) for phenotypic changes in
metabolic (glucose tolerance, plasma lipid panel) and learning and
memory traits (Barnes maze) as described in the Supplemental
Materials.
2.8. Statistics
For metabolic and behavior phenotypes, two-sided Student's t-test
was used to test statistical differences in between fructose-treated
mice and control mice, or between the fructose and fructose + DHA
Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
3
group, or between KO mice (Bgn KO or Fmod KO) and wild-type mice.
Two-way ANOVA with Holm-Sidak post hoc analysis for multiple com-
parisons was performed for the learning curves. For RNA-Seq, RRBS,
KDA, and SSEA analyses, the statistical tests and multiple testing correc-
tion methods were described in detail under the respective method
sections. All statistical and bioinformatics analyses were performed in
R (http://www.R-project.org/).
2.9. Study Approval
All experiments were performed in accordance with the National In-
stitutes of Health Guide for the Care and Use of Laboratory Animals and
were approved by the Chancellor's Animal Research Committee of the
University of California, Los Angeles.
3. Results
3.1. Induction of Metabolic and Behavioral Abnormalities by Fructose and
Normalization by DHA
Compared to the control group (n = 8), adult rats consuming fruc-
tose (n = 8) had MetD characteristics including signicantly increased
blood glucose, triglycerides, insulin, and insulin resistance index
(Fig. 2a). The animals on fructose also displayed impaired memory as
demonstrated by prolonged latency in the Barnes Maze test (Fig. 2a).
On the other hand, fructose-treated animals with DHA supplementation
(DHA + fructose) showed signicantly improved metabolic parameters
including lower serum triglycerides, insulin, insulin resistance index,
as well as improved memory as reected by the reduced latency
time in the Barnes Maze test (Fig. 2a). The measurements in the
DHA + fructose group were not signicantly different from the control
group with no fructose treatment for triglycerides, insulin, and latency
time, indicating that DHA supplementation counteracted the fructose
effects. Of note, total caloric intake between groups was similar between
groups.
3.2. Transcriptomic Changes in Hypothalamus and Hippocampus Induced
by Fructose and Normalization by DHA
Using RNA-Seq, we identied 581 and 146 differentially expressed
genes, 352 and 79 differentially expressed transcripts, and 99 and 53
genes showing alternative exon usage in hypothalamus and hippocam-
pus, respectively, at p b 0.01 (Dataset S1). These results support large-
scale alterations in transcriptional activities affecting both differential
expression and alternative splicing, and dene 734 unique hypothalam-
ic genes (312 passed FDR b 5%) and 206 hippocampal genes (56 passed
FDR b 5%) as fructose gene signatures. The reliability of the differen-
tially signals at p b 0.01 was supported by consistent qPCR experiments
on 12 selected genes (Table S1). Among the signatures were transcrip-
tion factors (e.g., Aff3, Junb, Zbtb16, Tcf7l2), epigenetic regulators
(e.g., Trmu, Suv420h2, Rgs9bp), and alternative splicing regulators
(e.g., Bicc1, Prpf31, Rbpms) (Dataset S1), which may partially explain
the large-scale transcriptional alterations observed.
The fructose signatures showed strong tissue-specicity, with 664
genes unique to the hypothalamus, 136 unique to the hippocampus,
and 70 overlapping between tissues. In particular, only in hypothalamus
we observed upregulation of Slc2a5 (encoding the major fructose trans-
porter GLUT5) and downregulation of Slc2a4 (encoding the glucose
transporter protein GLUT4). The signatures also differ by direction of
change between the two brain compartments: 68% of hypothalamic sig-
natures showed down-regulation compared to 42% in hippocampus.
Among the overlapping genes between the two brain regions (Dataset
S1), only 6 showed consistent upregulation (Atp5f1, Pnpla1, Pomc,
Sepw1, Slc39a12, and Ubc) and 5 showed consistent downregulation
(Alox15, Dstn, Rpl30, Slc39a4, and Slc5a7). The consistent genes are
4 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx

Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx 5

involved in mitochondria function, energy balance, antioxidation,
inammation, and cognition.
Compared to the fructose group, the DHA + fructose group showed
differential expression of 651 and 462 genes, 449 and 277 transcripts,
and alternative exon usage for 155 and 81 genes in hypothalamus and
hippocampus at p b 0.01, respectively (Dataset S1), which together de-
ne DHA gene signatures of 910 unique genes (382 passed FDR b 5%)
in hypothalamus and 569 unique genes (246 passed FDR b 5%) in hippo-
campus. Strikingly, these DHA signatures signicantly overlapped with
the fructose signatures (Table 1), with DHA reversing the expression of
the common signatures perturbed by fructose in both brain regions
(Fig. 2b; Dataset S1).
localized with DMLs within 50 kb distance (Table S3), suggesting cis-
regulation of gene expression by the local DMLs. Notably, some of the
genes with co-localizing DMLs are transcription factors (Aff3, Junb, and
Zbtb16) or epigenetic factors (Parp9) (Dataset S3), which could in turn
trans-regulate downstream target genes. Additionally, microRNAs adja-
cent (within 20 kb or 50 kb) to the DMLs had the potential to regulate
36%45% and 17%45% of hypothalamic and hippocampal signatures
which were the predicted target genes of the corresponding microRNAs
(Table S3; details of the microRNAs in Dataset S3). These results suggest
that fructose-induced DNA methylation changes could contribute
to transcriptional alteration through both cis and trans-regulatory
mechanisms.

3.3. Functional Categorization of Fructose and DHA Transcriptomic 3.6. Perturbation of Gene Subnetworks by Fructose and Normalization by
Signatures DHA

We found enrichment of both tissue-specic and shared biological
pathways between brain regions and between treatment groups (select
pathways shown in Fig. 2d; full details in Dataset S2). Fructose and DHA
signatures shared similar over-represented pathways, although the di-
rection of expression changes of the overlapping genes was predomi-
nantly opposite (exemplied in Fig. 2d insert; details in Dataset S1).
The pathways unique to hypothalamus included those highly relevant
to cardiometabolic, inammatory, and neuronal functions. The pathway
affected in the hippocampus was neurological processes. Those affect-
ed in both regions were related to extracellular matrix (ECM), adhesion,
complement, and insulin-like growth factor (IGF) signaling.
3.4. Induction of DNA Methylomic Changes by Fructose and Normalization
by DHA
To identify epigenomic changes which may play a role in the
transcriptomic alternation associated with fructose and DHA, we exam-
ined the DNA methylome using RRBS that measures millions of poten-
tial DNA methylation sites at single base resolution. We found that
fructose consumption induced a large-scale switch of the methylation
patterns in both brain regions: 734 and 810 of differentially methylated
loci (DMLs) showed hypomethylation and hypermethylation in hypo-
thalamus, respectively; 972 and 900 DMLs showed hypomethylation
and hypermethylation in hippocampus at a FDR b 5%, respectively
(Fig. 2c). DHA largely reversed the fructose-induced methylation
changes in these two brain regions (Fig. 2c; Table 1).
3.5. Relationship Between Differential Methylation and Differential
Expression
To explore the potential relationship between DMLs and gene ex-
pression, we mapped the fructose DMLs in each brain region to adjacent
genes within 10 kb distance. We found 50% of DMLs to be located in the
intergenic regions, 34% in the gene body (including 5-UTR, intron,
coding sequence, and 3-UTR), and 16% in the upstream or downstream
regions of the genes, with a paucity of DMLs in the promoter and tran-
scription start site (Fig. S1; Table S2). Assessing the methylome-
transcriptome relationship revealed co-localization of DMLs and gene
signatures: 74 out of the 734 (10.1%) hypothalamic signature genes
and 18 out of the 206 (8.7%) hippocampal signature genes co-
To explore the genegene relations among the fructose signature
genes and to identify key perturbation points of fructose, we employed
two types of networks - proteinprotein interaction (PPI) networks
(Keshava Prasad et al., 2009) that capture a majority of the well-
known interactions between proteins, and data-driven Bayesian
networks (BNs) (Zhu et al., 2004, 2008) that elucidate genegene regu-
latory relationships (detailed in Materials and Methods). Using these
networks and a network topology-based key driver analysis (KDA),
we identied multiple candidate key driver (KD) genes whose network
neighboring genes highly overlapped with the fructose signature genes
(Dataset S4). As shown in Fig. 3, the top 5 KDs for each brain region form
central hubs connecting many fructose signature genes in coherent
gene subnetworks, which contained genes involved in diverse process-
es such as the ECM (e.g., Col6a2, Lama4, Thbs1, Fmod, Bgn), small mole-
cule transporters (e.g., Slc6a13, Slc22a2, Slc22a6, Slc13a3), and forkhead
transcription factors (e.g., Foxc2, Foxd2). Between the subnetworks of
the two brain compartments, Fmod, Bgn, and Foxc2 were consistent
KDs revealed from our data-driven approach (Dataset S4). Remarkably,
these KDs and subnetworks were largely reversed by DHA supplemen-
tation (Fig. 3).
3.7. Validation of Bgn and Fmod as Key Regulators of MetDs and Behavioral
Aberrations in KO Mouse Models
To evaluate our network KD prediction, we used Bgn and Fmod KO
mice (Xu et al., 1998; Svensson et al., 1999) to test the effect of
perturbing these KDs on phenotypes related to MetDs and cognitive
function. Compared to wild type (WT) controls, we observed signi-
cantly elevated triglyceride, total cholesterol, HDL cholesterol, and un-
esteried cholesterol in the plasma of both Bgn and Fmod KO mice, sig-
nicant increases in LDL cholesterol in the Bgn KO, and elevated free
fatty acids in the Fmod KO (Fig. 4a). We also found decreased levels of
plasma glucose, insulin, as well as insulin resistance index in Fmod KO
(Fig. 4bd). The intraperitoneal glucose tolerance test of Bgn KO
demonstrated signicantly lower plasma glucose at 15 min after
glucose injection and a similar trend in the Fmod KO (Fig. 4e).
To assess the potential inuence of Bgn and Fmod for spatial learning
and memory, we tested the animals in the Barnes Maze. Because Fmod
and Bgn KOs have been previously shown to have impaired tendon
and joints which may affect latency time (Ameye et al., 2002,

Fig. 2. Changes in metabolic and behavior phenotypes, transcriptome, DNA methylome, and biological pathways in response to fructose treatment and DHA supplementation.
(a) Metabolic and behavior phenotypes. From left to right: blood glucose, serum triglycerides, serum insulin, insulin resistance index, and latency time in the memory retention probe
in the Barnes Maze test. Fructose group was compared with the control group, and the fructose + DHA group was compared to the fructose only group. p b 0.01 and #p b 0.05 by 2-
sided Student's t-test. Error bars in the plots are standard errors. N = 8/group. (b) Heatmap of gene expression changes in hypothalamus and hippocampus. Blue to red colors indicate
low to high expression values. (c) Heatmap of DNA methylation changes in hypothalamus and hippocampus. Blue to red colors indicate low to high methylation levels. (d) Select
biological pathways affected by fructose and DHA in hypothalamus and hippocampus. Bars are the log10 enrichment p values of the pathways. The insert plot on the right shows the
opposite direction of changes in genes in the focal adhesion pathway in the hypothalamus between fructose and DHA: fructose mostly inhibits whereas DHA reverses the expression
levels. In the insert plot Y-axis indicates log10 fold changes. The fructose + DHA group is labeled as F + D in panels bd. (For interpretation of the references to color in this gure
legend, the reader is referred to the web version of this article.)

Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
6 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx
Table 1
Overlap between fructose and DHA signatures. Numbers of signicant genes or methylation loci are shown. Enrichment p value was calculated using 2-sided Fisher's exact test.
Data Tissue Fructose signature DHA signature
RNAseq Hypothalamus 734 910
RNAseq Hippocampus 206 569
RRBS Hypothalamus 1544 1665
RRBS Hippocampus 1872 1957
Chakravarti, 2002; Jepsen et al., 2002), we analyzed the Barnes Maze
data by number of mistakes to eliminate the potential confound origi-
nated by differences in mobility. In addition, number of mistakes has
been demonstrated as a more accurate measure of cognitive abilities
in mice (O'leary and Brown, 2013). During the learning phase, there
was a signicant effect of animal type (F2,35 = 10.14; p = 0.0003 by
2-way ANOVA) and day (F3,105 = 40.44; p b 0.0001 by 2-way ANOVA)
on the number of mistakes made in the Barnes Maze (Fig. 4f). Post
hoc analysis revealed that both Bgn KO and Fmod KO made fewer
mistakes compared with WT during the learning and the memory
phases. Representative track plot analysis show a difference in strat-
egy to nd the escape hole (Fig. 4g, inset): the WT animals use a trial
and error approach while the Bgn and Fmod KO animals follow a
more direct path.
3.8. Relevance of Fructose Signatures and KDs to Human Diseases
To assess the relevance of our nutrigenomic signals from the rodent
models to human pathophysiology, we compared the fructose signature
genes as well as the KDs to human GWAS of MetDs and brain related
diseases or traits. We found numerous overlapping genes between fruc-
tose signature genes and top GWAS hits in the GWAS catalog for a broad
range of brain disorders and MetDs-related diseases or traits (Dataset
S4). Importantly, KDs such as Fmod, Col1a2, Col18a1, and Dcn were
GWAS candidate genes associating with obesity, cardiovascular disease,
metabolic syndrome, and brain morphology and cognition, respectively.
When using full sets of GWAS results we had accessed for plasma lipids,
T2D, systolic and diastolic blood pressure traits, cognitive traits, and bi-
polar disorder, we found that human orthologs of the fructose signature
genes derived from both hypothalamus and the hippocampus were sig-
nicantly enriched for genetic polymorphisms showing stronger associ-
ations with these diseases or traits (Table 2).
4. Discussion
In this systems nutrigenomics study, we show that fructose re-
models fundamental aspects of gene regulation such as differential
DNA methylation, differential gene expression, alternative splicing,
and implications for microRNA alterations, all of which have the poten-
tial to impact pathogenesis. In addition, we found that fructose alters
the organization of genes in brain region-specic networks interrelating
cell metabolism, immune function, inammation, and cell communica-
tion via key regulators such as the extracellular matrix genes Bgn and
Fmod. Remarkably, we found that DHA supplementation was capable
to reverse the genomic, epigenomic, network, and phenotypic perturba-
tions induced by fructose via the same key drivers.
Our ndings indicate that fructose and DHA induce transcriptomic
reprogramming by engaging core transcription factors, splicing factors,
and epigenetic modications. For instance, fructose altered classic
transcription factors such as Tcf7l2 (the most signicant and robust
signal for T2D from human GWAS) (Voight et al., 2010) and Zbtb16. Ad-
ditionally, N20% of the fructose and DHA signature genes only showed
changes at isoform or alternative splicing levels but not overall expres-
sion, and alternative splicing factors such as Bicc1, Prpf31, and Rbpms
were among the signature genes, supporting the importance of RNA
splicing in the pathogenesis of human diseases highlighted recently
(Xiong et al., 2015). The involvement of epigenetic mechanisms in
Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
Overlap Background Fold enrichment p value
374 17,435 9.76 7.6E305
118 17,411 17.53 1.9E124
381 6,686,093 984.54 0
557 8,742,773 1329.25 0
fructose and DHA actions was substantiated by the signicant alter-
ations in the methylome of hypothalamus and hippocampus. cis-Regu-
lation of transcriptome by DNA methylation was evidenced by co-
localization of a subset of signature genes with local DMLs; trans-
regulation mechanism was supported by the co-localization of DMLs
with a number of transcriptional regulators (e.g., Aff3, Junb, Zbtb16,
and Parp9) and microRNAs (e.g., rno-miR-421, rno-miR-143) whose
predicted target genes were enriched among the signature genes. An in-
creasing body of evidence indicates that neurological and psychiatric
disorders have an epigenetic component (Jakovcevski and Akbarian,
2012, Tsankova et al., 2007). In particular, disruption in cell metabolism
seems to be a driving force for epigenetic changes associated with cog-
nitive function (Tyagi et al., 2015). The current results showing that
DHA can counteract the broad transcriptomic and methylomic alter-
ations induced by fructose imply that diet has the ability to regulate
the susceptibility to disease by modulating the multifaceted transcrip-
tional machinery.
Our studies also revealed broad impact of fructose and DHA on fun-
damental cell processes encompassing metabolism, inammation, cell
communication, and neuronal signaling, as well as critical differences
between the two brain regions. For instance, the fructose and glucose
transporter genes Slc2a5 and Slc2a4, and a broader category of pathways
including insulin signaling, cardiomyopathy, transforming growth
factor (TGF) beta signaling, mitogen-activated protein kinases (MAPK)
signaling, platelet-derived growth factor (PDGF) signaling, toll like
receptor (TLR) signaling, brain-derived neurotropic factor (BDNF) sig-
naling, neural cell adhesion molecule (NCAM) signaling, and axon guid-
ance were perturbed only in the hypothalamus. In fact, most of the
hypothalamic pathways have been recently implicated in leptin resis-
tance, hypothalamic deregulation and peripheral metabolism impair-
ment (Milanski et al., 2012, Yan et al., 2014). The prominent response
of the hypothalamus to diets is consistent with its role as the master
metabolic sensor and regulator, and the current results suggest that
these actions intermingle with inammatory and cognitive signals. In
contrast, the effects of fructose on the hippocampus were prominent
in pathways engaged in cognitive function and neurological processes.
For instance, hippocampal signature genes Pdc, Chrnb4, and Gch1 were
previously implicated in schizophrenia (Sullivan et al., 2008), addictive
behavior (Picciotto and Kenny, 2013), attention and vigilance (Yasuda
et al., 2014), and early-onset Parkinson's disease (Cederfjall et al.,
2013). The biological pathways affected in both brain regions were
related to ECM, focal adhesion, complement cascade, smooth muscle
contraction, and IGF signaling, which have been implicated in both
metabolic and brain disorders (Soleman et al., 2013, Frischknecht and
Gundelnger, 2012, Chen et al., 2011, Phieler et al., 2013, Skaper, 2007).
A highlight of the study was the identication of the extracellular
matrix genes Bgn and Fmod as key intermediates of the fructose effects
on the gene network and phenotypic alterations, and these ndings
were corroborated in two knockout mouse models. Small leucine-rich
proteoglycans Fmod and Bgn consistently emerged as KDs in both
brain regions for both dietary components, and the expression levels
of both genes were also signicantly altered by fructose and DHA.
Fmod and Bgn are important structural components of the ECM. The
role of the ECM in the regulation of neurite outgrowth, and neural
functionality and plasticity (Dityatev et al., 2010) as well as metabolic
diseases (Soleman et al., 2013, Frischknecht and Gundelnger, 2012,
Chan et al., 2014) is becoming increasingly recognized. Our study
 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx 7

Fig. 3. Gene subnetworks and top network key drivers (KDs) of fructose and DHA. (a) KDs and gene subnetwork in hypothalamus. (b) KDs and gene subnetwork in hippocampus. Larger
nodes depict KDs; grey nodes are network genes in the neighborhood of KDs that are not affected by fructose or DHA; top and bottom halves of each node, if colored, denotes genes affected
by fructose (top) and DHA (bottom), respectively; red and blue colors denotes increased and decreased expression, respectively. Direction of change for the fructose group is determined
by comparison with the control group; direction of change for the DHA + fructose group is determined by comparison with the fructose group. (For interpretation of the references to color
in this gure legend, the reader is referred to the web version of this article.)

provides unique nutrigenomic-based evidence that proteoglycans are
likely main orchestrators of important metabolic and cognitive process-
es in the brain.
To assess the translatability of our rodent ndings to human patho-
physiology, we found signicant intersections between the fructose sig-
natures or the network KDs identied in our rodent models and human
GWAS hits for neurological and neurodegenerative diseases, psychiatric
diseases, and various metabolic diseases. As human GWAS implies a
causal role of the candidate genes based on genetic evidence, the over-
lapping signatures or network KDs in response to fructose likely play
causal roles in the development of MetDs or brain disorders. The same
genes also likely mediate the benecial actions of DHA.
The strength of our study lies in the use of powerful systems
nutrigenomic approach to comprehensively characterize phenotypic

Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
8 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx

Fig. 4. Phenotypic validation of Bgn and Fmod using knockout (KO) mice. (a) Lipid traits including triglyceride (TG), total cholesterol, HDL cholesterol, un-esteried cholesterol (UC), LDL
cholesterol and free fatty acids (FFA). (b) Glucose. (c) Insulin. (d) Insulin resistance (IR) measured by insulin sensitivity index. (e) Intraperitoneal glucose tolerance test (IPGTT).
(f) Mistakes made during the Barnes Maze test in four days of the learning phase. (g) Track plots and mistakes made during Barnes Maze test for spatial memory. Two-sided Student's
t-test was used to test statistical difference between knockout mice (Bgn KO or Fmod KO) and wild-type (WT) mice for analyses in (a)(g). Two-way ANOVA with Holm-Sidak post
hoc analysis for multiple comparisons was performed for the learning curves in panel (f) with *WT vs. Bgn KO; #Bgn KO vs. Fmod KO; and WT vs. Fmod KO. Error bars in the plots are
standard errors. p b 0.05. Sample size n = 816.

and multi-tissue transcriptomic and methylomic proles of fructose and evidence for the potential causal role of the identied genes and
DHA, followed by network modeling, in vivo validation, and human data pathways in human diseases.
integration. The highly integrative nature not only enables a compre- We note the following limitations of the current study. First, the Bgn
hensive and in-depth exploration of the molecular mechanisms mediat- and Fmod KO animals used are not brain-specic. It is possible that the
ing the actions of nutrition on health and disease, but also pinpoints phenotypic changes in peripheral metabolism observed in these models
critical regulators of the biological pathways and provides compelling are the results of perturbations of these genes outside of the brain.
Future tissue-specic studies are warranted to further investigate the
action sites of these ECM genes. Second, the behavioral changes ob-
Table 2 served in our systemic treatment studies of fructose and DHA as well
Enrichment of human GWAS signals in fructose signatures and KDs by SNP set enrichment
as in the genetically modied animal models of Bgn and Fmod could
analysis (SSEA).
be inuenced by numerous variables that are either not investigated
Tissue GWAS GWAS p-Value p-Value or difcult to tease out. Third, we only focused on two brain regions in
disease/trait studya (KS test)b (Fisher's the current study and future investigations of additional tissues and
exact testb
cell types are warranted.
Hypothalamus HDL cholesterol GLGC 4.35E05 1.25E12 The staggering increase in the prevalence of MetDs, its negative im-
LDL cholesterol GLGC 1.05E03 1.90E05
pact on brain function and pathogenesis of brain disorders, and the role
Total cholesterol GLGC 8.51E06 3.65E14
Triglycerides GLGC 5.88E04 3.09E09 played by fructose consumption as a dietary perturbation in these pro-
Diastolic blood pressure ICBP 2.40E05 7.55E05 cesses are becoming major concerns for public health and basic and
Systolic blood pressure ICBP 3.06E03 1.16E02 clinical science. Our study provides molecular evidence supporting the
Type 2 diabetes DIAGRAM+ 2.07E02 3.39E02
ability of fructose to disrupt critical genes and fundamental physiologi-
Cognitive function FHS 1.78E02 3.18E03
Bipolar disorder WTCCC 3.06E02 4.04E02
cal processes such as insulin, IGF, TLR, TGF beta, BDNF, MAPK, NCAM,
and PDGF signaling pathways and neurological processes. The majority
Hippocampus HDL cholesterol GLGC 6.48E05 2.02E07
of these pathways have been implicated in both MetDs and brain disor-
LDL cholesterol GLGC 2.98E05 2.16E04
Total cholesterol GLGC 2.37E05 4.19E13 ders. In addition, genes affected by fructose signicantly overlapped
Triglycerides GLGC 1.34E02 6.55E06 with human GWAS genes for brain disorders (e.g., Alzheimer's disease,
Diastolic blood pressure ICBP 4.52E01 4.59E03 attention decient hyperactive disorder, depression, addiction,
Cognitive function FHS 9.43E03 4.28E02
Parkinson's disease) and MetDs-related diseases (e.g., blood pressure,
Schizophrenia CATIE 2.26E02 3.02E01
cardiovascular disease, lipids, obesity, metabolic syndrome, and T2D).
a
GLGC: Global Lipids Genetics Consortium; ICBP: The International Consortium for Our results point to the capacity of fructose to reprogram molecular
Blood Pressure; DIAGRAM +: The Diabetes Genetics Replication And Meta-analysis
Consortium; FHS: Framingham Heart Study; WTCCC: Wellcome Trust Case Control
substrates underlying pathology. The remarkable capacity of DHA sup-
Consortium; CATIE: Clinical Antipsychotic Trials for Intervention Effectiveness. plementation to restore genomic, epigenomic, pathway, network, and
b
Bolded p values are those passed Bonferroni-corrected p b 0.05. phenotypic traits vulnerable to the effects of fructose provides multi-

Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx
level support for the potential of omega-3 fatty acids as a nutritional
remedy for metabolic and brain dysregulation. This points to an effec-
tive means to guide personalized medicine by matching the disease-
promoting molecular signatures with an agent capable of reversing
the patterns.
In summary, we show that DHA reverses biological pathways and
gene networks perturbed by fructose via transcriptional regulators
(such as transcription factors, epigenetic regulators, splicing factors)
and essential network regulators (such as Bgn and Fmod) that may con-
trol the balance between health and disease. Our study reveals critical
information supporting an action of ECM in orchestrating brain function
through gene network organizations. Our experimental validation high-
lights the predictive power of our systems genomics approaches in
identifying vulnerable disease genes and mechanisms. In addition, our
paradigm based on the integration between rodent studies and
human GWAS serves as a precedent to apply nutrigenomics principles
to predict disease susceptibility and to guide personalized medicine.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.ebiom.2016.04.008.
Role of Funding Sources
XY is supported by NIH grant R01DK104363, American Heart
Association Scientist Development Grant 13SDG17290032, Leducq
Foundation, Hellman Fellows Award, UCLA Faculty Research Grant,
and UCLA CTSI Grant UL1TR000124. FGP is supported by NIH grants
R01DK104363, R01NS050465 and The Letten Foundation. MFY was
supported in part by the Intramural Program of the NIDCR, NIH. XX is
supported by NIH grants R01HG006264 and U01HG007013, and NSF
grant 1262134. GW was supported by NSFC (91019016), NBRPC
(2012CB316503), and the China Scholarship Council. None of the
funding sources had any role in the design and conduct of the study,
the collection, analysis, or interpretation of the data, or the preparation,
review, or approval of the manuscript.
Conicts of Interests
The authors declare no conict of interests.
Author Contributions
QM contributed to study design, project coordination, experiments,
data analyses, writing and editing of the manuscript; XY and FGP de-
signed and directed the study, and contributed to data interpretation,
writing and editing of the manuscript; EN contributed to experiments,
data analysis, and writing of the manuscript; ZY, YZ, RA, YZhuang, ET,
AM, QZ, JL, MM, LO, WG, JZ, BZ, MP, TMK, XX, and MFY contributed to
experiments and/or data analyses, and reviewed and edited the
manuscript.
Acknowledgements
We thank Dr. Oldberg Ake from Experimental Medical Sciences, Uni-
versity of Lund, Lund, Sweden for kindly providing the Fmod knockout
mouse model for this study.
References
Agrawal, R., Gomez-Pinilla, F., 2012. Metabolic syndrome in the brain: deciency in
omega-3 fatty acid exacerbates dysfunctions in insulin receptor signalling and
cognition. J. Physiol. 590, 24852499.
Agrawal, R., Noble, E., Vergnes, L., Ying, Z., Reue, K., Gomez-Pinilla, F., 2015. Dietary
fructose aggravates the pathobiology of traumatic brain injury by inuencing energy
homeostasis and plasticity. J. Cereb. Blood Flow Metab. (Online before print).
Akalin, A., Kormaksson, M., Li, S., Garrett-Bakelman, F.E., Figueroa, M.E., Melnick, A.,
Mason, C.E., 2012. methylKit: a comprehensive R package for the analysis of
genome-wide DNA methylation proles. Genome Biol. 13, R87.
Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
9
Ameye, L., Aria, D., Jepsen, K., Oldberg, A., Xu, T., Young, M.F., 2002. Abnormal collagen
brils in tendons of biglycan/bromodulin-decient mice lead to gait impairment,
ectopic ossication, and osteoarthritis. FASEB J. 16, 673680.
Bomm, T.R., Forny-Germano, L., Sathler, L.B., Brito-Moreira, J., Houzel, J.C., Decker, H.,
Silverman, M.A., Kazi, H., Melo, H.M., Mcclean, P.L., Holscher, C., Arnold, S.E., Talbot,
K., Klein, W.L., Munoz, D.P., Ferreira, S.T., De Felice, F.G., 2012. An anti-diabetes
agent protects the mouse brain from defective insulin signaling caused by
Alzheimer's disease-associated Abeta oligomers. J. Clin. Invest. 122, 13391353.
Bremer, A.A., Stanhope, K.L., Graham, J.L., Cummings, B.P., Ampah, S.B., Saville, B.R., Havel,
P.J., 2014. Fish oil supplementation ameliorates fructose-induced hypertriglyc-
eridemia and insulin resistance in adult male rhesus macaques. J. Nutr. 144, 511.
Cederfjall, E., Nilsson, N., Sahin, G., Chu, Y., Nikitidou, E., Bjorklund, T., Kordower, J.H., Kirik,
D., 2013. Continuous DOPA synthesis from a single AAV: dosing and efcacy in
models of Parkinson's disease. Sci. Rep. 3, 2157.
Chakravarti, S., 2002. Functions of lumican and bromodulin: lessons from knockout
mice. Glycoconj. J. 19, 287293.
Chan, K.H., Huang, Y.T., Meng, Q., Wu, C., Reiner, A., Sobel, E.M., Tinker, L., Lusis, A.J., Yang,
X., Liu, S., 2014. Shared molecular pathways and gene networks for cardiovascular
disease and type 2 diabetes mellitus in women across diverse ethnicities. Circ.
Cardiovasc. Genet. 7, 911919.
Chassaing, B., Koren, O., Goodrich, J.K., Poole, A.C., Srinivasan, S., Ley, R.E., Gewirtz, A.T.,
2015. Dietary emulsiers impact the mouse gut microbiota promoting colitis and
metabolic syndrome. Nature 519, 9296.
Chen, Y., Zhu, J., Lum, P.Y., Yang, X., Pinto, S., Macneil, D.J., Zhang, C., Lamb, J., Edwards, S.,
Sieberts, S.K., Leonardson, A., Castellini, L.W., Wang, S., Champy, M.F., Zhang, B.,
Emilsson, V., Doss, S., Ghazalpour, A., Horvath, S., Drake, T.A., Lusis, A.J., Schadt, E.E.,
2008. Variations in DNA elucidate molecular networks that cause disease. Nature
452, 429435.
Chen, D.Y., Stern, S.A., Garcia-Osta, A., Saunier-Rebori, B., Pollonini, G., Bambah-Mukku, D.,
Blitzer, R.D., Alberini, C.M., 2011. A critical role for IGF-II in memory consolidation and
enhancement. Nature 469, 491497.
De Caterina, R., 2011. n-3 fatty acids in cardiovascular disease. N. Engl. J. Med. 364,
24392450.
De Graaf, A.A., Freidig, A.P., De Roos, B., Jamshidi, N., Heinemann, M., Rullmann, J.A., Hall,
K.D., Adiels, M., Van Ommen, B., 2009. Nutritional systems biology modeling: from
molecular mechanisms to physiology. PLoS Comput. Biol. 5, e1000554.
Dityatev, A., Schachner, M., Sonderegger, P., 2010. The dual role of the extracellular matrix
in synaptic plasticity and homeostasis. Nat. Rev. Neurosci. 11, 735746.
Farooqui, A.A., Farooqui, T., Panza, F., Frisardi, V., 2012. Metabolic syndrome as a risk factor
for neurological disorders. Cell. Mol. Life Sci. 69, 741762.
Frischknecht, R., Gundelnger, E.D., 2012. The brain's extracellular matrix and its role in
synaptic plasticity. Adv. Exp. Med. Biol. 970, 153171.
Guo, W., Fiziev, P., Yan, W., Cokus, S., Sun, X., Zhang, M.Q., Chen, P.Y., Pellegrini, M., 2013.
BS-Seeker2: a versatile aligning pipeline for bisulte sequencing data. BMC Genomics
14, 774.
Hindorff, L.A., Sethupathy, P., Junkins, H.A., Ramos, E.M., Mehta, J.P., Collins, F.S., Manolio,
T.A., 2009. Potential etiologic and functional implications of genome-wide association
loci for human diseases and traits. Proc. Natl. Acad. Sci. U. S. A. 106, 93629367.
Jakovcevski, M., Akbarian, S., 2012. Epigenetic mechanisms in neurological disease. Nat.
Med. 18, 11941204.
Jepsen, K.J., Wu, F., Peragallo, J.H., Paul, J., Roberts, L., Ezura, Y., Oldberg, A., Birk, D.E.,
Chakravarti, S., 2002. A syndrome of joint laxity and impaired tendon integrity in
lumican- and bromodulin-decient mice. J. Biol. Chem. 277, 3553235540.
Keshava Prasad, T.S., Goel, R., Kandasamy, K., Keerthikumar, S., Kumar, S., Mathivanan, S.,
Telikicherla, D., Raju, R., Shafreen, B., Venugopal, A., Balakrishnan, L., Marimuthu, A.,
Banerjee, S., Somanathan, D.S., Sebastian, A., Rani, S., Ray, S., Harrys Kishore, C.J., Kanth,
S., Ahmed, M., Kashyap, M.K., Mohmood, R., Ramachandra, Y.L., Krishna, V., Rahiman,
B.A., Mohan, S., Ranganathan, P., Ramabadran, S., Chaerkady, R., Pandey, A., 2009.
Human protein reference database2009 update. Nucleic Acids Res. 37, D767D772.
Lowette, K., Roosen, L., Tack, J., Vanden Berghe, P., 2015. Effects of high-fructose diets on
central appetite signaling and cognitive function. Front. Nutr. 2, 5.
Lustig, R.H., Schmidt, L.A., Brindis, C.D., 2012. Public health: the toxic truth about sugar.
Nature 482, 2729.
Lyssiotis, C.A., Cantley, L.C., 2013. Metabolic syndrome: F stands for fructose and fat. Na-
ture 502, 181182.
Makinen, V.P., Civelek, M., Meng, Q., Zhang, B., Zhu, J., Levian, C., Huan, T., Segre, A.V.,
Ghosh, S., Vivar, J., Nikpay, M., Stewart, A.F., Nelson, C.P., Willenborg, C., Erdmann, J.,
Blakenberg, S., O'donnell, C.J., Marz, W., Laaksonen, R., Epstein, S.E., Kathiresan, S.,
Shah, S.H., Hazen, S.L., Reilly, M.P., Lusis, A.J., Samani, N.J., Schunkert, H.,
Quertermous, T., Mcpherson, R., Yang, X., Assimes, T.L., 2014. Integrative genomics re-
veals novel molecular pathways and gene networks for coronary artery disease. PLoS
Genet. 10, e1004502.
Milanski, M., Arruda, A.P., Coope, A., Ignacio-Souza, L.M., Nunez, C.E., Roman, E.A.,
Romanatto, T., Pascoal, L.B., Caricilli, A.M., Torsoni, M.A., Prada, P.O., Saad, M.J.,
Velloso, L.A., 2012. Inhibition of hypothalamic inammation reverses diet-induced in-
sulin resistance in the liver. Diabetes 61, 14551462.
Narayanan, M., Huynh, J.L., Wang, K., Yang, X., Yoo, S., Mcelwee, J., Zhang, B., Zhang, C.,
Lamb, J.R., Xie, T., Suver, C., Molony, C., Melquist, S., Johnson, A.D., Fan, G., Stone,
D.J., Schadt, E.E., Casaccia, P., Emilsson, V., Zhu, J., 2014. Common dysregulation net-
work in the human prefrontal cortex underlies two neurodegenerative diseases.
Mol. Syst. Biol. 10, 743.
Newcomer, J.W., 2007. Metabolic syndrome and mental illness. Am. J. Manag. Care 13,
S170S177.
O'leary, T.P., Brown, R.E., 2013. Optimization of apparatus design and behavioral measures
for the assessment of visuo-spatial learning and memory of mice on the Barnes maze.
Learn. Mem. 20, 8596.
10 Q. Meng et al. / EBioMedicine xxx (2016) xxxxxx
Panagiotou, G., Nielsen, J., 2009. Nutritional systems biology: denitions and approaches.
Annu. Rev. Nutr. 29, 329339.
Phieler, J., Garcia-Martin, R., Lambris, J.D., Chavakis, T., 2013. The role of the complement
system in metabolic organs and metabolic diseases. Semin. Immunol. 25, 4753.
Picciotto, M.R., Kenny, P.J., 2013. Molecular mechanisms underlying behaviors related to
nicotine addiction. Cold Spring Harb. Perspect. Med. 3, a012112.
Rhinn, H., Fujita, R., Qiang, L., Cheng, R., Lee, J.H., Abeliovich, A., 2013. Integrative genomics
identies APOE epsilon4 effectors in Alzheimer's disease. Nature 500, 4550.
Skaper, S.D., 2007. The brain as a target for inammatory processes and neuroprotective
strategies. Ann. N. Y. Acad. Sci. 1122, 2334.
Smoot, M.E., Ono, K., Ruscheinski, J., Wang, P.L., Ideker, T., 2011. Cytoscape 2.8: new fea-
tures for data integration and network visualization. Bioinformatics 27, 431432.
Soleman, S., Filippov, M.A., Dityatev, A., Fawcett, J.W., 2013. Targeting the neural extracel-
lular matrix in neurological disorders. Neuroscience 253, 194213.
Steffen, B.T., Steffen, L.M., Zhou, X., Ouyang, P., Weir, N.L., Tsai, M.Y., 2015. n-3 Fatty acids
attenuate the risk of diabetes associated with elevated serum nonesteried fatty
acids: the multi-ethnic study of atherosclerosis. Diabetes Care 38, 575580.
Suez, J., Korem, T., Zeevi, D., Zilberman-Schapira, G., Thaiss, C.A., Maza, O., Israeli, D.,
Zmora, N., Gilad, S., Weinberger, A., Kuperman, Y., Harmelin, A., Kolodkin-Gal, I.,
Shapiro, H., Halpern, Z., Segal, E., Elinav, E., 2014. Articial sweeteners induce glucose
intolerance by altering the gut microbiota. Nature 514, 181186.
Sullivan, P.F., Lin, D., Tzeng, J.Y., Van Den Oord, E., Perkins, D., Stroup, T.S., Wagner, M., Lee,
S., Wright, F.A., Zou, F., Liu, W., Downing, A.M., Lieberman, J., Close, S.L., 2008.
Genomewide association for schizophrenia in the CATIE study: results of stage 1.
Mol. Psychiatry 13, 570584.
Svensson, L., Aszodi, A., Reinholt, F.P., Fassler, R., Heinegard, D., Oldberg, A., 1999.
Fibromodulin-null mice have abnormal collagen brils, tissue organization, and
altered lumican deposition in tendon. J. Biol. Chem. 274, 96369647.
Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D.R., Pimentel, H., Salzberg, S.L.,
Rinn, J.L., Pachter, L., 2012. Differential gene and transcript expression analysis of
RNA-seq experiments with TopHat and cufinks. Nat. Protoc. 7, 562578.
Tsankova, N., Renthal, W., Kumar, A., Nestler, E.J., 2007. Epigenetic regulation in psychiat-
ric disorders. Nat. Rev. Neurosci. 8, 355367.
Tyagi, E., Zhuang, Y., Agrawal, R., Ying, Z., Gomez-Pinilla, F., 2015. Interactive actions of
Bdnf methylation and cell metabolism for building neural resilience under the inu-
ence of diet. Neurobiol. Dis. 73, 307318.
Virtanen, J.K., Mursu, J., Voutilainen, S., Uusitupa, M., Tuomainen, T.P., 2014. Serum
omega-3 polyunsaturated fatty acids and risk of incident type 2 diabetes in men:
the Kuopio ischemic heart disease risk factor study. Diabetes Care 37, 189196.
Voight, B.F., Scott, L.J., Steinthorsdottir, V., Morris, A.P., Dina, C., Welch, R.P., Zeggini, E.,
Huth, C., Aulchenko, Y.S., Thorleifsson, G., Mcculloch, L.J., Ferreira, T., Grallert, H.,
Amin, N., Wu, G., Willer, C.J., Raychaudhuri, S., Mccarroll, S.A., Langenberg, C.,
Hofmann, O.M., Dupuis, J., Qi, L., Segre, A.V., Van Hoek, M., Navarro, P., Ardlie, K.,
Balkau, B., Benediktsson, R., Bennett, A.J., Blagieva, R., Boerwinkle, E., Bonnycastle,
L.L., Bengtsson Bostrom, K., Bravenboer, B., Bumpstead, S., Burtt, N.P., Charpentier,
G., Chines, P.S., Cornelis, M., Couper, D.J., Crawford, G., Doney, A.S., Elliott, K.S.,
Elliott, A.L., Erdos, M.R., Fox, C.S., Franklin, C.S., Ganser, M., Gieger, C., Grarup, N.,
Green, T., Grifn, S., Groves, C.J., Guiducci, C., Hadjadj, S., Hassanali, N., Herder, C.,
Isomaa, B., Jackson, A.U., Johnson, P.R., Jorgensen, T., Kao, W.H., Klopp, N., Kong, A.,
Kraft, P., Kuusisto, J., Lauritzen, T., Li, M., Lieverse, A., Lindgren, C.M., Lyssenko, V.,
Please cite this article as: Meng, Q., et al., Systems Nutrigenomics Reveals Brain Gene Networks Linking Metabolic and Brain Disorders,
EBioMedicine (2016), http://dx.doi.org/10.1016/j.ebiom.2016.04.008
Marre, M., Meitinger, T., Midthjell, K., Morken, M.A., Narisu, N., Nilsson, P., Owen,
K.R., Payne, F., Perry, J.R., Petersen, A.K., Platou, C., Proenca, C., Prokopenko, I.,
Rathmann, W., Rayner, N.W., Robertson, N.R., Rocheleau, G., Roden, M., Sampson,
M.J., Saxena, R., Shields, B.M., Shrader, P., Sigurdsson, G., Sparso, T., Strassburger, K.,
Stringham, H.M., Sun, Q., Swift, A.J., Thorand, B., et al., 2010. Twelve type 2 diabetes
susceptibility loci identied through large-scale association analysis. Nat. Genet. 42,
579589.
Xiong, H.Y., Alipanahi, B., Lee, L.J., Bretschneider, H., Merico, D., Yuen, R.K., Hua, Y.,
Gueroussov, S., Najafabadi, H.S., Hughes, T.R., Morris, Q., Barash, Y., Krainer, A.R.,
Jojic, N., Scherer, S.W., Blencowe, B.J., Frey, B.J., 2015. RNA splicing. The human splic-
ing code reveals new insights into the genetic determinants of disease. Science 347,
1254806.
Xu, T., Bianco, P., Fisher, L.W., Longenecker, G., Smith, E., Goldstein, S., Bonadio, J., Boskey,
A., Heegaard, A.M., Sommer, B., Satomura, K., Dominguez, P., Zhao, C., Kulkarni, A.B.,
Robey, P.G., Young, M.F., 1998. Targeted disruption of the biglycan gene leads to an
osteoporosis-like phenotype in mice. Nat. Genet. 20, 7882.
Yan, J., Zhang, H., Yin, Y., Li, J., Tang, Y., Purkayastha, S., Li, L., Cai, D., 2014. Obesity- and
aging-induced excess of central transforming growth factor-beta potentiates diabetic
development via an RNA stress response. Nat. Med. 20, 10011008.
Yang, X., Deignan, J.L., Qi, H., Zhu, J., Qian, S., Zhong, J., Torosyan, G., Majid, S., Falkard, B.,
Kleinhanz, R.R., Karlsson, J., Castellani, L.W., Mumick, S., Wang, K., Xie, T., Coon, M.,
Zhang, C., Estrada-Smith, D., Farber, C.R., Wang, S.S., Van Nas, A., Ghazalpour, A.,
Zhang, B., Macneil, D.J., Lamb, J.R., Dipple, K.M., Reitman, M.L., Mehrabian, M., Lum,
P.Y., Schadt, E.E., Lusis, A.J., Drake, T.A., 2009. Validation of candidate causal genes
for obesity that affect shared metabolic pathways and networks. Nat. Genet. 41,
415423.
Yang, X., Zhang, B., Molony, C., Chudin, E., Hao, K., Zhu, J., Gaedigk, A., Suver, C., Zhong, H.,
Leeder, J.S., Guengerich, F.P., Strom, S.C., Schuetz, E., Rushmore, T.H., Ulrich, R.G.,
Slatter, J.G., Schadt, E.E., Kasarskis, A., Lum, P.Y., 2010. Systematic genetic and genomic
analysis of cytochrome P450 enzyme activities in human liver. Genome Res. 20,
10201036.
Yasuda, Y., Hashimoto, R., Ohi, K., Yamamori, H., Fujimoto, M., Umeda-Yano, S., Fujino, H.,
Fukunaga, M., Horiguchi, M., Takeda, M., Ichinose, H., 2014. A functional polymor-
phism of the GTP cyclohydrolase 1 gene predicts attention performance. Neurosci.
Lett. 566, 4649.
Zhang, B., Gaiteri, C., Bodea, L.G., Wang, Z., Mcelwee, J., Podtelezhnikov, A.A., Zhang, C., Xie,
T., Tran, L., Dobrin, R., Fluder, E., Clurman, B., Melquist, S., Narayanan, M., Suver, C.,
Shah, H., Mahajan, M., Gillis, T., Mysore, J., Macdonald, M.E., Lamb, J.R., Bennett,
D.A., Molony, C., Stone, D.J., Gudnason, V., Myers, A.J., Schadt, E.E., Neumann, H.,
Zhu, J., Emilsson, V., 2013. Integrated systems approach identies genetic nodes
and networks in late-onset Alzheimer's disease. Cell 153, 707720.
Zhao, Y., Barrere-Cain, R.E., Yang, X., 2015. Nutritional systems biology of type 2 diabetes.
Genes Nutr. 10, 481.
Zhu, J., Lum, P.Y., Lamb, J., Guhathakurta, D., Edwards, S.W., Thieringer, R., Berger, J.P., Wu,
M.S., Thompson, J., Sachs, A.B., Schadt, E.E., 2004. An integrative genomics approach
to the reconstruction of gene networks in segregating populations. Cytogenet. Ge-
nome Res. 105, 363374.
Zhu, J., Zhang, B., Smith, E.N., Drees, B., Brem, R.B., Kruglyak, L., Bumgarner, R.E., Schadt,
E.E., 2008. Integrating large-scale functional genomic data to dissect the complexity
of yeast regulatory networks. Nat. Genet. 40, 854861.
Am J Physiol Endocrinol Metab 299: E685E694, 2010.
First published September 7, 2010; doi:10.1152/ajpendo.00283.2010. Review

Fructose: a highly lipogenic nutrient implicated in insulin resistance, hepatic

steatosis, and the metabolic syndrome
Mark J. Dekker,1 Qiaozhu Su,1 Chris Baker,1 Angela C. Rutledge,1,2 and Khosrow Adeli1,2
1
Molecular Structure and Function, Research Institute, The Hospital for Sick Children; and 2Department of Biochemistry,
University of Toronto, Toronto, Ontario, Canada
Submitted 10 May 2010; accepted in final form 31 August 2010

Dekker MJ, Su Q, Baker C, Rutledge AC, Adeli K. Fructose: a highly

lipogenic nutrient implicated in insulin resistance, hepatic steatosis, and metabolic
syndrome. Am J Physiol Endocrinol Metab 299: E685E694, 2010. First published
September 7, 2010; doi:10.1152/ajpendo.00283.2010.As dietary exposure to
fructose has increased over the past 40 years, there is growing concern that high
fructose consumption in humans may be in part responsible for the rising incidence
of obesity worldwide. Obesity is associated with a host of metabolic challenges,
collectively termed the metabolic syndrome. Fructose is a highly lipogenic sugar
that has profound metabolic effects in the liver and has been associated with many
of the components of the metabolic syndrome (insulin resistance, elevated waist
circumference, dyslipidemia, and hypertension). Recent evidence has also uncov-
ered effects of fructose in other tissues, including adipose tissue, the brain, and the
gastrointestinal system, that may provide new insight into the metabolic conse-
quences of high-fructose diets. Fructose feeding has now been shown to alter gene
expression patterns (such as peroxisome proliferator-activated receptor-! coactiva-
tor-1"/# in the liver), alter satiety factors in the brain, increase inflammation,
reactive oxygen species, and portal endotoxin concentrations via Toll-like recep-
tors, and induce leptin resistance. This review highlights recent findings in fructose
feeding studies in both human and animal models with a focus on the molecular and
biochemical mechanisms that underlie the development of insulin resistance,
hepatic steatosis, and the metabolic syndrome.
hyperlipidemia; inflammation; gene expression; triglyceride; intestine

AS THE GLOBAL INCIDENCE OF OBESITY continues to increase, the
search to identify the dietary components that contribute to this
phenomenon is intense and ongoing. A controversial topic to
be sure, fructose has garnered considerable attention in the last
decade or so as a possible contributor to this worldwide rise in
overweight and obesity rates (131). In addition to contributing
to the caloric overconsumption and the associated energy
imbalance that are currently rampant in developed countries,
the unique metabolism of fructose may hold intriguing insight
into why fructose has been implicated in the etiology of a
number of metabolic diseases (22, 126). It is, of course,
important to note that the human diet rarely (if ever) encounters
fructose as a single nutrient. In most cases, the dietary exposure
to fructose comes through the coingestion of glucose via
sucrose (glucose and fructose) or industrial blends of fructose
and glucose (high-fructose corn syrup, HFCS) in ratios very
similar to sucrose (50% fructose, 50% glucose). There is a
large amount of interest in the metabolic syndrome (MetS), the
collection of obesity-related risk factors associated with insulin
resistance (32), in experimental and clinical obesity research.
Initially, fructose drew great interest as a potentially beneficial
sweetener for patients with diabetes mellitus due to its much
Address for reprint requests and other correspondence: K. Adeli, The
Hospital for Sick Children, Rm 3652, 555 University Ave., Toronto, ON, M5G
1X8 Canada (e-mail: khosrow.adeli@sickkids.ca).
http://www.ajpendo.org 0193-1849/10 Copyright 2010 the American Physiological Society
lower glycemic index compared with glucose (14, 60). An
additional characteristic of fructose that might have been well
suited for diabetic patients is that fructose does not stimulate
insulin secretion or require insulin for the initial steps of its
hepatic metabolism. Following glycolysis to pyruvate, fructose
carbon is a substrate for pyruvate dehydrogenase (PDH), an
enzyme metabolically regulated by insulin. As insulin secre-
tion is not stimulated by fructose, substantial amounts of
fructose are anaerobically metabolized to lactate (1, 132).
However, fructose has been cast in a much more pejorative
light in recent years.
The contribution of fructose (and particularly HFCS) to the
increased incidences of obesity and the MetS is under consid-
erable scrutiny and is hotly debated. Critical evaluations of the
literature have emerging as being both for (35) and against (42)
dietary restriction. It is important to note that, in many of the
studies outlining the biological pathways of fructose, high
levels of pure fructose (60% of diet) have been utilized. The
outcomes observed are not necessarily applicable to the
amount of fructose currently encountered in the human diet,
particularly that fructose is typically consumed concomitantly
with glucose. Nonetheless, fructose consumption in human
subjects has been linked to each of the MetS identifying
features, which include dyslipidemia (69, 126), visceral adi-
posity (126), insulin resistance (126), and high blood pressure
(for an extensive review, see Ref. 131). Although calorically
E685
Review
E686 METABOLIC EFFECTS OF FRUCTOSE
identical to glucose, fructose metabolism (hepatic metabolism
in particular) differs from that of glucose in several important
ways. The rate of hepatic uptake of fructose from portal
circulation is greater than the rate of glucose uptake, and
because fructose metabolism bypasses phosphofructokinase,
fructose metabolism is not under the regulatory control of
insulin (140). An underinvestigated aspect of fructose is the
possibility that glucose, when consumed along with fructose,
may facilitate fructose uptake and vice versa, as suggested by
plasma glucose excursions measured during the consumption
of glucose-fructose-, sucrose-, and HFCS-sweetened beverages
(124). As lipids are strongly linked to the development of
obesity and insulin resistance (2, 21, 33), the prevailing hy-
potheses concerning the mechanisms by which fructose pro-
motes MetS focus on the lipogenic nature of the sugar. The
induction of lipogenesis would increase deposition of triglyc-
eride (TG) in adipose tissue (obesity) and ectopic tissues such
as liver (hepatic steatosis) and muscle, eventually resulting in
impaired insulin signaling and dyslipidemia (125). Fructose
feeding has therefore been historically utilized as a model for
studying various aspects of hepatic dyslipidemia and insulin
resistance (6, 13). The results of a recent study suggest that
fructose consumption may specifically promote lipid deposi-
tion in visceral adipose tissue, particularly in men, whereas
glucose consumption appears to favor lipid deposition in sub-
cutaneous adipose tissue (126). Interestingly, 24-hour plasma
TG profiles are increased to a greater extent following the
consumption of fructose- and HFCS-sweetened beverages
compared with glucose consumption (124). This suggests that
the coingestion of fructose with glucose may also elicit an
unfavorable TG profile similar to that of fructose alone. Al-
though lipogenesis is certainly at the core of the biochemical
changes induced by fructose, emerging evidence suggests that
the simultaneous stimulation of alternate physiological and
Fig. 1. Whole body effects of chronic fructose consumption.
Disturbances occur in multiple tissues, including liver, adipose,
the gastrointestinal system, and the central nervous system,
following chronic fructose consumption. This constellation of
abnormalities influences multiple aspects of the metabolic syn-
drome, including dyslipidemia, insulin resistance, and central
adiposity. Additional emerging complications, such as impaired
satiety, increased hepatic lipid deposition, inflammation, and
altered gastrointestinal integrity may be mechanistically respon-
sible for the clinical alterations observed with fructose consump-
tion [images taken from ADAM Education (http://adameduca-
tion.com)].
AJP-Endocrinol Metab VOL
signaling pathways may also contribute to the deleterious
effects of fructose (Fig. 1). The goals of this review are to
highlight emerging concepts in fructose metabolism, link these
concepts to current knowledge of MetS, and identify the truly
exciting areas with substantial research potential.
Fructose Takes Aim at the Liver
The liver is an essential organ for the maintenance of lipid,
glucose, and hormonal homeostasis. As such, the liver is at the
crossroads of metabolic health and disease. Since the identifi-
cation of sterol regulatory element-binding proteins (SREBPs)
by Browns and Goldsteins groups as transcription factors that
modulate lipid homeostasis (59, 145), there has been a consid-
erable focus on gene expression in the regulation of intracel-
lular carbohydrate and lipid concentrations. Fructose may ac-
tivate SREBP-1c independently of insulin, which activates
genes involved in de novo lipogenesis (DNL) (38, 81, 93).
When glucose is replaced with fructose, carbohydrate regula-
tory element-binding protein (ChREBP) activity and nuclear
SREBP-1 are increased following two weeks of dietary treat-
ment (66). Prolonged consumption of fructose, but not glucose,
also increased hepatic fractional DNL in humans when mea-
sured during energy-balanced feeding (126). The increased rate
of fructose-induced DNL generates fatty acids that can then be
incorporated into hepatic TG or other lipid species. Increased
hepatic lipid levels are associated with increased very-low-
density lipoprotein (VLDL) synthesis and secretion, specifi-
cally that of VLDL1 (4). Apolipoprotein B-100 (apoB) is
essential for the intracellular assembly of TG into VLDL, and
apoB degradation is reduced when hepatic lipid is increased
(100), which leads to the accumulation of apoB in the hepatic
endoplasmic reticulum (ER) and causes ER stress (130). Re-
cent experimental evidence has demonstrated that ER stress
299 NOVEMBER 2010 www.ajpendo.org
 Review
METABOLIC EFFECTS OF FRUCTOSE E687
promotes SREBP-1c activation and thus contributes to DNL observed with fructose feeding likely create a scenario ideal for
(63, 130). Reduction of ER stress markers, via overexpression lipid overloading, leading to impaired insulin sensitivity and
of chaperone glucose-regulated protein-78 (GRP78) in the hyperlipidemia.
livers of ob/ob mice, inhibited SREBP-1c cleavage and the
expression of SREBP-1c and SREBP-2 target genes. Further- Fructose-Induced Disturbances in Circulating Lipid
more, hepatic TG and cholesterol levels were reduced and and Lipoprotein Clearance
insulin sensitivity improved in GRP78-injected mice (63).
The significantly lower postprandial postheparin lipoprotein
Fructose feeding has also been shown to induce the activa-
lipase (LPL) activity in subjects consuming fructose suggested
tion of ChREBP and increase the expression of lipogenic genes
that reduced TG clearance also contributes to fructose-induced
such as fatty acid synthase (FAS), acyl coenzyme-A carboxy-
postprandial hypertriglyceridemia (126). Both reduced post-
lase (ACC), and stearoyl coenzyme-A desaturase-1 (108).
prandial exposure to insulin (132) and decreased insulin sen-
Interestingly, statin treatment in fructose-fed rats with statin
sitivity (83) have likely contributed to a lowered postprandial
(atorvastatin) completely attenuated fructose-induced increases
LPL activity in subjects consuming fructose compared with
in circulating TG and free fatty acids (FFA) and normalized
those consuming glucose (126). Subcutaneous adipose tissue is
elevated blood pressure (108). This may be due to activation of
more sensitive to the effects of insulin in activating LPL than
protein kinase A (PKA) by atorvastatin, resulting in the se-
visceral adipose tissue (44); thus, the differential LPL re-
questering of ChREBP away from the nucleus, and the activa-
sponses may direct fat deposition toward subcutaneous adipose
tion of carnitine palmitoyltransferase (CPT I) (108). The net
tissue following glucose consumption and toward visceral
result of PKA activation by statin treatment could therefore be
adipose tissue following fructose exposure (126). Apolipopro-
seen to promote the oxidation of fatty acids over synthesis and
tein CIII (apoCIII) is an inhibitor of LPL and hepatic lipase and
storage of TG, resulting in decreased steatosis and lipid inter-
plays a pivotal role in the hydrolysis and clearance of TG-rich
mediates and enhanced insulin signaling (108).
particles such as VLDL and chylomicrons (117). In vivo
Central abdominal obesity (as measured by waist circumfer-
studies with high-fructose feeding of hamsters demonstrated
ence) is a core feature of the MetS (95). Under experimental
that fructose stimulated forkhead box O1 (FOXO1) produc-
conditions, human consumption of beverages containing fruc-
tion and promoted its nuclear redistribution in liver (130).
tose, rather than glucose, was associated with increased vis-
This, in turn, augmented apoCIII production and impaired TG
ceral obesity (126). In humans and rodents, mesenteric adipose
hydrolysis (130), highlighting a possible mechanism for im-
tissue is the visceral adipose tissue depot that is situated in an
paired hepatic TG-rich lipoprotein clearance following fructose
important anatomic location to participate in the portal drain-
feeding.
age of the gastrointestinal system. Thus, during periods of
North Americans typically spend 18 hours each day in what
caloric excess, disturbances in this adipose tissue depot such as
is known as the postprandial period (47). Elevated postprandial
increased lipolysis of endogenous lipid stores would contribute
TG concentrations have been increasingly associated with
to the delivery of FFA (97) and inflammatory mediators (41)
proatherogenic conditions (12, 61, 78, 98, 123). This link may
directly to the liver. Collectively, this has been termed the
be due to lipoprotein remodeling induced by increased levels of
portal theory of insulin resistance (16, 31). Metabolically,
VLDL1 and mediated by cholesteryl ester transfer protein and
mesenteric fat has a high rate of lipolysis and low response to
hepatic lipase, which results in increased concentrations of
the inhibition of lipolysis by insulin (91). In fact, dysfunction
small dense low-density lipoproteins (LDL) and remnant-like
in lipolysis and expression of lipolytic genes have been dem-
lipoproteins (3, 17, 67, 80, 101). Small dense LDL particles are
onstrated in obese diabetic subjects (144).
more easily oxidized than larger LDL particles (102); accord-
It has been suggested that fructose consumption promotes
ingly, subjects consuming fructose also have been shown to
development of MetS through increased adiposity and insulin
have significantly increased concentrations of oxidized LDL.
resistance in adipose tissue, which lead to increased circulating
Fructose induced a remarkable doubling in both fasting and
and portal levels of FFA (111). The resultant increase in
postprandial small dense LDL concentrations in subjects with
hepatic FFA uptake increases hepatic lipid availability and
MetS compared with non-MetS subjects, which suggests that
promotes hepatic insulin resistance (16). However, Stanhope et
preexisting hypertriglyceridemia can exacerbate fructose-in-
al. (126) recently demonstrated that human subjects consuming
duced lipoprotein remodeling (126).
fructose had significantly higher rates of hepatic DNL than
those consuming glucose, accompanied by an elevated 23-hour Role Of !-Oxidation and Inflammatory Signaling in
postprandial TG level, altered lipid profile, lipoprotein remod- Fructose-Induced Insulin Resistance
eling, and decreased insulin sensitivity, while they maintained
similar concentrations of circulating FFA. The absence of an Peroxisome proliferator-activated receptor (PPAR)! coacti-
effect of fructose on systemic FFA suggests that fructose may vator-1# (PGC-1#) has been reported to coactivate its partners
promote insulin resistance by providing a more direct source of through augmentation of their transcriptional activity (76).
intrahepatic lipid via DNL (125). It is important to note that Though initially described to pair with PPAR!, the number of
FFA are not the only biologically relevant product of adipose partners for PGC-1 transcription factors is rapidly growing.
tissue, as will be discussed later. Hepatic DNL limits fatty acid Utilization of a PGC-1# antisense oligonucleotide (ASO) to
oxidation in the liver via production of malonyl coenzyme-A, knock down the expression of PGC-1# has been shown to
which reduces the entry of fatty acids into the mitochondria improve the metabolic phenotype induced by fructose feeding
(82) by inhibiting CPT I. Thus, fructose-induced DNL may by reducing expression of SREBP-1 and downstream lipogenic
increase hepatic lipid by supplying endogenous fatty acids. genes in the liver. PGC-1# ASO treatment reversed hepatic
Taken together, the increased visceral adiposity and DNL insulin resistance induced by fructose in both basal and insulin-

AJP-Endocrinol Metab VOL 299 NOVEMBER 2010 www.ajpendo.org

Review
E688 METABOLIC EFFECTS OF FRUCTOSE
stimulated states (94). This important study identified differ-
ential effects of PGC-1# ASO on the liver in regular-chow-fed
rats and high-fructose-fed rats. In regular-chow-fed rats,
PGC-1# ASO induced slight but statistically significant reduc-
tions in the sensitivity of hepatic glucose production to the
suppressive effects of insulin (94). The reduced insulin sensi-
tivity was due to decreased mitochondrial fatty acid oxidation
and the accumulation of hepatic diacylglycerol (DAG), which
in turn inhibited insulin signaling via activation of protein
kinase C$ (114). Metabolism in high-fat diet-fed rats was
unaffected when treated with the same PGC-1# ASO (94).
This evidence suggests that fructose may modulate many of its
metabolic effects, including stimulation of DNL, via activation
of specific metabolic targets by PGC-1#.
PGC-1" and -# have been reported to coactivate both
PPAR" and nuclear respiratory factor (NRF)-1 (115). PPAR"
plays a key role in the transcriptional control of genes encoding
mitochondrial fatty acid oxidation enzymes such as long-chain
acyl coenzyme-A dehydrogenase and medium-chain acyl co-
enzyme-A dehydrogenase. Fructose administration (10% wt/
vol) in the drinking water of rats reduced the transactivating
and transrepressing activity of PPAR" (109). The mRNA
levels of PPAR" target genes, such as CPT I (51) and PPAR"
itself (45), were decreased as a result of the reduced transcrip-
tional activity of PPAR". Decreased PPAR" activity compro-
mises fatty acid oxidation in mitochondria, a catabolic process
directly controlled by CPT I (51). The accumulation of FFA in
liver may then induce oxidative stress (109). The same study
also observed an increased activity of proinflammatory tran-
scription factor nuclear factor-%B (NF-%B). These changes
were not observed in rats given glucose in their drinking water.
The leptin-signal transducer and activator of transcription-3
pathway may be involved in this regulation (109).
Extrahepatic Effects of Fructose Feeding
In addition to well-characterized hepatic effects of fructose
feeding (such as steatosis, increased lipoprotein secretion,
hepatic insulin resistance), fructose has been identified to alter
biological pathways in several extrahepatic tissues, including
the central nervous system, adipose tissue, and the gut. These
new areas of research have expanded the physiological role of
high-fructose diets and now introduce a whole body approach
to investigating the metabolic fate of fructose.
Central Effects of Fructose
The role of insulin resistance in neuronal dysfunction has
been well documented, and the observation that fructose can
induce these changes is also now well documented. Several
studies have linked insulin resistance and type 2 diabetes to
deficits in hippocampal declarative memory (28, 86, 128), and
Stranahan et al. (129) have shown that mice on hypercaloric
diets have impaired memory, reduced dendritic spine density,
and impaired long-term potentiation. When fructose-specific
neuronal effects are evaluated, fructose-fed rats and hamsters
show hippocampal insulin resistance associated with impaired
memory retention and decreased long-term depression forma-
tion in hippocampal neurons (87, 110). These effects are likely
the result of systemic insulin resistance, as insulin is known to
play a key role in neuronal synaptic plasticity (96).
AJP-Endocrinol Metab VOL 299 NOVEMBER 2010
The effects of fructose on the central nervous system may be
more far reaching than its ability to elicit insulin resistance.
Recent studies have shown a more direct effect of fructose on
the brain and in the general function of neurons. The fructose
transporter GLUT5 showed increased mRNA expression in
brains of fructose-fed rats, accompanied by increased protein
levels of GLUT5 in the hippocampus (120), implying that
neuronal fructose uptake can be increased with fructose feed-
ing. Lindqvist et al. (77) have reported increased ghrelin as
well as leptin levels in fructose-fed rats. Increased circulating
leptin may relate to development of leptin resistance upon
high-fructose feeding (118). This would imply that despite
increases in leptin levels there is a decreased ability to respond
to leptin, leading to decreased satiety. This observation implies
the potential for increased food intake with fructose feeding. It
should be noted, however, that the changes in ghrelin and
leptin are not excusive to the one sugar, as both glucose and
sucrose overconsumption elicited similar increases in both
leptin and ghrelin. In addition to these observations, central
administration of fructose can cause increased appetite as well
as significant changes in neuronal function. Hyperphysiologi-
cal doses of fructose administered directly to the brain have
been shown to induce increased feeding, whereas glucose
causes a decrease in food consumption (89). This was found to
be associated with differential induction of malonyl-CoA be-
tween the two sugars. Malonyl-CoA is known to cause an
anorectic effect, which leads to the loss of appetite. Glucose
administration in the brain activates the glycolytic cycle, which
leads to increased neuronal ATP levels; this is associated with
a fall in AMP levels. Since AMP is an activator of AMP kinase
(AMPK), there is a decrease in AMPK phosphorylation/activ-
ity. AMPK activity, in turn, catalyzes the phosphorylation/
activation of acetyl-CoA carboxylase (ACC), glucose blocks
this activation and this leads to increased malonyl-CoA and
decreased food intake. Fructose administered centrally de-
creases ATP levels and leads to a subsequent decrease in
hypothalamic malonyl-CoA levels (27, 143). This is thought to
lead to a decrease in satiety. In contrast to the above obser-
vations, fructose given as a pre-meal load showed no dif-
ference in food intake compared with either glucose or
sucrose pre-meal loads (5), and all three sugars, when given
to rats in drinking water, showed the same effects on food
consumption (77).
While direct central administration of fructose may not be a
typical physiological condition, as dietary fructose is rapidly
taken up and metabolized by the liver, the signaling changes
induced by fructose are intriguing and may point to larger
systemic signaling changes that could lead to alterations in
satiety factors. For example, fructose feeding has a well-
documented effect of dramatically increasing hepatic lactate
production (1, 132). As neurons utilize lactate as well as
glucose as energy substrates, increased lactate and reduced
insulin responses seem likely signals from the periphery to the
brain conveying metabolic status following fructose intake.
Evidence for Fructose-Induced Leptin Resistance
Leptin is a well-characterized adipose-derived hormone with
central and peripheral effects on food intake and dietary sub-
strate handling (for review, see Ref. 20). Leptin receptor
knockout mice (db/db) and mice that secrete mutant leptin
www.ajpendo.org

METABOLIC EFFECTS OF FRUCTOSE
peptides (ob/ob) are well characterized for a phenotype that
includes obesity, hepatic steatosis, and insulin resistance (146).
Leptin stimulates fatty acid oxidation through the activation of
PPAR" (71) through the action of AMPK (90). However,
obese human subjects (116) and diet-induced obese mice (43)
and rats (141) demonstrate resistance to leptin, which has been
thought to contribute to lipid deposition in the liver (71) and
skeletal muscle (127). Acutely, fructose feeding decreases
leptin secretion in human subjects (132). As insulin-mediated
glucose metabolism has been shown to regulate leptin release
(92), this phenomenon can likely be attributed to a lack of
insulin response with fructose consumption. Importantly, four
weeks of fructose feeding induced hyperleptinemia in humans
(68). Along with decreased insulin sensitivity, hepatic steato-
sis, and increased circulating TG, rats fed fructose also exhib-
ited hyperleptinemia (75, 109, 138), making the fructose-fed
rat a useful tool for studying leptin resistance. These apparently
conflicting responses appear to reflect a difference in acute vs.
chronic exposure to fructose, where the development of leptin
resistance and hyperleptinemia is likely due to increased adi-
pose tissue mass induced by fructose feeding (54). The major
pathway proposed to be responsible for leptin resistance in-
volves the signaling molecule suppressor of cytokine signaling
3 (SOCS3). Fructose feeding induced SOCS3 expression and
impaired serine/threonine phosphorylation, resulting in leptin
resistance (75, 138). In addition, fructose can stimulate the
expression of protein tyrosine phosphatase-1B (75). Vila et al.
(138) observed that rats fed fructose exhibited impaired c-Jun
NH2-terminal kinase (JNK) and mitogen-activated protein ki-
nase signaling and increased expression of FOXO1 due to
SOCS3 expression. In turn, this led to decreased PPAR",
suggestive of impaired fatty acid oxidation, which would
contribute to TG accumulation in the liver. Conversely, the
activation of PPAR" reversed leptin resistance in fructose-fed
rats (75). Fructose feeding also increased hepatic ceramide
concentrations, leading Vila et al. to suggest that incomplete
fatty acid oxidation due to PPAR" impairments provided
substrate for ceramide synthesis (138). This could also lead to
activation of protein phosphatase-2A and thus contribute to
deficiencies in leptin signaling and exacerbate metabolic dis-
ease (138). Although hyperleptinemia may be a result of
increased adipose tissue mass, it appears that leptin resistance
precedes increased adiposity, elevated circulating leptin, and
changes in glucose metabolism in rats (118), which suggests
that leptin resistance may be an early hallmark of fructose
feeding-induced metabolic dysfunction.
Fructose-Induced Inflammation
Low-grade inflammation is now recognized as a common
feature of the metabolic abnormalities observed in obesity (31,
56, 107). TNF" is increased in obesity and has been exten-
sively characterized for its role in insulin resistance (36, 135).
The discoveries that increased adipose-derived TNF" in obe-
sity contributed to insulin resistance (57) and that a lack of
functional TNF" protected obese mice from diet-induced in-
sulin resistance (137) have stimulated substantial research in
diabetes and insulin resistance. Increased plasma concentra-
tions of TNF" have been observed following fructose feeding
in mice (122) and hamsters (136). Furthermore, TNF" mRNA
was increased in hepatic tissues in fructose-fed mice (53, 122).
AJP-Endocrinol Metab VOL 299 NOVEMBER 2010
Review
E689
The activation of inflammatory pathways by fructose feed-
ing can have a direct influence on hepatic and intestinal
secretion of lipoproteins. Classical inflammatory pathways,
such as NF-%B, are increased with fructose feeding (109) and
contribute to hepatic TG overproduction (136). To determine
the mechanism that underlies the dyslipidemic properties of
TNF", recent studies in our laboratory examined the effect of
TNF" infusion on lipoprotein and TG production from both
hepatic (103) and intestinal sources (104). In the liver, TNF"
infusion led to decreased tyrosine phosphorylation of the in-
sulin receptor and insulin signaling molecules insulin receptor
substrate-1 (IRS-1) and Akt upon insulin stimulation, indicat-
ing decreased insulin sensitivity. These changes accompanied
the increased production of VLDL1-type particles (103). Sim-
ilarly, TNF" infusion decreased tyrosine phosphorylation of
insulin receptor, IRS-1, and Akt in the intestine, leading to
increased production of apoB48-containing chylomicron par-
ticles (104). In both cases, decreased insulin sensitivity was
accompanied by increased mass and expression of microsomal
TG transfer protein, a key protein involved in the maturation
and lipid loading of TG-rich lipoproteins (103, 104).
Free radicals generated from increased oxidative stress are
linked to insulin resistance (37). It is perhaps not surprising,
then, that fructose feeding has been shown to increase oxida-
tive stress (30) and is associated with MetS in rodents, as
reviewed previously (88). Although not a substantial focus of
this review, fructose feeding in rats has been shown to increase
hydrogen peroxide generation and inflammatory markers (99),
associating fructose feeding with an additional emerging factor
associated with insulin resistance. Perhaps more convincing of
the direct effects of fructose on oxidative stress is the obser-
vation that treatment of fructose-fed rats with antioxidants
decreased reactive oxygen species generation and prevented
insulin resistance (10, 121). Oxidative stress and inflammation
converge in shared pathways connected to insulin resistance,
such as JNK1, making both stimuli important factors to con-
sider with fructose exposure.
Toll-like receptors (TLR) are important pattern recognition
receptors in the immune system that identify bacterial patho-
gens (for review, see Ref. 70). First identified in Drosophila
(7), a human analog of Toll (TLR-4, also present in rodents)
has also been described (84). TLR-4 recognizes lipopolysac-
caride (LPS) residues from endotoxic bacteria and activates the
inflammatory NF-%B pathway. Additional endogenous ligands
for TLR-4, such as ceramide (40) and FFA (119), have now
been identified. Importantly, increased TLR-4 expression has
been associated with insulin resistance (106). TLR-4 knockout
mice have been shown to be resistant to the impaired insulin
sensitivity induced by lipid infusion and are partially resistant
to the negative effects of a high-fat diet (119). Thus, it appears
that TLR-4 is at the crossroads of insulin resistance, lipids,
and inflammation (119). Recently, Spruss et al. (122) demon-
strated that TLR-4 knockout mice were resistant to fructose-
induced liver steatosis. Although fructose-fed wild-type mice
exhibited markedly increased portal LPS and MyD88 (a
marker of TLR-4 signaling), TLR-4 knockout mice fed fruc-
tose demonstrated levels similar to those of control mice. In the
liver, TNF" mRNA expression and insulin resistance were
induced in fructose-fed wild-type mice but not in TLR-4
knockout mice. The authors suggest that reductions in gut
integrity were responsible for an influx of gastrointestinal
www.ajpendo.org
Review
E690 METABOLIC EFFECTS OF FRUCTOSE
microbiota and increased LPS concentration (23, 25). This
hypothesis is strengthened by the findings that nonabsorbable
antibiotic treatment of mice consuming fructose prevented
fructose-induced increases in LPS, hepatic lipids, and hepatic
TG (15). Changes in gut microbiota are increasingly being
linked to obesity and the metabolic syndrome (for review, see
Refs. 26 and 74). As this is a rapidly expanding area of
research, future investigation into the role that fructose may
have in microbiota health and species colonization should
provide intriguing results. Although the traditional concept of
fructose-induced metabolic dysregulation has focused on lipo-
genesis and insulin regulation, these data demonstrate an al-
ternate method by which fructose induces insulin resistance,
through increased bacterial induction of an inflammatory re-
sponse accompanied by steatosis.
Finally, although there is a dramatic increase in hepatic TG
following fructose feeding, it is now well established that TG
per se are not responsible for the deleterious effects of hepatic
steatosis but that the accumulation of lipid species such as
DAG (114) and ceramide (55, 113) impair insulin signaling
and may induce inflammation. Although much of these data
were derived from high-fat feeding studies, Vila et al. (138)
have recently demonstrated that ceramides are significantly
increased in the livers of fructose-fed rats. As has already been
discussed, fructose is a potent lipogenic stimulus in the liver.
The study of lipid intermediates is an emerging area of re-
search, and more directed studies are required to unravel a
possible role for fructose.
Role of the Gut in Meditating the Fructose Response
Increased lipoprotein secretion associated with intestinal
insulin resistance has been identified as a factor strongly
associated with aspects of MetS, including central adiposity
(18, 19), insulin resistance (8, 9), and elevated fasting TG (29,
64). In fact, the gut may be a prime candidate for the conver-
gence of many of the emerging factors involved in fructose-
related disturbances. High-fructose feeding increased chylomi-
cron production and secretion (52) and induced an insulin-
resistant state characterized by impaired IRS-1 and Akt
phosphorylation in the hamster small intestine (39). The ad-
ministration of the PPAR! agonist rosiglitazone, an insulin
sensitizer, reduced the effect of fructose feeding in hamsters
(73). Interestingly, PPAR! agonist treatment may also increase
body weight and body fat distribution from visceral to subcu-
taneous adipose tissue depots (142). Further study on meta-
bolic consequences of PPAR! agonist treatment in MetS is
required. As has been previously mentioned, infusion of TNF"
increased the release of intestinal chylomicrons (104). Chylo-
microns are known to promote the intestinal absorption of LPS
(46); therefore, the increase in postprandial chylomicron levels
in insulin resistance may expose LPS to the portal circulation
(including mesenteric adipose tissue) and induce inflammation
and insulin resistance. This inflammatory environment may
propagate the secretion of TNF", as has been seen following
fructose feeding (122, 136), thus leading to enhanced chylo-
micron production and increased exposure to LPS, establishing
a vicious cycle. Conversely, treatment of fructose-fed rats with
the probiotic oligofructose, an oligosaccharide that has been
suggested to promote intestinal integrity and inflammatory
health (48, 72), blunted hepatic steatosis and circulating TG
AJP-Endocrinol Metab VOL 299 NOVEMBER 2010
concentrations (24). Intestinal inflammation, absorption of LPS
into circulation, and gastrointestinal integrity therefore repre-
sents potential contributors to the metabolic consequences of
fructose consumption.
The gut actively secretes hormones that maintain gut integ-
rity, modulate gastric emptying, and promote insulin secretion
(34). Glucagon-like peptide-1 (GLP-1) is secreted in response
to nutrient ingestion (133, 134, 139) and can abolish the
postprandial rise in TG following infusion in human subjects
(85). Consumption of a high-fructose beverage was associated
with increased TG and prolonged GLP-1 release during a
24-hour period (132). However, others have observed a de-
creased GLP-1 response in human subjects following fructose
ingestion compared with an equal amount of glucose (65).
Recent evidence from our laboratory demonstrates a role for
GLP-1 in the dyslipidemia observed following fructose feeding
of hamsters (58). Pharmaceutical inhibition of dipeptidyl pep-
tidase-4 (the enzyme responsible for GLP-1 breakdown) atten-
uated the fructose-induced rise in plasma TG, predominantly in
the VLDL fraction (58). As the postprandial response to GLP-1
is blunted in obesity (105) and type 2 diabetes (79), the
evidence from fructose feeding suggests that increased GLP-1
concentrations improve the metabolic response to a deleterious
diet. Although these data are intriguing, very few studies have
focused on fructose and gut hormone secretion. This makes it
difficult to draw a conclusion regarding the particular effects of
fructose, but it also represents an opportunity for further
investigation into potential mechanisms of action.
Very recent evidence suggests that a single bolus of fructose
initiates a positive regulatory control loop involving stimula-
tion of leptin secretion and intestinal carbohydrate transport
(112). During feeding, the stomach releases leptin (11), which
enters the intestinal lumen (49, 50). Sakar et al. (112) demon-
strated that oral fructose increased gastric juice leptin concen-
trations 15 minutes after gavage without altering circulating
leptin concentrations. When leptin was given orally at a dose
similar to the leptin concentrations in gastric juice following
fructose ingestion, increased GLUT2 and GLUT5 transport
activities and hepatic expression of SREBP-1c, ACC1, and
FAS and activation of ACC1 in the liver were observed.
Although these results have yet to be fully explained, they
highlight a diverse and integrated regulatory system that is
activated during the briefest of exposures to fructose. Taken
together, the role of fructose in the gut offers new and exciting
opportunities for future research and therapeutic target devel-
opment.
Concluding Remarks
This review outlines several emerging molecular mecha-
nisms and biochemical pathways that are stimulated following
fructose exposure. High quantities of each food item would
need to be consumed to achieve the dietary exposures repre-
sented in many of the rodent studies. However, as this review
article clearly demonstrates, dietary exposure to fructose can
potentially have substantial and profound metabolic conse-
quences that possibly predispose individuals to chronic condi-
tions such as type 2 diabetes and cardiovascular disease. A
combination of exposure to fructose as a component of com-
mon sweeteners and increased caloric intake may, in fact,
culminate in increased consumption of fructose. It is also
www.ajpendo.org

METABOLIC EFFECTS OF FRUCTOSE
important to note that in many cases it is difficult to delineate
between the effects of the fructose molecule and the impact of
increased concentrations of endogenous lipids. Carefully con-
trolled studies, such as the series of experiments by Nagai et al.
(94), that utilize high-fat and -fructose diets can provide new
and novel insight into the molecular impact of fructose intake.
Additionally, many of the deleterious effects attributed to
fructose via the consumption of sucrose and HFCS, particu-
larly in sweetened beverages, may indeed be related to the
consumption of excess calories, which has led the American
Heart Association to recommend limiting added sugar intake in
women to no more than 100 calories a day and men to no more
than 150 calories a day (62). It is imperative that, in addition to
in vivo and in vitro experiments, carefully designed dietary
exposure and dose-response studies in human subjects with
appropriate dietary controls (i.e., fat vs. glucose vs fructose) be
completed to fully elucidate the metabolic consequences of
fructose intake in humans.
GRANTS
This work is supported by operating grants to K. Adeli by the Canadian
Institutes of Health Research (CIHR) and the Heart and Stroke Foundation
(HSFO). M. J. Dekker is supported by a Research Fellowship from HSFO; Q.
Su is supported by a Research Fellowship from CIHR.
DISCLOSURES
The authors disclose no relevant financial or competing interests in the
material presented herein.
REFERENCES
1. Adams SH, Stanhope KL, Grant RW, Cummings BP, Havel PJ.
Metabolic and endocrine profiles in response to systemic infusion of
fructose and glucose in rhesus macaques. Endocrinology 149: 3002
3008, 2008.
2. Adeli K, Lewis GF. Intestinal lipoprotein overproduction in insulin-
resistant states. Curr Opin Lipidol 19: 221228, 2008.
3. Adiels M, Olofsson SO, Taskinen MR, Boren J. Overproduction of
very low-density lipoproteins is the hallmark of the dyslipidemia in the
metabolic syndrome. Arterioscler Thromb Vasc Biol 28: 12251236,
2008.
4. Adiels M, Taskinen MR, Packard C, Caslake MJ, Soro-Paavonen A,
Westerbacka J, Vehkavaara S, Hakkinen A, Olofsson SO, Yki-
Jarvinen H, Boren J. Overproduction of large VLDL particles is driven
by increased liver fat content in man. Diabetologia 49: 755765, 2006.
5. Akhavan T, Anderson GH. Effects of glucose-to-fructose ratios in
solutions on subjective satiety, food intake, and satiety hormones in
young men. Am J Clin Nutr 86: 1354 1363, 2007.
6. Allen RJ, Leahy JS. Some effects of dietary dextrose, fructose, liquid
glucose and sucrose in the adult male rat. Br J Nutr 20: 339 347, 1966.
7. Anderson KV, Jurgens G, Nusslein-Volhard C. Establishment of
dorsal-ventral polarity in the Drosophila embryo: genetic studies on the
role of the Toll gene product. Cell 42: 779 789, 1985.
8. Annuzzi G, De Natale C, Iovine C, Patti L, Di Marino L, Coppola S,
Del Prato S, Riccardi G, Rivellese AA. Insulin resistance is indepen-
dently associated with postprandial alterations of triglyceride-rich li-
poproteins in type 2 diabetes mellitus. Arterioscler Thromb Vasc Biol 24:
23972402, 2004.
9. Annuzzi G, Giacco R, Patti L, Di Marino L, De Natale C, Costabile
G, Marra M, Santangelo C, Masella R, Rivellese AA. Postprandial
chylomicrons and adipose tissue lipoprotein lipase are altered in type 2
diabetes independently of obesity and whole-body insulin resistance.
Nutr Metab Cardiovasc Dis 18: 531538, 2008.
10. Armutcu F, Coskun O, Gurel A, Kanter M, Can M, Ucar F,
Unalacak M. Thymosin alpha 1 attenuates lipid peroxidation and im-
proves fructose-induced steatohepatitis in rats. Clin Biochem 38: 540
547, 2005.
11. Bado A, Levasseur S, Attoub S, Kermorgant S, Laigneau JP, Bor-
toluzzi MN, Moizo L, Lehy T, Guerre-Millo M, Marchand-Brustel Y,
AJP-Endocrinol Metab VOL 299 NOVEMBER 2010
Review
E691
Lewin MJ. The stomach is a source of leptin. Nature 394: 790 793,
1998.
12. Bansal S, Buring JE, Rifai N, Mora S, Sacks FM, Ridker PM. Fasting
compared with nonfasting triglycerides and risk of cardiovascular events
in women. JAMA 298: 309 316, 2007.
13. Bar-On H, Stein Y. Effect of glucose and fructose administration on
lipid metabolism in the rat. J Nutr 94: 95105, 1968.
14. Bazzano LA, Li TY, Joshipura KJ, Hu FB. Intake of fruit, vegetables,
and fruit juices and risk of diabetes in women. Diabetes Care 31:
13111317, 2008.
15. Bergheim I, Weber S, Vos M, Kramer S, Volynets V, Kaserouni S,
McClain CJ, Bischoff SC. Antibiotics protect against fructose-induced
hepatic lipid accumulation in mice: role of endotoxin. J Hepatol 48:
983992, 2008.
16. Bergman RN, Kim SP, Hsu IR, Catalano KJ, Chiu JD, Kabir M,
Richey JM, Ader M. Abdominal obesity: role in the pathophysiology of
metabolic disease and cardiovascular risk. Am J Med 120: S3S8, 2007.
17. Berneis KK, Krauss RM. Metabolic origins and clinical significance of
LDL heterogeneity. J Lipid Res 43: 13631379, 2002.
18. Blackburn P, Lamarche B, Couillard C, Pascot A, Bergeron N,
Prudhomme D, Tremblay A, Bergeron J, Lemieux I, Despres JP.
Postprandial hyperlipidemia: another correlate of the hypertriglyceride-
mic waist phenotype in men. Atherosclerosis 171: 327336, 2003.
19. Blackburn P, Lamarche B, Couillard C, Pascot A, Tremblay A,
Bergeron J, Lemieux I, Despres JP. Contribution of visceral adiposity
to the exaggerated postprandial lipemia of men with impaired glucose
tolerance. Diabetes Care 26: 33033309, 2003.
20. Bluher S, Mantzoros CS. Leptin in humans: lessons from translational
research. Am J Clin Nutr 89: 991S997S, 2009.
21. Bonen A, Dohm GL, van Loon LJ. Lipid metabolism, exercise and
insulin action. Essays Biochem 42: 4759, 2006.
22. Bray GA. Fructose: should we worry? Int J Obes (Lond) 32, Suppl 7:
S127S131, 2008.
23. Brun P, Castagliuolo I, Di L V, Buda A, Pinzani M, Palu G, Martines
D. Increased intestinal permeability in obese mice: new evidence in the
pathogenesis of nonalcoholic steatohepatitis. Am J Physiol Gastrointest
Liver Physiol 292: G518 G525, 2007.
24. Busserolles J, Gueux E, Rock E, Demigne C, Mazur A, Rayssiguier
Y. Oligofructose protects against the hypertriglyceridemic and pro-
oxidative effects of a high fructose diet in rats. J Nutr 133: 19031908,
2003.
25. Cani PD, Bibiloni R, Knauf C, Waget A, Neyrinck AM, Delzenne
NM, Burcelin R. Changes in gut microbiota control metabolic endotox-
emia-induced inflammation in high-fat diet-induced obesity and diabetes
in mice. Diabetes 57: 1470 1481, 2008.
26. Cani PD, Delzenne NM. Interplay between obesity and associated
metabolic disorders: new insights into the gut microbiota. Curr Opin
Pharmacol 9: 737743, 2009.
27. Cha SH, Wolfgang M, Tokutake Y, Chohnan S, Lane MD. Differen-
tial effects of central fructose and glucose on hypothalamic malonyl-CoA
and food intake. Proc Natl Acad Sci USA 105: 1687116875, 2008.
28. Convit A. Links between cognitive impairment in insulin resistance: an
explanatory model. Neurobiol Aging 26, Suppl 1: 3135, 2005.
29. Dekker MJ, Wright AJ, Mazurak VC, Marangoni AG, Rush JW,
Graham TE, Robinson LE. Fasting triacylglycerol status, but not
polyunsaturated/saturated fatty acid ratio, influences the postprandial
response to a series of oral fat tolerance tests. J Nutr Biochem 20:
694 704, 2008.
30. Delbosc S, Paizanis E, Magous R, Araiz C, Dimo T, Cristol JP, Cros
G, Azay J. Involvement of oxidative stress and NADPH oxidase acti-
vation in the development of cardiovascular complications in a model of
insulin resistance, the fructose-fed rat. Atherosclerosis 179: 4349, 2005.
31. Despres JP, Lemieux I. Abdominal obesity and metabolic syndrome.
Nature 444: 881887, 2006.
32. Desroches S, Lamarche B. The evolving definitions and increasing
prevalence of the metabolic syndrome. Appl Physiol Nutr Metab 32:
2332, 2007.
33. Dresner A, Laurent D, Marcucci M, Griffin ME, Dufour S, Cline
GW, Slezak LA, Andersen DK, Hundal RS, Rothman DL, Petersen
KF, Shulman GI. Effects of free fatty acids on glucose transport and
IRS-1-associated phosphatidylinositol 3-kinase activity. J Clin Invest
103: 253259, 1999.
34. Drucker DJ. The biology of incretin hormones. Cell Metab 3: 153165,
2006.
www.ajpendo.org
 Whole-grain recipes, p. 8 Telomeres & omega-3s, p. 7 What to get at Starbucks, p. 12

APRIL 2010 $3.50

GST/HST INCLUDED

EB MAN
IE LI
ONN mong
BY B
o d p re ssure. A
blo
as high
n a d ia n adults h e ou t o
f ten.
fi ve C a it s n in
One out
of ties, t a t t ack
,
eir eigh o f hear
w h o r each th r r is k ts
those yo u . It acco
un
boost s ementia
e adly. It d d h
nsion is d a se , a n too ucm
H yper te r e, k id ney dise r a ilm ent. And
ilu th e
hear t fa an any o
stroke, c to r visits th
or e d o
for far m sure is
m a jo r culprit. ig h b lo o d pr e s
salt is a t of h nsion
u t 3 0 p er c en b e ll , a hyper te
s, ab o a mp od
analyse N orm C t o f B lo
In our in ta k e , says a n d p re siden
igh salt ry
due to h f C alga
e U n iv er sit y o
at th organs
ex p e r t da . d other
C a n a s , a n
P r e s s ur
e idne y atch:
e th e hear t, k e s s u r e . T he c
d a mag pr e
lt may n blood efore w
A nd s a d it s ef fec t o e d to foods b
o n d d
n d b ey eat is a
a b o ve a f th e salt we
cen t o
Photo: joanna wnuk/fotolia.com.

8 0 per .
the salt
buy the
m.
a n e e d s to cut o n p. 3
.
w h y C a nad C o n ti n u e d
d
h ow a n
Heres
 CO V E R S T O RY

H ypertension is the leading risk factor for death in the world, says a new report
from Statistics Canada.1 Thats because high blood pressure boosts your risk
of dying of a heart attack or stroke more than smoking, high cholesterol, obesity,
Thats becauseunless you live in a
society where people eat very little salt
blood pressure rises as you age. In the
U.S. Atherosclerosis Risk in Communities
or any other risk factor does. study, which followed more than 15,000
Americans aged 45 to 64, average sys-
Cutting salt would make a huge difference. The beauty of reducing salt is that it tolic blood pressure (the upper number)
jumped five points in five years.6
would reduce cardiovascular disease and reduce the number of people who are
Blood pressures drift upward as
required to take blood-pressure-lowering drugs, says hypertension authority Norm people get older and theyre exposed to
Campbell. long-term excess sodium, explains U.S.
physician Stephen Havas, a former vice
Heres why youand, more importantly, the food industryshould hold the salt. president at the American Medical As-
sociation. Thats why almost all adults
are going to get blood pressures that put
them at higher risk for heart disease and

1 2
stroke.
Less salt means lower blood Nearly everyone gets high

pressure and less disease.
Its no surprise that cutting salt lowers
blood pressure. That has been shown
in studies that compare higher- versus
lower-salt diets in both adults and chil-
dren.2,3
And a recent meta-analysis of 13
studies found not just lower blood pres-
sures, but a lower risk of heart attacks,
strokes, and other cardiovascular events
among people who cut their salt
intake.4
For example, in the U.S. Trials of
Hypertension Prevention studies, some
3
blood pressure. Risk rises before your blood
Why worry about salt if you havent been pressure is high.
diagnosed with high blood pressure?
Doctors consider prescribing drugs when
Odds are, you will be.
your blood pressure is highthat is, its
Over time, 90 per cent of people in this
at least 140 over 90 (see How High Is Too
country develop hypertension, says Norm
High?). But its a threat to your blood
Campbell, a professor of medicine, physi-
vessels before it crosses that line.
ology, and pharmacology at the University
People dont realize that blood pres-
of Calgary.
sure higher than 120 over 80 is associated
with increased risk, says Havas.
HOW HIGH IS TOO HIGH? Between normal and hyperten-
sion you have a huge number of heart
disease and stroke deaths attributable
SYSTOLIC

DIASTOLIC

If your systolic and diastolic to excess blood pressure, he explains.

2,400 people with pre-hypertension
were randomly assigned to either
cut their sodium by roughly 750 to
1,000 milligrams a day or to follow
general guidelines for healthy eating for
1 to 4 years.5
Ten to 15 years after the studies
ended, researchers found a 25 to 30
per cent lower risk of heart attacks,
strokes, or other cardiovascular events
Photo: joanna wnuk/fotolia.com.
pressures fall into different Experts call those in-between blood
categories, your risk depends pressures pre-hypertension, and 20
on the higher category. per cent of Canadians have it.
Researchers arent sure how elevated
blood pressure raises the risk of heart
Hypertension
attacks and strokes. One possibility: it
140 90
may accelerate the clogging of arteries.
Pre-hypertension The progression of atherosclerosis
is much higher in the face of hyper-
120 80 tension, explains researcher Norman

in the group that cut salt.

Normal Kaplan of the University of Texas
A decrease in sodium in the diet,
Southwestern Medical Center in Dallas.
even among those with only mod-
The heart beating at a higher
estly elevated blood pressure, lowers
pressure may lead to damage in the
risk of cardiovascular disease later in
Numbers (in millimetres of mercury) blood vessel wall, and that could allow
life, says investigator Nancy Cook,
apply to adults who arent taking cholesterol and inflammatory cells to
an associate professor of medicine at drugs to lower their blood pressure. enter.
Harvard Medical School in Boston,
Massachusetts. >>>>>
Source: www.statcan.gc.ca/pub/82-003-x/2010001/
article/11118-eng.pdf.

NUTRITION ACTION HEALTHLETTER O APRIL 2010 3

4 Hypertension harms
the heart, brain, and
kidneys.
High blood pressure doesnt just raise
the risk of heart attacks and strokes.
It also boosts the risk of heart failure,
which affects 400,000 Canadians. It
can mean that the hearts pump has
deteriorated and cant push the blood
out, says Kaplan.
And high blood pressure is a lead-
ing cause of chronic kidney disease,
which two million Canadians have
Go to sodium101.ca for the Canadian Stroke Networks
or are at risk for. Also troubling is
easy-to-swallow advice on salt and blood pressure.
the growing evidence that hyper-
tension raises the risk of dementia.
For example, in the U.S. Womens
Health Initiative Memory Study,
which took MRI brain scans of 1,400
women over age 65, those with high
blood pressure had more abnormal brain
lesions eight years later.7
Even moderately elevated blood pres-
sure is associated with silent vascular dis-
ease in the brain that contributes to risk of
dementia, conclude the studys authors.
5 Drugs havent solved the
problem.
So what if you get high blood pressure?
Cant you just take a drug to lower it?
You dont want to wait until your blood
pressure crosses that magic threshold of
140 over 90 because by that point youve
already done a fair amount of damage to
your heart, vascular system, kidneys, and
brain, says Havas.
Whats more, a third of Canadians with
hypertension still dont have their blood
pressure under control: an estimated 17 per
cent of those who have hypertension dont
know it, 4 per cent know they have it but
arent being treated, and 14 per cent are
being treated but not enough to get their
blood pressure below 140 over 90.1
Why? Hypertension is a chronic con-
dition that doesnt make the patient feel
anything, explains Kaplan. If people
with, say, rheumatoid arthritis dont take
their medication, they hurt. So theyll
take that medication. But people with
hypertension dont experience anything
obvious. So they may stop.
Doctors may also share some of the
blame. Some physicians will put pa-
tients on anti-hypertensive medication,
and then say, Okay, Im done, without
monitoring to see if the patients blood
pressure is still elevated, says Kaplan.
But eating less salt would make blood
pressure drugs more effective in those
who need them. With lower sodium
4 NUTRITION ACTION HEALTHLETTER O APRIL 2010
intakes, you see a greater fall in blood
pressure, says Kaplan. Thats particular-
ly true for diuretics, but its been shown
with other drugs as well.
In fact, researchers recently tested the
impact of a lower-salt diet on 12 people
who had resistant hypertensionthat is,
their blood pressures were still high even
though they were taking an average of
3 different drugs every day.
When doctors told them to not just take
the drugs but to also eat a lower-salt diet
(1,050 milligrams a day), their blood pres-
sures were dramatically lower (by an aver-
age of 22 over 9 points) than when they
were on a high-salt diet (5,750 mg a day).8
The authors conclusion: High dietary
salt ingestion is an important cause of
resistant hypertension.
6 Assume that youre sensi-
tive to salt.
For certain individuals who are salt sen- companies from dumping so much salt
sitive, excessive consumption of sodium
can increase blood pressure, says the U.S. ting sodium out of the food supply is the
Salt Institute, which represents the salt
industry. 9 Certain individuals?
Some people react to sodium more
quickly than others, says Havas. But
90 per cent of Canadians eventually get
high blood pressure, and a principal cause
is exposure to excess sodium.
So most people over time dont do
well with high salt loads, says Havas.
Whats more, there is no predictor or test
of salt sensitivity, he adds. So one has to
assume that almost all of us are sensitive
to long-term sodium exposure.
Thats why expert panels recommend
no more than 1,500 mg of sodium a
day if youre middle-aged or older, are
of African descent, or already have high
blood pressurea sizeable chunk of the
population. Everyone else should shoot
for 2,300 mg of sodium a day.
7 Other factors are no
excuse to ignore salt.
Cutting excess salt isnt the only
way to lower blood pressure. Get-
ting more potassium also helps,
and Canadians average far less
than the 4,700 mg people should
be eating. Eating a DASH diet
which is rich in fruits, vegetables,
and low-fat dairy foodsknocks
down blood pressure (see Nutrition
Action, Oct. 2009, cover story). So
does staying trim, daily exercise,
and limiting alcohol to no more
than two drinks a day (for men) or
one drink a day (for women).10
All of those factors affect
blood pressure, says Havas. I
dont think anyone would argue that you
should only work on one front. But cut-
ting salt is still key.
Sometimes, he adds, other factors are a
smokescreen that the food industry throws
out to confuse everybody. In fact, it may
be easier to change the populations salt
intake than anything else.
We cant get people to lose weight
and maintain the weight loss over time,
though we should try because some
people will do it, notes Havas. We cant
get people to exercise regularly and main-
tain it over time, though its worth trying
because some people will do it.
And its not easy to get people to eat
more vegetables and fruit. If the average
person eats 3 or 4 servings a day and we
need to get to 8 or 9 or more, its going to
take a long time.
In contrast, the government can stop
into packaged and restaurant foods. Get-
easiest because you can engineer that,
explains Havas. You cant engineer more
fruits and vegetables, greater weight loss,
and more exercise.
8
Salts harm goes beyond
blood pressure.
Salt appears to damage the heart and
blood vessels above and beyond its im-
pact on blood pressure.
With most animals, if you give them a
high intake of salt, their blood pressure will
go up, they develop cardiovascular disease,
and they die prematurely, explains the
University of Calgarys Norm Campbell.
But if you genetically breed animals so
you can feed them salt and they have no
increase in blood pressure, they still develop
cardiovascular disease prematurely. That
suggests that salt has direct toxic effects on
 C O V ER S T O R Y
the heart and blood vessels.
And theres evidence that salt is toxic to
humans. By and large, these human stud-
ies show increased cardiovascular event
rates on typical high-salt diets independent
of blood pressure, notes Campbell.
How might salt harm the body beyond
its impact on blood pressure?
O Left ventricular hypertrophy. High
blood pressure can thicken the muscle
in the chamber of the heart that pumps
blood throughout the body.
The presence of left ventricular thick-
ness is a very-high-risk situation because
it contributes to most cases of heart
failure, explains Kaplan.
But some studies suggest that a high-
salt diet worsens the damage caused
by high blood pressure.11 And cutting
back on salt may reverse the muscle
thickening, he adds.
O Stiff arteries. Stiff arteries are often
an early sign of heart disease. When
researchers put overweight or obese
people with normal blood pressure on a
diet with a typical sodium intake (about
3,500 mg a day), their arteries were stiffer
than when they ate a lower-salt diet
(1,150 mg a day).12
These findings suggest additional
cardioprotective effects of salt reduc-
tion beyond blood pressure reduction,
conclude the researchers.
O Kidney disease. High blood pressure
damages the kidneys. But salt may make
it worse. In some studies, people who
consume more salt excrete more protein
in their urine.13 Thats a sign that their
kidneys are under stress.
The presence of protein in the filter-
ing surfaces of the kidneys is associated
Photos: Namita Davis (packages), Stephen Schmidt (finished dishes).
with inflammation and damage, notes
Kaplan. For people who have kidney
disease, a reduction in sodium can
reduce protein in the urine.
O Osteoporosis. High-salt diets increase
calcium losses in urine. When theres
excess salt in the kidneys tubules, it
draws out the calcium, explains Kaplan.
But only a few studies have looked at
salts impact on bone.14 Stay tuned.
9 Small cuts in salt can save
lives and dollars.
The average Canadian consumes roughly
3,500 milligrams of sodium a day. What
would be saved by cutting that intake to
1,700 mg? (Thats a typical target intake
>>>>>
CO V E R S T O RY
HOW TO DEFUSE A SALT MINE
A s long as the food supply stays salt-laden, the easiest way to reach target sodium
levels is to make your own.
Okay, its not easy to bake your own raisin bran, but you can easily whip up your own
salad dressing and season your own rice, couscous, or pasta. You can also buy no-salt-
added canned beans (try Eden), tuna (try Presidents Choice), and tomatoes (try Unico).
And you can dust off your old copies of Nutrition Action and rediscover Kate Sherwoods
amazing Healthy Cook recipes, which trim sodium but not taste.
But on days when theres no way youre going to make your own, try this: add salt-free
vegetables, beans, or grains to packaged or restaurant foods.
You not only cut the salt in each serving, you boost the potassium. Whats more, you
save money by stretching costlier packaged foods or restaurant take-out. Its a win-win.
Here are a few examples:
Start With Add What You Get
Take-Out
ADD: 500 g steamed broccoli to
+ =
an order of Kung Pao chicken.
ALTERNATIVES: Add steamed or
sauted vegetables to any take-
out Chinese dish with a Szech-
uan, Hunan, garlic, or other
strong-flavoured sauce.
Packaged Grains
ADD: 1 package (250 g) Uncle
+ =
Bens Bistro Express Wholegrain
Brown Rice.
ALTERNATIVES: Add whole wheat
couscous, orzo, or bulgur to any
seasoned grains. Add whole-grain
pasta to any seasoned pasta dish.
Frozen Dishes
+ =
ADD: 250 g asparagus and 250 g
broccoli florets.
ALTERNATIVES: Add fresh (or
frozen) vegetables to any dish.
(Most packaged foods have too
few veggies.)
Indian Dishes
ADD: 1 can of unsalted canned
+ chick peas (garbanzos).
ALTERNATIVES: Add unsalted
garbanzos to kidney bean curry,
=
red lentils (they cook quickly) to
lentil curry or dal, or frozen peas
to curried vegetables.
NUTRITION ACTION HEALTHLETTER O APRIL 2010 5
 CO VE R S T OR Y
SALT ON THE MENU
in most studies, and its close to the 1,500 mg a day recommend-
Many of these typical restaurant dishes have at least a days
ed for most people.)
We estimated that it would lead to about an 8 per cent reduc- worth of sodium (1,500 mg). Tips: Ask the chef to cook yours
tion in heart attacks, a 12 per cent reduction in stroke, and a without salt. Check the chains Web site for lower-sodium op-
21 per cent reduction in heart failure, says the University of Cal-

)
(mg
tions. And go to sodium101.ca for more advice.
garys Norm Campbell. That would mean saving about 11,500

es

ium
ori
cardiovascular events per year. Food

Sod
Ca l
However, more Canadians are being treated for high blood Extreme Pita Market Fresh Veggie Pita (regular) 280 530
pressure now than when the initial estimates were made. In
Swiss Chalet Classic Chicken Breast & veg. medley 360 560
1999, we started a huge effort to improve hypertension control
Caseys Salmon Nioise Salad with dressing 760 760
and treatment in Canada, explains Campbell. Well rerun the
analyses when the new national control rates are released. Milestones Fire-Grilled Wild Coho Salmon & rice & veg. 740 780
But theres a downside to treating more people for high blood Boston Pizza Delicious Alternatives Pollo Pomodoro Linguini 510 820
pressure. Right now 48 per cent of women and 42 per cent of Boston Pizza Delicious Alternatives Spicy Garlic Chicken Pizza 650 860
men over 60 are on an anti-hypertensive drug, explains Camp- Starbucks Italian Cheese, Tomato & Pesto Sandwich 470 900
bell. Thats a success in the sense that those people are protect- Kelseys Veggie Burger 690 950
ed from the ravages of hypertension, but thats a huge portion of Mr. Greek Spanakopita 410 970
the population taking pharmacotherapy. Kelseys Broccoli Cheddar Soup 190 1,020
Cutting salt would trim the cost of that treatment, because Manchu Wok Green Bean Chicken with Mixed
lower blood pressure would curb the number of people who need Vegetables & rice 660 1,070
drugs.15 We would save between $430 and $540 million per Starbucks Reduced-Fat Chicken Bruschetta Sandwich 430 1,090
year purely related to a reduced need for hypertension treatment Kelseys Classic Sirloin Burger 780 1,140
and physician visits, says Campbell. Milestones Grilled Chicken Caesar Salad with dressing 580 1,160
Smaller cuts in sodium (say, to 2,300 mg a day rather than to
Mr. Greek Greek Salad with dressing 600 1,160
1,700 mg) would have less impact on heart attacks, strokes, and
Boston Pizza Vegetarian Pizza (individual) 680 1,160
heart failure.16 On the other hand, we only examined the effect
The Keg Grilled Top Sirloin (8 oz.) & buttered baked potato 1,400 1,230
of salt on blood pressure, not its direct effects on cardiovascular
disease and other health issues, says Campbell. So our num- Milestones Spinach & Feta Omelette 410 1,300
bers are an underestimate. Swiss Chalet Chalet Chicken Soup 170 1,380
Swiss Chalet Chicken Pot Pie 550 1,420

10 Its tough to cut salt on your own.

Im a physician, says researcher Kirsten Bibbins-Domin-
go of the University of California, San Francisco, and its always
East Side Marios Classic Lasagna
Extreme Pita Chef Inspired Chicken Souvlaki Pita (regular)
Kelseys Shepherds Pie
610 1,420
610 1,430
740 1,480
striking to me how hard it is for my patients to cut salt. Swiss Chalet Grilled Chicken Caesar Salad with dressing 680 1,550
Thats because roughly 80 per cent of the sodium consumed is Boston Pizza Pepperoni Pizza (individual) 750 1,650
added to food before we open a package or walk into a restau- Swiss Chalet Rotisserie Chicken Club Wrap 820 1,730
rant. So unless you make everythingincluding breads, crack- Montanas Cedar Roasted Salmon with rice & vegetables 780 1,780
ers, cereals, soup, pizza, spaghetti sauce, salad dressingfrom Kelseys Classic Wings (8) with medium sauce
scratch, you cant easily avoid the salt. & blue cheese dressing 1,300 1,870
You can take all the salty snacks out of your dietthe nuts Swiss Chalet Classic Chicken Breast
and the chips and everything else, but much still remains, & dipping sauce & rice 570 2,020
says Bibbins-Domingo. So many patients come to me thinking Boston Pizza Chicken Parmesan 740 2,050
theyve made healthier choices and theyre oftentimes consuming The Keg Teriyaki Sirloin (8 oz.) & buttered baked potato 1,440 2,270
the same, if not more, salt. Boston Pizza Homestyle Lasagna 930 2,280
And restaurants make grocery store salt levels look low (see Salt Swiss Chalet Chicken Stir Fry & rice 750 2,290
on the Menu). People eat more and more in restaurants, says
Montanas Roasted Turkey Club Sandwich 1,050 2,320
Havas. They have no idea how much sodium is in those foods.
Kelseys Stacked Pastrami Sandwich 650 2,500
Like the governments of the United Kingdom and Finland,
The Keg French Onion Soup 340 2,600
Ottawa could pressureand, if necessary, requirecompanies to
cut salt and could require warning labels on high-salt foods. Caseys Shrimp Pad Thai 720 2,810
In the meantime, the best advice is to make your own. Kelseys Sizzling Chicken Fajitas (4) with salsa & sour cream 910 3,190
Eat fresh fruits and vegetables and as much unprocessed food Kelseys Fully Loaded Nachos with salsa & sour cream 2,290 3,530
as possible, says Campbell. They have about a tenth the salt of Source: restaurant Web sites. Chart compiled by Namita Davis in Toronto with help
processed foods, and theyre higher in fibre and potassium and from Lauren Clark in Ottawa.
lower in sugars and saturated fat, so they make a healthier diet.

1 9 saltinstitute.org/Issues-in-focus/Food-salt-health/How-the-body-handles-salt.
Statistics Canada, Blood pressure in Canadian adults. Cat. no. 82-003-XPE, Health
10 JAMA 288: 1882, 2002.
Reports, Vol. 21, no. 1, Mar. 2010.
2 J. Human Hypertens. 16: 761, 2002. 11 Circulation 87: 476, 1993.
3 Hypertension 48: 861, 2006. 12 Am. J. Clin. Nutr. 89: 485, 2009.
4 BMJ 339: b4567, 2009. 13 J. Intern. Med. 256: 324, 2004.
5 BMJ 334: 885, 2007. 14 J. Bone Miner. Res. 23: 1477, 2008.
6 Circulation 106: 703, 2002. 15 Can. J. Cardiol. 23: 437, 2007.
7 J. Clin. Hypertens. 12: 203, 2010. 16 Can. J. Cardiol. 24: 497, 2008.
8 Hypertension 54: 475, 2009.

6 NUTRITION ACTION HEALTHLETTER O APRIL 2010

 The n e w e ng l a n d j o u r na l of m e dic i n e

Review article

Medical Progress

Salt in Health and Disease A Delicate

Balance
Theodore A. Kotchen, M.D., Allen W. Cowley, Jr., Ph.D.,
and Edward D. Frohlich, M.D.

T
he fact that salt (sodium chloride) is essential for life has From the Medical College of Wisconsin,
been recognized for millennia. Historically, the exchange value of salt played Milwaukee (T.A.K., A.W.C.); and Ochsner
Medical Center, New Orleans (E.D.F.).
an important role in establishing trade routes, securing alliances, and provok- Address reprint requests to Dr. Kotchen
ing revolutions. Homer referred to salt as a divine substance, and Plato described it at the Department of Medicine, Medical
as especially dear to the gods. Salt has been associated with sexual potency, fertil- College of Wisconsin, 8701 Watertown
Plank Rd., Milwaukee, WI 53226, or at
ity, and immortality. tkotchen@mcw.edu.
In sodium-deficient states, salt consumption is driven by salt appetite an
N Engl J Med 2013;368:1229-37.
innate and motivated behavioral response that drives a human or animal to seek DOI: 10.1056/NEJMra1212606
and ingest salt-containing foods and fluids.1 However, under usual circumstances, Copyright 2013 Massachusetts Medical Society.
the ambient salt diet is in excess of physiological need, and in humans, it has been
difficult to distinguish innate salt appetite and salt need from salt preference.2 The
hunger for salt is also influenced by taste, culture, social custom, the widespread
availability of salt, and habit independent of the need for salt.3 Despite its historical
value and physiological importance, high salt consumption has been recognized as
detrimental to health. In this article, we provide an overview of the current under-
standing of the relation of salt consumption to hypertension and cardiovascular
disease.

S A LT C ONSUMP T ION A ND A R TER I A L PR E SSUR E

A high-salt diet convincingly contributes to elevated arterial pressure in numerous

animal species, including genetic and acquired models of experimental hypertension.
People living in nonindustrial, unacculturated communities with low salt intake have
low average blood pressures that increase little with age. Blood pressure increases
when such populations adopt modern lifestyles.
Within populations, either slight but significant correlations or insignificant cor-
relations between blood pressure and dietary salt have been observed.4,5 A relatively
constricted range of sodium intakes in particular, high sodium intakes may
contribute to the underestimation of the association between blood-pressure level
and sodium intake within populations. Studies across populations provide more
convincing evidence than within-population studies of the association of salt intake
with both blood pressure and the age-related rise of blood pressure in adults.6-8
There is also a modest association between higher salt intake and higher blood
pressure in children and adolescents.9 Low dietary intake of potassium may in-
crease the effect of sodium on blood pressure, and the relationship between so-
dium and blood pressure becomes stronger if the urinary sodium:potassium ratio,
rather than simply the sodium excretion rate, is considered.7
Clinical trials provide definitive evidence of a direct cause-and-effect relation-

n engl j med 368;13 nejm.org march 28, 2013 1229

The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 1. Meta-Analyses of the Effect of Salt Reduction on Blood Pressure.*

Study Persons with Hypertension Persons with Normal Blood Pressure

Reduction Reduction Reduction Reduction Reduction Reduction

No. of No. of in Sodium in Systolic in Diastolic No. of No. of in Sodium in Systolic in Diastolic
Trials Persons Intake Pressure Pressure Trials Persons Intake Pressure Pressure
mmol/day mm Hg mmol/day mm Hg
Midgley et al.10 28 1131 95 3.7 0.9 28 2374 125 1.0 0.1
Cutler et al.11 22 1043 77 4.8 2.5 12 1689 76 2.3 1.4
Graudal et al.12 58 2161 118 3.9 1.9 56 2581 160 1.2 0.3
He and MacGregor13 17 734 78 5.0 2.7 11 2220 74 2.0 1.0

* The numbers in the sodium intake and blood pressure columns are average reductions. An amount of 17 mmol of sodium is equivalent to
400 mg of sodium or 1.0 g of sodium chloride.
The reduction in the blood-pressure value was significant.

ship between salt consumption and blood pres- not a binary, trait, depending on the methods
sure. Although meta-analyses are potentially used for assessment and the definition of salt
subject to criticism owing to variations in the sensitivity, approximately 30 to 50% of persons
inclusion and exclusion criteria of the trials and with hypertension and a smaller percentage of
other variations among the study protocols, sev- persons with normal blood pressure are thought
eral meta-analyses of randomized clinical trials to have salt-sensitive blood pressure.16 Pheno-
have consistently shown that persons with hy- types associated with salt-sensitive blood pres-
pertension have a greater response to a reduced sure include low-renin hypertension, older age,
salt intake than do persons with normal blood African American ethnicity, obesity, and the met-
pressure10-13 (Table 1). In a meta-analysis of 10 abolic syndrome.17,18
controlled trials involving a total of 966 children Blood-pressure responses to salt may be mod-
(median age, 13 years; range, 8 to 16), a 42% ified by other components of the diet. Low dietary
reduction in salt intake was associated with intakes of potassium and calcium potentiate the
small but significant reductions of both systolic salt sensitivity of blood pressure.19 Conversely,
pressure (1.17 mm Hg; 95% confidence interval high dietary intakes of potassium and calcium
[CI], 1.78 to 0.56) and diastolic pressure attenuate the development of salt-induced hyper-
(1.29 mm Hg; 95% CI, 1.94 to 0.65).14 tension in several animal models. In genetic ex-
Trials involving abrupt and severe salt restric-perimental models of hypertension, blood-pressure
tion have shown significant increases in plasma responses to salt are modulated by the protein,
renin activity, serum aldosterone, and plasma carbohydrate, and fat composition of the diet. In
levels of noradrenaline and adrenaline, total addition, the full expression of salt-sensitive hy-
cholesterol, and triglycerides.12,15 The implicationpertension depends on the concomitant intake of
is that these neural and hormonal responses may sodium and chloride, rather than sodium with
have adverse cardiovascular consequences. Stud- some other anion.20 However, in usual diets, it
ies assessing a long-term (>6 months) modest has been estimated that more than 85% of so-
reduction of salt intake have shown only small dium is consumed as sodium chloride.
increases in renin activity and little or no change Experimental models of hypertension provide
in sympathetic tone or plasma lipid levels. convincing evidence of a genetic susceptibility to
salt sensitivity. The most intensely studied experi-
S A LT SENSI T I V I T Y mental model of salt-sensitive hypertension has
OF BL O OD PR E SSUR E been the Dahl rat, developed by Lewis K. Dahl
nearly 50 years ago and inbred by John Rapp. In
Blood-pressure responses to salt are heteroge- consomic rats (in which otherwise genetically
neous and are normally distributed within popu- identical animals differ by one chromosome),
lations. Although salt sensitivity is a continuous, transfer of any one of several chromosomes from

1230 n engl j med 368;13 nejm.org march 28, 2013

The New England Journal of Medicine

Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
 Medical Progress
normotensive Brown Norway rats into Dahl salt-
sensitive rats attenuates or abolishes salt-induced
hypertension and proteinuria.21 Knockout mice
lacking genes for melanocytestimulating hor-
mone, atrial natriuretic peptide and its receptor,
the prostaglandin EP2 receptor, or the bradykinin
receptor all have salt-sensitive hypertension. The
heritability of salt-induced increases in blood
pressure may also be unrelated to genetic polymor-
phisms. In the normotensive SpragueDawley rat,
either a high or a low salt intake during pregnancy
is associated with a reduced number of renal
glomeruli and proteinuria in the offspring,22 and
animals with a reduced number of nephrons be-
come progressively more salt-sensitive with age
in terms of blood pressure.
Limited clinical data are available concerning
the heritability of salt sensitivity. As compared with
whites with normal blood pressure, blacks with
normal blood pressure have slower sodium excre-
tion after intravenous administration of a sodium
load and have greater increases in blood pressure
in response to an extremely high salt intake.23
Among both black families and white families,
the blood-pressure response to sodium loading
and sodium restriction is highly heritable.24,25 In
addition, among white twins with normal blood
pressure, there is a strong heritable influence on
plasma renin activity, the plasma aldosterone con-
centration, and the efficiency of sodium excretion
after infusion of a saline load.26
Monogenic renal tubular disorders resulting in
either sodium retention or renal sodium wasting
are associated with hypertension and hypoten-
sion, respectively.27 However, the causative mu-
tations of these monogenic syndromes of sodium
retention are not applicable to the general popu-
lation or to the vast majority of persons with
hypertension. A number of rare alleles in several
genes that alter renal salt handling are associ-
ated with blood-pressure variation in the general
population.28 Preliminary evidence in various pa-
tient populations has identified a number of DNA
polymorphisms associated with salt sensitivity in
genes that may contribute to the regulation of re-
nal sodium transport.29,30 Salt sensitivity is also
reportedly associated with single-nucleotide poly-
morphisms in at least a dozen genes that have
no apparent physiological basis for the regula-
tion of arterial pressure or sodium balance. For
the most part, these observations await confir-
mation.
n engl j med 368;13
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
S A LT C ONSUMP T ION
A ND C A R DIOVA SCUL A R DISE A SE
It has been projected that a reduction in dietary
salt intake by 3 g per day (on the basis of the cur-
rent average consumption in the United States)
would reduce the annual number of new cases of
coronary heart disease by 60,000 to 120,000,
cases of stroke by 32,000 to 66,000, and cases of
myocardial infarction by 54,000 to 99,000 and
would reduce the annual number of deaths from
all causes by 44,000 to 92,000.31 With notable
exceptions, results of observational studies gen-
erally support an association of high salt intake
with cardiovascular end points. In a 2009 meta-
analysis of 19 independent cohort samples from
13 studies involving a total of 177,025 participants
(follow-up, 3.5 to 19 years) and 11,000 cardiovas-
cular events, Strazzullo et al. reported that a high
salt intake is associated with increased risks of
stroke and total cardiovascular disease,32 although
an inverse trend with respect to the association
between salt intake and the risk of cardiovascular
disease was observed in three cohorts. Results
from several recent observational studies are con-
sistent with the overall conclusions of the meta-
analysis by Strazzullo et al.33-35
In contrast, a limited number of observational
studies have suggested either no association of
cardiovascular disease with salt intake or an in-
creased prevalence of cardiovascular disease with
low salt intake.36 On the basis of a post hoc analy-
sis of two populations enrolled in drug trials,
ODonnell et al. reported that both high and low
sodium intakes were associated with increased
cardiovascular events in a J-shaped curve (24-hour
sodium excretion was estimated on the basis of
the measured sodium concentration in a fasting
morning urine sample).37 As compared with par-
ticipants who had a baseline sodium excretion of
4 to 6 g per day (10 to 15 g per day of sodium
chloride), participants who excreted more than
6 g of sodium (15 g of sodium chloride) per day
and those who excreted less than 4 g of sodium
(10 g of sodium chloride) per day in that study
showed an increase in cardiovascular deaths,
strokes, or heart attacks. Studies with negative or
paradoxical outcomes have been criticized for a
number of methodologic deficiencies, including
confounding variables (e.g., coexisting condi-
tions and diuretic therapy) and a short duration
of follow-up.
nejm.org march 28, 2013 1231
 The n e w e ng l a n d j o u r na l
Results of epidemiologic studies and random-
ized trials suggest that potassium consumption
influences the effect of sodium on blood pres-
sure and the risk of cardiovascular disease. Low
potassium intake is associated with an increased
risk of hypertension, and a high ratio of sodium
intake to potassium intake is a more potent risk
factor for hypertension and cardiovascular dis-
ease than each factor alone.34,38 A high potassi-
um intake offers the greatest benefit when so-
dium intake is high.
All observational studies share intrinsic weak-
nesses and methodologic limitations. None were
designed to address the relation between daily so-
dium intake and the risk of cardiovascular disease.
In several long-term, prospective, randomized
clinical trials, reduced salt intake was reported to
result in a decreased incidence of cardiovascular
events.39-41 In contrast, on the basis of a meta-
analysis of seven randomized trials (involving a
total of 6250 participants) with at least 6 months
of follow-up, a 2011 Cochrane analysis conclud-
ed that reducing dietary salt intake did not de-
crease the risk of death or cardiovascular dis-
ease.42 One of the trials in the analysis included
patients with heart failure who were simultane-
ously receiving aggressive treatment with diuretic
agents. In addition, trials involving persons with
normal blood pressure and those involving per-
sons with hypertension were analyzed separately,
potentially resulting in lack of statistical power.
On the basis of a meta-analysis that excluded the
study in which patients received concomitant di-
uretic therapy and that combined the normoten-
sive and hypertensive study populations, He and
MacGregor concluded that decreased salt intake
was associated with a significant reduction in
cardiovascular events and a nonsignificant re-
duction in all-cause mortality.43
Results from trials in discrete patient popula-
tions suggest that caution is warranted in rec-
ommending rigorous sodium restriction for
specific patient groups. An observational study
involving 2807 adults with type 1 diabetes (mean
age, 39 years) showed that dietary sodium was
inversely associated with all-cause mortality and
the development of end-stage renal disease (me-
dian follow-up, 10 years).44 In that study, reduced
survival was also observed among adults with
high sodium intakes. In a related study involving
638 patients with long-standing type 2 diabetes
(mean age, 64 years), low urinary sodium excre-
1232 n engl j med 368;13
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
of m e dic i n e
tion was associated with increased all-cause and
cardiovascular mortality (median follow-up, 9.9
years).45 Notably, the patients in that study had
multiple coexisting conditions, including renal
impairment and cardiovascular disease, at base-
line. Clinical trials of a low-sodium diet in com-
bination with high-dose diuretic agents and
fluid restriction in patients with congestive heart
failure have been reported to show an increase
in hospital readmissions and deaths.46,47 How-
ever, in patients with chronic kidney disease,
modest salt reduction is associated with im-
proved clinical outcomes and greater reductions
in blood pressure in response to pharmacologic
inhibition of the reninangiotensin system.48
MECH A NISMS OF S A LT-INDUCED
H Y PER TENSION A ND TA RGE T- ORG A N
DA M AGE
In parallel with these observational studies and
clinical trials, mechanisms by which a high salt
intake may increase blood pressure and lead to
adverse cardiovascular outcomes have been studied
in the laboratory. Hypertension can be produced in
response to a high dietary sodium intake in a num-
ber of well-recognized experimentally induced
conditions, all of which have the common denom-
inator of a diminution in the renal capacity to
excrete sodium.49 This natriuretic handicap may
be due to an intrinsic renal defect. Alternatively,
stimuli that result in increased renal tubular re-
absorption of sodium chloride may reset the kid-
neys so that a higher level of renal-arterial perfu-
sion pressure is required to maintain a net
sodium balance.
As suggested by Guyton (cited in Cowleys
review50), impaired natriuresis may result in a
small increase in blood volume, and in response,
whole body autoregulation may explain the rise
of total peripheral resistance. Whether this se-
quence of events occurs in either the Dahl salt-
sensitive rat or in humans with salt-sensitive
hypertension is not clear. What is clear, however,
is that salt can activate a number of neural, en-
docrine or paracrine, and vascular mechanisms,
all of which have the potential to increase arte-
rial pressure (Table 2); for a list of relevant
references, see the Supplementary Appendix,
available with the full text of this article at
NEJM.org.
In rats, a high-salt diet leads to accumulation
nejm.org march 28, 2013
 Medical Progress
of hypertonic sodium in the interstitial space.51
This hypertonicity is sensed by macrophages,
which produce an angiogenic protein, vascular
endothelial growth factor, in the skin that stimu-
lates lymphatic-vessel growth, creating a third
fluid compartment that buffers sodium-induced
increases in vascular volume. It has been suggested
that failure of this extrarenal regulatory mecha-
nism may lead to salt sensitivity in rats with
deoxycorticosterone acetatesalt hypertension.52
However, this hypothesis remains speculative be-
cause there is no evidence that altering the dis-
tribution of salt and body fluids would by itself
affect the long-term regulation of arterial pressure.
Independently of its effect on arterial pressure,
prolonged salt loading in the rat causes altera-
tions of vascular endothelial-cell function and
promotes organ damage (Fig. 1).53-55 Adminis-
tration of excess dietary salt in rats with sponta-
neous hypertension causes perivascular fibrosis of
the coronary arteries, fibrosis of the noncardiac
ventricular interstitial matrix, ischemia of both
ventricles, and ventricular diastolic dysfunction.56
Severe proteinuria and end-stage renal failure
develop within 3 weeks and are associated with
interstitial fibrosis, renal arteriolar damage, in-
creased glomerular hydrostatic pressure, and glo-
merular hyalinization.57 Although treatment with
an angiotensin-receptor antagonist does not re-
duce arterial pressure in these rats, it prevents or
attenuates the salt-induced structural and func-
tional changes in the heart and kidney.58,59 High
salt intake also results in decreased elasticity and
fibrosis of large arteries, potentially worsening
hypertension and exacerbating cardiovascular
risk.60 Like salt excess, high levels of aldosterone
are associated with alterations of myocardial and
renal structure and function, owing to oxidative
stress and vascular inflammation.61 The proin-
flammatory effects of aldosterone are amplified
by salt, and clinically, target-organ damage has
been related to the interdependence of aldoste-
rone and dietary salt.62-65
The most frequent causes of hospitalization
among older persons in industrialized societies
are cardiac failure and end-stage renal disease.
Such target-organ disease may result from long-
term consumption of excess salt. Salt intake is an
independent predictor of left ventricular mass,
and left ventricular mass decreases in response to
dietary salt restriction.66,67 In patients with hyper-
tension, a high salt intake amplifies the effect of
n engl j med 368;13
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
Table 2. Interrelated Salt-Induced Alterations That May Impair Sodium
Excretion and Promote Vasoconstriction.*
Kidney
Increased activity of the sympathetic nervous system
Decreased renal medullary blood flow
Increased formation of reactive oxygen species
Low bioavailability of nitric oxide
Defective dopamine-receptor function
Enhanced vasoconstrictor effects of angiotensin II and vasopressin
Increased intrarenal generation of angiotensin II
Overexpression of angiotensinogen in proximal tubule
Increased expression of aldosterone synthetase
Activation of the mineralocorticoid receptor
Decreased production of 20-hydroxyeicosatetraenoic acid, an arachidonic
acid metabolite
Failure of atrial natriuretic peptide to potentiate marinobufagenin-induced
inhibition of Na+/K+ATPase
Nervous system
Increased sympathetic nervous system activity triggered by sodium concen-
tration in cerebrospinal fluid
Enhanced response of vasomotor neurons of the rostral ventrolateral medulla
to excitatory amino acids
Decreased release of nitric oxide in the paraventricular nucleus
Activation of glutamate receptors in the paraventricular nucleus
Decreased baroreceptor sensitivity
Increased oxidative stress in the brain
Production of a ouabain-like compound by the brain
Up-regulation of mineralocorticoid receptors in the central nervous system
Blood vessels
Reduced production of nitric oxide and impaired nitric oxidedependent
vasodilation
Increased production of reactive oxygen species
Reduced scavenging of free radicals by superoxide dismutase
Failure of atrial natriuretic peptide to potentiate marinobufagenin-induced
inhibition of Na+/K+ATPase
Potential vascular effect of salt-induced autoimmune inflammatory response
* For a list of relevant references, see the Supplementary Appendix, available at
NEJM.org. Na+/K+ATPase denotes the sodiumpotassium pump.
arterial pressure on target-organ damage, includ-
ing cardiac hypertrophy and microalbuminuria.68
Furthermore, in patients who have hypertension
with compensated heart failure and a normal
ejection fraction, dietary-salt restriction reduces
arterial pressure, arterial stiffness, and oxidative
stress.69 Multifactorial causation of prolonged
salt excess, including an interaction with tissue
reninangiotensin systems, may contribute to ma-
jor target-organ impairment.
nejm.org march 28, 2013 1233
 The n e w e ng l a n d j o u r na l of m e dic i n e

(9.5 g of sodium chloride) in 2001 to 3440 mg

per day (8.6 g of sodium chloride) in 2008.
Increased arterial In 2005, the U.S. Department of Health and
pressure
Human Services recommended that adults in the
United States consume no more than 2300 mg
of sodium per day (5.8 g of sodium chloride) and
that those in specific groups (persons 51 years
Prolonged high
Kidney Heart of age or older, persons with hypertension, dia-
sodium chloride intake
betes, or chronic kidney disease, and persons of
Glomerular injury Cardiac hypertrophy African-American ethnicity) consume no more
Renal failure Diastolic dysfunction
Systolic dysfunction than 1500 mg per day (3.8 g of sodium chlo-
Blood vessels ride).74 The 1500-mg recommendation applies to
about half the U.S. population. The same recom-
Oxidative stress
Endothelial dysfunction mendations were endorsed as part of the Dietary
Fibrosis Guidelines issued in 2011 by the U.S. Department
Decreased vascular elasticity
of Agriculture and the Department of Health and
Human Services.75 Numerous professional soci-
Figure 1. Target-Organ Damage Due to High Intake of Sodium Chloride.
COLOR FIGURE
eties, including the American Heart Association,
In addition to increasing arterial pressure, a prolongedDraft
high3 intake of2/15/13
sodium
have also endorsed recommendations to reduce
chloride has a direct effect on target-organ damage.Author Kotchen sodium intake to less than 1500 mg per day.76 In
Fig #1
Title
England and Wales, the government-recommend-
ed target was 2400 mg of sodium per day (6.0 g
ME Laurencot
of sodium chloride) by 2012.73 The global goal
R EC OM MENDAT
DE
IONS
Inglefinger
Artist Knoper
A ND S T R ATEGIE S AUTHOR
FORPLEASE S ANOTE:
LT set by the World Health Organization is to re-
R EDUC T ION Figure has been redrawn and type has been reset
Please check carefully
duce sodium intake to less than 2000 mg per
Issue date 3/28/13 day (5 g of sodium chloride) per person by 2025,
In view of the association of a high salt intake with some countries aiming for even lower levels
with hypertension and cardiovascular and renal in the long term.73
disease, many countries have introduced popula- Despite these recommendations, initiatives,
tion-based recommendations and initiatives to and early successes, sodium intake remains high.
reduce salt consumption.70 Beginning in the ear- Data from the National Health and Nutrition Ex-
ly 1970s, Finland implemented population-wide amination Survey suggest that, for over a decade,
initiatives to reduce salt intake.71 Between 1979 sodium consumption has been relatively con-
and 2002, the average 24-hour urinary sodium stant in the United States and well above recom-
excretion decreased from more than 5200 mg mended amounts (Fig. 2). Currently, Americans
per day (13.0 g of sodium chloride) to less than consume a mean of approximately 3400 mg of
4000 mg per day (10.0 g of sodium chloride) in sodium per day (8.5 g of sodium chloride), with
Finnish men and from nearly 4200 mg per day 77% of the sodium coming from packaged, pro-
(10.5 g of sodium chloride) to less than 3000 mg cessed, and restaurant foods.77 In 2010, the Insti-
per day (7.5 g of sodium chloride) in Finnish tute of Medicine recommended that sodium in-
women.72 Along with this reduction in sodium take be reduced gradually, and emphasized that
intake, there has been a reduction of more than voluntary approaches for reducing sodium levels
10 mm Hg in both systolic and diastolic blood in the food supply have not been successful.78
pressure and a corresponding decrease of 75 to Nevertheless, reflecting the difficulty of trans-
80% in the rate of death due to stroke and coro- lating science into public policy,79 there remain
nary heart disease.71 In 2004, with the voluntary outspoken critics of these population-based recom-
engagement of the food industry, the British gov- mendations to reduce sodium consumption.80-83
ernment introduced a population-based salt-reduc- Several specific concerns have been expressed.
tion program with the use of a media campaign to Critics point out that the influence of salt intake
increase public awareness and demand for change.73 on blood pressure is generally too small to man-
Sodium intake decreased from 3800 mg per day date policy decisions and that there is substan-

1234 n engl j med 368;13 nejm.org march 28, 2013

The New England Journal of Medicine

Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
 Medical Progress

tial variation from one person to another in the

blood-pressure response to salt reduction. In ad- Consumption recommendation Consumption recommendation
dition, critics note that results of studies of the for the general population for persons >50 yr of age, persons
with hypertension, diabetes, or
relationship of reduced sodium intake to mor- 4000
chronic kidney disease, and persons
bidity and mortality have been inconsistent and of African-American ethnicity
3500
that population-based estimates of the reduction
in cardiovascular disease that is related to an 3000

Sodium (mg/day)
effect of salt reduction on blood pressure are
2500
based on a surrogate end point. They also note
that a reduction in sodium intake may have an 2000
adverse effect on other end points, such as level of
1500
lipids, catecholamines, renin, and aldosterone, and
that in the general population, a J-shaped curve 1000
may characterize the relationship between salt 0
consumption and cardiovascular morbidity and
00

02

04

06

08

10
20

20

20

20

20

20
mortality. Critics also note that low salt intake
9

9
9

0
19

20

20

20

20

increases the risk of cardiovascular events in spe-
cific patient groups (e.g., patients with congestive
heart failure who are aggressively treated with
diuretic agents and patients with diabetes). In
response to concerns that a low level of sodium
intake may adversely affect blood lipids, insulin
resistance, and the risk of cardiovascular dis-
ease, the Institute of Medicine is undertaking a
study to evaluate the results, study design, and
methodological approaches that have been used
to assess the relationship between sodium and
health outcomes.84
C ONCLUSIONS
Although it has been difficult to separate salt
need from salt preference, current levels of salt
consumption exceed salt need and are associated
with adverse clinical outcomes. High salt intake
is associated with high blood pressure and in-
creased rates of cardiovascular disease. Experi-
mental studies continue to provide information
about mechanisms for these adverse effects of
salt. In clinical trials, a reduction in salt intake is
associated with reduced blood pressure, more so
in persons with hypertension than in those with
normal blood pressure. Although not discussed
in the present review, it should be noted that re-
duced salt intake is associated with greater blood-
20
Figure 2. Average Daily Sodium Consumption in the United States, 19992010.
Data are from the National Health and Nutrition Examination Survey. Current
recommendations of the U.S. Department of Health and Human Services
for the general population and various subgroups are shown.
pressure responses to antihypertensive drug ther-
apy, including drug therapy in patients with
resistant hypertension.85,86 Most, but not all, clin-
ical trials have shown that reduced salt intake is
also associated with decreased risks of cardio-
vascular events and death. Consequently, recom-
mendations for reducing the currently high levels
of salt consumption in the general population
seem justifiable, although in terms of safety, the
lower limit of salt consumption has not been
clearly identified. It may be premature to dis-
count the apparently paradoxical cardiovascular
outcomes associated with low salt intake, par-
ticularly in specific clinical conditions (e.g., type
1 or type 2 diabetes and congestive heart failure
that is aggressively treated with diuretic agents).
Less-rigorous targets for salt reduction may be
appropriate for these and other patient groups.
No potential conflict of interest relevant to this article was
reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.

REFERENCES
1. Daniels D, Fluharty SJ. Salt appetite: a
neurohormonal viewpoint. Physiol Behav
2004;81:319-37.
2. Morris MJ, Na ES, Johnson AK. Salt
craving: the psychobiology of pathogenic
sodium intake. Physiol Behav 2008;94:
709-21.
3. Leshem M. Biobehavior of the human
love of salt. Neurosci Biobehav Rev 2009;
33:1-17.
4. Frost CD, Law MR, Wald NJ. By how
much does dietary salt reduction lower
blood pressure? Analysis of observational
data within populations. BMJ 1991;302:
815-8.

n engl j med 368;13 nejm.org march 28, 2013 1235

The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF MELBOURNE on June 18, 2014. For personal use only. No other uses without permission.
Copyright 2013 Massachusetts Medical Society. All rights reserved.
http://www.kidney-international.org
& 2010 International Society of Nephrology
WASHWorld Action on Salt and Health
Feng J. He1, Katharine H. Jenner1 and Graham A. MacGregor1
1
Centre for Environmental and Preventive Medicine, Wolfson Institute of Preventive Medicine, Barts and The London School of Medicine
& Dentistry, Queen Mary University of London, London, UK
There is overwhelming evidence that our current high-salt
intake is the major factor increasing blood pressure (BP) and,
thereby, a major cause of cardiovascular disease and kidney
disease worldwide. A reduction in salt intake to the
recommended level of o56 g/day is very beneficial, and
could prevent millions of deaths each year and make
major savings for healthcare services. Several countries,
e.g., Finland and the UK, have already reduced the amount
of salt being consumed by a combined policy of getting the
food industry to decrease the amount of salt added to foods,
clear labelling on food products, and increasing public
awareness of the harmful effects of salt on health. Many
other developed countries, e.g., Australia, Canada, and the
US, are also stepping up their activities. The major challenge
now is to spread this out worldwide, particularly to
developing countries where E80% of global BP-related
disease burden occurs. In many developing countries, most
of the salt consumed comes from salt added during cooking
or from sauces; therefore, public health campaigns are
needed to encourage consumers to use less salt. A modest
reduction in salt intake across the whole population will
result in major improvements in public health and have huge
economic benefits in all countries around the world. World
Action on Salt and Health (WASH) is a coalition of health
professionals from different countries who know very well
the harm of high BP and has a major role in implementing
changes in their own countries. We welcome nephrologists
to join (http://www.worldactiononsalt.com).
Kidney International (2010) 78, 745753; doi:10.1038/ki.2010.280;
published online 18 August 2010
KEYWORDS: cardiovascular disease; cost-effectiveness; dietary salt; renal
disease; salt reduction activities
Correspondence: Feng J. He, Centre for Environmental and Preventive
Medicine, Wolfson Institute of Preventive Medicine, Barts and The London
School of Medicine & Dentistry, Queen Mary University of London,
Charterhouse Square, London EC1M 6BQ, UK. E-mail: f.he@qmul.ac.uk
Received 2 June 2010; accepted 9 June 2010; published online
18 August 2010
Kidney International (2010) 78, 745753
mini review
During evolution, the dietary intake of salt by the human
race and its forebears was o0.25 g/day (0.1 g sodium). The
addition of salt to foods began only about 5000 years ago
when the Chinese discovered that salt could be used to
preserve foods. Salt then became of great economic
importance, as it was possible to preserve foods during the
winter and allowed for the development of settled commu-
nities. Salt was the most taxed and traded commodity in the
world, with the intake reaching a peak around the 1870s.
However, with the invention of the deep freezer and the
refrigerator, salt was no longer required as a preservative. Salt
intake had been declining, but with the recent large increase
in the consumption of highly salted processed foods, salt
intake is now increasing again. The average salt intake in
most countries around the world is E912 g/day, with many
Asian countries having mean intakes over 12 g/day.1
Over several million years of evolution, humans have
become genetically programmed to a salt intake of o0.25 g/
day. The recent dietary changes (in evolutionary terms) to the
high-salt intake present a major challenge to the physiological
systems in excreting these large amounts of salt through the
kidneys. The consequence is that the high-salt intake causes a
gradual rise in blood pressure (BP),2,3 thereby increasing the
risk of cardiovascular disease (CVD)4 and renal disease.57
Furthermore, a high-salt intake may have direct effects on
stroke,8 left ventricular hypertrophy,9 progression of renal
disease, and proteinuria,5 independent of but additive to the
effect of salt on BP. There is also increasing evidence that salt
intake is indirectly related to obesity through soft drink
consumption,10,11 associated with an increased risk of renal
stones and osteoporosis,12 linked to the severity of asthma,13
and is probably a major cause of stomach cancer.14 The
evidence on these harmful effects of salt has been compre-
hensively examined in several recent review articles.3,15 In this
paper, we provide a brief update on the evidence, particularly
that relating salt to CVD and kidney disease. Additionally, we
provide an update on the activities of reducing salt intake
occurring around the world.
EVIDENCE RELATING SALT TO CVD
CVD is the leading cause of death and disability worldwide.
Raised BP is a major cause of CVD, responsible for 62% of
stroke and 49% of coronary heart disease.16 There is
overwhelming evidence from epidemiological, migration, inter-
vention, treatment, animal, and genetic studies demonstrating
745
mini review FJ He et al.: World Action on Salt and Health

1.6 0.20 TOHP I

Control
1.4 P = 0.007 P = 0.02
0.16
1.2
Relative risk

0.12
1.0

0.8 0.08

Cumulative incidence of CVD

Salt
0.6 0.04 reduction

0 0
Lower Higher Lower Higher TOHP II
0.10
salt salt salt salt

0.08 Control
Stroke CVD
Figure 1 | Relative risk for stroke and total cardiovascular Salt
disease (CVD) associated with a 5 g/day increase in salt 0.06
reduction
intake. Adapted from Strazzullo et al.19
0.04

that dietary salt is the major cause of raised BP and is largely
responsible for the rise in BP with age, and that a reduction
in salt intake lowers BP, thereby reducing BP-related
diseases.3 On the basis of the falls in BP from a meta-
analysis of randomized salt reduction trials,17 it was
estimated that a reduction of 6 g/day in salt intake would
reduce stroke by 24% and coronary heart disease by 18%.
This would prevent E35,000 stroke and coronary heart
disease deaths a year in the UK18 and E2.5 million deaths
worldwide.
Prospective cohort studies have shown that salt intake was
related to CVD. A recent meta-analysis of 13 cohort studies
with 177,025 participants for a follow-up period between 3.5
and 19 years showed that an increase of 5 g/day in salt intake
was associated with a 23% increase in the risk of stroke and a
14% increase in total CVD.19 After excluding one study
which had serious methodological problems,20 the pooled
analysis showed that a 5 g/day increase in salt intake was
associated with a 17% increase in CVD risk (Figure 1).19
Outcome trials have demonstrated that a reduction in salt
intake results in a decrease in CVD. A follow-up study of
individuals who took part in two large randomized salt
reduction trials, Trial of Hypertension Prevention (TOHP) I
and II, has shown a significant effect of salt reduction on
CVD.4 In the original TOHP trials, over 3000 participants
with an average baseline BP of 127/85 mm Hg were
randomized to a reduced salt group (for 18 months in
TOHP I and 3648 months in TOHP II) or to a control
group. Compared with the control group, individuals in the
intervention group reduced their salt intake by 2530% from
an average of E10 g/day. This resulted in a fall in BP of 1.7/
0.9 mm Hg at 18 months (TOHP I) and 1.2/0.7 mm Hg at 36
months (TOHP II). After the original trials were completed,
participants were not given further dietary advice. A follow-
up study at 1015 years post-trial showed that individuals
who were originally allocated to the reduced-salt group had a
0.02
0
2 4 6 8 10 12 14 16
Follow-up (years)
Figure 2 | Cumulative incidence of cardiovascular disease
(CVD) by salt intervention group in the Trial of Hypertension
Prevention (TOHP) I and II, adjusted for age, sex, and clinic.
Adapted from Cook et al.4
25% lower incidence of cardiovascular events after adjusting
for confounding factors (Figure 2).4 Another outcome trial of
over 2.5 years in elderly Taiwanese veterans (N 1981)
showed that switching from the usual salt (i.e., sodium
chloride) to potassium-enriched salt (49% sodium chloride,
49% potassium chloride, 2% other additives), with a
subsequent reduction of 17% in salt intake and an increase
of 76% in potassium intake, as measured by urinary sodium/
creatinine and potassium/creatinine ratio, resulted in a 40%
decrease in CVD mortality.21
SALT REDUCTION IS ONE OF THE MOST COST-EFFECTIVE
INTERVENTIONS TO REDUCE CVD
Several studies have demonstrated that a reduction in
population salt intake is very cost-effective in both developed
and developing countries.2228 Murray et al.25 carried out a
global analysis and showed that non-personal health inter-
ventions, including government action to stimulate a
reduction in the salt content of processed foods, were cost-
effective ways to limit CVD and could avert over 21 million
disability-adjusted life years per year worldwide. Studies in
Norway,22 Canada,23 and the US2628 have all demonstrated
that salt reduction not only saves lives, but also saves money.
For example, a recent study in the US showed that even a very
modest reduction in salt intake of only 10%, which could be

746 Kidney International (2010) 78, 745753

FJ He et al.: World Action on Salt and Health
15%
10 Reduction
9 in salt intake
deaths averted (106)
8 prevalence
Number of CVD
(20062015)
7
6
5 0.20
4
3 0.10
2
1 0.00
0
Figure 3 | Number of cardiovascular disease (CVD) deaths averted and the financial costs associated with implementation of salt
reduction and tobacco control in 23 low- and middle-income countries. Adapted from Asaria et al.24
easily achieved, as demonstrated in the UK, would prevent
hundreds of thousands of strokes and heart attacks over the
lifetimes of adults aged 4085 years, who are alive today, and
could save 4$32 billion in medical expenses in the US
alone.28 A larger decrease in salt intake would result in a
larger health improvement and greater cost savings.26 A
modest reduction in salt intake could be as beneficial as
smoking cessation.26
In developing countries, where the healthcare resources
are very limited, a reduction in salt intake is one of the most
cost-effective interventions to reduce CVD. Asaria et al.24
estimated the effects and cost of strategies to reduce salt
intake and to control tobacco use for 23 low- and middle-
income countries that account for 80% of chronic disease
burden in the developing world. They demonstrated that,
over 10 years (from 2006 to 2015), a 15% reduction in
population salt intake could avert 8.5 million CVD deaths
and a 20% reduction in smoking prevalence could avert 3.1
million CVD deaths. The modest reduction in salt intake
could be achieved by a voluntary reduction in the salt content
of processed foods and condiments by manufacturers, plus a
sustained mass-media campaign aimed to encourage dietary
change within households and communities. The cost for
implementing such salt reduction programmes was estimated
to be US$0.09 per person per year. The cost for tobacco
control including both price and non-price measures was
US$0.26 per person per year (Figure 3).24 These figures
clearly suggest that a reduction in salt intake is more or at the
very least just as cost-effective as tobacco control in terms of
reducing CVD on its own, the leading cause of death and
disability worldwide.
EVIDENCE RELATING SALT TO KIDNEY DISEASE
AND KIDNEY STONES
Dietary salt intake increases BP, which, in itself, can cause
kidney disease through hypertensive glomerulosclerosis or
through malignant hypertension. However, more frequently,
salt causes an increase in BP in those who have underlying
kidney diseases. This will accelerate the rate of deterioration
of renal function. There is also increasing evidence from
animal experiments demonstrating that salt may have a direct
effect on kidney function, independent of BP. For instance, in
animal models of renal impairment, a high-salt intake has
Kidney International (2010) 78, 745753
mini review
20%
15% 20%
Reduction
0.40 Reduction Reduction
in smoking
in salt intake in smoking
per person (US$)
0.30 prevalence
Annual cost
been shown to aggravate proteinuria and glomerulosclerosis,
and accelerate progression.29 In both partially nephrecto-
mized, spontaneously hypertensive rats and in the remnant
kidney model of renal failure, salt restriction slowed down
the deterioration of renal impairment to a greater extent than
thiazide diuretics.30
In humans, epidemiological studies have shown a direct
association between salt intake and urinary albumin excre-
tion, independent of BP.31,32 A randomized double-blind trial
in 40 Black hypertensive individuals demonstrated that a
reduction in salt intake from E10 to 5 g/day reduced 24-h
urinary protein by 19% (Po0.01; Figure 4).7 A more recent
double-blind trial in a larger number of individuals including
71 Whites, 69 Blacks, and 29 Asians with mildly raised BP
demonstrated that even a smaller reduction in salt intake, i.e.,
from an average of 9.7 to 6.5 g/day reduced 24-h urinary
albumin excretion significantly in all three ethnic groups.33
Other studies in patients with proteinuria or diabetes showed
that the antiproteinuric effects of an angiotensin-converting
enzyme inhibitor or angiotensin receptor blocker were
abolished by increasing salt intake.6,34 A retrospective analysis
of 57 chronic kidney disease patients, with an average
observation period of 3 years showed that a lower salt intake
reduced proteinuria and slowed down the progression of
renal disease despite a similar BP control between the two
groups on a high-and low-salt intake.5
In patients who are on dialysis, a reduction in salt intake
reduces the amount of fluid that they drink between dialyses.
This particularly applies to hemodialysis patients, where
controlling BP is a major problem. A lower salt intake
decreases weight gain between dialyses and BP is easier to
control.
Salt intake is one of the major dietary determinants of
urinary calcium excretion.12,35 As calcium is the main
component of most urinary stones, salt intake is therefore
an important cause of renal stones. Randomized trials have
demonstrated that a reduction in salt intake reduces calcium
excretion36,37 and reduces reoccurrences of kidney stones.37
WORLDWIDE ACTIONS ON SALT
The World Health Organization (WHO) has set a worldwide
target of a maximum intake of 5 g/day for adults.38 Many
countries have developed their own guidelines and set targets
747
mini review
Randomized cross-over
reduced salt diet
110
Slow
Urinary protein (mg per 24 hours)
sodium
100
90
80
70
60
Urinary sodium (mmol per 24 hours)
200
150
100
50
0 based on the UKs average salt intake of 9.5 g/day,45 as
** P < 0.01 compared with slow sodium period
Figure 4 | Change in urinary sodium and protein excretion
with a modest reduction in salt intake from E10 to 5 g/day in
40 hypertensive Blacks.
for salt intake, e.g., the UK and US guidelines recommend a
salt intake of o6 g/day for adults.39,40 These recommenda-
tions are based on the feasibility of reducing population salt
intake to these levels, but not on the potential maximum
beneficial effects of salt reduction. Therefore, these are viewed
as interim targets. Recently, the UK governments health
advisory agency, the National Institute for Health and
Clinical Excellence has recommended a reduction in the
populations salt consumption to 3 g/day by 2025.41
In most developed countries, E80% of salt consumed is
added to foods at the stage of manufacturing,42 and the
consumers have no say over how much salt is added.
Therefore, to achieve a reduction in population salt intake, it
is imperative that the food industry reduces the amount of
salt they add to all foods. Several countries, e.g., Finland, the
UK, have already successfully carried out salt reduction
programmes. Many other developed countries have also
launched national initiatives to reduce salt in foods.
The UK strategya model for other countries
The UK is one of the countries leading the way in salt
reduction, and is setting an example for other countries to
748
FJ He et al.: World Action on Salt and Health
Table 1 | UK strategy for reducing salt
Salt intake
Reduction Target intake
Placebo Source g/day needed (g/day)
Table/cooking (15%) 1.4 g 40% reduction 0.9 g
Natural (5%) 0.5 g No reduction 0.5 g
Food industry (80%) 7.6 g 40% reduction 4.6 g
Total: 9.5 g Target: 6.0 g
follow. In 1996, 22 experts on salt and BP set up an action
groupConsensus Action on Salt and Health (CASH)
** following the governments rejection of the 1994 COMA
(the Committee on Medical Aspects of Food and Nutrition
Policy) panels recommendation to reduce salt intake.43,44
CASH has waged a highly successful campaign to persuade
food manufacturers and suppliers to universally and
gradually reduce the salt content of processed foods, educate
the public in becoming more aware of salt in terms of
understanding the impact of salt on their health, and
translate the evidence into public health policy. CASH
persuaded the UK Department of Health to change its stance
on salt, finally resulting in the Chief Medical Officer
endorsing the original recommendations of the COMA
** report to reduce salt intake to o6 g/day in adults, and also
ensured that the UK Food Standards Agency (FSA) took on
the task of reducing salt intake.
A strategy to reduce population salt intake was developed
measured by 24-h urinary sodium (Table 1). It was estimated
that E15% of the salt consumed (i.e., 1.4 g) was added either
at the table or during cooking, 5% was naturally present in
the foods (0.5 g), and the rest 80% (7.6 g) was added by the
food industry in processed, canteen, restaurant, and takeaway
foods. To reach the target of 6 g, a total reduction of 3.5 g
(i.e., 40%) was needed. Therefore, the food industry would
need to reduce the amount of salt added to foods from 7.6 to
4.6 g (40% reduction) and the public would need to reduce
the amount of salt they add to foods themselves from 1.4 to
0.9 g (40% reduction).
It was estimated that, in the UK, only E15% of foods
were eaten outside the home, i.e., restaurant, canteen, etc,
and therefore, the main target in the initial phase of salt
reduction focused on foods that were bought in super-
markets. These foods, where salt was added, were split into
more than 80 categories. The FSA set target levels of salt for
each food category that the food industry needed to achieve
within a certain time period.46 The aim was to implement a
step-wise reduction in salt added to foods, i.e., 1020%
reduction and repeated at 12 year intervals. Such reductions
are not detectable by human salt taste receptors47 and cause
no technical or safety issues to the food in question,
therefore, posing no risk to manufacturers and no rejection
from consumers. Recently, the salt targets have been revised
down to ensure that salt intake will reach the target of 6 g/day
by 2012.48
Kidney International (2010) 78, 745753
FJ He et al.: World Action on Salt and Health
The UK salt reduction campaigns started in 2003/2004
and have been successful. The salt content in many food
categories has already been reduced, e.g., a one-third
reduction in the average amount of salt in branded, pre-
packed, sliced bread, and a reduction of over 40% in branded
breakfast cereals.46 Table and cooking salt sales have also been
reduced by E4050%. The average salt intake, as measured
by 24-h urinary sodium, in the general adult population in
the UK has fallen from 9.5 to 8.6 g/day by May 2008.49 This
change reverses the previous increasing trend for salt intake
and marks the beginning of a substantial reduction in salt
intake, as increasing reductions in salt added to foods are
being made by the food industry. The UK salt reduction
campaigns, which cost just d15 million, led to E6000 fewer
CVD deaths per year, saving the UK economy Ed1.5 billion
per annum.41,50
Clear labelling of the salt content of food is essential for
consumers to choose products with less salt. A front of pack
signpost labelling system51 has been developed in the UK,
which uses a combination of the traffic light system, where
there is a colour-coding of green, amber, and red for low,
medium, and high amounts of salt, fat, saturated fat, and
sugar, and the Guideline Daily Amount system where the
amount of salt per portion is expressed as a percentage of the
daily recommended maximum. This type of standardised
label is already being implemented by many supermarkets
and is much preferred by consumers,52 as they can see at a
glance whether a product has a little or a lot of salt. Signpost
labelling has already been shown to have a dramatic effect on
the purchase of foods, particularly of those in the red (high)
category. Despite this, perhaps owing to intense lobbying
from the food industry, the European Parliament has recently
voted against making this scheme mandatory in Europe, and
the shape of the final nutritional labelling legislation is
uncertain.
Finland
Finland was one of the first countries to initiate a systematic
approach to reduce salt intake, in the late 1970s, through
mass-media campaigns, cooperation with the food industry
and implementing salt labelling legislation.5355 Since the
1980s, many food companies have reduced the salt content of
their food products by replacing conventional table salt with
a sodium-reduced, potassium- and magnesium-enriched
mineral salt known as Pansalt. In the early 1990s, the
Ministry of Trade and Industry, and the Ministry of Social
Affairs and Health set salt labelling legislation for all food
categories, which made a substantial contribution to the salt
intake of the Finnish population. Foods that are high in salt,
e.g., 41.3% in bread, 1.8% in sausages, 1.4% in cheese, are
required to carry a high-salt content warning and if a food
product contains a low level of salt, e.g., o0.7% in breads,
1.2% in sausages, 0.7% in cheese, the product is allowed to
display a low-salt label. These different measures have been
very effective and led to a marked reduction in the salt
content of many food categories, e.g., the average salt content
Kidney International (2010) 78, 745753
mini review
in bread has been decreased by around 20% from E1.5%
to E1.2%.
The salt reduction campaign has resulted in a significant
reduction in the average salt intake of the Finnish popula-
tion53,55 from E14 g/day in 1972 to o9 g/day in 2002.53 The
reduction in salt intake was accompanied by a fall of over
10 mm Hg in both systolic and diastolic BP, a pronounced
decrease of 7580% in both stroke and coronary heart disease
mortality, and a remarkable increase of 56 years in life
expectancy.53 The reduction in salt intake was a major
contributory factor for these results, particularly the fall in
BP as both body mass index and alcohol consumption had
increased during that period. An increase in potassium intake
via the use of reduced-sodium, potassium- and magnesium-
enriched salt, an increased consumption of fruit and
vegetables, a reduction in fat intake, and a decrease in
smoking rate in men also played an important part in the fall
in CVD.
Japan
In the late 1950s, deaths from stroke in Japan were among the
highest in the world, and salt intake was also very high. It was
found that the number of stroke in different parts of Japan
was directly related to the amount of salt consumed. The
Japanese government initiated a campaign to reduce salt
intake. Over the following decade, the national salt intake was
reduced from an average of 13.5 to 12.1 g/day and in the
north, salt intake fell dramatically from 18 to 14 g/day.
Paralleling this reduction in salt intake, there were falls in BP
both in adults and children, and an 80% reduction in stroke
mortality56 despite large increases in fat intake, cigarette
smoking, alcohol consumption, and obesity. It would appear
that the Western influence, which was rapidly overtaking
Japan at that time, had little effect on BP, provided salt intake
was reduced. Overall the reduction in salt intake appeared to
be associated with the fall in deaths from stroke.
The US
In the US, there has been consistent advice to reduce salt
intake to o6 g/day since the 1980s. However, little action has
been taken until recently. In 2005, in a petition to the Food
and Drug Administration (FDA), the Center for Science in
the Public Interest called for tougher regulations on salt.57 In
2007, the American Medical Association published a report
calling for a major reduction in the salt content of processed
and restaurant foods.58 The FDA has been urged to cease the
rule that allows salt and its component sodium to be treated
as generally recognized as safe.57,58 In November 2008,
Michael Bloomberg, the Mayor of New York City, announced
his plan to cut salt levels in processed foods by 20% over
5 years. In January 2010, the National Salt Reduction
Initiative (NSRI), a New York City-led partnership of more
than 40 cities, states, and national health organizations,
unveiled its proposed targets to guide a voluntary reduction
of salt levels in packaged and restaurant foods with an aim of
reducing Americans salt intake by 20% over 5 years and 40%
749
mini review
Figure 5 | Action groups on salt reduction around the world. The red dot indicates the location of World Action on Salt and Health
(WASH) members.
over 10 years.59 To achieve these reductions, the salt content
of processed and packaged foods will need to decrease by
25% in 5 years and by 50% in 10 years. Similar to the UK
strategy, the US NSRI has developed voluntary 2- and 4-year
interim targets for cutting salt levels in restaurant and
packaged foods.59
More recently, the Institute of Medicine (IOM) released
the report Strategies to Reduce Sodium Intake in the United
States.60 The report recommends setting mandatory national
standards for the salt levels that food manufacturers,
restaurants, and food service companies can add to their
products in the US, and that all of the relevant health
agencies should be involved in achieving this. The FDAs role,
as envisioned in the report, will be to set the new standards,
which can be phased in gradually allowing tastes to adjust
over time, although no time scale has been given. Following
IOMs report, 16 food companies in the US, including food
manufacturers Kraft Foods and HJ Heinz Co, have pledged to
cut the amount of salt in some of their products.
WASH (World Action on Salt and Health)
Following the success of the UK campaign groupCASH,
a world action groupWASH was established in 2005.61
The aim of WASH is to set up similar groups modelled
on CASH suited to each individual country to reduce salt
intake with an appropriate strategy relevant to the needs of
that particular country, and to stimulate actions from the
government and/or department of health, the food industry,
750
FJ He et al.: World Action on Salt and Health
media, and public. WASH is supported by more than 400
members from 81 countries (Figure 5).
WASH works to reduce salt in the diet worldwide by
exerting pressure on multi-national food companies to
reduce the salt content of their products. In 2009, WASH
and the International WASH members conducted a survey of
over 260 branded products from KFC, McDonalds, Kelloggs,
Nestle, Burger King, and Subway in different countries and
revealed huge variations in salt content in global brands
unrelated to local traditional taste preferences. Not one
product surveyed had the same salt content around the world
and some displayed huge differences from one country to
another. For instance, Kelloggs All Bran contained 2.15 g of
salt per 100 g in Canada, but only 0.65 g of salt per 100 g just
over the border in the US, less than a third of the Canadian
level.61 This illustrates once again how easy it would be for
the food industry to reduce the amount of salt they add to
foods, particularly as they could do this straightaway to their
branded products. Pressure from campaigns such as this has
resulted in several large multi-national manufacturers pled-
ging to reduce the amount of salt added to foods across the
world. Those already committed to salt reduction worldwide
include Pepsico, Unilever, Kelloggs, Heinz, Campbells, and
Kraft.
WASH members in each country are encouraged to set up
their own country division of WASH to work together on a
local level to lower salt intake specifically in their own
population. For example, in 2007, an Australian Division of
Kidney International (2010) 78, 745753
FJ He et al.: World Action on Salt and Health
World Action on Salt and Health (AWASH) was established.62
AWASH has launched a national campaign to lower the salt
intake of the Australian population to 6 g/day by 2012. In
March 2010, the Australian government announced salt
targets for bread and breakfast cereals. AWASH is calling on
the government to take urgent action by setting salt target
levels for all processed and take-away foods.
Other developed countries
In Ireland, the Food Safety Authority set a goal to reduce salt
intake in the population to o6 g/day.63 Their strategic
approach includes consumer awareness efforts and actions by
the food industry to lower the salt content of their food
products, where to-date substantial reductions have been
made, e.g., the salt content of breakfast cereals has been
reduced by 11% since 2003.
In Portugal, in view of the high-salt intake and the high
prevalence of stroke, a group of doctors formed an action
group, the Portuguese Action Against Salt and Hypertension
(PAASH). They carried out a pilot study in almost 500 people
and showed that the average salt intake was 11.9 g/day,64
which is twice the recommended level for adults. PAASH
started a mass-media campaign about the harmful con-
sequences of consuming too much salt, along with an
educational campaign directed at food manufacturers, health
authorities, and policy makers. This has led to the Portuguese
Parliament approving a law restricting the salt content of
bread and processed foods.65
In Canada, the first Chair in Hypertension Prevention and
Control was appointed in 2006.66 One of the major aims of
the Chair was to get a reduction in salt intake among the
Canadians. The Chair is supported by a number of health
organizations and scientists. Together, they lobby the
government and the food industry for policies to reduce
the level of salt in foods.15 In October 2007, 17 health
organizations endorsed a national collaborative policy
statement urging reductions in salt in processed and
packaged foods, and the Canadian government established
an expert Sodium Working Group to oversee the imple-
mentation of salt reduction strategies.
Many other developed countries, e.g., New Zealand,
France, Italy, Holland, Belgium, Switzerland, Sweden, Den-
mark, Slovenia, Bulgaria, Croatia, Serbia, and Barbados, are
stepping up their activities to reduce salt intake. The WHO is
starting salt reduction strategies through its regional
directorates.67 The European Union is also following suit
and 11 countries have signed up to make a 16% reduction in
salt intake over 4 years. In February 2010, the Pan American
Health Organization/WHO Regional Expert Group on
cardiovascular disease prevention through dietary salt
reduction68 produced a policy statement outlining the
recommendations for a population-based approach to reduce
dietary salt intake in the Americans. The Policy Statement
provides countries with a roadmap for concerted actions by
governments, non-governmental organizations, and the food
industry. The goal is a gradual and sustained drop in dietary
Kidney International (2010) 78, 745753
mini review
salt intake to reach national targets or the internationally
recommended target of o5 g/day by 2020.
Developing countries
Approximately 80% of CVD deaths occur in developing
countries (i.e., low- and middle-income countries).69 Salt
intake is very high70 and the healthcare resources are very
constrained in many of these countries. A reduction in salt
intake is probably the most important strategy to combat the
epidemic of CVD in these countries.24
Despite the compelling evidence for a reduction in
population salt intake, little action is happening in develop-
ing countries. Although action groups have been set up in
several countries, e.g., China, Pakistan, Bangladesh, Nepal,
Cuba, Kenya, and Ghana, many developing countries have
not developed dietary guidelines or strategies to reduce salt
intake. It is important that each country determines what its
salt intake is and where the major sources of salt are in the
diet, and then implements a strategic approach to lower salt
intake in the population to the target level. In developing
countries, the major sources of salt consumption are additions
during cooking and in sauces (e.g., soy sauce), spice mixes,
seasonings, pickles, etc. rather than pre-packaged prepared
foods. Public health campaigns are needed to encourage
people to use less salt in their own food preparations.
In many developing countries, more and more processed
foods are being consumed, as their diets are becoming
westernized. The food industry in these countries is poorly
regulated with very little or no food-content labelling,
making informed eating almost impossible. Therefore, many
developing countries need a combined policy of getting the
public to use less salt at home and getting the food industry
to add less salt to foods and to adopt a clear labelling system
such as the signpost labelling system.51 At the same time,
developing countries need to ensure that all imported food
products are low in salt.
Additionally, on behalf of WASH, many members from
both developed and developing nations have supported the
last three World Salt Awareness Weeks held in February each
year. Activities from both developed and developing
countries have ranged from involvement with hypertension
and renal experts, representatives from the health and
education ministries, consumers, the media, and the food
industry. The Salt Awareness Weeks have generated signifi-
cant publicity and helped raise awareness of salt reduction
strategies in many countries.
CONCLUSIONS
There is now overwhelming evidence for a reduction in salt
intake in populations worldwide. Reducing salt from the
current intake of 912 g/day to the recommended level of
o56 g/day will have major beneficial health effects along with
major cost savings in all countries around the world. Clearly,
the time has now come for action and implementation.
All countries should adopt a coherent and workable
strategy to reduce salt intake. In most developed countries,
751
mini review
E80% of salt is hidden in foods, i.e., added by the food
industry.42 It is therefore vital to persuade the food industry
to make a gradual and sustained reduction in the amount of
salt they add to foods. Several countries, e.g., Finland, Japan,
and the UK, have successfully carried out salt reduction
programmes and salt intake has already fallen.49,53 Other
countries, e.g., Australia, Canada, the US, are also stepping
up their activities to reduce salt intake by following the UK
model. The global food industry is now poised to reduce the
amount of salt being added to foods and cannot justify the
inequality of reducing salt levels in some countries and
excluding others throughout the world.
In developing countries, where most of the salt consumed
comes from salt either added during cooking or from sauces,
public health campaigns are needed to encourage consumers
to use less salt. As there is an increasing trend in the
consumption of processed foods in many developing
countries, a combined policy of educating the public,
working with the food industry, and clear labelling, is
needed. A modest reduction in population salt intake
worldwide would result in a major improvement in public
healthsimilar to the provision of clean water and drains in
the late nineteenth century in Europe.
WASH would like to extend its membership to new
members, particularly nephrologists, who are well aware of
the dangers of salt. Joining WASH does not require any time
or financial commitments from the members, but only their
support to salt reduction worldwide. We value information
sharing, feedback, and input from your country and
encourage members to participate in activities such as World
Salt Awareness Week, in particular, by setting up an action
group in your own country. To join, please e-mail WASH
(wash@qmul.ac.uk) or visit the WASH website for more
information (http://www.worldactiononsalt.com).
DISCLOSURE
All the authors declared no competing interests.
REFERENCES
1. Brown IJ, Tzoulaki I, Candeias V et al. Salt intakes around the world:
implications for public health. Int J Epidemiol 2009; 38: 791813.
2. Intersalt Cooperative Research Group. Intersalt: an international study of
electrolyte excretion and blood pressure. Results for 24 h urinary sodium
and potassium excretion. BMJ 1988; 297: 319328.
3. He FJ, MacGregor GA. Reducing population salt intake worldwide:
from evidence to implementation. Prog Cardiovasc Dis 2010; 52:
363382.
4. Cook NR, Cutler JA, Obarzanek E et al. Long term effects of dietary
sodium reduction on cardiovascular disease outcomes: observational
follow-up of the trials of hypertension prevention (TOHP). BMJ 2007;
334: 885.
5. Cianciaruso B, Bellizzi V, Minutolo R et al. Salt intake and renal outcome in
patients with progressive renal disease. Miner Electrolyte Metab 1998; 24:
296301.
6. Heeg JE, de Jong PE, van der Hem GK et al. Efficacy and variability of the
antiproteinuric effect of ACE inhibition by lisinopril. Kidney Int 1989; 36:
272279.
7. Swift PA, Markandu ND, Sagnella GA et al. Modest salt reduction reduces
blood pressure and urine protein excretion in black hypertensives:
a randomized control trial. Hypertension 2005; 46: 308312.
8. Perry IJ, Beevers DG. Salt intake and stroke: a possible direct effect. J Hum
Hypertens 1992; 6: 2325.
752
FJ He et al.: World Action on Salt and Health
9. Kupari M, Koskinen P, Virolainen J. Correlates of left ventricular mass in a
population sample aged 36 to 37 years. Focus on lifestyle and salt intake.
Circulation 1994; 89: 10411050.
10. He FJ, Markandu ND, Sagnella GA et al. Effect of salt intake on renal
excretion of water in humans. Hypertension 2001; 38: 317320.
11. He FJ, Marrero NM, MacGregor GA. Salt intake is related to soft drink
consumption in children and adolescents: a link to obesity? Hypertension
2008; 51: 629634.
12. Cappuccio FP, Kalaitzidis R, Duneclift S et al. Unravelling the links
between calcium excretion, salt intake, hypertension, kidney stones and
bone metabolism. J Nephrol 2000; 13: 169177.
13. Burney P. A diet rich in sodium may potentiate asthma. Epidemiologic
evidence for a new hypothesis. Chest 1987; 91: 143S148S.
14. Joossens JV, Hill MJ, Elliott P et al. Dietary salt, nitrate and stomach cancer
mortality in 24 countries. European Cancer Prevention (ECP) and the
INTERSALT Cooperative Research Group. Int J Epidemiol 1996; 25:
494504.
15. Mohan S, Campbell NR. Salt and high blood pressure. Clin Sci (Lond) 2009;
117: 111.
16. World Health Organisation. World Health Report 2002: Reducing Risks,
Promoting Healthy Life. World Health Organisation: Geneva, Switzerland,
2002. http://www.who.int/whr/2002 (access verified 28 May 2010).
17. He FJ, MacGregor GA. Effect of modest salt reduction on blood pressure:
a meta-analysis of randomized trials. Implications for public health. J Hum
Hypertens 2002; 16: 761770.
18. He FJ, MacGregor GA. How far should salt intake be reduced?
Hypertension 2003; 42: 10931099.
19. Strazzullo P, DElia L, Kandala NB et al. Salt intake, stroke, and
cardiovascular disease: meta-analysis of prospective studies. BMJ 2009;
339: b4567.
20. Alderman MH, Madhavan S, Cohen H et al. Low urinary sodium is
associated with greater risk of myocardial infarction among treated
hypertensive men. Hypertension 1995; 25: 11441152.
21. Chang HY, Hu YW, Yue CS et al. Effect of potassium-enriched salt on
cardiovascular mortality and medical expenses of elderly men. Am J Clin
Nutr 2006; 83: 12891296.
22. Selmer RM, Kristiansen IS, Haglerod A et al. Cost and health consequences
of reducing the population intake of salt. J Epidemiol Community Health
2000; 54: 697702.
23. Joffres MR, Campbell NR, Manns B et al. Estimate of the benefits of a
population-based reduction in dietary sodium additives on hypertension
and its related health care costs in Canada. Can J Cardiol 2007; 23:
437443.
24. Asaria P, Chisholm D, Mathers C et al. Chronic disease prevention: health
effects and financial costs of strategies to reduce salt intake and control
tobacco use. Lancet 2007; 370: 20442053.
25. Murray CJ, Lauer JA, Hutubessy RC et al. Effectiveness and costs of
interventions to lower systolic blood pressure and cholesterol: a global
and regional analysis on reduction of cardiovascular-disease risk. Lancet
2003; 361: 717725.
26. Bibbins-Domingo K, Chertow GM, Coxson PG et al. Projected effect of
dietary salt reductions on future cardiovascular disease. N Engl J Med
2010; 362: 590599.
27. Palar K, Sturm R. Potential societal savings from reduced sodium
consumption in the U.S. adult population. Am J Health Promot 2009; 24:
4957.
28. Smith-Spangler CM, Juusola JL, Enns EA et al. Population strategies to
decrease sodium intake and the burden of cardiovascular disease: a cost-
effectiveness analysis. Ann Intern Med 2010; 152: 481487, W170-483.
29. Dworkin LD, Benstein JA, Tolbert E et al. Salt restriction inhibits renal
growth and stabilizes injury in rats with established renal disease. J Am
Soc Nephrol 1996; 7: 437442.
30. Benstein JA, Feiner HD, Parker M et al. Superiority of salt restriction over
diuretics in reducing renal hypertrophy and injury in uninephrectomized
SHR. Am J Physiol 1990; 258: F1675F1681.
31. du Cailar G, Ribstein J, Mimran A. Dietary sodium and target organ
damage in essential hypertension. Am J Hypertens 2002; 15: 222229.
32. Verhave JC, Hillege HL, Burgerhof JG et al. Sodium intake affects urinary
albumin excretion especially in overweight subjects. J Intern Med 2004;
256: 324330.
33. He FJ, Marciniak M, Visagie E et al. Effect of modest salt reduction on
blood pressure, urinary albumin, and pulse wave velocity in White, Black,
and Asian mild hypertensives. Hypertension 2009; 54: 482488.
34. Houlihan CA, Allen TJ, Baxter AL et al. A low-sodium diet potentiates
the effects of losartan in type 2 diabetes. Diabetes Care 2002; 25:
663671.
Kidney International (2010) 78, 745753
 Nutrition 26 (2010) 10441049

Contents lists available at ScienceDirect

Nutrition
journal homepage: www.nutritionjrnl.com

Review

Fructose and metabolic diseases: New findings, new questions

Luc Tappy M.D. a, b, *, Kim A. Le Ph.D. a, Christel Tran M.D. a, Nicolas Paquot M.D. c
a
Department of Physiology, University of Lausanne, Lausanne, Switzerland
b
Service of Endocrinology, Diabetes and Metabolism, CHUV, Lausanne, Switzerland
c
Service of Diabetology, CHU Sart Tilman, Lie`ge, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: There has been much concern regarding the role of dietary fructose in the development of
Received 21 October 2009 metabolic diseases. This concern arises from the continuous increase in fructose (and total added
Accepted 26 February 2010 caloric sweeteners consumption) in recent decades, and from the increased use of high-fructose
corn syrup (HFCS) as a sweetener. A large body of evidence shows that a high-fructose diet
Keywords: leads to the development of obesity, diabetes, and dyslipidemia in rodents. In humans, fructose has
Fructose
long been known to increase plasma triglyceride concentrations. In addition, when ingested in
Metabolic diseases
large amounts as part of a hypercaloric diet, it can cause hepatic insulin resistance, increased total
and visceral fat mass, and accumulation of ectopic fat in the liver and skeletal muscle. These early
effects may be instrumental in causing, in the long run, the development of the metabolic
syndrome. There is however only limited evidence that fructose per se, when consumed in
moderate amounts, has deleterious effects. Several effects of a high-fructose diet in humans can be
observed with high-fat or high-glucose diets as well, suggesting that an excess caloric intake may
be the main factor involved in the development of the metabolic syndrome. The major source of
fructose in our diet is with sweetened beverages (and with other products in which caloric
sweeteners have been added). The progressive replacement of sucrose by HFCS is however unlikely
to be directly involved in the epidemy of metabolic disease, because HFCS appears to have basically
the same metabolic effects as sucrose. Consumption of sweetened beverages is however clearly
associated with excess calorie intake, and an increased risk of diabetes and cardiovascular diseases
through an increase in body weight. This has led to the recommendation to limit the daily intake of
sugar calories.
! 2010 Elsevier Inc. All rights reserved.

Pure, white, and deadly: the dark side of sugar was suspected
many years ago, when an association between sugar consump-
tion and coronary heart diseases was recognized and empha-
sized by John Yudkin [1]. Sugar, a natural sweetener obtained
from either sugar cane or beets, is a disaccharide composed of
one glucose molecule linked through an a14 glycoside bond to
a fructose molecule. Fructose, besides contributing to half the
total content of sugar, can also be found as a hexose in fruits and
honey. More recently, sweeteners started to be produced from
corn through starch isolation and hydrolysis to glucose, followed
by enzymatic isomerization of part of the glucose into fructose
[2,3]. The resulting mixture, known as high-fructose corn syrup
(HFCS), has several industrial advantages over sugar, the most
important being its low price, and has progressively replaced
sugar consumption in North America over the past 30 years.
* Corresponding author. Tel.: ++41 21 692 55 41; fax: ++41 21 692 55 95.
E-mail address: luc.tappy@unil.ch (L. Tappy).
Fructose metabolism has been reviewed extensively else-
where [46] and will be only briefly outlined here. In the gut,
fructose is transported by specific transporters, GLUT5 [7,8]. In
some subjects, fructose absorption is quantitatively limited, and
some malabsorption occurs when large amounts of fructose are
ingested. This can cause abdominal discomfort and diarrhea, and
production of volatile fatty acids from colonic fructose fermen-
tation [9,10]. Fructose absorbed from the gut into the portal vein
is nearly completely metabolized in the liver through metabolic
pathways distinct from those of glucose; furthermore, the initial
steps of its metabolism are insulin-independent, and hence,
fructose is largely metabolized without requiring insulin secre-
tion and without increasing plasma glucose. This is due to the
fact that 1) part of the fructose appears to be directly metabo-
lized in enterocytes, where it is converted and into lactate and
glucose, and 2) the bulk of absorbed fructose is taken up by liver
cells, where it is rapidly converted into fructose 1-phosphate and
triose-phosphates through the sequential actions of fructokinase

0899-9007/$ see front matter ! 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.nut.2010.02.014
 L. Tappy et al. / Nutrition 26 (2010) 10441049
and aldolase B and triokinase. Fructokinase and aldolase B are
not inhibited by ADP and citrate and hence are not regulated by
the cellular energy status. In that, fructose differs from glucose,
because the ADP and citrate concentrations exert a negative
feedback control on the initial steps of glycolysis. As a conse-
quence of this absence of feedback inhibition, virtually all the
fructose ingested with a meal (whether under its pure, unbound
form, or bound to glucose in sucrose) is rapidly converted into
hepatic triose-phosphates [11]. These substrates are subse-
quently oxidized within the liver cells or converted into glucose
and lactate to be released into the bloodstream, or converted into
hepatic glycogen. A small, but significant amount of trisose-
phsophates is also converted into triacylglycerol in liver cells
through the process of de novo lipogenesis (TG) [12]. Although
quantitatively far less important than the other pathways of
fructose disposal, de novo lipogenesis appears to be closely
associated with the adverse metabolic effects of fructose.
Because fructose metabolism is not dependent on insulin
secretion, at least for its initial steps, and because fructose
ingestion causes only a limited rise in glycemia, fructose was
initially proposed as a natural substitute of sucrose for diabetic
patients. It however became rapidly apparent that an increased
dietary intake of fructose had serious adverse metabolic effects
in both rodents and humans. Thus it was recognized that a high-
fructose intake is associated with increased plasma triglyceride
concentrations, hepatic steatosis, impaired glucose tolerance and
insulin resistance, and even high blood pressure [4,5].
Besides these metabolic effects, fructose effects on mineral
metabolism have also been considered. Fructose forms
complexes with metal ions and hence may modulate the intes-
tinal absorption and bioavailability of minerals [13]. Fructose has
indeed been reported to decrease copper absorption in rats [14].
A diet containing up to 20% energy as fructose had however no
adverse effect on copper balance in humans [15]. Fructose also
increases iron absorption in rats [15] but does not appear to alter
zinc bioavailability [16]. The effects of fructose on calcium
metabolism have also been documented. The results indicated
that rats receiving glucose-sweetened beverages had lower
phosphate and calcium intake and increased urinary calcium
excretion compared to the rats receiving fructose-sweetened
beverages. These results suggest that fructose is not directly
involved in the negative association that was observed between
sugar intake and bone health [17]. Sweetened beverage
consumption however accounts for an important portion of
total fructose intake. In some populations, mainly children and
teenagers, a high consumption of sweetened beverages may
be associated with a lowered intake of milk, and hence, of
calcium [18].
Several important questions regarding the role of fructose in
the pathogenesis of metabolic diseases remain, however, only
partially addressed, as follows.
What are the mechanisms responsible for fructose-induced
metabolic alterations?
There is overwhelming evidence that, in rodents, a high-
sucrose or a high-fructose diet will lead to the development of
obesity, diabetes, and dyslipidemia, and to a substantial decrease
in both liver and muscle insulin sensitivity [19]. Oxidative stress
and endoplasmic reticulum stress appear to be involved in these
processes [2023]. Furthermore, the adverse metabolic effects
observed with high-sucrose diets appear directly related to the
fructose component of sucrose [24]. The adverse effects of fruc-
tose on glucose metabolism are closely linked to alterations of
1045
lipid metabolism. In rats, intrahepatic fat content and blood very
low density lipoprotein (VLDL)-triacylglycerol concentrations
increase within 6 wk of a high-fructose diet, while intramuscular
fat content increases within about 3 mo. Interestingly, the
development of insulin resistance follows a similar time course:
hepatic insulin resistance is observed early after switching to
a high-fructose diet, while the appearance of muscle insulin
resistance is delayed [19]. This suggests that fructose-induced
insulin resistance is closely linked to ectopic lipid deposition
and tissue-specific lipotoxicity [25].
In humans, the adverse metabolic effects of fructose are less
clearly documented. There is ample evidence that increasing the
fructose content of the diet increases plasma triglycerides,
through several mechanisms, among which are a stimulation of
hepatic de novo lipogenesis [26], and a decreased VLDL-
triacylglycerol clearance [27]. Based on animal experiments, an
increased VLDL-triacylglycerol secretion is also likely to be
involved [28]. The increase in plasma triglyceride induced by
fructose is significantly blunted in females, which suggests that
female sex hormones may exert a protective effect [29,30]. In
humans, a high-fructose, hypercaloric diet leads to ectopic fat
deposition in liver cells and skeletal muscle in humans already
after a 1-wk period [31]. There is strong evidence that fructose
decreases hepatic insulin sensitivity in humans [26,32,33], and
that a 10-wk high-fructose diet impairs glucose homeostasis in
overweight individuals [34]; however, few studies have directly
evaluated the effect of a high-fructose diet on whole body insulin
sensitivity. It has been reported that supplementation with
various amounts of fructose for up to 4 wk failed to decrease
whole body insulin sensitivity in healthy individuals [31,35].
Whether this may be different in obese insulin-resistant subjects
remains to be evaluated.
In animal models, high-fructose diets almost invariably lead
to the concomitant development of excess body fat and insulin
resistance; an increased body fat mass, body fat distribution, and
ectopic lipid deposition in liver and muscle can all play a role in
the development of the metabolic syndrome. In humans, it has
been recognized for several decades that abdominal fat distri-
bution is associated with metabolic and cardiovascular diseases
[36], and that intravisceral fat contributes much more to insulin
resistance than subcutaneous fat [37]. In this regard, the inter-
relationship between body fat distribution and intrahepatic fat
content bears a special interest. Many obese patients indeed do
not display significant metabolic disorders and have been coined
healthy obese or metabolically fit obese [38]. When
comparing these healthy obese subjects to obese subjects with
metabolic complications, it appears that both intravisceral fat
and intrahepatic fat content are tightly related with insulin
resistance. Furthermore, intrahepatic fat content appears more
specifically related to the metabolic syndrome [39,40]. In
humans, a high-fructose diet has been shown to increase
visceral fat [34] and intrahepatic fat [31], which raises the
possibility that the effects on hepatic fat metabolism and on
visceral fat deposits may be central in the adverse metabolic
effects of fructose. Based on these considerations, it is tempting
to speculate that a stimulation of hepatic de novo lipogenesis
induced by fructose leads to intrahepatic fat deposition,
increased VLDL-triacylglycerol secretion, and hepatic insulin
resistance; this may also secondarily lead to visceral fat depo-
sition through mechanisms that remain to be determined. With
time, the increased plasma VLDL-TG and an inhibition of lipid
oxidation induced by a high-fructose diet may favor ectopic fat
deposition in muscle, muscle lipotoxicity, and whole body
insulin resistance (Fig. 1).
1046 L. Tappy et al. / Nutrition 26 (2010) 10441049
Excess Fructose Calories
Hepatic de novo
lipogenesis
VLDL-triacylglycerol
secretion
+
VLDL- triacylglyerol
clearance
hypertriacylglycerolemia
Hepatic steatosis
Ectopic lipids
Visceral fat
in skeletal muscle
Insulin resistance
Impaired glucose tolerance
Metabolic syndrome
Fig. 1. Putative mechanisms linking excess fructose consumption to the metabolic
syndrome. In liver cells, fructose stimulates de novo lipogenesis, leading to
increased hepatic fatty acids, which can be deposited as ectopic liver fat (hepatic
steatosis) or be secreted as VLDL-triacylglycerols. In addition, fructose impairs the
extrahepatic clearance of VLDL-triacylglycerols. When excess fructose calories are
consumed, this leads to hepatic steatosis and to hyper-triacylglycerolemia, which in
turn favors visceral fat accumulation and ectopic lipid deposition in skeletal muscle.
The metabolic syndrome develops in the long run as a consequence of hepatic and
muscle lipotoxicity and of visceral obesity.
Fructose is also known to increase plasma uric acid, and this
effect may be involved in the development of insulin resistance.
In rats, fructose-induced hyperuricemia results in inhibition of
NO synthase. Because insulin-induced glucose utilization
involves not only the stimulation of key metabolic pathways in
insulin-sensitive cells but also a NO-dependent increase in
muscle blood flow [41], it was proposed that inhibition of the
vascular effects of insulin by uric acid was involved in fructose-
induced insulin resistance. In support of this hypothesis the
development of insulin resistance was prevented by lowering
uric acid concentrations with an uricosuric agent in fructose-fed
rats [42].
Are the effects of fructose different from those of glucose?
In our everyday diet, naturally occurring, free fructose
(essentially with fruits and honey) is a modest component of
energy intake. Furthermore, there is some evidence that
consumption of fructose with fruits or honey does not produce
the same adverse metabolic effects as added fructose. This may
be due to the presence of natural antioxidants and/or dietary
fibers with fruits and honey [20,21]. The vast majority of fructose
in our diet corresponds to added sugars, the two main sources
being sucrose (containing 50% fructose) and HFCS (containing
42%55% fructose). As a consequence, the intakes of fructose and
glucose always vary simultaneously, and therefore, high-fructose
consumers are also high-glucose consumers. Few studies have
however compared the effects of high-fructose versus high-
glucose diets. In rodents, a high-fructose diet was shown to
increase the expression of lipogenic genes, to suppress peroxi-
some proliferator activated receptor-alpha (PPAR-a)-dependent
lipid oxidation genes, and to cause intrahepatic fat deposition
and hypertriglyceridemia. In contrast, a high-glucose diet also
stimulated lipogenic gene expression, but failed to suppress lipid
oxidation genes or to produce hepatic steatosis or dyslipidemia
[43]. In humans, it has been shown that fructose, when
substituted for starch, increases plasma triglycerides [44]. It has
also been shown that, in obese hyperinsulinemic women,
administration of equivalent amounts of pure fructose, sucrose,
and HFCS led to identical increases in plasma triglyceride,
although the intake of fructose was about 50% less with sucrose
and HFCS compared to pure fructose [30]. In normal weight and
obese women overfed for 4 d with 50% glucose or fructose above
their energy requirement, de novo lipogenesis was found to be
identical under both conditions. There was also no significant
difference in plasma glucose, triacylglycerol, or insulin concen-
trations [45]. In healthy young normal weight subjects, a
6-d overfeeding with 3.5 g/kg fat-free mass/d (corresponding
roughly to 30% energy requirements) fructose or glucose
increased both VLDL-triacylglycerol concentrations and intra-
hepatic fat, with large interindividual variations, however [46].
In overweight subjects who received glucose or fructose drinks
corresponding to 30% of their energy requirements with ad
libitum food intake, body weight increased to the same extent
with both sugars, while impaired glucose tolerance developed
only in subjects receiving fructose. Interestingly, fructose
significantly increased visceral adipose tissue as estimated by
magnetic resonance imaging, while the increase with glucose
was of smaller magnitude and failed to reach statistical
significance [34].
Because all these experiments involved hypercaloric condi-
tions, one might wonder whether the observed effects can be
specifically attributed to sugars or are the mere consequence of
hypercaloric feeding. Similar short-term overfeeding experi-
ments were performed with a saturated fat supplementation
corresponding to 30% energy requirements: they demonstrated
that fat overfeeding also increased intrahepatic fat content, but
failed to increase plasma VLDL-triacylglycerol. Altogether, one
may conclude that excess calories, whether as simple sugars or as
saturated fat, increases intrahepatic fat content, and that sugars
specifically increase plasma VLDL-triglyceride. Fructose however
appears to exert this latter effect more potently than glucose.
Such short-term experiments remain however difficult to
extrapolate to dietary alterations of longer durations, for, with
time, significant changes in body composition are likely to occur.
Are the effects of free fructose different from bound
fructose?
In North America, HFCS has replaced a substantial portion of
sucrose over the past three decades. The concomitant increases
in the HFCS intake and in the prevalence of obesity have raised
concern regarding a possible causal role of HFCS [47]. In rodents,
feeding a diet rich in HFCS increased body weight and body fat
and caused dyslipidemia and insulin resistance. As for fructose,
HFCS feeding elicited an endoplasmic reticulum stress response
in hepatocytes. These effects of HFCS appear therefore very
comparable to those of sucrose [48,49]. HFCS differs from sucrose
by a slightly higher fructose content, and by providing glucose
and fructose as monosaccharides. There is however no evidence
that providing fructose as a hexose has different metabolic
effects than when fructose is consumed bound to glucose in
sucrose. In patients with type 2 diabetes, the glucose and insulin
responses to administration of 35 g of sucrose or HFCS were quite
similar [50]. Based on the fact that fructose inhibits less ghrelin
and increases less leptin than glucose [51], one may have
hypothesized that HFCS would have a different effect than
 L. Tappy et al. / Nutrition 26 (2010) 10441049
sucrose on these hormones and hence would have a lesser
satiating effect. It was however documented that it was not the
case, and that HFCS and sucrose produced similar leptin
increases and ghrelin suppression in healthy female volunteers
[52]. Furthermore, HFCS, sucrose, or equimolar glucose-fructose
mixtures elicited similar satiety responses [53] or energy intake
at a subsequent meal [54]. There is therefore no hint that the
effects of free fructose may differ from those of fructose bound to
glucose.
Do changes in fructose consumption explain the current
epidemics of metabolic disorders?
There is compelling evidence that a hypercaloric, high-
fructose diet can induce, not only in animal models, but also in
humans, a whole range of metabolic alterations, the most
prominent being a disturbance of hepatic lipid metabolism and
of plasma lipid profile. This, together with the observation that
total sugar and fructose intakes have increased significantly over
the past three decades, has led to the speculation that a high-
fructose intake may bear a direct, causal effect in the current
epidemics of obesity and related metabolic disorders [47,55].
Several surveys, using various epidemiologic methods to assess
dietary intakes, have consistently reported that total sugar
consumption, added sugar intake, and fructose have indeed
increased worldwide over the past three decades [56], reviewed
in [6]. In the US, average fructose consumption has increased
from about 37 g/d in the late 1970s [56] to 49 g/d in the 1999
2004 period [57]. Not only fructose, but also total energy intake,
increased during the same period.
Consumption of sweetened beverages accounts for the major
portion of total fructose consumption, the remaining being
mainly added sugar. As a result, glucose intake always covaries
with fructose intake, and epidemiologic studies cannot differ-
entiate between the effects of fructose per se and those specifi-
cally attributable to glucose. In this regard, it is therefore more
appropriate to evaluate the potential effect of fructose by
comparing the evolutions over time of sweetened beverages and
metabolic diseases. While doing so, one has however to keep in
mind that sweetened beverage consumption impacts on both
sugar consumption and total energy intake at the same time.
Several studies have assessed the relationships between
sweetened beverages consumption and energy intake, and
a meta-analysis clearly documented that sweetened beverages
consumption was associated with an excess energy intake [18].
Many cross-sectional studies have shown a positive association
between consumption of sweetened beverages, mainly in teen-
agers, and body weight [5865] (reviewed in [66]). Interven-
tional studies further showed that adding sweetened beverages
to the usual diet led to an increase in body weight [67,68], while
reducing sweetened beverages intake in overweight subjects
decreased body weight [6973]. Several large studies also
reported that sweetened beverage consumption increased the
risk of developing type 2 diabetes, but this effect was essentially
linked to body weight changes [7476]. A significant association
was observed between sweetened beverage consumption and an
increased incidence of heart disease. Here again, this was mainly
explained by a higher body weight in those consuming sweet-
ened beverages. The relationship nonetheless remained signifi-
cant after adjusting for body weight and may be explained either
by the high-glycemic index or by the high-fructose content of
sweetened beverages [77]. Finally it was reported that the
consumption of sweetened beverages was higher in patients
with non-alcoholic steatohepatitis than in healthy controls [78].
1047
Perspectives and conclusions
There is considerable evidence that a high-fructose intake can
indeed produce adverse metabolic alterations, the most prom-
inent ones being an increase in plasma triglycerides, hepatic
insulin resistance, and hepatic steatosis. These effects are
consistently observed in rodents fed a high-fructose diet and are
generally concomitant with an increased body mass. In humans,
many of these alterations (hepatic steatosis, hepatic insulin
resistance) are observed when fructose is experimentally
administered in amounts largely exceeding usual fructose intake,
and under hypercaloric conditions. When administered as part of
a weight-maintenance diet, fructose can indeed increase plasma
triglyceride. A recent meta-analysis [79] concluded that a fruc-
tose intake >50 g/d was already associated with altered plasma
triglyceride concentrations. Fructose consumption and sweet-
ened beverages are closely linked, and there is overwhelming
evidence that sweetened beverage intake favors weight gain and
that excess body weight in turn increases the risk of diabetes and
coronary heart diseases. It appears therefore sound at this stage
to advise limiting consumption of sugar calories to less than 140
kcal/d for men and 100 kcal/d for women (corresponding to
about one can of sweetened beverage/d), as recently proposed by
the American Heart Association [80]. Several important ques-
tions remain nonetheless unaddressed, and further studies are
clearly required to better delineate the potential adverse meta-
bolic effects of fructose when included in a weight-maintenance,
non-hypercaloric diet, to document the interactions between
fructose and other nutrients, and to evaluate whether there exist
subgroups of individuals (obese insulin-resistant patients,
offsprings of patients with type 2 diabetes) in whom the adverse
metabolic effects of fructose may possibly be enhanced.
Acknowledgments
The work performed on this topic in the authors laboratory
has been supported by Grants 067787, 109737, and 121995 from
the Swiss National Foundation for Science.
References
[1] Yudkin J. Dietary carbohydrate and ischemic heart disease. Am Heart J
1963;66:8356.
[2] Wolf A, Bray GA, Popkin BM. A short history of beverages and how our body
treats them. Obes Rev 2008;9:15164.
[3] White JS. Straight talk about high-fructose corn syrup: what it is and what
it aint. Am J Clin Nutr 2008;88:1716S21S.
[4] Le KA TL. Metabolic effects of fructose. Curr Opin Clin Nutr Metabol Care
2007;10:2104.
[5] Havel PJ. Dietary fructose: implications for dysregulation of energy
homeostasis and lipid/carbohydrate metabolism. Nutr Rev 2005;63:
13357.
[6] Tappy L, Le KA. Metabolic effects of fructose and the worldwide increase in
obesity. Physiol Rev 2010;90:2346.
[7] Douard V, Ferraris RP. Regulation of the fructose transporter GLUT5 in
health and disease. Am J Physiol Endocrinol Metab 2008;295:E22737.
[8] Corpe CP, Burant CF, Hoekstra JH. Intestinal fructose absorption: clinical
and molecular aspects. J Pediatr Gastroenterol Nutr 1999;28:36474.
[9] Kneepkens CM, Vonk RJ, Fernandes J. Incomplete intestinal absorption of
fructose. Arch Dis Child 1984;59:7358.
[10] Ravich WJ, Bayless TM, Thomas M. Fructose: incomplete intestinal
absorption in humans. Gastroenterology 1983;84:269.
[11] Mayes PA. Intermediary metabolism of fructose. Am J Clin Nutr
1993;58(suppl):754S65S.
[12] Parks EJ, Skokan LE, Timlin MT, Dingfelder CS. Dietary sugars stimulate
fatty acid synthesis in adults. J Nutr 2008;138:103946.
[13] ODell BL. Fructose and mineral metabolism. Am J Clin Nutr 1993;58:
771S8S.
[14] Johnson MA. Interaction of dietary carbohydrate, ascorbic acid and copper
with the development of copper deficiency in rats. J Nutr 1986;116:80215.
 REVIEW

CURRENT
OPINION Resistant starch: a promising dietary agent for the
prevention/treatment of inflammatory bowel disease
and bowel cancer
Janine A. Higgins a,b,c and Ian L. Brown a,b,c,d

Purpose of review
Resistant starch represents a diverse range of indigestible starch-based dietary carbohydrates. Resistant
starch has been investigated in the past for its effects on bowel health (pH, epithelial thickness, and
apoptosis of colorectal cancer cells); reduction in postprandial glycemia; increased insulin sensitivity; and
effects on the gut microbiome. This review highlights advances as resistant starch gains clinical relevance
as a potential treatment/preventive tool for diseases such as colorectal cancer (CRC) and diabetes.
Recent findings
Recent articles have evaluated the comparative physiological effects of different types of resistant starch
and investigated the effects of resistant starch on blood lipids, body weight, and defining resistant starch-
induced changes to the micriobiome that may be important in health and disease. The most novel and
relevant recent data describe a role for resistant starch in ameliorating inflammation; the use of resistant
starch for optimal bowel health and prevention of CRC; and, further, that the systemic effects of resistant
starch may be important for the treatment of other forms of cancer, such as breast cancer.
Summary
This review describes advances in resistant starch research highlighting the gastrointestinal effects that are
now being linked to systemic, whole body effects with clinical relevance. These effects have important
implications for overall health and the prevention or amelioration of various chronic diseases.
Keywords
cancer, inflammation, lipids, microbiome, resistant starch, weight control

INTRODUCTION It should be noted that it is rare for any starch or

Resistant starch is any starch or starch-derived
material that is not digested and absorbed in the
upper digestive tract and, so, passes into the large
bowel [1]. Its importance to health was first
suggested in the early 1980s and it has since been
the subject of extensive research. Resistant starch
comes in many forms with widely varying physical,
organoleptic, and physiological properties. Resist-
ant starch has been broadly described in five general
categories. RS1 represents starch physically inaccess-
ible to digestive enzymes due to the presence of seed
coats, germ, and so on (e.g. whole grains); RS2 are
starch granules which are inaccessible to amylases
due to granule starch structure or conformation (e.g.
high amylose maize starch); RS3 is the product of
retrograded starch (e.g. starch cooked then cooled);
RS4 encompasses starches that are chemically
modified to be resistant to digestion; and RS5 are
inclusion complexes formed by amylose with polar
lipids [2,3].
starch-derived material to be comprised of only
resistant starch; usually there is also a digestible
component. New forms of RS2 are, or will in the
near future, be available from high amylose varieties
of barley, wheat, and rice produced by transgenic or
conventional breeding techniques [3].
A current area of debate is whether RS4 com-
pounds have the same physiological effects as the
extensively studied RS2 and RS3 starches. This is a
topic of some importance as RS4 products are
versatile and can be easily and inexpensively
a
Center for Human Nutrition, bDepartment of Pediatrics, Anschutz
Medical Campus, University of Colorado Denver, cChildrens Hospital
Colorado, Aurora, Colorado, USA and dSchool of Medicine, Faculty of
Health Sciences, Flinders University, South Australia, Australia
Correspondence to Ian Brown, P.O. Box 405, Gymea, NSW 2227,
Australia. E-mail: Ian.brown@clovercorp.com.au
Curr Opin Gastroenterol 2013, 29:190194
DOI:10.1097/MOG.0b013e32835b9aa3

www.co-gastroenterology.com Volume 29 ! Number 2 ! March 2013

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Resistant starch for CRC and weight management Higgins and Brown
KEY POINTS
! Resistant starch has a positive impact on bowel health
and is associated with increased apoptosis of
colorectal cancer (CRC) tumors.
! Resistant starch reduces body weight and prevents
weight regain after weight loss in obese rodents.
! Resistant starch decreases inflammation which is
important due to the presence of inflammation in many
chronic diseases.
! Resistant starch, between and within each category,
may have significantly different physiological
properties.
engineered to deliver resistant starch and other
compounds to specific areas of the colon. The pub-
lications dealing with this topic are pointed out in
this review.
BODY WEIGHT
Many studies have previously tried to define the
effect of resistant starch consumption on body
weight and adiposity. The physiological effects of
resistant starch make it extremely difficult to assess
its impact on weight. Resistant starch increases stool
bulk, luminal thickness, and bacterial numbers in
the colon. These factors add to overall body weight,
perhaps masking any differences that do exist due to
resistant starch ingestion. In animal studies, this can
be overcome at the end of the study by measuring
body weight with and without the gut and excising
individual fat pads for weighing solutions that
cannot be employed in a human population. Recent
data provide contradictory evidence which is similar
to that previously published. In rats, resistant starch
had no effect on total body, colon, liver, or epidi-
dymal fat pad weights [4]. In old mice, resistant
starch also had no effect on total body weight [5].
However, other studies show decreased body weight
[6,7] and liver weight with no change in the weight
of epididymal fat pads following resistant starch
feeding [7]. In a study which lends credence to
the idea that individual responses to resistant starch
are highly variable in part because of the plant
source of resistant starch, RS2 added to wheat bread
had no effect on body weight, but, added to maize
bread, RS2 caused decreased weight gain in rats [8].
A novel area of research was pursued recently by
examining the effect of resistant starch and exercise
on weight regain following weight loss in obese rats
[9 ]. Resistant starch was almost as effective as
&&
exercise at ameliorating the rate of weight regain
0267-1379 ! 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
after weight loss. In addition, the weight regained by
the exercise group was all fat at the expense of lean
body mass, whereas the addition of a resistant starch
diet to exercise caused accrual of lean body mass
with much less fat deposition which would confer
several metabolic advantages and make weight loss
easier to maintain [9 ]. Resistant starch lowered
&&
cortisol concentrations regardless of exercise con-
dition and prevented weight regain even if exercise
was stopped as commonly occurs with humans on
weight loss regimens. Preliminary data from a
microarray study in mice show that RS2 affects
the expression of genes associated with energy
metabolism which may be part of the mechanism
whereby colonocytes might exert influence on
whole body energy metabolism [10]. These exciting
data need to be confirmed in a thorough clinical
trial.
GLUCOSE RESPONSE AND INSULIN
SENSITIVITY
Very little novel work remains in this field due to
the exhaustive nature of previous studies. Several
studies utilizing RS2 and RS3 have confirmed the
many previous studies showing that resistant starch
decreases postprandial glycemia [6,11,12], although
one study reported no difference in glycemic
response between RS2 and digestible starch con-
sumption [13]. The most novel research in this field
used a well designed trial to show that RS2 and
several forms of RS4 exhibited the same reduction
in postprandial glucose responses relative to diges-
tible starch [14 ]. Neither RS2 nor RS4 had any effect
&
on satiety over 2 h [14 ]. This is important infor-
&
mation as different RS4 sources can be quickly and
inexpensively engineered to achieve specific results
in terms of fermentation profile. It remains an
important area of investigation to define short chain
fatty acid (SCFA) profile and the subsequent meta-
bolic effects of different forms of RS4.
Previous studies have shown that resistant
starch consumption improves insulin sensitivity
in rats and humans. New data using RS2 in pigs
[15] and overweight men with high waist circum-
ference [16] corroborate these earlier findings. In a
result that warrants further investigation, a sex
difference was noted in the latter study. The
improvement in insulin sensitivity in response to
resistant starch consumption shown in men was not
apparent in an equivalent cohort of women [16]. In
insulin-resistant adults, resistant starch intake had
no effect on whole body insulin sensitivity, but did
improve glucose uptake in forearm muscle and
increased free fatty acid suppression in adipose tis-
sue [17 ], indicating that resistant starch improves
&
www.co-gastroenterology.com 191
 Nutrition
peripheral insulin sensitivity, but not to a magni-
tude that is detectable on the whole body level. As
insulin resistance is associated with the develop-
ment and complications of diabetes and cardio-
vascular disease (CVD), it remains important to
characterize this parameter in different populations,
especially obese individuals who are at highest risk
for these diseases.
INFLAMMATION
Only one current publication specifically addresses
the issue of resistant starch ingestion and inflam-
mation. As inflammation is a characteristic feature
of many chronic diseases such as obesity, diabetes,
some cancers, and CVD, this is an area that has the
potential to expand rapidly. RS3 ingestion in wild-
type mice decreased ileal and colonic inflammatory
lesions and increased IL-10-expressing cells and
peroxisome proliferator-activated receptor gamma
expression relative to a low fat diet [18 ]. In IL-10-/-
&&
mice (model of cholitis), under the same conditions,
RS3 caused lower ileal and colonic inflammatory
lesions and lower IFNg in lesions [18 ]. These are
&&
promising data which provide evidence that dietary
resistant starch may play a role in attenuating some
of the inflammatory responses in chronic disease
and provide a therapeutic role for resistant starch
consumption in the future.
LIPIDS
The effect of resistant starch on circulating lipids has
long been an area that is understudied and subject to
contradictory evidence. Data from rodents show
quite convincingly that chronic resistant starch
ingestion reduces circulating cholesterol and trigly-
ceride concentrations. However, data from human
studies have always been equivocal, possibly due to
the length of resistant starch feeding and the study
populations examined. Most recently, pig studies
show that RS2 feeding reduces fasting serum trigly-
ceride and cholesterol concentrations [7,15]. In
addition, resistant starch caused lower liver triacyl-
glycerol [15] and cholesterol concentrations [7],
and lower adipose triglyceride concentrations [7].
Hepatic expression of lipogenic genes was lower,
whereas expression of genes of the lipolytic pathway
was higher in response to resistant starch [15].
Thus, this is the field that requires more clinical
data before a sound judgment can be made regard-
ing any effect of resistant starch on lipid concen-
trations.
BOWEL EFFECTS: FERMENTATION AND
MICROBIOME
Previous work in the area of bowel health has
focused on measuring pH and SCFA concentrations
192 www.co-gastroenterology.com
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
in different segments of the bowel and correlating
these with improvements in bowel function and
decreases in pathogenic bacteria. However, with
the advent of cheaper and faster microbiome
analysis, more recent work has focused on fermen-
tation and the abundance of different bacterial
species. This work currently stands in a descriptive
phase with the future open to describe which bac-
terial populations and what interactions between
different species and strains confer which pheno-
typic and health benefits.
In a study confirming previous rodent work,
resistant starch increased intestinal nutrient flow
and caused selective proliferation of Bifidobacterium
[19] and Bacillus species [20] in pigs. In rats, RS2
selectively increases propionate concentrations and
Ruminococcus bromii abundance, whereas RS4 feed-
ing increases butyrate and propionate concen-
trations and is associated with proliferation of
Lactobacillus gasseri and Parabacteroides distasonis
[21]. A human study has recently confirmed that
RS4, in the form of butyrylated starch, leads to
higher butyrate delivery to the colon than RS2,
which, overall, increases both butyrate and propio-
nate concentrations in the colon compared to an
equivalent dose of RS2 [22].
A problem in this arena of research has been
the variability of response to resistant starch
consumption. Since a particular type of resistant
starch can have a specific effect on bacteria, the
description of the source of resistant starch in a
particular experiment is essential for the interpret-
ation of the data. Individual SCFA responses to
resistant starch are highly variable [23] as is the
amount of energy that is salvaged from fermenta-
tion [24 ]. Two recent publications give a clue to
&
the source of this variability. The first describes
in-vitro experiments showing that resistant starch
from the same class (RS3) but derived from differ-
ent plant sources in conjunction with different
bacteria yield very different SCFA levels [25]. The
second indicates that R. bromii is a key species for
fermentation of resistant starch in the human
colon and that R. bromii facilitates resistant starch
fermentation by other species, even by bacteria
that are weak RS2 and RS3 fermenters in isolation
(Eubacterium rectal, Bacteroides thetaiotaomicron)
[24 ]. As every individual has a unique microbiome
&
footprint and consumes resistant starch from a
variety of classes and plant sources, individual
responses to resistant starch consumption would
be expected to be very high. Perhaps, in future
studies, assessing the basal microbiome of the
patient and using very well characterized resistant
starch sources would glean some more consistent
and interpretable data.
Volume 29 ! Number 2 ! March 2013
 Resistant starch for CRC and weight management Higgins and Brown
CANCER
Due to the effects of resistant starch that promote
bowel health, its usefulness as part of colorectal
cancer (CRC) prevention and treatment plans has
been investigated for many years. These efforts con-
tinue and remain a sizable portion of all research
published over the past 18 months. In vitro, the
fermented, nondigestible starch portion of a com-
mon bean, very high in resistant starch, decreased
survival of HT-29 human colon adenocarcinoma
cells in a dose-dependent manner [26]. In addition,
DNA fragmentation was similar between the high
resistant starch bean fraction and a SCFA mixture
similar to that produced through the beans fermen-
tation, indicating that SCFA could be mediating this
effect. In studies that confirm previous reports, the
carcinogenic effects of red meat consumption on
the colon are significantly reduced by the inclusion
of resistant starch in the diet of rodents [27,28 ].
&
This may be because the addition of resistant starch
to high red meat diets changes the pattern of fer-
mentation from protein substrates to carbohydrate
substrates thereby decreasing the production of pro-
mutagenic adducts that arise during protein fermen-
tation [28 ] and/or via an increase in telomere
&
length which protects against DNA damage [27].
Colon damage is often induced in rats by inject-
ing the carcinogen azoxymethane (AOM) and
assessing the health of the epithelium and rates of
apoptosis 6 h later. This classical experimental para-
digm is currently being used to test the efficacy of
RS4 products. Such studies show that butyrylated
RS4 yielded the highest butyrate and propionate
concentrations in the serum and along the length
of the colon and the highest rates of apoptosis,
relative to RS2 and control starches [29], and that,
in the raw form, RS2 causes the greatest fermenta-
tion and lowest aberrant crypt formation, but, under
cooked conditions, RS4 yields a greater amount of
fermentation and a lower aberrant crypt formation
[30]. In a study comparing colon damage from
moderate fat and protein (Western diet) feeding
in rats, RS2 and RS4 had equivalent effects, causing
less single-stranded DNA breaks and higher SCFA
concentrations in the portal vein and cecum than
rats fed a low fat or Western diet containing diges-
tible starch [4].
The only clinical trial to examine the role of
resistant starch in CRC prevention was designed to
test the effects of both aspirin and resistant starch on
polyp number and size in individuals with familial
adenomatous polyposis, which is a rare and aggres-
sive form of CRC. This study showed no effect of
resistant starch supplementation on either polyp
load or maximal polyp size over 19 years [31].
The length of the study and the high dose of
0267-1379 ! 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
resistant starch used are strengths of the study.
However, this trial is not definitive and more data
need to be gathered in humans as there were few
individuals who had biopsies after 1 year and the
numbers were not distributed evenly across all
groups which greatly reduced statistical power after
the first year. This was an international multicenter
study which decreased the consistency of endos-
copy reports and the areas of the colon which could
be examined across individuals. Therefore, this
study provides interesting preliminary data which
need to be explored more carefully in the future.
An area of usefulness for resistant starch is its use
in targeted drug delivery which may aid the effec-
tiveness of CRC chemotherapy approaches. Both
small and large compounds can be taken per orum
and delivered to the bowel using resistant starch
acetate coating [32]. The thickness and composition
of the coating can be changed to achieve targeted
drug release [32,33], which carries important ram-
ifications for CRC treatment. Undoubtedly, this is
an area of research which is in its infancy, but will
expand rapidly in the near future.
The impact of resistant starch in the cancer
field is set to expand rapidly beyond CRC as the
first report of resistant starch feeding on reducing
breast cancer proliferation had positive results. In
this study, ovariectomized nude mice were supple-
mented with estrogen, at levels approximately
equivalent to postmenopausal women, then
injected with Michigan Cancer Foundation-7 (estro-
gen responsive human breast cancer) tumors sub-
cutaneously [34 ]. Twelve weeks of resistant starch
&&
feeding significantly reduced mean tumor size,
lowered circulating estradiol concentrations, and
decreased proliferation and increased apoptotic
protein and mRNA expression, relative to controls
[34 ]. There was a positive correlation between
&&
circulating estrogen and mean tumor size. These
are exciting new data which pave the way for future
resistant starch research beyond the realm of CRC.
CONCLUSION
Much recent data corroborate earlier evidence for
the reduction in glycemia, improvement in insulin
sensitivity, and positive bowel effects, including
amelioration of CRC, that occur with resistant
starch ingestion. The most prominent recent work
includes the first data from breast cancer studies in
rodents, the positive effects of resistant starch in
preventing weight regain after weight loss in obese
rodents, and an initial study defining the effects of
resistant starch on systemic inflammation. These
three areas are of potential clinical relevance and
warrant further examination. Should the descriptive
www.co-gastroenterology.com 193
 Nutrition
data regarding the microbiome effects of resistant
starch translate in prospective, intervention studies,
further uses for resistant starch in the treatment of
colitis and other inflammatory bowel diseases might
develop in the future. Resistant starch sources can be
manipulated to achieve the optimal desired physio-
logical response or to target drug delivery to the
colon. Resistant starch can be incorporated into
foods such as breakfast cereals, pizza bases, breads,
candy, pasta, yogurt, cheeses, and smoothies. Thus,
resistant starch is a versatile carbohydrate that
potentially offers an inexpensive, easy, and practical
way to impact health and prevent or treat many
chronic diseases in a broad population.
Acknowledgements
None.
Conflicts of interest
There are no conflicts of interest.
Supported in part by the National Institutes of Health
(DK 038088 and DK43607; Colorado CTSI Grant UL1
RR025780). Contents are the authors sole responsibility
and do not necessarily represent official NIH views.
REFERENCES AND RECOMMENDED
READING
Papers of particular interest, published within the annual period of review, have
been highlighted as:
& of special interest
&& of outstanding interest
Additional references related to this topic can also be found in the Current
World Literature section in this issue (p. 237).
1. Asp NG, Bjorck I, Holm J, et al. Enzyme resistant starch fractions and dietary
fibre. Scand J Gastroenterol Suppl 1987; 129:2932.
2. Englyst KN, Liu S, Englyst HN. Nutritional characterization and measurement
of dietary carbohydrates. Eur J Clin Nutr 2007; 61 (Suppl 1):S1939.
3. Topping D. The contribution of resistant starch to health benefits of dietary
fibre. J Japan Assoc Dietary Fiber Res 2012; 16 (Suppl A):s11s19.
4. Conlon MA, Kerr CA, McSweeney CS, et al. Resistant starches protect
against colonic DNA damage and alter microbiota and gene expression in rats
fed a Western diet. J Nutr 2012; 142:832840.
5. Zhou J, Keenan MJ, Keller J, et al. Tolerance, fermentation, and cytokine
expression in healthy aged male C57BL/6J mice fed resistant starch. Molec
Nutr Food Res 2012; 56:515518; Epub 2011/12/17.
6. Zhu L, Gu M, Meng X, et al. High-amylose rice improves indices of animal
health in normal and diabetic rats. Plant Biotechnol J 2012; 10:353362.
7. Ha AW, Han GJ, Kim WK. Effect of retrograded rice on weight control, gut
function, and lipid concentrations in rats. Nutr Res Pract 2012; 6:1620;
Epub 2012/03/14..
8. Brites CM, Trigo MJ, Carrapico B, et al. Maize and resistant starch enriched
breads reduce postprandial glycemic responses in rats. Nutr Res 2011;
31:302308.
9. Higgins JA, Jackman MR, Brown IL, et al. Resistant starch and exercise
&& independently attenuate weight regain on a high fat diet in a rat model of
obesity. Nutr Metab 2011; 8:49; Epub 2011/07/09..
This is the first study to comprehensively show the independent and additive effects
of resistant starch consumption on weight maintenance after weight loss in obesity.
10. Keenan MJ, Martin RJ, Raggio AM, et al. High-amylose resistant starch
increases hormones and improves structure and function of the gastrointest-
inal tract: a microarray study. J Nutrigenet Nutrigenom 2012; 5:2644; Epub
2012/04/21.
11. Hallstrom E, Sestili F, Lafiandra D, et al. A novel wheat variety with elevated
content of amylose increases resistant starch formation and may beneficially
influence glycaemia in healthy subjects. Food Nutr Res 2011; 55:22.
12. Karupaiah T, Aik CK, Heen TC, et al. A transgressive brown rice mediates
favourable glycaemic and insulin responses. J Sci Food Agric 2011;
91:19511956.
194 www.co-gastroenterology.com
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
13. Bodinham CL, Smith L, Wright J, et al. Dietary fibre improves first-phase
insulin secretion in overweight individuals. PloS One 2012; 7:e40834; Epub
2012/07/21..
14. Haub MD, Louk JA, Lopez TC. Novel resistant potato starches on glycemia and
& satiety in humans. J Nutr Metab 2012; 2012:478043; Epub 2012/06/02..
This is the only study which describes the glycemic and satiety effects of several
forms of RS4 vs. RS2.
15. He J, Chen D, Zhang K, Yu B. A high-amylopectin diet caused hepatic
steatosis associated with more lipogenic enzymes and increased serum
insulin concentration. Br J Nutr 2011; 106:14701475.
16. Maki KC, Pelkman CL, Finocchiaro ET, et al. Resistant starch from high-
amylose maize increases insulin sensitivity in overweight and obese men.
J Nutr 2012; 142:717723.
17. Robertson MD, Wright JW, Loizon E, et al. Insulin-sensitizing effects on
& muscle and adipose tissue after dietary fiber intake in men and women with
metabolic syndrome. J Clin Endocrinol Metab 2012; 97:33263332; Epub
2012/06/30..
This study eloquently describes the tissue-specific effect of resistant starch on
insulin sensitivity in a patient population which is at greatest risk of diabetic
complications and CVD.
18. Bassaganya-Riera J, DiGuardo M, Viladomiu M, et al. Soluble fibers and
&& resistant starch ameliorate disease activity in interleukin-10-deficient mice
with inflammatory bowel disease. J Nutr 2011; 141:13181325.
This is the first comprehensive investigation of resistant starch and inflammation
which paves the way for future studies that have ramifications for the prevention
and/or treatment of inflammatory bowel disease.
19. Regmi PR, Metzler-Zebeli BU, Ganzle MG, et al. Starch with high amylose
content and low in vitro digestibility increases intestinal nutrient flow and
microbial fermentation and selectively promotes bifidobacteria in pigs. J Nutr
2011; 141:12731280.
20. Han GQ, Xiang ZT, Yu B, et al. Effects of different starch sources on Bacillus
spp. in intestinal tract and expression of intestinal development related genes
of weanling piglets. Mol Biol Rep 2012; 39:18691876.
21. J Abell GC, Christophersen CT, McOrist AL, Clarke JM. Dietary resistant and
butyrylated starches have different effects on the faecal bacterial flora of azox-
ymethane-treated rats. Br J Nutr 2011; 105:14801485 [Epub 2011/01/25].
22. Clarke JM, Topping DL, Christophersen CT, et al. Butyrate esterified to starch
is released in the human gastrointestinal tract. Am J Clin Nutr 2011;
94:12761283.
23. McOrist AL, Miller RB, Bird AR, et al. Fecal butyrate levels vary widely among
individuals but are usually increased by a diet high in resistant starch. J Nutr
2011; 141:883889.
24. Ze X, Duncan SH, Louis P, Flint HJ. Ruminococcus bromii is a keystone
& species for the degradation of resistant starch in the human colon. ISME J
2012 [Epub 2012/02/22].
This publication identifies one key species of bacteria that could alter the
fermentation profile of many types of resistant starch and, therefore, alter SCFA
production.
25. Purwani EY, Purwadaria T, Suhartono MT. Fermentation RS3 derived from
sago and rice starch with Clostridium butyricum BCC B2571 or Eubacterium
rectale DSM 17629. Anaerobe 2012; 18:5561.
26. Cruz-Bravo RK, Guevara-Gonzalez R, Ramos-Gomez M, et al. Fermented
nondigestible fraction from common bean (Phaseolus vulgaris L.) cultivar Negro
8025 modulates HT-29 cell behavior. J Food Sci 2011; 76:17503841.
27. OCallaghan NJ, Toden S, Bird AR, et al. Colonocyte telomere shortening is
greater with dietary red meat than white meat and is attenuated by resistant
starch. Clin Nutr 2012; 31:6064.
28. Winter J, Nyskohus L, Young GP, et al. Inhibition by resistant starch of red
& meat-induced promutagenic adducts in mouse colon. Cancer Prev Res (Phila)
2011; 4:19201928; Epub 2011/09/03.
This study provides mechanisic data on the role of resistant starch in ameliorating
the proliferation of CRC cells.
29. Clarke JM, Young GP, Topping DL, et al. Butyrate delivered by butyrylated
starch increases distal colonic epithelial apoptosis in carcinogen-treated rats.
Carcinogenesis 2012; 33:197202.
30. Zhao Y, Hasjim J, Li L, et al. Inhibition of azoxymethane-induced preneoplastic
lesions in the rat colon by a cooked stearic acid complexed high-amylose
cornstarch. J Agric Food Chem 2011; 59:97009708; Epub 2011/07/26.
31. Burn J, Bishop DT, Chapman PD, et al. A randomized placebo-controlled
prevention trial of aspirin and/or resistant starch in young people with familial
adenomatous polyposis. Cancer Prev Res 2011; 4:655665.
32. Pu H, Chen L, Li X, et al. An oral colon-targeting controlled release system
based on resistant starch acetate: synthetization, characterization, and pre-
paration of film-coating pellets. J Agric Food Chem 2011; 59:57385745.
33. Chen L, Pu H, Li X, Yu L. A novel oral colon-targeting drug delivery system
based on resistant starch acetate. J Control Release 2011; 30:e51e52.
34. Kondegowda NG, Meaney MP, Baker C, Ju YH. Effects of nondigestible
&& carbohydrates on the growth of estrogen-dependent human breast cancer
(MCF-7) tumors implanted in ovariectomized athymic mice. Nutr Cancer
2011; 63:5564.
This is the first article which describes an effect of resistant starch on a cancer
outside of the colon, namely breast cancer. These data provide evidence that the
systemic effects of resistant starch ingestion may be useful to treat many types of
cancer and chronic diseases.
Volume 29 ! Number 2 ! March 2013
 www.thelancet.com

Maternal and Child Nutrition

Executive Summary of The Lancet Maternal and Child Nutrition Series

Nutrition is crucial to both individual and national development. The evidence in

this Series furthers the evidence base that good nutrition is a fundamental driver
of a wide range of developmental goals. The post-2015 sustainable development
agenda must put addressing all forms of malnutrition at the top of its goals.
 Executive Summary
Maternal and Child Nutrition
Maternal and child undernutrition, consisting of
stunting, wasting, and deficiencies of essential
vitamins and minerals, was the subject of a Series
of papers in The Lancet in 2008.15 In the Series, we
quantified the prevalence of these issues, calculated
their short-term and long-term consequences, and
estimated their potential for reduction through
high and equitable coverage of proven nutrition
interventions.
The 2008 Series identified the need to focus on the
crucial period from conception to a childs second
birthdaythe 1000 days in which good nutrition and
healthy growth have lasting benefits throughout life.
The Series also called for greater priority for national
nutrition programmes, stronger integration with health
programmes, enhanced intersectoral approaches, and
more focus and coordination in the global nutrition
system of international agencies, donors, academia, civil
society, and the private sector.
5 years after the initial series, we re-evaluate the
problems of maternal and child undernutrition and
also examine the growing problems of overweight
and obesity for women and children and their
consequences in low-income and middle-income
Benefits during the life course
Morbidity and Cognitive, motor,
mortality in childhood socioemotional development
(
Nutrition specific Optimum fetal and child nutrition and development
interventions
and programmes
countries (LMICs). Many of these countries are
said to have the double burden of malnutrition
continued stunting of growth and deficiencies of
essential nutrients along with the emerging issue of
obesity. We also assess national progress in nutrition
programmes and international eorts toward previous
recommendations.
The first paper6 examines the prevalence and
consequences of nutritional conditions during the life
course from adolescence (for girls) through pregnancy
to childhood and discusses the implications for
adult health. The second paper7 covers the evidence
supporting nutrition-specific interventions and the
health outcomes and cost of increasing their population
coverage. The third paper8 examines nutrition-sensitive
interventions and approaches and their potential
to improve nutrition. The fourth paper9 discusses
the features of an enabling environment that are
needed to provide support for nutrition programmes,
and how they can be favourably influenced. A set of
Comments1015 examine what is currently being done,
and what should be done nationally and internationally
to address nutritional and developmental needs of
women and children in LMICs.
School performance Adult stature Work capacity
and learning capacity and productivity
( ).
$

Nutrition sensitive
programmes and approaches
1
Agriculture and food security
1
Adolescent health and
preconception nutrition
1
)'$"
dietary
supplementation
1
 '%$*)' $)

Breastfeeding, nutrient-
rich foods, and eating
routine
 $
$
' + $ Low burden of 1
% "
().
$)(
practices, parenting, infectious diseases 1
'".
 "
development
stimulation 1
)'$"
#$)"
")
1
Womens empowerment
1
Child protection
supplementation or %%
(*' )., including  $
$
' + $ Access to and use of 1
Classroom education
fortification availability, economic resources (maternal, health services, a safe and 1
Water and sanitation
1
'() $
$
access, and use of food household, and hygienic environment 1

")
$
# ".
&"$$ $
('+ (
complementary feeding community levels)
1

)'.
(*&&"#$)) %$

for children
1

)'.
diversification Building an enabling environment
Knowledge and evidence 1
 %'%*(
+"*) %$(
1
 $
+ %*'(
$
Politics and governance 1
Advocacy strategies
stimulation
Leadership, capacity, and financial resources 1

%' /%$)"
$
+') "
%%' $) %$
1
Treatment of severe acute
Social, economic, political, and environmental context (national and global) 1
Accountability, incentives regulation,
malnutrition
1

((
&'+$) %$
$ legislation
management 1
'( &
&'%'##(
1
*)' ) %$
$)'+$) %$(
$ 1
Capacity investments
emergencies 1

%#() 
'(%*'
#% " () %$

Figure 1: Framework for actions to achieve optimum fetal and child nutrition and development

2 www.thelancet.com
 Executive Summary

A new conceptual framework Panel 1: Definition of nutrition-specific and nutrition-sensitive interventions

The present Series is guided by a framework and programmes
(figure 1) that shows the means to optimum fetal
Nutrition-specific interventions and programmes
and child growth and development.6 This framework   of fetal and
outlines the dietary, behavioural, and health child nutrition and developmentadequate food and nutrient intake, feeding,
determinants of optimum nutrition, growth, and caregiving and parenting practices, and low burden of infectious diseases
development, and how they are aected by underlying  

food security, caregiving resources, and environmental

conditions, which are in turn shaped by economic 
and social conditions, national and global contexts, 
capacity, resources, and governance. The Series focuses nutrition in emergencies
on how these determinants can be changed to enhance Nutrition-sensitive interventions and programmes
growth and development, including the nutrition-
specific interventions that address the immediate
causes of suboptimum growth and development
and the potential eects of nutrition-sensitive
interventions that address the underlying determinants
of malnutrition and incorporate specific nutrition goals
and actions (panel 1). It also shows how an enabling
environment can be built to support interventions and
programmes to enhance growth and development.
An unfinished agenda for undernutrition
The publication of The Lancet Maternal and Child
Undernutrition Series 5 years ago stimulated a
tremendous increase in political commitment to
reduction of undernutrition at global and national
levels. Most development agencies have revised their
strategies to address undernutrition focused on the
1000 days during pregnancy and the first 2 years of life,
as called for in the 2008 Series. One of the main drivers
of this new international commitment is the Scaling Up
Nutrition (SUN) movement.18,19 National commitment
in LMICs is growing, donor funding is rising, and civil
society and the private sector are increasingly engaged.
However, this progress has not yet translated
into substantially improved outcomes globally.
Improvements in nutrition still represent a massive
unfinished agenda. The 165 million children with
stunted growth have compromised cognitive
development and physical capabilities, making yet
another generation less productive than they would
otherwise be.6 Countries will not be able to break out
of poverty and sustain economic advances without
ensuring that their populations are adequately
nourished. Undernutrition reduces a nations economic
advancement by at least 8% because of direct
www.thelancet.com
 


safe and hygienic environmentand incorporate specific nutrition goals and actions
 
interventions, potentially increasing their scale, coverage, and eectiveness
 


Adapted from Scaling Up Nutrition16 and Shekar and colleagues, 2013.17
productivity losses, losses via poorer cognition, and
losses via reduced schooling.20 We cannot aord for
nothing to change.
Burden of nutritional conditions
Undernutrition in LMICs
Stunted linear growth has become the main indicator of
childhood undernutrition, because it is highly prevalent
in nearly all LMICs, and has important consequences for
health and development. It should replace underweight
as the main anthropometric indicator for children. The
prevalence of stunting in children younger than 5 years
in LMICs in 2011 was 26%, a decrease from 40% in
1990, and 32% in 2005, the estimate in the previous
nutrition Series.1,6 The number of stunted children has
also decreased globally, from 253 million in 1990, to 178
million in 2005, to 165 million in 2011. This represents
an average annual rate of reduction of 21%.6
The World Health Assembly (WHA) called for a 40%
reduction in the global number of children younger
than 5 years who are stunted by 2025 (compared with
the baseline of 2010).21 This aim would translate into a
39% reduction per year and imply reducing the number
of stunted children from 171 million in 2010, to about
100 million in 2025.6 At the present rate of decline,
3
 Executive Summary

stunting is expected to reduce to 127 million, a 25% from 56 to 61 million, whereas Asia is projected to show
 a substantial decrease in stunting prevalence.
        The prevalence of wasting was 8% globally in 2011,
the largest number of children aected by stunting, aecting 52 million children younger than 5 years, an
69 million, live in south-central Asia. In Africa, only small 11% decrease from an estimated 58 million in 1990.6
improvements are anticipated on the basis of present The prevalence of severe wasting was 29%, aecting
trends, with the number of aected children increasing 19 million children.6 70% of the worlds children with
wasting live in Asia, mostly in south-central Asia, where
Attributable Proportion of Attributable Proportion of an estimated 15% (28 million) are aected.6
deaths with total deaths deaths with total deaths of
UN of children NIMS children
Deficiencies of essential vitamins and minerals
prevalences* younger than prevalences younger than are widespread and have substantial adverse eects
5 years 5 years
on child survival and development.6 Deficiencies of
Fetal growth restriction (<1 month) 817 000 118% 817 000 118%
vitamin A and zinc adversely aect child health and
Stunting (159 months) 1 017 000* 147% 1 179 000 170%
survival, and deficiencies of iodine and iron, together
Underweight (159 months) 999 000* 144% 1 180 000 170%
Wasting (159 months) 875 000* 126% 800 000 115% with stunting, contribute to children not reaching their
Severe wasting (159 months) 516 000* 74% 540 000 78% developmental potential. Much progress has been made
Zinc deficiency (1259 months) 116 000 17% 116 000 17% in addressing vitamin A deficiency but eorts must
Vitamin A deficiency (659 months) 157 000 23% 157 000 23% continue at present coverage levels to avoid regressing
Suboptimum breastfeeding 804 000 116% 804 000 116%
(023 months)
because dietary intake of vitamin A is still inadequate.
Joint eects of fetal growth restriction 1 348 000 194% 1 348 000 194% Additionally, micronutrient deficiencies have an
and suboptimum breastfeeding in important part to play in maternal health.6
neonates
Joint eects of fetal growth restriction, 3 097 000 447% 3 149 000 454% Breastfeeding practices are far from optimum,
suboptimum breastfeeding, stunting, despite improvements in some countries. Suboptimum
wasting, and vitamin A and zinc
deficiencies (<5 years) breastfeeding results in an increased risk for mortality
in the first 2 years of life and results in 800 000 deaths
Data are to the nearest thousand. *Prevalence estimates from the UN. Prevalence estimates from Nutrition Impact
Model Study (NIMS). annually.6
Table 1: Global deaths in children younger than 5 years attributed to nutritional disorders
Maternal, newborn, and child nutrition
New evidence further reinforces the importance of the
Key messages on disease burden due to nutritional conditions
nutritional status of women at the time of conception
 Iron and calcium deficiencies contribute substantially to maternal deaths and during pregnancy, both for the health of the mother
 Maternal iron deficiency is associated with babies with low weight (<2500 g) at birth
and for ensuring healthy fetal growth and development.
 Maternal and child undernutrition, and unstimulating household environments,
contribute to deficits in childrens development and health and productivity in adulthood 32 million babies are born small-for-gestational-age
 Maternal overweight and obesity are associated with maternal morbidity, preterm (SGA) annuallyrepresenting 27% of all births in LMICs.
birth, and increased infant mortality Fetal growth restriction causes more than 800 000
 Fetal growth restriction is associated with maternal short stature and underweight deaths each year in the first month of lifemore than
and causes 12% of neonatal deaths
a quarter of all newborn deaths.6 This new finding
 Stunting prevalence is slowly decreasing globally, but aected at least 165 million
 contradicts the widespread assumption that babies who
 Suboptimum breastfeeding results in more than 800 000 child deaths annually are born SGA, by contrast with preterm babies, are not
 Undernutrition, including fetal growth restriction, suboptimum breastfeeding, at a substantially increased risk of mortality. Neonates
stunting, wasting, and deficiencies of vitamin A and zinc, cause 45% of child deaths, with fetal growth restriction are also at substantially
resulting in 31 million deaths annually
increased risk of being stunted at 24 months and of
 Prevalence of overweight and obesity is increasing in children younger than 5 years
globally and is an important contributor to diabetes and other chronic diseases development of some types of non-communicable
in adulthood diseases in adulthood.6
 Undernutrition during pregnancy, aecting fetal growth, and the first 2 years of life is Undernutrition (fetal growth restriction, suboptimum
a major determinant of both stunting of linear growth and subsequent obesity and breastfeeding, stunting, wasting, and deficiencies
non-communicable diseases in adulthood
of vitamin A and zinc) causes 45% of all deaths of

4 www.thelancet.com

children younger than 5 years, representing more
than 3 million deaths each year (31 million of the
69 million child deaths in 2011).6 Fetal growth
restriction and suboptimum breastfeeding together
cause more than 13 million deaths, or 194% of all
deaths of children younger than 5 years, representing
435% of all nutrition-related deaths (table 1).
Good nutrition early in life is also essential for children
 maternal morbidity and infant mortality.6
nutrition often coincides with other developmental
risks, in particular inadequate stimulation during early
childhood.6 Interventions to promote home stimulation
and learning opportunities in addition to good nutrition
will be needed to ensure optimum early development
and longer-term gains in human capital.6
This new evidence strengthens the case for a
continued focus on the crucial 1000 day window during
pregnancy and the first 2 years of life. It also shows the
importance of intervening early in pregnancy and even
before conception. Because many women do not access
nutrition-promoting services until month 5 or 6 of
pregnancy, it is important that women enter pregnancy
in a state of optimum nutrition. The emerging
platforms for adolescent health and nutrition might
oer opportunities for enhanced benefits.7
There is a growing interest in adolescent health as
an entry point to improve the health of women and
children, especially because an estimated 10 million
girls younger than 18 years are married each year.6
 Furthering the evidence to improve maternal
pre-conception period and in adolescents in countries
with a high burden of undernutrition and young age
       have been successfully implemented at scale, and the
sucient number of those in need may be a challenge.
Prevention of maternal deaths
Iron and calcium deficiencies contribute substantially
to maternal deaths. Previously reported analyses,
confirmed by this Series, showed that anaemia is a
risk factor for maternal deaths, probably because of
haemorrhage, the leading cause of maternal deaths
(23% of total deaths). Additionally there is now sound
evidence that calcium deficiency increases the risk of
pre-eclampsia, currently the second leading cause of
maternal death (19% of total deaths). Thus, addressing
deficiencies of these two minerals could result in
substantial reduction of maternal deaths.
www.thelancet.com
Executive Summary
Emerging burden of obesity
Overweight in adults and increasingly in children
constitutes an emerging burden that is quickly
establishing itself globally, aecting both poor and rich
populations. The prevalence of maternal overweight
has increased steadily since 1980, and exceeds that
of maternal underweight in all regions of the world.
Maternal overweight and obesity result in increased
Overweight and obesity prevalence is increasing in
children younger than 5 years globally, especially in
developing countries, and is becoming an increasingly
important contributor to adult obesity, diabetes, and
non-communicable diseases.6 Although the prevalence
of overweight in high-income countries is more than
double that in LMICs, most aected children (76% of the
total number) live in LMICs. The trends in early childhood
overweight are a probably a consequence of changes in
dietary and physical activity patterns over time overlaid on
risks attributable to fetal growth restriction and stunting.
If trends are not reversed, increasing rates of childhood
overweight and obesity will have vast implications, not
only for future health-care expenditures but also for the
overall development of nations. These findings confirm
the need for eective interventions and programmes
      
of excessive weight gain relative to linear growth is
essential.
and child nutrition
Since the 2008 Series, many nutrition interventions
evidence base for eective interventions and delivery
strategies has grown. At the same time, coverage rates
for other interventions are either poor or non-existent.
We modelled ten nutrition-specific interventions
across the lifecycle to address undernutrition and
micronutrient deficiencies in women of reproductive
age, pregnant women, neonates, infants, and children
to assess the eects and cost of scaling up (figure 2).7
     
supplementation, maternal balanced energy protein
supplementation, maternal calcium supplementation,
multiple micronutrient supplementation in pregnancy,
promotion of breastfeeding, appropriate complementary
feeding, vitamin A administration and preventive
5
 Executive Summary
zinc supplementation in children aged 659 months,
management of severe acute malnutrition (SAM), and
management of moderate acute malnutrition.
Continued investment in nutrition-specific inter-
ventions and delivery strategies to reach poor segments
of the population at greatest risk can make a substantial
dierence. If these ten proven nutrition-specific
interventions were scaled-up from existing population
coverage to 90%, an estimated 900 000 lives could be
saved in 34 high nutrition-burden countries (where 90%
of the worlds stunted children live, figure 3) and the
prevalence of stunting could be reduced by 20% and that
of severe wasting by 60%. This would reduce the number
of children with stunted growth and development by
33 million.7 On top of existing trends, this improvement
would comfortably reach the WHA targets for 2025.
Cost of scaling up proven interventions
We estimate that the cost of scaling-up this package
of ten essential nutrition-specific interventions to
Preconception care: family /

90% coverage in 34 countries is Int$96 billion per
year (table 2).7 Of the $96 billion, $37 billion (39%)
is for micronutrient interventions, $09 billion (10%)
for educational interventions, and $26 billion (27%)
for management of SAM. The remaining $23 billion
(24%) accounts for provision of food for pregnant
women and children aged 623 months in poor
households. Since many interventions are being scaled
       
the cost per discounted life-year saved is about
$370 ($213 per undiscounted life-year saved).
More than half the $96 billion is accounted for by
two large countries which will rely heavily on domestic
resources (India and Indonesia). Consumables (drugs,
or other items such as for transport or administration)
account for a little less than half of the $96 billion, and
all but the poorest countries can be expected to cover
most of the expenditures on personnel. Therefore,
$34 billion from external donors could make a
substantial dierence to child nutrition
# 

&($$ !"''#" +
(
 "
! /
+ (&)
%&'
"

planning, delayed age at first /
( '$ 
!%#"('%"' +
"y initiation of breast /

#!$ !"'%,
"
pregnancy, prolonging of &($$ !"''#" feeding /
'!"

&($$ !"''#"
inter-pregnancy interval, /

 (!
&($$ !"''#" +
Vitamin K administration .
!#"'&
abortion care, psychosocial care /
 "
"%,
$%#'" +
 $
&$

/
%)"')
-"
&($$ !"''#" #%!!$$  &($$ !"''#"
+

" 
"
" 
!%#
 $e +
"
 $"
"e +
%$!
" %$"t
+


#%!!$$  #%!!$$ s
+
Tobacco cessation +

" 
#%!!$$ 
%&
!'%"

"
 ##

!#%',
"
!#%' ',
# &"' 
"
$%"", #"'& ""'&
"
 %"
!$%#)
#"'#"

%#*'
"

"(%#) #$!"'
#('#!&
&&
$%)"'#"
" &&
$%)"'#"
"
'%'!"' '%'!"'
+
"
!"&$ 
 "!"'
#
 "%&
*#%

women "!"'
#
 $',
+
$"
' " +
"!%$
)
 "
"
$%#(')',
+
#$(
!"&$  diarrhoea
+
W #"#!

+


""  ) #$!"'
+
"
!"&$ 

in children
+
' "

"
+
#$(
!"&$ n

Delivery platforms: Community delivery platforms, integrated management of childhood illnesses, child health days, school-based
delivery platforms, financial platforms, fortification strategies, nutrition in emergencies

# =Interventions modelled

$s=Other interventions reviewed

Figure 2: Conceptual framework

WRA=women of reproductive age. WASH=water, sanitation, and hygiene. SAM=severe acute malnutrition. MAM=moderate AM.

6 www.thelancet.com
 Executive Summary

Afghanistan
Pakistan
Iraq Bangladesh
Egypt
Chad Nepal Myanmar

Burkina Faso India

Mali Niger Philippines
Sudan Yemen
Guatemala Vietnam
Cte dIvoire Ethiopia
Ghana
Nigeria Kenya Uganda
Cameroon Indonesia
Tanzania Rwanda
DR Congo
Angola Madagascar
Zambia Malawi
Mozambique
South Africa
High burden countries
Other countries

Figure 3: Countries with the highest burden of malnutrition

These 34 countries account for 90% of the global burden of malnutrition.

The promise of emerging interventions and delivery
strategies and platforms
Delivery strategies are crucial to achieving coverage
with nutrition-specific interventions and reaching
populations in need. A range of channels can provide
opportunities for scaling up and reaching large
population segments, such as fortification of staple
foods and conditional and unconditional cash transfers.7
Community delivery platforms for nutrition education
and promotion, integrated management of childhood
illness, school-based delivery platforms, and child health
days are other possible channels.
Innovative delivery strategiesespecially community-
based delivery platformsare promising for scaling
up coverage of nutrition interventions and have the
potential to reach poor and dicult to access populations
through communication and outreach strategies.7 These
could also lead to potential integration of nutrition with
maternal, newborn, and child health interventions,
helping to achieve reductions in inequities.
Unlocking the potential of nutrition-sensitive
programmes
In addition to nutrition-specific interventions,
acceleration of progress in nutrition will also require
increases in the nutritional outcomes of eective, large-
scale, nutrition-sensitive development programmes.8
Nutrition-sensitive programmes address key underlying
determinants of nutritionsuch as poverty, food
insecurity, and scarcity of access to adequate care
resourcesand include nutrition goals and actions. They
can therefore help enhance the eectiveness, coverage,
and scale of nutrition-specific interventions.
Our review of potentially nutrition-sensitive
programmes in agriculture, social safety nets, early child
Number of lives Cost per life-year
saved* saved
Optimum maternal nutrition during pregnancy
Maternal multiple micronutrient supplements to all 102 000 $571 (3981191)
Calcium supplementation to mothers at risk of low intake (49 000146 000)
Maternal balanced energy protein supplements as needed
Universal salt iodisation
Infant and young child feeding
Promotion of early and exclusive breastfeeding for 6 months and 221 000 $175 (132286)
continued breastfeeding for up to 24 months (135 000293 000)
Appropriate complementary feeding education in food secure
populations and additional complementary food supplements in
food insecure populations
Micronutrient supplementation in children at risk
Vitamin A supplementation between 6 and 59 months age 145 000 $159 (106766)
Preventive zinc supplements between 12 and 59 months of age (30 000216 000)
Management of acute malnutrition
Management of moderate acute malnutrition 435 000 $125 (119152)
Management of severe acute malnutrition (285 000482 000)
Data are number (95% CI) or cost in 2010 international dollars (95% CI). 
packages are scaled up at once. Cost per life-year saved assumes that a life saved of a child younger than 5 years saves on
average 59 life-years, based on WHO data (2011188) that life expectancy at birth on average in low-income countries is 60,
and that most deaths of children younger than 5 years occur in the first year of life. To convert to cost per discounted life-
year saved multiply these estimates by 59/32 (ie, 184). Intervention has eect on maternal or child morbidity, but no
direct eect on lives saved. Cost per life-year saved by management of severe acute malnutrition only, costs for
supplementary feeding for moderate acute malnutrition are currently unavailable.
Table 2: Eect of packages of nutrition interventions at 90% coverage

www.thelancet.com 7
 Executive Summary
development, and schooling confirms that programmes
in these sectors are successful at addressing several of
the underlying determinants of nutrition, but evidence
of their nutritional eect is still scarce.
Targeted agricultural programmes have an important
role in support of livelihoods, food security, diet
quality, and womens empowerment, and complement
global eorts to stimulate agricultural productivity
and thus increase producer incomes while protecting
consumers from high food prices.8    implementation also contribute to the limited evidence
on nutrition outcomes, however, is inconclusive, with
Key messages on nutrition-specific interventions
 A clear need exists to introduce promising evidence-
based interventions in the preconception period and in
adolescents in countries with a high burden of
 in measurement and and reporting of outcomes.
however, targeting and reaching a sucient number of
those in need will be challenging.
 Promising interventions exist to improve maternal
nutrition and reduce intrauterine growth restriction and
small-for-gestational-age (SGA) births in appropriate
settings in developing countries, if scaled up before and
during pregnancy. These interventions include balanced
energy protein, calcium, and multiple micronutrient
supplementation and preventive strategies for malaria in
pregnancy.
 Replacement of iron-folate with multiple micronutrient
supplements in pregnancy might have additional benefits
for reduction of SGA in at-risk populations, although
further evidence from eectiveness assessments might be
needed to guide a universal policy change.
 Strategies to promote breastfeeding in community and
facility settings have shown promising benefits on
 
evidence for long-term benefits on nutritional and
developmental outcomes is scarce.
  breastfeeding promotion and promotion of
strategies is insucient, with much the same benefits
noted from dietary diversification and education and
food supplementation in food secure populations and
slightly greater eects in food insecure populations.
Further eectiveness trials are needed in food insecure
populations with standardised foods (pre-fortified or
non-fortified) to assess duration of intervention,
outcome definition, and cost eectiveness.
 Treatment strategies for severe acute malnutrition with
recommended packages of care and ready-to-use
therapeutic foods are well established, but further
evidence is needed for prevention and management
strategies for moderate acute malnutrition in population
settings, especially in infants younger than 6 months.
8 www.thelancet.com
the exception of eects on vitamin A intake and status
from homestead food production programmes and
distribution of biofortified vitamin A-rich orange sweet
     
programmes are more successful when they incorporate
strong behaviour change communications strategies
and a gender-equity focus. Although firm conclusions
have been hindered by a dearth of rigorous programme
evaluations, weaknesses in programme design and
of nutritional outcomes so far.
 Data for the eect of various nutritional interventions

studies should focus on these aspects with consistency
 Conditional cash transfers and related safety nets can
address the removal of financial barriers and promotion
of access of families to health care and appropriate
foods and nutritional commodities. Assessments of the
feasibility and eects of such approaches are urgently
needed to address maternal and child nutrition in well
supported health systems.
 Innovative delivery strategies, especially
community-based delivery platforms, are promising for
scaling up coverage of nutrition interventions and have
the potential to reach poor populations through demand
creation and household service delivery.
 Nearly 15% of deaths of children younger than 5 years
can be reduced (ie, 1 million lives saved), if the ten core
nutrition interventions we identified are scaled up.
 The maximum eect on lives saved is noted with
management of acute malnutrition (435 000
(135 000293 000) lives would be saved with delivery of
an infant and young child nutrition package, including

could save 145 000 (30 000216 000) lives.
 These interventions, if scaled up to 90% coverage,
could reduce stunting by 203% (33.5 million fewer
stunted children) and can reduce prevalence of severe
wasting by 614%.
 The additional cost of achieving 90% coverage of these
proposed interventions would be US$96 billion
per year.

Social safety nets provide cash and food transfers
to a billion poor people and reduce poverty. They also
have an important role in mitigation of the negative
eects of global changes, conflicts, and shocks by
protecting income, food security, and diet quality. When
targeted to women, they enhance several aspects of
womens empowerment. Pooled evidence, however,
shows limited eects of these programmes on child
nutrition, although some individual studies showed
eects in younger and poorer children exposed for
longer durations.8 Absence of clarity in nutrition goals,
weaknesses in design, and poor quality services probably
account for the limited nutritional eects.
Child stunting and impaired cognitive development
share many of the same risk factors including nutritional
deficiencies, intra-uterine growth restriction, and social
and economic conditions, such as poverty and maternal
depression.6 Linear growth and cognitive development
also share the same period of peak vulnerability
the first 1000 days of life. Combination of early child
development and nutrition interventions therefore
makes sense biologically and programmatically, and
evidence from mostly small-scale programmes suggests
additive or synergistic eects on child development and
in some cases on nutrition outcomes.8
Interventions to improve maternal mental health also
have high potential for nutritional eects and should
be incorporated in nutrition-sensitive programmes.8
Maternal depression is an important determinant of
suboptimum caregiving and health-seeking behaviours
and is associated with poor nutrition and child
development outcomes.
Parental schooling is consistently associated with
improved nutrition outcomes and schools provide an
opportunity, so far largely untapped, to include nutrition
in school curricula for prevention and treatment
of undernutrition or obesity.8 Nutrition-sensitive
programmes also oer a unique opportunity to reach girls
in adolescence (preconception) and possibly to achieve
scale either through school-linked programmes with
conditions or home-based programmes.
The potential of nutrition-sensitive programmes to
improve nutrition outcomes is clear, but it has yet to
be unleashed. Importantly, several of the programmes
documented in our analysis8 were not originally
designed with clear nutrition goals and actions from the
outset and were retrofitted to be nutrition-sensitive. The
www.thelancet.com
Executive Summary
Key messages on nutrition-sensitive interventions and programmes
 
child development, and education have enormous potential to enhance the scale and

nutrition-sensitive programmes achieve their own goals.
 
mitigation of potentially negative eects of global changes and man-made and
environmental shocks, in supporting livelihoods, food security, diet quality, and womens
empowerment, and in achieving scale and high coverage of nutritionally at-risk
households and individuals.
 

and actions and rigorous eectiveness assessments are needed.
 

the eectiveness of biofortification continues to grow for other micronutrient and crop
combinations.
 

programme nutrition goals and interventions, and quality of services need to be
strengthened.
 
or synergistic eects on child development, and in some cases, nutrition outcomes.
Integration of stimulation and nutrition interventions makes sense programmatically
and could save cost and enhance benefits for both nutrition and development outcomes.
 
schools provide an opportunity, so far untapped, to include nutrition in school curricula
for prevention and treatment of undernutrition or obesity.
 
health-seeking behaviours and is associated with poor nutrition and child development

nutrition-sensitive programmes.
 
preconception and possibly to achieve scale, either through school-linked conditions and
interventions or home-based programmes.
 

womens physical and mental health, nutrition, time allocation, and empowerment.
nutrition-sensitivity of programmes can be enhanced
       
     
     
optimisation of womens nutrition, time, physical and
mental health, and empowerment.
With guidance on how nutrition-sensitivity can be
enhanced and a new generation of nutrition-sensitive
programmes, stronger evidence should emerge in the
near future. Currently, new agriculture, social safety
net programmes, and joint nutrition and early child
development programme designs, methods, and
packages of interventions are being tested, several of
9
 Executive Summary
which integrate complementary inputs that address other
constraints to optimum nutritionsuch as maternal
depression, or scarcity of access to water, sanitation, and
hygiene servicesand are strengthening links with health
services. Rigorous impact evaluations are underway,
many of which are based on strong programme theory
and impact pathway analysis. They are also addressing
key weaknesses encountered in previous evaluations and
are assessing outcomes on a range of nutrition and child
development outcomes as well as several household and
gender outcomes along the impact pathway. The body
of evidence generated by these enhanced programmes
and evaluations in the next 510 years will be of crucial
importance to inform future investments in nutrition-
sensitive programmes from many sectors.
Building an enabling environment to deliver
nutrition results
The nutrition landscape has shifted fundamentally since
2008. The 2008 Series showed that the stewardship
of the nutrition system was dysfunctional and deeply
Key messages on enabling environments for nutrition
  global agenda. Nearly every major development agency
accelerated with deliberate action
 
human suering should prioritise investment in scale-up of nutrition-specific
interventions, and should maximise the nutrition sensitivity of national
development processes
  more prominent on the agendas of the UN, the G8 and
 G20, and supporting civil society.
governance, and capacity and resources
 
defeating. Political calculations are at the basis of eective coordination between
sectors, national and subnational levels, private sector engagement, resource
mobilisation, and state accountability to its citizens
  breastfeeding in the first 6 months of life can be
squanderedconversion to results needs a dierent set of strategies and skills
 
fundamentally important for creating and sustaining momentum and for conversion
of that momentum into results on the ground.
  translated into outcomes.
national and global support to a long-term process of strengthening systemic and
organisational capacities
 
improvements in nutrition, but eorts to realise this have to date been hindered by a
scarcity of credible evidence and trust. Both these issues need substantial attention if
the positive potential is to be realised
  and other actors to be responsive to challenging
contextual analyses about how to shape and sustain enabling environments, is
essential as the focus shifts toward action
10
fragmented in terms of messaging, priorities, and
funding.5 Much progress has been made since then,
largely driven by the new evidence introduced in the
2008 Series, which identified the first 1000 days of life
as the window for outcomes, pinpointed a package
of highly eective interventions for reduction of
undernutrition, and proposed a group of high-burden
countries as priorities for increased investment.
The launch of the SUN movement in 2010 represented
a major step toward improved stewardship of the global
nutrition architecture.18,19 SUN brings together more
than 100 entities across the organisational spectrum
of the nutrition community. Up to now, more than 30
countries (representing 35% of the global child stunting
burden) have joined SUN, committing to scaling-up direct
nutrition interventions and advancing nutrition-sensitive
development. Although it is too soon to evaluate SUNs
eect on rates of reduction of undernutrition, it is clear
that through SUN, many countries have made advances
in building multistakeholder platforms across sectors,
aligning nutrition-relevant programmes within a common
results framework, and mobilising national resources.
Additionally, nutrition has been greatly elevated on the
has published a policy document on undernutrition, and
donors have increased ocial development assistance
to basic nutrition by more than 60% between 2008 and
2011, in a very dicult fiscal climate. Nutrition is now
Nowadays, the impetus for improving nutrition is
even stronger than it was 5 years ago. The WHA targets
for reducing stunting, wasting, low birthweight,
anaemia, and overweight, and increasing exclusive
achieved by 2025 with sucient support.21 Central to
this scaled-up support is the creation of an enabling
environment to build commitment and ensure that it is
Improvement of data, research, and accountability for
results
The availability of timely and credible nutrition data,
presented in accessible ways, can help governments
circumstances, and help civil society organisations
to hold them accountable for the eectiveness of
www.thelancet.com

their interventions.9 Advances in health management
information systems and the growing availability
of newer technologies can help with the real-time
monitoring of nutrition outcomes and programme
coverage and quality, and should be researched.
Additionally, although much progress has been made
to work out the costs of addressing undernutrition,
continued work to contextualise and specify these costs
for dierent countries is essential, along with stronger
designation of donor and government spending to
improve tracking of investments and results in nutrition.
Improved data for micronutrient deficiencies and
other nutritional conditions are needed at national and
subnational levels. This improvement should involve
the development and use of improved biomarkers
that could be used to describe nutritional conditions
and increase knowledge of how they aect health and
development. Such information is needed to guide
intervention programmes in countries and priorities for
support globally.
Although substantial progress has been made
to establish the needs around nutrition, no
systematic process exists for bringing together the
implementation-related evidence for how to scale
up the vast array of nutrition-specific and nutrition-
sensitive interventions with quality and equity (so-
called implementation science). This evidence is
essential to ensure that future investments are directed
toward proven pathways to outcomes.
Beyond this evidence, service providers, governments,
donors, and the private sector need strong national
monitoring and assessment platforms to hold them
accountable for the quality and eectiveness of
their investments in nutrition.9 Boosting nutrition
commitment and accountability can be achieved
through assessing and implementing innovative new
instruments and mechanisms, including computer-
based monitoring systems, commitment indices, and
social accountability mechanisms.
Engagement and regulation of the private sector
The scale, know-how, reach, financial resources, and
existing involvement of the private sector in actions that
aect nutrition status is well known.9 Yet there are still
too few independent and rigorous assessments of the
eectiveness of involvement of the commercial sector
in nutrition. Distrust of the private sectorespecially
www.thelancet.com
Executive Summary
the food industryremains high and is linked, partly,
to the decades-long tussle related to the marketing
of breastmilk substitutes in developing countries and
around continued marketing of sugar-sweetened
beverages and fast foods worldwide.
This troubled history has made it more dicult for the
private sector to be a major contributor to the collective
creation and sustenance of momentum for reduction
of malnutrition. In view of the needs and substantial
resources, influence, and convening power of the
private sector, it might represent a missed opportunity.
Opportunities exist for collaboration around advocacy,
monitoring, value chains, technical and scientific
collaboration, and staple-food fortification that are
uncontentious and deserve further exploration. Know-
ledge in this area must be expanded rapidly to guide the
private sector toward more positive eects for nutrition.
Regulatory and fiscal eorts are essential when the
private sector is involved in marketing of products that
are detrimental to optimum nutrition. The experience
gained with the International Code of Marketing
of Breastmilk Substitutes should be applied to the
promotion of other harmful, widely-consumed food
products that are being marketed for young children.
Mobilisation of resources
High-burden countries, together with donors,
multilaterals, and the private sector, have a responsibility
to increase allocations to nutrition-specific and
nutrition-sensitive programmes. Meeting the estimated
$96 billion financing gap will require an increase in
donor spending, alongside an equal or greater increase
of spending by LMICs and the establishment of nutrition
budget lines in all high-burden countries.7 To achieve this
aim will be politically challenging, hence the need to build
leadership, commitment, and accountability at national
and international levels.9 However, the financing gap is
unlikely to be closed by these sources alone. Innovation
is needed across all sectors to leverage private-sector and
public-sector resources and generate additional funding.
The nutrition sector can draw on several innovative ideas
from other sectors, including advance market contracts
to promote investment, market levies, and taxes in the
eort. Additional resources must be directed not only to
interventions, but also to the creation of environments
to enable advancement of nutrition, including capacity
and leadership at all levels of government.9 A political
11
 Executive Summary

economy approach to prioritisation of such investments 8 Ruel MT, Alderman H, and the Maternal and Child Nutrition Study Group.

is crucial if sustainable, supportive environments for accelerate progress in improving maternal and child nutrition. Lancet
long-term nutrition agendas are to be created.  
6736(13)60843-0.
Nutrition is crucial to both individual and national 9 Gillespie S, Haddad L, Mannar V, Menon P, Nisbettt N, and the Maternal


development. The evidence in this Series furthers the building commitment and accelerating progress. Lancet
evidence base that good nutrition is a fundamental  
 

driver of a wide range of development goals. The building momentum for impact. Lancet 
post-2015 sustainable development agenda must put 
 
addressing all forms of malnutrition at the top of its Lancet 
goals. 6736(13)61100-9.
   
Now is our crucial window of opportunity to scale- Lancet 
6736(13)61054-5.
up nutrition.22 National and international momentum
13 Taylor A, Dangour AD, Reddy KS. Only collective action will end
to address human nutrition and related food security undernutrition. Lancet 
org/10.1016/S0140-6736(13)61084-3.
and health needs has never been higher. We must work 14 Nabarro D. Global child and maternal nutritionthe SUN rises. Lancet
together to seize this opportunity.  
6736(13)61086-7.
References
 
 
action. Lancet 


S0140-6736(13)61053-3.
regional exposures and health consequences. Lancet371: 24360.
16 Scaling Up Nutrition. Progress report from countries and their partners in
 

Undernutrition Study Group. What works? Interventions for maternal and
child undernutrition and survival. Lancet371: 41740.   
Improving nutrition through multisectoral appraoches. Washington, DC,
 
 
International Bank for Reconstruction and Development, International
consequences for adult health and human capital. Lancet371: 340-57.
Development Association of The World Bank, 2013.
4 Bryce J, Coitinho D, Darnton-Hill I, et al, for the Maternal and Child
 


Announcements/Scaling_Up_Nutrition-A_Framework_for_Actionpdf
action at national level. Lancet371: 51026.
(accessed April 2, 2013).
5 Morris SS, Cogill B, Uauy R, for the Maternal and Child Undernutrition
 Food


Nutr Bull31: 17886.
has it proven so dicult and what can be done to accelerate progress?
Lancet371: 60821. 20 Horton S, Steckel R. Global economic losses attributable to malnutrition

 


Group. Maternal and child undernutrition and overweight in low-income
2013.
and middle-income countries. Lancet 
dx.doi.org/10.1016/S0140-6736(13)60937-X. 21 WHO. Proposed global targets for maternal, infant and young child


7 Bhutta ZA, Das JK, Rizvi A, et al, The Lancet Nutrition Interventions Review
2012.



 22 Dube L, Pingali P, Webb P. Paths of convergence for agriculture, health, and
be done and at what cost? Lancet  wealth. Proc Natl Acad Sci USA 109: 12294301.
doi.org/10.1016/S0140-6736(13)60996-4.

Acknowledgments
Maternal and Child Nutrition Study Group:  Funding: Funding for the preparation of the Series was provided to the Johns
School of Public Health, USA), Harold Alderman (International Food Policy Research Hopkins Bloomberg School of Public Health through a grant from the Bill &
Institute, USA), Zulfiqar A Bhutta (Aga Khan University, Pakistan), Stuart Gillespie Melinda Gates Foundation. The sponsor had no role in analysis or interpretation of
(International Food Policy Research Institute, USA), Lawrence Haddad (Institute of the evidence.
Development Studies, UK), Susan Horton (University of Waterloo, Canada),
 

Anna Lartey (University of Ghana, Ghana), Venkatesh Mannar (The Micronutrient
Initiative, Canada), Marie Ruel (International Food Policy Research Institute, USA),
Second printing with corrections.
Cesar Victora (Universidade Federal de Pelotas, Brazil), Susan Walker (The University
of the West Indies, Jamaica), Patrick Webb (Tufts University, USA)

12 www.thelancet.com
 Series

Maternal and Child Undernutrition 1

Maternal and child undernutrition: global and regional
exposures and health consequences
Robert E Black, Lindsay H Allen, Zulfiqar A Bhutta, Laura E Caulfield, Mercedes de Onis, Majid Ezzati, Colin Mathers, Juan Rivera, for the Maternal
and Child Undernutrition Study Group*

Maternal and child undernutrition is highly prevalent in low-income and middle-income countries, resulting in Lancet 2008; 371: 24360
substantial increases in mortality and overall disease burden. In this paper, we present new analyses to estimate the Published Online
eects of the risks related to measures of undernutrition, as well as to suboptimum breastfeeding practices on January 17, 2008
DOI:10.1016/S0140-
mortality and disease. We estimated that stunting, severe wasting, and intrauterine growth restriction together were
6736(07)61690-0
responsible for 22 million deaths and 21% of disability-adjusted life-years (DALYs) for children younger than 5 years.
See Comment pages 179, 180,
Deficiencies of vitamin A and zinc were estimated to be responsible for 06 million and 04 million deaths, and 181
respectively, and a combined 9% of global childhood DALYs. Iron and iodine deficiencies resulted in few child deaths, See Perspectives page 197
and combined were responsible for about 02% of global childhood DALYs. Iron deficiency as a risk factor for maternal This is the first in a Series of five
mortality added 115 000 deaths and 04% of global total DALYs. Suboptimum breastfeeding was estimated to be papers about maternal and child
responsible for 14 million child deaths and 44 million DALYs (10% of DALYs in children younger than 5 years). In an undernutrition

analysis that accounted for co-exposure of these nutrition-related factors, they were together responsible for about 35% *Members listed at end of paper

of child deaths and 11% of the total global disease burden. The high mortality and disease burden resulting from Johns Hopkins Bloomberg
School of Public Health,
these nutrition-related factors make a compelling case for the urgent implementation of interventions to reduce their Baltimore, MD, USA
occurrence or ameliorate their consequences. (Prof R E Black MD,
Prof L E Caulfield PhD); USDA,
Introduction Undernutrition encompasses stunting, wasting, and ARS Western Human Nutrition
Research Center, Davis, CA, USA
Maternal and child undernutrition remain pervasive and deficiencies of essential vitamins and minerals (col- (Prof L H Allen PhD);
damaging conditions in low-income and middle-income lectively referred to as micronutrients) as one form of the Aga Khan University, Karachi,
countries. A framework developed by UNICEF recognises condition known as malnutrition, with obesity or Pakistan (Prof Z A Bhutta, MD);
the basic and underlying causes of undernutrition, over-consumption of specific nutrients as another form. World Health Organization,
Geneva, Switzerland
including the environmental, economic, and The term hunger, which literally describes a feeling of (M de Onis MD, C Mathers PhD);
sociopolitical contextual factors, with poverty having a discomfort from not eating, has also been used to describe Harvard School of Public
central role (figure 1). Although addressing general Health, Boston, MA, USA
deprivation and inequity would result in substantial (M Ezzati PhD); and Mexico
Key messages National Institute of Public
reductions in undernutrition1 and should be a global Health, Cuernavaca, Mexico
priority, major reductions in undernutrition can also be Maternal and child undernutrition is the underlying cause (Prof J Rivera PhD)
made through programmatic health and nutrition of 35 million deaths, 35% of the disease burden in Correspondence to:
interventions. This paper is the first in a Series of five children younger than 5 years and 11% of total global Robert Black, Johns Hopkins
papers that focus on the disease burden attributable to Bloomberg School of Public
DALYs Health, Baltimore, MD, USA
undernutrition and the interventions aecting household The number of global deaths and DALYs in children less rblack@jhsph.edu
food availability and use, maternal and child care, and than 5 years old attributed to stunting, severe wasting,
control of infectious diseases. The first two papers and intrauterine growth restriction constitutes the largest
quantify the prevalence of maternal and child percentage of any risk factor in this age group
undernutrition and consider the short-term consequences Vitamin A and zinc deficiencies have by far the largest
in terms of deaths and disease burden, as measured by remaining disease burden among the micronutrients
disability-adjusted life-years (DALYs). They also discuss considered
the long-term educational and economic eects and Iodine and iron deficiencies have small disease burdens,
associations with adult chronic diseases, particularly as partly because of intervention programmes, but sustained
countries go through the demographic, epidemiological, eort is needed to further reduce their burden
and nutritional transitions.26 The third paper estimates Suboptimum breastfeeding, especially non-exclusive
the potential benefits of implementing health and breastfeeding in the first 6 months of life, results in
nutrition interventions that current evidence indicates 14 million deaths and 10% of disease burden in children
are eective and applicable in low-income and younger than 5 years
middle-income countries. The final two papers consider Maternal short stature and iron deficiency anaemia
the current state of such interventions and how they increase the risk of death of the mother at delivery,
could be implemented fully through actions at national accounting for at least 20% of maternal mortality
and global levels.

www.thelancet.com Vol 371 January 19, 2008 243

 Series

undernutrition, especially in reference to food insecurity, even though there are no MDG indicators for these
wherein people do not have physical and economic conditions. This Series also examines the consequences of
access to sucient, safe, nutritious, and culturally low body-mass index and short stature in women.
acceptable food to meet their dietary needs.7,8 This paper consists mainly of new analyses of the
Undernutrition is an important determinant of maternal prevalence of nutritional conditions, risk factors, and
and child health.912 consequent disease burden; if this was not possible or
The Millennium Development Goals (MDGs) state as necessary, previously published results are presented.
the first goal to halve between 1990 and 2015 the proportion The burden of disease attributable to maternal and child
of people who suer from hunger.7 One indicator to undernutrition is presented for three world regions
monitor progress for this target is the proportion of Africa, Asia, and Latin Americathat include primarily
children who are underweightie, low weight compared low-income and middle-income countries; only 1% of
with that expected for a well-nourished child of that age deaths in children younger than 5 years occur outside
See Online for webtable 1 and sex. This anthropometric indicator can indicate these regions. UN regions and subregions (webtable 1)
wasting (ie, low weight-for-height, indicating acute weight were used.
loss), or much more commonly, stunting (ie, low
height-for-age, indicating chronic restriction of a childs Prevalence and consequences
potential growth). Those two conditions can have dierent Maternal short stature and low body-mass index in
determinants and respond to dierent interventions.13 pregnancy and lactation
Therefore, consideration of wasting and stunting is more Maternal undernutrition, including chronic energy and
useful than consideration of underweight. This series micronutrient deficiencies, is prevalent in many regions,
primarily uses these two indicators, but also presents especially in south-central Asia, where in some countries
information on underweight because weight-for-age has more than 10% of women aged 1549 years are shorter
See Online for webtable 2 been used in previous analyses.911 Reduction of fetal than 145 cm (webtable 2). Maternal undernutritionie,
growth restriction and micronutrient deficiencies is also body-mass index of less than 185 kg/mranges
essential to achieving the MDGs and deserves high priority, from 10% to 19% in most countries. A serious problem of
maternal undernutrition is evident in most countries in
Short-term consequences: Long-term consequences:
sub-Saharan Africa, south-central and southeastern Asia,
Mortality, morbidity, disability Adult size, intellectual ability,
and in Yemen, where more than 20% of women have a
economic productivity, body-mass index of less than 185 kg/m. With a
reproductive performance, prevalence of low body-mass index around 40% in
metabolic and cardiovascular disease
women, the situation can be considered critical in India,
Maternal and child Bangladesh, and Eritrea. Maternal short stature and low
undernutrition body-mass index have independent adverse eects on
pregnancy outcomes.
Immediate
The nutritional status of a woman before and during
Inadequate dietary intake Disease
causes pregnancy is important for a healthy pregnancy
outcome.14,15 Maternal short stature is a risk factor for
caesarean delivery, largely related to cephalopelvic
disproportion. A meta-analysis of epidemiological studies
Unhealthy household
Household food
Inadequate care environment and lack
found a 60% (95% CI 5070) increased need for assisted
insecurity
of health services delivery among women in the lowest quartile of stature
(146 cm to 157 cm, depending on the region) compared
with women in the highest quartile.16 If operative delivery
to ensure a healthy birth is not available to women who
Income poverty: need it, both mother and baby are at risk.17 Even if
Underlying employment, self-employment,
causes dwelling, assets, remittances, operative delivery is accessible, aordable, and safe,
pensions, transfers etc anaesthesia and laparotomy increase the risk of maternal
morbidity.18 Low maternal body-mass index does not
seem to increase the risk of pregnancy complications and
Lack of capital: financial, human, assisted delivery.16 Rather, there seems to be a synergistic
physical, social, and natural
positive eect of short stature and higher maternal
Basic
causes body-mass index on increasing these complications.19,20
Social, economic, Low maternal body-mass index is associated with
and political context intrauterine growth restriction.11 Previous analyses
estimated the disease burden of low maternal body-mass
Figure 1: Framework of the relations between poverty, food insecurity, and other underlying and immediate index as a risk factor for perinatal conditions,11 whereas
causes to maternal and child undernutrition and its short-term and long-term consequences the estimates presented in this paper consider intrauterine

244 www.thelancet.com Vol 371 January 19, 2008

 Series

Children Percentage Number stunted in Percentage Number severely Percentage Number

<5 years in stunted (95% CI) millions (95% CI) severely wasted wasted in millions underweight underweight in
millions27 (95% CI) (95% CI) (95% CI) millions (95% CI)
Africa 141914 401 (368434) 569 (522616) 39 (2257) 56 (3080) 219 (198240) 311 (281340)
Eastern 48807 500 (423579) 244 (207283) 36 (1584) 18 (0741) 280 (236329) 137 (115161)
Middle 20197 415 (383448) 84 (7791) 50 (20120) 10 (0,424) 225 (192261) 45 (3953)
Northern 22171 245 (173339) 54 (3875) 33 (1289) 07 (0320) 68 (28153) 15 (0634)
Southern 6075 302 (254356) 18 (1522) 27 (1068) 02 (0604) 114 (80157) 07 (0510)
Western 44663 377 (335421) 168 (150188) 43 (1896) 19 (0843) 239 (210269) 107 (94120)
Asia 356879 313 (275351) 1116 (9811251) 37 (1262) 133 (44223) 220 (185256) 786 (659913)
Eastern 95070 145 (135155) 137 (128147) 07 (0316) 07 (0316) 51 (4854) 48 (4551)
South-central 181481 407 (342477) 738 (620865) 57 (24128) 103 (44233) 331 (266403) 601 (483730)
Southeastern 54970 343 (265435) 189 (145239) 36 (1488) 20 (0849) 207 (172246) 114 (95135)
Western 25358 206 (100388) 52 (2598) 16 (0458) 04 (0115) 89 (28242) 22 (0761)
Latin America 56936 161 (94228) 92 (53130) 06 (0210) 03 (0106) 48 (3164) 27 (1837)
Caribbean 3657 82 (39167) 03 (0106) 10 (0425) 003 (00109) 51 (2796) 02 (0104)
Central 16161 231 (139364) 37 (2259) 06 (0217) 01 (00403) 62 (34110) 10 (0518)
America
South America 37118 138 (69263) 51 (2698) 06 (0216) 02 (00706) 41 (2567) 15 (0925)
All developing 555729 320 (293346) 1777 (16291925) 35 (1851) 193 (100286) 202 (179226) 1124 (9931255)
countries

Stunting=height-for-age less than 2 SD. Severe wasting=weight-for-length or weight-for-height less than 3 SD. Underweight=weight-for-age less than 2 SD.

Table 1: Childhood stunting, severe wasting, and underweight estimates and numbers aected in 2005 based on the WHO Child Growth Standards by
UN regions and subregions

growth restriction to be the risk factor for neonatal

conditions. Additional work is needed to quantify the
relative eects of low maternal body-mass index, extent
of weight gain in pregnancy, and maternal micronutrient
deficiencies on the occurrence and severity of intrauterine
growth restriction.
Maternal undernutrition has little eect on the
volume or composition of breast milk unless
malnutrition is severe. The concentration of some
micronutrients (vitamin A, iodine, thiamin, riboflavin,
pyridoxine, and cobalamin) in breast milk is dependent No data
<20%
on maternal status and intake, so the risk of infant
20299%
depletion is increased by maternal deficiency.21 This 30399%
factor is most evident in the case of vitamin A, where 40%
the content in breast milk is the main determinant of
infant status because stores are low at birth. Maternal
supplementation with these micronutrients increases
the amount secreted in breast milk, which can improve Figure 2: Prevalence of stunting in children under 5 years
infant status.
For all developing countries, an estimated 32%
Childhood underweight, stunting, and wasting (178 million) of children younger than 5 years had a
The prevalences of underweight, stunting, and wasting height-for-age Z score of less than 2 in 2005 (table 1).22,23
worldwide and for UN regions are based on analysis Eastern and middle Africa have the highest prevalence
of 388 national surveys from 139 countries, applying estimates in UN subregions with 50% and 42%,
comparable methods, including use of the new WHO respectively; the largest number of children aected by
Child Growth Standards.2226 In 2005, 20% of children stunting, 74 million, live in south-central Asia.
younger than 5 years in low-income and middle-income Of the 40 countries with a child stunting prevalence
countries had a weight-for-age Z score of less than 2 of 40% or more, 23 are in Africa, 16 in Asia, and one in
(table 1). The prevalences were highest in south-central Latin America; and of the 52 countries with prevalence of
Asia and eastern Africa where 33% and 28%, respectively, less than 20%, 17 are in Latin America and the Caribbean,
were underweight. 16 in Asia, 11 in Europe, and four each in Africa and

www.thelancet.com Vol 371 January 19, 2008 245

 Series

Panel 1: Disparities in stunting and dietary diversity

This Series of papers on maternal and child nutrition builds
on a framework that recognises the underlying causes of
undernutrition, including the environmental, economic, and
political contextual factors, in particular poverty. Figure 4A
shows the disparities in stunting prevalence between the top
and bottom wealth quintiles of the population, based on data
from Demographic and Health Surveys for 11 of the 36 focus
No data countries. In most countries the poor children have about
2029%
3039%
twice as much stunting as the wealthier children.
4049% The three proximate determinants of child nutritional status
5059% include food security, adequate care, and health. Each of
60%
these is strongly aected by poverty. For example,
animal-source foods are an important component of child
Figure 3: Prevalence of stunting among children under 5 years old in India by
state diets, as a major source of protein and micronutrients; low
intake of these foods is a risk factor for stunting. Figure 4B
Oceania (figure 2). Including only countries with a shows the gap between the proportions of children in the top
stunting prevalence of 20% or more, 36 countries account and bottom wealth quintiles who received these foods in the
See Online for webtable 3 for 90% of all stunted children worldwide (webtable 3). 24 h before the survey, in selected countries with available
These countries will be the focus of estimations of the data. Wide gaps are evident in most countries.
eects of interventions on disease burden.28 21 of these Like stunting, micronutrient deficiencies are also linked to
countries are in Africa and, although fewer countries are poverty. In India, for example, anaemia aects 79% of
in Asia (13), they account for 61% of the total stunted children in the lowest wealth quintile, compared with the still
children because of their large populations. India, with high prevalence of 64% in the top quintile.29
an estimated prevalence of 51%, has more than 61 million
stunted children, 34% of the global total; however, the A Yemen
prevalence varies substantially by state within India Tanzania
(figure 3). Within countries the prevalence of stunting is Uganda
generally highest for the poorest segments of the Pakistan
population (panel 1, figure 4).29 Nigeria
The global estimate of wasting (weight-for-height Nepal
Z score of less than 2) is 10% (55 million children). Madagascar
South-central Asia is estimated to have the highest Kenya
prevalence (16%) and numbers aected (29 million). The India
same regional pattern is seen for severe wasting Egypt
(weight-for-height Z score of less than 3), often used as Bangladesh
a criterion for therapeutic feeding interventions, with a
0% 20% 40% 60%
prevalence of 35% or 19 million children (table 1). The
highest percentages of children with severe wasting are B Uganda
seen in south-central Asia and middle Africa. Of the Tanzania
36 countries with 90% of stunted children, the prevalence South Africa
Phillipines
See Online for webtable 4 of severe wasting varies from 01% to 12% (webtable 4).
Nigeria
For the disease burden estimations, it is important to Nepal
know the extent of overlap of the populations of children Madagascar
younger than 5 years who are stunted (height-for-age Kenya
Z score <2) and those who have severe wasting. To Indonesia
examine this question, data from 19 Demographic and Ethiopia
Egypt
Health Surveys were analysed.30 The surveys were done
Bangladesh
between 1998 and 2005, and included two from south Asia,
0% 20% 40% 60% 80% 100%
ten from Africa, and seven from Latin America and the
Caribbean. The prevalence of severe wasting was higher at Figure 4: Stunting and dietary diversity
younger ages and declined by 24 months. Conversely, (A) Children aged 059 months who were stunted. Bars show the gap
stunting prevalence increased progressively until reaching between prevalence in the poorest (red) and least poor (blue) wealth
quintiles. (B) Children aged 1223 months who ate meat, fish, poultry, or
a plateau around 24 months. In these countries, severe eggs in the 24 h before the survey. Bars show the gap between intake in
wasting was not accompanied by stunting in 80100% of the poorest (red) and least poor (blue) wealth quintiles.
younger children and 4050% of older children. Thus,

246 www.thelancet.com Vol 371 January 19, 2008

 Series

similar stunting prevalence can have a several-fold

<3 3 to <2 2 to <1 More
(95% CI) (95% CI) (95% CI) than 1 dierence in the prevalence of severe wasting.31
Eight data sets from low-income countries (Ghana,
Weight-for-age (Z score)
Guinea Bissau, Senegal, the Philippines, Nepal, Pakistan,
Overall* 97 (52179) 25 (1836) 18 (1227) 10
India, and Bangladesh)3238 were used for analysis of
Diarrhoea* 95 (55165) 34 (2744) 21 (1627) 10
disease risks associated with childhood undernutrition.
Pneumonia* 64 (39104) 13 (0920) 12 (0719) 10
These data were from specific study populations, but
Malaria 16 (1027) 12 (0535) 08 (0232) 10
were broadly representative of these countries and
Measles 64 (4691) 23 (1732) 13 (1115) 10
low-income countries in general. Children were grouped
Height-for-age (Z score)
by Z score as less than 3, 3 to less than 2, 2 to less
Overall* 41 (2664) 16 (1322) 12 (0915) 10
than 1, and 1 or more for each of these indices.
Diarrhoea* 46 (2781) 16 (1125) 12 (0917) 10 Generalised linear mixed models were used to estimate
Pneumonia* 32 (1567) 13 (0921) 1 (0616) 10 the risk of all-cause mortality, as well as of death due to
Malaria 21 (0949) 10 (0424) 07 (0509) 10 diarrhoea, pneumonia, malaria, and measles for each of
Measles 28 (1458) 17 (0836) 07 (0509) 10 the groups with Z scores less than 1. The estimated
Weight-for-height (Z score) odds ratios were then adjusted for confounding due to
Overall* 94 (53168) 30 (2045) 15 (1219) 10 socioeconomic factors that aect mortality through other
Diarrhoea* 63 (27147) 29 (1845) 12 (0719) 10 pathways, such as non-nutritional determinants of
Pneumonia* 87 (48156) 42 (3255) 16 (1124) 10 infection or access to better clinical care. The adjustment
Malaria 23 (1632) 30 (1089) 09 (0326) 10 used data sets with robust measurement of socioeconomic
Measles 60 (4382) 37 (2555) 18 (0936) 10 status from Nepal36,37 and Honduras39 and showed that
odds ratios were attenuated by 10% and 20%, respectively.
*Ghana, Senegal, Guinea Bissau, the Philippines, India, Nepal, Bangladesh,
Pakistan. Ghana, Senegal, and Guinea Bissau. Nepal, Ghana, Senegal, Guinea On the basis of these results, 15% attenuation was applied
Bissau, and the Philippines. to the odds ratios calculated with the generalised linear
mixed models for underweight, stunting, and wasting.
Table 2: Odds ratio for mortality by weight-for-age, height-for-age and
weight-for-height by cause of death The odds ratios derived for cause-specific mortality were
also used for diarrhoea, pneumonia, and malaria
morbidity.
identifying children who are stunted will not capture most The risk of death increases with descending Z scores
of those with severe wasting. Furthermore, stunting and for underweight, stunting, or wasting (table 2). The
severe wasting are not necessarily associated on a increased risks are all significant for the below 3 category,
geographical or ecological basisie, countries with a as are many in the other two categories.

Low IUGRLBW IUGRLBW IUGRLBW Livebirths Number Number

birthweight (estimated %)40 (% 20002499 g) (% 15001999 g) (thousands)45 20002499 g 15001999 g
(% <2500 g)45 (thousands) (thousands)
Africa 143 889 785 104 30 305 26669 280
Eastern 135 827 730 097 10 649 8719 85
Middle 123 724 640 084 4413 3170 27
Northern 153 98 866 114 4587 4445 51
Southern 146 921 814 107 1243 1132 12
Western 154 989 874 115 9412 9200 107
Asia 183 1239 1094 145 77 490 94638 1388
Eastern 59 179 158 021 20 537 3661 08
South-central 271 1987 1755 232 39 937 77498 1840
Southeastern 116 665 587 078 11 743 7745 61
Western 154 989 874 115 5273 5154 60
Latin America 10 529 467 062 11 671 6139 38
Caribbean 137 844 746 098 754 630 06
Central America 101 537 474 063 3423 1825 12
South America 96 494 436 058 7494 3682 21
All developing 16 1081 955 126 119 466 12 7523 1629
countries

IUGR-LBW=Intrauterine growth restriction-low birthweight.

Table 3: Prevalence of intrauterine growth restriction-low birthweight and its components by UN region in 2004

www.thelancet.com Vol 371 January 19, 2008 247

 Series
Term low-birthweight
Babies born at term (ie, who have completed 37 weeks of
gestation), but of low birthweight (<2500 g) are likely to
have had intrauterine growth restriction; we will refer to
this group as intrauterine growth restriction-low
birthweight. Various steps are used to estimate the
prevalence of this condition,40 which in developing
countries is present in 108% of livebirths each year. The
proportions of infants born at term weighing 15001999 g
and those weighing 20002499 g were estimated with
data sets from five countries (Bhutta Z, unpublished).4144
These proportions were applied to regional (table 3) and
See Online for webtable 5 national (webtable 5) data to estimate that, of babies born
at term globally, 955% weigh 20002499 g, and
126% weigh 15001999 g.45
Poor fetal growth is rarely a direct cause of death, but
rather can contribute indirectly to neonatal deaths,
particularly those due to birth asphyxia and infections
(sepsis, pneumonia, and diarrhoea), which together
account for about 60% of neonatal deaths. To quantify
the risk of neonatal death associated with intrauterine
growth restriction-low birthweight, data from five
community-sampled prospective birth cohorts were
analysed: two from Nepal,42,44 and one each from India,43
Pakistan (Bhutta Z, unpublished), and Brazil.41 Each
cohort provided data for gestational age, birthweight,
survivorship, and cause of death during the first 28 days
of life.
The risks of all-cause mortality and death due to birth
asphyxia and infections for newborn babies weighing
15001999 g, 20002499 g, and 2500 g or more (reference
group) were examined. Babies weighing less than 1500 g
were excluded from the analysis because they were likely
to have been born preterm. Birth asphyxia and infections
as causes of death were selected for analysis because of
strong biological associations with impaired fetal
growth.4652 With weighted linear regression techniques
Vitamin A deficient
Figure 5: Prevalence of vitamin A deficiency in children under 5 years
248
to estimate the relation between birthweight category
and the log mortality rate, the relative risks for all-cause
mortality (five studies) and for deaths due to birth
asphyxia (two studies) and infection (four studies) were
calculated. Infants born at term weighing 15001999 g
were 81 (95% CI 33193) times more likely to die, and
those weighing 20002499 g were 28 (95% CI 1844)
times more likely to die from all causes during the
neonatal period than were those weighing more than
2499 g at birth. Based on two studies from South Asia,
for deaths due to birth asphyxia, the relative risks were 54
(95% CI 18168) for those weighing 15001999 g
and 23 (1341) for those weighing 20002499 g at
birth. For infectious causes, the relative risks were 42
(15117) for those weighing 15001999 g and 20
(1234) for those weighing 20002499 g.
Vitamin A deficiency
Countries were classified (figure 5) as having vitamin A
deficiency on the basis of the most recent data for several
indicators (serum retinol, conjunctival impression cytology,
and xerophthalmia) with criteria previously described.53
Brazil and China have subnational areas of deficiency so
only populations from these areas were considered at risk.
For these calculations all children in a country (or portions
of the country for Brazil and China) classified as having
vitamin A deficiency were considered at riskie, analogous
to the baseline status of populations in which vitamin A
intervention trials have been done. This prevalence was
then reduced on the basis of the latest figures for coverage
available (2005) from UNICEF of children receiving two
doses of vitamin A supplements in the past year (considered
to be fully protected from deficiency) or one dose
(considered to be protected for half the year).
Blindness from corneal scarring that is directly due to
xerophthalmia has an estimated disease burden from
direct sequelae.54 The relative risks of cause-specific
mortality as a result of vitamin A deficiency were derived
from a meta-analysis55 of the nine randomised
placebo-controlled trials in children 659 months showing
risk reduction with supplementationie, by inverting the
reduced risk shown by intervention.55,56 This calculation
yields a relative risk of 147 (95% CI 125175) for
diarrhoea mortality and 135 (096189) for measles
mortality related to vitamin A deficiency in the
non-supplemented population as a whole. Additionally,
the findings from three trials43,57,58 of vitamin A
supplementation of newborn infants in Asia show
reductions in mortality in the first 6 months of life. The
results from these trials are applied in the first 6 months
of life to indicate a relative risk of 125 for all deaths due
to infection and two-thirds of deaths due to prematurity
ie, excluding a third of early neonatal deaths from extreme
prematurity. There has been an elevated risk of morbidity
from diarrhoea in observational studies;59 however,
placebo-controlled trials have shown an eect only on
diarrhoea severity, not incidence.55,60 One trial reported an
www.thelancet.com Vol 371 January 19, 2008

eect of vitamin A on malaria morbidity,61 but another did
not;62 thus, this eect is not included in our estimates. An
increased risk of maternal mortality due to vitamin A
deficiency was not included because results from trials in
Nepal63 and Bangladesh64 are inconsistent, leading to a
non-significant pooled estimate.
Zinc deficiency
The International Zinc Nutrition Consultative Group has
proposed a method for assessment of the populations
risk of zinc deficiency based on indirect indicatorsie,
the prevalence of stunting, one of the clinical
manifestations of zinc deficiency, and the adequacy of
absorbable zinc in the food supply at the country level.65,66
The populations risk of zinc deficiency was estimated
for the 178 countries for which information is available
from the UN Food and Agriculture Organization. The
latest prevalence of stunting for 131 countries was
obtained from the WHO Global Database on Child
Growth and Malnutrition.24 Data from other sources were
used to classify 35 additional countries by stunting
prevalence. Each country was classified into three
categories of risk of zinc deficiency on the basis of the
combination of stunting prevalence and adequacy of zinc
in the food supply.65,66 Countries at high risk of zinc
deficiency are those with a stunting prevalence of more
than 20% and estimated prevalence of inadequate zinc
intake of more than 25%; countries at low risk of zinc
deficiency are those with stunting prevalence of less
than 10% and inadequate zinc intake of less than 15%;
countries at medium risk of zinc deficiency are those
with all other combinations of the categories of stunting
prevalence and adequacy of zinc in the food supply. All
the trials of zinc supplementation included in the disease
burden risk estimates were done in countries classified
as having medium-high zinc deficiency. The national
prevalence of zinc deficiency is high in south Asia, most
of sub-Saharan Africa, and parts of Central and South
America (figure 6). In countries classified as having zinc
deficiency, all children were considered to be at risk.
Zinc deficiency in children results in increased risk of
diarrhoea, pneumonia, and malaria, as evidenced by
many randomised placebo-controlled trials done in
various populations in all regions of the world.6771 The
relative risk of deficiency for outcomes in the at-risk
populations can be calculated by inverting the reduction
in morbidity and mortality seen in the trials of zinc
supplementation. The pooled relative risk for morbidity
associated with zinc deficiency is 109 (95% CI 101118)
for diarrhoea, 125 (109143) for pneumonia, and 156
(129189) for malaria.72,73 For mortality in infants aged
159 months the relative risk is estimated to be 127
(096163) for diarrhoea, 118 (090154) for The prevalence of sequelae of iodine deficiency, such as
pneumonia, and 111 (094130) for malaria, based on a
meta-analysis (S Sazawal, unpublished) of randomised
controlled trials in Bangladesh,74 Nepal,71 and Zanzibar.69
These values were used in the estimates of disease
www.thelancet.com Vol 371 January 19, 2008
Series
No data
Low
Medium
High
Figure 6: National risk of zinc deficiency in children under 5 years
burden for the age group 659 months because this age
group has a more consistent benefit in supplementation
trials than infants younger than 6 months.
Iron deficiency anaemia
According to a WHO review of nationally representative
surveys from 1993 to 2005, 42% of pregnant women and
47% of preschool children worldwide have anaemia.75 For
these analyses, 60% of this anaemia was assumed to be
due to iron deficiency in non-malaria areas and 50% in
malaria areas.76
The major cause of iron deficiency anaemia is low
consumption of meat, fish, or poultry, especially in poor
people (panel 1).77 In young children the peak prevalence
of iron deficiency anaemia occurs around 18 months of
age and then falls as iron requirements decline and iron
intake is increased through complementary foods.
Women of childbearing age are at high risk for negative
iron balance because of blood loss during menstruation
and the substantial iron demands of pregnancy.
Previous analyses examined the relation between
anaemia in pregnancy and risk of maternal mortality.12
An odds ratio of 08 (95% CI 070091) for maternal
mortality was found for a 10 g/L increase in mean
haemoglobin in late pregnancy. These analyses also
assessed the eects of anaemia on child cognition: the
combined analysis of the five available trials found 173
(95% CI 104241) lower IQ points per 10 g/L decrease
in haemoglobin.12 A separate meta-analysis of iron
supplementation trials an overall benefit of 12 IQ points,
in children receiving iron, but there was no eect in
children younger than 27 months.78
Iodine deficiency
goitre, congenital hypothyroidism, and developmental
disability, have been estimated and assigned disability
weights to calculate disease burden.79 More recently the
prevalence of iodine deficiency has been measured by an
249
 Series
Outcome 05 months 623 months
Predominant Partial breastfeeding Not breastfeeding Not breastfeeding
breastfeeding
All cause 148 (113192)34,88,89 285 (159510)34,88,89 1440 (6093405)34,88 368 (146929)34,90
mortality
Diarrhoea 228 (085611)34,88 462 (1811177)34,88 1053 (2803964)34,88 283 (0155482)34
mortality
Pneumonia 175 (048643)34,88 249 (103604)34,88 1513 (06137384)34,88 152 (0092706)34
mortality
Diarrhoea 126 (081195)91 304 (132700)91,92 365 (169788)91,92 120 (105138)9193
incidence
Pneumonia 179 (129248)92 248 (0232715)92 207 (0192264)92 117 (037365)92
incidence
Data are point estimate (95% CI), references.
Table 4: Relative risk of suboptimum breastfeeding (compared with exclusive breastfeeding from 0 to
5 months and any breastfeeding from 6 to 23 months)
indicator of low urinary iodine concentration; however,
the rates of resulting sequelae have not been linked to
this new exposure measure. Thus, previously used
methods have been retained for the analysis in this paper.
Iodine deficiency has adverse eects on both pregnancy
outcome and child development. Even mild, subclinical
maternal iodine deficiency during pregnancy impairs
motor and mental development of the fetus and increases
risk of miscarriage and fetal growth restriction. Maternal
See Online for webtable 6 supplementation with iodine in more severe deficiency
improves pregnancy outcomes especially if done by the
second trimester, and improves neurological and
cognitive development of the infant.80 There are
insucient data on benefits to pregnancy outcomes from
interventions when deficiency is mild.81 Breast milk
iodine content is very low in areas of endemic iodine
deficiency, exacerbating depletion in infants and
increasing their risk of impaired development. A
meta-analysis showed that populations with chronic
deficiency have a 135 point reduction in IQ.82
Folic acid, vitamin B12, and other micronutrient
deficiencies
Other micronutrient deficiencies of concern in maternal
and child health include calcium, iodine, the B vitamins
(especially folic acid and vitamin B12), and vitamin D.
Calcium deficiency is recognised as the main cause of
rickets in Africa and some parts of tropical Asia, and
increasingly in other parts of the world.83 An
estimated 3580% of children in countries such as Turkey,
India, Egypt, China, Libya, and Lebanon are vitamin D
deficient (serum 25[OH]D <15 ng/mL) owing to the
practice of shrouding, avoidance of skin exposure to
sunlight, and the fact that few foods are fortified with
vitamin D.84 Vitamin D deficiency in utero can cause poor
fetal growth and skeletal mineralisation and is followed by
lower concentrations of the vitamin in breast milk. Rickets
and poor bone mineralisation subsequently appear during
the first years of life. The fetus is relatively protected from
250
maternal calcium deficiency, but calcium deficiency
rickets can result from low intakes in young children.
Poor folate status at conception, especially in the
subgroup of women who are genetically susceptible,
increases risk of neural tube and other birth defects, and
possibly pre-eclampsia and other adverse outcomes.85 In
some but not all studies maternal vitamin B12 deficiency is
a risk factor for neural tube defects and early fetal loss.86 In
women with deficiency the content of B12 in breast milk
can be so low that symptoms of deficiency appear in their
breastfed infants, including failure-to-thrive, stunting,
poor neurocognitive function, and global developmental
delays, all of which can be irreversible. Adverse pregnancy
outcomes have been seen in association with thiamin,
riboflavin, and vitamin B6 deficiencies in the few published
studies, and a recently reported trial of a supplement
containing B vitamins along with vitamins C and E found
a reduction in births with intrauterine growth restriction.87
Risks associated with child feeding practices
Suboptimum breastfeeding
The recommended feeding of children is exclusive
breastfeeding for the first 6 months of life and continued
breastfeeding through the second year of life. In Africa,
Asia, and Latin America and the Caribbean only 4757%
of infants younger than 2 months are exclusively
breastfed, an estimate that is based on an analysis of
recent national survey data (figure 7 and webtable 6 for
36 countries). For children 25 months of age this
percentage falls to 2531%. For children aged
611 months, 6% in Africa and 10% in Asia have stopped
breastfeeding, as have 32% in Latin America and the
Caribbean.
The risk of morbidity and mortality from suboptimum
breastfeeding in young children has been documented in
observational studies.34,88 A random eects meta-analysis,
which included all identified papers with appropriate
data, was used to estimate the increased risk of
cause-specific morbidity and mortality in relation to four
patterns of breastfeeding in children younger than
6 months (exclusiveie, nothing but breastmilk;
predominantonly water or teas in addition to
breast-milk; partialother liquids or solids in addition to
breast-milk; and not breastfeeding), and two patterns
(breastfeeding or not) in children aged 623 months.34,8893
Previous estimates did not separate predominant and
partial breastfeeding.94 In the first 6 months of life, the
relative risks were increased for each of the three patterns
that were compared with the reference patternie,
exclusive breastfeeding, for diarrhoea and pneumonia
morbidity and mortality (table 4). The relative risks were
significant for predominant breastfeeding for all-cause
mortality and pneumonia incidence, and there were
similar, but not significant, point estimates for diarrhoea
and pneumonia mortality and diarrhoea incidence.
Compared with exclusive breastfeeding, partial
breastfeeding had moderately higher relative risks than
www.thelancet.com Vol 371 January 19, 2008

predominant breastfeeding, and not breastfeeding had
very high relative risks. In infants aged 623 months
there was a statistically raised risk of not breastfeeding
for all-cause mortality and diarrhoea incidence, but there
was no significant raised risk for other outcomes.
Complementary feeding
Even with optimum breastfeeding children will become
stunted if they do not receive an adequate quantity and
quality of complementary foods after 6 months of age.
Most incident stunting (and wasting outside of famine
situations) happens in the first 2 years of life when children
have a high demand for nutrients and there are limitations
in the quality and quantity of their diets, especially after the
period of exclusive breastfeeding.95 This age group (younger
than 2 years) also has a high rate of infectious diseases
such as diarrhoea that adversely aect growth and
nutritional status. A study that quantified the relative
eects of poor complementary diet and infectious diseases
on reduced growth found that they were of roughly equal
importance in one setting in rural Bangladesh,96 but this
result might vary in other settings. Suboptimum
complementary feeding is clearly a determinant of
stunting, and improvements in most settings need to focus
on both feeding frequency and energy density, and ensure
an adequate quality diet, including sucient
micronutrients.97 There is a clear disparity in dietary quality,
as indicated by the lower consumption of animal source
foods by poor children (panel 1).
Information derived from ecacy trials that attempted
to improve complementary feeding through nutrition
education, food supplementation, or both, can provide a
perspective on the importance of suboptimum
complementary feeding as a determinant of stunting.
The eects on linear growth seem to be best with
interventions that use specific educational messages
eg, on consumption of animal-source foods emphasise
energy density of the diet, and, in areas with food
insecurity or low consumption of sources of
micronutrients, provide food supplements with
micronutrient fortification.98,99
Contribution of infectious diseases to stunting and
intrauterine growth restriction
Infectious diseases are important determinants of
stunting.100 Although there can be contributions to growth
faltering from respiratory illnesses101 or malaria,102 the
role of diarrhoea seems to be particularly important,103106
perhaps because of its association with malabsorption of
nutrients, as well as anorexia and catabolism.107
To quantify the contribution of diarrhoea to the
occurrence of stunting, data sets from Bangladesh (1),104
Brazil (2),103,108 Guinea-Bissau (2),109111 Ghana (1),112,113 and
Peru (3)105,114,115 were analysed. Data were pooled to estimate
the eects of diarrhoeal incidence on the odds of stunting
at 24 months of age. All were longitudinal studies that
enrolled children at or near birth and followed them with
www.thelancet.com Vol 371 January 19, 2008
Series
<2 months 25 months
100 Not
90 Partial
Predominant
80 Exclusive
70
60
50
40
30
20
10
0
Africa Asia Latin America Africa Asia Latin America
and the Caribbean and the Caribbean
Figure 7: Percentage of children by breastfeeding pattern, age group, and region
regular anthropometric measurements and daily records
of diarrhoeal surveillance. Because children were
measured at dierent ages, the measurement of
height-for-age Z score at the oldest date in the interval
between 18 and 24 months of age was accepted as the
Z score at 24 months. Logistic regression was used to
model the stunting at 24 months as a function of diarrhoeal
incidence. Children were only included if they contributed
at least 250 days of follow-up, up to the date when the
outcome was measured. The eect of diarrhoeal incidence
on stunting did not dier by study (p=0409; likelihood
ratio test), allowing us to pool across studies (webfigure). See Online for webfigure
By contrast, our analysis required study-specific intercepts
and study-specific gender eects. The odds of stunting
increased multiplicatively with each episode of diarrhoea.
The adjusted odds of stunting at 24 months of age
increased by a factor of 105 (odds ratio 105, 95% CI
103107) with each episode of diarrhoea in the first
24 months. The magnitude of this eect was not aected
when the children who were stunted in the first 6 months
of life were excluded.
Malaria in pregnancy is associated with intrauterine
growth restriction which increases the risk of death in
infancy.116 Infection at the end of pregnancy seems to be
particularly harmful.117
Global and regional disease burden from
undernutrition
The global burden of disease project
The global burden of disease project reports mortality
and burden of disease for 136 comprehensive categories
of specific diseases and injuries, and their sequelae.54 The
burden of disease measures the gap between the current
health of a population and an ideal situation where
everyone in the population lives into old age in full health,
in DALYs. DALYs combine years of life lost due to
premature death and years of life lived with disabilities
(YLD) into one indicator allowing assessment of the total
loss of health from dierent causes. One DALY can be
251
 Series
regarded as roughly 1 lost year of so-called healthy life.
Methods and data for estimating mortality and DALYs in
the global burden of disease project are described
elsewhere.54
WHO has made an incremental revision of the global
burden of disease for 2004, which included updating
mortality and causes of death, and revising YLD estimates
for selected causes if new data were available. Estimates
of child deaths by cause have been revised to take into
account new estimates of the total numbers of neonatal,
infant, and child deaths for 2004,118 updated
country-specific estimates of deaths in children younger
than 5 years by cause based on models developed by the
WHO Child Epidemiology Reference Group,119,120 and
updated UNAIDS and WHO estimates for HIV/AIDS,
tuberculosis, and vaccine-preventable child deaths.121,122
YLD estimates for protein-energy malnutrition have been
updated to take into account latest WHO estimates of
stunting and wasting according to the new WHO
Panel 2: Specific assumptions for calculating the population
attributable fraction for multiple nutritional risks
The hazardous eects of one risk are not mediated
through other risks
This assumption might not hold for zinc deficiency because
some of its hazardous eects are mediated through reducing
growth. On the basis of the eects of zinc supplementation on
linear growth,127 and the model used to estimate intervention
eects,28 it is estimated that around 46% of the excess risk for
zinc deficiency is mediated through stunting. There would be
no mediation via stunting for vitamin A deficiency128 or iron
deficiency,129 because of their lack of eects on growth.
The proportional risks for one risk do not depend on
exposure to other risks
This assumption seems to be valid in that there has been no
eect modification found in factorial design trials of
vitamin A, zinc supplementation, or both,130,131 and no
dierence in eect of vitamin A supplementation by baseline
nutritional status by various anthropometric measures.37,132
Likewise, no significant dierences have been seen with the
eects of zinc supplementation on infectious disease
morbidity in children with better or worse nutritional status.133
Exposures to these risks are uncorrelated
The data from the cohort studies used to estimate odds ratios
show that there was a small positive correlation between
severe wasting and stunting (the prevalences of severe
wasting in children with height-for-age Z scores below and
above 2 were 47% and 29%, respectively). There was also a
positive correlation between stunting and zinc deficiency, but
this is accounted for in the mediated eect mentioned above.
Correlation was not incorporated in the calculations, which
results in a slight underestimation of the combined eects,
but previous sensitivity analyses have shown that the
underestimation is small for these risks.126
252
standards,23 and YLD estimates for iron deficiency
anaemia have been updated using the most recent data
from the WHO Global Database on Anaemia.123 All
estimates of death and disease burden attributable to
undernutrition in this analysis are for the year 2004.
The 136 diseases and injuries in the global burden of
disease project follow the rules of the international
classification of disease (ICD) system. The ICD considers
four groups of nutritional deficiencies, and their sequelae,
as possible direct causes of death: protein-energy
malnutrition, iron deficiency anaemia, vitamin A
deficiency, and iodine deficiency. For this reason,
estimates were made of the number of deaths and
amount of disability that could be directly attributed to
these conditions. As described above, in addition to these
direct deaths, undernourished children are at an
increased risk of illness and death from many infectious
diseases, also included in the global burden of disease
categories. Indeed, substantially more children die as a
result of the synergy between infectious diseases and
undernutrition than from the direct sequelae of the
nutritional conditions themselves.11 Like the ICD, global
burden of disease also includes low birthweight as a
direct cause of death; nearly all of these deaths in
developing countries are due to preterm births, not
intrauterine growth restriction among births at term.119
Estimating deaths and burden of disease attributable to
nutritional risk factors
To estimate the number of deaths from various infectious
diseases attributable to dierent forms of undernutrition,
the methods of the comparative risk assessment project
were used.124,125 In brief, for each risk factor and for each of
the diseases that are aected by it, population attributable
fractions were computed using the following equation,
which is for a risk factor with n levels of exposure.
Population attributable fractions estimate the proportional
reduction in mortality that would be seen if risk factor
exposure had been reduced to an alternative (counterfactual)
distribution that would result in the lowest population risk,
irrespective of whether currently attainable in practice.
This optimum exposure, referred to as the theoretical-
minimum-risk exposure distribution, results in estimates
of deaths and disease burden that could be avoided if all
levels of suboptimum nutrition were addressed, in a
consistent and comparable way across risk factors.
n n
P RR P RR
i i i i
PAF= i=1
n
i=1
P RR i i
i=1
PAF=population attributable fractions. RRi=relative
risk of disease or mortality for the ith exposure category.
Pi=proportion of children/women of child-bearing age
in the ith exposure category. Pi=proportion of
children/women of child-bearing age in the ith exposure
www.thelancet.com Vol 371 January 19, 2008
 Series

category in alternative/counterfactual exposure

Deaths Percentage of Disease burden Percentage of
scenarios. n=number of exposure categories. deaths in children (1000 DALYs) DALYs in children
Deaths and disease burden associated with under 5 years under 5 years
protein-energy malnutrition (an ICD condition), and its Underweight* 1 957 530 190 81 358 187
direct sequelae as estimated in global burden of disease,54 Stunting 1 491 188 145 54 912 126
were added to the burden associated with underweight. Wasting* 1 505 236 146 64 566 148
In estimating the disease burden for stunting and Severe wasting* 449 160 44 25 929 60
wasting it is appropriate to add the direct protein-energy Intrauterine growth restriction- 337 047 33 13 536 31
malnutrition burden to only one; it was added to wasting low birthweight
(or severe wasting) because this condition is most Total of stunting, severe wasting, 2 184 973 214 90 962 212
consistent with mortality from severe acute malnutrition, and intrauterine growth
even though YLDs for chronic eects were attributed to restriction-low birthweight

both stunting and wasting.
The nutritional status measuresunderweight,
stunting, and wastingwere treated as risk factors whose
exposures and relative risks applied to all children in the
age group 159 months, and intrauterine growth
restriction-low birthweight only for the first month of life.
The analyses for vitamin A and zinc deficiencies and
suboptimum breastfeeding, for which either exposure or
increased risk were age-dependent, had to be done for age
groups other than all children under 5 or neonatal, the
two groups that are readily available from global burden
of disease databases. The division of cause-specific deaths
between 4 weeks and 59 months by age (eg, for the interval
15 months, 611 months, 1223 months) were from data
for total (all-cause) deaths by month of age estimated
using demographic techniques from complete birth
histories in demographic and health surveys.
Exposures and relative risks for the calculation of
infectious disease deaths and burden from
nutrition-related risk factors were largely newly derived,
as described in this paper; however, the estimation of the
direct sequelae was done using previous methods with
the results updated to 2004. A summary of the source of
the inputs is provided in webtable 7.
Cause-specific deaths can be caused by multiple risk
factors acting simultaneously, and hence can be prevented
by intervention on each of the risks. For example, some
deaths from diarrhoea could be prevented by removing
exposure to stunting, zinc deficiency, vitamin A deficiency,
or suboptimum breastfeeding. Such deaths are attributable
to all four risks when considered individually. As a result
of multi-causality, the population attributable fractions for
multiple risk factors that aect the same disease outcome
overlap and cannot be combined by simple addition;
rather their combined eect is generally less than the
crude sum of individual ones, because some deaths are
attributed to multiple exposures.126 Under specific
assumptions (panel 2), the combined (joint) population
attributable fraction that avoids double-counting can be
estimated from those of individual risks.126
Deaths and burden of disease attributable to nutritional
risk factors
Of the nutritional status measures for children,
underweight was responsible for the largest disease burden
*Deaths (138 739) and DALYs (14 486 400) directly attributed to protein energy malnutrition included Included in
wasting. Total takes into account the joint distribution of stunting and severe wasting.
Table 5: Global deaths and disease burden measured in disability-adjusted life-years (DALYs) in children
under 5 years of age attributed to nutritional status measures in 2004
Deaths Percentage of deaths in Disease burden Percentage of DALYs in
children under 5 years (1000 DALYs) children under 5 years
Vitamin A deficiency 667 771 65 22 668 53
Zinc deficiency 453 207 44 16 342 38
Iron deficiency 20 854 02 2 156 05
Iodine deficiency 3 619 003 2 614 06
Table 6: Global deaths and disability-adjusted life-years (DALYs) in children under 5 years of age
attributed to micronutrient deficiencies in 2004
(table 5). The burden due to severe wasting from infectious
diseases is modest compared with the other measures,
largely because most children with non-optimum weight
for height have a Z score of 1 to 3 and substantially fewer
have a score less than 3. The burden associated with all
levels of wasting was slightly less than that of stunting.
Stunting, severe wasting, and intrauterine growth
restriction-low birthweight together were responsible for See Online for webtable 7
21 million deaths (21% of worldwide deaths in children
under 5) and 910 million DALYs (21% of global DALYs
for children under 5; 7% of global total DALYs). Of the
UN subregions, the disease burden attributed to these
anthropometric measures is highest in south-central
Asia (webtable 8), where India alone has 06 million See Online for webtable 8
deaths and 246 million DALYs attributed to stunting,
severe wasting, and intrauterine growth restriction-low
birthweight. The undernutrition burden is also high in
eastern, middle and western Africa; these subregions,
which together have an under-5 population of 111 million
(93% of India), 11 million deaths, and 423 million
DALYs, showing higher relative and absolute eects
among African children.
Among the deficiencies of vitamins and minerals
examined, the largest disease burdens were attributed to
vitamin A and zinc deficiencies (table 6). Vitamin A
deficiency in newborn babies, infants, and children
resulted in about 6% of under-5 deaths, 5% of under-5
DALYs, and 17% of total DALYs. Zinc deficiency resulted
in about 4% of under-5 deaths and DALYs and 1% of total

www.thelancet.com Vol 371 January 19, 2008 253

 Series
DALYs. The regional patterns for disease burden
attributed to vitamin A and zinc deficiencies are similar.
The highest burden for each is in South-central Asia,
See Online for webtable 9 followed by several subregions of Africa (webtable 9).
Iron deficiency resulted in a relatively small number of
under-5 deaths and DALYs (table 6). This burden is directly
attributed to the sequelae of iron deficiency anaemia in
children including cognitive impairment.76 In addition,
iron deficiency anaemia in pregnancy is considered a risk
factor for maternal mortality; 115 000 deaths per year from
maternal causes, resulting in 34 million DALYs, have
been attributed to this risk factor.12 Iodine deficiency
results in a very small number of deaths, which is
attributed to an excess risk of death with congenital
hypothyroidism, and a modest number of DALYs that
primarily result from multiple sequelae from iodine
deficiency, including cognitive and motor impairment and
hearing loss.79
In an analysis taking into account the joint distribution
of nutritional status risk factors (intrauterine growth
restriction, stunting, severe wasting, and deficiencies of
vitamin A or zinc) and the fact that some of the eect of
zinc deficiency is mediated through stunting, it is
estimated that there were 28 million child deaths
(28% of under-5 deaths) and 1140 million DALYs (27% of
under-5 DALYs) attributable to these forms of childhood
undernutrition. This constitutes 85% of the total global
disease burden.
There was a large disease burden attributed to
suboptimum breastfeeding, including 14 million deaths
(12% of under-5 deaths) and 435 million DALYs, which
is 10% of global under-5 DALYs and 3% of total DALYs.
Most of the attributable deaths (106 million) and DALYs
(370 million) were due to non-exclusive breastfeeding in
the first 6 months of life, accounting for 77% and 85%,
respectively, of deaths and DALYs attributed to
suboptimum breastfeeding. Again the highest disease
burden estimates are for south-central Asia and several
See Online for webtable 10 sub-regions of Africa (webtable 10). This risk factor was
combined with anthropometric status and deficiencies of
vitamin A and zinc in an analysis allowing for co-exposure
and avoiding double counting of disease burden. If one
assumes that the risks of suboptimum breastfeeding and
other nutritional factors are independent, the combined
mortality eects of all risk factors were 36 million child
deaths (35% of under-5 deaths) and 1405 million DALYs
(35% of under-5 DALYs); this is 10% of the total global
disease burden. These estimates change only slightly if
the burden of disease directly attributed to iron and
iodine deficiencies are added. These results are robust to
assumptions about some of the eects of suboptimum
breastfeeding being mediated through other nutritional
exposures. For example, if 25% of the hazardous eects
of suboptimum breastfeeding are mediated through
other nutritional risks, the total number of attributable
child deaths is reduced to 35 million, still about
35% deaths and DALYs in this age group. Adding the
254
maternal deaths and DALYs due to iron deficiency
anaemia increases the total global disease burden
attributed to undernutrition to 11%.
Discussion
The attribution of more than a third of child deaths and
more than 10% of total global disease burden to maternal
and child undernutrition demonstrates the importance of
these prevalent risk factors. The 21% of global deaths and
DALYs in children younger than 5 years old attributed to
stunting, severe wasting, and intrauterine growth
restriction-low birthweight, largely because of their
synergistic relationship with infectious diseases, constitutes
the largest percentage for any risk factor in this age
group.125
The estimated total burden attributed to the com-
bination of underweight and intrauterine growth
restriction-low birthweight is about the sameie, 22% of
deaths, and disease burden in children younger than 5
yearsas the burden attributed to stunting, severe
wasting, and intrauterine growth restriction-low
birthweight. This result is lower than our previous
estimate that indicated that 35% of child deaths could be
attributed to childhood underweight and maternal low
body-mass index operating through intrauterine growth
restriction to aect low birthweight.11 There are several
reasons for this dierence. The new estimates were done
for a dierent base year (2004 vs 2000) with modest
reductions in overall child mortality and in some major
causes of death, such as diarrhoea, malaria, and measles,
that are aected by the nutritional risk factors. Further,
there have been declining trends in the prevalence of
stunting in most regions.134 More importantly the current
estimates used the new WHO growth standards for all
nutritional status prevalence and mortality and morbidity
risk estimates. These factors, along with the use of
dierent data sets for much of the risk analysis, resulted
in somewhat lower relative risks of mortality than in
previous analyses,10,11 especially for children with Z scores
between 1 and 2. Thus, some of the dierence in
estimated disease burden is due to changes in methods,
but it also is likely to indicate progress in preventing and
managing undernutrition and reducing some important
causes of death.
The methods used in these analyses indicate several
advances from those used previously.11,12,56,73 Most of the
risk relations were re-examined, as were the exposure
definitions and prevalences, with the most current
evidence. The latest information on total and cause-specific
child mortality was used. The new WHO growth standards
were used for all analyses. Importantly, the prevalence of
stunting is about 15% higher with the new standards than
with the previous growth reference.135 The attributable
deaths and DALYs were estimated for intrauterine growth
restriction, stunting, and severe wasting, separately and
together, as well as for all wasting and underweight.
Co-exposures were considered, and the resulting estimate
www.thelancet.com Vol 371 January 19, 2008

for the anthropometric and micronutrient factors, as well
as suboptimum breastfeeding, provides a total disease
burden without double counting that results from simple
addition of the burdens from individual factors.
These estimations using a risk factor method show that
449 000 child deaths can be attributed to severe wasting.
Estimates by others136 have ranged up to 17 million deaths
using a methodology based on actual mortality rates in
studies done an average 18 years (maximum 25 years)
ago, when overall child mortality rates were much higher
than today. The case fatality rates implied by these two
calculations would be 2% versus 9%. True case-fatality
rates for all cases of severe wasting identified in a
cross-sectional population survey are unknown, but are
almost certainly lower than these of severely malnourished
children presenting to hospitals or feeding centres in
emergency settings. Futhermore, over time case fatality
rates have decreased because of better treatment. Since
the prevalence of HIV infection has increased in some
populations, some of the deaths reported to have been
due to protein-energy malnutrition in civil registration
systems could actually have been due to AIDS, but the
magnitude of this possible misclassification is unknown.
Finally, direct application of case fatality (versus
proportional risk) overlooks the fact that some of the
exposed children would have non-zero risk of death
regardless of their severe wasting status. Irrespective of
whether the true number of deaths attributable to severe
wasting is closer to a half million or to 15 million a year,
therapeutic feeding interventions are important com-
ponents of nutrition programmes in settings where
severe wasting is common.
Deficiencies of vitamin A and zinc were confirmed to
be important risk factors in terms of their eects on total
child mortality. The current estimate of disease burden
due to vitamin A deficiency is similar to previous
estimates despite increases in the population coverage
with vitamin A supplementation, because of inclusion of
risk in the 05 month age group for the first time. In
other words, the true decline in the burden in children
above 6 months of age has been numerically compensated
by the finding that vitamin A is important in an even
younger age group. That the estimates of disease burden
due to zinc deficiency are about 40% lower than
previously reported73 is because of the lower relative risks
based on analysis of additional studies.72 The
WHO/UNICEF recommendation that zinc be used
along with oral rehydration therapy for all childhood
diarrhoea is just beginning to be implemented in many
countries, and preventive use of zinc supplementation
or fortification is uncommon. The high residual burden
of those micronutrient deficiencies indicates that
interventions need to be expanded, especially in
south-central Asia and Africa.
Iodine and iron deficiencies currently result in only
small disease burdens. For iodine deficiency, there has
been substantial success with promotion of the use of
www.thelancet.com Vol 371 January 19, 2008
Series
iodised salt; this and other interventions should be
continued to prevent resurgence of the related disease
burden. For iron deficiency the disease burden in
children is relatively small, accounting for only 05% of
under-5 DALYs, despite inclusion in the calculations of
life-long developmental disability in 20% of children who
have ever had severe anaemia.76 This postulated
permanent disability has not been seen in all studies and
its existence might need to be reconsidered in the
future.137 Nevertheless, children with iron deficiency
anaemia should be treated to prevent any possible current
and future adverse eects. Iron deficiency anaemia is an
important contributor to maternal mortality, increasing
the risk of dying with blood loss during delivery.12
Suboptimum breastfeeding has large mortality
consequences worldwide, similar to those of stunting.
The estimate reported here is the same (14 million
deaths) as a previous estimate using a dierent
categorisation of breastfeeding practices (combining
partial and predominant breastfeeding) and dierent
mortality risks.94,138 The timing of initiation of
breastfeeding was not considered in these estimates.
There is epidemiological evidence to suggest that
beginning breastfeeding within the first day post partum
would have additional benefit with regard to mortality
even in exclusively breastfed infants, rearming
recommendations to begin breastfeeding immediately
after delivery.89 More than three quarters of the burden
attributed to suboptimum breastfeeding is due to
non-exclusive breastfeeding in the first 6 months of life
when even provision of water or teas leads to an increased
risk of death. These estimates do not consider the adverse
eect of transmission of HIV in breast milk. Recent
evidence indicates that conditions in which breast milk
substitutes are recommended for babies of HIV-positive
mothers are rarely met, and there is a net benefit of
breastfeeding in terms of HIV-free survival even in
populations with a high prevalence of HIV infection.139141
Exclusive breastfeeding results in lower rates of HIV
transmission than partial breastfeeding with rates
of 1%142,143 and 4%140 being reported from studies in Africa.
With around 2 million babies born to HIV-positive
mothers every year,144 these transmission rates would
result in 20 00080 000 HIV-infected infants in the first
6 months of life. Antiretroviral treatment for pregnant or
lactating women will substantially reduce these resulting
infections, and such treatment is increasing in Africa.144
Because HIV transmission continues with partial
breastfeeding after the first 6 months of life,145 early
breastfeeding cessation has been considered an option.
Evidence is still incomplete; however, a recent trial found
no dierence in HIV-free survival at 24 months of age in
children who were randomly assigned to abrupt weaning
at 4 months of age or continued breastfeeding.146
Inadequate complementary feeding contributes to
stunting as do infectious diseases, especially repeated
episodes without adequate case management. These
255
 Series
exposures, which are arguably the key determinants of
stunting, were not modelled as risk factors in this paper.
However, the eects on stunting and related disease
burden of interventions to improve complementary
feeding and to reduce diarrhoea incidence are calculated
in a subsequent paper in the Series.28 Results presented
here on the eects of diarrhoea on the risk of stunting are
incorporated in these calculations of intervention eects,
but other infectious diseases and perhaps asymptomatic
infections are also likely to contribute to stunting.
The nutrition-related risk factors individually result in
large disease burdens, but often coexist both in individuals
and in populations. Stunting and severe wasting in
children 159 months of age are largely uncorrelated, and
the estimation of the joint eects of these two risk factors
resulted in a slightly smaller total disease burden than
simply adding their eects because even in the absence of
statistical correlation, some children would be exposed to
both conditions. Intrauterine growth restriction-low
birthweight was considered for these estimations to aect
only infants younger than 1 month, but in reality could
aect later ages. Vitamin A, zinc, iron, and iodine
deficiencies and the anthropometric measures as risk
factors are largely uncorrelated, but there is still overlap in
their risk because there is multicausality, such as diarrhoea
for vitamin A, zinc, and stunting. In addition to simple
multicausality, about half of the eect of zinc deficiency is
mediated through stunting; the rest is a more direct eect
on morbidity and mortality, probably as a result of reduced
immune function.147 The risk related to suboptimum
breastfeeding might in part be due to micronutrient
deficiencies resulting from inadequate dietary intake, but
is also due to avoidance of infection. Calculating the
combined eects of suboptimum breastfeeding and the
other nutrition-related risk factors found that, even after
accounting for overlapping eects from multicausality of
diarrhoea and pneumonia, about half of the disease
burden attributable to suboptimum breastfeeding was
added to the other risk factors. Therefore, even if all other
Panel 3: Research needs
Development of methods to assess nutritional status and
its determinants
Prevalence of deficiencies of vitamin A, zinc, iron, and
iodine in subnational populations
Consequences of nutritional deficiencies for mortality
from HIV/AIDS, malaria, and other important infectious
diseases
Consequences of nutritional deficiencies for immune
competence, brain development, cognitive ability, and
other possible eects
Overlap of micronutrients and their joint eects on
mortality and morbidity
Development of international fetal and newborn growth
standards
256
nutritional risks were addressed, a substantial number of
child deaths would still require interventions related to
breastfeeding practices.
Additional research will be important for improving
understanding of the distribution of nutritional de-
ficiencies in populations and their risk relations. Some
specific research topics are listed in panel 3. Despite the
remaining research needs, these results, which show
the high prevalence of various forms of maternal and
child undernutrition and their health consequences,
present a compelling justification for implementation
of eective nutrition-related interventions. The potential
eects of these interventions are considered in the third
paper in this Series, with particular attention to their
contribution to achievement of MDG 1 with reduced
undernutrition and also MDGs 4 and 5 regarding child
and maternal mortality.
Contributors
RB conceptualised and coordinated the analyses and preparation of the
paper. Primary responsibility for specific topics were as follows: LHA,
maternal undernutrition and consequences of selected micronutrient
deficiencies; ZB, complementary feeding; LEC, analysis of intrauterine
growth restriction-low birthweight and childhood undernutrition risks
for mortality; MdO, stunting, wasting and low birthweight prevalence by
country and region; ME, analysis of risk factor attributable mortality and
disease burden; CM, update of country, regional, and global
disease-specific mortality and burden of disease for 2004; and JR,
exposure to zinc deficiency by country. All authors contributed to the
final paper.
Other contributors
Analyses of national surveys for prevalence of underweight, stunting and
wasting based on new WHO Child Growth Standards: Monika Blssner,
Yang Hong, Allen Shoemaker (WHO, Switzerland). Provision of
estimates of prevalence of breastfeeding status: Ana Betrn, Jeremy Lauer,
(WHO, Switzerland). Review of relative risks of suboptimum
breastfeeding: Ai Koyanagi, Sunil Sazawal (Johns Hopkins Bloomberg
School of Public Health, USA). Provision of panel 1 (equity gaps in regard
to stunting and dietary diversity): Satoru Shimokawa (Cornell University,
USA), Cesar Victora (University of Pelotas, Brazil). Provision of data for
analysis of mortality risks related to anthropometric indicators:
Keith West, Luke Mullany, Parul Christian, Larry Moulton, James Tielsch
(Johns Hopkins Bloomberg School of Public Health, USA), Linda Adair
(University of North Carolina School of Public Health, USA), Kre Mlbak
(Statens Serum Institut, Denmark), Zulfiqar A Bhutta (Aga Khan
University, Pakistan), A M Menezes (University of Pelotas, Brazil),
Michel Garenne (Institute Pasteur, France), Shams El Arifeen
(International Centre for Diarrhoeal Diseases Research, Bangladesh),
M Andersen (University of Copenhagen, Denmark). Provision of data for
analysis of socioeconomic confounding: Saul S Morris (London School of
Hygiene and Tropical Medicine, United Kingdom), Keith West (Johns
Hopkins Bloomberg School of Public Health, USA). Review of risk related
to maternal body-mass index: Anju Aggarwal (Johns Hopkins Bloomberg
School of Public Health, USA). Provision of data on vitamin A
supplementation coverage: Nita Dalmiya (UNICEF, USA). Assistance in
classifying countries as vitamin A deficient: Keith West (Johns Hopkins
Bloomberg School of Public Health, USA). Analysis of the contribution of
diarrhoea to stunting: William Checkley, Gillian Buckley (Johns Hopkins
Bloomberg School of Public Health, USA). Analyses of mortality and
burden of disease attributable to nutrition risk factors: Rodrigo Dias
(Harvard University Initiative for Global Health, USA). Analyses of age
patterns of deaths in children younger than 5 years: Kenneth Hill,
Thomas Laakso (Harvard University Initiative for Global Health, USA).
Analyses of contribution of undernutrition to mortality by cause at birth
and during childhood, confounding by socioeconomic factors and of
correlation between stunting and wasting: Ping Chen, Carmen Carrillo
(Johns Hopkins Bloomberg School of Public Health, USA).
www.thelancet.com Vol 371 January 19, 2008

Maternal and Child Undernutrition Study Group
Series steering committeeRobert E Black (Johns Hopkins Bloomberg
School of Public Health, USA), Zulfiqar A Bhutta (Aga Khan University,
Pakistan), Jennifer Bryce (Johns Hopkins Bloomberg School of Public
Health, USA), Saul S Morris (London School of Hygiene and Tropical
Medicine, UK), Cesar G Victora (Federal University of Pelotas, Brazil).
Other membersLinda Adair (University of North Carolina, USA),
Tahmeed Ahmad (ICDDR,B, Bangladesh), Lindsay H Allen (USDA ARS
Western Human Nutrition Research Center, USA), Laura E Caulfield
(Johns Hopkins Bloomberg School of Public Health), Bruce Cogill
(UNICEF, USA), Denise Coitinho (WHO, Switzerland), Simon Cousens
(London School of Hygiene and Tropical Medicine, UK), Ian Darnton-Hill
(UNICEF, USA), Mercedes de Onis (WHO, Switzerland); Kathryn Dewey
(University of California, Davis, USA), Majid Ezzati (Harvard School of
Public Health, USA), Caroline Fall (University of Southhampton, UK),
Elsa Giugliani (Federal University of Rio Grande de Sul, Brazil),
Batool A Haider (Aga Khan University, Pakistan), Pedro Hallal (Federal
University of Pelotas, Brazil), Betty Kirkwood (London School of Hygiene
and Tropical Medicine, UK), Reynaldo Martorell (Emory University,
Rollins School of Public Health, USA), Colin Mathers (WHO,
Switzerland), David Pelletier (Cornell University, USA),
Per Pinstrup-Andersen (Cornell University, USA), Linda Richter (Human
Sciences Research Council, South Africa), Juan A Rivera (Mexico National
Institute of Public Health), Harshpal Singh Sachdev (Sitaram Bhartia
Institute of Science and Research, India), Meera Shekar (World Bank,
USA), Ricardo Uauy (Institute of Nutrition, Chile).
Conflict of interest statement
We declare that we have no conflict of interest. As corresponding author,
R Black states that he had full access to all data and final responsibility to
submit for publication.
Acknowledgments
Funding for the preparation of the Series was provided by the Bill &
Melinda Gates Foundation. Meetings were hosted by the UNICEF
Innocenti Research Centre and the Rockefeller Foundation Bellagio
Conference Center. The sponsors had no role in the analysis and
interpretation of the evidence nor in writing the report and the decision
to submit for publication. We thank Barbara Ewing and Mary Rybczynski
for administrative assistance with the Series.
References
1 Haddah H, Alderman H, Appleton S, Song L, Yisehac Y. Reducing
child malnutrition: How far does income growth take us?
World Bank Economic Review 2003; 17: 10731.
2 Ezzati M, Vander Hoorn S, Lawes CMM, et al. Rethinking the
diseases of auence paradigm: global patterns of nutritional risks
in relation to economic development. PLoS Medicine 2005; 2: e133.
3 Prentice AM, Moore SE. Early programming of adult diseases in
resource poor countries. Arch Dis Child 2005; 90: 42932.
4 Caballero B, Popkin BE, eds. The nutrition transition: diet and
disease in the developing world. London: Academic Press, 2002.
5 Salomon JA, Murray CJL. The epidemiologic transition
re-examined: compositional models for causes of death by age and
sex. Population and Development Review 2002; 28: 20528.
6 Grantham-McGregor S, Cheung YB, Cueto S, et al. Developmental
potential in the first five years for children in developing countries.
Lancet 2006; 369: 6070.
7 UN Millennium Project 2005. Halving hunger: it can be done.
London and Sterling, VA: Task Force on Hunger, 2005.
8 The World Bank. Repositioning nutrition as central to development.
A strategy for large-scale action. Washington DC: The World Bank,
2006.
9 Caulfield LE, de Onis M, Blossner M, Black RE. Undernutrition as
an underlying cause of child deaths associated with diarrhea,
pneumonia, malaria, and measles. Am J Clin Nutr 2004; 80: 19398.
10 Pelletier DL, Frongillo EA Jr, Habicht JP. Epidemiologic evidence
for a potentiating eect of malnutrition on child mortality.
Am J Public Health 1993; 83: 113033.
11 Fishman SM, Caulfield L, de Onis M, et al. Childhood and maternal
underweight. In: Ezzati M, Lopez AD, Rodgers A, Murray CLJ, eds.
Comparative quantification of health risks: global and regional
burden of disease attributable to selected major risk factors.
Geneva: World Health Organization, 2004: 39161.
www.thelancet.com Vol 371 January 19, 2008
Series
12 Stoltzfus RJ, Mullany L, Black RE. Iron deficiency anaemia. In:
Ezzati M, Lopez AD, Rodgers A, Murray CLJ, eds. Comparative
quantification of health risks: global and regional burden of disease
attributable to selected major risk factors. Geneva: World Health
Organization, 2004: 163209.
13 Waterlow JC. Classification and definition of protein-calorie
malnutrition. BMJ 1972; 3: 56669.
14 Kramer MS. Determinants of low birth weight: methodological
assessment and meta-analysis. Bull World Health Organ 1987;
65: 663737.
15 Kramer MS, Victora C. Low birth weight and perinatal mortality. In:
Semba RD, Bloem MW, eds. Nutrition and health in developing
countries. Humana Press, 2001.
16 WHO. Maternal anthropometry and pregnancy outcomes: A WHO
Collaborative Study World Health Organ Suppl 1995; 73: 3237.
17 Ronsmans C, Holtz S, Stanton C. Socioeconomic dierentials in
caesarean rates in developing countries: a retrospective analysis.
Lancet 2006; 368: 151623.
18 Villar J, Valladares E, Wojdyla D, et al. Caesarean delivery rates and
pregnancy outcomes: the 2005 WHO global survey on maternal and
perinatal health in Latin America. Lancet 2006; 367: 181929.
19 Cnattinguis R, Cnattinguis S, Notzon F. Obstacles to reducing
cesarean rates in a low-cesarean setting: the eect of maternal age,
height and weight. Obstet Gynecol 1998; 92: 50106.
20 Dempsey J, Ashiny Z, Qiu C, et al. Maternal pre-pregnancy
overweight status and obesity as risk factors for cesarean delivery.
Matern Fetal Neonatal Med 2005; 17: 17985.
21 Allen LH. Maternal micronutrient malnutrition: eects on breast
milk and infant nutrition, and priorities for intervention. SCN News
1994; 11: 2124.
22 WHO. WHO Child Growth Standards: Length/height-for-age,
weight-for-age, weight-for-length, weight-for-height and body mass
index-for-age: methods and development. Geneva: World Health
Organization, 2006.
23 de Onis M, Garza C, Onyango AW, Martorell R. WHO Child Growth
Standards. Acta Paediatr Suppl 2006; 450: 1101.
24 de Onis M, Blossner M. The World Health Organization Global
Database on Child Growth and Malnutrition: methodology and
applications. Int J Epidemiol 2003; 32: 51826.
25 de Onis M, Blossner M, Borghi E, Morris R, Frongillo EA.
Methodology for estimating regional and global trends of child
malnutrition. Int J Epidemiol 2004; 33: 126070.
26 de Onis M, Blossner M, Borghi E, Frongillo EA, Morris R. Estimates
of global prevalence of childhood underweight in 1990 and 2015.
JAMA 2004; 291: 260006.
27 UN Department of Economics and Social Aairs, Population
Division. World Population Prospects, the 2004 revision. New York:
United Nations, 2005.
28 Bhutta ZA, Ahmad T, Black RE, et al, for the Maternal and Child
Undernutrition Study Group. What works? Interventions for
maternal and child undernutrition and survival. Lancet 2008;
published online Jan 17. DOI: 10.1016/S0140-6736(07)61693-6.
29 Gwatkin DR, Rutstein S, Johnson K, Suliman EA, Wagsta A.
Initial country-level information about socio-economic dierences
in health, nutrition, and population 2nd edn. Washington DC:
World Bank, 2003.
30 Measure DHS. Demographic and Health Surveys. About DHS
homepage. http://www.measuredhs.com/aboutdhs (accessed
Nov 21, 2007).
31 Victora CG. The association between wasting and stunting: an
international perspective. J Nutr 1992; 122: 110510.
32 Adair LS, Popkin BM, VanDerslice J, et al. Growth dynamics in the
first two years of life: a prospective study in the Philippines.
Eur J Clin Nutr 1993; 47: 4251.
33 WHO/CHD Immunisation-Linked Vitamin A Supplementation
Study Group. Randomised trial to assess benefits and safety of
vitamin A supplementation linked to immunisation in early
infancy. Lancet 1998; 352: 125763.
34 Arifeen S, Black RE, Antelman G, et al. Exclusive breastfeeding
reduces acute respiratory infection and diarrhea deaths among
infants in Dhaka slums. Pediatrics 2001; 108: E67.
35 Garenne M, Maire B, Fontaine O, Dieng K, Briend A. Risks of dying
associated with dierent nutritional status in pre-school aged
children. Dakar: ORSTOM, 1987 (reprinted by CEPED, Paris, 2000).
257
 Series

Maternal and Child Nutrition 1

Maternal and child undernutrition and overweight in
low-income and middle-income countries
Robert E Black, Cesar G Victora, Susan P Walker, Zulfiqar A Bhutta*, Parul Christian*, Mercedes de Onis*, Majid Ezzati*,
Sally Grantham-McGregor*, Joanne Katz*, Reynaldo Martorell*, Ricardo Uauy*, and the Maternal and Child Nutrition Study Group

Maternal and child malnutrition in low-income and middle-income countries encompasses both undernutrition and Lancet 2013; 382: 42751
a growing problem with overweight and obesity. Low body-mass index, indicative of maternal undernutrition, has Published Online
declined somewhat in the past two decades but continues to be prevalent in Asia and Africa. Prevalence of maternal June 6, 2013
http://dx.doi.org/10.1016/
overweight has had a steady increase since 1980 and exceeds that of underweight in all regions. Prevalence of stunting
S0140-6736(13)60937-X
of linear growth of children younger than 5 years has decreased during the past two decades, but is higher in south
This publication has been
Asia and sub-Saharan Africa than elsewhere and globally aected at least 165 million children in 2011; wasting corrected. The corrected version
aected at least 52 million children. Deficiencies of vitamin A and zinc result in deaths; deficiencies of iodine and first appeared at thelancet.com
iron, together with stunting, can contribute to children not reaching their developmental potential. Maternal on June 20, 2013
undernutrition contributes to fetal growth restriction, which increases the risk of neonatal deaths and, for survivors, See Comment page 371
of stunting by 2 years of age. Suboptimum breastfeeding results in an increased risk for mortality in the first 2 years This is the first in a Series of
of life. We estimate that undernutrition in the aggregateincluding fetal growth restriction, stunting, wasting, and four papers about maternal and
child nutrition
deficiencies of vitamin A and zinc along with suboptimum breastfeedingis a cause of 31 million child deaths
*Members listed alphabetically
annually or 45% of all child deaths in 2011. Maternal overweight and obesity result in increased maternal morbidity
Members listed at end of paper
and infant mortality. Childhood overweight is becoming an increasingly important contributor to adult obesity,
diabetes, and non-communicable diseases. The high present and future disease burden caused by malnutrition in Johns Hopkins University,
Bloomberg School of Public
women of reproductive age, pregnancy, and children in the first 2 years of life should lead to interventions focused on Health, Baltimore, MD, USA
these groups. (Prof R E Black MD,
Prof P Christian DrPH,
Introduction continuing stunting of growth and deficiencies of essential Prof J Katz ScD); Universidade
Federal de Pelotas, Pelotas,
Maternal and child malnutrition, encompassing both nutrients along with obesity in national populations and Rio Grande do Sol, Brazil
undernutrition and overweight, are global problems with within families. We also want to assess national progress (Prof C G Victora MD); The
important consequences for survival, incidence of acute in nutrition programmes and international actions con- University of the West Indies,
Tropical Medicine Research
and chronic diseases, healthy development, and the sistent with our previous recommendations.
economic productivity of individuals and societies.
Maternal and child undernutrition, including stunting,
wasting, and deficiencies of essential vitamins and Key messages
minerals, was the subject of a Series15 in The Lancet in Iron and calcium deficiencies contribute substantially to maternal deaths
2008, which quantified their prevalence, short-term and Maternal iron deficiency is associated with babies with low weight (<2500 g) at birth
long-term consequences, and potential for reduction Maternal and child undernutrition, and unstimulating household environments,
through high and equitable coverage of proven nutrition contribute to deficits in childrens development and health and productivity in adulthood
interventions. The Series identified the need to focus on Maternal overweight and obesity are associated with maternal morbidity, preterm
the crucial period of pregnancy and the first 2 years of birth, and increased infant mortality
lifethe 1000 days from conception to a childs second Fetal growth restriction is associated with maternal short stature and underweight
birthday during which good nutrition and healthy growth and causes 12% of child deaths
have lasting benefits throughout life. The 2008 Series also Stunting prevalence is slowly decreasing globally, but aected at least 165 million
called for greater national priority for nutrition pro- children younger than 5 years in 2011; wasting aected at least 52 million children
grammes, more integration with health programmes, Suboptimum breastfeeding results in more than 800 000 child deaths annually
enhanced intersectoral approaches, and more focus Undernutrition, including fetal growth restriction, suboptimum breastfeeding,
and coordination in the global nutrition system of inter- stunting, wasting, and deficiencies of vitamin A and zinc, cause 45% of child deaths,
national agencies, donors, academia, civil society, and the resulting in 31 million deaths annually
private sector. 5 years after that series, we intend not Prevalence of overweight and obesity is increasing in children younger than 5 years
only to reassess the problems of maternal and child globally and is an important contributor to diabetes and other chronic diseases
undernutrition, but also to examine the growing problems in adulthood
of overweight and obesity for women and children and Undernutrition during pregnancy, aecting fetal growth, and the first 2 years of life is
their consequences in low-income and middle-income a major determinant of both stunting of linear growth and subsequent obesity and
countries (LMICs). Many of these countries are said to non-communicable diseases in adulthood
suer the so-called double burden of malnutrition, with

www.thelancet.com Vol 382 August 3, 2013 427

 Series

Institute, Mona Campus,
Kingston, Jamaica
(Prof S P Walker PhD);
The Aga Khan University and
Medical Center, Department of
Pediatrics, Karachi, Pakistan
(Prof Z A Bwhutta PhD); World
Health Organization,
Department of Nutrition for
Health and Development,
Geneva, Switzerland
(M de Onis MD); Imperial College
of London, St Marys Campus,
School of Public Health,
MRC-HPA Centre for
Environment and Health,
Department of Epidemiology
and Biostatistics, London, UK
(Prof M Ezzati PhD); Institute of
Child Health, University College
London, London, UK (Prof
S Grantham-McGregor FRCP);
The University of the West
Indies, Mona, Jamaica
(Prof S Grantham-McGregor);
Emory University, Atlanta, GA,
USA (Prof R Martorell PhD); and
London School of Hygiene and
Tropical Medicine, London, UK
(Prof R Uauy PhD)
Correspondence to:
Prof Robert Black, Johns Hopkins
University, Bloomberg School of
Public Health, Baltimore,
MD 21205, USA
rblack@jhsph.edu
See Online for appendix
The present Series is guided by a framework (figure 1)
that shows the means to optimum fetal and child growth
and development, rather than the determinants of
undernutrition as shown in the conceptual model
developed by UNICEF and used in the 2008 Series.1 This
new framework shows the dietary, behavioural, and
health determinants of optimum nutrition, growth, and
development and how they are aected by underlying
food security, caregiving resources, and environmental
conditions, which are in turn shaped by economic and
social conditions, national and global contexts, resources,
and governance. This Series examines how these
determinants can be changed to enhance growth and
development. These changes include nutrition-specific
interventions that address the immediate causes of
suboptimum growth and development. The framework
shows the potential eects of nutrition-sensitive inter-
ventions that address the underlying determinants of
malnutrition and incorporate specific nutrition goals
and actions. It also shows the ways that an enabling
environment can be built to support interventions and
programmes to enhance growth and development and
their health consequences. In the first paper we assess
the prevalence of nutritional conditions and their health
and development consequences. We deem a life-course
perspective to be essential to conceptualise the nutritional
eects and benefits of interventions. The nutritional
status of women at the time of conception and during
pregnancy is important for fetal growth and development,
and these factors, along with nutritional status in the first
2 years of life, are important determinants of both
undernutrition in childhood and obesity and related
diseases in adulthood. Thus, we organise this paper to
consider prevalence and consequences of nutritional
conditions during the life course from adolescence to
pregnancy to childhood and discuss the implications for
adult health. In the second paper, we describe evidence
supporting nutrition-specific interventions and the health
eects and costs of increasing their population coverage.
In the third paper we examine nutrition-sensitive inter-
ventions and approaches and their potential to improve
nutrition. In the fourth paper we examine the features of
an enabling environment that are needed to provide
support for nutrition programmes and how they can be
favourably changed. Finally, in a Comment6 we will
examine the desired national and global response to
address nutritional and developmental needs of women
and children in LMICs.
Prevalence and consequences of nutritional
conditions
Adolescent nutrition
Adolescent nutrition is important to the health of girls
and is relevant to maternal nutrition. There are 12 billion
adolescents (aged 1019 years) in the world, 90% of whom
live in LMICs. Adolescents make up 12% of the population
in industrialised countries, compared with 19% in LMICs
(appendix p 2 shows values for ten countries studied in
depth).7 Adolescence is a period of rapid growth and
maturation from childhood to adulthood. Indeed, some
researchers have argued that adolescence is a period
with some potential for height catch-up in children with

Benefits during the life course

Morbidity and Cognitive, motor, School performance Adult stature Work capacity
mortality in childhood socioemotional development and learning capacity and productivity
Obesity and NCDs

Nutrition specific Optimum fetal and child nutrition and development Nutrition sensitive
interventions programmes and approaches
and programmes Agriculture and food security
Adolescent health and Breastfeeding, nutrient- Feeding and caregiving Low burden of Social safety nets
preconception nutrition rich foods, and eating practices, parenting, infectious diseases Early child development
Maternal dietary routine stimulation Maternal mental health
supplementation Womens empowerment
Micronutrient Child protection
supplementation or Food security, including Feeding and caregiving Access to and use of Classroom education
fortification availability, economic resources (maternal, health services, a safe and Water and sanitation
Breastfeeding and access, and use of food household, and hygienic environment Health and family planning services
complementary feeding community levels)
Dietary supplementation
for children
Dietary diversification Building an enabling environment
Knowledge and evidence Rigorous evaluations
Feeding behaviours and
Politics and governance Advocacy strategies
stimulation
Leadership, capacity, and financial resources Horizontal and vertical coordination
Treatment of severe acute
Social, economic, political, and environmental context (national and global) Accountability, incentives regulation,
malnutrition
Disease prevention and legislation
management Leadership programmes
Nutrition interventions in Capacity investments
emergencies Domestic resource mobilisation

Figure 1: Framework for actions to achieve optimum fetal and child nutrition and development

428 www.thelancet.com Vol 382 August 3, 2013

 Series

stunting from early childhood.8 Adolescent fertility is
three times higher in LMICs than in high-income coun-
tries. Pregnancies in adolescents have a higher risk of
complications and mortality in mothers9 and children10
and poorer birth outcomes than pregnancies in older
women.10,11 Furthermore, pregnancy in adolescence will
slow and stunt a girls growth.12,13 In some countries, as
many as half of adolescents are stunted (height-for-age
Z score [HAZ] <2), increasing the risk of poor perinatal
outcomes in their ospring (appendix p 2). We used
age-specific, low body-mass index (BMI) cutos (BMI
Z score [BMIZ] <2) from the WHO reference for
children aged 519 years to examine ten selected
countries; in these locations as many as 11% (India) of
adolescent girls are thin. In these countries, prevalence of
high BMI for age, defined as BMIZ >2 (obesity), is as
high as 5% (Brazil; appendix p 2). Adolescents have as
high a prevalence of anaemia as women aged 2024 years.
In India, for example, 558% of adolescents aged
1519 years and 567% of women aged 2024 years were
anaemic;14 corresponding values for Guatemala were
210% and 204 %, respectively.15
Maternal nutrition
Prevalence of low BMI (<185 kg/m) in adult women
has decreased in Africa and Asia since 1980, but remains
higher than 10% in these two large developing regions
(figure 2). During the same period, prevalence of
overweight (BMI 25 kg/m) and obesity (BMI 30 kg/m)
has been rising in all regions, together reaching more
than 70% in the Americas and the Caribbean and more
than 40% in Africa by 2008.16,17
Few studies have examined the risk of maternal
mortality in relation to maternal anthropometry with a
prospective design. In one study in Nepal18 of about
22 000 women, mid upper arm circumference during

Africa Americas and the Caribbean

600 1980 2000
1985 2005
500 1990 2008
1995
400
Prevalence (%)

300

200

100

Asia Europe
600

500

400
Prevalence (%)

300

200

100

Oceania Global
600

500

400
Prevalence (%)

300

200

100

0
BMI 185 BMI 25 BMI 30 BMI 185 BMI 25 BMI 30

Figure 2: Trends in thinness (BMI <185 kg/m), overweight (BMI 25 kg/m), and obesity (BMI 30 kg/m), using population weighted average prevalences
for women aged 2049 years UN regions and globally, 19802008
Error bars are 95% CIs. BMI=body-mass index.

www.thelancet.com Vol 382 August 3, 2013 429

 Series

pregnancy was inversely associated with all-cause mater-
nal mortality up to 42 days post-partum after adjusting
for numerous factors. An inverse association exists
between maternal height and the risk of dystocia (dicult
labour), as measured by cephalopelvic disproportion or
assisted or caesarean deliveries.1923
Figure 2 also shows trends for overweight and obesity
in women aged 2049 years in dierent UN regions.
Oceania, Europe, and the Americas had the highest
proportion of overweight and obese women; however
northern and southern Africa, and central and west Asia
also had high prevalences (appendix p 3).
Maternal obesity leads to several adverse maternal and
fetal complications during pregnancy, delivery, and
post-partum.24 Obese pregnant women (pre-pregnant
BMI 30 kg/m) are four times more likely to develop
gestational diabetes mellitus and two times more likely to
develop pre-eclampsia compared with women with a
BMI 185249 kg/m2).2528 During labour and delivery,
maternal obesity is associated with maternal death,
haemorrhage, caesarean delivery, or infection;2931 and a
higher risk of neonatal and infant death,32 birth trauma,
and macrosomic infants.3337 In the post-partum period,
obese women are more likely to delay or fail to lactate and
to have more weight retention than women of normal
weight.38 Obese women with a history of gestational
diabetes have an increased risk of subsequent type
two diabetes, metabolic syndrome, and cardiovascular
disease.39 The early intrauterine environment has a role in
programming phenotype, aecting health in later life.
Maternal overweight and obesity at the time of pregnancy
increases the risk for childhood obesity that continues
into adolescence and early adulthood, potentiating the
transgenerational transmission of obesity.40,41
Maternal vitamin deficiencies
Anaemia and iron
Anaemia (haemoglobin <110 g/L), which might be
attributable to low consumption or absorption in the diet
or to blood loss, such as from intestinal worms, is highly
prevalent during pregnancy. This Series focuses on
anaemia amenable to correction with iron supplemen-
tation.42 To establish the importance of iron deficiency as
a cause of maternal anaemia we used the results of trials
of iron supplementation to work out the shift in the
population haemoglobin distribution. A meta-analysis of
the eects of iron supplementation trials showed that,
among pregnant women with anaemia at baseline, iron
supplementation led to a 102 g/L increase in haemo-
globin.43 The corresponding figure for children was
80 g/L.44 We applied these shifts to the present distri-
butions of haemoglobin estimated by Stevens and
colleagues42 and calculated the proportion of pregnant
women with anaemia whose blood haemoglobin would
increase to at least 110 g/L. We likewise calculated the
proportion of severe anaemia that would increase to at
least 70 g/L. These results constitute the prevalence of
iron amenable or iron deficiency anaemia (IDA) or severe
IDA, defined as the proportion of anaemia or severe
anaemia that would be reduced if only iron was provided,
holding other determinants of anaemia unchanged. With
this approach in 2011, Africa had the highest proportion
of IDA for pregnant women followed closely by Asia
(table 1,4547 appendix p 4). Likewise, Africa had the highest
prevalence of severe IDA, but in all regions prevalence
was less than 1%.
Our previous analyses48 showed that anaemia in preg-
nancy increased the risk of maternal mortality. An
updated analysis49 with ten studies (four more than the

Vitamin A deficiency45 Iodine Zinc deficiency47 Iron deficiency anaemia (haemoglobin

deficiency46 (weighted <110 g/L)
(UIC <100 g/L) average of
country means)
Children <5 years Pregnant women Children <5 years Pregnant women
Night Serum retinol Night Serum retinol
blindess <070 mol/L blindness <070 mol/L
Global 09% 333% 78% 153% 285% 173% 181% 192%
(0118) (294371) (6591) (60246) (282289) (159188) (156208) (171215)
Africa 21% 416% 94% 143% 400% 239% 202% 203%
(1031) (344449) (81107) (97190) (394406) (211268) (186217) (183224)
Americas 06% 156% 44% 20% 137% 96% 127% 152%
and the (0013) (66245) (2762) (0436) (125148) (68124) (98160) (117186)
Carribean
Asia 05% 335% 78% 184% 316% 194% 190% 198%
(0013) (307363) (6690) (54314) (307325) (169220) (145234) (158235)
Europe 07% 149% 29% 22% 442% 76% 121% 162%
(0015) (01297) (1146) (0043) (435450) (6291) (78162) (126197)
Oceania 05% 126% 92% 14% 173% 57% 154% 172%
(0110) (60192) (03182) (0040) (166181) (10103) (70252) (97256)

Data are % (95% CI). UIC=urine iodine concentration.

Table 1: Prevalence of vitamin A deficiency (19952005), iodine deficiency (2013), inadequate zinc intake (2005), and iron deficiency anaemia (2011)

430 www.thelancet.com Vol 382 August 3, 2013


previous analysis) showed that the odds ratio (OR) for
maternal deaths was 071 (95% CI 060085) for a
10 g/L greater mean haemoglobin in late pregnancy.
Only two of the ten studies adjusted for socioeconomic
confounding variables; one showed no attenuation and
the other a 20% attenuation of the eect.
There is strong biological plausibility for a causal link
between maternal IDA and adverse birth outcomes includ-
ing low birthweight and increased perinatal mortality.5052 A
meta-analysis53 that included 11 trials indentified a
significant 20% reduction in the risk of low birthweight
associated with antenatal supplementation with iron alone
or combined with folic acid (relative risk [RR] 080, 95% CI
071090). A previous Cochrane review54 had much the
same findings. Dibley and colleagues55 pooled data from
demographic and health surveys from Indonesia for 1994,
1997, 200203, and 2007 and showed that risk of death of
children younger than 5 years was reduced by 34% when
the mother consumed any iron-folic acid supplements
(hazard ratio [HR] 066; 95% CI 053081). Dibley and
colleagues further showed that the protective eect was
greatest for deaths on the first day of life (040; 021077),
but the protective eect was also shown for neonatal
deaths (069; 95% CI 049097) and post-neonatal deaths
(074; 056099). A randomised controlled trial56 from
China showed a significant 54% (RR 046, 95% CI
021098) reduction in neonatal mortality with antenatal
iron and folic acid supplementation compared with folic
acid alone as control. In Nepal, mortality from birth to
7 years was reduced by 31% (HR 069, 95% CI 049099)
in the ospring of mothers who had received iron and folic
acid during pregnancy compared with controls who
received vitamin A only.57
Randomised controlled trials from high-income
countries have shown benefits of iron supplementation
for improved maternal mental health and reduced
fatigue.58 Evidence from LMICs for the eect of mater-
nal IDA on mothers mental health and mother-child
interactions is scarce. In a small South African trial,59
iron supplementation of women with IDA from
10 weeks post-partum to 9 months led to lower maternal
depression and perceived stress at 9 months compared
with placebo. At 9 months, iron supplemented mothers
had better maternal-child interactions.60 By contrast, in
Bangladesh higher levels of maternal iron supple-
mentation decreased the quality of maternal-child
interaction at age 34 months and had no eect on
maternal distress (anxiety and depression).61
There is some evidence for whether maternal IDA
aects child development. Infants of mothers identified
as having IDA at 68 weeks post-partum had lower
developmental levels at 10 weeks and 9 months compared
with infants of control mothers without IDA.60 In Nepal,
children whose mothers received iron and folate supple-
mentation during pregnancy had better general intelli-
gence and cognitive functioning at age 79 years
compared with children of mothers receiving placebo,
www.thelancet.com Vol 382 August 3, 2013
Series
suggesting that benefits can be detected in later childhood
when more complex tasks can be measured.62
Vitamin A
Maternal vitamin A deficiency can cause visual impair-
ment and possibly other health consequences. Deficiency
is assessed in pregnant women as either a history of night
blindness or serum or plasma retinol concentrations of
less than 070 mol/L (subclinical vitamin A deficiency).
WHO provides prevalence estimates for 19952005 from
64 countries, which we used for estimates for the UN
world regions (table 1).45 Globally, the prevalence of night
blindness in pregnant women is estimated to be 78%
(95% CI 7087), aecting 97 million women. An
estimated 153% (74232) of pregnant women globally
(191 million women) have deficient serum retinol con-
centrations. The degree to which night blindness and low
serum retinol overlap is not accounted for in this esti-
mation, but night blindness is known to be associated
with a four-times higher odds of low serum retinol
(OR 402, 95% CI 2274).63 Night blindness is reduced
by vitamin A supplementation in pregnancy.63,64 Maternal
night blindness has been associated with increased low
birthweight64 and infant mortality,65 yet trials of vitamin A
in pregnancy have not showed significant eects on
these outcomes.6669
Zinc
Zinc is a key micronutrient with a ubiquitous role in
biological functions, including protein synthesis, cellular
division, and nucleic acid metabolism. Estimates revised
in 2012 suggest that 17% of the worlds population is at
risk of zinc deficiency, on the basis of an analysis of
national diets.47 Excess losses of zinc during diarrhoea
also contribute to zinc deficiency. The eect of subclinical
zinc deficiency (defined as low plasma zinc concentration
without obvious signs of zinc deficiency) in women of
reproductive age and during pregnancy on health and
development outcomes is poorly understood, although
zinc deficiency has been suggested as a risk factor with
adverse long-term eects on growth, immunity, and
metabolic status of surviving ospring.70 Zinc deficiency
due to a rare genetic abnormalityacrodermatitis entero-
pathicain pregnancy results in a high risk of preterm
and prolonged labour, post-partum haemorrhage, and
fetal growth restriction.70,71 A review of supplementation
trials with zinc in pregnancy showed a significant 14%
reduction in preterm births in women in low-income
settings, but no significant eect on low birthweight.72
Iodine
Maternal iodine deficiency is of concern in regard to
adverse eects on fetal development, yet few countries
have nationally representative data from large-scale
surveys of urinary iodine concentration in pregnant
women. Because of the correlation between urinary
iodine concentration in pregnant women and children
431
 Series
aged 612 years (r 069),73 status assessment in school-
age children is used to estimate country, regional, and
global prevalence of iodine deficiency. Global estimates
of iodine deficiency suggest that 285% of the worlds
population or 19 billion individuals are iodine-
deficient (table 1).46,74 This figure represents largely mild
deficiency (defined as urinary iodine concentration of
5099 ug/L).
Severe iodine deficiency in pregnancy causes cretin-
ism, which can be eliminated with iodine supplemen-
tation before conception or in the first trimester of
pregnancy.75 Furthermore, two meta-analyses showed
average deficits of 125135 intelligence quotient (IQ)
points in children associated with iodine deficiency of
their mothers in pregnancy; however, they controlled
for only a limited number of socioeconomic con-
founders.76,77 A review of the eects of iodine supple-
mentation in deficient populations showed a small
increase in birthweight.78 Eects of mild or moderate
iodine deficiency on brain development are not well
established.78,79 The index of iodine deficiency (urinary
iodine concentration) is a population measure and not
an individual one,80 therefore some individuals in
regions of mild to moderate deficiency might have more
severe deficiency.
Folate
The global prevalence of folate deficiency has not been
estimated because of the scarcity of suitable population-
based data.81 A substantial proportion of neural tube
defects (congenital malformations of the spinal cord and
brain) are related to inadequate consumption of folic
acid around the time of conception, in some populations
associated with genetic factors that increase the need for
dietary folic acid. A Cochrane review82 in 2010 included
five trials of folic acid (a synthetic form of folate)
supplementation and identified a 72% (RR 028; 95% CI
015052) reduction in the risk of neural tube defects.
A more recent systematic review had much the same
findings and estimated that in 2005 56 000 deaths were
attributable to insucient dietary folic acid.83 These
deaths were not added to the total deaths associated with
undernutrition in the present analysis, because of the
uncertainty about this estimate.
Calcium and vitamin D
Calcium is an essential nutrient for several body func-
tions, including enzymatic and hormonal homoeostasis.
Evidence for the association between maternal dietary
calcium intake and maternal bone density and fetal
mineralisation is inconsistent.84 Epidemiological evi-
dence does show an inverse association between calcium
intake and development of hypertension in pregnancy.85,86
Gestational hypertensive disorders are the second
leading cause of maternal morbidity and mortality and
are associated with increased risk of preterm birth and
fetal growth restriction.87,88
432
Substantial evidence suggests that calcium supple-
mentation in pregnancy is associated with a reduction
in gestational hypertensive disorders and preterm
birth.89,90 However, the eect varies according to the
baseline calcium intake of the population and pre-
existing risk factors. A review of 1591 randomised con-
trolled trials suggested that calcium supplementation
during pregnancy was associated with a reduction in the
risk of gestational hypertension and a 52% reduction
in the incidence of pre-eclampsia, along with a
24% reduction in preterm birth and an increase in birth-
weight of 85 g. There was no eect on low birthweight
or perinatal or neonatal mortality. The eect was mainly
noted in populations with low calcium intake.92 The
eects of calcium supplementation interventions are
described in the accompanying report by Zulfiqar A
Bhutta and colleagues.93
The US Institute of Medicine has defined adequate
vitamin D status as having serum 25-hydroxyvitamin D
([OH]D) concentrations greater than 50 nmol/L in both
the general population and pregnant women;94 serum
concentrations of less than 25 nmol/L denote vita-
min D deficiency whereas concentrations of less than
50 nmol/L denote vitamin D insuciency.95 Although
few nationally representative surveys exist for vitamin D
status, an estimated 1 billion people globally residing in
diverse geographies, many in LMICs, might be vitamin D
insucient or deficient.96103
Vitamin D has an essential role in fetal development,
ensuring fetal supply of calcium for bone development,
enabling immunological adaptation required to main-
tain normal pregnancy, preventing miscarriage, and
promoting normal brain development.99,104108 Poor
maternal vitamin D status has been associated with
severe pre-eclampsia (new-onset gestational hyper-
tension and proteinuria after 20 weeks of gestation) in
turn leading to an increased risk of perinatal morbidity
and mortality.99,109,110 Maternal vitamin D deficiency,
especially in early pregnancy, has been associated
with risk of pre-eclampsia (OR 209, 95% CI
150290), preterm birth (158, 108231), and small-
for-gestational age (SGA; 152, 108225).109,111 A
systematic review of three trials of vitamin D in
pregnancy showed an overall reduction of low
birthweight of borderline significance (relative risk
[RR] 048; 95% CI 023101).111 These associations
need to be better quantified before they can be included
in the global disease burden related to undernutrition.
Eect of maternal stature or BMI on fetal
growth restriction or postnatal growth
Maternal characteristics
The 2008 Maternal and Child Undernutrition Series
examined the association of maternal nutritional status
(BMI and short stature) and fetal growth restriction,
defined as low birthweight at term. Here, we use data
from nine (height) and seven (BMI) population-based
www.thelancet.com Vol 382 August 3, 2013
 Series

cohort studies and WHO perinatal facility-based data for
24 countries to examine associations separately for term
and preterm SGA in LMICs.112121 Maternal stunting
(height <145 cm) put infants at risk of term and preterm
SGA (appendix p 5). Low maternal BMI in early pregnancy
also put infants at higher risk of SGA (appendix p 5). BMI
of 25 or greater was somewhat protective against term and
preterm SGA (appendix p 5). Notably, most women in the
BMI category of greater than 25 kg/m were very mildly
overweight with very few obese women, which might
explain the protective eect.
Maternal stature is a composite indicator representing
genetic and environmental eects on the growing period
of childhood. In a study122 involving 109 Demographic
studies that use dierent reference populations for SGA.
Previous studies have also not separated SGA into term
and preterm. In a pooled-analysis130 of 22 population-
based cohort studies in LMICs in Asia, sub-Saharan
Africa, and Latin America, the RR for neonatal (128 day)
mortality associated with SGA (<10th centile)
was 183 (95% CI 134250) and for post-neonatal
(29365 day) mortality was 190 (132273), compared
with appropriate-for-gestational-age (AGA) infants. The
RR for term SGA was 306 (95% CI 221423) for
neonatal mortality and 198 (139281) for post-neonatal
350 Prevalence preterm SGA
Prevalence term SGA

Health Surveys, analyses adjusted for wealth, education, 300

and urban or rural residence showed that the absolute
risk of dying among children younger than 5 years born 250

to the tallest mothers (160 cm) was 0073 (95% CI

Proportion (%)

200
00720074) and to those born to the shortest mothers
(<145 cm) was 0128 (01260130). The correspond- 150
ing absolute risk for a child being stunted was
0194 (01920196) for the tallest mothers and 100
0682 (06730690) for the shortest. The association
with wasting was significant but much weaker.122 50

Fetal growth restriction 0

Africa Asia Latin America Oceania
Previous global and regional estimates of fetal growth and the Caribbean
restriction used the term low birthweight as a proxy for
Figure 3: Prevalence of small-for-gestational-age births, by UN regions
being SGA in the absence of population-based birthweight
and gestational age data at that time.1 Through recent
Attributable Proportion of Attributable Proportion of
analyses, we now have estimates of SGA prevalence from
deaths with UN total deaths deaths with NIMS total deaths
22 population-based cohort studies and 23 countries with prevalences* of children prevalences of children
facility-based data,114 which were used to model SGA as a younger than younger than
function of low birthweight and other covariates (neonatal 5 years 5 years
mortality rate, representativeness of facility delivery) to Fetal growth restriction 817 000 118% 817 000 118%
obtain country-specific SGA prevalence for 2010.123 The (<1 month)
numbers on which the model is based include all livebirths, Stunting (159 months) 1 017 000* 147% 1 179 000 170%
but exclude babies who died so soon after birth that they Underweight (159 months) 999 000* 144% 1 180 000 170%
were not weighed. Imputing birthweight for these infants Wasting (159 months) 875 000* 126% 800 000 115%
did not change the estimation of SGA prevalence, although Severe wasting 516 000* 74% 540 000 78%
(159 months)
it did increase the mortality risk associated with being
Zinc deficiency 116 000 17% 116 000 17%
born SGA. Figure 3 shows prevalence of SGA separated
(1259 months)
into term and preterm SGA. By our estimate, in 2010
Vitamin A deficiency 157 000 23% 157 000 23%
324 million babies were born SGA, 27% of all births in (659 months)
LMICs regions. When comparing the estimated numbers Suboptimum breastfeeding 804 000 116% 804 000 116%
of children with SGA with those aected by stunting or (023 months)
wasting, it is important to note that the cuto for SGA is Joint eects of fetal growth 1 348 000 194% 1 348 000 194%
the 10th centile of a reference population, whereas the restriction and suboptimum
breastfeeding in neonates
cutos for wasting or stunting are two Z scores below the
Joint eects of fetal growth 3 097 000 447% 3 149 000 454%
median, or the 23rd centile. Appendix p 8 shows prevalence
restriction, suboptimum
with 95% CIs of term and preterm SGA by UN subregions. breastfeeding, stunting,
Notably, only 20% of preterm births in LMICs were also wasting, and vitamin A and
SGA. Appendix p 8 show the ten countries with the highest zinc deficiencies (<5 years)
number of SGA births in 2010. Data are to the nearest thousand. *Prevalence estimates from the UN. Prevalence estimates from Nutrition Impact
Associations between fetal growth restriction and Model Study (NIMS).
infant survival have been previously reported,124129
Table 2: Global deaths in children younger than 5 years attributed to nutritional disorders
although it is dicult to compare associations across

www.thelancet.com Vol 382 August 3, 2013 433

 Series
mortality, relative to term AGA infants. Infants born
preterm and SGA were at highest risk with RR of 1542
(9112612) for neonatal mortality and 522 (283964) contributor to stunting and wasting in children. To
for post-neonatal mortality, relative to term AGA.
By applying the attributable fractions of deaths to the
total neonatal and post-neonatal deaths for 2011 obtained
from the UN Interagency Group on Mortality Esti-
mation,131 we estimated that the number of deaths
attributed to SGA in 2011 was 817 000 in neonates and
418 000 in infants aged 111 months using standard
methods.1 The largest number of attributed deaths were
in Asia (appendix p 22). For the calculation in the group
who were both SGA and preterm, the counterfactual was
AGA and preterm so that the SGA eect was separated
from that of being preterm. In the estimation of the
deaths attributed to several nutritional conditions, we
attributed only the neonatal deaths to SGA; for children
aged 159 months, we attributed deaths to wasting and
stunting, to which SGA contributed (table 2).
Panel 1: Determinants of childhood stunting and overweight
The determinants of optimum growth and development
(figure 1) consist of factors operating at dierent levels of
causation, ranging from the most distal socioeconomic and
political determinants to the proximate level where food,
disease, and care have a crucial role. A mirror image of figure 1
would show the determinants of linear growth failurethe
process leading to child stuntingand overweight. The large
socioeconomic inequalities in stunting prevalence in almost all
low-income and middle-income countries (LMICs) show the
great importance of distal determinants. In particular, maternal
education is associated with improved child-care practices
related to health and nutrition and reduced odds of stunting,
and better ability to access and benefit from interventions.
Almost all stunting takes place in the first 1000 days after
conception. The few randomised controlled trials of
breastfeeding promotion145 that included nutritional status
outcomes did not show any eects on the weight or length of
infants. By contrast, there is strong evidence that the
promotion of appropriate complementary feeding practices
reduces the incidence of stunting.93 A meta-analysis of zinc
supplementation trials146 has shown a significant protective
eect against stunting.
Severe infectious diseases in early childhoodsuch as
measles, diarrhoea, pneumonia, meningitis, and malaria
can cause acute wasting and have long-term eects on linear
growth. However, studies have consistently shown that
diarrhoea is the most important infectious disease
determinant of stunting of linear growth. In a pooled
analysis147 of nine community-based studies in low-income
countries with daily diarrhoea household morbidity and
longitudinal anthropometry, the odds of stunting at
24 months of age increased multiplicatively with each
diarrhoea episode or day of diarrhoea before that age. The
434
Causes of childhood growth faltering are multifactorial,
but fetal growth restriction might be an important
quantify the association between fetal growth restriction
and child undernutrition, we did a systematic scientific
literature search to identify longitudinal studies that had
taken measurements of birthweight, gestational age, and
child anthropometry. Data could be obtained from 19 birth
cohort studies which were submitted to a meta-analysis to
examine the odds of stunting, wasting, and underweight
in children 1260 months of age associated with SGA and
preterm birth.132 Four mutually exclusive exposure cate-
gories were created: AGA and preterm; SGA and term;
SGA and preterm; and term AGA (as the reference group).
The meta-analysis showed that SGA alone and preterm
alone were associated with increased overall ORs of 243
(95% CI 222266) and 193 (171218), respectively,
whereas both SGA and preterm increased the OR to 451
(342593) for stunting (appendix p 23). These raised
proportion of stunting attributed to five previous episodes of
diarrhoea was 25% (95% CI 838).
Environmental (or tropical) enteropathy is an acquired
disorder, characterised by reduced intestinal absorptive
capacity, altered barrier integrity, and mucosal inflammation,
occurring in young children living in unsanitary settings.148
These children also have high rates of symptomatic and
asymptomatic infections with enteric pathogens, but the exact
association of these infections or of other possible toxic enteric
exposures with enteropathy is unclear. Some researchers have
suggested that these functional changes and the associated
inflammation have significant adverse eects on the growth of
children. Alternatively, these changes might be a consequence
of nutritional deficits very early in life, including in utero, that
lead to intestinal microbial colonisation.
Optimum growth in the first 1000 days of life is also essential for
prevention of overweight. Whereas attained weight at any age in
early life is positively associated with adult body-mass index in
LMIC cohorts,2,149,150 rapid weight gains in the first 1000 days are
strongly associated with adult lean mass, whereas weight gains
later in childhood lead mainly to adult fat mass. In particular,
evidence suggests that infants whose growth faltered in early life,
and who gained weight rapidly later in childhood, might be at
particular risk of adult obesity and non-communicable diseases.2
Child overweight is also related to growing up in an obesogenic
environment, in which population changes in physical activity
and diet are the main drivers. Modifiable risk factors for
childhood obesity are maternal gestational diabetes; high levels
of television viewing; low levels of physical activity; parents
inactivity; and high consumption of dietary fat, carbohydrate,
and sweetened drinks, yet few interventions have been
rigorously tested.151153
www.thelancet.com Vol 382 August 3, 2013
 Series

ORs were also noted in the Asian, African, and Latin
American UN regions (data not shown).
We estimated population attributable risk for childhood
stunting for the risk categories of SGA and preterm
birth. Because risk estimates were derived as ORs using
logistic regression analysis, we used a method to
approximate the risk ratio proposed by Zhang and Yu133
for estimation of the population attributable risk. Using
the approximated RR estimates across all 19 cohorts,
population attributable risk for SGA-term for stunting
was 016 (012019), that for SGA-preterm was 004
(002005), and that for AGA-preterm was 004
(002006). The combined population attributable risk
related to SGA for stunting was 020 and that for preterm
birth was 008. Thus, overall we estimate that about a
fifth of childhood stunting could have its origins in the
fetal period, as shown by being born SGA.
Most studies of fetal growth restriction and childhood
cognitive and motor development in LMICs involve term
infants of low birthweight or examine birthweight
adjusted for gestational age. Consistent evidence exists
for associations of fetal growth restriction with lower
psychomotor development levels in early childhood (up
to age 36 months) with small to moderate eect sizes
compared with infants of normal birthweight.134,135 A
study136 from Bangladesh showed much the same
associations for both mental and motor development.
Evidence suggests that there is an eect on development
that is attributable to birth size, independent of that
attributable to poor postnatal growth.137
Evidence for eects of fetal growth restriction on
cognition and behaviour after early childhood is less
consistent. Birthweight was associated with attained
schooling in the COHORTS analyses;138 however, this
was unadjusted for gestational age. Size at birth was
not related to womens educational achievement in
Guatemala139 and term low birthweight was not associated
with IQ and behaviour in school-aged children in Brazil140
and Jamaica,141 or behaviour in South Africa.142 In Taiwan,
term infants of low birthweight had lower academic
achievement at age 15 years than did infants of normal
birthweight,143 and in Thailand birthlength was associated
with IQ at age 9 years independent of postnatal growth to
age 1 year;144 however, in both cases eect sizes were small.
Childhood nutrition
Stunting, underweight, and wasting
Panel 12,145153 describes the determinants of stunting and
overweight in children. Estimates of the prevalence of
stunting, underweight, and wasting worldwide and for
UN subregions are based on analyses jointly done by
UNICEF, WHO, and the World Bank154 of 639 national
surveys from 142 countries in the WHO database, using
standard methods.1,155,156 In 2011, globally, 165 million
children younger than 5 years had a height-for-age
Z score (HAZ) of 2 or lower (stunted) on the basis of

Stunting (HAZ <2) Wasting (WHZ <2) Severe wasting (WHZ <3) Underweight (WAZ <2)
UN155 NIMS158 UN NIMS UN NIMS UN NIMS
Pro- Number Pro- Number Pro- Number Pro- Number Pro- Number Pro- Number Pro- Number Pro- Number
portion (millions) portion (millions) portion (millions) portion (millions) portion (millions) portion (millions) portion (millions) portion (millions)
Africa 356% 563 355% 566 85% 134 79% 125 35% 55 28% 44 177% 279 184% 290
(333 (525 (344 (543 (74 (116 (73 (115 (29 (4564) (25 (4052) (157 (247 (174 (275
380) 600) 366) 578) 96) 152) 86) 136) 41) 33) 197) 311) 191) 301)
Asia 268% 958 295% 1035 101% 361 112% 392 36% 129 36% 127 193% 691 219% 766
(232 (828 (264 (925 (79 (282 (95 (334 (24 (84 (27 (94 (146 (521 (188 (659
305) 1088) 313) 1098) 123) 440) 127) 446) 48) 173) 47) 164) 241) 861) 240) 841)
Latin 134% 71 146% 78 14% 07 15% 08 03% 02 04% 02 34% 18 37% 20
America (94 (4894) (136 (7382) (09 (0510) (13 (0710) (02 (0102) (04 (0203) (2345) (1224) (35 (1822)
and the 177) 155) 19) 18) 04) 06) 41)
Carib-
bean
Oceania 355% 05 347% 04 43% 01 51% 01 07% 00 15% 00 140% 02 139% 02
(160 (0208) (278 (0305) (30 (0001) (33 (0001) (05 (0000) (09 (0000) (80 (0103) (107 (0102)
614) 395) 62) 68) 11) 24) 232) 168)
LMICs 28% 1597 299% 1683 88% 503 93% 526 33% 185 31% 173 174% 990 194% 1091
(256 (1459 (279 (1573 (74 (421 (84 (474 (25 (140 (26 (144 (143 (817 (173 (972
304) 1734) 310) 1746) 103) 584) 104) 585) 40) 231) 38) 215) 204) 1163) 205) 1154)
High- 72% 51 17% 12 03% 02 24% 17
income (41 (2989) (08 (0625) (00 (0009) (1734) (1224)
countries 126) 35) 13)
Global 257% 1648 80% 515 29% 187 157% 1007
(235 (1508 (68 (433 (22 (142 (130 (833
279) 1788) 93) 596) 36) 232) 184) 1180)

Data are % (95% CI). HAZ=height-for-age Z score. WHZ=weight-for-height Z score. WAZ=weight-for-age Z score. LMICs=low-income and middle-income countries.

Table 3: Prevalence and numbers of children younger than 5 years with stunting, wasting, severe wasting, and underweight using estimates from UN and NIMS, by UN regions for 2011

www.thelancet.com Vol 382 August 3, 2013 435

 Series

the WHO Child Growth Standards (table 3)a However, with the increase in population in Africa, this is
35% decline from an estimated 253 million in 1990. The the only major world region with an increase in the
prevalence decreased from an estimated 40% in 1990, to number of stunted children in the past decade.
an estimated 26% in 2011an average annual rate of The complex interplay of social, economic, and
reduction of 21% per year (figure 4154). East and west political determinants of undernutrition (figure 1)
Africa, and south-central Asia have the highest preva- results in substantial inequalities between population
lence estimates in UN subregions with 42% (east subgroups. In our analysis, using previously described
Africa) and 36% (west Africa and south-central Asia); methods,158 of 79 countries with population-based
the largest number of children aected by stunting, surveys since the year 2000 (figure 5), stunting
69 million, live in south-central Asia (appendix p 9). prevalence among children younger than 5 years was
The surveys in the WHO database and other population- 247 times (range 100764) higher in the poorest
representative data were also analysed with a Bayesian quintile of households than in the richest quintile. Sex
hierarchical mixture model to estimate Z-score distri- inequalities in child nutrition tend to be substantially
butions of anthropometric indices by the Nutrition smaller than economic inequalities (appendix p 24). In
Impact Model Study (NIMS).157 These distributions were 81 countries with data, stunting prevalence is slightly
used to assess trends in stunting and underweight in higher (114 times, range 083153) in boys than in
children, the present prevalence of these measures, and girls. This finding is consistent with the higher
the present prevalence of wasting. These methods have mortality in children younger than 5 years in boys than
the advantage of estimating the full distribution of in girls in most countries in the world. Place of
anthropometric variables and therefore measure the full residence is also an important correlate of the risk of
extent of mild-to-severe undernutrition without restrictive stunting (appendix p 24). In 81 countries with data,
assumptions. They also allow for non-linear time trends. stunting was 145 times higher (range 094 to 294) in
NIMS analyses resulted in much the same estimates of rural than in urban areas.
the prevalence of stunting, underweight, and wasting According to UN estimates, globally in 2011, more than
as those of the UN in 2011 (table 3, appendix p 10). 100 million children younger than 5 years, or 16%, were
The complete trend analysis showed that the largest underweight (weight-for-age Z score [WAZ] <2 on the
reductions in stunting since 1985 have been in Asia, basis of the WHO Child Growth Standards), a 36%
whereas Africa had an increase until the mid-1990s and decrease from an estimated 159 million in 1990.154
subsequently a modest reduction in the prevalence.157 Estimated prevalences in NIMS were slightly higher at

Global Asia
300 Numbers Prevalence 60 300 60
270 2531 55 270 55
2263 Global target 50 50
240 240
2010 =100 million 45 45
Numbers (millions)

Numbers (millions)

210 210 1887

1812 40 40
Prevalence (%)

Prevalence (%)

180 1671 180 1615 984

1535 35 35
1402 1356 833
150 1274 30 150 692 30
120 25 120 1139 565 25
20 20
90 90
15 15
60 60
10 10
30 5 30 5
0 0 0 0

Africa Latin America and the Caribbean

300 60 300 60
270 55 270 55
50 50
240 240
45 45
Numbers (millions)

Numbers (millions)

210 210
40 40
Prevalence (%)

Prevalence (%)

180 35 180 35
150 30 150 30
120 25 120 25
20 20
90 90
533 558 581 599 606 15 15
60 457 479 502 60
10 10
30 5 30 137 121 104 90 74 62 5
51 42
0 0 0 0
1990 1995 2000 2005 2010 2015 2020 2025 1990 1995 2000 2005 2010 2015 2020 2025
Year Year

Figure 4: Trends in prevalence and numbers of children with stunted growth (HAZ <2), by selected UN regions and globally, 19902010, and projected to
2025 on the basis of UN prevalence estimates
HAZ=height-for-age Z score. Data from UNICEF, WHO, World Bank.154

436 www.thelancet.com Vol 382 August 3, 2013

 Series

110 million (194%).155 Prevalences were highest in south-
central Asia and western Africa where 30% and 22%,
respectively, were underweight (appendix p 9).
The UN estimate for wasting (weight-for-height Z
score [WHZ] <2 on the basis of WHO Child Growth
Standards) was 8% (52 million) globally in 2011, an 11%
decrease from an estimated 58 million in 1990.154 70% of
the worlds children with wasting live in Asia, most in
south-central Asia, where an estimated 15% (28 million)
are aected. Much the same regional pattern occurs for

Stunting (HAZ <2) Overweight (BAZ >2)

Guatemala (DHS 1998)
Timor-Leste (DHS 2009)
India (DHS 2005)
Madagascar (DHS 2005)
Laos (MICS 2006)
Niger (DHS 2006)
Nepal (DHS 2011)
Malawi (DHS 2010)
Peru (DHS 2004)
Bangladesh (DHS 2007)
Rwanda (DHS 2010)
Chad (DHS 2004)
Nigeria (DHS 2008)
Somalia (MICS 2006)
Cameroon (MICS 2006)
Mozambique (MICS 2006)
Honduras (DHS 2005)
Benin (DHS 2006)
Cambodia (DHS 2010)
Central African Republic (MICS 2006)
Ethiopia (DHS 2011)
Guinea Bissau (MICS 2006)
Tanzania (DHS 2010)
Zambia (DHS 2007)
Burkina Faso (MICS 2006)
Cte dIvoire (MICS 2006)
Democratic Republic of the Congo (DHS 2007)
Bolivia (DHS 2008)
Lesotho (DHS 2009)
Guinea (DHS 2005)
Mali (DHS 2006)
Kenya (DHS 2008)
Liberia (DHS 2007)
Uganda (DHS 2006)
Nicaragua (DHS 2001)
Gabon (DHS 2000)
Haiti (DHS 2005)
Tajikistan (MICS 2005)
Namibia (DHS 2006)
Belize (MICS 2006)
So Tom and Prncipe (DHS 2008)
Togo (MICS 2006)
Congo (Brazzaville) (DHS 2005)
Swaziland (DHS 2006)
The Gambia (MICS 2005)
Syria (MICS 2006)
Morocco (DHS 2003)
Zimbabwe (DHS 2010)
Senegal (DHS 2010)
Sierra Leone (DHS 2008)
Mauritania (MICS 2007)
Ghana (DHS 2008)
Azerbaijan (DHS 2006)
Mongolia (MICS 2005)
Guyana (DHS 2009)
Turkey (DHS 2003)
Egypt (DHS 2008)
Vanuatu (MICS 2007)
Palestinians in Lebanon (MICS 2006)
Albania (DHS 2008)
Armenia (DHS 2010)
Georgia (MICS 2005)
Maldives (DHS 2009) Figure 5: Prevalence of
Uzbekistan (MICS 2006) stunting (HAZ <2 Z scores
Kazakhstan (MICS 2006)
Thailand (MICS 2005) below median) and
Kyrgyzstan (MICS 2005) overweight (BAZ >2 Z scores
Colombia (DHS 2010) above median) for highest
Bosnia and Herzegovina (MICS 2006)
Jordan (DHS 2007) and lowest wealth quintiles
Suriname (MICS 2006) in selected countries
Dominican Republic (DHS 2007) Red circles are lowest wealth
Macedonia (MICS 2005)
Moldova (DHS 2005) quintiles, blue circles are
Montenegro (MICS 2005) highest wealth quintiles.
Serbia (MICS 2005)
Belarus (MICS 2005)
HAZ=height-for-age Z score.
Brazil (DHS 2006) DHS=Demographic and Health
Survey. MICS=Multiple
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Indicator Cluster Survey.
Prevalence (%) BAZ=body-mass index for
age Z score.

www.thelancet.com Vol 382 August 3, 2013 437

 Series
severe wasting (WHZ <3), with a global prevalence in
2011 of 3% or 19 million children. The highest percen-
tages of children with severe wasting are in south-central
Asia (51%) and central Africa (56%).
Suboptimum growth, according to anthropometric
measures indicative of stunting, wasting, and under-
weight, has been shown to increase the risk of death from
infectious diseases in childhood.159,160 This association
has been recently re-examined with the pooled analysis
of individual-level data from ten longitudinal studies
involving more than 55 000 child-years of follow-up and
1315 deaths in children younger than 5 years.161 As with
previous analyses, all degrees of stunting, wasting, and
underweight had higher mortality and the risk increased
as Z scores decreased (appendix p 11). Undernutrition
can be deemed the cause of death in a synergistic
association with infectious diseases; if the undernutrition
did not exist, the deaths would not have occurred.1 All
anthropometric measures of undernutrition were asso-
ciated with increased hazards of death from diarrhoea,
pneumonia, and measles; the association was also noted
for other infectious diseases, but not malaria. We
calculated the population attributable fractions for stunt-
ing, underweight, wasting, and its subset of severe
wasting using the UN and NIMS prevalence data with
standard methods.1 These fractions were multiplied by
the corresponding age-specific and cause-specific deaths162
to estimate the number of deaths attributable to each
anthropometric measure (table 2).162 For the percentage of
total deaths the denominator was 6934 million.131 Details
of these estimates for UN subregions and causes of death
are in appendix pp 1213. Stunting and underweight have
the highest proportions of attributed child deaths, about
14% for both; wasting accounts for 126% (severe wasting
74%) of child deaths. Table 2 and appendix pp 1415
show estimations using the NIMS prevalence data. In
these estimates stunting and underweight each account
for 17% of child deaths and wasting for 115% (severe
wasting 78%).
Stunting is a well established risk factor for poor child
development with numerous cross-sectional studies
showing associations between stunting and motor and
cognitive development. Several longitudinal studies
show stunting before age 23 years predicts poorer cog-
nitive and educational outcomes in later childhood and
adolescence.135,163 Eect sizes for the longitudinal studies
comparing children with HAZ of 2 or lower with non-
stunted children (HAZ 1) are moderate to large.163
Length-for-age Z score (LAZ) at age 2 years was con-
sistently associated with higher cognitive Z scores in
children aged 49 years (017019 per unit change LAZ)
across four cohorts with moderate (2432%) or high
(6786%) stunting prevalence.164 Associations with
underweight have also been reported.163
Stunted children show behavioural dierences in
early childhood including apathy, more negative aect,
and reduced activity, play, and exploration.165,166 The first
438
2 years of life are a crucial period linking growth and
development; growth from birth to 24 months but not
from 24 to 36 months was associated with child develop-
ment in Guatemala,167 and weight gain in the first
2 years predicted school outcomes in five cohorts.138
Analyses from the COHORTS group that are presented
here suggest that growth in the first 2 years of life, but
not at later ages, is associated with achieved school
grades in adults.168 However, some evidence suggests
that growth after 24 months of age might also be
associated with lower cognitive ability, but with a
smaller eect size than for early growth.169 In a Malawi
cohort, height gain from 18 to 60 months predicted
mathematics ability at 12 years. Height gain at 1 month
and change from 1 to 6 months and 6 to 18 months
were not significant predictors.170
We analysed changes in stunting prevalence between
1996 and 2008 in Bangladesh, Brazil, and Nigeria,
according to wealth and urban or rural status (panel 2,
figure 6).
Overweight and obesity
The prevalence of overweight worldwide and for UN
regions is based on the joint analyses done by UNICEF,
WHO, and the World Bank.154 In 2011, globally, an
estimated 43 million children younger than 5 years, or
7%, were overweight (ie, WHZ greater than two Z scores
above the median WHO standard), on the basis of the
WHO Child Growth Standards (appendix p 9)a 54%
increase from an estimated 28 million in 1990. This
trend is expected to continue and reach a prevalence of
99% in 2025 or 64 million children (figure 7). Increasing
trends in child overweight are taking place in most world
regions, not only in high-income countries, where preva-
lence is the highest (15% in 2011). However, most
overweight children younger than 5 years (32 million in
2011) live in LMICs. In Africa, the estimated prevalence
increased from 4% in 1990 to 7% in 2011, and is expected
to reach 11% in 2025 (figure 7). Prevalence of over-
weight is lower in Asia (5% in 2011), but the number of
aected children is higher compared with Africa (17 and
12 million, respectively).
Dierences in childhood overweight prevalence
between the richest and poorest quintiles are small in
most countries (figure 5), and in general prevalence
tends to be higher in the richest quintile than in the
poorest. In 78 countries with data, prevalence in the
richest quintile was on average 131 (range 055360)
times higher than in the poorest quintile. Overweight is
much the same between the sexes (appendix p 24) and
slightly more prevalent in urban than in rural areas
(appendix p 25). In 81 countries with data, urban preva-
lence was 108 times higher on average (range 044146)
than rural prevalence.
Childhood overweight results in both immediate and
longer-term risks to health. Among the immediate risks
are metabolic abnormalities including raised cholesterol,
www.thelancet.com Vol 382 August 3, 2013
 Series

triglycerides, and glucose, type 2 diabetes, and high

blood pressure.171 Childhood overweight is also a strong
risk factor for adult obesity and its consequences.2,172
Childhood vitamin deficiencies
Anaemia and iron
The percentages of children with anaemia (haemoglobin
<110 g/L) and severe anaemia (haemoglobin <70 g/L) due
to inadequate iron, (ie, anaemia that is correctable by oral
iron supplements, calculated as described earlier) are
181% and 15%, respectively. The prevalence is highest
in Africa and Asia for all IDA and in Africa for severe
IDA (table 1). However, the proportion of all childhood
anaemia corrected by iron supplementation ranges from
63% in Europe to 34% in Africa where there are other
major causes of anaemia; the proportion of severe
anaemia corrected by iron supplementation in Africa is
57% (appendix p 4).
Iron supplementation in children aged 5 years and
older with IDA generally benefits their cognition, but
studies of children younger than 3 years have had
mostly negative findings.135,173175 Most cohort and cross-
sectional studies of children younger than 3 years with
IDA find developmental deficits and studies from the
past 15 years provide evidence of neurophysiological
changes suggestive of delayed brain maturation.135
However, IDA is associated with many social dis-
advantages that also aect child development and, thus,
Stunting prevalence (%)
Panel 2: How do inequalities in stunting evolve with time?
Increased availability of survey data, including several surveys from dierent years, allowed
the analyses of time trends in nutritional indicators according to population subgroups.
Figure 6A compares trends in stunting by wealth quintile in three countries. In Nigeria,
there was almost no change in stunting prevalence from 2003 to 2008, and the degree of
inequality remained almost unchanged. In Bangladesh, stunting prevalence decreased in all
subgroups, but inequality also remained at the same magnitude. In Brazil, where prevalence
of stunting is much lower, equity improved because of a substantial decrease in stunting
the poorest populations.
Figure 6B shows the corresponding results for urban and rural dierences over time. In
Nigeria, rural prevalence was higher than urban prevalence in 2003, and both remained at
similar levels by 2008. In Bangladesh, both urban and rural rates decreased, but the gap
was reduced over time. In Brazil, where there was a two-times rural-urban gap in 1996,
full equality had been reached by 2006.
Increased data availability has led to the ability to study trends over time for subnational
groups. Such data should be used for advocacy purposes, showing which population
groups require closer attention, and also as a means to assess the eect of
nutrition-specific and nutrition-sensitive interventions, as well as of broader
developmental programmes and initiatives.
A
100 Wealth quintiles
Q1: poorest 20% Q2 Q3 Q4 Q5: richest 20%
90
80

randomised controlled trials are necessary to establish 70

a causal association. A previous systematic review 60

showed that iron supplementation resulted in a small 50
improvement in mental development scores in children 40
with IDA aged older than 7 years, but had no eect in 30
children younger than 27 months.173 To further assess 20
this scientific literature, we identified seven double- 10
blind randomised controlled trials176182 of iron lasting at
0
least 8 weeks in children younger than 4 years. Five
trials177180,181,182 showed benefits in motor development B
and two did not.176,180 Only one showed benefits in 100 Area of residence
Rural Urban
language,178 and a small study showed benefits in 90
mental development.177 In an eighth randomised 80
controlled trial,183 children given iron in infancy showed
Stunting prevalence (%)

70
no cognitive benefit when followed up at age 9 years. 60
Four additional randomised controlled trials184187 50
examined the combined eects of iron and folate
40
supplementation in children younger than 36 months
30
of age. One showed benefits to motor milestones,184
20
others showed no benefits to motor185,186 or language186
10
milestones, and one showed no benefit to cognitive
function.187 Thus, there is some evidence that iron 0
96

99

04

07

96

06

03

08

deficiency aects motor development in children

20

20
20

20

20
19
19

19

younger than 4 years, but no consistent evidence for an Bangladesh Brazil Nigeria
eect on mental development. However, many of the
supplementation trials produced only small dierences
Figure 6: Changes in stunting over time, in Bangladesh, Brazil, and Nigeria
in iron status between the treated and control groups, (A) Stunting prevalence by wealth quintile. The longer the line between
possibly limiting their ability to aect development. It two groups, the greater the inequality. (B) Stunting prevalence by urban or
is also possible that mental development takes longer to rural location.

www.thelancet.com Vol 382 August 3, 2013 439

 Series

Global Africa
65 Numbers Prevalence 637 14 65 14
60 60
552
55 12 55 12

50 478 50
10 10
45 45
412

Numbers (millions)
Numbers (millions)

40 40

Prevalence (%)
Prevalence (%)
353 8 8
35 317 35
284 296
30 30
6 6
25 25 214
20 20 173
4 4
15 15 138
110
10 10 87
2 69 2
46 56
5 5
0 0 0 0

Asia Latin America and the Caribbean

65 14 65 14
60 60
55 12 55 12

50 50
10 10
45 45
Numbers (millions)
Numbers (millions)

40 40

Prevalence (%)
Prevalence (%)

8 8
35 35
30 30
6 6
25 147 231 25
144 141 143 205
20 184 20
165 4 4
15 15
10 2 10 2
5 5 36 37 38 39 38 38 38 38

0 0 0 0
1990 1995 2000 2005 2010 2015 2020 2025 1990 1995 2000 2005 2010 2015 2020 2025
Year Year

Figure 7: Trends in prevalence and numbers of overweight (WHZ >2) children, by selected UN regions and globally, 19902010, and projected to 2025, on the
basis of UN prevalence estimates
WHZ=weight-for-height Z score.

improve than the duration of the trials or that the has declined, probably because of large-scale vitamin A
eects of iron deficiency early in life are irreversible. supplementation programmes in many countries, sub-
clinical vitamin A deficiency aects high proportions of
Vitamin A children in Africa and southeast Asia (table 2).
Clinical assessment of ocular symptoms and signs of Many randomised controlled trials have been done
xerophthalmia and biochemical assessment of serum to examine the eect of supplementation every
concentration of retinol are the two common methods 46 months and fortification on survival of children
in population surveys for estimation of prevalence of aged 6 months and older; these studies provide the
vitamin A deficiency. WHO provides prevalence esti- best evidence for deaths attributable to vitamin A
mates of vitamin A deficiency in preschool children deficiency.188191 Meta-analyses of these trials show a
(<5 years) for 19952005 from 99 countries.45 Globally, mortality reduction of 23%,188 30%,189 and 24%190 in
09% (95% CI 0315) or 517 million preschool age children aged 659 months. With publication of a large
children are estimated to have night blindness and programme eectiveness study from India, a revised
333% (311354) or 90 million to have subclinical meta-analysis shows a mortality eect of 11%, still a
vitamin A deficiency, defined as serum retinol concen- statistically significant benefit.192 In our calculations we
tration of less than 070 mol/L. Vitamin A deficiency use only the eects in the trials on particular causes of
using night blindness prevalence can be defined as a death, not the eects on overall deaths because the
global problem of mild public health importance,45 causes of death at the time of the trials and nowadays
although the prevalence in Africa (2%) is higher than are probably dierent (eg, diarrhoea and measles
elsewhere. Although prevalence of clinical symptoms account for a much smaller proportion of child deaths

440 www.thelancet.com Vol 382 August 3, 2013


now than 1020 years ago).162 To derive the risk of
vitamin A deficiency the inverse (1/risk reduction) of
the cause-specific mortality reduction identified in the
trials was deemed to be the risk of deficiency and
adjusted with the assumption that all the eect was in
the subset of the trial population with low serum retinol
(appendix p 16). Much the same adjustment was done
for the eect of vitamin A deficiency on diarrhoea
incidence (appendix p 17). This allows the adjusted RR
to be applied to the present prevalence of low serum
retinol to estimate the attributable deaths or disease
episodes. Although results of trials193195 in south Asia
show a newborn supplementation eect on mortality in
the first 6 months of life, the evidence from Africa is
less clear, and further studies are underway. Therefore,
we do not estimate deaths attributable to vitamin A
deficiency in the first 6 months of life. Child deaths
attributable to vitamin A deficiency for 2011 are
estimated to be 157 000 (table 2).
Zinc
Zinc deficiency in populations could be assessed by a
shift of the population distribution of serum zinc con-
centrations to lower values as recommended by the
International Zinc Nutrition Consultative Group;196 how-
ever, insucient data exist to classify countries or sub-
national populations. Instead the proportion of the
national population estimated to have an inadequate
zinc intake on the basis of national food availability and
dietary requirements is used.70 According to this method,
an estimated 17% of the worlds population has an
inadequate zinc intake; substantial regional variation
exists, with Asia and Africa having the highest preva-
lences (table 1). A systematic review197 showed that zinc
supplementation resulted in a 9% reduction (RR 091,
95% CI 082101) of borderline significance in all-
cause child mortality. A separate analysis of available
trials showed a significant 18% reduction (RR 082,
070096) in all-cause mortality in children aged 14
years.198 There were suggestive benefits on diarrhoea-
specific (RR 082, 95% CI 064105) and pneumonia-
specific (085, 065111) mortality.197 This and previous
analyses have included cause-specific mortality eects
even when they were not statistically significant when
the eect on all-cause mortality was statistically
significant.199,200 These trials were not powered for cause-
specific mortality eects. Supporting evidence for the
cause-specific mortality eects comes from randomised
controlled trials that showed significant reductions in
diarrhoea incidence (RR 087, 95% CI 081094) and
pneumonia incidence (081, 073090).197 To derive the
risk of zinc deficiency, we adjusted the inverse of the
cause-specific mortality (appendix p 18) and incidence
(appendix p 19) reductions noted in the trials with the
assumption that all the eect was in the subset of the
trial population at risk of zinc deficiency, as estimated
from the availability of food in national diets.70 This
www.thelancet.com Vol 382 August 3, 2013
Series
method allows the adjusted RRs to be applied to the
present prevalence of inadequate zinc intake in countries
to estimate the attributable deaths. The population
attributable fractions for diarrhoea and pneumonia were
multiplied by the number of these deaths in 2011.162 The
number of child deaths attributed to zinc deficiency in
2011 is 116 000 (table 2).
Zinc deficiency also has a small negative eect on
growth. A meta-analysis of randomised controlled trials of
zinc supplementation showed a significant benefit for
linear growth in children aged 05 years.201 The eect was a
gain of 037 cm in zinc-supplemented children. Trials that
used a dose of zinc of 10 mg per day for 24 weeks, rather
than lower doses, showed a larger benefit of 046 cm.
Breastfeeding practices
Present recommendations are that babies should be
put on the breast within 1 h after birth, be exclusively
breastfed for the first 6 months, and for an additional
18 months or longer, be breastfed along with comple-
mentary foods. There are no recently published system-
atic compilations of data for breastfeeding patterns so
we analysed data from 78 countries with surveys done
in LMICs during 200010 (appendix p 20). Early
initiation of breastfeeding (within 1 h) is highest in
Latin America (mean 58%, 95% CI 5067), intermediate
in Africa (50%, 4555) and Asia (50%, 4258), and
lowest in eastern Europe (36%, 2350). Except for
eastern Europe, where the lowest rates of breastfeeding
are recorded, globally about half of children younger
than 1 month, and three in every ten children aged
15 months are exclusively breastfed. Breastfeeding in
623 month olds is most frequent in Africa (mean 77%,
95% CI 7381) followed by Asia (62%, 5471) and Latin
America (60%, 5069), with lower occurrence in eastern
Europe (33%, 2442).
The risks of increased mortality and morbidity due to
deviation from present breastfeeding recommendations
are well documented.1 Updated systematic reviews of
these risks have results that are much the same as our
previous estimates (appendix p 21).1,202,203 The number of
child deaths attributed to suboptimum breastfeeding in
2011 is 804 000 or 116% of all deaths (table 2).
Three prospective case-cohort studies provide data for
the association of early breastfeeding initiation (within
24 h) with neonatal mortality.204 Although early initiation
was associated with lower neonatal mortality (RR 056,
95% CI 046079), in babies who were exclusively
breastfed the mortality risk was not significantly
reduced (069, 027175). The possible benefit of early
initiation of breastfeeding was therefore not deemed to
be additive to the eects of exclusive breastfeeding in
our analyses.
A systematic review shows that breastfeeding is con-
sistently associated with an increase in IQ of about three
points,205 even after adjustment for several confound-
ing factors including maternal IQ. Evidence for the
441
 Series
protection aorded by breastfeeding against risk factors
for non-communicable diseases is less consistent. A
series of meta-analyses,205 based mainly on studies in
adults from high-income settings, showed no evidence of
protection against total cholesterol levels or diastolic
blood pressure. When the meta-analysis was restricted to
high-quality studies, breastfeeding was associated on
average with a 1 mm reduction in systolic blood pressure,
and with a 12% reduction in the risk of overweight or
obesity. Studies of diabetes or glucose levels were too few
to allow a firm conclusion.
Joint eects of nutritional conditions on child
mortality
To estimate the population attributable fraction and the
number of deaths attributable to several risk factors, we
used Comparative Risk Assessment methods.206 We
calculated attributable deaths from four specific causes
of mortality (diarrhoea; measles; pneumonia; and other
infections, excluding malaria) associated with dierent
nutritional status measures. In the neonatal period all
deaths were deemed to be associated only with sub-
optimum breastfeeding and fetal growth restriction. The
Comparative Risk Assessment methods allow the
estimation of the reduction in death that would take
place if the risk factors were reduced to a minimum
theoretical level or counterfactual level of exposure.206 For
the assessment the following formula was used:
n conditional growth variables that remove the correlation
P ( RR i i 1)
i =1
PAF = n to be made about the relative association between growth
P (RR i i 1) + 1
i =1
For the purpose of this analysis, RRi equals the RR of
mortality for the ith exposure category. Pi equals the
proportion of children in the ith exposure category, for
dierent levels of fetal growth restriction, stunting,
wasting, deficiencies of vitamin A and zinc, and sub-
optimum breastfeeding practices in their relevant age
groups in specific UN regions.
Because specific causes of deaths could potentially be
caused by more than one factor, the population
attributable fractions for multiple risk factors that aect
the same disease outcome overlap and cannot be
combined by simple addition. For this reason, the joint
population attributable fraction was estimated with the
formula: Joint PAFi=1product (1PAFi), where PAFi
equals the population attributable fraction of the dierent
risk factors.207 All analyses were done for relevant age
groups of that risk factor and then the results from the
age groups were aggregated. For stunting and wasting
we did calculations separately with UN and NIMS
prevalence estimates.
The resulting 31 million deaths constitute 45% of
global deaths in children younger than 5 years in 2011,
442
attributed jointly to fetal growth restriction, suboptimum
breastfeeding, stunting, wasting, and deficiencies of
vitamin A and zinc. The overall results were the same
using UN or NIMS prevalence estimates. As part of this
total, the joint distribution of suboptimum breastfeeding
and fetal growth restriction in the neonatal period
contributes 13 million deaths or 19% of all deaths of
children younger than 5 years.
Eects of fetal and early childhood
undernutrition on adult health
In The Lancets 2008 Series on maternal and child
undernutrition,2 consequences of early childhood
nutrition on adult health and body composition were
assessed by reviewing the scientific literature and doing
meta-analyses of five birth cohorts from LMICs (India,
the Philippines, South Africa, Guatemala, Brazil), an
eort that gave rise to the COHORTS collaboration.208 On
the one hand, the conclusions were that small size at
birth and at 2 years of age (particularly height) were
associated with reduced human capital: shorter adult
height, less schooling, reduced economic productivity,
and for women, lower ospring birthweight. On the
other hand, larger child size at 24 months of age was a
risk factor for high glucose concentrations, blood
pressure, and harmful lipid levels once adjustment for
adult BMI was made, suggesting that rapid weight gain,
especially after infancy, is linked to these conditions.
The COHORTS group later did pooled analyses that use
between growth measures across ages, allowing inferences
during specific age intervals and outcomes.149 Also,
because gains in height and weight are correlated, these
analyses use conditionals that separate linear growth from
weight gain. The conclusions were that heavier birthweight
and faster linear growth from 0 to 2 years lead to large
gains in human capital, but have little association with
adult cardiovascular risk factors. Also, faster weight gain
independent of linear growth has little benefit for human
capital. After the age of 2 years, and particularly after the
age of 4 years, rapid weight gains show adverse eects on
adult cardiovascular risk. The COHORTS analyses control
for confounding and the findings are broadly much the
same across the five cohorts.
Additional longitudinal studies report eects on later
mental health with higher levels of depression and
anxiety and lower self-esteem in adolescents who were
stunted by age 2 years compared with non-stunted;209
increased depression in adolescents who had severe
acute malnutrition in infancy;210 increased risk of suicidal
ideation associated with lower HAZ at 24 months,211 and
higher levels of hyperactivity in late adolescence and
attention deficit in adults.209,212 Although observational,
the findings are persuasive and consistent with the little
quasi-experimental evidence available. Follow-up studies
of a community randomised nutrition trial in Guatemala
www.thelancet.com Vol 382 August 3, 2013

have shown long-term eects of exposure to improved
nutrition during the first 23 years, but not after 3 years,
on education,213 and wages.214 Eects of improved nutri-
tion in the first several years of life on risk factors for
chronic diseases were minor but in some cases bene-
ficial.215 Exposure to improved nutrition during child-
hood aected the growth of the next generation in girls
and their future children.2
Famines are another source of information about long-
term eects of poor nutrition in early life. The Dutch
famine of 194445 (brief, intense, but aecting a pre-
viously well-nourished population) suggests eects of
prenatal exposure on schizophrenia, no eects on human
capital (height, cognitive function), and weak and in-
consistent eects on cardiovascular risk factors.216 The
195961 Chinese famine (prolonged, severe, and aecting
an already malnourished population) suggests eects of
exposure during pregnancy and the first 2 years of life
on height, wealth, income, mental health, and inter-
generational eects on birthweight.217220 Surprisingly,
some findings suggest protective eects of famine expo-
sure, which might be explained by high mortality and
intense selection of the hardiest.219221
Discussion
Nutrition has profound eects on health throughout the
human life course and is inextricably linked with cog-
nitive and social development, especially in early child-
hood. In settings with insucient material and social
resources, children are not able to achieve their full
growth and developmental potential. Consequences
range broadly from raised rates of death from infectious
diseases and decreased learning capacity in childhood to
increased non-communicable diseases in adulthood.
Nutrition and growth in adolescence is important for a
girls health and adult stature. Women with short stature
are at risk of complications in delivery, such as obstructed
labour. Nutritional status at the time of conception and
during pregnancy is crucial for fetal growth. Babies with
fetal growth restriction, as shown by being SGA, are at
increased risk of death throughout infancy. We estimate
that 32 million babies are born SGA, 27% of births
in LMICs, and about 800 000 neonatal deaths and
400 000 post-neonatal infant deaths can be attributed to
the increased risk associated with having fetal growth
restriction. The use of more appropriate methods than
those in our previous review1 to understand the preva-
lence and risk of SGA presented here resulted in
estimates that are more than double our previous
estimate of attributable neonatal deaths related to term
low birthweight. This new finding contradicts the
widespread assumption that SGA infants, by contrast
with preterm babies, are not at a substantially increased
risk of mortality. Additionally, babies who are SGA have
an increased risk of growth faltering in the first 2 years of
life; our estimates suggest that 20% of stunting might be
attributable to fetal growth restriction. The links between
www.thelancet.com Vol 382 August 3, 2013
Series
fetal growth restriction and maternal nutritional con-
ditions, short stature, and low BMI, should lead to more
emphasis on nutritional interventions before and during
pregnancy. These interventions would have benefits for
health of adolescents and women, could reduce compli-
cations of pregnancy and delivery for the mother, and
enhance fetal growth and development.
Stunted linear growth has become the main indicator
of childhood undernutrition, because it is highly
prevalent in all developing regions of the world, and has
important consequences for health and development. It
should replace underweight as the main anthropometric
indicator for children. Underweight indices include chil-
dren who are short, but who can have an increased
WHZ, being therefore at increased risk of long-term
adverse health outcomes. Linear growth assessment in
primary care is an essential component of country eorts
to reduce childhood stunting. More experience is needed
in the operational aspects of the assessment and inter-
pretation of linear growth by health workers and in the
eective intervention responses.
Prevalence of stunting in children younger than 5 years
in developing countries in 2011 was about 28%, a
decrease from 40% or more in 1990 and the 32% estimate
in our 2008 nutrition Series for 2004.1 The number of
stunted children globally has decreased from 253 million
in 1990 to 165 million in 2011. Another reported estimate
was 167 million in developing countries for 2010.157 The
13-year Comprehensive Implementation Plan (201225)
on Maternal, Infant and Young Child Nutrition, endorsed
at the 2012 World Health Assembly, includes six global
nutrition targets, the first of which calls for a 40%
reduction of the global number of children younger than
5 years who are stunted by 2025 (compared with the
baseline of 2010).222 This goal would translate into a 39%
relative reduction per year and imply reducing the
number of stunted children from the 171 million in 2010
to about 100 million. However, at the present rate of
decline, prevalence of stunting is expected to reach 20%,
or 127 million, in 2025. In Africa, only small reductions
in prevalence are anticipated on the basis of present
trends. However, in view of the rising number of births,
the actual number of stunted children will increase from
56 to 61 million. By contrast, Asia is projected to show a
substantial decrease in stunting prevalence and in the
absolute number of children aected.
Stunted, underweight, and wasted children have an
increased risk of death from diarrhoea, pneumonia,
measles, and other infectious diseases. Recalculated
risks of cause-specific mortality161 confirmed previous
estimates1,159,160 and were used in estimation of the deaths
attributable to these conditions. The attributable fraction
of deaths for these nutritional status measures has
remained nearly the same as in previous calculations,
but the number of attributed deaths has declined because
of the decrease in the prevalence of these measures and a
decline in the aected causes of death. By our estimates
443
 Series
more than 1 million deaths can be attributed to stunting
and about 800 000 to wasting, about 60% of which are
attributable to severe wasting. These attributable deaths
cannot be added because of the overlap of these and
other nutritional conditions, but are instead included in
the calculation of the deaths attributed to the joint
eects of all nutritional conditions. Our estimate of
about 1 million child deaths due to underweight is higher
than the recently published 860 000 deaths in the Global
Burden of Disease (GBD).200 The GBD Study did not
estimate deaths attributed to stunting or wasting.
Deficiencies of vitamins and minerals have important
health consequences, both through their direct eects,
such as iron deficiency anaemia, xerophthalmia due to
vitamin A deficiency, and iodine deficiency disorders,
and because they increase the risk of serious infectious
diseases. In the latter category, vitamin A and zinc
deficiencies have been shown to have the greatest eects
among the micronutrients. Our estimate of the nearly
157 000 child deaths attributed to vitamin A deficiency is
smaller than our previous estimate (668 000).1 Notably,
we did not include the risk of mortality in the first
6 months of life, as we did last time for Asia, because of
more uncertainty about this risk; several trials of neonatal
vitamin A supplementation are underway in Asia and
Africa to assess the benefits. Likewise, our estimate of
116 000 child deaths attributed to zinc deficiency is much
smaller than the 450 000 in our previous estimates, partly
because we now estimate the risk of pneumonia and
diarrheal mortality beginning at 12 rather than 6 months
of age. In both cases there have also been changes in our
methods of estimating the prevalence of the conditions
and the risk associations, and combining these factors to
get attributed fractions of deaths. These methods are
much the same as those reported by the GBD Study.200
Our estimates are 20% higher for zinc and 30% higher
for vitamin A, than the GBD estimates. The reduction
from our estimates for the year 2004 is partly explained
by decreases in the prevalences of these deficienciesfor
vitamin A mostly because of large-scale implementation
of high-dose supplementation programmes, and a
reduction in deaths from diarrhoea and measles aected
by vitamin A deficiency and in deaths from diarrhoea
and pneumonia aected by zinc deficiency. Some of
this mortality reduction (eg, for diarrhoea) could be
attributable to the vitamin A supplementation, but might
be also related to improvement in nutritional status
measures (eg, stunting), water and sanitation, and illness
treatment, and for measles is largely attributable to
improved coverage with measles vaccine. Although the
number of deaths attributed to the present prevalence of
vitamin A deficiency is relatively small, importantly, a
reduction in present coverage of vitamin A interventions
would probably result in an increase in mortality, because
in most LMICs dietary intake is still inadequate.
Breastfeeding exclusively for the first 6 months and
then along with complementary food to 24 months of age
444
is highly beneficial. Data from developing regions show
that present practices are far from optimum, despite
improvements in some countries. Our previous estimates
suggested that 14 million deaths were attributed to
suboptimum breastfeeding, especially related to non-
exclusive breastfeeding in the first 6 months of life. Our
new estimates are 804 000 deaths, a substantial reduction
since our estimate for the year 2004. Although present
estimates of the risks associated with suboptimum
breastfeeding practices are almost unchanged from what
was used previously, the reduction in infectious disease
deaths in the past 10 years results in a lower number of
preventable deaths.173 Our number is higher than the
estimated 545 000 deaths attributed to suboptimum
breastfeeding in the GBD Study.200 Insucient methods
are provided by GBD to understand the reasons for
dierences, but one reason is probably the lower
estimates of aected diarrhoea and pneumonia deaths
than the UN estimates we used.223
To work out the total deaths attributed to nutritional
conditions, we calculated the joint distribution of
stunting, wasting, fetal growth restriction, deficiencies of
vitamin A and zinc, and suboptimum breastfeeding. The
resulting 31 million deaths constitute 45% of the
69 million global child deaths in 2011. The total number
of attributed deaths is reduced from the 35 million we
estimated for 2004, despite the population attributable
fraction increasing from 35% to 45%. In this period the
mortality in children younger than 5 years decreased
from 10 million to 69 million deaths and there were
even larger decreases in the causes of deathsuch as
measles, diarrhoea, and pneumonia162that are most
aected by nutritional conditions. The increase in the
percentage of attributed deaths compared with the
previous series is largely because of the more appropriate
inclusion of the eects of SGA (about 800 000 deaths)
instead of term low birthweight (337 000 deaths) and the
inclusion of all wasting (WHZ <2) instead of severe
wasting (WHZ <3) in the joint calculations.
The number of children aected by excessive body-
weight relative to length or height is increasing globally.
Although the prevalence of overweight in high-income
countries is more than double that in LMICs, three
quarters of the global total live in LMICs. The recorded
trends in the prevalence of early childhood overweight are
probably a consequence of changes in dietary and physical
activity patterns over time. These behavioural changes
are aected by many social and environmental factors,
including interpersonal (family, peers, and social net-
works), community (school, workplace, and institutions),
and governmental (local, state, and national policies).224
Studies show that the trend toward childhood obesity can
start as early as age 6 months.225,226 In LMICs, rapid weight
gain after 2 years of age is particularly associated with
adult fat mass.2,168 Intrauterine, infant, and preschool
periods are deemed to be possible crucial periods for
programming long-term regulation of energy balance.227229
www.thelancet.com Vol 382 August 3, 2013

If trends are not reversed, increasing rates of childhood
overweight and obesity will have vast implications not
only for future health-care expenditures but also for the
overall development of nations. These findings confirm
the need for eective interventions and programmes to
reverse anticipated trends. The early recognition of exces-
sive weight gain relative to linear growth is essential.
Routine assessment of both weight and length in all
children needs to become standard clinical practice from
very early childhood.
We have reported previously and confirmed in a new
analysis that anaemia that is reduced by supplemental
iron in pregnancy is a risk factor for more than a
quarter of maternal deaths. Anaemia is probably a
particularly important risk factor for haemorrhage, the
leading cause of maternal deaths (23% of total deaths).223
Additionally, there is now sound evidence that calcium
deficiency increases the risk of pre-eclampsia, now the
second most important cause of maternal death (19% of
total deaths).223 Thus, addressing deficiencies of these
two minerals could result in substantial reduction of
maternal deaths.
Nutrition is crucial for optimum child development
throughout the first 1000 days of life and beyond.
Maternal nutrition aects fetal growth and brain develop-
ment with clear evidence for the need to continue and
strengthen programmes to prevent maternal iodine
deficiency. Fetal growth restriction and postnatal growth
aect motor and cognitive development, with the largest
eects from stunting before age 23 years. Growth later
in childhood might also aect development though with
a smaller eect size. Iron deficiency anaemia aects
motor development; eects on cognitive development
have been more clearly shown in children older than
5 years than in younger children.
Promotion of good early nutrition is essential for
children to attain their developmental potential; however,
poor nutrition occurs with other risks for development,
in particular inadequate stimulation during early child-
hood. Interventions to promote home stimulation and
learning opportunities in addition to good nutrition will
be needed to ensure good early development and longer-
term gains in human capital, with integration and coord-
ination of programmes and policies across sectors.
Panel 3 lists research priorities. Additionally, there is a
general need for better data on micronutrient deficiencies
and other nutritional conditions at national and sub-
national levels. This work should involve the development
and use of improved biomarkers that could be used to
describe nutritional conditions and increase knowledge
about how they aect health and development. Such
information is needed to guide intervention programmes
in countries and priorities for support globally.
In the 5 years since our last series there have
been some improvements in nutritional conditions,
especially for child growth. Nonetheless, the extent of
these conditions remains high with serious detrimental
www.thelancet.com Vol 382 August 3, 2013
Series
health consequences, with a high child mortality
burden related to both stunting and wasting. New
evidence in this paper supports the focus on pregnancy
and the first 2 years of life, the crucial 1000 days, which
we called for in the previous series, but adds more
emphasis to the nutritional conditions in adolescence,
at the time of conception, and during pregnancy, as
important for maternal health and survival, fetal growth
and subsequent early childhood survival, growth, and
development. Fetal growth restriction and poor growth
early in infancy are now recognised as important
determinants of neonatal and infant mortality, stunting,
and overweight and obesity in older children and
adults. Preventive eorts should continue to focus on
the 1000 days, while therapeutic eorts continue to
target severe wasting.
Panel 3: Research priorities
How much of the eect of the underlying economic and social determinants of
optimum growth and development is mediated through known, measurable
proximate determinants?
What are the consequences of calcium or vitamin D deficiencies for maternal health,
fetal growth, birth outcomes, and infant health?
What is the importance of zinc deficiency or other micronutrient deficiencies in the
risk of preterm delivery?
What is the role of nutrition in adolescence, at the time of conception, and in
pregnancy in healthy fetal growth and development?
What factors are associated with regional variation in fetal growth restriction and do
its consequences on growth and mortality vary by setting?
What is the interaction of fetal growth restriction and post-partum infections and diet
in contributing to stunting?
What is the role of vitamin A deficiency in newborn babies or the first 6 months of life
in neonatal and infant mortality?
Are there benefits from iron supplementation for mental development in children
younger than 5 years with iron deficiency anaemia?
Do deficiencies of any micronutrients other than vitamin A and zinc increase the risk
of mortality from infectious diseases?
Can the risk of stunting be explained by specific dietary factors, micronutrient
deficiencies, and clinical or subclinical infections?
How can the rapid decreases in stunting noted in selected countries be explained by
changes in intervention coverage and in distal determinants?
Are there benefits of early initiation of breastfeeding that are independent of the
practice of exclusive breastfeeding?
How can developmental dierences associated with nutritional deficiencies at
various ages of childhood be best quantified and related to functioning and
productivity in adulthood?
What are the optimum physical growth rates in the first year of life to avoid both
undernutrition and obesity?
What are the eects of physical growth rates and stunting on cognitive abilities and
psychological functioning?
Can health workers eectively and eciently assess and interpret height-for-age
measures of linear growth in health and nutrition programmes?
How important is optimum growth and nutritional status in the fetal and early
childhood period (first 1000 days of life) in the development of adult obesity and
non-communicable diseases?
445
 Series
Contributors
REB conceptualised and coordinated the analyses, did the first draft of the
paper, responded to reviewer comments and incorporated all revisions
until publication. CGV contributed with analyses of inequalities,
breastfeeding patterns, and long-term consequences of early nutrition,
and served as a coordinator for this paper. SPW contributed sections on
the eect on child development of maternal IDA, fetal growth restriction,
and childhood undernutrition, maternal IDA, and mental health and
served as a coordinator for this paper. ZAB assisted with conceptualising
the paper and contributed sections on micronutrients and adolescent
nutrition. PC contributed to writing the sections on maternal and
childhood micronutrient deficiencies (vitamin A and iodine), mortality
and morbidity eects of maternal low BMI and short stature, and
micronutrient deficiencies and eects of fetal growth restriction on
childhood stunting. MdO contributed global and regional analyses on
childhood stunting, wasting, underweight, and overweight. ME
contributed to exposures for various child and maternal nutritional
indicators, to selected aetiological eect sizes for childhood exposures,
and provided advice on methods for calculation of attributable burden.
SG-MG reviewed the scientific literature on iron and iodine deficiency
and contributed to these sections. JK contributed analyses and text on
maternal risk factors for, prevalence of, and mortality consequences of
fetal growth restriction and the mortality eect of vitamin A
supplementation. RM contributed the sections on adolescent nutrition,
the consequences of short maternal stature, and long-term consequences
of early nutrition. RU contributed to sections on maternal obesity,
considering causes and short-term and long-term consequences to
mothers, neonates, and infants. All authors contributed to the final paper.
Maternal and Child Nutrition Study Group
Robert E Black (Johns Hopkins Bloomberg School of Public Health,
USA), Harold Alderman (International Food Policy Research Institute,
USA), Zulfiqar A Bhutta (Aga Khan University, Pakistan),
Stuart Gillespie (International Food Policy Research Institute, USA),
Lawrence Haddad (Institute of Development Studies, UK),
Susan Horton (University of Waterloo, Canada), Anna Lartey
(University of Ghana, Ghana), Venkatesh Mannar (The Micronutrient
Initiative, Canada), Marie Ruel (International Food Policy Research
Institute, USA), Cesar Victora (Universidade Federal de Pelotas, Brazil),
Susan Walker (The University of the West Indies, Jamaica),
Patrick Webb (Tufts University, USA)
Series Advisory Committee
Marc Van Ameringen (Gain Health Organization, Switzerland),
Mandana Arabi (New York Academy of Sciences, USA), Shawn Baker
(Helen Keller International, USA), Martin Bloem (United Nations
World Food Programme, Italy), Francesco Branca (WHO, Switzerland),
Leslie Elder (The World Bank, USA), Erin McLean (Canadian
International Development Agency, Canada), Carlos Monteiro
(University of So Paulo, Brazil), Robert Mwadime (Makerere School of
Public Health, Uganda), Ellen Piwoz (Bill & Melinda Gates Foundation,
USA), Werner Schultink (UNICEF, USA), Lucy Sullivan (1000 Days,
USA), Anna Taylor (Department for International Development, UK),
Derek Yach (The Vitality Group, USA). The Advisory Committee
provided advice in a meeting with Series Coordinators for each paper at
the beginning of the process to prepare the Series and in a meeting to
review and critique the draft reports.
Other contributors
Tanya Malpica-Llanos, Li Liu, and Jamie Perin assisted with analyses of
attributable deaths and Rachel White provided administrative support
for this paper and the series. Aluisio Barros, Giovanny Frana, and
Maria Clara Restrepo contributed to the analyses of inequalities and of
breastfeeding patterns. Anne CC Lee and Naoko Kozuki contributed to
analyses on fetal growth restriction. Monika Blssner and Elaine Borghi
contributed to the analyses of global and regional estimates of stunting,
wasting, underweight, and overweight. Gretchen Stevens and Yuan Lu
conducted analysis of maternal BMI; Yuan Lu analysed the studies on
the eects of vitamin A and zinc deficiencies.
Conflicts of interest
REB serves on the Boards of the Micronutrient Initiative, Vitamin
Angels, the Child Health and Nutrition Research Initiative, and the
446
Nestle Creating Shared Value Advisory Committee. VM serves on the
Nestle Creating Shared Value Advisory Committee. MdO is a sta
member of the World Health Organization. MdO alone is responsible
for the views expressed in this publication; they do not necessarily
represent the decisions or policies of the World Health Organization.
The other authors declare that they have no conflicts of interest. As
corresponding author Robert E Black states that he had full access to all
data and final responsibility for the decision to submit for publication.
Acknowledgments
Funding for the preparation of the Series was provided to the Johns
Hopkins Bloomberg School of Public Health through a grant from the
Bill & Melinda Gates Foundation. The sponsor had no role in the
analysis and interpretation of the evidence or in writing the paper and
the decision to submit for publication.
References
1 Black RE, Allen LH, Bhutta ZA, et al, for the Maternal and Child
Undernutrition Group. Maternal and child undernutrition: global
and regional exposures and health consequences. Lancet 2008;
371: 24360.
2 Victora CG, Adair L, Fall C, et al, for the Maternal and Child
Undernutrition Group. Maternal and child undernutrition:
consequences for adult health and human capital. Lancet 2008;
371: 34057.
3 Bhutta ZA, Ahmed T, Black RE, et al, for the Maternal and Child
Undernutrition Group. What works? Interventions for maternal
and child undernutrition and survival. Lancet 2008; 371: 41740.
4 Bryce J, Coitinho D, Darnton-Hill I, et al, for the Maternal and
Child Undernutrition Group. Maternal and child undernutrition:
eective action at national level. Lancet 2008; 371: 51026.
5 Morris SS, Cogill B, Uauy R. Eective international action against
undernutrition: why has it proven so dicult and what can be done
to accelerate progress? Lancet 2008; 371: 60821.
6 Maternal and Child Nutrition Study Group. Maternal and child
nutrition: building momentum for impact. Lancet 2013; published
online June 6. http://dx.doi.org/10.1016/S0140-6736(13)60988-5.
7 Cappa C, Wardlaw T, Langevin-Falcon C, et al. Progress for children:
a report card on adolescents. Lancet 2012; 379: 232325.
8 Prentice AM, Ward KA, Goldberg GR, et al. Critical windows for
nutritional interventions against stunting. Am J Clin Nutr 2013;
93: 91118.
9 UNICEF. Progress for children: a report card on maternal mortality.
New York: United Nations Childrens Fund, 2008.
10 Kozuki N, Lee AC, Vogel J, et al. The association of parity and
maternal age with small-for-gestational age, preterm and neonatal
and infant mortality: a meta-analysis. BMC Public Health (in press).
11 Gibbs CM, Wendt A, Peters S, et al. The impact of early age at first
childbirth on maternal and infant health. Paediatr Perinat Epidemiol
2012; 26 (suppl 1): 25984.
12 Gigante DP, Rasmussen KM, Victora CG. Pregnancy increases BMI
in adolescents of a population-based birth cohort. J Nutr 2005;
135: 7480.
13 Rah JH, Christian P, Shamim AA, et al. Pregnancy and lactation
hinder growth and nutritional status of adolescent girls in rural
Bangladesh. J Nutr 2008; 138: 150511.
14 Ministy of Health and Family Welfare Government of India.
International Institute for Population Sciences, Macro International.
National family health survey (NFHS-3), 200506: India: Volume I.
Mumbai: IIPS, 2007. http://www.measuredhs.com/pubs/pdf/
FRIND3/FRIND3-Vol1AndVol2.pdf (accessed May 20, 2013).
15 Ministerio de Salud Pblica y Asistencia Social, Instituto Nacional
de Estadstica, Centros de Control y Prevencin de Enfermedades.,
Encuesta Nacional de Salud Materno-infantil 2008
(ENSMI-2008/2009). Guatemala, 2010.
16 Finucane MM, Stevens GA, Cowan MJ, et al, on behalf of the
Global Burden of Metabolic Risk Factors of Chronic Diseases
Collaborating Group (Blood Glucose). National, regional, and global
trends in body-mass index since 1980: systematic analysis of health
examination surveys and epidemiological studies with 960
country-years and 91 million participants. Lancet 2011; 377: 55767.
17 Stevens GA, Singh GM, Lu Y, et al. National, regional, and global
trends in adult overweight and obesity prevalences.
Popul Health Metr 2012; 10: 22.
www.thelancet.com Vol 382 August 3, 2013
 Comment
Published Online
June 6, 2012
http://dx.doi.org/10.1016/
S0140-6736(13)61086-7
For the Lancet Series on
Maternal and Child Nutrition
see http://www.thelancet.com/
series/maternal-and-child-
nutrition
For the Scaling Up Nutrition
policy brief see http://
scalingupnutrition.org/
wp-content/uploads/pdf/
SUN_Framework.pdf
For the Lancets 2008 Series adult earnings, and economic growth. A second Series
on maternal and child
undernutrition see http://www.
thelancet.com/series/maternal-
and-child-undernutrition
For the SUN Movement Revised
Roadmap see http://
scalingupnutrition.org/wp-
content/uploads/2012/10/SUN-
Movement-Road-Map-
Septemeber-2012_en.pdf
For the 1000 days partnership
see http://www.thousanddays.
org
666
*Mark A Boyd, Basil Donovan
The Kirby Institute for infection and immunity in society,
University of New South Wales, Sydney, NSW 2052, Australia
mboyd@kirby.unsw.edu.au
MAB has received research support from AbbVie, Gilead Sciences, and Merck;
and honoraria for serving on advisory boards and speaking from AbbVie,
Boehringer Ingelheim, Bristol-Myers Squibb, Gilead Sciences, Janssen
Pharmaceuticals, and Merck. BD declares that he has no conflicts of interest.
1 Lennox JL, DeJesus E, Lazzarin A, et al; STARTMRK investigators. Safety and
ecacy of raltegravir-based versus efavirenz-based combination therapy in
treatment-naive patients with HIV-1 infection: a multicentre, double-blind
randomised controlled trial. Lancet 2009; 374: 796806.
2 Steigbigel RT, Cooper DA, Kumar PN, et al; BENCHMRK Study Teams.
Raltegravir with optimized background therapy for resistant HIV-1
infection. N Engl J Med 2008; 359: 33954.
3 Sax PE, DeJesus E, Mills A, et al, for the GS-US-236-0102 study team.
Co-formulated elvitegravir, cobicistat, emtricitabine, and tenofovir versus
co-formulated efavirenz, emtricitabine, and tenofovir for initial treatment
of HIV-1 infection: a randomised, double-blind, phase 3 trial, analysis of
results after 48 weeks. Lancet 2012; 379: 243948.
4 DeJesus E, Rockstroh JK, Henry K, et al, for the GS-236-0103 Study Team.
Co-formulated elvitegravir, cobicistat, emtricitabine, and tenofovir
disoproxil fumarate versus ritonavir-boosted atazanavir plus
co-formulated emtricitabine and tenofovir disoproxil fumarate for initial
treatment of HIV-1 infection: a randomised, double-blind, phase 3,
non-inferiority trial. Lancet 2012; 379: 242938.
Global child and maternal nutritionthe SUN rises
In April, 2010, a policy briefScaling Up Nutrition:
A Framework for Actionwas released at the spring
meetings of the World Bank and International
Monetary Fund. It was a collective eort stimulated
by the publication in January, 2008, of The Lancets
Series on undernutrition. The Lancet Series encouraged
an emphasis on the 1000 day window from the
start of pregnancy to a childs second birthday, with
interventions that are both cost-eective and yield
high returns for cognitive development, individual
on nutrition, published in The Lancet,14 now explicitly
shows that the solution to malnutrition relies on a
collective eort in which all stakeholdersgovernments,
academia, civil society, UN system organisations,
foundations, development banks, and businessescarry
out specific roles in ensuring that interventions are
delivered equitably and at scale.
The policy brief, which quickly became known as the
SUN Framework, set the stage for the transformation
that is now happening in global nutrition. It called for
country-owned nutrition strategies and programmes;
urgent scaling up of evidence-based and cost-eective
interventions; integration of nutrition within national
strategies for gender equality, agriculture, food
5 Ra F, Rachlis A, Stellbrink HJ, et al, on behalf of the SPRING-2 study group.
Once-daily dolutegravir versus raltegravir in antiretroviral-naive adults
with HIV-1 infection: 48 week results from the randomised, double-blind,
non-inferiority SPRING-2 study. Lancet 2013; 381: 73543.
6 Walmsley S, Antela A, Clumeck N, et al. Dolutegravir (DTG; S/GSK1349572)
+ abacavir/lamivudine once daily statistically superior to tenofovir/
emtricitabine/efavirenz: 48-week resultsSINGLE (ING114467).
52nd International Conference on Antimicrobials and Chemotherapy;
Sept 912, 2012; San Francisco, CA, USA. Presentation H-556b.
7 Cahn P, Pozniak AL, Mingrone H, et al, on behalf of the extended SAILING
Study Team. Dolutegravir versus raltegravir in antiretroviral-experienced,
integrase-inhibitor-naive adults with HIV: week 48 results from the
randomised, double-blind, non-inferiority SAILING study. Lancet 2013;
published online July 3. http://dx.doi.org/10.1016/S0140-6736(13)61221-0.
8 Underwood MR, Johns BA, Sato A, Martin JN, Deeks SG, Fujiwara T.
The activity of the integrase inhibitor dolutegravir against HIV-1 variants
isolated from raltegravir-treated adults. J Acquir Immune Defic Syndr 2012;
61: 297301.
9 Madruga JV, Cahn P, Grinsztejn B, et al; DUET-1 study group. Ecacy and
safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected
patients in DUET-1: 24-week results from a randomised, double-blind,
placebo-controlled trial. Lancet 2007; 370: 2938.
10 Nichols G, Grossberg R, Lazzarin A, et al. Antiviral activity of dolutegravir in
subjects with failure on an integrase inhibitor-based regimen: week 24
phase 3 results from VIKING-3. Abstracts of the 11th International Congress
on Drug Therapy in HIV Infection. J Int AIDS Soc 2012; 15 (suppl 4): 18112.
security, social protection, education, water supply,
sanitation, and health care; and a substantial increase
in domestic support and external assistance for
nutrition within the food security, social protection,
and health sectors. The SUN Road Map, prepared later
in 2010 and revised in 2012, set out ways for a wide
range of groups to work together in sharpening, scaling
up, and aligning their responses to peoples nutritional
needsand achieving results.
Alongside the 1000 days advocacy partnership,
the SUN Movement was launched at the UN General
Assembly in September, 2010. It takes forward the SUN
Road Map by encouraging coherence and eectiveness
among all groups working for better nutrition; it is not
an initiative, project, or programme. By April, 2013,
35 countries had joined the SUN Movement with
commitments that are in line with the SUN Framework
and Road Map. These countries nutrition solutions
show the commitments of political leaders, whole-of-
society responses, careful tracking of progress, and the
benefits of shared experience.
The second Lancet nutrition Series provides a range
of valuable insights as the SUN Movement moves
through 2013, a year dense with events that will
move nutrition to the heart of the development
www.thelancet.com Vol 382 August 24, 2013

agenda. It calls for a substantial increase of political
commitment in responding to the complex causes of
undernutrition. It recognises that the SUN Movement
has the potential to harness such change and yield
durable results.
The first paper1 of the Series leaves no doubt as to
why nutrition is key for sustainable development
and the wellbeing of entire populations. The second
paper2 strengthens the arguments of the 2008 Lancet
Series by bringing additional evidence to support the
focus on ensuring that all women, girls, and young
children are able to access specific interventions of
good quality; they should be included in mainstream
eorts for public health, family planning, and water
and sanitation. The third paper3 sets the foundation
for evidence-based research into achieving outcomes
through nutrition-sensitive strategies in four key
areas: agriculture and food, social security, early child
development, and classroom education. It draws on
the experience of countries that have made great
progress when stressing that gender and social
equality are the cornerstones of nutritional success.
By focusing on the political context for eective
action, the fourth paper4 recalls that the realisation of
human rights, a commitment to equity, and gender
equality should be properly prioritised. It includes
important proposals for ways in which business can
best be engaged (and the challenges of doing so), and
encourages increased involvement of civil society at all
levels. It likewise underlines the need for governments
to increase their own accountability for ensuring
that people are able to achieve good nutrition and
to ensure the existence of a fair and transparent
framework for regulating any entity that mighteven
unwittinglyundermine nutritional justice.
At the May, 2012, World Health Assembly, govern-
ment representatives agreed ambitious goals for reduc-
tion of all forms of malnutrition, including obesity.
This Lancet Series points out that these goals can only
be achieved through improved nutrition governance,
more human resources, better demonstration of
results, and increased investments from domestic
and international sources. In the past 3 years the SUN
Movement has provided a platform to enable leaders
to pledge to intensify eorts for improved nutrition.
www.thelancet.com Vol 382 August 24, 2013
Comment
Bloomberg via Getty Images
2013 provides a once-in-a-lifetime opportunity to
strengthen worldwide resolve for improved nutrition,
through commitments being made in a series of
international and regional events.
The publication of the second Lancet Series is
timely and reinforces the urgency for transformation
of political commitment into actions that lead to
improvements in nutrition. The Series brings scientific
rigour to the challenge of equitable delivery of eective
services at scaleboth now and in the years to come.
David Nabarro
SUN Movement Secretariat, Villa La Pelouse (2nd Floor), Palais Des
Nations, 1201 Geneva, Switzerland
david.nabarro@undp.org
I declare that I have no conflicts of interest.
1 Black RE, Victora CG, Walker SP, et al, and the Maternal and Child Nutrition
Study Group. Maternal and child undernutrition and overweight in
low-income and middle-income countries. Lancet 2013; published online
June 6. http://dx.doi.org/10.1016/S0140-6736(13)60937-X.
2 Bhutta ZA, Das JK, Rizvi A, et al, The Lancet Nutrition Interventions Review
Group, and the Maternal and Child Nutrition Study Group. Evidence-based
interventions for improvement of maternal and child nutrition: what can
be done and at what cost? Lancet 2013; published online June 6. http://
dx.doi.org/10.1016/S0140-6736(13)60996-4.
3 Ruel MT, Alderman H, and the Maternal and Child Nutrition Study Group.
Nutrition-sensitive interventions and programmes: how can they help to
accelerate progress in improving maternal and child nutrition?
Lancet 2013; published online June 6. http://dx.doi.org/10.1016/
S0140-6736(13)60843-0.
4 Gillespie S, Haddad L, Mannar V, Menon P, Nisbett N, and the Maternal and
Child Nutrition Study Group. The politics of reducing malnutrition:
building commitment and accelerating progress. Lancet 2013; published
online June 6. http://dx.doi.org/10.1016/S0140-6736(13)60842-9.
667
 Series

Maternal and Child Nutrition 3

Nutrition-sensitive interventions and programmes: how can
they help to accelerate progress in improving maternal and
child nutrition?
Marie T Ruel, Harold Alderman, and the Maternal and Child Nutrition Study Group*

Lancet 2013; 382: 53651 Acceleration of progress in nutrition will require eective, large-scale nutrition-sensitive programmes that address
Published Online key underlying determinants of nutrition and enhance the coverage and eectiveness of nutrition-specific
June 6, 2013 interventions. We reviewed evidence of nutritional eects of programmes in four sectorsagriculture, social safety
http://dx.doi.org/10.1016/
nets, early child development, and schooling. The need for investments to boost agricultural production, keep prices
S0140-6736(13)60843-0
low, and increase incomes is undisputable; targeted agricultural programmes can complement these investments by
This online publication has
been corrected. The corrected supporting livelihoods, enhancing access to diverse diets in poor populations, and fostering womens empowerment.
version first appeared at However, evidence of the nutritional eect of agricultural programmes is inconclusiveexcept for vitamin A from
thelancet.com on June 20, 2013 biofortification of orange sweet potatoeslargely because of poor quality evaluations. Social safety nets currently
This is the third in a Series of provide cash or food transfers to a billion poor people and victims of shocks (eg, natural disasters). Individual studies
four papers about maternal and show some eects on younger children exposed for longer durations, but weaknesses in nutrition goals and actions,
child nutrition
and poor service quality probably explain the scarcity of overall nutritional benefits. Combined early child development
*Members listed at end of paper
and nutrition interventions show promising additive or synergistic eects on child developmentand in some cases
Poverty, Health and Nutrition
nutritionand could lead to substantial gains in cost, eciency, and eectiveness, but these programmes have yet to
Division, International Food
Policy Research Institute, be tested at scale. Parental schooling is strongly associated with child nutrition, and the eectiveness of emerging
Washington, DC, USA school nutrition education programmes needs to be tested. Many of the programmes reviewed were not originally
(M T Ruel PhD, H Alderman PhD) designed to improve nutrition yet have great potential to do so. Ways to enhance programme nutrition-sensitivity
Correspondence to: include: improve targeting; use conditions to stimulate participation; strengthen nutrition goals and actions; and
Dr Marie T Ruel, Poverty, Health
optimise womens nutrition, time, physical and mental health, and empowerment. Nutrition-sensitive programmes
and Nutrition Division,
International Food Policy can help scale up nutrition-specific interventions and create a stimulating environment in which young children can
Research Institute, Washington, grow and develop to their full potential.
DC 20006, USA
m.ruel@cgiar.org
Introduction increasing incomes, and urbanisationshifts that raise
The food system is threatened by food and oil price concerns about diet quality and food safety, while
volatility, diversion of resources from production of food threatening water, land, and other finite natural
to biofuels, climate change and related water short- resources.58 In view of these challenges, protection of
ages, persistent conflicts and emergencies, and natural nutrition, let alone acceleration of progress, will entail
disasters aecting agriculture production and yields.14 more than bringing nutrition-specific interventions to
These challenges are compounded by changes in demand scale. It will require a new and more aggressive focus on
for food that are brought about by growing populations, coupling eective nutrition-specific interventions (ie,
those that address the immediate determinants of
Key messages nutrition) with nutrition-sensitive programmes that
address the underlying causes of undernutrition
Nutrition-sensitive interventions and programmes in agriculture, social safety nets, (panel 19,10).
early child development, and education have enormous potential to enhance the scale Nutrition-sensitive programmes draw on comple-
and eectiveness of nutrition-specific interventions; improving nutrition can also help mentary sectors such as agriculture, health, social pro-
nutrition-sensitive programmes achieve their own goals. tection, early child development, education, and water
Targeted agricultural programmes and social safety nets can have a large role in and sanitation to aect the underlying determinants of
mitigation of potentially negative eects of global changes and man-made and nutrition, including poverty; food insecurity; scarcity of
environmental shocks, in supporting livelihoods, food security, diet quality, and access to adequate care resources; and to health, water,
womens empowerment, and in achieving scale and high coverage of nutritionally and sanitation services.11 Key features that make pro-
at-risk households and individuals. grammes in these sectors potentially nutrition-sensitive
Evidence of the eectiveness of targeted agricultural programmes on maternal and are: they address crucial underlying determinants of
child nutrition, with the exception of vitamin A, is limited; strengthening of nutrition nutrition; they are often implemented at large scale and
goals and actions and rigorous eectiveness assessments are needed. can be eective at reaching poor populations12 who
(Continues on next page) have high malnutrition rates; and they can be leveraged
to serve as delivery platforms for nutrition-specific

536 www.thelancet.com Vol 382 August 10, 2013


interventions. Nutrition-sensitive programmes might
therefore help to accelerate progress in improving
nutrition by enhancing the household and community
environment in which children develop and grow, and
by increasing the eectiveness, coverage, and scale of
nutrition-specific interventions.
Nutrition-sensitive programmes can help protect poor
populations from the negative consequences of global
food security threats and mitigate the eects of financial,
weather-related, and man-made shocks (eg, conflicts).
Such shocks make poor populations increasingly vul-
nerable to undernutrition, as shown by food and fuel
price crises in the past 6 years,4 and documented eects
of conflicts on morbidity and mortality among aected
populations.13,14 Climate change and the expected
increased frequency of droughts and flooding are likely
to reduce food availability and dietary diversity, and
increase rates of infectious diseases such as diarrhoea or
malaria.15 Under these circumstances, nutrition-sensitive
programmes can help to protect the assets and welfare of
poor people and their investments in the health, nutri-
tion, and development of their children.
Nutrition-sensitive programmes are likely to aect
nutrition through changes in food and non-food prices
and income, and through womens empowerment.
Panel 216,17 and figures 118,19 and 218,19 show results of
analyses of the links between income growth and
maternal and child anthropometry and anaemia
(appendix p 1). Appendix p 2 summarises evidence
regarding the association between womens empower-
ment and child nutrition.
We review evidence of the nutritional eect of pro-
grammes from dierent sectors, and discuss how such
investments could be made more nutrition-sensitive.
We selected sectors on the basis of: relevance for
nutrition (eg, address crucial underlying determinants
of nutrition); availability of assessments of their
nutritional eect; high coverage of poor populations;
and targeting (programmes are, or could be, targeted to
reach nutritionally vulnerable groups). The two sectors
that most closely meet these criteria are agriculture and
social safety nets. Early child development programmes
do not meet the high coverage criteria but they are
included because child development and nutrition
outcomes share many of the same risk factors, and
there is a growing interest in examination of potential
integration and synergies in programming and
outcomes.20,21 Schooling is also included, despite failing
to meet all criteria, because of the importance of
parental education for child nutrition and development.
Health, water and sanitation, and family planning are
covered in the accompanying report by Zulfiqar Bhutta
and colleagues.22 Investments and policies in several
other sectors (eg, transportation; communication and
information technology; and global food, agriculture,
and trade) have the potential to aect nutrition, as do
more targeted policies (eg, maternity leave); however,
www.thelancet.com Vol 382 August 10, 2013
Series
(Continued from previous page)
The feasibility and eectiveness of biofortified vitamin A-rich orange sweet potato
for increasing maternal and child vitamin A intake and status has been shown;
evidence of the eectiveness of biofortification continues to grow for other
micronutrient and crop combinations.
Social safety nets are a powerful poverty reduction instrument, but their potential
to benefit maternal and child nutrition and development is yet to be unleashed; to
do so, programme nutrition goals and interventions, and quality of services need to
be strengthened.
Combinations of nutrition and early child development interventions can have
additive or synergistic eects on child development, and in some cases, nutrition
outcomes. Integration of stimulation and nutrition interventions makes sense
programmatically and could save cost and enhance benefits for both nutrition and
development outcomes.
Parental schooling is consistently associated with improved nutrition outcomes and
schools provide an opportunity, so far untapped, to include nutrition in school
curricula for prevention and treatment of undernutrition or obesity.
Maternal depression is an important determinant of suboptimum caregiving and
health-seeking behaviours and is associated with poor nutrition and child development
outcomes; interventions to address this problem should be integrated in
nutrition-sensitive programmes.
Nutrition-sensitive programmes oer a unique opportunity to reach girls during
preconception and possibly to achieve scale, either through school-linked conditions
and interventions or home-based programmes.
The nutrition-sensitivity of programmes can be enhanced by improving targeting; using
conditions; integrating strong nutrition goals and actions; and focusing on improving
womens physical and mental health, nutrition, time allocation, and empowerment.
See Online for appendix
Panel 1: Definition of nutrition-specific and nutrition-sensitive interventions
and programmes
Nutrition-specific interventions and programmes
Interventions or programmes that address the immediate determinants of fetal and
child nutrition and developmentadequate food and nutrient intake, feeding,
caregiving and parenting practices, and low burden of infectious diseases
Examples: adolescent, preconception, and maternal health and nutrition; maternal
dietary or micronutrient supplementation; promotion of optimum breastfeeding;
complementary feeding and responsive feeding practices and stimulation; dietary
supplementation; diversification and micronutrient supplementation or fortification for
children; treatment of severe acute malnutrition; disease prevention and management;
nutrition in emergencies
Nutrition-sensitive interventions and programmes
Interventions or programmes that address the underlying determinants of fetal and
child nutrition and developmentfood security; adequate caregiving resources at
the maternal, household and community levels; and access to health services and a
safe and hygienic environmentand incorporate specific nutrition goals and actions
Nutrition-sensitive programmes can serve as delivery platforms for nutrition-specific
interventions, potentially increasing their scale, coverage, and eectiveness
Examples: agriculture and food security; social safety nets; early child development;
maternal mental health; womens empowerment; child protection; schooling; water,
sanitation, and hygiene; health and family planning services
Adapted from Scaling Up Nutrition9 and Shekar and colleagues, 2013.10
537
 Series
Panel 2: How responsive is nutrition to income growth?
As economies grow, stunting rates typically decrease, but the predicted decrease is far
slower than the corresponding poverty reduction associated with economic growth
(figure 1). Country fixed-eects regressions show that a 100% increase in gross domestic
production (GDP) per person predicts a 59% (95% CI 4176) reduction in stunting and an
110% (86134) decrease in the World Banks poverty measure of individuals living on
$125 per person, per day. The eect of growth in gross national product on nutrition
comes from a combination of increased household resources, and improved infrastructure
and nutrition-relevant services. Much unexplained variability exists in the eect of national
income on stunting. As shown in figure 1, countries such as Guatemala, South Africa, and
India have higher stunting rates than expected for their income levels. By contrast, the
Dominican Republic, Senegal, Ghana, China, and Sri Lanka are among the best performers.
The association between prevalence of child underweight and GDP growth is stronger than
for stunting, with the rate of decrease with 100% GDP growth being 70% (95% CI 5388;
appendix p 1). This estimate is larger than reported with earlier datasets.16 Anaemia
defined as haemoglobin concentrations below 109 g/Ldecreases at a slower rate; a
100% improvement in income would decrease child anaemia by only 24% (1336) and
maternal anaemia by 18% (0431). However, severe anaemiadefined as haemoglobin
below 70 g/Ldecreases at a much higher rate with income growth for both mothers (65%;
95% CI 4288) and children (90%; 51129).17 Although data for low birthweights are not
as reliable as those for other nutritional indicators, estimates using World Bank data suggest
that a 100% increase in GDP per person typically reduces low birthweight prevalence by only
23% (95% CI 0841). Bangladesh, India, Sudan, and Haiti have particularly high rates of low
birthweight prevalence relative to their national levels of income. For women underweight
(body-mass index <185 kg/m2), a 100% growth in national income results in a 40%
(95% CI 1758) decrease in underweight prevalence, a rate substantially lower than the
reduction in child underweight.
The association between national income growth and women overweight and obesity is
much stronger than for women underweight: a 100% increase in GDP per person is
estimated to increase prevalence of overweight and obesity in women by 70% (95% CI
40100; figure 2). These findings show that when GDP per person increases, prevalence
of womens overweight or obesity increases faster than prevalence of women
underweight decreases. Again, some countries are clear outliers, with Egypt and several
Latin American countries having very high levels and several Asian countries (eg,
Vietnam, India, China, Thailand) having lower than expected levels of overweight and
obesity in view of their GDP per person.
we excluded these sectors because of the absence of
assessments of nutritional eects.
Consistent with the Maternal and Child Nutrition
Series, we focus on adolescent girls and women,
infants, and young children during the first 1000 days
of life (period from conception to a childs second
birthday). Interventions to improve nutrition and child
development during this period have high rates of
return because of their importance in enhancing
economic productivity later in life23 fostered by a
combination of improved health, nutrition, and
cognition, which lead to more schooling, higher-paying
jobs, and overall enhancement of physical, cognitive,
and reproductive performance.24
The programmes we reviewed generally have several
objectives, including improving income, food security,
womens empowerment, and nutrition. For this reason
538
cost-eectiveness studies cannot be easily applied to
assess or rank these programmes. Similarly, although
costbenefit analysis can be used in principle, this
analysis needs a common metric for all outputs, generally
in monetary terms. However, a conversion of a death
averted into monetary values requires an arbitrary
assessment of the value of premature deaths averted.
Similarly, although equity is usually deemed socially
desirable, its value cannot be easily quantified.25
Therefore, the nutrition outcomes in the programmes
we discuss cannot be directly compared with those in the
accompanying report by Zulfiqar Bhutta and colleagues.22
However, as we explain, the programmes we review are
an integral component of an overall strategy to improve
global nutrition.
Agriculture
Agriculture systems have a crucial role in provision of
food, livelihoods, and income.1 Agriculture is the main
occupation of 80% of poor populations in rural areas,
including women. In Africa, women account for 70% of
agricultural labour and 80% of food processing labour.26,27
Growing concerns about how to meet the food needs of
an estimated global population of 9 billion by 2050 have
spurred renewed eorts to boost agriculture production
and productivity in the face of increasing threats that
aect the global food system. Agriculture growth has
been shown to reduce undernutrition;28 an additional
investment of US$8 billion per year globally would reduce
the number of underweight children by 10 million and of
hungry people by 201 million by 2050, and raise the
income of many of the worlds poorest people.28 Moreover,
the economic returns to investments in agriculture are
high compared with many other economic investments.29
Although investments to enhance agriculture produc-
tivity and boost global food supply are crucial for long-
term reductions in poverty, hunger, and malnutrition,
they might not solve the problem of scarcity of access to
nutritious and diverse diets (as opposed to scarcity of
calories) that poor people face. A new emphasis on
making agricultural systems and food and agriculture
policies more nutrition-sensitive is called for and several
reports discuss approaches and instruments to do so.1,3033
An approach that can complement eorts to raise agri-
cultural productivity and food supply globally is targeted
agricultural programmes aimed at enhancing poor
households income and access to high-quality diets.
Our review focuses on these types of programmes, most
specifically homestead food production systems, and the
biofortification of staple crops, because they both meet
our selection criteria, with the exception of scale.
Targeted agricultural programmes can aect nutrition
through several pathways (panel 334,35). Despite variations
in the way researchers use these pathways, all concur
that womentheir social status, empowerment, con-
trol over resources, time allocation, and health and
nutritional statusare key mediators in the pathways
www.thelancet.com Vol 382 August 10, 2013
 Series

between agriculture inputs, intra-household resource

allocation, and child nutrition.26,30,3438
Burundi
The recognised importance of development of new Afghanistan
60
approaches to stimulate agricultures contribution to nutri- Niger Guatemala
Yemen
tion has led to an increased interest in examination of the MDG

Stunting prevalence in children aged 05 years

ETH Angola
Nepal
so far untapped potential of leveraging value chains to SDN
BGD
improve nutrition. Since this approach is still at a nascent DRC PAK India
stage, experience and evidence of eectiveness are scarce. 40 MMR NGA
VTN
Panel 43946 provides a brief overview of the approach. CIV
KEN PHN Peru South Africa
Syria Botswana
Haiti IDN
Home gardens and homestead food production systems Ghana IRQ Egypt Ecuador
Predicted stunting line
Several scientific literature reviews of homestead food Senegal China
20
production systems have been done in the past Kyrgyzstan UzbekistanSri Lanka
Algeria Kazakhstan
Thailand
decade.34,38,4752 These reviews focused on dierent types of Turkey
Moldova
programmes and nutritional outcomes and used dierent Colombia
DOM
Brazil
search strategies and inclusion and exclusion criteria. Predicted poverty line
Despite these dierences, key conclusions were largely 0
Belarus

consistent across all reviews (appendix pp 34). First, these
reviews note that there is little evidence of eectiveness of
homestead food production programmes on maternal or
child nutritional status (anthropometry or micronutrient
status), with the possible exception of vitamin A status. For
child anthropometry, a few studies reported an eect on at
least one indicator,5357 but eects were generally small.
Although meta-analysis might not be the method of choice
for synthesis of evidence from such diverse programmes,
the results of a four-study meta-analysis52 showed no
overall eect of targeted agricultural programmes on
underweight, wasting, or stunting. Another four-study
meta-analysis51 for vitamin A status, however, reports a
small overall dierence in serum retinol between
intervention and control areas (008 mol/L); a cluster-
randomised eectiveness assessment of a biofortified
orange sweet potato intervention in Uganda also showed a
95 percentage point reduction in the prevalence of low
0 2000 4000 6000 8000 10 000
GDP per person (2005 international $)
Figure 1: Prevalence of stunting in children aged 05 years and GDP per person
Most observations for prevalence of stunting are from 200008. The fitted curves are locally weighted regressions
of prevalence of stunting in children aged 05 years and poverty (<$125 per person, per day), against GDP per
person. The adjustment to international dollar units converts income expressed in nominal dollars to one that is
expressed in terms of international dollars, which have the same estimated purchasing power as a dollar in the
USA, accounting for local prices. The size of the circles represents the estimated population of stunted children
aged 05 years, in about 2005, on the basis of multiplication of stunting prevalence by UN estimates of the
population of children aged 05 years. Data are sourced principally from the Demographic and Health Surveys,18
with observations for some countries sourced from WHO.19 GDP=gross domestic product. BGD=Bangladesh.
CIV=Cte dIvoire. DOM=Dominican Republic. DRC=Democratic Republic of the Congo. ETH=Ethiopia.
IDN=Indonesia. IRQ=Iraq. MDG=Madagascar. MMR=Myanmar (Burma). KEN=Kenya. NGA=Nigeria. PAK=Pakistan.
PHN=Philippines. SDN=Sudan. VTN=Vietnam.
80
Egypt
Jordan Bosnia-Herzegovina
Women overweight prevalence, BMI>25 (%)

serum retinol (<105 mol/L) in intervention compared 60

with control children aged 35 years at baseline.58 The Swaziland Peru Serbia
Bolivia
second consistent message is that nutritional eect is Nicaragua South Africa
Turkey
more likely when agriculture interventions target women Lesotho Honduras
Guatemala

and include womens empowerment activities, such as 40 Morocco

Brazil
Kyrgyzstan Dominican
improvement in their knowledge and skills through Moldova Republic
Ghana Armenia
behaviour-change communications or promotion of their Cameroon
Sierra Thailand
increased control over income from the sale of targeted Leone Kenya
Philippines
Nigeria
commodities. No studies, however, have specifically com- 20
Indonesia China
Pakistan
pared targeting of men versus women, or mainstreaming
gender versus not doing so in the programmes reviewed. India
Congo
The third key message is that, with the exception of two (DRC)
Bangladesh
Ethiopia Vietnam
studies of biofortified orange sweet potato,58,59 impact 0
0 2000 4000 6000 8000 10 000
evaluation studies have generally been too poor and GDP per person (2005 PPP$)
sample sizes often too small to draw definite conclusions
Figure 2: Prevalence of women overweight (BMI>25) and GDP per person, for low-income and
about eects on nutritional status.
middle-income countries
One review,38 which specifically looked at eects of Most observations for prevalence of women overweight are from 200010. The fitted curve is a locally weighted
homestead food production systems on intermediary regression of prevalence of women overweight against GDP per person. The correlation between prevalence of
outcomes along the impact pathway, concluded that, women underweight and the log of GDP per person is 071 and is significant at the 1% level. The size of the circles
represents the estimated population of overweight women aged 1549 years, in about 2005, on the basis of
when measured, positive eects are shown for several multiplication of prevalence of women overweight by the UN population estimates of the female population aged
underlying determinants of nutrition, including house- 1549 years. Data are sourced principally from the Demographic and Health Surveys18 and WHO.19 DRC=Democratic
hold production and consumption, maternal and child Republic of the Congo. GDP=gross domestic product. PPP=purchasing power parity.

www.thelancet.com Vol 382 August 10, 2013 539

 Series
Panel 3: Pathways by which agriculture can aect nutrition outcomes
As a source of food: increases household availability and access to food from own
production
As a source of income: increases income from wages earned by agricultural workers or
through the marketing of agriculture commodities produced
Food prices: agricultural policies (national and global) aect a range of supply and
demand factors that establish the price of marketed food and non-food crops; this
price in turn, aects the income of net seller households, the purchasing power of net
buyers, and the budget choices of both
Womens social status and empowerment: womens participation in agriculture can
aect their access to, or control over, resources and assets, and increase their decision-
making power regarding intra-household allocation of food, health, and care
Womens time: womens participation in agriculture can aect their time allocation income, control over resources, or influence in decision
and the balance between time spent in income generating activities and time
allocated to household management and maintenance, caregiving, and leisure
Womens own health and nutritional status: womens participation in agriculture can
aect their health (eg, through exposure to agriculture-associated diseases) and
nutritional requirements (eg, through increased energy expenditure); their health and
nutritional status can, in turn, aect their agricultural productivity and hence their
income from agriculture
Adapted from the World Bank34 and Gillespie and colleagues, 201235
Panel 4: Value chains for nutrition
Food supply chains are defined as the series of processes and actors that take a food
from its productionincluding inputs into productionto consumption and disposal
as waste.41 Broadly defined steps along the supply chain include production, processing,
distribution, retailing, promotion, labelling, and consumption. The concept of value
chain refers to the addition of value (usually economic) for chain actors at dierent
steps along the chain.
In the past 5 years, value chains have been singled out as one potential strategy to leverage
agriculture to improve nutrition.3941 The approach could be particularly relevant for
traditional value chains for micronutrient-rich foods such as dairy, meat, fish, poultry, and
fruits and vegetables, which are generally lacking in the diets of low-income households
because of scarce availability, perishability, and high prices often compounded by a scarcity
of information and knowledge about their health and nutritional benefits. Food value
chains are therefore a possible entry point to stimulate both supply and demand
(especially among poor populations) for micronutrient-rich foods.
Value chain concepts and analysis have unique features that make them a promising
approach for tackling both undernutrition and overnutrition: 1) they focus on
coordination between actors, because all value chain processes and actors are tightly
linked by each action aecting the others along the chain; 2) they are analytical, versatile,
and solution-oriented and can therefore be used to assess the constraints that aect
availability, aordability, acceptability, or quality of nutritious foods in a given context,
and identify and test solutions that can be implemented at specific leverage points along
the chain; 3) they focus on addition of economic value, and could therefore be used to
identify points before, during, and after production at which nutritional (and economic)
value could be added, or losses in nutrients prevented. In view of the importance of
coordination across sectors and of development of joint solutions to stimulate agriculture
and nutrition linkages, value chain concepts and analysis might provide a useful
framework and platform to achieve these goals.
(Continues on next page)
540
intake of target foods and micronutrients, and overall
dietary diversity. This finding is consistent with results
from an impact-pathway focused assessment of a home-
stead food production system in Cambodia, which
showed no eect on child anthropometry or anaemia
despite eects on household production, consumption,
and dietary diversity.60
Despite explicit targeting of women in many agri-
cultural programmes, few studies have measured
specific aspects of womens empowerment as a pathway
to improved nutrition, and results are mixed. Assess-
ments of homestead food production systems in
Bangladesh and Nepal report positive eects on womens
making on a range of issues.53,6163 In Kenya, a project
promoting orange sweet potato production among
women farmers showed that women gained control over
selling the product, whereas men maintained control
over income.64 Livestock and dairy projects in Kenya and
Bangladesh report increases in womens income or
influence in decision making,65,66 whereas in India,
mens but not womens income improved as a result of a
dairy project.67 Very few studies have measured the eect
of agriculture interventions on womens time, know-
ledge, practices, health, or nutritional status and none
have modelled the potential mediating role of these
maternal resources on child nutrition.38
Biofortification
Biofortification is a uniquely nutrition-sensitive agriculture
intervention because it focuses on breeding of staple crops
that are rich in essential micronutrients.68 The many
advantages of the approach are well documented.68,69 Bio-
fortification, however, cannot achieve the high con-
centrations of micronutrients needed to treat severe
deficiencies or to fulfil the high requirements (eg, for iron
and zinc) of pregnant and lactating women and infants; it
is more suited for provision of a daily dose of micronutrients
(about 50% or more of daily needs) to help prevent
deficiencies in individuals throughout the lifecycle, outside
of the 1000 days window. As is true for all approaches,
biofortification should be considered as one component of
a larger strategy to eliminate micronutrient deficiencies,
and the optimum mix of supplementation, dietary diversi-
fication, fortification, biofortification, and health services
should be defined depending on local context.
Three broad milestones need to be achieved for bio-
fortification to succeed: 1) breeding objectives (minimum
target concentration for each micronutrient) must be met;
2) retention and bioavailability of micronutrients must be
satisfactory so that intake leads to expected improvements
in status; and 3) farmer adoption rates and intakes by
target populations must be adequate. HarvestPlus, a
programme that has led a global eort to breed and
disseminate biofortified staple food crops since 2003, has
made substantial progress in research to test these three
steps for vitamin A, zinc, and iron in seven crops: cassava,
www.thelancet.com Vol 382 August 10, 2013

maize, sweet potato, bean, pearl millet, rice, and
wheat (table). In addition to achieving major progress in
breeding and releasing crops, the programme will
complete all planned retention and bioavailability studies
in 2013, and 11 ecacy trials in 2014. Two eectiveness
trials,58,59 which assessed the rollout of orange sweet potato
in Uganda and Mozambique (milestone three), have been
completed. They showed high farmer adoption and sig-
nificant increases in vitamin A intakes in both countries
and in child vitamin A status in Uganda.58,59 Eectiveness
trials for the other target crops are expected to be
completed by 2018.
Thus, present evidence regarding biofortification is
concentrated on the first two milestonesproof of
concept that breeding for micronutrient-rich crops is
feasible and that micronutrients are retained and bio-
availableand a growing evidence of ecacy. Results on
bioconversion of carotene to retinol in humans and
bioavailability of zinc and iron from biofortified com-
pared with common varieties are very encouraging,
suggesting that extra minerals will lead to net increases
in quantities absorbed. Ecacy studies also confirm that
intakes of iron-biofortified rice70 and beans71,72 improve
iron status, and that all biofortified crops released so far
have favourable agronomic qualities, including equal or
higher yields than common varieties, and greater disease
resistance and drought tolerance. Evidence regarding the
eectiveness of biofortification, however, is still confined
to vitamin A in orange sweet potato, and the scalability of
delivery is yet to be shown.
Social safety nets
Social safety nets are programmes that distribute
transfers to low-income households. These programmes
raise income among vulnerable groups and enhance
resilience by preventing destitution brought about by
loss of assets or reduced investment in human capital
during times of crises. Transfers can be in the form of
cash or food, although with improved technology for
tracking income transfers, cash transfers are increasingly
the preferred means to support chronically poor house-
holds. Between 075 and 10 billion people in low-
income and middle-income countries currently receive
cash support.73 Although many transfer programmes
reach only a small share of the vulnerable population,
some have extensive coverage, such as Ethiopias
Productive Safety Net Programme, which reaches 10% of
the countrys population,74 and transfer programmes in
Brazil and Mexico that reach 25%, and in Ecuador 40%,
of their populations.75 The generosity of transfers varies
widely, ranging from transfers that increase total income
marginally to those that boost income by up to a third for
the poorest recipients.75
The main goal of social transfers is to augment income,
but programmes sometimes include additional inter-
ventions or conditions that can enhance their nutrition
sensitivity such as: linking of transfers to health and
www.thelancet.com Vol 382 August 10, 2013
Series
(Continued from previous page)
Value chains also have important limitations. First, they focus on one food at a time, as
opposed to the whole diet and the many nutrients required for healthy living. Eorts to
integrate nutrition into value chains should therefore focus on complementary value
chains to fill the specific dietary and nutrient gaps identified in target populations. Second,
the focus on addition of economic value plus incorporation of nutrition goals might create
insurmountable trade-os for value chains actors. Third, although they might be
well-suited to enhance access to micronutrient-rich foods for girls and women during the
reproductive period, their role for addressing the special needs of young children might be
limited to fortified complementary foods or products, and a few target foods such as dairy
products and biofortified crops (eg, biofortified orange sweet potatoes).
Case studies4143 and a review of on-going programmes44 suggest that several research
initiatives and value chain actors are currently exploring the potential of value chains to
improve nutrition. One such initiative is homegrown school feeding programmes, which
use value chains to link agriculture and nutrition, with potential livelihood and income
benefits for farmers and nutrition benefits for young children and their families.45,46
Existing eorts to incorporate nutrition in value chains should also consider addressing
food safety issues, especially since most of the micronutrient-rich foods of interest are
also highly perishable and susceptible to food safety problems. Tackling of food-borne
diseases would improve nutrition.
nutrition services (eg, through conditionality); targeting For more on the HarvestPlus
of households with nutritionally vulnerable members, on programme see http://www.
harvestplus.org
the basis of age or physiological status; inclusion of
nutrition-specific interventions for selected individuals
within the household (eg, nutrition behaviour-change
communications or distribution of fortified foods or
supplements); administration of transfers in a sex-
sensitive manner (eg, by directing transfers to women or
designing them to accommodate time constraints of
caregivers); and targeting of populations facing climatic
or economic stress related to seasonality or other shocks,
or focusing on emergencies.
Conditional cash transfers
Conditional cash transfers aim to stimulate households
to invest in the health, nutrition, and education of their
children (enhancing human capital) by promotion of the
use of these services as conditions (conditionalities) for
receipt of transfer. Most conditional cash transfers target
transfers to women, on the premise that increasing
womens control over resources will lead to greater
investments in children (appendix p 2). Although
conditional cash transfers are implemented worldwide,
experimental evidence of eectiveness comes mostly
from Latin America. In addition to their positive eects
on poverty reduction, household food consumption, and
dietary diversity,76,77 almost all programmes assessed
increased the use of preventive and curative health and
nutrition services.78,79 The Mexico, Brazil, and Nicaragua
programmes8083 also showed improvements in womens
control over additional resources, enhanced self-esteem,
heightened knowledge and awareness of health and
nutrition, and increased opportunities for women to
strengthen their social networks.
541
 Series
Country (year of first release)*
Vitamin A crops (released)
Cassava Nigeria, Democratic Republic of Congo (2012)
Maize Nigeria, Zambia (2012)
Orange sweet potato Uganda (2007), Mozambique (2002)
Iron crops (released)
Bean Rwanda (2012)
Pearl Millet India (2012)
Zinc crops (under developmentto be released in 2013)
Rice Bangladesh and India (2013)
Wheat India and Pakistan (2013)
References are provided in appendix pp 56. *Approved for release by National Governments after intensive multi-location testing for agronomic traits and micronutrient
performance. Completed though not necessarily reported.
Table: Release schedule for biofortified crops and status of related nutrition studies
Despite the many benefits of conditional cash
transfers for households and women, evidence of
eects on nutritional outcomes is mixed.76,78 A review77
using pooled estimates shows that, overall, conditional
cash transfers have had a small, but not statistically
significant, eect on child anthropometry. A forest plot
analysis of 15 programmes,77 combining conditional
cash transfers and unconditional cash transfers, shows
an average eect of 004 in height-for-age Z score, an
eect size that is neither statistically significant nor
biologically meaningful; similarly, no significant eect
was identified for conditional cash transfers only. In
view of the heterogeneity of populations and pro-
gramme designs and methods, meta-analyses might
not be the most appropriate approach for assessment of
eect, but analyses of individual studies are consistent
with the findings. Only a few conditional cash transfer
studies show eects on anthropometry, and these
eects are shown in the youngest or poorest children, or
those exposed to the programme for long durations.76,77,84
Evidence of eects on micronutrient nutrition is equally
scant and comes from only a few studies that have
looked at these outcomes.76 The Mexico conditional
cash transfer programme, which distributed a
micronutrient-fortified food to beneficiary mothers and
children, showed a positive eect on child intake of
iron, zinc, and vitamin A among those who consumed
the product, but only a small eect on mean
haemoglobin or anaemia reduction.76 Two other
programmes, in Honduras and Nicaragua, that assessed
eect on haemoglobin showed no eect.76 The Mexico
programme showed reductions in low birthweight
attributed to changes in womens empowerment, which
in turn were attributed to womens increased demand
for better quality prenatal care as a result of participation
in the programme.85 Evidence is also emerging of small
eects of conditional cash transfers on child develop-
ment outcomes.86
542
Status of nutrition studies
Dietary intake Bio-availability Ecacy Eectiveness
and retention
P P 201314 201315
P P Continuing 201315
P P P P
P P Continuing
P P P 201315
P 2013 201314 201416
P P 201314 201416
School feeding programmes
School feeding programmes are a type of conditional
transfer, albeit in kind. Similar to other transfers, they
are mainly a form of social assistance for consumption.
The links to nutrition are less direct than transfers
targeted to mothers and children during the first
1000 days, but school feeding can reduce hunger and
stimulate learning.87 These programmes, however, are
implemented in nearly every country in the world.88
Results from a meta-analysis show that school feeding
programmes have small eects on school-age childrens
anthropometry, particularly in low-income settings.89
Major eects on height are not expected in school-age
children and weight gains can be either positive (in
underweight populations) or negative (when risks of
obesity are high). In middle-income countries, school
meals might also serve as an opportunity to combat
obesity; Brazil and Chile have redesigned their pro-
grammes with this risk in mind.90
School meals might also benefit other members of the
household when the food provided is shared or when the
school-aged childs intake at home is reduced.91,92
Randomised controlled trials in Burkina Faso and
Uganda showed eects on weight among preschool-
aged boys (ie, <5 years) whose siblings received school
meals or take-home rations compared with a control
group93 (Gilligan D, International Food Policy Research
Institute, personal communication). Another oppor-
tunity oered by school feeding programmes is to aect
iron nutrition, especially for adolescent girls. A review92
of randomised evaluations of iron-rich school meals
(fortified or providing animal-source foods) documents
that three of four studies improved iron status, irre-
spective of initial status. The addition of a micronutrient
mix to school meals in India improved total body iron,
but not anaemia, possibly because of worm loads.94
Deworming can, however, be included as part of a larger
school health programme, although the timing of
www.thelancet.com Vol 382 August 10, 2013

delivery diers from the daily meal programme. Since
school-aged children are the main reservoir of worm
loads in a population, such an intervention could benefit
younger children as well.
Unconditional transfers
Unconditional transfers, either as cash or in kind, have
also been popular, particularly outside of Latin America.
Households commonly spend more on food and health
with cash transferseven when they are only indirectly
linked to nutrition and healththan they spend out of
other increases in income.95,96 Moreover, some uncon-
ditional cash transfers use so-called soft conditions in the
form of broadly targeted behaviour-change communi-
cations or social marketing to encourage health-seeking
behaviour. In Africa, soft conditions or unconditional cash
transfers are more common than conditional cash
transfers.97 One randomised trial showed that a con-
ditional cash transfer with health conditionality increased
clinic visits in Burkina Faso, whereas an unconditional
cash transfer did not,98 showing the importance of the
condition for achieving health-seeking behaviour changes
in this setting. Evidence, however, shows an absence of
overall eect of both unconditional cash transfers and
conditional cash transfers on child nutritional status.77
Costs will dier between the two approaches as will the
distribution of benefits for dierent outcomes; as such, no
method dominates in all situations.
In-kind household food distributions
In-kind household food distributions are currently less
prominent than they were in previous decades, mostly
because of cost considerations. They are now largely used
as part of an emergency response or in places where the
logistics of cash transfers are constrained. Evidence from
Mexico suggests that in-kind food transfer programmes
might have unintended eects on overweight and obesity
when the energy contribution of the food basket exceeds
the energy gap in the targeted population (panel 5,99,100
figure 3). In addition to general family rations, food
distribution programmes often provide micronutrient-
fortified foods (eg, corn-soy or wheat-soy blend) to
mothers and young children. In Haiti, distribution of
such food rations to all mothers and children within the
first 1000 days of life had a greater eect on child growth
than did targeting of underweight children younger than
5 years.101 In view of the severity of food insecurity in this
population, no unintended eects on overweight or
obesity were identified. Complementation of this pro-
gramme with the distribution of iron-fortified micro-
nutrient powders reduced anaemia prevalence by half in
as little as 2 months.102
Transfer programmes in emergencies
Transfer programmes in emergencies also usually
combine nutritionally enhanced complementary foods for
pregnant and lactating women and their young child with
www.thelancet.com Vol 382 August 10, 2013
Series
Panel 5: Unintended, negative eects of in-kind and cash transfers in Mexico
Social safety nets can reduce poverty and increase use of health and education services.79
Depending on their target populations, however, these interventions can have unintended
negative consequences. The eects of Mexicos Programa de Apoyo Alimentario (PAL; Food
Support Programme) on excess weight gain in women is one such example. PAL is a dual
conditional cash and in-kind transfer programme targeted to poor and remote
communities in rural Mexico. An assessment of PAL99 showed improved household dietary
quality, but also increases in total energy consumption in a population that was not
energy-deficient and had a high prevalence of overweight and obesity among women at
baseline (65%). PAL also increased the already steep annual weight gain in adult women in
the control group (425 g, SD 80) by 291 g (111) per year in the food basket group
(a 68% increase) and by 222 g (122) per year in the cash group (a 52% increase). The most
substantial eect was recorded in adult women who were already obese before the
programme started (518 g [153] per year in the food basket group and 354 [169] per year
in the cash group; figure 3).100 The PAL programmes food basket included several
energy-dense staple and basic food products including oil, cookies, and whole milk, and
provided an additional 450 kcal per day per adult. To avoid negative eects on populations
experiencing the double burden of child stunting and adult obesity, transfer programmes
should be designed to respond to the identified needs of target populations, and for food
transfers, their specific nutrient gaps.
7000 Food basket Cash
Estimated eect on annual weight gain (g)
6000
5000
*
*
4000
3000
2000
1000
0
All women <25 25 to 30 >30
BMI group
Figure 3: Estimated eect of Mexicos Programa de Apoyo Alimentario
programme on annual weight gain in women, by initial BMI
BMI=body-mass index. *p<005. p<001.
family rations or cash. Disasters, particularly sudden
onset emergencies such as earthquakes and hurricanes,
often disrupt normal market channels, which might
dampen the logistical advantages of cash compared with
food transfers. Although food aid deliveries overall
declined from 15 million metric tonnes (t) in 1999 to
41 million t in 2011, emergency deliveries have remained
almost constant; they now account for more than 67% of
total food aid.103 Even when targeted towards overall
household subsistence, aid during disasters can prevent
major deteriorations in child undernutrition.104,105 Age-
based targeting of fortified foods can help prevent under-
nutrition and complement eorts to tackle cases of severe
acute malnutrition with specially formulated products.
The nutritional eects of emergency deliveries can be
543
 Series
enhanced by inclusion of lipid-based nutrient supplements
in the package of assistance to families.106108
Early child development
Stunting and impaired cognitive development share
several of the same risk factors, including deficiencies in
protein, energy, and some micronutrients, intrauterine
growth retardation, and social and economic conditions,
such as maternal depression and poverty.109 Some of the
key phases of brain growth and development also
encompass the first 1000 days of life, the period of peak
susceptibility to nutritional insults. Therefore, some key
interventions can protect children from both nutritional
and developmental risks; these include core maternal
and child nutrition interventions, psychosocial stimu-
lation and responsive parenting, and interventions to
alleviate poverty, food insecurity, maternal depression,
and gender inequity.21
Evidence of the eect of early child development inter-
ventions, with or without a nutrition component, on child
development outcomes has been extensively reviewed in
two previous series in The Lancet.20,21 We focus on evidence
of how child stimulation and nutrition interventions can
have complementary eects on nutrition outcomes
(appendix pp 79). The most comprehensive, long-term
study110 of interventions that provided both child stimu-
lation and food supplementation to stunted children aged
924 months in Jamaica showed an additive eect of the
two interventions on cognitive development, but not on
growth. At adolescence, the additive eects on cognition
were not sustained, but the group having received
stimulation had long-term benefits, which ranged from
improved development outcomes to educational attain-
ment and social behaviour.111,112 In Bangladesh, addition of
stimulation and home visits to standard nutrition and
health care for severely malnourished children improved
development outcomes and weight-for-age Z score
(WAZ).113 Another trial in Bangladesh, which added
responsive parenting (including feeding) to an informal
nutrition and child development education programme,
showed benefits on several feeding and parenting
behaviours, child self-feeding, and development out-
comes; addition of iron-fortified micronutrient powders to
the intervention improved weight gain and WAZ but had
no additional eect on development outcomes.114 A zinc
supplementation and responsive stimulation intervention
in underweight children in Jamaica showed synergistic
eects on child development between the two interven-
tions: greater benefits on development outcomes were
identified in the zinc and stimulation group, compared
with little or no eect in the groups receiving either.115 This
synergistic eect, however, was not noted for morbidity
(reduced only in zinc group) or growth (no eect in either
group). A randomised controlled trial of cash transfers to
households linked to preschool enrolment in Uganda
provides an example of joint benefits on cognition and
nutrition; findings showed significant eects of cash
544
transfers on child cognitive development, which were
mediated by increased preschool participation, improved
diets, and reduced anaemia (Gilligan D, Roy S, Inter-
national Food Policy Research Institute, personal
communication).
Other trialsin India, Pakistan, and Bangladesh116118
although successful at improving child development or
nutrition outcomes, or both, failed to show additive or
synergistic eects between nutrition and stimulation
interventions. In India,116 beneficiaries of the Integrated
Child Development Services (ICDS) programme were
allocated to groups receiving breastfeeding and comple-
mentary feeding counselling, or this package plus
responsive feeding and psychosocial stimulation skills.
Compared with ICDS only, both intervention packages
improved child dietary intake and haemoglobin and
reduced morbidity, but only the nutrition intervention
increased length gain, and only the full package including
stimulation benefited development outcomes.116 A factorial
design trial in Pakistan117 showed no evidence of additive
or synergistic eects of a nutrition (counselling and
micronutrient powders) and stimulation intervention
(monthly group meetings and home visits for children
aged 024 months) on child development or nutrition
outcomes. Preliminary results show eects of all three
intervention packages on developmental scores compared
with control, with larger eect sizes among the two
stimulation groups. In Bangladesh,118 psychosocial
stimulation with or without food supplements among
severely underweight children aged 624 months on
discharge from hospital had an eect on mental
development and a small eect on WAZ, but no additive
or synergistic eects were noted between the two
interventions. One intervention that closely ties feeding
practices with child stimulation is responsive feeding.
Few studies of this approach, however, have been designed
to distinguish messaging on complementary feeding
from those on psychosocial care114 and few so far have
shown a clear association with nutritional outcomes.119
Reduction of maternal depression is another way to
address risk factors common to both nutrition and child
development.120,121 Eorts are being made to link basic
health services with a wide range of social support for
women. Ecacy trials.122,123 show that the benefits might
accrue to both mothers and their newborn babies,
justifying ongoing eorts to bring these initiatives to scale.
Schooling
Although children are beyond the crucial 1000 days
window when they enter school, their schooling
experience might be a strong determinant of the nutrition
of the next generation. Parental schooling has been
consistently associated with child nutritional status, with
maternal education often, but not always, having a larger
explanatory power than paternal education, controlling
for income and schooling choices.124,125 The positive global
trends in schooling are, therefore, encouraging for
www.thelancet.com Vol 382 August 10, 2013

nutrition. Data from developing countries show an
average increase in years of schooling from 260 to 762
for boys and from 150 to 664 for girls between 1950
and 2010.126 The girl to boy ratio shows a substantial
improvement over this period, from 577% to 859%.
Enrolment data show parity in girls primary schooling in
most countries;127 moreover, in many countries more
girls than boys are now in secondary school. Still, only
about a fifth of adolescent girls in sub-Saharan Africa
and two fifths of girls in south Asia are enrolled in
secondary education.128
We assessed the level of parental education necessary
for a meaningful reduction in child undernutrition by
analysing 19 datasets from the Demographic and Health
Survey (collected since 1999) and derived estimates of the
risk of child stunting associated with maternal and
paternal primary and secondary education, controlling for
household wealth, rural versus urban residence, and child
age and sex.The analysis showed that the risk of stunting
is significantly lower among mothers with at least some
primary schooling (odds ratio [OR] 089, 95% CI
085093), and even lower (p<0001) among mothers nutritional contribution of programmes in four sectors
with some secondary schooling (075, 071079). Paternal and the potential for enhancing their nutrition-sensitivity.
education at both the primary and secondary levels also
reduced the risk of stunting although the respective ORs
(096, 093101; and 085, 081089) are smaller than nutrition-sensitive programmes have intensified in the
for maternal schooling. Despite this overall association,
there is appreciable heterogeneity in eect sizes for both
maternal and paternal education in individual countries,
probably indicative of dierences in both quality of
education and quality of data.
Schooling directly increases individual earnings and
national income and, through these pathways, can aect
nutrition in the long term. Thus, programmes to
increase schooling via the supply of inputs or through
fee waivers or cash transfers can be expected to reduce
the risk of undernutrition for the next generation. There
is a lack of clarity, however, about which aspects of
schooling, beyond the income eect, benefit nutrition.
At least five overlapping pathways have been suggested,
but not formally tested. Schooling might: 1) transmit
information about health and nutrition directly; 2) teach
numeracy and literacy, thereby assisting caregivers in
acquiring information and possibly nutrition know-
ledge;129,130 3) expose individuals to new environments,
making them receptive to modern medicine; 4) impart
self-confidence, which enhances womens roles in
decision making, and their interactions with health-care
professionals; and 5) provide women with the
opportunity to form social networks, which can be of
particular importance in isolated rural areas. The
question remains as to whether schooling could do even
more to directly aect nutrition, both in the short term
for school children and in the long term as they
transition into their parental role. Although nutrition
modules are available in some school health education
programmes,131,132 assessments of the eect of a school
www.thelancet.com Vol 382 August 10, 2013
Series
health and nutrition curriculum in developing countries
on undernutrition or health knowledge, let alone on
parenting skills decades later, are absent.
Schools are also suitable venues to introduce pro-
grammes to combat obesity. Such programmes can focus
on healthy diets and promotion of physical activity. A
systematic review133 of 22 studies in low-income and
middle-income countries noted that 82% of such
programmes had a favourable eect on physical activity,
diet, or both.
Discussion
In 2008, The Lancet Maternal and Child Undernutrition
Series included conditional cash transfer programmes
and dietary diversification approaches as general
nutrition support strategies, and noted small positive
eects of conditional cash transfers on child anthrop-
ometry in three Latin American countries, and an
absence of a statistically significant of eect of dietary
diversification strategies on child nutrition outcomes.134
In the present series, we discuss evidence regarding the
Although the concept of nutrition sensitivity is not new,
investments in development and implementation of
past few years, prompted by the 2008 series, and
spearheaded by the Scaling Up Nutrition movement.9 It For more on the Scaling Up
is important to recognise, when interpreting the results Nutrition movement see http://
scalingupnutrition.org
of our review, that most of the programmes included
were retrofitted and tagged as nutrition sensitive without
having been originally designed as such.
Targeted agricultural programmes have an important
role in supporting livelihoods, improving household
food security and healthy diets, and in fostering
womens empowerment. Yet, our review shows
inconclusive evidence of eects on child nutritional
status, with the possible exception of benefits on
vitamin A intake and, to a lesser extent, vitamin A
status. These findings are probably the result of a
combination of factors, including: weaknesses in
programme design and implementation (especially the
nutrition, behaviour-change communication, and
health components);60 inclusion of households with
children outside of the 1000 days window with little
potential to benefit in linear growth; and the fact that
other pressing constraints to nutritionsuch as infec-
tious diseases, helminths, and environmental enter-
opathy associated with scarcity of access to appropriate
water, sanitation, and hygienemight not be addressed
by the programmes. Additionally, the assessments of
most of the programmes we reviewed had crucial
weaknesses such as an absence of valid comparison
and control groups, a possibly too-short duration of
intervention, small sample sizes, the inclusion of the
wrong age group in eectiveness assessments, and the
failure to control for potential confounding factors in
545
 Series
Panel 6: Research priorities
Rigorous, theory-based eectiveness and cost-eectiveness assessments of complex
and large scale nutrition-sensitive programmes. These assessments should include:
Use of experimental randomised controlled trials, where feasible, to test dierent
methods of delivery and joint packages of interventions.
Careful assessment of programme impact pathways and quality of service delivery,
use of process evaluation instruments and mixed methods, including assessment
of the capacity and eciency of front-line health workers.
Measurement of gender-disaggregated impact indicators; these indicators should
be carefully selected on the basis of nutrition goals and interventions included in
programmes and could include: anthropometry, micronutrient status biomarkers,
child development outcomes, child morbidity.
Measurement of intermediary outcomes along the impact pathway (eg, household
consumption; food security and dietary diversity; dimensions of womens
empowerment, maternal physical, and mental health; detailed dietary intake or
simpler measures such as dietary diversity for target individuals).
Detailed costing for assessment of cost-eectiveness.
Development of methods to allow comparison of the social benefits of complex
programmes with many objectives and joint outcomes, with the benefits of
single-outcome programmes.
Formative research and focused ethnographic studies to guide selection, design, and
implementation of nutrition interventions to be integrated in nutrition-sensitive
programmes, and for overall design of nutrition-sensitive programmes.
Qualitative research to understand barriers to participation, adoption and use of
programme inputs and services (eg, agricultural inputs, compliance with conditions in
conditional transfers, recommended feeding or caregiving practices).
Research to rigorously test the feasibility and desirability of integration of
interventions from several sectors versus co-location. Such research would establish
whether programme implementers should develop new instruments and methods for
joint planning, implementation, monitoring, and assessment, or whether
investments should focus on eective programme co-location and implementation.
Research to test and document the scalability of newly released biofortified crops.
Assessment of eectiveness large-scale programmes combining early child
development and nutrition interventions in dierent contexts and assessment of
synergies both in programming and outcomes.
Research to test dierent delivery platforms for programmes to reduce maternal
depression.
Research to test dierent delivery systems for reaching adolescent girls (eg, school
programmes, social safety nets with conditions to keep girls in school, agricultural
programmes targeting adolescent girls at home).
Assessments of school nutrition programmes and their short term eect on
knowledge of school children and long-term eect on parenting skills.
the analysis. All these assessment design flaws reduce
the ability to detect an eect even if one exists.
In view of the complexity and diversity of agricultural
programmes, their many goals, and their long impact
pathways, some argue that impact assessments should
focus on outcomes such as food security and diet quality,
rather than child nutritional status during the 1000 days
window. Although complexity is a valid concern, the
many pathways by which agriculture can improve
nutritionand evidence that eects on several indicators
along these pathways are achievedsupport giving these
programmes a fair chance to deliver on child nutrition
546
outcomes. Future work should include testing of pro-
grammes with stronger designs, nutrition goals, and
interventions; use of rigorous programme-theory based
impact and impact pathway assessments; and assessment
of cost and cost-eectiveness (panel 6).
Biofortification has made substantial progress by
establishing proof of concept for orange sweet potato and
showing eects, through rigorous assessments, on
maternal and child intake of vitamin A and child vitamin
A status. The challenges for biofortification now rest in
showing eectiveness of new crops, refining delivery and
marketing strategies, scaling up successfully, and
integrating new varieties into national agricultural
research systems.
Social safety nets are a powerful way to reduce poverty,
and currently provide income support to a billion
chronically poor individuals and to shock victims. They
have been shown to improve household food availability
and dietary quality and to foster certain aspects of womens
empowerment, and for conditional cash transfers in
particular, to stimulate demand for health and education
services. Despite these many benefits, pooled evidence
shows little eect on nutritional outcomes. By contrast
with agriculture programmes, several eectiveness assess-
ments of social safety nets have used rigorous randomised
controlled trial designs, although most of these studies
were done in middle-income countries and might under-
estimate the magnitude of eect that could be achieved in
poorer settings. Additionally, some aspects of programme
design or implementation might have diluted their nutri-
tional eect, including poor timing and short duration of
maternal and child exposure, absence of clear nutrition
goals, and poor selection or implementation of nutrition
interventions in some programmes. Also, the gap between
increased use of health and nutrition services and
nutrition benefits has been attributed, at least partly, to the
poor quality of services provided.78 Conditional cash
transfers are designed to increase health awareness and
service demand, but ultimately their nutritional eect
rests on the quality of public health services.
Our review of early child development interventions
provides little evidence that stimulation alone has a direct
eect on nutrition outcomes, but it suggests that
combined early child development and nutrition inter-
ventions can have additive or synergistic eects on
development outcomes, and in some cases on nutrition.
The examples of successful joint delivery of these services
point to an area in which programmatic synergies, cost
savings, and potential benefits for both child development
and nutrition might be identified.
Girls schooling is increasing in many countries, largely
as a result of government interventions to change incen-
tives and reduce barriers to girls school enrolment and
participation. Increases in parental schooling have con-
tributed to reductions in stunting, but larger eects could
probably be achieved if eective nutrition education
programmes were incorporated into school curricula.
www.thelancet.com Vol 382 August 10, 2013

Our review shows the potential of programmes in the
four sectors reviewed to improve the lives of poor
households and individuals, both in the short-term and the
long-term. It also shows, however, that more needs to be
done to increase the nutrition sensitivity of programmes
so that their potential to deliver on maternal and child
nutrition outcomes is unleashed. The nutrition sensitivity
of promising programmes can be enhanced in several
ways. First, targeting on the basis of nutritional
vulnerability (eg, age, physiological status) in addition to
geographic targeting on the basis of poverty, food
insecurity, or location can help reach households and
individuals most likely to benefit from the programme.
Alternatively, targeting of nutritionally vulnerable indi-
viduals could be used as a second level of targeting for
subgroups of programme beneficiaries who meet pre-
established criteria. For example, targeting agricultural
programmes to households with pregnant or lactating
women or children younger than 2 years might be neither
logistically feasible nor optimum for community develop-
ment; however, geographic community-level targeting
could be used as a first targeting criteria, with a second
level focusing on reaching mothers and young children
with a specific package of preventive nutrition and health
interventions. Another key target group for nutrition-
sensitive programmes is adolescent girls; conditions or
other incentives can be used to keep girls in school, help
delay first pregnancy, address HIV risk factors,135 and
improve adolescent girls' nutrition knowledge and micro-
nutrient status to prepare them for motherhood.
Second, evidence shows that nutrition improvements
are not automatic even with programmes that are success-
ful at reducing poverty, food insecurity, and sex inequalities.
To reach their full potential, programmes such as those
reviewed need careful identification of nutrition goals and
appropriate design and eective implementation of
interventions to achieve them. A third way to enhance the
nutrition sensitivity of programmes is to engage women
and include interventions to protect and promote their
nutritional wellbeing, physical and mental health, social
status, decision making, and their overall empowerment
and ability to manage their time, resources, and assets. A
fourth promising, yet underused approach to enhance the
nutrition sensitivity of programmes is to use them as
delivery platforms for various nutrition-specific inter-
ventions. Nutrition behaviour-change communications,
which are incorporated in several agriculture, social safety
nets, early child development, and school health pro-
grammes are one example of such use. Other opportunities
include addition of the distribution of micronutrient-
fortified products to nutritionally vulnerable adolescent
girls, mothers, and young children,101,136 or of preventive
health inputs to agriculture, social safety net, early child
development, or school programmes.
Finally, a crucial question that remains to be addressed
when designing nutrition-sensitive programmes is the
degree to which programmes should indeed integrate
www.thelancet.com Vol 382 August 10, 2013
Series
actions from several sectors, or co-locate programmes
managed by dierent sectors so that they reach and
saturate the same communities, households, and indi-
viduals. In view of the complexity of integration,33,137,138
especially across many sectors, it is important to carefully
assess whether investments should focus on joint
planning, implementation, monitoring, and assessment,
or on eective programme co-location (panel 6).
Conclusions
Nutrition-sensitive programmes hold great promise for
supporting nutrition improvements and boosting the
scale, coverage, and benefits of nutrition-specific actions.
New incentives are needed to support innovations in
nutrition-sensitive programmes and unleash their
potential to tackle nutrition while also achieving their own
goals. New nutrition-sensitive agriculture44 and social
safety net programme designs, methods, and packages of
interventions are being tested and are strengthening links
with health services. Rigorous impact evaluations, many
of which are based on strong programme-theory and
impact pathway analysis, are addressing key weaknesses
encountered in previous evaluations and are assessing
impacts on a range of nutrition and child development
outcomes and several household and gender outcomes
along the impact pathway. Evidence generated by these
enhanced programmes and assessments in the next
510 years will be of crucial importance to inform future
investments in agriculture and social safety net pro-
grammes to improve nutrition.
The potential benefits of integration of early child
development and nutrition programming include cost
savings and gains in both child development and nutrition
outcomes. Leveraging health, agriculture, or social safety
net platforms for joint early child development and
nutrition programming during the first 1000 days of life
would help focus on the crucial period of peak vulnerability
for both nutrition and development. Current work is
exploring such approaches. Benefits from psychosocial
interventions on cognition, however, extend well beyond
the first 2 years, and therefore, continued child develop-
ment support is required throughout the entire preschool
period. Early child development programmes, possibly
linked to conditions in transfer programmes or delivered
through preschool or community settings, could oer
psychosocial stimulation and parenting interventions,
while also providing relevant nutrition interventions for
children 2 years and older, focusing on micronutrients,
healthy diets, and obesity prevention. With improved
guidance and curricula for nutrition education in schools,
a new emphasis on using schools to improve nutrition
knowledge and practices and preparing school children
for their future parenting roles should also emerge.
The immense potential of programmes addressing
the underlying determinants of undernutrition to
complement and enhance the eectiveness of nutrition-
specific interventions is real, but is yet to be unleashed.
547
 Series
Investments in nutrition-sensitive programmes can
have a pivotal role in prevention of the excess stunting,
wasting, and impaired child development that the scale-
up of nutrition-specific interventions cannot resolve on
its own.
Contributors
MTR and HA conceptualised the report, reviewed the literature, and
wrote all drafts of the report.
Maternal and Child Nutrition Study Group
Robert E Black (Johns Hopkins Bloomberg School of Public Health,
USA), Harold Alderman (International Food Policy Research Institute,
USA), Zulfiqar A Bhutta (Aga Khan University, Pakistan),
Stuart Gillespie (International Food Policy Research Institute, USA),
Lawrence Haddad (Institute of Development Studies, UK),
Susan Horton (University of Waterloo, Canada), Anna Lartey
(University of Ghana, Ghana), Venkatesh Mannar (The Micronutrient
Initiative, Canada), Marie Ruel (International Food Policy Research
Institute, USA), Cesar Victora (Universidad de Federal de Pelotas,
Brazil), Susan Walker (The University of the West Indies, Jamaica),
Patrick Webb (Tufts University, USA).
Series Advisory Committee
Marc Van Ameringen (Gain Health Organization, Switzerland),
Mandana Arabi (New York Academy of Sciences, USA), Shawn Baker
(Helen Keller International, USA), Martin Bloem (United Nations World
Food Programme, Italy), Francesco Branca (WHO, Switzerland),
Leslie Elder (The World Bank, USA), Erin McLean (Canadian
International Development Agency, Canada), Carlos Monteiro
(University of So Paulo, Brazil), Robert Mwadime (Makerere School of
Public Health, Uganda), Ellen Piwoz (Bill & Melinda Gates Foundation,
USA), Werner Schultink (UNICEF, USA), Lucy Sullivan (1000 Days,
USA), Anna Taylor (Department for International Development, UK),
Derek Yach (The Vitality Group, USA). The Advisory Committee
provided advice in a meeting with Series Coordinators for each paper at
the beginning of the process to prepare the Series and in a meeting to
review and critique the draft reports.
Other contributors
Derek Headey, research fellow at IFPRI, did the analysis and prepared
the panel on the relation between income and nutrition. Joe Green,
independent consultant, did the analysis of the role of education.
Mara van den Bold, research analyst at IFPRI, reviewed the literature on
womens empowerment, provided background text, and wrote the panel
on womens empowerment and nutrition. Jef Leroy, research fellow at
IFPRI, did the analysis and wrote the panel on the cash transfer
programme in Mexico. Erick Boy, senior research fellow at IFPRI,
provided background text and other material for the section on
biofortification. Sivan Yosef, programme manager at IFPRI, provided
research and editing assistance during all steps of the process.
Conflicts of interest
REB serves on the Boards of the Micronutrient Initiative, Vitamin
Angels, the Child Health and Nutrition Research Initiative, and the
Nestle Creating Shared Value Advisory Committee. VM serves on the
Nestle Creating Shared Value Advisory Committee. All other authors
declare that they have no conflicts of interest. As corresponding author,
Marie Ruel states that she has full access to all data and final
responsibility for the decision to submit for publication.
Acknowledgments
Funding for the preparation of the Series was provided to the Johns
Hopkins Bloomberg School of Public Health through a grant from the
Bill & Melinda Gates Foundation. The sponsor has no role in the
analysis and interpretation of the evidence or in writing the paper and
the decision to submit for publication. We thank John Hoddinott, senior
research fellow at IFPRI, who reviewed an early version of the paper, and
to four anonymous reviewers for their insightful comments.
References
1 Pinstrup-Andersen P, ed. The African food system and its
interactions with human health and nutrition. Ithaca: Cornell
University Press, 2010.
548
2 IFPRI. 2011 Global food policy report. Washington, DC:
International Food Policy Research Institute, 2012.
3 Nelson GC, Rosegrant MW, Koo J, et al. Climate change. Impact on
agriculture and costs of adaptation. Washington, DC: International
Food Policy Research Institute, 2009.
4 Webb P. Medium-to long-run implications of high food prices for
global nutrition. J Nutr 2010; 140: S14347.
5 FAO, WFP, IFAD. The state of food insecurity in the world, 2012.
Economic growth is necessary but not sucient to accelerate
reduction of hunger and malnutrition. Rome: Food and Agriculture
Organization, 2012.
6 World Bank. Global economic prospects: commodities at the
crossroads. Washington, DC: World Bank, 2009.
7 OECD-FAO. OECD-FAO agricultural outlook, 2012. DOI:10.1787/
agr_outlook-2012-en.
8 Pinstrup-Andersen P, Watson DDI. Food policy for developing
countries. The role of government in global, national and local food
systems. Ithaca: Cornell University Press, 2011.
9 Scaling Up Nutrition. Progress report from countries and their
partners in the Movement to Scale Up Nutrition. New York: UN, 2011.
10 Shekar M, Ruel-Bergeron J, Herforth A. Module A. Introduction.
In: Improving nutrition through multisectoral approaches.
Washington, DC, International Bank for Reconstruction and
Development, International Development Association of The
World Bank, 2013.
11 Black RE, Victora CG, Walker SP, and the Maternal and Child
Nutrition Study Group. Maternal and child undernutrition and
overweight in low-income and middle-income countries. Lancet
2013; published online June 6. http://dx.doi.org/10.1016/S0140-
6736(13)60937-X.
12 Grosh M, Del Ninno C, Tesliuc E, Ouerghi A. For protection and
promotion: the design and implementation of eective safety nets.
Washington, DC: World Bank, 2008.
13 Maxwell D, Webb P, Coates J, Wirth J. Rethinking food security in
humanitarian response. Food Policy 2010; 35: 9197.
14 Connolly MA, Gayer M, Ryan MJ, Salama P, Spiegel P,
Heymann DL. Communicable diseases in complex emergencies:
impact and challenges. Lancet 2004; 364: 197483.
15 Alderman H. Safety nets can help address the risks to nutrition
from increasing climate variability. J Nutr 2010; 140: S14852.
16 Haddad L, Alderman H, Appleton S, Song L, Yohannes Y.
Reducing child malnutrition: how far does income growth take us?
World Bank Econ Rev 2003; 17: 10731.
17 Alderman H, Linnemayr S. Anemia in low-income countries is
unlikely to be addressed by economic development without
additional programs. Food Nutr Bull 2009; 30: 26569.
18 Demographic Health Surveys. Measure DHS Stat-Compiler. http://
www.statcompiler.com/ (accessed Nov 22, 2012).
19 WHO. Global database on child Growth and malnutrition. World
Health Organization, 2012. http://www.who.int/nutgrowthdb/
database/en/ (accessed Dec 12, 2012).
20 Engle PL, Black MM, Behrman JR, Cabral de Mello M, Gertler PJ,
Kapiriri L. Strategies to avoid the loss of developmental potential in
more than 200 million children in the developing world. Lancet
2007; 369: 22942.
21 Engle PL, Fernald LCH, Alderman H, et al. Strategies for reducing
inequalities and improving developmental outcomes for young
children in low-income and middle-income countries. Lancet 2011;
378: 133953.
22 Bhutta ZA, Das JK, Rizvi A, et al, The Lancet Nutrition Interventions
Review Group, and the Maternal and Child Nutrition Study Group.
Evidence-based interventions for improvement of maternal and
child nutrition: what can be done and at what cost? Lancet 2013;
published online June 6. http://dx.doi.org/10.1016/S0140-
6736(13)60996-4.
23 Almond D, Currie J. Killing me softly. J Econ Perspect 2011;
25: 125.
24 Hoddinott J, Maluccio J, Behrman JR, et al. The consequences of
early childhood growth failure over the life course. IFPRI
Discussion Paper 01073. Washington, DC: International Food Policy
Research Institute, 2011.
25 Alderman H. The economic cost of a poor start to life.
J Dev Origins Health Dis 2010; 1: 1925.
www.thelancet.com Vol 382 August 10, 2013
 SCIENTIFIC AMERICAN

S ECU R I T Y

IS YOUR
FOOD T

CONTAMINATED?
H

New approaches are needed I

to protect the food supply E

By Mark Fischetti C

G
E

IVEN THE BILLIONS OF FOOD ITEMS THAT ARE PACKAGED, PURCHASED AND CONSUMED
every day in the U.S., let alone the world, it is remarkable how few of them O
are contaminated. Yet since the terrorist attacks of September 11, 2001, F
food defense experts have grown increasingly worried that extremists
might try to poison the food supply, either to kill people or to cripple the F

economy by undermining public confidence. At the same time, production O

O
of edible products is becoming ever more centralized, speeding the spread
D
of contaminantsnatural ones or those introduced purposelyfrom farms or processing
plants to dinner tables everywhere. Mounting imports pose yet another rising risk.

Can we protect food from tainting? And
if toxic substances or pathogens do slip
into the supply chain, can they be quickly
detected to limit their harm to consum-
ers? Tighter production procedures can
go a long way toward protecting the pub-
lic, and if they fail, smarter monitoring
technologies can at least limit injury.
TIGHTEN SECURITY
PREVENTING A TERRORIST or a disgruntled
employee from contaminating milk, juice,
produce, meat or any other type of comes-
tible is a daunting problem. The food sup-
ply chain comprises a maze of steps, and
virtually every one of them presents an
opportunity for tampering. Blanket solu-
tions are unlikely to work because the
chain differs from commodity to com-
modity, says David Hennessy, an eco-
nomics professor at Iowa State Universi-
tys Center for Agricultural and Rural De-
velopment. Protecting dairy products is
different from protecting apple juice,
which is different from protecting beef.

IN BRIEF

Terrorist plots, greater imports and more Safety procedures adapted from mili- New technologies such as micro uidic Streamlining government regulations
centralized food production are raising tary methods of vulnerability analysis chips, RFID tags and edible markers can could help, but closer oversight by brand-
the chances that food will be tainted by can tighten the physical security of food- detect contamination, help trace it back name sellers of their suppliers will do
natural or man-made contaminants. production facilities. to its source and speed recalls. more to ensure safer fare.

Photograph by Aaron Goodman 59

2015 Scientific American

 SCIENTIFIC AMERICAN

THE PROBLEM Even within a given supply chain, tech-

nology offers few quick fixes. Preventing
Vulnerabilities Everywhere contamination largely comes down to
tightening physical plant security and pro-
The food supply chain includes numerous steps where natural pathogens can multiply cessing procedures at every turn. Each
or where terrorists or disgruntled employees could introduce toxic substances. The main farmer, rancher, processor, packager, ship-
countermeausures are to impose tight security and production procedures at each facility. per, wholesaler and retailer has to identify
every possible vulnerability in the facility
and in its procedures and close up every
hole, says Frank Busta, director emeritus
of the National Center for Food Protection
and Defense at the University of Minneso-
ta. The effort begins with standard facility
access controls, which Busta often refers
to as gates, guns and guards, but extends
to thoroughly screening employees and
T RANCHES FARMS carefully sampling products at all junc-
H tures across the facility at all times.
E
That advice seems sound, of course,
but the challenge for operators is how
S
best to button down procedures. Several
systems for safeguarding food production
C
have been rolled out in recent years. Al-
I FEEDLOTS though these are not required by any reg-
E STORAGE SITES ulatory agency, Busta strongly recom-
N
mends that producers implement them.
In the U.S., that impetus was made stron-
C
ger by legislation such as the 2002 Bioter-
E rorism Act and a 2004 presidential direc-
tive, both of which enhanced scrutiny of
SLAUGHTERHOUSES
ingredient suppliers and tightened con-
MILLS trol of manufacturing procedures.
O
The primary safeguard systems Busta
F recommends borrow from military prac-
tices. One tool, which the FDA and the
F U.S. Department of Agriculture are now
O using, carries the awkward name of
PACKAGING OPERATIONS CARVER+Shock. It was adapted from De-
O
fense Department procedures for identify-
D WAREHOUSES
ing a military services greatest vulnerabili-
PROCESSING PLANTS ties. CARVER+Shock is essentially a com-
plete security audit, says Keith Schneider,
a professor at the University of Floridas
department of food science and human
nutrition. The approach analyzes every
SHIPPING node in the system for factors that range
from the likely success of different kinds of
attacks to the magnitude of public health,
economic and psychological effects (to-
IMPORTS (SEE P. 62) gether, the shock value) that a given type
of infiltration could cause.
WHOLESALERS
TRACK CONTAMINANTS
NO MATTER how tightly procedures might
be controlled, determined perpetrators
could still find ways to introduce
pathogens or poisons. And natu-
ral pathogens such as Salmonel-
RESTAURANTS/STORES la are always a concern. Detect-

60 Illustrations by Lucy Reading-Ikkanda

2015 Scientific American

 SCIENTIFIC AMERICAN

COMING TECHNOLOGIES

Detect, Track and Trace

When a natural pathogen, or a perpetrator, contaminates food, lives can be saved if the tainting can be quickly detected and traced to
its origin, so that the rest of the batch can be destroyed or recalled. Several technologies, in various stages of development, could help:

Microfluidic detectorsBotulinum bacteria produce the most poisonous toxin known. They and
similar agents, such as Escherichia coli could be detected during ood rocessing by micro uidic
chi s sel contained diagnostic labs the si e o a finger esearch grou s ha e cra ted chi s
in which channels are lined with antibodies that can detect botulism or E. coli during milk
production. Samples could be tested either before or after the milk is piped into tanker trucks
that leave the dairy and again before or after it is pasteurized at a production plant. Similar
chi s could detect other to ic substances at uid rocessing lants that roduce a le
juice, soup, baby formula and other liquid foods.

Active packagingE. coli, Salmonella and other pathogens could be WARNING!

detected by small windows in packaging, such as the cellophane around
meat or the plastic jar around peanut butter. The intelligent window would
contain antibodies that bind to enzymes or metabolites produced by the
microorganism; if binding occurred, the patch would turn color. The challenge is to craft the
windows from materials and reactants that can safely contact food. Similar biosensors could react
if the contents reached a certain pH level or were exposed to high temperature, indicating spoilage.
And they could sense if packaging was tampered with, for example, by reacting to the pressure
imposed by a syringe or to oxygen seeping in through a puncture hole.

RFID tags allets or cases o a ew select oods now s ort radio re uency identification
(RFID) tags that, when read by a scanner, indicate which farm or processing plant the batch
came from. Future tokens that are smaller, smarter and cheaper could adorn individual packages
and log every facility they had passed through and when. Researchers have been develop-
ing tags that could be read through uid (traditional designs cannot) and thus
could be embedded inside the wall of sour cream or yogurt containers. Active
tags could record the temperatures a package had been exposed to, but they
are more expensive to produce.

Edible tagsManufacturers often combine crops from many growers, such as spinach leaves, into a retail
Orchard
ac age so tags a ed to bags might not hel in estigators trac contamination bac to a s ecific source 132
Edible inks can now be used to print markers that indicate site of origin directly onto a package of foodor
60 microns
in some cases, onto the food itself.

ing these agents, tracing them back to
the spot of introduction, and tracking
which grocery stores and restaurants
ended up with tainted products are para-
mount. Putting such systems in place is
just as important as prevention, Schnei-
der says.
Here new technology does play a ma-
jor role, with various sensors applied at
different points along the chain. You
cant expect one technology to counter all
the possible taintings for a given food,
notes Ken Lee, former director of Ohio
State Universitys department of food sci-
ence and technology.
A variety of hardware is being devel-
oped [see box above], although little has
been deployed commercially thus far. Ra-
dio-frequency identification (RFID) tags
are furthest along, in part because the
Defense Department and Walmart have
required their main suppliers to attach
the tokens to pallets or cases of food-
stuffs. The Metro AG supermarket chain
in Germany has done the same. The ulti-
mate intent is for automated readers to
scan the tags at each step along the sup-
ply chainfrom farm, orchard, ranch or
processor, through packaging, shipping
and wholesaleand to report each items
location to a central registry. That way if
a problem surfaces, investigators can
quickly determine where the batch origi-
nated and which stores or facilities
might have received goods from that
batch and when. Retailers can also read
the tags on their items to see whether
they received a product that was later
identified as suspicious.
As RFID tags get smaller and cheap-
er, they may be placed directly on indi-
vidual itemson every bag of spinach,
jar of peanut butter, container of shrimp
and sack of dog food. That way if a re-
call is issued, the items can be found as
they run past a scanner at the checkout
counter, says Jean-Pierre mond, of the
electro-optical systems laboratory at the
Georgia Tech Research Institute.
Universities and companies are devel-
oping all kinds of other tags, some that
are very inexpensive and others that cost
more but supply extensive information.
Some tokens, for example, can sense

61

2015 Scientific American


Making Imports Safer
Regardless of how safe domestic food production is, imports are
our Achilles heel. Realistically, no technology can ensure that
imports are safe. The food in every shipping container entering
a U.S. port or border crossing could be pulled and irradiated,
and some comestibles such as spices are already processed this
way ut industry e erts say the ste would add significant
costs for producers and shipping delays for middlemen. And the
public continues to be wary of the technology. Furthermore,
although irradiation would kill pathogens, it would have no
e ect on oisons or adulterants
Inspecting all incoming food would require vast increases
in FDA and U.S. Department of Agriculture budgets; the agencies
T
currently inspect a meager 1 percent of imports. But a 2011 law,
H the Food Safety Modernization Act, mandates that this will
E soon change. The FDA is now finali ing rules that when they
go into e ect in and will o erhaul its ood sa ety
S program, including its protection against tainted imports. The
C
N
C
Form
er U.S.S.R.: 4
E 31
Japa
E. As n: 827
ia: 1,10
China: 6 6
,679
S. E. Asia: 12,0
O
13
F
s: 5
land
Paci fic Is
Australia,
F me
O
D
U.S. FOOD IMPORTS, 2013
Millions of U.S. dollars
whether food has been exposed to warm
temperatures and thus might be more
likely to harbor Escherichia coli or Salmo-
nella bacteria. Other tags could track how
long items spent in transit from node to
node in the supply chain, which could in-
dicate unusually long delays that might
raise suspicion about tampering. So-called
active packaging could detect contamina-
tion directly and warn consumers not to
eat the product they are holding.
The big impediment for any marker, of
course, is the price. Today it costs around
15cents to put an RFID tag on a case of let-
SCIENTIFIC AMERICAN
POLICY SOLUTIONS
agency says it has already started ramping up inspection of
food-export facilities and now looks at about 1,200 a year.
ut the law re uires it to set u a erification system or oreign
food supplierswhich includes more than 100,000 facilities
that will require importers and U.S. manufacturers that use
imported ingredients to provide safety plans that include
more detailed information about production processes at
foreign suppliers.
The best recourse, Lee says, is for companies to insist that
suppliers impose strict standards and that the companies send
inspectors overseas to verify compliance. Other experts agree,
adding that go ernment edicts are not as e ecti e oo o ten
import requirements are used as trade barriers, and they just
escalate, says David Hennessy, an economics professor at Iowa
State University. The food companies themselves have a lot to
lose, however. When they source a product in a country, they
ought to impose tough procedures there. M.F.
22 ,55 5
Canada
151
19,
U.S. pe
:
:
ro
Eu
394
rica:
exi N. Af
M 18,0
c
0 7 o: Sub
-Sa M. East:
hara 940
n Afri
ca: 1,837
Am
C.
,16
3 erica: 4,970 S. Asia: 4,793
Caribbean: 562
S. A
rica
: 14, 5 87
Total U.S. Imports: $114,025 million
tuce. For some growers, mond notes, that nies via a secure Internet site hosted by
could equal the profit they are going to VeriSign, the data security firm. The com-
make on that case. To be embraced widely, pilation allowed the participants to more
therefore, he says tags will have to provide quickly resolve order discrepancies, to log
additional value to suppliers or buyers. how long food sat idle and to reveal ways
The University of Florida conducted a to raise shipping efficiency.
project with Publix Super Markets and
SOURCE: U.S. DEPARTMENT OF AGRICULTURE
produce suppliers in Florida and Califor- CONTROL SUPPLIERS
nia to assess such possibilities. In initial COSTS WILL NOT DROP until new techno-
trials, tags tracked crates and pallets that logies are deployed widely. But food de-
were shipped from the growers to several fense analysts say adoption is unlikely to
of Publixs distribution centers. Informa- occur until clear, streamlined regulations
tion gleaned from scanning tags at vari- are enacted. And that prospect, in turn,
ous points was available to all the compa- seems likely to be remote until the high-
62 Illustration by George Retseck and Lucy Reading-Ikkanda
2015 Scientific American
 SCIENTIFIC AMERICAN

S A F E T Y AT H O M E

The Vigilant Kitchen

If contaminated food does make it into your grocery bag, smart appliances could still prevent it from reaching your mouth. Innovations
that could reach commercial introduction are described here by Ken Lee of Ohio State University. None of this technology
would be visually obtrusive, he says, and all of it would be easy to clean.
Pulsed Light
When homeowners are asleep,
fi tures underneat cabinets
emit pulses of ultraviolet light
that kill germs on countertops
and other surfaces.

Microwave Refrigerator I

An infrared sensor gauges internal A built-in reader scans RFID tags on food and checks E
food temperature and compares it for recalls over a wireless Internet connection. (A
with safety guidelines, indicating homeowner could hold nonrefrigerator items under N

when the proper value has been it, too.) The reader also notes expiration dates written C
reached. Instead of entering a into the tags and tracks when containers such as milk
cooking time, a user enters the cartons are removed and put back, to see if they have E

food type or target temperature. been out for too long and therefore might be spoiled.
A red light warns of trouble.

est levels of government are reformed.
There are more than a dozen differ-
ent federal agencies that oversee some as-
pect of food safety, Lee points out, noting
that simple coordination among them is
difficult enough, and efficient approval of
F
proposed unifying responsibility for pre- the latest security procedures and detec-
vention of food contamination and re- tion, track and trace technologiesand F
sponse to outbreaks of foodborne illness drop those that do not. O
in a new Food Safety Administration. Brand-name providers, after all, have
O
More effective government is good, the most to lose. If a natural or man-
but experts say food suppliers also need made toxin is found in, say, a bag of Dole D

sensible requirements is even harder to
come by. The FDA regulates pizza with
cheese on it, but the USDA regulates pizza
if it has meat on it, quips Jacqueline
Fletcher, a professor of entomology and
plant pathology at Oklahoma State Uni-
versity. The requirements for organic
farmers are different from those for non-
organic farmers.
Spurred by recalls and outbreaks, Con-
gress passed the Food Safety Moderniza-
tion Act in 2011, which directed the FDA to
refocus its efforts on prevention of con-
tamination rather than recalls of bad food,
strengthen inspections of both domesti-
cally produced and imported foods [see
box on opposite page], and make better
use of risk analysis in setting its priorities.
In February, the Obama administration
to improve vigilance. The strongest tool spinach or a container of Safeway milk,
for stopping intentional contamination consumers will shun that particular la-
is supply-chain verification, says Shaun bel. If a brand-name company wants to
Kennedy, director of the National Center protect its products, Kennedy says, it
for Food Protection and Defense. Each should validate every participant in the
brand-name provider such as Dole and chain, all the way back to the farm.
each grocery store conglomerate such as
Safeway must verify that every company Mark Fischetti is a senior editor
involved in its supply chain implements at Scientific American.
M O R E TO E X P L O R E
Terrorist Threats to Food. World Health Organization, May 2008.
The Case of Botulinum Toxin in Milk: Experimental Data. Oliver G. Weingart et al. in Applied and Environmental
Microbiology, Vol. 76, No. 10, pages 32933300; May 2010.
Emerging Global Food System Risks and Potential Solutions. Shaun Kennedy in Improving Import Food Safety.
Edited by Wayne Ellefson, Lorna Zach and Darryl Sullivan. John Wiley & Sons, 2013.
National Center for Food Protection and Defense: www.ncfpd.umn.edu
s c i e n t i f i c a m e r i c a n . c o m /m a g a z i n e /s a

63

2015 Scientific American

 Articles
Lancet Infect Dis 2014;
14: 97681
Published Online
September 4, 2014
http://dx.doi.org/10.1016/
S1473-3099(14)70883-7
See Comment page 907
Division of Viral Hepatitis,
National Center for HIV,
Hepatitis, STD and TB
Prevention (M G Collier MD,
Y E Khudyakov PhD,
N P Nelson MD, A Suryaprasad MD,
J Drobeniuc MD,
S D Holmberg MD, F Xu MD),
Epidemic Intelligence Service
(E Epson MD), Division of
Foodborne, Waterborne, and
Environmental Diseases,
National Center for Emerging
and Zoonotic Infectious Diseases
(S E Lance DVM, M E Wise PhD),
Centers for Disease Control and
Prevention, Atlanta, GA, USA;
New Mexico Department of
Health, Santa Fe, NM, USA
(D Selvage MHS,
M Adams-Cameron MPH);
Colorado Department of Public
Health and Environment,
Denver, CO, USA (E Epson,
A Cronquist RN, R H Jervis MPH);
California Department of Public
Health, Sacramento, CA, USA
(K Lamba MPH, A C Kimura MD);
Nevada Division of Public and
Behavioral Health, Carson City,
NV, USA (R Sowadsky MSPH);
Arizona Department of Health
Services, Phoenix, AZ, USA
(R Hassan MSPH); Hawaii State
Department of Health, Honolulu,
HI, USA (S Y Park MD); Texas
Department of State Health
Services, Austin, TX, USA
(E Garza MPH) Food and Drug
Administration, Silver Spring,
MD, USA (A J Elliott MS,
D S Rotstein DVM, J Beal MPH,
T Kuntz MS, S E Lance,
R Dreisch MS); and Oakridge
Institute for Science and
Education, Oakridge, TN, USA
(A J Elliott)
976
Outbreak of hepatitis A in the USA associated with frozen
pomegranate arils imported from Turkey: an epidemiological
case study
Melissa G Collier, Yury E Khudyakov, David Selvage, Meg Adams-Cameron, Erin Epson, Alicia Cronquist, Rachel H Jervis, Katherine Lamba,
Akiko C Kimura, Rick Sowadsky, Rashida Hassan, Sarah Y Park, Eric Garza, Aleisha J Elliott, David S Rotstein, Jennifer Beal, Thomas Kuntz,
Susan E Lance, Rebecca Dreisch, Matthew E Wise, Noele P Nelson, Anil Suryaprasad, Jan Drobeniuc, Scott D Holmberg, Fujie Xu, for the Hepatitis A
Outbreak Investigation Team
Summary
Background In May, 2013, an outbreak of symptomatic hepatitis A virus infections occurred in the USA. Federal, state,
and local public health ocials investigated the cause of the outbreak and instituted actions to control its spread. We
investigated the source of the outbreak and assessed the public health measures used.
Methods We interviewed patients, obtained their shopping information, and did genetic analysis of hepatitis A virus
recovered from patients serum and stool samples. We tested products for the virus and traced supply chains.
Findings Of 165 patients identified from ten states, 69 (42%) were admitted to hospital, two developed fulminant hepatitis,
and one needed a liver transplant; none died. Illness onset occurred from March 31 to Aug 12, 2013. The median age of
patients was 47 years (IQR 3558) and 91 (55%) were women. 153 patients (93%) reported consuming product B from
retailer A. 40 patients (24%) had product B in their freezers, and 113 (68%) bought it according to data from retailer A.
Hepatitis A virus genotype IB, uncommon in the Americas, was recovered from specimens from 117 people with
hepatitis A virus illness. Pomegranate arils that were imported from Turkeywhere genotype IB is commonwere
identified in product B. No hepatitis A virus was detected in product B.
Interpretation Imported frozen pomegranate arils were identified as the vehicle early in the investigation by combining
epidemiologywith data from several sourcesgenetic analysis of patient samples, and product tracing. Product B was
removed from store shelves, the public were warned not to eat product B, product recalls took place, and postexposure
prophylaxis with both hepatitis A virus vaccine and immunoglobulin was provided. Our findings show that modern
public health actions can help rapidly detect and control hepatitis A virus illness caused by imported food. Our findings
show that postexposure prophylaxis can successfully prevent hepatitis A illness when a specific product is identified.
Imported food products combined with waning immunity in some adult populations might make this type of intervention
necessary in the future.
Funding US Centers for Disease Control and Prevention, US Food and Drug Administration, and US state and local
public health departments.
Introduction retailer A and who had hepatitis A illness with symptom
Infection with the hepatitis A virus causes an acute viral onset dates within 1 week of each other. The Department
illness characterised by fever, abdominal pain, high contacted states in the same distribution area for retailer A,
concentrations of aminotransferase enzymes, and and the Colorado Department of Public Health and
jaundice. Hepatitis A vaccine or immunoglobulin given Environment identified four additional retailer A shoppers
within 2 weeks of exposure can prevent disease.1 In the with hepatitis A illness. None of these patients had typical
USA, hepatitis A vaccination has been recommended for risk factors for hepatitis A (eg, international travel, injection
all children aged 1223 months since 2006,2 and drug use). Initial interviews using a hypothesis generating
consequently the number of acute cases has greatly questionnaire, which asks about all possible food exposures,
fallen.3 Hepatitis A virus is transmitted by person-to- identified some similar eating patterns, including fruits
person contact or ingestion of contaminated food or and vegetables consumed in smoothies. Four patients from
water, and is primarily a human pathogen. International two states had had product Ba frozen mix of cherries,
travel was reported in about half of all reported US cases.4 strawberries, raspberries, blueberries, and pomegranate
The most recent common source outbreak of foodborne arilsin their freezers and used it in smoothies. Retailer A
hepatitis A in the USA happened in 2005.5 confirmed that people with hepatitis A purchased product B
On May 13, 2013, the New Mexico Department of Health from retailer A. This report describes the investigation of
notified the Centers for Disease Control and Prevention the multistate outbreak associated with product B and the
(CDC) of two people in the same city who had shopped at public health actions taken to halt the outbreak.
www.thelancet.com/infection Vol 14 October 2014
 Articles

Methods final phylogenetic tree based on maximum likelihood Correspondence to:

Patients
We defined acute hepatitis A as an acute illness with
hepatitis symptoms (eg, fever, headache, malaise, anorexia,
nausea, vomiting, diarrhoea, and abdominal pain), and
either jaundice, or high serum alanine aminotransferase or
aspartate aminotransferase concentrations6 and positive for
IgM antibody to hepatitis A virus. An outbreak-related case
was confirmed if the patient met acute hepatitis A criteria
with symptom onset between March 15, 2013, and Aug 12,
2013, and either who reported consuming product B within
their illness incubation period (1550 days before symptom
onset) irrespective of hepatitis A virus genotype or who had
hepatitis A virus genotype IB isolated from a clinical
specimen. Secondary cases had close contact with a
confirmed patient but did not report consuming product B
during the incubation period. State and local health
departments collected information about demographic,
clinical, and food exposure and clinical specimens from
people with reported acute hepatitis A with symptom onset
during March 15 to Sept 5, 2013. This study was done in
response to a public health emergency and was thus
exempt from institutional review board approval.
Procedures
Serum or stool samples from potential outbreak-related
cases were collected within 28 days of symptom onset,
frozen at 20C as soon as possible after collection, and
shipped overnight on dry ice to the CDC Division of Viral
Hepatitis laboratory for genetic sequencing. Hepatitis A
virus RNA was extracted from serum samples and used to
amplify and sequence a 315 bp fragment of the VP1/P2B
region.7 We did preliminary sequence analysis with
DNAStar Lasergene 11 (DNASTAR, Madison, WI, USA).
We did further sequence analysis with the Geneious 7.0.5
(Biomatters, Auckland, New Zealand). We used Geneious
Alignment to do sequence alignment. We constructed a
algorithms.8 To improve case-finding, in addition to Dr Melissa Collier, Division of
Viral Hepatitis, CDC Mailstop
outbreak-related laboratory activities, we also solicited G-37, Atlanta, GA 30329, USA
specimens nationwide through the CDCs Epidemic mgcollier@cdc.gov
Information Exchange (Epi-X) on June 26, 2013 from
people with acute hepatitis A and no history of international
travel during the incubation period. These data were used
in a sensitivity analysis to assess the inclusion of hepatitis
A virus genotype IB in our case definition by recording
how many genotype IB infections were detected among
non-outbreak related cases. Samples of product B from
patients freezers were collected for hepatitis A virus
testing, and tested with a method modified from the US
Food and Drug Administration (FDA) bacteriological
analytical manual for hepatitis A virus in green onions.9
The FDA investigated the facility of manufacturer C,
which made product B. All product B ingredients were
traced back to their source, and frozen pomegranate arils
from importer D were traced forward.
Information about the ongoing outbreak was sent
through Epi-X to notify all states about the potential
outbreak on May 22, 2013. The public was notified of the
outbreak on May 31, 2013, through state and local health
department press releases and the CDCs website.10
Information from membership cards for retailer A was
obtained to identify which patients bought product B from
retailer A during Dec 21, 2012May 31, 2013; patients were
then contacted from May 31, 2013, through an automated
calling system with information about the risk of hepatitis A
associated with product B. Product B was removed from
shelves of retailer A on May 30, 2013. CDC, state, and local
health ocials recommended that anyone who had eaten
product B in the previous 2 weeks be assessed for
hepatitis A virus postexposure prophylaxis, which included
vaccination and immunoglobulin, given either through
state or local health departments or their private physician.
CDC provided guidance to state and local health

8 New Mexico reports Public notified

first case of the outbreak
7 Voluntary recall
Colorado and of product
Nevada report FDA issues two import
6 cases alerts for importer D
pomegranate arils
5 Voluntary recall
Number of illnesses

of product E
4

0
31

14

21

28

12

19

26

16

23

30

14

21

28

g4

1
g1
ril

ly
ay

ne

ne
ch

ril

ay

ly
ne
ril

ly
ay

ne

ne

Au
ril

ay

ly
Ju
Ap

Au
Ju

Ju

Ju
Ju

Ju
Ap
Ap

Ap

M
ar

Ju
Ju

Ju
M

Date of illness onset (2013)

Figure 1: Confirmed outbreak-related cases of hepatitis A

www.thelancet.com/infection Vol 14 October 2014 977

 Articles

departments about who should receive what postexposure back to their originating firms through investigation of
prophylaxis according to the Advisory Committee for purchasing and invoice documents.
Immunization Practices guidelines;1 however, final Strawberries and blueberries came from multiple
decisions were made locally based on local resources. importers from more than one location, and so were not
considered the likely vehicle. Cherries from the USA and
Results strawberries from Argentina in product B came from a
We identified 165 people from ten states meeting the single supplier but in several dierent lots and so were
confirmed case definition; eight of whom met the unlikely to be the vehicle. Product B cherries and
secondary case definition. Dates of illness onset ranged strawberries were also used in other widely distributed
from March 31, 2013, to Aug 12, 2013 (figure 1). 69 (42%) products made by manufacturer C and no other outbreak-
of 165 patients were admitted to hospital, two (1%) associated cases were reported among consumers of other
developed fulminant hepatitis, and one secondary case manufacturer C products. Frozen pomegranate arils
patient with fulminant hepatitis needed a liver transplant; imported from Turkey by importer D from a single lot
no deaths were reported. Most patients were aged were identified in product B (figure 2). The same lot of
4064 years and most were female (table). Few patients frozen pomegranate arils were also used in product E
were aged 18 years or younger, none of whom were produced by manufacturer F, and one person who had
vaccinated (two children were from one family who consumed product E had an infection of hepatitis A virus
refused postexposure prophylaxis). IB matching the outbreak strain. The first lot to contain
Product B was a mixture of frozen strawberries, red the Turkish arils was produced on Jan 24, 2013, and the
raspberries, blueberries, cherries, and pomegranate arils. bags had a shelf life of 2 years.
The workers processing product B wore gloves, limiting Of 157 patients, 152 (97%) reported consuming
bare hand contact with the product, and no workers product B or had purchased it according to member card
reported symptoms of acute hepatitis. Through lot information. Of five people who did not recall consuming
numbers of product B recovered from four patients product B, three were members of retailer A but did not
freezers, the five ingredients of product B were traced purchase product B, one person did not consume or
purchase product B but consumed product E (which
contained the same arils), and one person did not
Patients (n=165)
consume either product B or E. According to member
Female patients 91 (55%) card information, 113 patients (72%) bought product B,
Age, years and 40 had product B in their freezers (of whom, we
Median (IQR) 47 (3558) could not find purchase records for three). 71 (63%) of
018 11 (7%) 113 patients bought product B more than once between
1839 42 (25%) Jan 24 and May 31, 2013. The person identified in New
4064 95 (58%) Jersey was a secondary case exposed to a primary case in
65 17 (10%) Colorado, the two people from Wisconsin consumed
Ethnic origin product B while living in California before relocating to
White 123 (74%) Wisconsin, and the person from New Hampshire
Hispanic or Latino 13 (8%) traveled to Nevada during the illness incubation period
Asian 6 (4%) but did not recall consuming product B.
Black 2 (1%) Of the 165 patients, 120 had clinical specimens available
Other 3 (2%) for testing, including 119 serum samples and one stool
Unknown 18 (11%) specimen. 117 (98%) tested positive for hepatitis A virus
State of residence genotype IB. Genotype IA virus was recovered from two
Arizona 24 (15%) patients who reported consuming product B, and one
California 80 (48%) patient was hepatitis A virus negative. Phylogenetic analysis
Colorado 29 (18%)
of the outbreak specimens showed that the sequences from
Hawaii 8 (5%)
99 genotype IB specimens (85%) were identical in the 315
New Hampshire 1 (<1%)
bp segment of VP1/P2B (figure 3); the remaining 18
outbreak specimens belonged to minor genotype IB
New Jersey 1 (<1%)
strains. To ensure that other hepatitis A cases related to this
New Mexico 11 (7%)
outbreak were detected, CDC issued a national request
Nevada 6 (4%)
through Epi-X for state and local health departments to
Utah 3 (2%)
submit specimens from people with hepatitis A symptom
Wisconsin 2 (1%)
onset from March 1 to Sept 5, 2013. Specimens from an
Data are n (%) unless otherwise stated. additional 131 people were submitted as a result; of these,
47 (36%) were PCR negative for hepatitis A virus, 82 (62%)
Table: Patient characteristics
were positive for genotype IA, of which 66 belonged to the

978 www.thelancet.com/infection Vol 14 October 2014

 Articles

Importer U, Chile
Raspberry, multiple lots
Manufacturer C

Product B, lot X Importer T, Chile

Retailer A (branch 1),
Nevada
Product B, lot X
Ship date: March 18, 2013
Retailer A (branch 2),
Nevada
Product B, lot X
Retailer A, New Mexico
Product B, lot Y
Ship date: March 15, 2013
Retailer A, New Mexico
Product B, lot Z
Cherry, strawberry, Raspberry, multiple lots
Distributor W, blueberry, raspberry,
California pomegranate
Importer S, Chile
Lots X and Y Raspberry, multiple lots
Importer Q, Chile
Product B, lot Y Blueberry, multiple lots
Cherry, strawberry,
blueberry, raspberry,
pomegranate Importer P, Argentina
Blueberry, multiple lots
Distributor V,
Importer O, Argentina
Arizona
Strawberry, multiple lots
Product B, lot Z
Lots Y and Z Cherry, strawberry,
blueberry, raspberry, Supplier N, USA (WA)
pomegranate Cherry, multiple lots

Importer D, Turkey
Raspberries Strawberries Pomegranates Pomegranate, single lot
Blueberries Cherries

Figure 2: Tracing the ingredients of product B

subtype IA1 cluster, and two (2%) were positive for

genotype IB. We could not isolate hepatitis A virus from
samples of product B.
Product B was bought 283 times by 113 patients (an
average of 25 purchases per person) between Dec 21,
2012, and May 31, 2013. 218 (77%) of these purchases were
made between Feb 17, 2013, and May 12, 2013. Product B
was distributed through retailer A in Alaska, Arizona,
California, Colorado, Hawaii, Idaho, Montana, New
Mexico, Nevada, Oregon, Utah, and Washington between
Jan 21, 2013, and May 30, 2013; however, no cases were
reported from residents of Alaska, Idaho, Montana,
Oregon, or Washington. During Jan 21May 30, 2013,
409 286 units were sold. Of those units, 364 076 (89%)
were sold in the seven states from which the cases were
reported and where product B was sold. The product also
was distributed on a much smaller scale (1616 units) in Major outbreak strains (N=99)
Minor outbreak strains (N=18)
the northeastern USA with dierent packaging under Nucleotide USA surveillance reference
brand G; no cases were reported from consumers of variation Egypt, Israel, Jordan, and other
05% countries references
brand G.
On June 4, 2013, manufacturer C issued a voluntary recall Figure 3: Hepatitis A virus genotype IB genetic analysis
of specific product B lots, including selected lots of brand G; Compares the sequence of the 315 bp segment (VP1/P2B) among outbreak related and non-outbreak related
this recall was expanded on June 28, 2013, to include genotype IB positive specimens.
additional lots.11 Retailer A used automated phone calls to
notify more than 250 000 customers who bought product B same lot of frozen Turkish pomegranate arils as product B
about the link to hepatitis A. State and local health (figure 1). Although product E was recalled, CDC did not
departments gave postexposure prophylaxis to thousands recommend postexposure prophylaxis for people who had
of people. Retailer A paid for vaccination for more than consumed the product; however, some states did oer
10 000 people at more than 200 stores, which was done by residents postexposure prophylaxis if they had consumed
licensed professionals. On June 26, 2013, a voluntary recall the product. FDA placed importer D on two separate
was issued by manufacturer F for product E, which FDA import alerts on June 29, halting importer D shipments
identified through traceforward activities to contain the into the USA.

www.thelancet.com/infection Vol 14 October 2014 979

 Articles
Discussion
This foodborne outbreak of hepatitis A virus was the first
in the USA for almost 10 years, during an era of universal
childhood vaccination. Its investigation and control were
unique in several ways (panel): the implicated source
originated from another continent, showing the
international distribution of foods and associated food
safety problems; genetic information from patient
specimens suggesting infection with IB genotype was
used to link cases; the product was sold by a single retailer
with modern information systems, including itemised
sales data, with exact purchase dates, and customer
information; and the retailer had the willingness and
resources to give hepatitis A vaccine to those aected.
Hepatitis A, unlike other foodborne pathogens, can be
prevented if vaccination can be provided within 2 weeks of
exposure; in one instance during this outbreak, parents
refused vaccination for their children who then became ill.
Public health action was not delayed to wait for results
of a case-control study because hepatitis A has a long
incubation period (making patient recall of dietary
exposures questionable), the shelf life of the implicated
product was 2 years, and there was a continuing health
risk. Although cherries and strawberries were also
common among lots of product B, both the cherries and
strawberries were used in many other products sold by
product Bs manufacturer and no outbreak-associated
cases were recorded among people who had consumed
those products. This finding, combined with the case of
IB genotype infection matching the outbreak strain
associated with product E (which contains no cherries or
strawberries), supported the conclusion that pomegranate
arils were the source.
Product B was packaged in large bags (3 lb or 15 kg)
and contained many servings. Hepatitis A virus is
resistant to environmental conditions such as freezing15
and can be infectious for up to a month on environmental
surfaces.16 This persistence in the environment was
Panel: Research in context
Systematic review
Outbreaks of infectious diseases are investigated with a
combination of methods, including descriptive epidemiology,
and for foodborne outbreaks, shopping data and product
tracing once a food item is identified as a potential vehicle.
We did not do a systematic review of methods. Case-control
or cohort studies5,1214 are done if a specific food cannot be
identified through descriptive epidemiology alone, which was
not necessary for this investigation.
Interpretation
This investigation is the first to our knowledge in which a
food item linked to hepatitis A infection from a common
source was identified rapidly enough for eective
post-exposure prophylaxis interventions, potentially
preventing hepatitis A illnesses.
980
shown in two previous outbreaks of hepatitis A in which
strawberries were harvested a year or more before
causing illnesses, suggesting that these contaminated
frozen foods can be a source of hepatitis A virus illness
for a long period if not recalled.12,13 Hepatitis A virus is
inactivated by heat, but needs high temperatures to be
completely inactivated (up to 90C for 180 s).17
To our knowledge, this outbreak is the first US multistate
foodborne hepatitis A outbreak in which vaccine and
immunoglobulin postexposure prophylaxis were
recommended and provided by local health departments
and where vaccine was provided by a retailer. The
infrastructure provided by retailer A (in-store pharmacies,
vaccine administration sta, and automated calling
system) enabled the vaccine to be delivered quickly and
eciently, and increased the uptake of postexposure
prophylaxis, as evidenced by the large number of people
vaccinated. These measures probably prevented additional
cases: the number of new cases decreased soon after the
public was notified on May 31, 2013. Illness among
unvaccinated children that occurred after one family
declined postexposure prophylaxis could have been
prevented. Postexposure prophylaxis, both active and
passive, is an eective illness prevention strategy. However,
the burden and cost of providing it must be balanced with
the risk of disease. For this reason and lack of cases
associated with product E, CDC did not recommend
postexposure prophylaxis for product E consumers.
This outbreak primarily involved genotype IB, and very
few non-outbreak related specimens identified during the
outbreak period were genotype IB. These findings assisted
the detection of the outbreak and show the usefulness of
molecular typing techniques. Most cases of acute
hepatitis A reported in the USA are genotype IA, which is
endemic in Mexico; genotype IB is predominant in the
Middle East.1820 Several recent outbreaks of hepatitis A
virus genotype IB illnesses associated with food from the
Middle East have been reported outside the USA: frozen
strawberries from Egypt and Morocco were associated
with illness in Europe,21 frozen pomegranate seeds from
Egypt were associated with illness in Canada in 2012,22 and
semidried tomatoes from Turkey were associated with
illnesses in Europe and Australia in 200911.14,24,25 These
outbreaks show the international nature of hepatitis A, the
diculties of identification and control of outbreaks
related to foods of foreign origin,26 and the importance of
ensuring adequate hygiene during the production of
imported foods.
The age-specific seroprevalence of antibody against
hepatitis A virus among US adults has decreased since
the mid-1990s because of herd immunity provided by the
universal vaccination of children.27,28 In this outbreak,
most patients were aged 4064 years and of white ethnic
origin, either because they were more likely to consume
product B or because they were less likely to be immune
to hepatitis A virus, or a combination of both. As more
food is imported from distant regions of the world where
www.thelancet.com/infection Vol 14 October 2014

adequate processing methods are dicult to verify, the
eectiveness of vaccination for postexposure prophylaxis
in people older than 40 years needs to be studied.
We encountered challenges during this investigation.
First, we were unable to confirm pomegranate aril
contamination through food testing because of the
diculty of isolating hepatitis A virus from food.9,29 Second,
public health surveillance in the USA for hepatitis A is
passive and specimens are not routinely submitted for
public health analysis. Cases could have been missed if not
reported or the outbreak link was not considered. Third,
we were unable to inspect importer D because of political
unrest in the region.
In conclusion, public notification, product recall, and
hepatitis A postexposure prophylaxis within 2 weeks of
exposure can prevent illnesses during outbreaks; these
eorts are helped by retailers with reliable sales data
who are willing to work with public health agencies. The
foreign origin of the contaminated ingredient in this
outbreak underscores the importance of the new import-
related regulations in the Food Safety Modernization
Act. Under this law, importers are required to have a
programme to verify food product safety, and will be
required to implement risk-based preventive controls.
Contributors
All authors collected, analysed, and interpreted data, and wrote and
revised the report.
Hepatitis A outbreak investigation team
Ronald Balajadia, Steany J Cavallo, Lei Chen, Elizabeth R Daly,
Paul Ettestad, Julianna Ferreira, Eric Garza, Gemechu Gerbi, Debra Gilliss,
Rikita I Hatia, Michael Hernandez, James Kazmierczak, Maja Kodani,
Ben Kupronis, Yulin Lin, Trudy Murphy, Jennifer F Myers, Vi Peralta,
Tasha Poissant, Suzanne Rouleau, Eyasu Teshale, Daulati Thakare,
John W Ward, Joli R Weiss, Kathleen Winter, and Guo-liang Xia.
Declaration of interests
We declare no competing interests.
Acknowledgments
This study was funded by the US Centers for Disease Control and
Prevention, the Food and Drug Administration, and other state and
local public health departments. We thank the following people for their
contributions to the investigation: Nicole Comstock, Alycia Downs,
Kathleen Harriman, Cynthia Jorgensen, Bill Keene, Kenneth Komatsu,
Paul Peterson, Devin Raman, and Phil Spradling. The findings and
conclusions in this report are those of the authors and do not
necessarily represent the ocial position of the Centers for Disease
Control and Prevention.
References
1 Novak R, Williams I, Bell B. Advisory Committee on Immunization
Practices (ACIP) Centers for Disease Control and Prevention
(CDC). Update: prevention of hepatitis A after exposure to
hepatitis A virus and in international travelers. Updated
recommendations of the Advisory Committee on Immunization
Practices (ACIP). MMWR Recomm Rep 2007; 56: 108084.
2 Fiore AE, Wasley A, Bell BP. Advisory Committee on
Immunization Practices (ACIP). Prevention of hepatitis A
through active or passive immunization: recommendations of the
Advisory Committee on Immunization Practices (ACIP).
MMWR Recomm Rep 2006; 55: 123.
3 Viral hepatitis surveillance United States, 2011. http://www.cdc.
gov/hepatitis/Statistics/2011Surveillance/index.htm (accessed
July 23, 2014).
4 Klevens RM, Miller J, Iqbal K, et al. The evolving epidemiology of
hepatitis A in the United States: incidence and molecular
epidemiology from population-based surveillance.
Arch Intern Med 2010; 170: 181118.
www.thelancet.com/infection Vol 14 October 2014
Articles
5 Bialek SR, George PA, Xia GL, et al. Use of molecular
epidemiology to confirm a multistate outbreak of hepatitis A
caused by consumption of oysters. Clin Infect Dis 2007; 44: 83840.
6 NNDSS hepatitis A, acute. http://wwwn.cdc.gov/NNDSS/script/
casedef.aspx?CondYrID=703&DatePub=1/1/2012 (accessed
July 23, 2014).
7 Nainan OV, Xia G, Vaughan G, Margolis HS. Diagnosis of hepatitis A
virus infection: a molecular approach. Clin Microbiol Rev 2006; 19: 6379.
8 Felsenstein J. Maximum-likelihood estimation of evolutionary trees
from continuous characters. Am J Hum Genet 1973; 25: 47192.
9 Food and Drug Administration bacteriological analytical manual
(BAM). Detection of hepatitis A in foods. http://www.fda.gov/
Food/FoodScienceResearch/LaboratoryMethods/ucm374006.htm
(accessed July 23, 2014).
10 Multistate outbreak of hepatitis A virus infections linked to
pomegranate seeds from Turkey (final update). http://www.cdc.gov/
hepatitis/outbreaks/2013/a1b-03-31/index.html (accessed July 23, 2014).
11 FDA investigates multistate outbreak of hepatitis A illnesses
associated with pomegranate seeds from Turkish importer. http://
www.fda.gov/Food/RecallsOutbreaksEmergencies/Outbreaks/
ucm354698.htm (accessed July 23, 2014).
12 Hutin YJ, Pool V, Cramer EH, et al. A multistate, foodborne
outbreak of hepatitis A. N Engl J Med 1999; 340: 595602.
13 Niu MT, Polish LB, Robertson BH, et al. Multistate outbreak of
hepatitis-A associated with frozen strawberries. J Infect Dis 1992;
166: 51824.
14 Petrignani M, Harms M, Verhoef L, et al. Update: A food-borne
outbreak of hepatitis A in the Netherlands related to semi-dried
tomatoes in oil, January-February 2010. Euro Surveill 2010; 15: 19572.
15 Shieh YC, Stewart DS, Laird DT. Survival of hepatitis A virus in spinach
during low temperature storage. J Food Protect 2009; 72: 239093.
16 McCaustland KA, Bond WW, Bradley DW, Ebert JW, Maynard JE.
Survival of hepatitis A virus in feces after drying and storage for
1 month. J Clin Microbiol 1982; 16: 95758.
17 Sow H, Desbiens M, Morales-Rayas R, Ngazoa SE, Jean J. Heat
inactivation of hepatitis A virus and a norovirus surrogate in soft-
shell clams (Mya arenaria). Foodborne Pathog Dis 2011; 8: 38793.
18 Normann A, Badur S, Onel D, et al. Acute hepatitis A virus
infection in Turkey. J Med Virol 2008; 80: 78590.
19 Nejati A, Makvandi M, Samarbafzadeh A, Neisi N,
Moradzadegan H. Molecular epidemiology of hepatitis A virus in
patients in the Ahwaz region of Iran. J Med Virol 2012; 84: 58286.
20 Faber MS, Stark K, Behnke SC, Schreier E, Frank C. Epidemiology
of hepatitis A virus infections, Germany, 2007-2008.
Emerg Infect Diseases 2009; 15: 176068.
21 Nordic outbreak investigation team. Joint analysis by the Nordic
countries of a hepatitis A outbreak, October 2012 to June 2013:
frozen strawberries suspected. Euro Surveill 2013; 18: 20520.
22 Swinkels HM, Kuo M, Embree G, et al. Hepatitis A outbreak in
British Columbia, Canada: the roles of established surveillance,
consumer loyalty cards and collaboration, February to May 2012.
Euro Surveill 2014; 19: 20792.
23 Carvalho C, Thomas HL, Balogun K, Tedder R, Pebody R,
Ramsay M, Ngui SL. A possible outbreak of hepatitis A associated
with semi-dried tomatoes, England, JulyNovember 2011.
Euro Surveill 2012; 17: 20083.
24 Donnan EJ, Fielding JE, Gregory JE, et al. A multistate outbreak of
hepatitis A associated with semidried tomatoes in Australia, 2009.
Clin Infect Dis 2012; 54: 77581.
25 Gallot C, Grout L, Roque-Afonso AM, et al. Hepatitis A associated
with semidried tomatoes, France, 2010. Emerg Infect Diseases 2011;
17: 56667.
26 Holmberg SD. Hepatitis A epidemiology goes global. Clin Infect Dis
2012; 54: 78283.
27 Armstrong GL, Billah K, Rein DB, Hicks KA, Wirth KE, Bell BP.
The economics of routine childhood hepatitis A immunization in
the United States: the impact of herd immunity. Pediatrics 2007;
119: e2229.
28 Klevens RM, Kruszon-Moran D, Wasley A, et al. Seroprevalence of
hepatitis A virus antibodies in the U. S.: results from the National
Health and Nutrition Examination Survey. Public Health Rep 2011;
126: 52232.
29 Koopmans M, Duizer E. Foodborne viruses: an emerging problem.
Int J Food Microbiol 2004; 90: 2341.
981
The Conversation
Hep A scare: buy local to avoid
virus-laced berries

Feb 16, 2015

JACKSON STILES Life Editor

20

Fruit lovers infected with hepatitis A have been betrayed by poor hygiene overseas.

Do your bit and bin these berries.

The danger of uncooked foreign food has been exposed after berries contaminated with
a hard-to-detect virus infected at least nine Australians.
Frozen mixed berries from Chile and China sold by the Nanna and Creative Gourmet
brands have been recalled nationwide, prompting experts and industry voices to
promote buying local.

More than a dozen cases of hepatitis A linked to the imported berries have been
confirmed around the country.

An industry spokesman says he is amazed that consumers continue to buy frozen

berries from overseas, given the unsanitary conditions.

Dont eat it, says Australian Blueberry Growers Association president Greg McCulloch.

They might think its just a germ or something, but hepatitis A is actually a germ in s**t,
which might make them think twice.

Food standards in China may not match those to which Australian farmers
adhere. Photo: Getty

Eating foreign imports without cooking or boiling is dangerous because only heat
cleanses food tainted by faecal matter. Australia is the latest to be stricken by frozen
berry imports, after similar infections in Ireland, Italy and several Scandinavian countries
since 2013.

Poor hygiene is the most likely cause of the contamination, The University of Melbourne
agricultural expert Dr Said Ajlouni agrees.

The virus probably came from infected people who went to the toilet and came back to
the factory without appropriate hand washing, he says.
Because we dont have any control over good hygiene and sanitation practices in these
countries, they can be very risky.

Raw human sewage, which may contain viruses, bacteria and industrial chemicals, is
widely used as fertiliser for a wide variety of crops in China and Latin America, says a
2008 report by Sri Lanka-based International Water Management Institute (IWMI).

Pregnant women and young children are particularly susceptible to food

contamination. The New Dailyunderstands that at least one Victorian primary school has
removed all frozen berries from its menu after the recall.

It would be unfair to boycott foreign fresh food based on one incident, says Dr Ajlouni,
but he confirms that buying local is safer.

Farmers are calling for an overhaul of labelling laws to help consumers identify
Australian grown and packaged foods following the berry scare.

Victorian Farmers Federation president Peter Tuohey says loopholes in labelling laws
have allowed many imported foods to be passed off as if they are Australian made.

Weve been pushing for truth-in-labelling laws for years and it hasnt happened yet, he
says.

Australian berries are grown under stringent standards, says another industry voice.

The fresh Australian berry is certainly grown under some of the strictest food safety
standards in the world, says Raspberries and Blackberries Australia spokesman
Jonathan Eccles.

Consumers should be confident that when theyre buying a fresh berry it is safe.

The impediment to safety is price. Most local berries are sold fresh because Australias
high labour costs make it impossible to compete in the frozen fruit market.

For example, it costs berry grower Phil Rowe more to transport his fresh berries from his
farm in West Gippsland to Melbourne than to ship a frozen load from Chile.

If you prefer the frozen variety, homegrown berries can just as easily be stored in the
freezer, says the Sunny Creek Farm owner.

Fresh is always best, Mr Rowe says.

The recalled products are 1kg packs of Nannas Frozen Mixed Berries, and 300g and
500g packs of Creative Gourmet Mixed Berries.
Hepatitis symptoms usually begin two to seven weeks after exposure and include fever,
loss of appetite, nausea, fatigue, abdominal pain, diarrhoea and jaundice and can last
for months.
 EFSA Journal 2011;9(10):2390

SCIENTIFIC REPORT OF EFSA

Shiga toxin-producing E. coli (STEC) O104:H4 2011 outbreaks in Europe:

Taking Stock1
European Food Safety Authority2, 3

European Food Safety Authority (EFSA), Parma, Italy

ABSTRACT
On 21 May 2011, Germany reported an ongoing outbreak of Shiga-toxin producing Escherichia coli
(STEC), serotype O104:H4. From an initial case control study, the outbreak was associated with the
consumption of fresh salad vegetables. Subsequent investigations showed that the risk of infection was
significantly associated with the consumption of fresh sprouted seeds rather than with other fresh
vegetables. A tracing back and tracing forward study showed that all of the clusters for which there
was sufficient information, could be attributed to consumption of sprouted seeds from a single
sprouted seed producer in Germany. Investigation of the production site showed no evidence of
environmental contamination. Employees were found to be infected, but since they had not become ill
prior to the outbreak, it was concluded that they were not the source of the food contamination. Hence,
contaminated seeds used for the sprout production were the most likely source. Several varieties of
seeds were used, and sprouts thereof were sold as a mixture. Subsequently, a cluster of patients with
bloody diarrhoea was reported, after having participated in a local event in France on 8 June.
Consumption of sprouted seeds was also associated with occurrence of the disease in this cluster.
Furthermore, the STEC isolates responsible for the outbreaks in France and Germany were found to be
indistinguishable. It was therefore concluded that there was a common source for both outbreaks. A
comparison of the back tracing information on the seeds from the French and German outbreaks led to
the conclusion that a specific consignment (lot) of fenugreek seeds imported from Egypt was the most
likely link between the outbreaks, although it could not be excluded that other lots from the same
exporter and importer were also implicated. STEC O104 is a very rare serogroup in humans in the EU
and worldwide. Sporadic cases in the EU have been previously linked to travel to North Africa, the
Middle East and Central Asia. On 26 July, the Robert Koch Institute declared the outbreak finished. A
total of 3911 cases have been reported to the ECDC and WHO, linked to the outbreaks, to date.

European Food Safety Authority, 2011

KEY WORDS
E. coli, outbreak, sprouted seeds, fenugreek, STEC, VTEC, EHEC, EAggEC, ESBL

1
On request from EFSA, Question No EFSA-Q-2011-00958, issued on 16 September 2011.
2
Correspondence: emrisk@efsa.europa.eu
3
Acknowledgement: EFSA wishes to thank Alfredo Caprioli, Alexandra Fetsch, Matthias Greiner, Stefan Gross, Arie Havelaar,
Petra Luber, Anika Schielke and Armin Weiser for the preparatory work on this scientific output and EFSA staff: Pierre-Alexandre
Beloeil, Ernesto Liebana Criado, Hubert Deluyker, Olaf Mosbach-Schulz, Tobin Robinson, Didier Verloo for the support provided
to this scientific output. This Scientific Report has been reviewed by John Cowden, Tine Hald and James Mclauchlin.

Suggested citation: European Food Safety Authority; Shiga toxin-producing E. coli (STEC) O104:H4 2011 outbreaks in Europe:
Taking Stock. EFSA Journal 2011;9(10):2390. [22 pp.] doi:10.2903/j.efsa.2011.2390.
Available online: www.efsa.europa.eu/efsajournal

European Food Safety Authority, 2011

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

SUMMARY

On 21 May 2011, Germany reported an ongoing outbreak of Shiga-toxin producing Escherichia coli
(STEC), serotype O104:H4. As of 27 July, 3126 cases of diarrhoeal disease caused by STEC E. coli
O104:H4 (probable and confirmed), including 17 deaths linked to the outbreak in Germany and
occurring in the EU (including Norway) have been reported to the European Centre for Disease and
Control (ECDC). In addition, in the EU 773 cases of haemolytic uraemic syndrome (HUS) caused by
this bacterium were also reported, including 29 deaths, and linked to the German outbreak. At this
time of reporting, a further 119 suspect cases including 4 deaths are also associated with this outbreak.
In addition, outside the EU 8 cases of STEC and 5 cases of HUS, including 1 death have been reported
in the USA, Canada and Switzerland through the international health regulations (IHR), all with recent
travel history to Germany. Based on the most recently reported information (last updated on 26 July
2011), the clinical onset of the last outbreak-related case in Germany was 4 July 2011.

Shortly after the onset of the outbreak in Germany, case-control studies conducted by the Robert Koch
Institute (RKI) demonstrated that clinical disease was statistically significantly associated with the
consumption of fresh salad vegetables. The high proportion of adult women among cases, was
consistent with fresh salad vegetables as the source of infection. Later, a detailed cohort study
demonstrated an association with sprouted seeds.

A tracing back and tracing forward study showed that most of the clusters, and all of the clusters for
which there was sufficiently detailed food consumption data, could be attributed to consumption of
sprouted seeds from one producer in Germany. Investigation of the production site showed no
evidence of environmental contamination. Some employees were found to be infected, but reported
not to have become ill prior to the outbreak and hence it was concluded that they were not the source
of the contamination. This left the seeds used for the sprout production as the prime suspect vehicle of
infection; however it was not possible to identify a single seed source since different species were used
to produce the sprouted seeds, which were sold as several different mixtures.

On the 24 June, France reported to the Rapid Alert System for Food and Feed (RASFF4), a cluster of
patients with bloody diarrhoea, after having participated in an event in the Commune of Bgles near
Bordeaux on the 8 June. At the time of issue of this report, there were 2 confirmed STEC cases and 9
cases with HUS reported to ECDC, with 4 further cases of non-HUS suspected. Eleven of these
patients, 7 women and 4 men, between 31 and 64 years of age, had attended the same event in Bgles.
Infection with E. coli O104:H4 was confirmed for 12 of the 15 cases. Epidemiology studies on the
French outbreak also implicated sprouted seeds as the outbreak vehicle.

The phenotypic and genotypic characterisation of the E. coli O104:H4 indicated that the isolates from
the French and German outbreaks were common to both incidents. Hence, it was concluded that the
same strain was involved in the outbreaks both in Germany and in France, strongly indicating a
common source.

A tracing back initiated to find the common food source for both outbreaks revealed fenugreek seeds
to be common to the 2 outbreaks. The comparison of the back tracing information on the seeds from
the French and German outbreaks led to the conclusion that a specific consignment (lot) of fenugreek
seeds imported from Egypt was the most likely link between the outbreaks, although it could not be
excluded that other lots imported by the same importer/exporter might be implicated.

Data concerning the trace back and trace forward were exchanged through the RASFF, allowing the
Member States and European institutions to receive up to date information.

The actual cause or route of contamination of the seed has not been demonstrated. However, based on
epidemiological and microbiological investigations within the EU as well as from previous sprouting

4
RASFF: http://ec.europa.eu/food/food/rapidalert/index_en.htm

EFSA Journal 2011;9(10):2390 2

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

seed-related outbreaks, it is likely that contamination occurred during seed production. This part of the
investigation would need to extend beyond the point of EU import to include the site(s) of production.

The published data for STEC O104 are scarce as this is a very rare serogroup infecting humans in
Europe and globally. According to the information reported to ECDC, there were 10 reported cases of
STEC O104 infection in the EU Member States and Norway during 2004-2010 from: Austria (1 case
in 2010), Belgium (1 case in 2008), Denmark (1 case in 2008), Finland (1 case in 2010), France (1
case in 2004), Norway (1 case in 2006, 3 cases in 2009), and Sweden (1 case in 2010). Moreover, a
paediatric case of HUS which occurred in Italy in 2009 has now been associated with STEC O104, so
giving a total of 11 cases.

Five of the 10 cases between 2004 and 2010 were related to travel outside the EU; the countries of
origin of the infection being Afghanistan (2008), Egypt (2010), Tunisia (2009, 2010) and Turkey
(2009). Only 3 of the STEC O104 strains isolated from these cases were of serotype O104:H4 (in
Finland in 2010, in Italy in 2009, and in France in 2004). As for the outbreak strain, the STEC
O104:H4 strains isolated in Italy and in Finland were both positive for genetic markers of
enteroaggregative adhesion, but differed from the 2011 epidemic strain in that they were negative for
extended-spectrum beta-lactamase production. The Finnish case was travel-related, with infection
acquired in Egypt; whereas the Italian case had a recent history of travel to Tunisia. The origin of the
source of infection for the French case was not reported.

In addition to those cases reported to ECDC, a review of the scientific literature revealed that STEC
O104:H4 has been isolated twice in Germany in 2001 and once in Korea in 2005. The German isolates
differed from the 2011 outbreak strain.

E. coli O104:H4 was not isolated from any batches of the suspect fenugreek seeds. The inability to
demonstrate the presence of E. coli O104:H4 in the suspect seeds is not unexpected. It is possible that
contaminated seeds were no longer in stock when sampling took place, or even if present were
contaminated at a level which made isolation of the organism impossible. However, this does not
mean that enterobacteriaceae would not have been present in seeds and sprouted seeds. Previous
studies have shown enterobacteriaceae to be present on the surface of the tissue of the plants and that
they can also be internalised within the plant (e.g. at primary production, through irrigating with
contaminated water or application of organic fertilizer not properly treated and still containing enteric
pathogens). In this regard, it is important to underline that a negative laboratory test does not prove the
absence of a pathogen, which may be unevenly distributed within the food matrix, perhaps at low
levels. This is particularly true when dealing with seeds: the matrix is made up of particles (seeds)
which individually may become contaminated and dispersed in large lots. Similarly, favourable
physico-chemical conditions to support survival or growth may not be homogeneously distributed.

The preparation of fresh sprouted seeds seldom includes a step where bacterial contamination is
eliminated. Hence, food preparation of fresh sprouted seeds is based on the understanding that they are
sold as ready-to-eat, i.e. safe to eat as is, or following only minimal preparation. For fresh produce,
this assumes and relies on a production process which prevents contamination and an ability to detect
contamination when it occurs. These conditions have proven not to be satisfied in this case. The fact
that the sampling and bacteriological methods might fail to detect the presence of pathogens such as
STEC O104:H4 and Salmonella spp. emphasises the importance of good production and handling
practices, as the public health protection associated with such a criterion is questionable given the
difficulty of tracing pathogens in seeds.

EFSA Journal 2011;9(10):2390 3

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

TABLE OF CONTENTS
Abstract .................................................................................................................................................... 1
Summary .................................................................................................................................................. 2
Table of contents ...................................................................................................................................... 4
Background as provided by EFSA ........................................................................................................... 5
Terms of reference as provided by EFSA ................................................................................................ 5
Assessment ............................................................................................................................................... 6
1. Introduction ..................................................................................................................................... 6
2. Characterisation of the E. coli O104:H4 strain isolated in the current outbreak ............................. 7
3. Previous worldwide occurrence of E. coli O104:H4 ....................................................................... 8
4. Background on the outbreaks in the EU .......................................................................................... 9
5. Epidemiological investigations into the vehicle, source and cause of the outbreak ........................ 9
6. Microbiological investigations in food and the environment ........................................................ 12
7. Discussion ...................................................................................................................................... 14
7.1. Public health aspects ............................................................................................................. 14
7.2. Epidemiological aspects........................................................................................................ 15
7.3. Microbiological aspects ........................................................................................................ 15
Conclusions ............................................................................................................................................ 17
References .............................................................................................................................................. 19
Appendix I .............................................................................................................................................. 21
Glossary .................................................................................................................................................. 22

EFSA Journal 2011;9(10):2390 4

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

BACKGROUND AS PROVIDED BY EFSA

On the 21 of May 2011, Germany reported an ongoing outbreak of Shiga-toxin producing Escherichia
coli- bacteria (STEC56), serotype O104:H4. As of 27 July, 3126 cases of diarrhoeal disease caused by
STEC E. coli O104:H4 (probable and confirmed), including 17 deaths linked to the outbreak in Germany
and occurring in the EU (including Norway) have been reported to the ECDC. In addition, 773 cases of
haemolytic uraemic syndrome (HUS) caused by this bacterium were also reported, including 29 deaths, and
linked to the German outbreak. At this time of reporting, a further 119 suspect cases including 4 deaths are
also associated with this outbreak7.

On the 24 of June, France reported8 a cluster of patients with bloody diarrhoea, after having participated in
an event in the Commune of Bgles near Bordeaux on the 8 of June. At the time of issue of this report,
there were 2 confirmed STEC cases and 9 cases with HUS reported to ECDC, with a further 4 suspected
cases of non-HUS. Eleven of these patients had attended the same event in Bgles. Infection with E. coli
O104:H4 was confirmed for 12 of the 15 cases. Epidemiology studies on the French outbreak also
implicated sprouted seeds as the outbreak vehicle.

Both events have resulted in numerous investigations to establish the extent, vehicle, source, and cause of
this outbreak. These investigations have led to the conclusion that fenugreek seeds imported into Europe
from Egypt were the most likely source of contamination, and sprouted seeds the most likely vehicle. The
objective of this review is to provide an overview of key information that was gathered and to reference
where this information is presented in detail.

TERMS OF REFERENCE AS PROVIDED BY EFSA

The Emerging Risks unit (EMRISK) is requested to coordinate the following activities in collaboration
with the Scientific Support and Assessment Unit, Biological Hazards Unit, Biological Monitoring Unit and
Communications;

1) In collaboration with the French and German authorities, and the EU Reference Laboratory on E.
coli, prepare a Scientific Report summarising the strain characteristics, epidemiological
investigations, and analytical methods for foods linked to this outbreak. Make extensive reference
to the reports produced by the national authorities and European institutions concerning this
outbreak.

2) In collaboration with the units involved in EFSAs outbreak response, produce a discussion paper
summarising EFSAs contribution to resolving the outbreak, and highlighting lessons learnt for
EFSA, for collaboration at MS level and at the level of EU coordination. This paper will address
both EFSAs scientific work and risk communications activities undertaken in coordination with
ECDC, Member States and the European Commission.

This report only addresses the first of the terms of reference (ToR). The second ToR will be addressed in a
separate document.

5
European Food Safety Authority; Urgent advice on the public health risk of Shiga-toxin producing
Escherichia coli in fresh vegetables. EFSA Journal 2011; 9(6):2274. [50 pp.]
6
In the EU and as reflected in EFSAs work on zoonoses, Shiga-toxin producing Escherichia coli are referred to as
VTEC (verotoxin-producing E. coli), but the term STEC is used for this outbreak as it is in line with terminology used
by WHO and other organisations.
7
http://www.ecdc.europa.eu/en/activities/sciadvice/Lists/ECDC%20Reviews/ECDC_DispForm.aspx?List=512ff74f-
77d4-4ad8-b6d6-bf0f23083f30&ID=1166&RootFolder=/en/activities/sciadvice/Lists/ECDC Reviews&MasterPage=1
8
RASFF Alert Notification 2011.08428 RASFF Alert Notification 2011.0842, Available at
https://webgate.ec.europa.eu/rasff-window/portal/index.cfm?event=notificationsList

EFSA Journal 2011;9(10):2390 5

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

ASSESSMENT

1. Introduction

On the 21 of May 2011, Germany reported an ongoing outbreak of Shiga-toxin producing Escherichia
coli- bacteria (STEC910), serotype O104:H4. As of 27 July, 3126 cases of diarrhoeal disease caused by
STEC E. coli O104:H4 (probable and confirmed), including 17 deaths linked to the outbreak in
Germany and occurring in the EU (including Norway) have been reported to the European Centre for
Disease Prevention and Control (ECDC). In addition, 773 cases of haemolytic uraemic syndrome
(HUS) caused by this bacterium were also reported, including 29 deaths, and linked to the German
outbreak. At this time of reporting, a further 119 suspect cases including 4 deaths are also associated
with this outbreak 11. Further details on the German cases are available from the Robert Koch
Institute12 (RKI, 2011).

On the 24 of June, France reported13 a cluster of patients with bloody diarrhoea, after having
participated in an event in the Commune of Bgles near Bordeaux on the 8 of June. At the time of
issue of this report, there were 2 confirmed STEC cases and 9 cases with HUS reported to ECDC, with
a further 4 suspected cases of non-HUS. Eleven of these patients, 7 women and 4 men, between 31
and 64 years of age, had attended the same event in Bgles. Infection with E. coli O104:H4 was
confirmed for 12 of the 15 cases. Epidemiological studies on the French outbreak also implicated
sprouted seeds as the outbreak vehicle.

Both events have resulted in investigations to find the source of this outbreak which resulted in the
conclusion that contaminated fenugreek seeds imported into Europe from Egypt were the most likely
source. The objective of this review is to provide an overview of key information that was gathered
and to reference where this information is presented in detail. The paper covers what is known about:

- the hazard, in this case the bacteria causing the illness;

- the vehicle of infection, i.e. what people ate that made them ill;
- the source, i.e. where the bacterium came from ;
- the cause, i.e. how sufficient contamination to cause the outbreak got from the source to the
vehicle.

9
European Food Safety Authority; Urgent advice on the public health risk of Shiga-toxin producing
Escherichia coli in fresh vegetables. EFSA Journal 2011; 9(6):2274. [50 pp.]
10
In the EU and as reflected in EFSAs work on zoonoses, Shiga-toxin producing Escherichia coli are referred to as
VTEC (verotoxin-producing E. coli) but the term STEC is used for this outbreak as it is in line with terminology used
by WHO and other organisations.
11
http://www.ecdc.europa.eu/en/activities/sciadvice/Lists/ECDC%20Reviews/ECDC_DispForm.aspx?List=512ff74f-
77d4-4ad8-b6d6-bf0f23083f30&ID=1166&RootFolder=/en/activities/sciadvice/Lists/ECDC Reviews&MasterPage=1
12
http://www.rki.de/cln_169/nn_217400/EN/Home/PM__EHEC.html
13
RASFF Alert Notification 2011.084213 RASFF Alert Notification 2011.0842, Available at
https://webgate.ec.europa.eu/rasff-window/portal/index.cfm?event=notificationsList

EFSA Journal 2011;9(10):2390 6

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

2. Characterisation of the E. coli O104:H4 strain isolated in the current outbreak

For a general recent brief review of characteristics of E. coli O104:H4 that are particularly relevant to
this outbreak, see EFSA (2011a).

Germany

The outbreak strain is a Shiga toxin producing Escherichia coli (STEC) that belongs to serotype
O104:H4. Detailed information on the outbreak strain characteristics can be found at the Robert Koch-
Institut (RKI) website14. Briefly, its microbiological characteristics are as follows:

o Shigatoxin 1 (stx1) not present

o Shigatoxin 2 (stx2a) present
o Intimin (eae gene) not present
o Enterohemolysin (hlyA) not present
o EAggEC (enteroaggregative E. coli) virulence plasmid:
aatA present (ABC-transporter protein gene);
aggR present (master regulator gene of virulence-plasmid genes);
aap present (secreted protein dispersin gene);
aggA present (AAF/I-fimbral subunit-gene);
aggC present (AAF/I-fimbral operon-gene);
o Multi-locus sequence typing (MLST) Sequence Type: ST678 (adk 6, fumC6, gyrB 5, icd 136,
mdh 9, purA 7, recA 7);
o Antimicrobial resistance profile: resistant to ampicillin, amoxicillin/clavulanic acid,
piperacillin/sulbactam, piperacillin/tazobactam, cefuroxime, cefuroxime-axetil, cefoxitin,
cefotaxime, cetfazidime, cefpodoxime, streptomycin, nalidixic acid, tetracyclin,
trimethoprim/sulfamethoxazol;
o The strain carries plasmid-borne blaCTX-M-15 and a blaTEM-1 genes;
o The macrorestriction-pulsed field gel electrophoresis (PFGE)-pattern (XbaI) of the current
STEC O104:H4 is different compared to HUSEC 041 (Robert Koch Institute, Personal
Communication);

In conclusion, the STEC O104:H4 outbreak strain shows an unusual combination of virulence factors
of STEC and EAggEC which has only been reported for a group of strains of serotype O111:H2
(Morabito et al., 1998) isolated in France during an outbreak of HUS (Boudailliez et al., 1997). As in
the present outbreak, the association of the O111 strains with severe disease supports the view that this
unusual combination of virulence factors might confer a very high degree of virulence. It is also
noteworthy that the outbreak described by Boudailliez et al. (1997), was ascribed to a person-to-
person transmission, without any evidence of foodborne transmission.

Whole genome sequencing data confirms that the German outbreak strain is an EAggEC which can be
distinguished from those of other O104:H4 strains because it contains a prophage encoding Shiga
toxin 2 and a distinct set of additional virulence and antibiotic-resistance factors which have been
acquired by horizontal genetic exchange (Rasko et al., 2011).

14

http://www.rki.de/cln_160/nn_217400/EN/Home/EHECO104,templateId=raw,property=publicationFile.pdf/EHECO1
04.pdf

EFSA Journal 2011;9(10):2390 7

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

France

The strain of E. coli O104:H4, isolated from 5 HUS cases, all of whom consumed sprouted seeds
served at an event in Bgles (near Bordeaux, South-west France) on 8 June 2011, exhibits the
following characteristics (ANSES, 2011):

o The strain contains the stx2 and aggR genes, but not the genes coding for intimin (eae),
haemolysin A (hlyA) and EAST1 toxin (astA);
o The antimicrobial susceptibility pattern of the strain is as follows: resistant to ampicillin,
cefotaxime, ceftazidime, cotrimoxazole, streptomycin, trimethoprim, sulfamethoxazole,
tetracyclin, and nalidixic acid. Sensitive to chloramphenicol, ciprofloxacin, gentamicin,
imipenem, and kanamycin;
o The strain contains extended-spectrum beta-lactamase (ESBL) blaCTX-M-15 (group 1) gene and
the penicillinase blaTEM gene.

Thus, information on the microbiological characterisation of the isolates implicated in the French
outbreak indicate that many characteristics (stx2 positive, eae negative, hlyA negative, multi-resistance
pattern to antimicrobials) are common with the German outbreak strain. In addition, the 2 molecular
techniques (Repetitive sequence based Polymerase Chain Reaction (Rep-PCR) and pulsed-field gel
electrophoresis (PFGE)) used to compare the outbreak strain in France and isolates of E. coli O104:H4
from 2 imported cases in France linked to the E. coli O104:H4 outbreak in Germany in May and June
2011, showed that the isolates were indistinguishable by these methods (Gault et al., 2011).

3. Previous worldwide occurrence of E. coli O104:H4

ECDC and EFSA reported on the previous occurrence of STEC O104:H4 (ECDC, EFSA 2011). The
text below is extracted from this document.

The published data for STEC O104 is scarce as this is a very rare serogroup in humans in Europe and
globally. According to the information reported to ECDC, there were 10 reported cases infected with
STEC O104 in the EU Member States and Norway in the period 2004-2009. These STEC O104 cases
have been reported from Austria (1 case in 2010), Belgium (1 case in 2008), Denmark (1 case in
2008), Finland (1 case in 2010), France (1 case in 2004), Norway (1 case in 2006, 3 cases in 2009),
and Sweden (1 case in 2010) (ECDC and EFSA, 2011). Moreover, a paediatric case of HUS which
occurred in Italy in 2009 has now been associated with STEC O104 (Scavia et al., 2011), so giving a
total of 11 cases.

Based on the known data, 56% of the cases were male and the age ranged from <1 year to 76 years.
Two of the cases (18%) developed HUS. Five of the cases were travel related; the countries of origin
of the infection being Afghanistan (2008), Egypt (2010), Tunisia (2009 and 2010) and Turkey (2009).
Only 3 of the STEC O104 strains isolated from these cases were of serotype O104:H4 (in Finland in
2010, in Italy in 2009, and in France in 2004). Similar to the outbreak strain, the STEC O104:H4
strains isolated in Italy and in Finland were both positive for genetic markers of enteroaggregative
adhesion, but negative for ESBL (extended-spectrum beta-lactamase) production. As mentioned, the
Finnish case was travel-related, with the infection acquired in Egypt; whereas the Italian case had a
recent history of travel to Tunisia (Scavia et al., 2011). The origin of the source of infection for the
French case was not reported.

From a review of the scientific literature, STEC O104:H4 has been isolated twice in Germany in 2001
(Mellmann et al, 2008) and once in Korea in 2005 (Bae et al, 2006). The case in Korea was a 29-year
old woman who developed HUS. The source of the infection was unknown. The STEC O104:H4
from Germany, with a MLST ST678, was isolated from an HUS case (Mellmann et al., 2008). Like

EFSA Journal 2011;9(10):2390 8

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

the current outbreak strain, the isolate was stx2-positive, eae-negative, and positive for
enteroaggregative adhesion and EAggEC genetic markers. However, it was negative for ESBL
production15, and some other differences with the outbreak strain were reported (Bielaszewska et al.,
2011).

4. Background on the outbreaks in the EU

This section complements the ECDC/EFSA rapid risk assessment16 published on 29 June, 2011, and
the ECDC rapid risk assessment update published on 8 July17.

As of 27 July, 3126 cases of STEC caused by E. coli O104:H4 (probable and confirmed), including 17
deaths in the EU (including Norway), linked to the outbreak in Germany, have been reported to the
ECDC. In addition, in the EU 773 cases of HUS have been reported, including 29 deaths, also linked
to the German outbreak18. A further 119 cases of STEC and 4 deaths are suspected at this time of
reporting. Switzerland has so far reported 5 non-HUS STEC cases through the international health
regulations (IHR), the United States reported 5 HUS cases including 1 death and 2 non-HUS cases, all
with recent travel history to Germany. Canada has also reported 1 non-HUS case with travel history to
Germany.

As based on the last reported information (26 July 2011), the clinical onset of the last outbreak-related
case in Germany was the 4 July 201119.

On the 24 June, France reported to the RASFF20 a cluster of patients with bloody diarrhoea, after
having participated in an event in the Commune of Bgles near Bordeaux on the 8 June. At the time of
issue of this report, there were 2 confirmed STEC cases and 9 cases with HUS reported to ECDC, with
4 further cases of non-HUS suspected. Eleven of these patients, 7 women and 4 men, between 31 and
64 years of age, had attended the same event in Bgles. Infection with E. coli O104:H4 was confirmed
for 12 of the 15 cases. Epidemiology studies on the French outbreak also implicated sprouted seeds as
the outbreak vehicle.

An important development since 29 June comes from the results of the screening of children and staff
in a school in Germany where 3 cases of HUS and a further 4 cases of STEC related diarrhoea had
been identified. As most recently reported (5 August 2011)21, the epidemic strain of E. coli was
isolated from 22 of the 30 children tested, indicating a high level of occurrence of asymptomatic
carriage. Three teachers at the school were found to be asymptomatic carriers of STEC as were 2
employees of the catering company supplying the school.

5. Epidemiological investigations into the vehicle, source and cause of the outbreak

Early after the onset of the outbreak in Germany, case-control studies conducted by the RKI
demonstrated that the onset of clinical disease was linked to the consumption of fresh salad vegetables.
Circumstantial evidence, such as the high proportion of adult women among the ill, was also

15
http://www.ehec.org/index.php?hid=43&lang=de&pid=HUSEC
16
http://ecdc.europa.eu/en/publications/Publications/2011June29_RA_JOINT_EFSA_STEC_France.pdf
17
http://www.ecdc.europa.eu/en/publications/Publications/110712_TER_Risk_Assessement_Ecoli.pdf
18
http://www.ecdc.europa.eu/en/activities/sciadvice/Lists/ECDC%20Reviews/ECDC_DispForm.aspx?List=512ff74f-
77d4-4ad8-b6d6-bf0f23083f30&ID=1166&RootFolder=/en/activities/sciadvice/Lists/ECDC Reviews&MasterPage=1
19
http://www.rki.de/cln_110/nn_205760/DE/Content/InfAZ/E/EHEC/Info-
HUS,templateId=raw,property=publicationFile.pdf/Info-HUS.pdf
20
RASFF: http://ec.europa.eu/food/food/rapidalert/index_en.htm
21
http://www.kreis-paderborn.org/kreis_paderborn/presse/2011/entries/erster-todesfall-ehec.php

EFSA Journal 2011;9(10):2390 9

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

consistent with vegetables being the vehicle of infection. Later, a detailed cohort study demonstrated
an association with sprouted seeds22.

In an attempt to identify the food source through microbiological examination, the responsible
competent authorities of the German Federal States performed risk-based inspections of clusters
(where a common meal or place of eating was identified) and took samples from foods and the
environment at clusters, suppliers, whole sale markets, and at retail23. So far (26 July, 2011), none of
the food samples taken during this monitoring activity, for which cross-contamination (from infected
persons in the household) could be excluded, has been found positive for the STEC serotype O104:H4
(BfR, 2011b).

Consequently, it was decided that a co-ordinated initiative to identify the food source and stop the
outbreak was needed. For this purpose, the German Task Force EHEC (Enterohaemorrhagic
Escherichia coli) was established by the German Federal Ministry of Food, Agriculture and Consumer
Protection (BMELV). Members of the Task Force EHEC were experts from the Federal States in
Germany, from the Federal Office of Consumer Protection and Food Safety (BVL), the Federal
Institute for Risk Assessment (BfR), the RKI and EFSA scientists. The Task Force EHEC was hosted
in the crisis centre at the BVL (Task Force EHEC, 2011).

In a first phase, 2 different strategies were followed to identify the STEC contaminated food. The first
approach examined 5 clusters in Germany (associated with hotels, restaurants or canteens), which
were prioritized due to the availability of precise information on the food intake of the affected cases
(e.g. tour parties, family celebrations etc.). A detailed trace back analysis was conducted for all salad
ingredients (including sauces and dressings etc.) and toppings that had been served to the customers.
The second approach was a trace forward analysis of the supply chains of a horticultural farm in
Lower Saxony (Establishment A) in order to find out if any of the clusters were linked to sprouting
seeds originating from there. The production site was suspected early on by the local authorities as a
possible source of infection for several clusters. Furthermore, STEC O104:H4 was isolated from the
faeces of employees of Establishment A, including 2 asymptomatic cases.

A total of 41 clusters were identified for which there was sufficient information on food intake in
order to carry out the tracing. The analysis of the sprouted seed supply chains showed that for all 41
clusters, there was a common link to sprouted seeds from Establishment A (the trace forward
approach). The results of the detailed trace back analysis at 5 clusters confirmed the findings of the
trace forward analysis. It was shown that sprouted seeds from Establishment A had been delivered to
all 5 clusters. Thus, the epidemiological investigations provided strong evidence that sprouted seeds
from Establishment A were the vehicle of infection of the German STEC O104:H4 outbreak24.

Further epidemiological investigations revealed that the disease occurrence coincided with
consumption of either of 2 different sprouted seed mixtures, the Keimspross-Mischung / Milde
Mischung (germ sprout or mild blend) or the Wrz-Mischung (spicy blend). The mild blend
contained 4 and the spicy 3 different kinds of sprouted seeds. Only lentil and fenugreek sprouts were
present in both mixtures. Whereas lentil sprouts were also used in other sprouted seed mixtures,
fenugreek sprouts were used for Keimspross-Mischung / Milde Mischung and Wrz- Mischung
only25. Hence, fenugreek sprouts were almost exclusively used for the 2 mixtures and only small
amounts were sold unmixed.

The sprouting procedure of fenugreek seeds was examined in detail. Fenugreek seed sprouting was
initiated twice per week, every week. The sprouting process took 3.5 days. After sprouting, the
22

http://www.rki.de/cln_117/nn_217400/EN/Home/EHEC__Report,templateId=raw,property=publicationFile.pdf/EHEC
_Report.pdf
23
www.bvl.bund.de/taskforce_en
24
www.bvl.bund.de/taskforce_en
25
www.bvl.bund.de/taskforce_en

EFSA Journal 2011;9(10):2390 10

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

fenugreek seeds were mixed with other sprouted seeds to generate the Keimspross-Mischung / Milde
Mischung and the Wrz-Mischung. It was assumed that fresh sprouted seeds were eaten after a
maximum of 14 days following their production. Starting with the known dates of exposure of the
cases from the selected 5 clusters and the delivery dates for the sprouted seeds to the place of exposure
(hotels, restaurants or canteens) the possible production dates at Establishment A were assessed. A
comparison with the total epidemiological curve of the outbreak confirmed the isolated production
cycles as consistent with the whole outbreak. Establishment A used lot nos. 48088 and 8266 of
fenugreek seeds for their production during April to mid-May 2011. The earlier one (lot no. 48088)
was totally used up when sampling by the local authorities took place at Establishment A (Task Force
EHEC 2011). Both lots were traced back to a single importer in Germany, who imported the seeds
from Egypt.

Three of the personnel of Establishment A developed diarrhoeal symptoms consistent with an STEC
infection, with earliest onset of disease on 6 May 2011. Two of the clinically ill staff worked part-
time, but included days when fenugreek seeds and sprouts were handled. The outbreak strain STEC
O104:H4, however, could only be detected in 1 of the 3 staff members. Moreover, within the course of
the investigation of the environment of Establishment A, 2 more staff members were identified
shedding the outbreak strain, both without any diarrhoeal symptoms. It was therefore questioned
whether the personnel could be the origin of the infection. While the onset of disease of the first staff
member who became ill occurred early in the outbreak, it occurred within days of the first occurrences
of illness in other people having eaten sprouted seeds originating from Establishment A. Personnel
from Establishment A were given a weekly package of sprouted seeds for private use, however, this
may explain the early exposure of personnel from Establishment A, but does not support that they
introduced the contamination into the production chain (BfR, 2011a). This does not exclude that they
may have amplified the spread of the infection in Establishment A, or that an unidentified carrier or
infected individual was responsible.

It should be noted that the tracing back investigations in Germany were done for all the types of seeds
used in Keimspross-Mischung / Milde Mischung and Wrz-Mischung. Additionally, an inventory
of all sprouted seed producers in Germany was established to gather information on the simultaneous
use of the suspected seeds. Hence, intensive sampling and microbiological analysis of seeds and
sprouts were performed. In the case of the occurrence of additional clusters, this information could
then be used to refine the identification of the source of infection, as well as, if seed contamination
were detected, a more rapid recall of all material.

The occurrence of the outbreak in France, without any direct link to the outbreak in Germany, but
caused by the same STEC strain as in the German outbreak (Gault et al., 2011) and the fact that
sprouting seeds were also strongly suspected as the source of the French outbreak (ANSES, 2011),
considerably strengthened the hypothesis that, among possible sources, seeds were the most likely
vehicle of the contamination. A cohort study of those attending the event was initiated to identify
which specific foods were significantly associated with the onset of clinical disease in the French
cluster and preliminary results revealed a significant association between consumption of sprouted
seeds (fenugreek, rocket and white mustard) and infection. In addition, the study showed that none of
the food handlers at the event or people who served the dishes had recently travelled to Germany, and
none of the guests had recent contact with someone coming back from Germany26.

Following this outbreak in France, EFSA was asked by the European Commission to support the
Member States and coordinate activities to investigate the source of the outbreaks in France and
Germany in order to allow risk managers to take the necessary and appropriate risk mitigating
measures towards prevention of further outbreaks. For this purpose EFSA set up a task force
composed of experts from the European Commission, affected EU Member States, the ECDC, WHO
and FAO, as well as EFSA staff members. Data concerning the trace back and trace forward were

26
http://www.invs.sante.fr/behweb/2011/03/pdf/n3.pdf

EFSA Journal 2011;9(10):2390 11

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

exchanged through the RASFF27, allowing the MS and European institutions to receive up to date
information. The study established one common link between all 41 of the identified German clusters
and the French cluster through the import of fenugreek from Egypt by a single importer in Germany.
The Importer had bought several large lots of these seeds over the last 48 months and distributed more
than one lot at the same time.

The technical report published by this ESFA task force on 5 July (EFSA, 2011b) concluded that a
specific lot (no. 48088) of fenugreek seeds imported from Egypt was the most likely common link;
however the implication of other lots cannot be excluded. The exact point of contamination in the food
chain has not been established, but took place at some point prior to leaving the Importer. The
contamination of seeds with the STEC O104:H4 strain reflects a production or distribution process
which allowed for contamination with faecal material of human and/or animal origin. Where exactly
this took place has not been established. Typically such contamination occurs during production at the
farm level, due to the use of natural fertiliser (e.g. animal manure or slurry) or contaminated irrigation
water. In general, such types of contamination lead to bigger (e.g. transnational) outbreaks, while local
contamination during storage or transport is more likely to lead to more confined outbreaks.

6. Microbiological investigations in food and the environment

Sample selection and preparation of seeds

A sample size of 50 g (rather than the 25 g required for other matrices) is recommended to increase the
sensitivity of the subsequent diagnostic testing (ANSES, 2011). The seeds are ground then suspended
in buffered peptone water and incubated at 37C for 24 h. 5 ml of the suspension (including some
ground seed debris) are vortexed to loosen bacteria adhering to the seed debris. The suspension is
centrifuged and then the supernatant is used for DNA extraction (ANSES, 2011).

Detection and identification of STEC O104:H4

The EU Reference Laboratory (EU-RL) for E. coli has proposed a method for the detection and
identification of STEC O104:H4 in food by Real Time PCR28. The method aims at extending the
scope of the CEN/ISO Technical Specification 13136 for the detection, in food and feed, of STEC
belonging to the 5 serogroups most commonly involved in severe human disease (O157, O103, O111,
O145, O26).

According to the approach described in the CEN/ISO TS 13136, the sample is subjected to the Real-
Time PCR screening for the presence of the stx and eae genes followed by, in the stx-positive and eae-
negative samples, the detection of the O104 and H4 antigen-associated genes. Finally, a confirmation
step aiming at isolating the STEC strain responsible for the positive PCR reactions is performed. It is
noted that if a sample is co-infected with an eae positive strain and O104, the latter would be missed
by not testing for it. The Real Time PCR screening is performed on DNA extracted from enrichment
cultures obtained by incubating a portion of the food sample in an appropriate medium, as described in
the CEN/ISO TS 13136. The antimicrobial resistance characteristics of the outbreak strain are
exploited for the isolation by plating PCR-positive enrichment cultures onto media supplemented with
antibiotics. The suspected colonies are tested for stx genes and then for the serotype-specific genes
wzxO104 and fliCH4.

27
RASFF: http://ec.europa.eu/food/food/rapidalert/index_en.htm
28
http://www.iss.it/vtec/work/cont.php?id=152&lang=2&tipo=3

EFSA Journal 2011;9(10):2390 12

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

Microbiological investigations

After identification of Establishment A and sprouting seeds as the food vehicle for the outbreak,
further investigations were carried out to identify where STEC O104:H4 may have been introduced
into the production and/or distribution chain (the cause). This included studies to assess whether the
contamination could have come from the environment (esp. through water), the seeds used for
sprouting, personnel or by cross-contamination. Microbiological analysis of the water has thus far
proved negative (BfR, 2011a). Several batches of the different seed varieties that were used by
Establishment A and were still present at the production site were sampled for microbiological
analysis, but none were positive for STEC O104:H4. Exhaustive inspection of all parts of the supply
chain showed no likely source of infection after the sprouting at Establishment A.

A recent overview of investigations carried out at MS level and transmitted through the RASFF,
shows that a total of 10392 samples were tested, 41 samples were found to be stx positive, and 8 STEC
O104 positive (i.e. 5 from food, 1 from irrigation and processing water, and 2 from the environment,
see Appendix I). Thus far E. coli O104:H4 has not been isolated from foods where cross-
contamination from cases or infected household members could not be excluded. In total, there have
been 5 reported isolations from foods, from locations linked to cases. One was from a small remaining
part of a cucumber which had been collected from a communal organic waste bin. A second sample
was from leftover seed sprouts (one of the mixtures with fenugreek seeds produced by Establishment
A) in an open package that was collected from a communal waste bin. In both cases, the vegetable
leftovers were taken by household members out of the waste bins. A third sample was from a bell
pepper. The other 2 samples were from raw and cooked salmon. These last 3 samples were all
obviously contaminated during the incubation period in the laboratory carrying out the analysis
(personal communication).

EFSA Journal 2011;9(10):2390 13

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

7. Discussion

7.1. Public health aspects

STEC strains with similarities to the outbreak strain (serotype O104:H4, Stx2 production, positive for
the genetic markers of enteroaggregative adhesion) have been sporadically isolated in Europe since
2001. The O104:H4 outbreak strain is a typical EAggEC strain that has acquired bacteriophages
encoding for Shiga toxin (Scheutz et al., 2011; Rasko et al., 2011). If it proves to be of human origin,
as for the other EAggEC, it is unlikely that it would have maintained itself in the EU population for
such a long time without being recognised as causing severe human health consequences. Instead, it
has appeared causing sporadic cases typically with a recent travel history outside the EU. Therefore,
we could assume that the O104:H4 clone, during the intervals between the sporadic cases, has
maintained itself outside the EU. If it is a pathogen with a human reservoir (rather than zoonotic), it is
conceivable that such a reservoir may be present in a population where the acquired immunity towards
EAggEC might limit the severity of the clinical manifestation of the infection, and/or the surveillance
systems in place may not be adequate to reveal the disease prevalence. If this is correct, the strain may
be imported into the EU by returning travellers or migrants, resulting in a few sporadic cases. The
Finnish and Italian cases being linked with North Africa support this hypothesis. However, the
zoonotic potential of this strain has yet to be characterized, so it remains possible that reservoirs of this
pathogen could be maintained in both wild and domestic animals, including those within the EU.

This outbreak, originating from contaminated seeds rather than, sporadic individual cases, allowed the
agent to enter into the food chain and resulted in significant exposure and a major outbreak, which had
not been the case for previous sporadic cases reported in the EU. While the epidemiological evidence
strongly points to contaminated sprouting seeds as the main vehicle for the outbreaks observed in
Germany and in France, as well as the related cases observed outside Germany, but with previous
recent travel to Germany; the most recent cases observed in Germany and in other countries may result
from vehicles or modes of transmission other than directly from contaminated seeds and sprouts
thereof. Since in Germany thousands of people were infected, it is not surprising that secondary cases
occur (Hauri et al., 2011; Kuijper et al., 2011) (see also Aldabe et al., 2011, for reporting on household
transmission in the French outbreak), or even secondary clusters when infected food handlers are
involved (see Grieg et al., 2007 for a review of outbreaks where food handlers are implicated). In
addition, it is well known that enteric pathogens for which there is a high risk of infection from low
levels of exposure cause outbreaks with person-to-person transmission, as occurs for shigellosis and,
for secondary cases, for STEC O157 (Snedeker et al., 2009).

Given the considerable number of asymptomatic carriers, person-to-person transmission is likely to

play a role in the spread of the outbreak, resulting in the recognition of apparently sporadic cases
without any identified link to a known case, vehicle or cluster. This mode of transmission may result
in outbreaks in institutions such as day-care centres, schools or nursing homes.

The limited information available on the infectivity makes it difficult to quantify the role that person-
to-person transmission may play in the future. From a bacteriological perspective, a key issue that
needs to be considered is that the outbreak strain is an enteroaggregative E. coli that produces Stx, and
not a "true" STEC. However, EAggEC are also known to have caused foodborne outbreaks in Europe:
links with restaurants have been frequently reported as well as a possible role of food handlers, as in a
recent outbreak in Italy (Scavia et al., 2008).

In addition, if such asymptomatic food handlers contaminate food, as probably happened in a school in
Germany, new clusters may appear with different vehicles of infection29.

In summary, following a dramatic decrease in the number of new cases during 28 June 7 July, new
cases and clusters continued to occur despite the identification of the original vehicle (sprouting seeds)

29
http://www.kreis-paderborn.org/kreis_paderborn/presse/2011/entries/erster-todesfall-ehec.php

EFSA Journal 2011;9(10):2390 14

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

and the communication of public health advice to avoid consuming sprouted seed. New cases and
clusters were most probably due to secondary STEC infections, through different vehicles or by
person-to-person transmission. The emerging evidence of a substantial proportion of subclinical
infections is important as they are generally unrecognised and unreported, and may contribute to
further transmission of the infection. This may have an impact on the epidemiology of the disease and
should be taken into account when defining public health measures to prevent further spread.

Regarding seeds already imported, trace-forward activities were advised. During the period that these
activities are under way, the advice to consumers not to grow sprouted seeds for consumption and not
to eat sprouts or sprouted seeds, unless they have been cooked thoroughly30, was maintained. Once
completed the trace forward operation, and the implicated lot(s) removed from the market, such advice
should be revised.

7.2. Epidemiological aspects

The investigations conducted thus far have demonstrated that there was a single hazard (STEC
O104:H4) involved in the outbreak. In Germany, the clusters were all linked to sprouted seeds
produced by one establishment. The tracing back demonstrated a probable common source for the
German and French outbreaks, a particular lot of fenugreek seeds imported from Egypt.

The cause of the outbreak i.e. how, when and where the seeds became contaminated has not (yet) been
demonstrated. Assuming this took place during the primary production process, this part of the
investigation would need to extend beyond the point of EU import to include the site(s) of production.

7.3. Microbiological aspects

The inability thus far to demonstrate the presence of E. coli O104:H4 in the suspected seeds is not
unexpected. It is, of course, possible that contaminated seeds were no longer in stock when sampling
took place. However, it is important to realise that a negative laboratory test does not prove the
absence of a pathogen. This is because sample-based sampling always involves some statistical
uncertainty, where there is a risk for not collecting samples of the contaminated lots or part of a lot
that is contaminated. This risk increases when the contamination occurs at a very low prevalence (i.e.
few units are contaminated). Besides the statistical uncertainty, the test sensitivity plays a role for
detecting a positive sample. The test sensitivity is a measure of how good a detection method is to
detect microorganisms in a contaminated sample. The rule is that no method is 100% effective. So
even though samples of contaminated material have been collected, it may be that the amount of
pathogens in the sample is so low that the diagnostic method is unable to detect them.

The uncertainties described above may be particularly true when dealing with seeds, because the
matrix is made up of particles, which may become singularly contaminated and dispersed in lots of
large dimensions. In addition, the contamination may be heterogeneously distributed within the food
matrix as well as, and maybe as a result of, varying physico-chemical conditions for supporting the
survival or growth of pathogens. Previous studies have shown enterobacteriaceae to be present on the
tissue surface of plants as well as internalised within various the plant tissues including seeds. This
may happen at primary production, through irrigating with contaminated water or application of
untreated or insufficiently treated fertilizer still containing enteric pathogens) (EFSA, 2011a).

The difficulties of detecting STEC in seeds and sprouts thereof with the applied sampling plans and
analytical methods when the contamination is low and unevenly distributed have been observed in

30
http://www.efsa.europa.eu/en/topics/topic/ecolioutbreak2011.htm

EFSA Journal 2011;9(10):2390 15

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

previous outbreaks. For example, in an outbreak of Salmonella enterica serotype Enteritidis phage
type 4b associated with bean sprouts in the Netherlands in 2000 (Van Duynhoven et al., 2002), it was
noted that the outbreak occurred even though at the implicated production facility, every new batch of
seeds and samples from each batch of sprouted seeds harvested each day was routinely tested for
Salmonella spp.

The proposed molecular analytical method is characterised by a high sensitivity. However, as seeds
are generally contaminated at very low level (0.1 to 1.8 cfu/g as previously assessed for Salmonella
(Liao and Fett, 2003)), test portions would then need to weigh more than the recommended 50g, which
may not be the case for leftover samples collected during outbreak investigation. In addition, bacteria
on or in seeds may undergo stress induced by very low water activity and therefore exhibit low
culturability during the enrichment phase which precedes the PCR step (i.e. viable but non-culturable
(VBNC) state).

Furthermore, seeds, as a biological matrix, present particular characteristics (e.g. very low water
activity, biofilm etc.) which must be considered when defining the analytical phases preceding the
PCR step. The method proposed by the EU-RL includes an annex specific for seed testing. However,
it has been recently suggested, elaborating from methods typically used in phytopathology for testing
for bacterial pathogens from seeds, to perform a preliminary step of seed soaking, to facilitate the
revivification of stressed bacteria (ANSES, 2011). Further investigations are still needed to determine
an optimal test portion of seeds, fully define preliminary analytical steps and assess the impact on the
sensitivity of the test method. Potential growth and PCR-inhibitory effects of the seeds may also have
to be taken into account.

Work has been initiated by the EU-RL to evaluate the test methods for seeds in use by coordinating a
dedicated working group and organising an inter-laboratory study on seeds naturally contaminated
with STEC other than O104. Seed lots naturally contaminated with STEC represent a valuable
material for conducting studies to evaluate and validate specific laboratory methods. Indeed, it is very
difficult to define spiking strategies that could reliably mimic the contamination naturally occurring in
field conditions.

According to the protocol for seed testing proposed by the EU-RL, the reporting of a negative result or
a presumptive positive result after the first screening step (detection of stx genes) will require 26-28
hours. The isolation of STEC O104:H4 from PCR-positive enrichment samples could require 48-72
hours. It is also possible that the positive results obtained in the PCR screening step, both for STEC in
general or specifically for STEC O104, are not confirmed by the isolation of the STEC strain. The
meaning of such a result will depend on the epidemiological setting driving the sampling: in a crisis
setting and with samples presumably related to an outbreak.

EFSA Journal 2011;9(10):2390 16

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

CONCLUSIONS

Description of the outbreak

Initial epidemiological studies of the German outbreaks indicated fresh salad vegetables as the
probable vehicle of infection. Further, more detailed studies implicated fresh sprouting seeds.

Tracing forward and backward the dissemination of the implicated sprouting seeds demonstrated that
all of the clusters for which there was sufficient information, could be linked to a single sprouted seed
producer in Germany.

No evidence for environmental contamination was found at the premises of the sprouted seed
producer. Employees found to be infected were not thought to be the origin of the contamination due
to the comparative dates of onset of illness with those of consumers. As a consequence, and taking
into account experience from other outbreaks of enteric pathogens through consumption of
contaminated sprouting seeds, contaminated seeds were hypothesized as being the most likely source.
Fenugreek sprouts were mostly sold as mixtures of sprouts, so that at this stage, it was not possible to
identify the specific sprout (and seed) type responsible.

The STEC strain responsible (O104:H4) for both the German and French outbreaks was found to be
indistinguishable, strongly indicating that there was a common source for both outbreaks.

The comparison of the back tracing information from the French and German outbreaks led to the
conclusion that a specific lot of fenugreek seeds imported from Egypt was the most likely source of
both outbreaks, although it could not be excluded that other lots imported by the same importer from
the same exporter may be implicated.

The cause of the outbreak, i.e. how, when and where the seeds became contaminated has not (yet)
been demonstrated. Assuming this took place during the primary production process, this part of the
investigation would need to extend beyond the point of EU import to include the site(s) of production.

Prior to the 2011 outbreak, STEC O104 was a very rare serogroup in humans in the EU and
worldwide. Sporadic cases in the EU have been linked to travel to North Africa, the Middle East and
Central Asia.

Implementation of optimised testing of seeds and sprouted seeds

Work has been initiated by the EU-RL to evaluate the test methods for seeds in use by coordinating a
dedicated working group and organising an inter-laboratory study.

Prevention of seed contamination during production and distribution

The preparation of fresh sprouted seeds does not always include a step where pathogen contamination
is eliminated. The consumption of fresh sprouted seeds however, is based on the understanding that
they are sold as ready to eat. For fresh produce this assumes and relies on a production process which
prevents contamination and ability to detect contamination when it occurs. These conditions have
proven not to have been met in the case of the 2011 outbreaks.

Previous food-borne outbreaks due to sprouting seeds have led public authorities in some countries
(FSANZ, 2010) to carry out risk analyses to identify critical steps and effective measures in the

EFSA Journal 2011;9(10):2390 17

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

production and processing of sprouted seeds to prevent contamination of the food chain (EFSA-Q-
2011-0087731).

Prevention of further exposure to the lots of contaminated seeds and sprouts thereof

In the short term all efforts were focused on the prevention of any further exposure of the consumer to
sprouts of seeds from the lots of concern. EFSA and ECDC strongly recommended advising consumers not
to grow sprouted seed for their own consumption and not to eat sprouted seeds unless they have been
cooked thoroughly (ECDC and EFSA, 2011).

Furthermore, it was important that a trace back investigation be initiated on the incriminated lot(s) of
fenugreek seeds in the country from which they were exported to the EU. In addition, Member States and
third countries were encouraged to initiate or complete forward tracing of companies receiving the suspect
lot(s).

Prevention of secondary infections via food and through person-to-person transmission

To complement public health measures regarding food handling, information to the public should
stress the need for proper hand washing, in addition to the specific advice related to sprouted seeds.

Since there is evidence of asymptomatic carriers of STEC in humans, screening of humans involved in
food handling is relevant. The monitoring and/or exclusion of STEC carriers from food handling
should be considered as a mitigation option.

It is extremely important to further strengthen epidemiologic surveillance of human cases. Germany

and France have developed protocols to thoroughly investigate the likely mode of transmission of new
cases. The ECDC will be compiling results from these investigations using the EPIS platform32.

31
http://registerofquestions.efsa.europa.eu/roqFrontend/?wicket:interface=:1::::
32
http://www.ecdc.europa.eu/en/activities/epidemicintelligence/Pages/Activities_EpidemicIntelligence.aspx

EFSA Journal 2011;9(10):2390 18

 E. coli (STEC) O104:H4 2011 outbreaks in Europe

GLOSSARY
EAggEC Enteroaggregative Escherichia coli

EHEC Enterohaemorrhagic Escherichia coli

ESBL Extended-spectrum beta-latamase

EU-RL European Union Reference Laboratory

HUS Hemolytic-uremic syndrome

PCR Polymerase chain reaction

RASFF Rapid Alert System for Food and Feed

STEC Shiga toxin-producing Escherichia coli (synonym for VTEC)

VTEC Vero cytotoxin-producing Escherichia coli (synonym for STEC)

EFSA Journal 2011;9(10):2390 22

Rising Food Prices
CAUSES AND CONSEQUENCES
 EXECUTIVE SUMMARY

Recent steep price increases of major crops (cereals, oilseeds) were triggered by a combination of
production remaining somewhat below trend and strong growth of demand.

A low and declining level of stocks has added to the price rise, as has probably a significant
increase in investments in agricultural derivative markets.

The OECD-FAO Agricultural Outlook expects prices to come down again, but not to their
historical levels. On average over the coming ten year period, prices in real terms of cereals, rice
and oilseeds are projected to be 10% to 35% higher than in the past decade.

The acute price hike adds to inflationary pressures in developed countries. Poor consumers in
developing countries, and food importing developing countries overall, will have to spend an
even higher share of their limited income on food.

In the short term, humanitarian aid is required, where appropriate in the form of cash or vouchers
so as to strengthen, rather than undermine domestic markets in recipient countries.

In the medium term, there is a need to foster growth and development in poor countries, to
improve the purchasing power of food buyers. Agricultural trade policies require further reform
in order to ensure an effective supply response. Investments in productivity growth, particularly
in less developed countries, should also strengthen the supply side of global agriculture. On the
demand side, policies that encourage increased production and use of biofuels from agricultural
commodity feed stocks warrant review.

Introduction

World prices of wheat, coarse grains, rice and oilseed crops all nearly doubled between the 2005 and
2007 marketing years and continued rising in early 2008. These increases in agricultural commodity prices
have been a significant factor driving up the cost of food and have led to a fuller awareness and a
justifiably heightened concern about problems of food security and hunger, especially for developing
countries.

The causes of this price spike are complex and due to a combination of mutually reinforcing factors,
including droughts in key grain-producing regions, low stocks for cereals and oilseeds, increased feedstock
use in the production of biofuels, rapidly rising oil prices and a continuing devaluation of the US dollar, the
currency in which indicator prices for these commodities are typically quoted. This turmoil in commodity
markets has occurred against the backdrop of an unsettled global economy, which in turn appears to have
contributed to a substantial increase in speculative interest in agricultural futures markets.

Tight market conditions for essential agricultural commodities pose policy challenges for national
governments as well as for international organisations. In order to take the right policy decisions, we need
to understand what caused the current price spike, what the implications may be for prices and price
volatility in the future, and how various countries and members of society may be affected. This note aims
to improve this understanding and thereby to contribute to sound policy formulation.

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 2

Are food price hikes unprecedented?
The commodity price developments witnessed recently are certainly unusual when viewed from the
perspective of the last decade or so, but less so when seen in a longer historical context. Figure 1 shows the
evolution in nominal and in real terms of annual average world prices of wheat, coarse grains, rice and
oilseeds from 1971 to 2007 with projections from 2008 to 2017. While spot prices for April-May 2008 are
not shown, for these commodities price levels greatly exceeded the expected annual average for 2008. Two
points are clear: first, agricultural commodity markets are notoriously volatile; second, the current price
spike is neither the only nor even the most significant one to occur in the last forty years.

Figure 1. Food commodity price trends 1971 2007, with projections to 2017

700 500
Wheat Coarse Grains
450
600
400

500 350

300
400
USD/t
USD/t

250
300
200

200 150

100
100
50

0 0
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017

Nominal Real Nominal Real

1600 1200
Rice Oilseeds
1400
1000

1200
800
1000
USD/t
USD/t

800 600

600
400

400
200
200

0 0
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017

Nominal Real Nominal Real

Note: Real prices deflated by USA GDP deflator 2007 = 1.

Source: OECD-FAO Agricultural Outlook, 2008-2017

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 3

 It is also important to recall that prices for meat and poultry have seen no or modest increases during
this same period. There have been substantial increases in the prices of dairy products, at least in part due
to recent policy developments, though more recently prices have again declined somewhat. As the
international debate has focused on the implications of increases in prices for crop markets this note
focuses primarily on prices for cereals and oilseeds.

How can the current price spike be explained?

Prices rise when markets get tight. Between 2005 and 2007 there were coinciding spells of
unfavourable weather in major producing regions in the world, pushing crop yields in those areas below
long term average levels. World cereal output in 2007 was just 3% larger than in 2005 while there was a
decline in overall oilseeds output; nevertheless, vegetable oil production rose by 7% due to rapid growth in
palm oil output.

At the same time, there was strong demand growth. Demand for wheat and coarse grains grew almost
twice as much as did production, and demand for vegetable oil increased two percentage points more than
output. More than half of the increase in use of both coarse grains and vegetable oil was due to higher use
in the biofuels industry. Use of cereals and vegetable oil for food also continued to grow, as did cereal use
for feed. This food and feed demand growth came primarily from countries outside the OECD area and
accounted for the remaining nearly 50% of the total increase in demand.

The production shortfall, relative to trend, would in itself have been enough to send prices higher,
although under normal conditions stocks would have buffered the market and dampened the price rise.
But stocks were already low and they kept declining in 2006 and 2007 because of bad weather and low
yields in major exporting countries. Supply shortfalls, the absence of a sufficient buffer, the continued
increase in food and feed use, and the high growth in relatively price-insensitive demand for biofuels all
coincided to make the price increases exceptional. More recently, there has also been a significant increase
in investments in agricultural derivative markets from non-traditional sources, whether for portfolio
diversification or speculation. It is likely that this has contributed to the rise in short term futures prices and
is an additional factor in the current spike in spot market prices.

Some of the factors behind the current price hike are transitory while others may be more permanent.
Making that distinction is an important ingredient in projecting market developments over the coming ten
year period, as is done in the OECD-FAO Agricultural Outlook. This differentiation is also important in
designing good policy to deal with the implications of the price increases.

What factors will shape future prices and price volatility?

The recent negative yield shocks in key agricultural commodity producing regions that have
contributed to price increases should be viewed as temporary. Barring any underlying climate change or
water constraints that could lead to permanent reductions in yield, normal higher output can be expected in
the very short term.

Macroeconomic conditions that favour economic growth, increases in purchasing power, and stronger
demand for agricultural commodities are expected to continue, at least for many non-OECD economies.
This is a permanent factor in future price determination, but not a new one: strong GDP growth in
developing countries has been a feature of commodity markets for many years. Thus, this factor should
slow the decline in real prices in the future, but not lift average prices to permanently higher levels.

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 4

 The oil price, and energy prices more generally, is a critically important contributing factor to the
increase in production costs for agricultural commodities and food and ultimately in the market prices for
these goods. Price projections discussed here reflect the widely held belief that the oil price increases are
permanent, lifting future prices to higher average levels.

Feedstock demand for biofuel production is expected to increase further, albeit at a slower rate than in
the past three years, and under current policy settings appears to represent a permanent factor in price
formation. Unlike strong income growth in developing countries, this is a new source of demand which is
seen as one of the factors lifting prices to higher average levels in the future.

Stocks of wheat, coarse grains and vegetable oil have fallen to low levels relative to use, reducing the
buffer against shocks in supply and demand. Stocks are not expected to be fully replenished over the
coming ten years, implying that tight markets may be a permanent factor in the period to 2017. This should
not lead to permanently higher prices, but provides the background for more price volatility in the future.

The surge of investment in futures commodity markets from non-traditional sources may have short
term price effects. But relative to the ten year outlook period these may prove temporary, given adjustment
in markets and participants behaviour: funds can move rapidly in and out of commodity markets as profit
opportunities dictate. Given their size, this may well be a new and permanent element in future price
volatility.

A more general point concerning price volatility relates to the thinness of markets, or the share of
imports and exports relative to the size of global consumption or production. When markets are thinner and
prices in domestic markets do not follow those in international trade because of insulating policies or
market imperfections, world market prices must change more to accommodate an external shock to traded
quantities, all else equal. Such market characteristics are expected to remain a permanent feature in the
volatility of prices.

Finally, the nature and composition of demand are factors that may increase the future variability in
world prices. First, industrial demand for grains and oilseeds and in particular policy-driven demand for
biofuels production is generally considered less responsive to prices than traditional food and feed demand.
Second, food demand becomes less responsive to price changes as incomes rise and the commodity share
in the food bill falls. Such changes are permanent factors that may lead to greater volatility in future world
prices.

What are current expectations for future prices?

The current outlook for crop production in 2008 is generally positive. The April forecast for global
cereal production suggests an increase of over 3% from 2007, with the bulk of the growth in wheat due to a
substantial recovery in production in major exporting countries. Rice harvests in major producing countries
are just beginning. Wheat production in major exporting countries is expected to be above 2005 levels. The
increased supply should bring some relief to the market in the course of 2008, but with use forecast to
again exceed supply and stocks forecast to be further drawn down, the overall market situation is likely to
remain tight throughout the year.

Over the 2008-2017 period covered by the OECD-FAO Agricultural Outlook, a strong combination of
supply response and continued growth in demand is expected to keep prices above historical levels, but
well below the peaks experienced today.

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 5

 Looking ahead to 2017, the average level of wheat and coarse grain prices is expected to remain
higher than in 2005, but well below levels in 2007-2008. World wheat and coarse grain areas are expected
to increase somewhat and yields are expected to grow along historical trends. Oilseeds prices are expected
to remain strong, though slightly lower than today. Current high prices are expected to bring about a supply
response that results in more land allocated to this sector and good yield growth. In addition, palm oil
production is expected to increase 40%. Rice production is expected to grow modestly with continued
productivity growth offsetting a small decline in the area planted.

Demand for cereals for use as feed stocks in biofuel production is projected, under current policies, to
almost double between 2007 and 2017, but the largest part of future growth in total use is explained by
rising food and feed demand, particularly in countries outside the OECD area that are experiencing strong
economic growth. Little rice is used for feed and almost none in biofuel production. Demand for rice,
almost all for food use, is expected to increase by less than 1% per year and is dominated entirely by
growth in developing countries. Biofuel use of vegetable oils is forecast to account for more than a third of
the expected growth in vegetable oil use from 2005 to 2017, and other uses are also expected to grow
substantially. Income growth drives much of this expansion in demand, with countries outside the OECD
area increasing their consumption of vegetable oils by 50%.

Based on these market developments, the Agricultural Outlook forecasts relatively tight markets to
continue, with prices down from current peaks but remaining higher on average than prices experienced
over the past decade, as shown in Figure 2. On average over the coming ten year period, nominal prices for
cereals, rice and oilseeds are expected to be 35% to be 60% higher than on average in the past ten years.
Prices in real terms are projected to be 10% to 35% higher than in the past decade. Productivity gains and
increasing competition in trade from countries outside the OECD area will eventually overtake stronger
demand. As that happens, prices will resume their decline in real terms, though more gradually than in the
past (see Figure 1).

Figure 2. Nominal prices fall but stay above average levels of the past

900
800
700
600
US$/tonne

500
400
300
200
100
0
Wheat Maize Rice Oilseeds

1998-2007 highest 2007-2008 2008-2017 2017

Which factors might change these price projections?

The recent spikes in food commodity prices surprised most economic forecasters, reminding us of the
inherent vulnerability of projections to unanticipated developments. The price projections discussed here
assume normal weather, unchanged policies, and stable economic performance. Looking at alternatives to
these assumptions provides additional insights regarding the factors influencing future prices. Key results
of four alternative scenarios are illustrated in Figure 3 and summarised below:

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 6

 If biofuel production is assumed to remain at 2007 levels, rather than doubling over the next 10
years as expected, the projected prices for coarse grains would be 12% lower and vegetable oil
15% lower in 2017 than currently expected.

If it is also assumed that oil prices stay at their 2007 level over the next decade, projected prices
for wheat and maize fall by a further 10% and for vegetable oil by a further 7%.

If the rates of growth in GDP in Brazil, China, India, Indonesia and South Africa are then
reduced to half the rate assumed in the Outlook with exchange rates and inflation also changed
consistently projected wheat and coarse grains prices would fall by a modest 1 to 2% more but
vegetable oil prices by a further 10%.

Finally, if it is also assumed that cereals and oilseeds yields are 5% higher than expected,
projected wheat and maize prices would be a further 6 and 8% lower, respectively, but there is
little further change in vegetable oil prices.

All these assumptions tend in the same direction, to lower prices, and taken together would lead to
prices for wheat, coarse grains and vegetable oils that are 20 to 35% lower in 2017 than what is now
projected. It is unlikely that these factors would all combine in the configuration described here, but it
serves to illustrate the relative significance of some of the factors that will determine future price levels. Of
course, alternative scenarios that could push prices back up could also be envisaged an equivalent
production shortfall in a major exporting country again this year, for example, would be expected to lead to
prices that remain near current levels.

Figure 3. Sensitivity of projected world prices to changes in four key assumptions

Percentage difference from baseline values, 2017

Wheat Maize Vegetable oil

0%

-5%

-10%

-15%

-20%

-25%

-30%

-35%

-40%
Scenario 1: Biofuel production constant at 2007 level

Scenario 2: Scenario 1 + Oil price constant at 2007 level (72$)

Scenario 3: Scenario 2 + Lower income growth in EE5 countries (half annual growth rate)

Scenario 4: Scenario 3 + Yield for wheat, oilseeds and coarse grains 5 % higher than over the projection period

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 7

What are the impacts of high food prices?

The impact of high food prices on developing countries depends on the interplay of various factors. In
general, commercial producers of these commodities will benefit directly from higher prices, as will in
many cases the people they employ (assuming, of course, that governments do not prevent higher prices on
world markets from being transmitted to domestic markets). Livestock producers, on the other hand, are
squeezed by both higher feed and energy costs and relatively flat prices. For farm households producing
mainly for their own consumption or for local markets insulated from price fluctuations on national and
international markets, the impacts will be mitigated. But for the urban poor and the major food importing
developing countries, the impacts will be strongly negative as an even higher share of their limited income
will be required for food. Each 10% increase in the prices of all cereals (including rice) adds nearly USD
4.5 billion to the aggregate cereals import bill of those developing countries that are net importers of
cereals.

The impact of high agricultural commodity prices on developed countries is relatively modest, overall.
The agricultural commodity price component of final food product prices is relatively small (often 35% or
less), as is the proportion of disposable income spent on food (10-15% for most OECD countries). Of
course these averages mask much more significant impacts on lower income consumers who spend a larger
share of their expenditure on food. In addition, and to the extent that high prices persist and hence do not
reduce the future rate of inflation, indirect economic impacts might also be important.

What are appropriate policy responses?

In the short term, humanitarian aid is required. Before recent price increases, hundreds of millions of
people were going hungry because they could not afford food. With higher prices, the numbers of people
suffering from extreme hunger has increased even further the 1st UN MDG has become an even greater
challenge. Immediate and substantial aid is required, where appropriate in the form of cash or vouchers so
as to strengthen, rather than undermine domestic markets in the recipient countries.

In the medium term, there is a real need to improve the purchasing power of poor food buyers so they
can acquire enough food even at the higher prices, relative to past averages, that are expected to prevail in
the future. Fundamentally that requires to foster growth and development in poor countries. In some of the
poorest countries, investment in agriculture, including in agricultural research, extension and education,
may be the best way to cut poverty and stimulate economic activity. In other situations, investment in
agriculture may also be helpful, but there may equally be a need to diversify the structure of the economy.
In many cases, investments in improving the overall environment in which agriculture operates may be
most appropriate improving basic governance systems, macroeconomic policy, infrastructure,
technology, education, health, etc. In other words, a tailored approach is needed, one that builds upon the
capacity and potential of individual countries, rather than a generalized rush to develop agriculture.

Agricultural trade policies require further reform. Trade restricting policies whether they restrict
exports or imports have undesirable and often unintended impacts, especially in the medium and long
term. Subsidies that distort markets are equally unhelpful. Export taxes and embargos may in the short
term provide some relief to domestic consumers, though such measures do not distinguish between low
and high income consumers, and they also impose a burden on domestic producers and limit their supply
response. Export restrictions contribute to global commodity market uncertainty and drive international
market prices further up. On the import side, protecting domestic producers of agricultural commodities
by providing high price support and border protection restricts growth opportunities for producers abroad
and imposes a burden on domestic consumers. A swift and ambitious conclusion of the Doha Round of
WTO negotiations could make an important contribution to exploiting the potential of markets to balance
global supply and demand.

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 8

 It is also instructive to look closely at the causes of recent price increases. On the supply side, the link
between production and yield shortfalls and climate change might be further explored. Investments in
R&D, technology transfer and extension services, particularly in less developed economies, could do much
to increase productivity and output. The use of genetic modification (GMOs) also offers potential that
could be further exploited, to improve productivity, to enhance the attributes of crops destined for either
food or non-food uses, and to enhance the resilience of crops against stress such as drought. On the
demand side, policies that encourage increased production and use of biofuels warrant a close review.
OECD/IEA analysis to date suggests that the energy security, environmental, and economic benefits of
biofuels production based on (first generation) agricultural commodity feed stocks are modest and are
unlikely to be delivered by current policies. Alternative approaches (for example, that encourage reduced
energy demand and GHG emissions, provide for freer trade in biofuels, and accelerate introduction of
second generation production technologies that do not rely upon current commodity feed stocks) offer
potentially greater benefits without the unintended impact on food prices.

High agriculture commodity prices also have an impact on close substitutes, such as fish, and could
contribute to even further pressure on already depleted fish stocks, as well as to increased demand for fish
from aquaculture. Policies that ensure the sustainable and responsible use of ocean resources have a key
role to play, both within national boundaries and on the high seas; concerted action to control illegal
fishing is needed. Options to improve the business environment for private investment in aquaculture
might also be explored.

The current hike in food prices is an issue of a truly global nature. It has complex causes and impacts,
and requires a complex response at the international level. Current developments on global food markets
are having dramatic implications for food security among poor people. At the same time, speculative
factors and inward looking policy actions contribute to the nervousness and volatility of markets. What is
needed now is an objective, effective and coherent global response to avoid making a difficult situation
worse.

Now, more than ever, it is important to counter growing calls for trade protectionism. Closing markets
to either imports or exports will have exactly the wrong result. More secure global food supplies will only
come from competitive producers around the world being provided the freedom to respond to current
market opportunities. Continuation or introduction of policies that create distortions and that undermine the
appropriate market responses should be avoided. The OECD will continue to monitor commodity market
developments and government initiatives over the coming months, and report periodically on its findings.

RISING FOOD PRICES: CAUSES AND CONSEQUENCES OECD 2008 9

Price Volatility in Food and Agricultural Markets:
Policy Responses

Policy Report including contributions by

FAO, IFAD, IMF,OECD, UNCTAD, WFP, the World Bank,

the WTO, IFPRI and the UN HLTF

2 June 2011

UN-HLTF on
Global Food Security
6 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

1. Introduction

1.1 Scope

1. Under the Food Security pillar of the Seoul Multi-year Action Plan on Development, the G20
request that FAO, IFAD, IMF, OECD, UNCTAD, WFP, the World Bank and the WTO work with key
stakeholders to develop options for G20 consideration on how to better mitigate and manage the risks
associated with the price volatility of food and other agriculture commodities, without distorting market
behaviour, ultimately to protect the most vulnerable. This report has been prepared by FAO, IFAD,
IMF, OECD, UNCTAD,WFP, the World Bank, the WTO, IFPRI and the UN HLTF.

2. The approach taken in this report reflects the view of the collaborating international
organisations that price volatility and its effects on food security is a complex issue with many
dimensions, agricultural and non-agricultural, short and long-term, with highly differentiated impacts on
consumers and producers in developed and developing countries. The report begins with a discussion of
volatility and of the ways in which volatility affects countries, businesses, consumers and farmers.
Lessons learned from recent experiences are briefly reviewed as well as the factors determining likely
levels of volatility in future. This report offers suggestions for a systematic and internationally
coordinated response building on the lessons learned as a result of the 2007-2008crisis.

3. It is important to distinguish between policy options designed to prevent or reduce price

volatility and those designed to mitigate its consequences. Both types of intervention are explored in
detail. Scope is identified for actions at individual, national, regional and international level. Some would
help to avert a threat, others are in the nature of contingency plans to improve readiness, while still others
address long-term issues of resilience. Finally, the report explores mechanisms of international
cooperation to implement this reports recommendations and to monitor progress.1

1.2 What is volatility?

4. In a purely descriptive sense volatility refers to variations in economic variables over time,
(more technical definitions of volatility and related terms are put forward in Annex A) Here we are
explicitly concerned with variations in agricultural prices over time. Not all price variations are
problematic, such as when prices move along a smooth and well-established trend reflecting market
fundamentals or when they exhibit a typical and well known seasonal pattern. But variations in prices
become problematic when they are large and cannot be anticipated and, as a result, create a level of
uncertainty which increases risks for producers, traders, consumers and governments and may lead to
sub-optimal decisions. Variations in prices that do not reflect market fundamentals are also problematic
as they can lead to incorrect decisions. These implications of volatility will be explored in detail in
Chapter 2.

5. Behind concerns about volatility lie concerns about price levels and behind both, lie concerns
about food security. While producers benefit (or at least those who are net producers and whose asset
base and knowledge enable them to respond effectively), consumers, especially poor consumers, are
severely adversely affected by high prices2. Food accounts for a very high share of the total budget of the
poorest households. And because poor households often consume foods that are less processed the effect
of rises in commodity prices is felt more strongly. These households find their nutrition status (especially
of pregnant women, children and those affected by long-term diseases such as HIV), as well as their
capacity to purchase education, health care, or other basic needs compromised, when food prices are
high.

6. Producers are more concerned about low prices, which may threaten their living standards as
well as their longer term viability when income is too low to provide for the farm family or for the

6
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 7

operational needs of the farm. Uncertainty may result in less than optimal production and investment
decisions3. In developing countries, many households are both producers and purchasers of agricultural
products. For this group the impacts of price volatility are complex, with net outcomes depending on a
combination of many factors.4

7. No attempt is made here to define extreme or excessive price volatility. Suffice it to say that
volatility becomes an issue for concern and for possible policy response when it induces risk averse
behaviour that leads to inefficient investment decisions and when it creates problems that are beyond the
capacity of producers, consumers or nations to cope. What constitutes excessive volatility depends very
much on the situation of the individual or nation. Poor consumers in less developed countries without
access to adequate social support are most immediately affected by price surges. Small resource limited
farmers face particularly severe problems when prices fall. The episode of volatility that occurred during
the 2007-2008 period, resulted in poor, vulnerable consumers and producers and poorer developing
countries dependent on food imports experiencing severe economic, social and political stress because of
high prices and fears of scarcity. Lessons learned concerning appropriate national and international
response are instructive as we enter 2011 with many commodity prices again increasing sharply.

1.3 Trends in volatility

8. When looked at in the long term there is little or no evidence that volatility in international
agricultural commodity prices, as measured using standard statistical measures is increasing and this
finding applies to both nominal and real prices5. Volatility has, however, been higher during the decade
since 2000 than during the previous two decades and this is also the case of wheat and rice prices in the
most recent years (2006-2010) compared to the nineteen seventies.6 Another conclusion that emerges
from the study of long term trends in volatility is that periods of high and volatile prices are often
followed by long periods of relatively low and stable prices. Finally, it is well established that
agricultural markets are intrinsically subject to greater price variation than other markets, for reasons that
are outlined in the introduction to Chapter 2.

9. International commodity prices since 1970 are presented in Figure 1 and commodity price
movements during the past decade as shown in Figure 2.
Figure 1. Agricultural commodity prices in real terms Figure 2. Monthly commodity price indices
(2005=100) (2002-04=100)
IMF food Index Beef Butter Maize Rice Wheat
1000
Food Prices Meat Dairy Sugar Wheat Rice
800

900

700
800

600 700

600
500

500

400
400

300
300

200
200
100

100
0
avr.-04

avr.-05

avr.-06

avr.-07

avr.-08

avr.-09

avr.-10
oct.-04

oct.-05

oct.-06

oct.-07

oct.-08

oct.-09

oct.-10
juil.-04

juil.-05

juil.-06

juil.-07

juil.-08

juil.-09

juil.-10
janv.-04

janv.-05

janv.-06

janv.-07

janv.-08

janv.-09

janv.-10

10. Since 1990, as shown in Figure 3, the implied volatility for major crops has increased
significantly.7 Implied volatility reflects the expectations of market participants on how volatile prices
will be and is measured as a percentage of the deviation in the futures price (six months ahead) from
underlying expected value (for a more detailed explanation of implied volatility see Annex A). Broadly

7
8 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

speaking, increases in implied volatility reflect how market conditions and unpredictable events translate
to higher uncertainty ahead for traders and other market participants.

Figure 3. Implied volatilities (annual)

1990-2010*
%
40

30

20

10

0
90 92 94 96 98 00 02 04 06 08 10

Wheat Maize Soybeans

* FAO (2010), Food Outlook, November. See also Annex A for an explanation of implied volatility and a
description of the assumptions that underpin the measure.

11. Irrespective of any conclusion that might be drawn concerning the long term trends, there is no
doubt that the period since 2006 has been one of extraordinary volatility. Prices rose sharply in 2006 and
2007, peaking in the second half of 2007 for some products and in the first half of 2008 for others. For
some products the run-up between the average of 2005 and the peak was several hundred percent. On the
rice market the price explosion was particularly pronounced. The price rises caused grave hardship
among the poor and were a major factor in the increase in the number of hungry people to more than one
billion.8 Prices then fell sharply in the second half of 2008, although in virtually all cases they remained
at or above the levels in the period just before the run-up of prices began. Market tensions emerged again
during 2010 and there have been sharp rises in some food prices. By early 2011, the FAOs food price
index was again at the level reached at the peak of the crisis in 2008 and fears emerged that a repeat of
the 2008 crisis was underway.

1.4 Volatility in global versus national markets

12. The trends and fluctuations described in the previous paragraphs relate to international prices.
Domestic price movements can be different. The extent to which global prices are transmitted to
domestic markets depends on how strongly integrated the latter are with the former. Measures such as
import duties, export taxes, non-tariff barriers or domestic policies such as price support all influence the
extent to which price changes in domestic markets mirror those on international markets. Market
structure is also important. In monopsonistic markets, whether private or state controlled, higher
international prices may not always result in better prices for producers Countries that insulate their own
markets export instability onto international markets, especially if they are major players in terms of
consumption or production. The degree of processing of final consumption goods also affects price
transmission. Lack of domestic infrastructure and generally undeveloped or inefficient market structures
can also significantly obstruct price transmission due to high transport and transactions costs.

8
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 9

13. Developing country markets often lack the capacity to absorb domestic shocks, and can be
subject to high domestic price volatility even during periods of calm international markets. Attention also
needs to be paid to volatility at local and national levels, and to its consequences for poor rural people
including small farmers. The causes may relate to climate shocks, pests or other natural calamities,
exacerbated by the fact that farmers may have poor access to technologies and generally poor
management of soil and water. Poor infrastructure, high transport costs, absence of credit or insurance
markets and various policy and governance failures may compound the initial difficulty. A relatively
minor climatic incident in these conditions can become a serious food crisis at local or regional level.
Again those most affected will be poor consumers and rural dwellers, mainly smallholders in less
developed countries or regions, heavily dependent on their own production.

14. During the 2007-2009 price spike and subsequent decline, there were quite significant
differences among regions and products in the speed and degree to which world price movements were
felt in regional or local markets. Many factors explain these differences including policy responses,
exchange rate movements, competition policy, market structure and degree of market openness.9 These
differences are important because they suggest that both price levels and degrees of volatility may differ
significantly from place to place at any given time and, therefore, that the level of hardship and
disruption being experienced may also differ. The international community needs timely and
differentiated information about the situation in different places in order to respond appropriately.

2. Price volatility in food and agriculture, potential developments and impacts

15. Are recent events random resulting from an unusual coincidence of different factors or are
there reasons to believe that the world is entering into a period of recurrent episodes of extreme price
volatility? It is not possible to have a view on the appropriate policy responses to volatility without first
exploring this question in some detail. In this context too, it is worth recalling that behind the expressed
concerns about volatility is a concern about price levels, particularly the impact of high prices on the
food security of the most vulnerable households and countries and of low prices on vulnerable producers.

16. Most agricultural commodity markets are characterized by a high degree of volatility. Three
major market fundamentals explain why that is the case. First, agricultural output varies from period to
period because of natural shocks such as weather and pests. Second, demand elasticities are relatively
small with respect to price and supply elasticities are also low, at least in the short run. In order to get
supply and demand back into balance after a supply shock, prices therefore have to vary rather strongly,
especially if stocks are low. Third, because production takes considerable time in agriculture, supply
cannot respond much to price changes in the short term, though it can do so much more once the
production cycle is completed. The resulting lagged supply response to price changes can cause cyclical
adjustments (such as the often referenced hog cycle) that add an extra degree of variability to the
markets concerned. Business cycle fluctuations in demand for agricultural non-food commodities (such
as cotton) from rapidly growing, industrializing economies may also be contributing to increased
volatility.

17. As of Spring 2011, world price levels as reflected in various measures, including the FAOs
world food price index, have once again reached the levels of 2007/08, giving rise to concerns that a
repeat of the earlier crisis is underway. Several of the same factors known to have contributed to the
2007/08 crisis are also present weather-related crop losses, export restrictions, high oil prices, and a
depreciating US dollar, against a background of a continuing tight supply-demand balance. The debate
on the impact of financial investment in commodity markets also continues. On the other hand, the
2010/11 situation differs from the earlier episode in some important respects. Firstly, the 2010 harvests in
many food importing countries in Africa were above average or very good, so that prices in the region
have been more stable. Stocks were higher at the outset which has also helped to mitigate the price rises.
Finally, the price increases have been differently distributed among commodities. Meats, sugar and dairy

9
10 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

products are all affected, and these are commodities that are less important in the food bills of the most
vulnerable. It should be noted also that while the index of prices for cereals has come close to its 2008
level on average, and prices of vegetable oils are also very high, contrary to the 2007/08 situation the
price rises have not affected rice. As rice is the staple food of many millions of the worlds most
vulnerable consumers, this means that the incidence of current price increases is somewhat different.
Nevertheless, there are serious risks to food security and the situation needs to be kept under close
review by national governments, and by international organisations and non-governmental agencies. .

2.1 The determinants of future increases in food prices and of volatility

18. Growing population and income in emerging and developing countries will add significantly to
the demand for food in the coming decades. By 2050 the worlds population is expected to have reached
about 9 billion people and the demand for food to have increased by between 70% and 100%. This alone
is sufficient to exert pressure on commodity prices. According to the latest OECD/FAO medium term
outlook projections, prices of crops and most livestock products will be higher in both real and nominal
terms during the decade to 2019 than they were in the decade before the 2007/08 price spikes. If the rate
of growth of agricultural production does not keep pace with demand, upward pressure on prices will
result. A demand or supply shock in a situation where the supply-demand balance is already tight, can,
for the reasons explained in the previous paragraph, result in increased volatility around the upward
trend.

19. The demand for food and feed crops for the production of biofuels is another significant factor.
During the 2007-2009 period biofuels accounted for a significant share of global use of several crops
20% for sugar cane, 9% for vegetable oil and coarse grains and 4% for sugar beet. Projections encompass
a broad range of possible effects but all suggest that biofuel production will exert considerable upward
pressure on prices in the future. For example, according to one study international prices for wheat,
coarse grains, oilseeds and vegetable oil could be increased by 8%, 13%, 7% and 35% respectively10.
Moreover, as long as governments impose mandates (obligations to blend fixed proportions of biofuels
with fossil fuels, or binding targets for shares of biofuels in energy use), biofuel production will
aggravate the price inelasticity of demand that contributes to volatility in agricultural prices.

20. Agricultural commodity prices are becoming increasingly correlated with oil prices. Oil prices
affect agricultural input prices directly and indirectly (through the price of fuel and fertiliser, for
example). In addition, depending on the relative prices of agricultural crops and oil, biofuel production
may become profitable (without government support) in some OECD countries. Financial investment in
commodities may also have contributed to an increasing correlation between oil and non-oil commodity
prices because of the significant share of such investment that tracks indexes containing a basket of
different commodities. High and volatile oil prices (if that is what is expected) could therefore contribute
to higher and more volatile agricultural prices, through higher input costs, higher demand for the
commodities used in the production of biofuels (sugar, maize, vegetable oils), through competition for
land with commodities that are not used directly for the production of fuel, and possibly through
financial investment in commodity baskets.

21. Low stocks relative to use, and uncertainty about stock levels in some parts of the world
contributed to the 2007/2008 price spike. Stocks can be drawn down in response to a supply or demand
shock, but once they have been depleted, supply can no longer be increased until new production comes
on board. Even expectations of depleted stocks may lead prices to rise sharply. The low stock levels
observed in recent years have been attributed to the partial dismantling of price support and intervention
purchase schemes in some OECD countries, as well as to correction of the quality of information on
private and government held stocks in important producing and consuming countries. Stocks were rebuilt
during 2009 and the first part of 2010 but currently stocks are again being depleted. If stock levels

10
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 11

remain low in major markets, and projections based on existing knowledge of market conditions and
policy settings suggest that they may, the risk of volatility in prices will remain high.

22. Climatic factors have indisputably contributed to the price rises in 2007/2008 and again in
2010. In 2008, an already tight market situation for wheat was aggravated by drought in Australia, which
is an important supplier of wheat to world markets. Canada, another important supplier, also experienced
weather related low yields for several crops. More recently, drought followed by fire in the Russian
Federation, fears about the Australian and Argentinean crops, and several downward revisions of US
crop forecasts in late 2010 and early 2011 have brought strong market reactions and soaring prices. It is
not clear whether these weather-related events are transitory in nature, cyclical (El Nino and La Nina) or
the harbingers of long term climate change. Experts concur broadly that climate change will, in the
longer term, lead to worsening conditions in some arid and semi-arid regions where agricultural
production is already difficult, while temperate regions in particular, but not exclusively, may benefit. It
is also thought that climate change will lead to more frequent extreme events such as droughts, heat
waves and floods. Clearly, climate change will provoke some adjustment of production patterns around
the world, as well as increased risks of local or regional supply problems that could add to future
volatility.

23. Stronger demand for food crops and animal products in conjunction with slow growth in
agricultural productivity and low stocks results in upward pressure on prices. Recent years have also seen
some shift in production patterns, particularly of food and feed grains, and world markets are more
dependent on supplies from the Black Sea region and other, newer, agricultural production regions than
in the past. Yields in these regions are less stable and supply more variable than in some other parts of
the world where natural conditions are better and where application of the most up-to-date technologies
and management practices have increased and stabilised yields. As the geographical distribution of
production changes, supply may therefore become more variable, in turn leading to increased price
volatility.

24. The same underlying factors that are leading to increased demand for food growth in
population, affluence leading to increased demand for animal protein, urbanisation, and biofuels are
also increasing pressure on finite resources such as land and water. While such resource constraints are,
thus far, more local than global in nature, growing concern is evident and the associated uncertainty may
imply upward pressure on prices and continuing or increased volatility.

25. During the 2007-2008 period, some policy measures put in place by a number of governments
contributed directly and indirectly to the crisis (export restrictions, hoarding), increasing the amplitude of
price movements and in some cases provoking price increases that were otherwise inexplicable in terms
of the market fundamentals. Inappropriate policy responses also contributed to volatility and could
continue to do so unless the international community is able to take steps to avoid such actions.
Additionally, private and public actors responding to the general nervousness of the markets, or for
speculative reasons, engaged in hoarding or precipitated purchases in an already tense market situation.
How to avoid repetition of these types of damaging private and public reactions is addressed later in this
report.

26. Trade in many agricultural commodities is denominated in USD. A depreciating USD, as

occurred in the years before and up to the peak of the price rises, causes dollar denominated international
commodity prices to rise, although not to the full extent of the depreciation. The opposite occurs when
the dollar appreciates as was the case from mid-2008 onwards. These currency movements added to the
amplitude of the price changes observed. (They also help to explain why demand remained strong in
countries where the currency was appreciating against the dollar and why falling prices were not fully
felt in the same countries once the dollar began to appreciate again.) Exchange rate volatility per se is

11
12 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

beyond the scope of this report but if the future is marked by increased exchange rate volatility this will
also have repercussions for the volatility of international prices of commodities.

27. There is no doubt that investment in financial derivatives markets for agricultural commodities
increased strongly in the mid-2000s, but there is disagreement about the role of financial speculation as a
driver of agricultural commodity price increases and volatility. While analysts argue about whether
financial speculation has been a major factor, most agree that increased participation by non-commercial
actors such as index funds, swap dealers and money managers in financial markets probably acted to
amplify short term price swings and could have contributed to the formation of price bubbles in some
situations. Against this background the extent to which financial speculation might be a determinant of
agricultural price volatility in the future is also subject to disagreement. It is clear however that well
functioning derivatives markets for agricultural commodities, could play a significant role in reducing or
smoothing price fluctuations indeed, this is one of the primary functions of commodity futures markets.
This topic will be taken up in more detail in Chapter 3.

28. This catalogue of factors points to a likelihood of higher real prices and a risk of increased
volatility in future years. While it is not possible to forecast future prices or future returns, there may be
ways in which the international community could be alerted to a risk that a period of excessive volatility
is in the offing.11 Various tools and mechanisms that could assist in this respect are described in later
sections.

29. There are also a number of factors that could act to mitigate increased price pressure and
increased volatility. Most important will be the supply response? Analysts have consistently
underestimated the capacity of producers to respond to positive market signals, as well as the potential
for higher yields and increased acreage. Successful conclusion to the WTO Doha Development Agenda
negotiations would be an important step, along with complementary policies that improve supply
capacity and ensure the benefits of open and competitive markets are widely spread. Deeper integration
of global and regional markets, better defined safeguard mechanisms and improvements in the
competitive environment will bring increased trade volume and more suppliers and buyers to markets
that are currently very shallow. Local or regional supply shocks could more easily be absorbed leading to
lower volatility on domestic and international markets and food could more easily flow from surplus
areas to rapidly urbanising food-importing areas

30. The extent of potential future increases in prices and volatility cannot be estimated accurately,
but the risks are sufficiently large to warrant serious reflection about what can be done to mitigate it
when the nature of the underlying causes makes mitigation possible and to manage the consequences,
when, as is inevitable, episodes of high volatility occur. The remainder of this report takes up this crucial
topic.

2.2 Why does agricultural price volatility matter?

31. At the macro level it is useful to distinguish between long and short run effects of commodity
price volatility and between importing and exporting countries. The assumption is that countries likely to
be most concerned by macro-economic impacts of agricultural price volatility are developing or
emerging economies that are dependent on agricultural commodities for a large share of their export
revenues, or whose food imports are significant in balance of payment or government finance terms. For
exporting countries heavily dependent on agricultural commodities, exceptionally low prices will have
immediate balance of payments impacts, but beyond that, uncertainty may curtail investment and affect
capacity utilization and there is some evidence of long-lasting significant negative effects on growth.
Importing countries faced with exceptionally high prices may also experience deterioration in the balance
of payments and deterioration in their public finances. Food price increases can have major repercussions
on the whole economy. For low-income food-importing countries, high food prices can result in inflation

12
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 13

and high import bills which in turn worsen the current account balance. As countries have to export more
to pay for imports, such deficits may result in the depreciation of the exchange rate. Fiscal measures,
such as cuts in import tariffs and in taxes on food, subsidization of food consumption, and increased
demands on risk management instruments entail increased budgetary costs that will have to be met by
increased government borrowing and budgetary discipline.

32. Looked at from the demand side, significantly higher food prices are disastrous for the poor
especially in developing countries where up to three-quarters of their total income may be spent on basic
foodstuffs. Immediate impacts are obvious, but there are also longer term costs imposed on the poorest
and most vulnerable as spending is switched to less nutritious foods and away from other basic needs
such as education or health. Typically the effects are felt more strongly by women and children.
Particularly severe are the effects on children stunting and cognitive loss often occurring as a result of
inadequate nourishment during the first 1 000 days after conception. The consequences are tragic for
individuals and for future prosperity in the countries where they live. This is not only an issue of concern
for the affected countries but also for humanitarian agencies and the international community.

33. Food price inflation can also be a serious issue in middle income countries, where many
consumers expend as much as half of their budget on basic foods. Even in the developed countries
significantly higher food prices can create hardship for the least well-off, who tend also to devote a larger
share of household spending to food. Nevertheless, consumers in developed countries face wider choices
in terms of their ability to adjust spending on different types of foods and most developed countries have
safety net mechanisms that are well suited to delivering targeted assistance to the most affected.

34. Looked at from the supply side, high prices benefit net producers of these commodities and
signal a need for increased production. Price increases affect mainly grains and oilseeds which is a high
proportion of total costs in intensive production systems. Profitability of livestock enterprises will be
affected especially if these costs cannot be fully passed on to consumers. Volatile feed prices are also
problematical for livestock producers; such uncertainty is detrimental to investment and production
decisions, particularly where the physical production cycle is long.

35. Low or volatile prices pose significant problems for farmers and other agents in food chains
who risk losing their productive investments if price falls occur while they are locked into strategies
dependent on higher price levels to be viable. Farmers who have already planted their crop are the classic
example. Poor smallholders who do not have access to credit may have difficulty financing the crucial
inputs needed to plant again and stay in business. This kind of problem may be particularly severe for the
female smallholders who are in the majority in many countries. Many farmers in developing (and even
some in more advanced) economies may not be operating on a sufficiently large scale to be able to carry
over income from one season to another. Thus, both the welfare of the family and the viability of the
farm may be threatened by excessive volatility. Uncertainty may also result in sub-optimal investment
decisions in the longer term.

2.3. Lessons learned from recent experiences

36. It is beyond the scope of this paper to undertake an exhaustive account of the ways in which
national governments and international institutions responded to the price volatility during 2007-2008. It
is generally agreed, however, that policy responses were mainly ad hoc in nature, that some decisions
were taken hastily, and that measures were somewhat inconsistent and largely uncoordinated at
international level. The speed and strength of the price rises also took the international agencies by
surprise and, they too, had to resort to ad hoc measures. Developed countries relied mainly on already
existing safety net mechanisms while developing countries took new measures or adjusted the parameters
of existing instruments.

13
14 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

37. Of 81 developing countries surveyed by the FAO, 43 reduced import taxes and 25 either
banned exports or increased taxes on them.12 A large number of developing countries implemented
measures to provide relief or partial relief from high prices to consumers 45 in all. Measures consisted
of cash transfers, direct food assistance or increases in disposable income (by reducing taxes or other
charges), or some combination of these measures. A significant number of countries also granted support
to producers in order to offset rapidly rising input costs, as prices for fertilizer also surged as did feed
costs for livestock producers. Several countries went to the international markets to procure supplies of
basic foodstuffs for their populations, believing that high prices would persist and that scarcity was
imminent, notwithstanding the fact that they did not have any immediate or short term need to do so.

38. The extremely rapid run-up in food prices eroded the capacity of the national and international
relief organisations to purchase food in the most hard hit countries and regions. With prices doubling or
tripling within a few months, their purchasing power was dramatically reduced. While response to
appeals made, for example, by the World Food Programme were both rapid and generous, crucial weeks
and months were lost as international organisations and humanitarian NGOs scrambled to raise funds or
divert monies from other uses to address the crisis. This situation revealed deficiencies in international
readiness to deal with such a widespread problem.

39. The events of 2007-2008 also revealed serious deficiencies in the quality of the information
base, and in particular concerning short-term forecasts and the level of stocks. More timely, complete and
accurate information and improved capacity to identify and analyse early warning signs might have
calmed the markets, re-assured populations and resulted in better readiness.

40. The different measures taken by individual governments in response to the crisis had different
degrees of effectiveness.13 The scale of the price increases was such that for many countries reducing
import tariffs had a relatively modest impact because the initial tariffs were low or the scale of the price
increases was so large. In any event, this instrument was quickly exhausted as tariffs were reduced to
zero. Some of these countries suffered steep falls in tariff revenues and deterioration in their fiscal
situation. Export taxes and restrictions differed between countries in their effectiveness in keeping
domestic prices lower and in some cases had only a relatively minor effect. Export restrictions by major
food exporters had strong destabilising effects on international markets. As more countries followed the
first movers, volatility was exacerbated and the upward price movement was amplified. Export
restrictions proved extremely damaging to third countries, especially the poorest import dependent
countries, and to the efforts of humanitarian organisations to procure supplies, despite various ad hoc
exemptions and exceptions which were put in place in order to mitigate the worst of these beggar thy
neighbour effects.

41. Targeted assistance to those most in need, either using cash transfers or direct food assistance,
may be the most effective and equitable way of reaching those affected by a food price crisis and several
countries have successfully used this kind of instrument. However, many countries did not have the
administrative frameworks in place to be able to implement safety-net measures at short notice. Neither
did they have the fiscal capacity. They therefore made blanket market and trade interventions that
sometimes proved ineffective or costly or both. Such measures, when they delivered some relief did so
irrespective of need. This revealed the importance of contingency planning to better equip countries to be
able to deliver targeted assistance where it is most needed.

42. Estimated numbers of hungry people in the world rose from 820 million in 2007 to more than a
billion in 2009, proof that neither national nor international responses were able to fully cope with the
scale of the problem. Deficiencies in information, communication, and in readiness contributed, as did
uncoordinated measures that may have actually aggravated the problem for people and countries less
well able to cope. The numbers of hungry people have since dropped to about 900 million. These events
have drawn increased attention to the fact that a significant proportion of humanity (about 16%) remains

14
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 15

chronically under-nourished, even during periods of relatively normal prices and low volatility. The
overarching goal of actions with respect to food price volatility should be to ensure that the most
vulnerable people have access to sufficient, nutritious food. All policy interventions should have as their
ultimate aim, the elimination of all food insecurity, whatever its cause.

3. Measures to increase productivity, sustainability and resilience of agriculture

43. Sections 4 and 5 of this report address policy solutions that aim to reduce price volatility and to
deal with its consequences, particularly for the most vulnerable and food insecure, respectively. A key
element in any long term solution is investment in increasing the productivity and resilience of
developing country agriculture. This can contribute to improving food security in two ways. It can reduce
food price volatility, for example through increased productivity and improved technical management of
production and of risk, and it can help farmers and households to cope better with the effects of
volatility, once it occurs. The set of recommendations put forward here, if implemented, would probably
constitute the single most important contribution to an enduring solution to global food insecurity. While
the benefits would accrue in the longer term, actions are needed immediately.

44. Agriculture is a source of livelihood for about 86% of rural people 1.3 billion smallholders
and landless workers it provides farm-financed social welfare when there are urban shocks, and a
foundation for viable rural communities.14 Long run projections are fraught with difficulty and estimates
can vary widely. According to FAO, the rate of growth in agricultural production is expected to fall to
1.5% between now and 2030 and further to 0.9% between 2030 and 2050, as compared with 2.3% per
year since 1961. Population estimates suggest that by 2050 the planet will be home to 9.1 billion persons,
up from the current population of 6.8 billion. This represents a 34% increase over the next 41 years.15
These particular estimates suggest that in the future, with the supply of food not growing at the same
pace with demand, upward pressure on prices could be a principal attribute of world food markets. In
addition to high price levels, shocks, due to climatic or other reasons, can create wide price movements,
as the food market may lack the capacity to absorb them. This adds to vulnerability and underlines the
importance for supply to keep up with growing demand.

45. Investing in agricultural productivity growth and resiliency in low income countries is
paramount to addressing local food price volatility. FAO estimates indicate that agricultural production
would need to grow globally by 70% over the same period, and more specifically by almost 100% in
developing countries, to feed the growing population. In the medium and longer term only investment in
developing countries agricultural sectors will result in sustainable increases in productivity, healthy
markets, increased resilience to international price spikes and improved food security. Investments in
infrastructure, extension services, education, as well as in research and development, can increase food
supply in developing countries and improve the functioning of local agricultural markets, resulting in less
volatile prices. In this way, markets can work for the poor people who bear the burden of food price
volatility.

46. Waste, due to post harvest losses, inadequate storage and infrastructure as well as under-
developed markets in general are a huge issue in agriculture in developing countries, amounting to a
significant proportion of production. The investments proposed here would contribute to reducing or
even eliminating avoidable waste of this type and in so doing would provide a significant boost in the
quantity of food actually reaching consumers. The increase in production needed to meet the anticipated
future demand would therefore be less than what is currently estimated. Similar efforts to avoid waste
which in the developed countries occurs mainly at retail level and in homes and restaurants would also
have large beneficial effects. This aspect is discussed in section 4.5.

47. The investments required in developing countries to support this expansion in agricultural
output amount to an average annual net investment of USD 83 billion (in 2009 USD).16 This total

15
16 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

includes investment needs in primary agriculture and necessary downstream services such as storage and
processing facilities, but does not include public goods like roads, large scale irrigation projects,
electrification and others that are also needed.

48. Most of the investment, both in primary agriculture and downstream sectors, will have to come
from private sources, primarily farmers themselves purchasing implements and machinery, improving
soil fertility, etc. For a better functioning agricultural system and improved food security, three kinds of
public investments are also needed.

Direct investment in agricultural research and development particularly on practices that

enhance the resilience of small-scale agriculture towards climate change and resource scarcity.
Investment in sectors strongly linked to agricultural productivity growth and to strengthening
the integration of smallholders into markets, such as agricultural institutions, extension
services, roads, ports, power, storage and irrigation systems.
Non-agricultural investment to enhance the rural institutional environment and bring about
positive impacts on human wellbeing, like investment in education, particularly of women,
sanitation and clean water supply, and health care.

49. Farmers and prospective farmers will invest in agriculture only if their investments are
profitable. Many types of public goods, such as the above mentioned, that make private investments
financially viable can only be provided by the governments. Private sector investment also needs to be
encouraged at all stages in the value chain upstream of the farm, in seed and fertilizer production and
distribution, and downstream in processing, marketing and distribution. Underlying competition
problems that have led to the development of cartels or of monopsonistic/monopolistic market structure
should also be tackled.

50. However, the capacity of the poorer developing countries to fill the investment gap is limited.
The share of public spending on agriculture has fallen to an average of around 7% in developing
countries, even less in Africa, and the share of official development assistance going to agriculture has
fallen to as little as 3.8%. Commercial bank lending to agriculture in developing countries is also small
less than 10% in sub-Saharan Africa. Private investment funds targeting African agriculture are an
interesting recent development but current investments are still small.17

51. Investments in agricultural research and development have been shown to have very high rates
of return and have an important role to play in fighting hunger and poverty. Bridging the gap between
research and development in the main cereals and staples that are most important for small farmers in
regions with high prevalence of hunger is an important challenge.

52. The agricultural sector is very green-house-gas (GHG) intensive: it accounts for about 13%-
33% of global GHG-emissions, but only for about 4% of global output. Agriculture will therefore be
called upon to contribute significantly to mitigation. At the same time, agriculture will be affected in
ways that are not fully understood or fully predictable, but there is little doubt that some regions,
principally arid and semi-arid zones, will come under increasing pressure. Climate change will lead to
more frequent extreme events such as droughts, heat waves and floods. These incidents will affect not
just production and the volatility of production, they may also create new difficulties related to water
quality, storage and related food safety issues. Complex demands for mitigation and adaption will
therefore be made on the sector during a period when significantly increased production is needed in
response to projected needs. In-depth research on the link between climate change and agricultural
production will be needed and would best be undertaken in coordination with the IPCC, while building
agricultural resilience through increased funding for climate change adaptation measures in developing
countries will be extremely important for example, under the newly created Green Climate Fund.18 19

16
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 17

53. Agricultural research is increasingly being delivered by the private sector with technologies
being developed for larger, commercial farming operations. The adoption of such technologies requires
increased management skills and effective learning, so that small farms too can have access to innovative
inputs. There is need to improve agricultural technologies specific for, and well targeted to small-scale
agriculture and for appropriate production policies and practices aimed at increasing smallholder
productivity in a sustainable manner.20

54. At present, much public research is carried out by the International Research Centres of the
Consultative Group on International Agricultural Research (CGIAR). There is general recognition of the
utility and benefits provided by this system of international research bodies and affiliated organizations
which have enormously contributed to the global pool of available agricultural technology and
knowledge. A new CGIAR Multi-Donor Trust Fund is established to harmonize donor investments in
key global challenges on agriculture and is being hosted and managed by the World Bank. New results-
oriented research programs focus on climate change mitigation and adaptation policies and technologies
and include a broad group of partners. There is a need to increase and sustain the financing of such
bodies in order that they may continue to invest today in the techniques and innovations that will be
needed to deal with the food security and climate challenges that have been defined elsewhere in this
report.

55. Increasing public investment in transport and productive infrastructure, as well as in human
capital, is central in stimulating productivity and reducing post-harvest wastage. Improving
infrastructure, in particular rural roads and market facilities such as warehouses, storage facilities and
market-information systems are important in reducing transport costs and integrating smallholders to
markets. Investing in, and improving irrigation facilities, and market institutions and mechanisms will
result in increased quantities of food produced, better quality and more stable prices.21 Improving
extension, education and health, targeting small producers but also other value chain actors, are key
elements of a sound policy approach to increase productivity and enhance food security and the well-
being of farmers. Annex B to this report contains a more complete treatment of the role of smallholders,
describing their role in production and consumption, how they are impacted by volatility and further
developing some of the policy recommendations made here to apply more specifically to small scale
agriculture.

56. All these responses to increase the resilience of agriculture and stabilize prices require public
interventions. Government expenditure on agriculture can have a significant positive impact on
productivity. Foreign direct investment also has a positive impact on productivity growth, but only if
carried out responsibly in combination with efficient bureaucracy, a lack of corruption, and democratic
political structures.22 23

57. More and better support for public investment in agriculture public goods will allow private
sector actors, including smallholders and small-scale market agents, to respond more profitably to rising
prices, both increasing local food supply and boosting the incomes of the poor.

58. Priority interventions include support for generation, adaptation, and adoption of improved
technology; improved agricultural water management, tenure security and land markets; and
strengthening agricultural innovation systems. Not only must there be far more investment in public
goods in these areas to facilitate smallholder and other private sector supply response, but investment
must be better.

59. The Global Agriculture and Food Security Program (GAFSP), launched in April 2010, provides
an important avenue for public investment. The GAFSP has pledged USD 925 million from a number of
donors. To date, investment programmes are assisted by the World Bank, the African Development
Bank, the Inter-American Development Bank, the International Fund for Agricultural Development

17
18 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

(IFAD) and FAO. However, there are unfunded country proposals to GAFSP, amounting to
approximately USD 800 million.

Recommendation 1

G20 governments commit to take comprehensive action to strengthen the longer term productivity,
sustainability and resilience of the food and agriculture system world-wide, encompassing several elements.
Improve food and agriculture innovation systems, encompassing public and private investments in scientific
research and development, technology transfer, and education, training and advisory services and ensure that
successful practices are scaled up.
Strengthen the CGIAR system to support technological innovation and global dissemination of technology, in
particular to improve productivity performance in less developed countries taking into account the needs of
smallholder and especially women farmers.
Support the development of technologies and provide the appropriate incentives to address challenges specific
to climate change and sustainable resource use (land and water).
Increase public (ODA and national governments) investment in developing country agriculture, and in activities
strongly linked to agricultural productivity growth, such as agricultural institutions, extension services, roads,
ports, power, storage, irrigation systems and information and communication technology, where appropriate.
link public investment to the provision of sustainable public-private-civil society partnerships.
Support comprehensive national food security strategies that are country-owned and led, evidence-based and
inclusive of civil society and farmer organizations. In this respect, follow up on previous G 20 commitments,
such as the Pittsburgh summit commitment, to fund the Global Agriculture and Food Security Program.
Provide the enabling environment for farmers and other private sector actors to scale up investments, above
and beyond ODA and national government spending, to achieve the increased productivity and enhanced
resilience on which long term food security will depend. To elicit the needed level of private sector investment,
less developed countries in particular will need to support introduction of effective governance systems and
institutions, stable macroeconomic conditions, sound structural policies, human capital development and public
services.
Annex C to this report contains selected project descriptions furnished by the international organisations, to
illustrate the kinds of practical, action-oriented initiatives that are needed in order to implement the
recommendations made here.

4. Policy options to reduce price volatility

60. There are many factors that contribute to high and volatile agricultural prices, making
necessary a combination of policy responses. In order to meet their objectives, policies need to be
legitimate and broadly owned by relevant stakeholders, particularly those policies that aim to restore trust
in markets and avoid panic-driven behaviour. The goal of the policies recommended is not to eliminate
agricultural price volatility, but rather to reduce uncertainty, and perhaps also the amplitude of variations
by smoothing out the extremes. Most importantly, price volatility should reflect market fundamentals as
accurately as possible and not convey incorrect signals as a result of missing or wrong information,
speculation, panic or other disruptive factors.

4.1 Market information, transparency and policy response

61. A lack of reliable and up-to-date information on crop supply, demand, stocks and export
availability contributed to recent price volatility.24 Information on the current situation and outlook for
global agriculture shapes expectations about future prices and allows markets to function more
efficiently. Lack of accurate information on market fundamentals may reduce efficiency and accentuate
price movements.25 At a regional level (mainly in Africa, but also in Haiti, Afghanistan and some Central
American countries), a few successful efforts, such as the Famine Early Warning System Network, have

18
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 19

increased the availability of information to governments and market participants. Better information and
analysis of global and local markets and improved transparency could reduce the incidence and
magnitude of panic-driven price surges.

62. Recent events have revealed weaknesses in the capacity of nations and international
organizations to produce consistent, accurate and timely agricultural market data and analysis, especially
in response to weather shocks. Action is needed to increase capacity to undertake more frequent and
systematic monitoring of the state of crops, and to develop mechanisms for improved short-run
production forecasts, able to translate crop growth, meteorological and remote sensing data into yield and
production expectations. Greater use could be made of satellite data and geo-information systems and, in
this context, international co-ordination and exchange of technologies and information could be
enhanced.

63. Information on stocks is an essential component of a global food market information system,
yet reliable data on stocks of grains and oilseeds are often not collected or, if collected are not reported
publicly. The reasons for poor stock data are multiple: some countries no longer hold public stocks
because the policy measures that created them have been removed or reformed; stocks can be very
dispersed among farmers, traders and other actors and difficult to track; and some information on stocks
is commercially sensitive. Generally, international agencies estimate net changes in stocks as a residual
on the basis of data on production, consumption and trade. As a result it is not possible to have complete
confidence in world food stock estimates. International cooperation could redress this situation and
ensure that reliable information on global stocks becomes widely available. This would in turn better
inform market participants and help avoid mis-informed panic-induced price surges. A first step could be
an audit and assessment of available information, identifying gaps and proposing ways in which they
could be filled.

64. Monitoring food prices, both on cash and futures markets, is essential in a food market
monitoring system. In a similar manner, assessing changes in oil prices and analysing their impact on
food markets is important.26 Better information about domestic price movements is necessary to better
understand how international price changes affect domestic markets in developing countries. Such
information is important for early warning systems, such as the FAO Global Information Early Warning
System and for Vulnerability Analysis and Mapping of WFP.27 It is also crucial for policy making and
designing effective risk management instruments for developing countries.

65. For developing countries, enhanced market information and early warning systems would
enable both governments and the private sector to plan ahead. Governments would be able to more
accurately assess needs, make budgetary provision for producer and consumer safety nets and better
position emergency food security reserves. Better market information and analysis could reduce
uncertainties and assist producers, traders and consumers to make better decisions.

66. Over the last decade a great deal of baseline information on food security vulnerability has
been developed. WFP support of national Food Security Monitoring Systems already provides a
monitoring and decision support tool to help governments manage and respond to risk related to price,
weather or other hazards. The reliability and timeliness of such early warning systems needs to be
improved, and capacity to develop and utilise them could be strengthened at both the national and
regional level. The focus should be on countries that are particularly vulnerable to price shocks and food
emergencies.

67. The experience of the 2007-08 food price crisis and the current excess price volatility in many
international food markets have exposed weaknesses in relation not only to the provision of market
information at the global level but also to the coordination of policy responses to food price volatility.
There is need to ensure better preparedness and more rapid and consistent policy responses in times of

19
20 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

crisis. Building on and complementing existing systems, improvements in global market information and
policy guidance could be achieved through a collaborative food information and policy initiative, the
Agricultural Market Information System (AMIS). Such initiative would improve data reliability,
timeliness and frequency, as well as enhance policy coordination in times of crisis.

68. AMIS could be built on the model of JODI (the Joint Oil Data Initiative), launched in 2000 to
improve information about oil markets. However, it would have the additional function of issuing global
food price surge alerts and promoting policy coherence. AMIS would involve the major food producing,
exporting and importing countries. It would also involve a joint Secretariat comprising of international
organizations with capacity to collect, analyse and disseminate information on a regular basis regarding
the food situation and outlook, as well as food policies.

69. The structure of AMIS would include two groups to effectively perform two important
functions: a Global Food Market Information Group would be responsible for food market information
collection and analysis, while the promotion of international policy coordination would be the objective
of a Rapid Response Forum.

70. Increased and regular exchange of information and collaboration between market experts from
participating countries and organizations in the AMIS Global Food Market Information Group could
result in more complete and reliable data on consumption, production, trade and stocks, increasing
market transparency and curbing food price volatility that is not based on underlying market conditions.

71. Through the comprehensive coverage of global major food markets and the close monitoring of
prices in combination with food security assessments across vulnerable countries AMIS will also provide
a mechanism for global early warning. This will increase the scope for more automated systems of
evaluating food security implications of changing market situations whereby an indicator of different
degrees of severity can be calculated routinely and where appropriate trigger an alert.

72. The AMIS Rapid Response Forum would provide policy advice and promote policy
coordination when the market situation and outlook indicates a high food security risk. Through the
participation of policy experts from the major producing and importing countries AMIS Rapid Policy
Response Forum will be able to mobilise political support to achieve agreement on appropriate policy
response and actions in times of crisis.

73. The Rapid Response Forum will meet in response to a food crisis alert. Its actions would be as
follows:

receive and assess information and analyses from the AMIS Secretariat on the current global
market situation and outlook and issue regular statements on the ensuing implications for food
security; receive information and assessments for particularly vulnerable countries.
provide appropriate policy guidance and promote policy coordination when the market
situation and outlook indicates a high food security risk. Such guidance will encourage the
implementation of efficient and effective policies, the avoidance of potentially damaging policy
choices, and will ensure that humanitarian responses are rapid and appropriate.
work closely with the Committee on World Food Security (CFS) to promote greater policy
convergence and strengthen policy linkages at global level.

20
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 21

Recommendation 2

Building upon existing mechanisms, establish an Agricultural Market Information System encompassing four elements.
G20 governments commit to instruct statistical or other relevant agencies to provide timely and accurate data on
food production, consumption, and stocks. Where the mechanisms and institutions are not in place nationally to
do so, G20 governments should undertake to create them.
International Organizations, with broad involvement of countries (G20 and other relevant players) commit to
undertake monitoring, reporting and analysing of current conditions and policy developments in major markets as
well as to enhance global food security by encouraging information sharing, improving data reliability and
increasing transparency, and introducing a global early warning system.
G-20 governments support the establishment of a Rapid Response Forum, with broad involvement of countries
(G20 and other relevant players) building on the proposed Agricultural Markets Information System to promote
policy coherence and coordination in times of crisis.
International Organizations support the improvement of national or regional systems to monitor stocks,
production, forecasts (with improved modelling and weather forecasting), food and nutrition security and
vulnerability, in order to enhance Early Warning Systems in vulnerable developing countries and regions.
Concrete proposals on the implementation of this Agricultural Market Information System are detailed in a
comprehensive scoping proposal made available as a separate document.

4.2 International food stocks

74. Buffer stocks attempt to influence prices rather than to provide emergency relief in a crisis. At
the international level buffer stocks have been an important characteristic of commodity markets in the
past. However, the various international commodity agreements which provided for stockholding or
supply controls to stabilise prices have either collapsed or been replaced by agreements whose main role
is market information provision.

75. Historically, international buffer stock mechanisms are widely judged to have had limited
success in reducing the volatility of prices. They have been more effective in moderating downward price
movements than price surges. In the case of a price surge, a buffer stock agency can only release in the
market what it has previously bought, and once its stock is exhausted there are no further means to curb
price increases.28

76. Attempting to stabilise prices using buffer stocks is potentially very costly. Stabilising world
prices around a level either lower or higher than that determined by market fundamentals requires
significant resources. Attempts to defend a price ceiling and reduce the average world level of food
prices over time can lead to substantial costs. Buffer stocks set to defend against price spikes are also
vulnerable to speculative attacks. If speculators perceive that the stocks held by the stabilization agency
are insufficient to maintain the target lower price level, they will compete to buy the entirety of the stock
in order to take advantage of likely profits.29

4.3 Futures markets

77. Futures markets perform several functions: they provide the instruments to transfer price risk,
they facilitate price discovery and arguably, increasingly in recent years they are offering commodities as
an asset class for financial investors, such as fund and money managers who had not previously been
present in these markets.

78. Commercial participants utilize futures contracts to hedge, or insure their crops or inventories
against the risk of fluctuating prices. For example, processors of agricultural commodities, who need to
obtain inventories, buy futures contracts to guard against future price rises. If the price rises, they use the

21
22 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

increased value of the futures contract to offset the higher cost of the physical quantities they need to
purchase.

79. Speculators also trade in the futures markets; they buy and sell futures contracts and take on the
risk of future price fluctuations to gain a risk premium. They are non-commercial participants as they
have no involvement in the physical commodity trade in contrast to commercial participants, such as
farmers, traders and processors.30

80. Since the beginning of the last decade, commodity derivative markets, including those for
agricultural commodities, have experienced significant inflows of funds from non-traditional investors,
such as commodity index funds, swap dealers and money managers. These financial investors hold large
futures positions including in basic food commodities such as wheat, maize and soybeans as well as in
cocoa, coffee and sugar.

81. Another essential function of futures markets is to facilitate price discovery. Price discovery is
the continuous process by which futures prices are reassessed by buyers and sellers as new information
becomes available. Market participants continuously update their expectations as both public and private
information become available. They adjust their market behaviour and through their transactions,
information is incorporated into the price.

82. Speculators are necessary for the performance of both these functions. They buy and sell
futures contracts and take on the risk of price fluctuations to earn a profit on price movements. By doing
so, they provide the market liquidity which enables commercial hedgers to find counterparties in a
relatively costless manner. Too little non-commercial participation results in low liquidity and potentially
in large seasonal price swings. 31 Too much non-commercial participation can cause frequent and erratic
price changes. This is the case when speculators assume that past developments carry information on
future price movements, giving rise to trend chasing. This will result in buying after prices rise and
selling after prices fall, independently of any changes in market fundamentals.32

83. The debate on whether speculation stabilizes or destabilizes prices resumes with renewed
interest and urgency during high price episodes. Some analysts purport that the influx of financial
investors in commodity futures markets has scant impact on market prices.33 Other analysts stress that the
large amount of money invested in commodity futures by financial investors has amplified price
movements to an extent which cannot be explained by market fundamentals.34 More research is needed
to clarify these questions and in so doing to assist regulators in their reflections about whether regulatory
responses are needed and the nature and scale of those responses.35

84. Despite these differences, there is widespread agreement that for agricultural commodity
derivatives markets to function well, and as intended in terms of hedging and price discovery,
appropriate regulation needs to be in place across all relevant futures exchanges and markets. In
particular, there is need for greater transparency about transactions across futures markets and especially
across over-the-counter (OTC) markets, where transactions take place off the regulated commodity
exchanges. Comprehensive trading data need to be reported to enable regulators and participants to
monitor information about the frequency and the volume of transactions to understand what is driving
commodity prices.36 Such data exist for some commodity exchanges (though perhaps in an aggregate
form which makes the identification of various participants difficult), but are currently unavailable for
off-exchange trading.

85. The specific nature of the regulatory framework for futures exchanges and OTC markets,
whether for agriculture or other commodities, is an issue best addressed by financial market regulators.
And indeed significant regulatory changes have already been decided or are under consideration in
several important jurisdictions. In the United States, the Dodd-Frank Wall Street Reform and Consumer
Protection Act (2010) has mandated a tightening of financial market regulation to improve transparency

22
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 23

and reduce systemic default risk in the over-the-counter (OTC) derivatives trade. In the European Union,
with the same objectives, the Commission has adopted a proposal for regulation of OTC derivatives
trading and is currently reviewing several key directives that regulate financial markets including the
Market Abuse Directive and the Markets in Financial Instruments Directive.

86. In addition to the long-established markets in the United States and Europe, agricultural
commodity futures exchanges also exist in some emerging-market members of the G20, including Brazil,
China, India and South Africa.37 Price developments in most of the contracts traded on these exchanges
closely follow the developed-country exchanges where price discovery provides global benchmarks.
Trading on local platforms allows exchange-rate risk to be avoided and reduces basis risk stemming from
a variety of factors: climatic conditions and different seasonal timings (South Africa), restrictions on
international and domestic trade (China), differences in quality specifications and difficulty in delivering
to overseas markets (Brazil). Some exchanges, e.g. in India, offer exchange trading for commodities
(e.g. cardamom and mentha oil) for which contracts exist nowhere else. All of these futures exchanges
are established venues for price-risk management through futures contracts on internationally traded
commodities and they have a highly although not necessarily heavily regulated environment.38

87. More generally, debate is on-going at national and international level about the possible merits
of the following actions in terms of transparency and improved market functioning:39
Establish a trade depository to register OTC contracts, in line with earlier decisions in the G20
Summit in 2009 in Pittsburgh.
Use of speculative position limits on commodity futures contracts to ensure control of undue
market influence.
Use of maximum limits to daily price changes to reduce volatility.
Use of limits on inventories held in delivery warehouses by non-commercial entities to limit
market manipulation possibilities.
Introduction of provisions for high volume and frequency trading into the regulatory regime.40
Ensuring that changes in regulation are adopted across commodity exchanges and across
countries in order to avoid the migration of participants and regulatory arbitrage.41
88. Beyond regulatory concerns, new futures instruments for mitigating commodity price risk
exposure might be explored. For example, a global wheat contract that would specify export delivery
points in the major producing regions has been proposed.42 The potential advantages of a global futures
contract with compulsory delivery include: identifying cheapest to deliver sources by designating
delivery points all over the world; acting as a global signalling system of both price and regional supply
availabilities; and attracting well-informed commercial entities while deterring non-commercial entities
from investing on such contracts. The development of such new instruments lies with the futures
exchanges and the market participants.

Recommendation 3
G20 governments recognize the need to improve information and transparency in futures and over-the-counter
markets and encourage appropriate rules to enhance their economic functions paying attention to the need for
harmonization across exchanges in order to avoid regulatory arbitrage.

Proposed changes should be considered in light of the on-going review of regulatory oversight of all financial
markets and not solely agricultural commodity markets, in particular by G20 Finance Ministers and Central Bank
Governors.

The G20 supports the efforts made by the United States, the European Commission and others in addressing
transparency and efficiency issues in futures markets.

23
24 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

4.4 Domestic and trade policies

Reducing import barriers, trade distorting domestic support, and all forms of export subsidies

89. Price volatility may originate from either domestic or international markets.43 Trade is an
excellent buffer for localised fluctuations originating in the domestic market. Though stockholding is a
necessary component of a well functioning market, in particular to smooth out seasonal fluctuations and
time lags in trade, year-to-year variations in domestic production can be more effectively and much less
expensively buffered by adjustments in the quantities imported or exported.44 To the extent that shocks
tend to be specific to individual regions of the globe, and to partly cancel out on a worldwide level, world
output of a given agricultural product is far less variable than output in individual countries. International
trade is therefore a potentially powerful engine to even out supply fluctuations across the globe, and as a
result to reduce market volatility. To fulfil this beneficial pooling function to the maximum degree, trade
has to be able to flow between nations and the tendency which has emerged, in recent crises, for
countries to try to insulate themselves from international markets needs to be reversed.

90. More generally and in the longer term context, trade is an essential component of any food
security strategy. There is significant potential for increased production in many parts of the world, but
not all countries everywhere can or should aspire to supplying all their own needs. Doing so is
excessively costly, and will reduce choice and quality, without providing the reliability needed to achieve
food security. Moreover, the impact that climate change will have on food production is uncertain, but
many experts concur that it will lead to a worsening of conditions for agricultural production in some
countries or regions already facing difficult climatic and natural conditions. Experts also agree that there
will be an increase in the incidence of extreme events such as drought, heat-waves, and floods. A
combination of better functioning and deeper markets for agricultural commodities and improved supply
capacity and resilience will allow countries in the most vulnerable zones to overcome these problems.

91. Trade policy restrictions are not the only impediment to the free flow of goods and services.
Market and transportation infrastructure, the capacity to meet sanitary and phyto-sanitary requirements
and many other factors will determine a countrys capacity to export. Initiatives such as the Aid-for
Trade programme being implemented by the WTO and the OECD are contributing to overcoming some
of these domestic barriers to trade. A stepping up of this effort could bring significant benefits to
developing countries with export potential in agriculture.

92. Policies that distort production and trade in agricultural commodities potentially impede the
achievement of long run food security, by stimulating or conserving production in areas where it would
not otherwise occur and by distorting, obscuring or impeding the transmission of price signals to
competitive producers elsewhere. The efficient functioning of food supply chains domestically and
internationally also requires attention to be paid to competition policy issues upstream and downstream
of the farm sector, as a necessary complement to agricultural and trade policy reforms.

93. Despite on-going reforms there are still significant barriers to trade in agricultural commodities
among developing countries and between developing and OECD countries. They contribute to the
thinness of international markets that has been blamed for some of the volatility experienced in recent
years. Average tariffs on agricultural and food are high for middle income and high income countries,
25% and 22% respectively.45 Protectionism on agricultural products is not only higher than on non
agricultural products (by a factor of four), it is also much more volatile.46 Agricultural trade policies are
designed to insulate domestic prices from world markets and lead to pro-cyclic effects: protection
decreases when prices are high, increasing demand on world markets, and increases when world prices
are low effectively operating as a variable levy. Therefore, large country trade policies increase world
price volatility and create negative externalities for smaller countries.47 A conclusion of the Doha Round
will reduce the scope to implement destabilizing policies on world markets by reducing the bound level
of tariffs and subsidies.

24
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 25

94. OECD monitors support and protection of agriculture in its own member countries and in some
emerging countries that are major players in global food production and consumption. While trends in
the indicators measuring support and protection all point towards a continuing reduction in the levels and
the intensity of distorting interventions, much still needs to be done. Latest data for the OECD countries
indicate that government support still accounts for 22% of the total receipts of agricultural producers and
that more than half of that support is delivered in ways that are highly distorting of trade and
competition48.

Clarifying and strengthening provisions concerning export restrictions

95. Under WTO disciplines, quantitative restrictions are generally prohibited by Article XI of
GATT 1994 Agreement but an exception allows governments to prohibit or restrict exports on the
condition that these measures are [...] temporarily applied to prevent or relieve critical shortages of
foodstuffs or other products essential to the exporting contracting party.

96. Export prohibitions or restrictions relating to foodstuffs must also conform with the provisions
of the Agreement on Agriculture, that requires WTO Members to give due consideration to the effects of
such prohibition or restriction on importing Members' food security, give notice in writing, as far in
advance as practicable, and consult, upon request, with other WTO Members. These provisions do not
apply to a developing country Member, unless the measure is taken by a developing country Member
which is a net-food exporter of the specific foodstuff concerned.

97. These disciplines are considered to have been insufficient and weak during the 2007-2009
period, when export restrictions exacerbated or even, according to most experts, caused severe disruption
and a collapse in confidence on international markets. Export restrictions have also contributed to the
price increases and general market nervousness currently being experienced.49 It has been estimated that
if countries are free to implement export taxes a 10 percentage point increase in world prices can be
amplified to between 20 and 50 percentage points. In addition, the risk of export restrictions, and the
asymmetry between international disciplines (e.g. in WTO agreements) on export restrictions (unbound)
and import restrictions (bound) is a severe barrier to increasing trust in international markets. To be sure
that international trade is a reliable source of food supply net food importers should benefit from much
stronger guarantees from their trading partners. A first best option would be to a ban on export
restrictions. Countries would address domestic food security issues with direct and targeted support.
However, it is most unlikely that a ban on export restrictions would be agreed and, even if agreed, that it
would be enforced during a food crisis. On the other hand, reinforced rules, in particular in terms of
transparency, are both possible and useful.

98. As agricultural markets become more open, alternative measures addressing market
imperfections, many of which are policy driven, are needed. This would ensure that the potential new
opportunities created can actually be exploited by competitive suppliers. In the case of less developed
countries in particular, investments in improving supply capacity, including Aid for Trade will be
important.

25
26 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

Recommendation 4

G20 governments demonstrate leadership in on-going WTO DDA negotiations, moving immediately to strengthen
international disciplines on all forms of import and export restrictions, as well as domestic support schemes, that
distort production incentives, discourage supply in response to market demand, and constrain international trade of
food and agriculture products. Specifically,
substantially improve market access, while maintaining appropriate safeguards for developing countries,
especially the most vulnerable ones;
substantially reduce trade distorting domestic support, especially by developed countries; and,
eliminate export subsidies.
Taking existing WTO rules into account and the state of play in the DDA negotiations G20 governments should:
develop an operational definition of a critical food shortage situation that might justify consideration of an export
restricting measure. An export ban would be defined as a time-limited measure of last resort, allowed only
when other measures, including triggering domestic safety net measures for the poorest, have been exhausted,
and taking into account, in particular, the food security needs of least developed countries and net food
importing developing countries.

widen, strengthen and enforce consultation and notification processes currently in place at the WTO. The
intention to impose an export restriction would have to be notified in advance of the action being applied and a
fast track consultation process could be put in place to discuss whether the measure can be avoided and
how. Consultation should be on-going and regular with a view to ensuring that the measure, once in place, is
removed at the earliest possible moment.

99. With respect to export restrictions nations have agreed to commit to make humanitarian
exemptions, first, at the G8 Summit in LAquila in July 2009, and then at the World Summit on Food
Security in Rome in November 2009, where all FAO member states agreed to remove food export
restrictions or extraordinary taxes for food purchased for non-commercial humanitarian purposes, and to
consult and notify in advance before imposing any such new restrictions. If honoured these
commitments would allow food to be shipped rapidly to where it is needed in an emergency.

100. Some nations that imposed export restrictions during 2008 and 2010 made exemptions for
purchases of humanitarian food, including those by the WFP. However, others have not made such
exemptions, forcing in-country and international humanitarian agencies to purchase food from more
distant sources. And most exemptions, if made, are on a case-by-case basis after concern has been raised
and the exemption requested. Valuable emergency response time and resources are lost, as procurement
teams have to spend time negotiating, or find alternative suppliers from other regions.

Recommendation 5

G20 governments strengthen the commitments made at the LAquila and Rome Summits, calling on all nations
to allow purchases of humanitarian food, especially by WFP, to be exempted from food export restrictions
and/or extraordinary taxes, so that humanitarian food can be purchased, exported and/or transited regardless
of any prohibitions, restrictions or extraordinary taxes imposed; and resolve to bring this commitment and call to
the UN General Assembly and to the WTO.

Reducing policy conflicts between food and fuel

101. Between 2000 and 2009, global output of bio-ethanol quadrupled and production of biodiesel
increased tenfold; in OECD countries at least this has been largely driven by government support
policies.50 Moreover, trade restrictions by favouring domestic sources of raw material for biofuels do not
maximise expected environmental benefits. Biofuels overall now account for a significant part of global
use of a number of crops. On average, in the 2007-09 period that share was 20% in the case of sugar

26
 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES 27

cane, 9% for both oilseeds and coarse grains (although biofuel production from these crops generates by-
products that are used as animal feed), and 4% for sugar beet.51 With such weights of biofuels in the
supply-demand balance for the products concerned, it is not surprising that world market prices of these
products (and their substitutes) are substantially higher than they would be if no biofuels were produced.
Biofuels also influence products that do not play much of a role as feedstocks, for example wheat,
because of the close relations between crops on both the demand side (because of substitutability in
consumption) and the supply side (due to competition for land and other inputs).

102. At the international level, crop prices are increasingly related to oil prices in a discrete manner
determined by the level of biofuel production costs. Increases in the price of oil enhance ethanols
competitiveness relative to petrol and strengthen its demand. Since both energy and food/feed utilise the
same input, for example grain or sugarcane, increases in the production of ethanol reduce the supply of
food and result in increases in its price. This relationship between the prices of oil, biofuels and crops
arises due to the fact that, in the short run, the supply of crops cannot be expanded to meet the demand by
both food and energy consumers.

103. If oil prices are high and a crops value in the energy market exceeds that in the food market,
crops will be diverted to the production of biofuels which will increase the price of food (up to the limit
determined by the capacity of conventional cars to use biofuels - in the absence of flexfuel cars and a
suitable distribution network). Changes in the price of oil can be abrupt and may cause increased food
price volatility. Support to the biofuel industry also plays a role. Subsidies to first-generation biofuel
production lower biofuel production costs and, therefore, increase the dependence of crop prices on the
price of oil. Such policies warrant reconsideration.52 53

Recommendation 6
G20 governments remove provisions of current national policies that subsidize (or mandate) biofuels production or
consumption. At the same time, governments should:
Open international markets so that renewable fuels and feed stocks can be produced where it is economically,
environmentally and socially feasible to do so, and traded more freely.
Accelerate scientific research on alternative paths to reduced carbon emissions and to improved sustainability
and energy security.
Encourage more efficient energy use, including in agriculture itself, without drawing on finite resources,
including those needed for food production.
Failing a removal of support, G20 governments should develop contingency plans to adjust (at least temporarily)
policies that stimulate biofuel production or consumption (in particular mandatory obligations) when global markets
are under pressure and food supplies are endangered.
Some ideas for how this could be done are explored in Annex D.

27
28 PRICE VOLATILITY IN FOOD AND AGRICULTURAL MARKETS: POLICY RESPONSES

4.5 Dealing with waste

104. Waste has been identified as an important issue which affects the underlying supply-demand
balance for food.

In developing countries, post-harvest and post production losses due to inadequate

infrastructure, poor storage facilities, inadequate technical capacity and under-developed markets
are the main causes of waste.

In OECD countries and increasingly in emerging economies waste occurs in the distribution
system, in the restaurant sector and at home,54including parts of food products which are not
economical to use; food that does not meet cosmetic standards, plate waste; food that is
discarded because it spoiled, and inefficient use of food , contributing to obesity.

105. Most losses are avoidable to some degree and some types of waste could be almost entirely
eliminated. Reducing waste could be an important part of a strategy to improve food security while
reducing environmental and resource pressures. If food waste can be reduced, the increase in production
estimated to be needed to meet the increase in demand over the next 40 years would be smaller.
Reducing waste would also help to reduce the pressure on land, water stress, soil degradation, and
greenhouse gas emissions.55 In developing countries, the measures proposed in Chapter 3 to improve the
overall resilience and productivity of agriculture should address much of the problem of waste from post-
harvest losses. In developed countries, possible avenues for policy action could include engaging with
the private sector, to increase awareness and develop voluntary agreements, reviewing regulations that
may inadvertently generate avoidable waste, supporting research to improve storage, prolong shelf life
and better detect deterioration, implementing public education campaigns, and investment in better
assessment and monitoring.

5. Policy options to deal with the consequences of price volatility, particularly for the most
vulnerable

5.1. Coping with volatility in the short run: buffer stocks, emergency food reserves, international
and national safety nets
National buffer stocks

106. Buffer stocks are an important policy instrument in a number of emerging economies and
developing countries, though they have been virtually abandoned in developed countries. Some rice
producing Asian countries rely on a combination of rice reserves, import or export monopolies, and
domestic procurement to stabilise prices within a pre-determined band. These measures aim to stabilise
domestic rice prices and, in some cases, have stimulated agricultural growth. In Africa, the experience
with maize buffer stocks is mixed.

107. The operational costs of buffer stocks are significant. Appropriate storage infrastructure is
extremely costly to acquire, and buying the food stock and holding it is also very expensive. Domestic
procurement, food releases from buffer stocks and trade programmes require continuing budgetary
allocations to cover any operational losses occurring in domestic and international trading. Losses
incurred on behalf of policy-dictated objectives for price stabilization may be viewed as direct subsidies.
Although expenditures associated with the acquisition and holding of stocks for food security purposes
can qualify under the WTO Green Box,56 from a WTO point of view, such price stabilisation
mechanisms could also be considered as trade distorting support. In times of price increases, such costs
can escalate to significant levels, rendering buffer stocks ineffective in containing price surges.

28
Food Sec.
DOI 10.1007/s12571-015-0427-z

ORIGINAL PAPER

Feeding 9 billion by 2050 Putting fish back on the menu

Christophe Bn & Manuel Barange & Rohana Subasinghe &
Per Pinstrup-Andersen & Gorka Merino &
Gro-Ingunn Hemre & Meryl Williams

Received: 11 January 2015 / Accepted: 4 February 2015

# The Author(s) 2015. This article is published with open access at Springerlink.com

Abstract Fish provides more than 4.5 billion people with at
least 15 % of their average per capita intake of animal protein.
Fishs unique nutritional properties make it also essential to
the health of billions of consumers in both developed and
developing countries. Fish is one of the most efficient con-
verters of feed into high quality food and its carbon footprint is
lower compared to other animal production systems. Through
fish-related activities (fisheries and aquaculture but also pro-
C. Bn (*)
Institute of Development Studies, University of Sussex, Sussex, UK
e-mail: c.bene@cgiar.org
M. Barange
Plymouth Marine Laboratory, Plymouth, UK
e-mail: MABA@pml.ac.uk
R. Subasinghe
Food and Agriculture Organization, Rome, Italy
e-mail: Rohana.Subasinghe@fao.org
P. Pinstrup-Andersen
Cornell University Ithaca, Ithaca, USA
e-mail: pp94@cornell.edu
cessing and trading), fish contribute substantially to the in-
come and therefore to the indirect food security of more than
10 % of the world population, essentially in developing and
emergent countries. Yet, limited attention has been given so
far to fish as a key element in food security and nutrition
strategies at national level and in wider development discus-
sions and interventions. As a result, the tremendous potential
for improving food security and nutrition embodied in the
strengthening of the fishery and aquaculture sectors is missed.
The purpose of this paper is to make a case for a closer inte-
gration of fish into the overall debate and future policy about
food security and nutrition. For this, we review the evidence
from the contemporary and emerging debates and controver-
sies around fisheries and aquaculture and we discuss them in
the light of the issues debated in the wider agriculture/farming
literature. The overarching question that underlies this paper
is: how and to what extent will fish be able to contribute to
feeding 9 billion people in 2050 and beyond?
Keywords Fish . Food security and nutrition . Micro-nutrient
deficiency . Fisheries . Aquaculture

G. Merino
AZTI, Pasaia, Gipuzkoa, Spain
e-mail: gmerino@azti.es
Introduction

G.<I. Hemre Fish is critically important to food security and good nutrition
Institute of Nutrition and Seafood Research, Bergen, Norway (Allison 2011; Thilsted 2012; Beveridge et al. 2013). Fish and
e-mail: Gro-Ingunn.Hemre@nifes.no
other aquatic foods are high in protein and contain many es-
M. Williams sential micronutrients. The fishery and aquaculture sectors are
17 Agnew Street, Aspely, Queensland, Australia the source of income for millions of women and men in low-
e-mail: meryljwilliams@gmail.com income families (Bn 2006), thus contributing directly and
indirectly to their food security (Bn et al. 2007; Allison
Present Address:
C. Bn 2011; World Bank/FAO/WorldFish 2012). Yet, the potential
International Center for Tropical Agriculture, Cali, Colombia contributions of fish to food security and nutrition (FSN) are

all but ignored in the international debate, as if a firewall stood
between the discussions about the role of fish and the broader
debates about FSN issues. A recent review of international
development and research agencies working on FSN revealed
for instance that Bfish is strikingly missing from strategies for
reduction of micronutrient deficiency, precisely where it could
potentially have the largest impact^ (Allison et al. 2013: 45).
The purpose of this paper is to make a case for a closer
integration of fish into the overall FSN debate. As such the
discussion is directed not to those in the fisheries and aqua-
culture communities who have been relentlessly advocating
for fish as an entry point in this debate (see e.g., Bn et al.
2007; Hall et al. 2011; Thilsted 2012), but to the vast majority
of the international experts who are influencing the wider
debate and policies on FSN.
The World Committee on Food Security (CFS) took an
important step in the right direction in 2012 by requesting that
the High Level Panel on Food Security (HLPE) undertook an
in-depth study of the role of sustainable fisheries and aquacul-
ture for FSN. The report (HLPE 2014) presents the strongest
case yet for incorporating fish into the debates on how to
achieve FSN for all and suggests a set of activities and policies
to be pursued.
Building on this report and a large body of literature, our
ambition is to bring fish onto the table. A number of papers
have succeeded in presenting a very comprehensive evidence-
based argument for supporting fish as a central element in the
FSN debate (see e.g., Prein and Ahmed 2000; Roos et al.
2003; 2007; Kawarazuka and Bn 2011; Beveridge et al.
2013). Here, we aim to review the evidence from the contem-
porary and emerging debates and controversies around fisher-
ies and aquaculture. In doing this we do not go into the micro-
level issues concerning social, gendered and micro-economic
aspects of fish, FSN at the local levels for those we refer
readers to the HLPE report (HLPE 2014).
Some of the underlying questions in the present study are
closely linked to the issues discussed in the wider agriculture/
farming literature. The latest estimate suggests for instance
that, in 2009, fish accounted for 17 % of the global popula-
tions intake of animal protein and 6.5 % of all protein con-
sumed (FAO 2014). In the context of this Special Issue on
BFeeding 9 Billion^, the key questions are: a) Can we main-
tain these consumption rates, given the projected growth in
human population and the growing environmental challenges
facing the earth? and b) Could fish - one of the most efficient
converters of feed into high quality food be a substitute for
other sources of animal protein? Finally, in the context of the
current discourse about future resource scarcity, should fish be
considered a more environmentally friendly source of protein
than the other livestock production systems?
Fish is more than just a source of animal protein. Fish
contains several essential amino acids, especially lysine and
methionine. The lipid composition of fish, with the presence
C. Bn et al.
of long-chain, poly-unsaturated fatty acids (LC-PUFAs), is
unique. Fish is also an important source of essential
micronutrients vitamins D, A and B, and minerals (calcium,
phosphorus, iodine, zinc, iron and selenium), which makes it
particularly attractive in the current fight against malnutrition
in low income and food deficient countries (LIFDCs). Some
countries (e.g., Zambia, Brazil, or Chile) have already recog-
nized this potential and have included fish in their national
school-feeding programmes. Should this be more systemati-
cally considered in countries with high levels of malnutrition
to complement or substitute the technology-dependant (and
expensive) bio-fortification programmes?
In addressing these issues we must first acknowledge that
fish production sectors and associated value chains have been
remarkably dynamic in the last three decades what some
refer to as the blue revolution. Every second fish we con-
sume is now produced in aquaculture. This new situation is a
result of the stagnation of capture fisheries following decades
of expansion, some of it carried out in an unsustainable man-
ner (FAO 2012), combined with impressive growth rates in
the aquaculture industry, subject to some major adjustments to
overcome the early challenges such as fish disease and nega-
tive environmental impacts. Critics of aquaculture have also
pointed to the use of fishmeal and fish oil produced from wild
fish, to feed farmed fish. These concerns are valid and need to
be taken (along with their solutions) into account in this
discussion.
Understanding fish production in relation to food security
and nutrition
The fastest growing food-supply industry in the world
In 2011, 173 Mt of fish were extracted from the global
marine and inland water ecosystems, of which 7 to 10
Mt were discarded prior to landing and 12 Mt were lost
at the post-harvest stage. The production from capture
fisheries and aquaculture available amounted therefore
to 154 Mt, of which about 131 Mt were utilized directly
for human consumption (Fig. 1) (HLPE 2014). Thanks
to aquaculture and fisheries, the global supply of fish
has grown by a factor of 8 since 1950. By comparison,
even after the Green Revolution the world rice produc-
tion increased only by a factor of 3. In effect, fish
production has been the fastest growing food industry
in the world for the last 40 years and is expected to
remain so in the near future. The fish supply per capita
has more than tripled in the last half century, from
6 kg/year in 1950 to 18.8 kg/year in 2011 Fig. 2.
In fact, the world fish supply has effectively been grow-
ing faster than the worlds population (FAO 2012).
These global figures mask, however, some important
Feeding 9 billion by 2050 Putting fish back on the menu

Fig. 1 World fish utilization. Discard 7-10 mill tonnes

Notes: a 2011 estimates. b 2010 Post-harvest losses 12 mill tonnes
estimates. Source data from FAO
(2012) and Shepherd and Jackson
(2013)

3.2% growth (since 1950)

173 mill tonnes

131 mill tonnes Direct Human Food
18.8 kg/capita/year (a)

Fish 73%
Pig 20%
17 mill tonnes Fishmeal, Fish oil (b) Poultry 05%
Others 02%
6 mill tonnes Ornamental, bait, etc.

Notes: (a) 2011 estimates. (b) 2010 estimates.

Source data from FAO (2012) and Shepherd and Jackson (2013).

regional variations: fish consumption is the lowest in
Africa (9.1 kg per capita in 2009), while Asia accounts
for almost two-thirds of total consumption (20.7 kg per
capita).
Fish as the largest source of animal protein
Fish is a major source of animal protein, overshadowing
most other sources. In 2010 it represented a source
twice as important as poultry, and three times larger
than cattle (Fig. 3). Today capture fisheries and aqua-
culture provide 3 billion people with almost 20 % of
their average per capita intake of animal protein, and a
further 1.3 billion people with about 15 % of their per
capita intake (HLPE 2014). This share can exceed 50 %
in some countries. In West Africa, Asian coastal coun-
tries, and many small island states, the proportion of
total dietary protein from fish can reach 60 % or more
(e.g., Gambia, Sierra Leone, Ghana, Cambodia,
Bangladesh, Indonesia, Sri Lanka, or the Maldives)
(FAO 2014).
The geography of fish as a source of protein is also signif-
icant in the FSN discussion. Twenty-two of the 30 countries
where fish contribute more than one-third of the total animal
protein supply were officially referred to as LIFDCs in 2010
(Kawarazuka and Bn 2011). In other words, almost three-
quarters of the countries where fish is an important source of
animal protein are poor (income-wise) and food-deficient. Yet

Fig. 2 Relative contribution of

aquaculture and capture fisheries
to production and food fish
supply. Source: FAO Statistics
and Information Branch of the
Fisheries and Aquaculture
Department

Source: FAO Statistics and Information Branch of the Fisheries and Aquaculture Department.

Fig. 3 World Production of the
main sources of animal protein
over the period 19602010.
Source: FAO Stat
in these LIFDCs, even if fish is a substantial proportion of the
food intake, undernourishment can still occur as total food
intake is often insufficient.
Fish - beyond protein
Some point out, however, that the main contribution of fish to
FSN may not be in relation to its protein content, but its lipid
and micro-nutrient content.
The lipid composition of fish is unique, having LC-PUFAs,
with many beneficial effects for child development and adult
health (Thilsted et al. 1997; Larsen et al. 2011; Richardson and
Montgomery 2005). Among fish species that are cheaper in
developing countries, small pelagic fish such as anchovy and
sardine are some of the richest in LC-PUFAs (USDA 2011),
especially compared to large freshwater fish such as carp and
tilapia. When its rich nutrient content is preserved, fish can
provide protective effects against a wide range of health is-
sues. LC-PUFAs for instance provide protection against dis-
eases such as stroke, high blood pressure or coronary heart
disease (Miles and Calder 2012; Rangel-Huerta et al. 2012).
Complementing its fatty acid content, fish is also known to
be an important source of essential micronutrients vitamins
D and B, and minerals (Roos et al. 2003; 2007; Bonham et al.
2009). Lipid-rich fish also contain vitamin A. Recent research
showed that fish species consumed whole with bones, heads,
and viscera play a critical role in micronutrient intakes of
people as these parts are where most micronutrients are con-
centrated. Some of these small fish (such as mola
(Amblypharyngodon mola), darkina (Esomus danricus), sar-
dines and pilchards, anchovy, seabass, tilapia) contain high
levels of minerals such as calcium, phosphorus, iodine, zinc,
iron and selenium, which are low in other foods. The potential
contribution that fish (even in small quantity) can therefore
C. Bn et al.
offer to address multiple micronutrient deficiencies is now
being recognized (e.g., Roos et al. 2007; Kawarazuka and
Bn 2011; Thilsted 2012). For instance, the high level of
iodine found in fish can help prevent iodine deficiency which
is known to cause cretinism (stunted growth and mental
retardation).
Current issues in relation to fish contribution to food
security
Fishmeal and fish oil in aquaculture
In addition to being used directly as human food, fish also
contributes indirectly to human nutrition when it is used as
fishmeal for aquaculture and poultry/livestock feeds (Tacon
and Metian 2009). In 2011, 23 Mt of fish essentially small
pelagic fish species such as anchovy, herring, mackerel and
sardine have been destined to non-direct human consump-
tion, of which 75 % (17 Mt) was reduced to fishmeal and fish
oil for aquaculture, poultry and other livestock feeding (cf.
Fig. 1). In 2010, 73 % of the total world fishmeal was used
to feed farmed fish, followed by pigs (20 %), poultry (5 %)
and others (2 %) (Shepherd and Jackson 2013).
From a FSN perspective, the use of fishmeal for farmed
fish (and livestock) raises important issues. Leaving aside the
debate on the role of small pelagic fish in supporting larger
fish, birds and marine mammals in the ecosystem (Smith et al.
2010), is fishmeal the most efficient way to use fish (especial-
ly low-cost small pelagic fish rich in LC-PUFA) or would
these fish contribute more to food security if a larger share
of them was eaten directly by people? Indeed, despite some
substantial improvement in the last decade, the rate of conver-
sion of fishmeal to fish is still a source of concern (Troell et al.
Feeding 9 billion by 2050 Putting fish back on the menu
2014). On average, for every kg of farmed fish produced,
0.7 kg of wild fish is needed (Tacon and Metian 2008). This
average figure, however, masks important differences: while
for omnivorous farmed fish, the rate is down to an acceptable
0.2 to 1.41 kg of wild fish per 1.0 kg of farmed fish, for
carnivorous farmed fish, the figure is higher: 1.35 to 5.16 kg
to produce one kg of farmed fish (Boyd et al. 2007).
Fish losses and implications on food security and nutrition
The global discards of fish (fish caught but dumped overboard
due to low quality, damage or spoilage, non-targeted species
or below regulation size) were estimated to be around 7.3 Mt
in 2005, 80 % of which coming from industrial fleets. In
contrast, small-scale fisheries generate less wastage in the
form of discards (about 2 Mt a year, that is, 4 % of their
estimated landing1) (Kelleher 2005).
With increasing fish scarcity and increasing fish prices,
species previously deemed commercially inferior are progres-
sively integrated into consumer eating habits and markets.
Most shrimp trawlers, which used to discard up to 95 % of
their catch, are now landing more bycatch for human con-
sumption (Bn et al. 2007). In small-scale fisheries where
discarding fish is rare, substantial quantities and quality are
lost due to post-harvest mishandling during transport, storage,
processing, on the way to markets and waiting to be sold.
Especially in developing countries, where access to electricity
and cold chain can still be an issue, fish post-harvest losses can
be significant. Estimated at 1012 Mt in 2005, the total ac-
counts for 10 % of global capture and culture fisheries pro-
duction (Bn et al. 2007). While physical losses in small-
scale fisheries are less than 5 %, economic losses can be sub-
stantial. In Africa, some estimates (FAO Focus nd) put post-
harvest losses in some cases to levels as high as 2025 %. In
aquaculture, waste streams in value chains have been large in
the past but are tending to decline rapidly as competitive pres-
sures force innovation (Arthur et al. 2013).
Sustainability of fisheries and implications for food security
Because total fish production (availability) is an important
dimension of food security, a key issue in this debate is the
environmental sustainability of capture fisheries. The extent to
which capture fisheries have exceeded sustainable levels has
generated strong expert and public opinion debates (Worm
et al. 2006; Pauly et al. 2005; Hilborn 2013) and many media
headlines, scientific papers and environmental campaigns
have been in the last two decades framed around the idea that
world fisheries resources are in crisis due to overfishing (see
e.g., Pauly et al. 1998; Myers and Worm 2003). FAO has
1
Overall, small scale fisheries land approximately 40 Mt annually, com-
pared to 50 Mt by large scale fisheries (HLPE 2014).
expressed a more nuanced but nonetheless concerned view
about the state of world marine fisheries (e.g., FAO 2012),
acknowledging the granularity in the state of resources world-
wide. The current consensus is that global fisheries would be
more productive if the levels of overfishing were reduced
(Srinivasan et al. 2010), and the environmental sustainability
of fisheries were recognized to be a sine qua none condition
for FSN (HLPE 2014).
An important point in this debate, however, is that improv-
ing FSN through fisheries would depend on stock recovery
and also on access to and distribution of the harvest, as well as
gender consideration (de Schutter 2012; Williams 2010).
Indeed overfishing per se is only one aspect of the problem.
Other economic activities, such as oil drilling, coastal devel-
opment, pollution, or dams and water flow management have
significant negative impact on aquatic habitats (Halpern et al.
2008).
Indirect contribution to food security and nutrition
through livelihood support
A critical pathway to enhance FSN is through the income that
people generate from engaging in remunerated activities. In
this respect aquaculture and fisheries (especially through the
number of small-scale operators engaged in fishing, aquacul-
ture, processing, and trading businesses see e.g., Bn et al.
2010) play a critical role in low income and emergent coun-
tries. Altogether it is estimated that between 660 and 820
million people (fishers, fish-farmers, fish traders, workers in
fish processing factories, and their families) depend on fish-
related activities as a source of income (Allison et al. 2013;
HLPE 2014). This represents more than 10 % of the world
population. For most of these households, the revenues gen-
erated may not necessarily be very high (Neiland and Bn
2004; Allison et al. 2011; Bn and Friend 2011), but it is
often the main component of their livelihood, which allows
them to secure accessibility to food (Heck et al. 2007; Bn
et al. 2009; Eide et al. 2011).
Feeding 9 billion by 2050: where does fish stand?
The global human population is expected to exceed 9 billion
by 2050 (UN estimates), increasing the pressure on the food
sectors to maximize production and reduce waste. Production
increase must occur in a sustainable way and in a context
where key resources, such as land and water, are likely to be
scarcer and where climatic change impact will intensify. The
fish-production sector is no exception.
In this context two key questions emerge. First, will fish-
eries and aquaculture be able to maintain the current global
fish consumption rate of 18 kg per capita per year, and the
equivalent regional values, and if not, how will society

address the needs of expected winners and losers (Barange
et al. 2014)? So far technological and institutional innovations
have ensured that the combined production of fish through
fisheries and aquaculture has been faster than the world pop-
ulations demand of fish. The question is now whether we can
keep up this pace with another 2 billion people added in the
next 35 years, and how the four dimensions of food security
(availability, accessibility/affordability, utilization and stabili-
ty) will be balanced to ensure that fish go to those who need
them most (Merino et al. 2012).
The second key question is whether sustainable fisheries
and aquaculture will be able to help address the bigger food
security issue that will affect the world in the coming decades?
In particular, could aquaculture become a substitute form of
protein for some of the less efficient food production systems,
or even be used to compensate for the decline in farming
systems productivity that is predicted as a consequence of
the impact of climate change?
How much fish do we need?
Global drivers of fish demand
World population is often presented as a key driver for the
growth in seafood demand and for fisheries development. In
reality, a more important driver for fish (and other animal-
source food) consumption is income (Speedy 2003).
Demand for fish as food is particularly high in the wealthier
strata of societies, including in the low-income countries, and
as income will continue to increase in highly populated coun-
tries such as China and India, demand levels are likely to rise
more strongly (Garcia and Rosenberg 2010). Overall, a large
increase in the number of people moving into the middle class,
particularly but not exclusively in Asia, is likely to result in a
very large expansion in the demand for fish.
Income is, however, not the only driver of fish demand. It is
recognized that urbanization is also an important factor in-
creasing animal-source food consumption in general and fish
consumption in particular. Delgado et al. (1997) suggest that
changes in food preference driven by urbanization alone have,
in the past, accounted for an extra 5.79.3 kg per capita con-
sumption of meat and fish per year. Similarly, Betru and
Kawashima (2009) and Toufique and Belton (2014) present
data from Ethiopia and Bangladesh, indicating that urbaniza-
tion affects animal food consumption rates independently of
income.
These different factors explain the rapid increase in
demand for meat, milk, and fish in the emerging econ-
omies of Asia. In China, for example, the demand for
fish is likely to increase from 24.4 kg per person per
year in 2000 to 41 kg per person per year by 2030
(World Bank 2014).
C. Bn et al.
Demand projection
Modelling exercises have been conducted recently with the
objective of estimating the projections of fish demand and
supply. These modelling exercises include the World Bank-
FAO-IFPRI Fish 2030 analysis elaborated on the IMPACT
model developed by the International Food Policy Research
Institute (IFPRI) (World Bank 2014); the OECD-FAO
Agricultural outlook model built on the combined
multimarket, partial equilibrium AgLink-CoSiMo model
(OECD-FAO 2013; Lem et al. 2014); and a series of peer-
reviewed articles relying on various types of modelling and
projection tools (Rice and Garcia 2011; Merino et al. 2010,
2012; Barange et al. 2014). The time horizon of these different
analyses is not always the same. For instance, the OECD-FAO
outlook model works over a 10 year-projection period (i.e., up
to 2023 as per the last iteration of the model), while the World
Bank-FAO-IFPRI model runs until 2030; Merino et al. 2010
used a 20 year simulation, calibrated on a 19972004 data set
(meaning technically that their projection runs until 2024),
while Merino et al. (2012) and Barange et al. (2014) proposed
a projection up to 2050. A strict comparison of the different
projections is therefore difficult.
A bigger issue is that very few of these studies integrated
information on the drivers of changes (the combination of
urbanization and increase in income), to estimate with accu-
racy the future demand for fish across the world. Many studies
assume constant consumption rates in the future (e.g., Barange
et al. 2014) or fixed nutritional targets (Rice and Garcia 2011).
Others worked directly with projected fish consumption (e.g.,
OECD-FAO 2013; Merino et al. 2012), that is, by dividing the
projected supply by the projected population. None of these
approaches therefore offers an appropriate basis for estimating
the actual demand for fish. The World Bank-FAO-IFPRI study
uses regional fish consumption rates to estimate the global
demand but the report does not clarify how these figures are
constructed.2 Aggregating the regional figures at the global
level, the report estimates that the world demand for fish will
be around 152 Mt in 2030 (World Bank 2014).
Merino et al. (2012) estimated the expected production
capacity of marine ecosystems exploited under maximum sus-
tainable yield principles, and projected aquaculture production
requirements to achieve a range of food consumption targets.
They concluded that between 125 and 210 Mt of fish by 2050
will be necessary to maintain fish consumption at around 15
20 kg per capita per year. Starting from a different angle, Rice
2
The report explains that BFor the subsequent years in the simulation
[after 2000 which was used as the base year for the calibration], these
intercept values [between supply and demand] are changed according to
the exogenous growth rates specified for each of the supply and demand
functions^ (World Bank 2014: 23, our emphasis). It is not clear from the
report how the exogenous growth rates in regional demand have been
computed in the model.
Feeding 9 billion by 2050 Putting fish back on the menu
and Garcia (2011) sought to estimate the need for additional
fish necessary to supply 20 % of the dietary protein require-
ment to feed a 9 billion population by 2050. On this basis, an
additional production of 75 Mt of fish from fisheries and
aquaculture would be needed above the 2006 production level
(144 Mt), that is, approximately 215 Mt. This represents an
almost 50 % increase in production with respect to the 2006
level. While this figure is above other projections, it has the
advantage that it starts from an actual estimate of the future
needs (as opposed to the estimate of apparent consumption
rates), but without factoring in future production potentials
or market responses to the fishmeal/fish oil demand (see
below).
In summary, the current understanding of the global drivers
of fish demand (urbanization and increase in living standards
in developing and emerged countries) is relatively well
established but not all of the current models have integrated
these drivers comprehensively. In comparison, the efforts to
better understand the ability of the world to produce fish in the
future (i.e., the supply side of the equation) have been more
elaborate.
The future of fisheries production and the impact of climate
change
A consensus has emerged in the literature that the doom-and-
gloom rhetoric that had driven the discussion surrounding the
state of marine fisheries in the late 2000s (Garcia and Grainger
2005; Caddy and Seijo 2005) was exaggerated (Grafton et al.
2010; Hilborn 2010) and that although the situation remains
concerning in respect to many stocks, we are not likely to face
the global collapse that had been announced by some biolo-
gists (e.g., Myers and Worm 2003; Worm et al. 2006; Pauly
2009). Instead, the downward trend of overfished stocks may
have been reined in (Fig. 4).3 Reflecting this, most of the
projections proposed in the recent literature estimate that the
global fisheries landings are likely to be stable in the short to
medium term. The OECD-FAO model for instance estimates
that capture fisheries will be 5 % higher by 2024 than it is was
in 2013, that is, around 96 Mt (OECD-FAO 2013) while the
World Bank-FAO-IFPRI model estimates that this will be
around 93 Mt in 2030. These figures are at a global scale
however, and some regional outlooks are for good stock re-
building, while others are for a worsening in overfishing.
Yet another key factor for which much uncertainty remains
is the impact of climate change. Unlike most terrestrial ani-
mals, aquatic animal species are poikilothermic (cold-
blooded) and changes in aquatic habitat temperatures will
more rapidly and significantly influence distribution, prey
availability, metabolism, growth and reproduction, with
3
According to the FAO, the number of stocks fished at unsustainable
levels decreased from 32.5 % in 2008 to 28.8 % in 2011 (FAO 2014).
stronger impact on fishing and aquaculture distribution and
productivity (Cheung et al. 2009). At the same time however,
the interconnectedness of aquatic systems allows many spe-
cies to change spatial distribution more easily as ecosystems
shift, to remain in their zones of preference. Clearly, therefore,
the impact of global climate change on ocean capture fisheries
will be important. Biological predictions based on ocean-
atmosphere general circulation models (OA-GCMs) have al-
ready demonstrated that the physical and chemical properties
of the oceans will be modified, affecting the productivity,
distribution, seasonality and efficiency of food webs, from
primary producers to fish (Steinacher et al. 2010; Cheung
et al. 2009, 2011). Some of these earlier GCM models how-
ever, were limited by their coarse resolution, too low to cap-
ture the processes that dominate the dynamics of the worlds
coastal and shelf regions, where most fisheries operate. More
advanced models are now available (e.g., Merino et al. 2012;
Blanchard et al. 2012; Barange et al. 2014).
Overall, and with few exceptions, the conclusion of all
these models is that although climate change will alter the
present geographical distribution of shelf-sea ecosystems pro-
ductivity, in most of the regions and EEZs, the overall poten-
tial impact is projected to be low to moderate. Barange et al.
(2014), for instance, used a high resolution shelf-sea physical-
biological model that allowed them to scale down the analysis
and gave greater confidence in predicting the consequences at
national scales. They conclude that by 2050, estimates of na-
tional fish production should remain on average within 10 %
of the present yields.
The contribution of aquaculture to future fish supply
The second element on the supply side is aquaculture.
Discussion of the rise of aquaculture has so far largely focused
on its contribution to global aquatic animal food supplies,
ignoring the resultant changes in species composition of the
fishes consumed, how it is farmed, and the implications for
food and nutrition security (Kawarazuka and Bn 2011;
UNHRC 2012). As a consequence, our understanding mainly
concerns the question of the ability of the aquaculture industry
to maintain its rate of growth. In this regard, most of the recent
analyses agree that the era of exponential growth is over and
while the sector will still continue to grow, the projected rate
of growth is expected to decelerate. The main causes of this
slower growth are likely to be freshwater scarcity, lower avail-
ability of locations for optimal production, and high costs of
fishmeal, fish oil and other feeds (FAO 2012). Nonetheless,
the World Bank-FAO-IFPRI model suggests this rate will still
be sufficient to maintain a steady rise, reaching the point
where it will equal global fisheries production by 2030
around 93 Mt (World Bank 2014) (Fig. 5). Technical innova-
tion, improved farm and animal health management, and im-
proved and more efficient germplasm will be responsible for
 C. Bn et al.

Fig. 4 Global trends in the state

of world marine fish stocks
(19742011). Source: FAO
(2012)

Source: FAO (2012).

this increased growth, in combination with the continued ex- Putting all the pieces together
pansion of fish-farming. Combined with a projected capture
fish production that will remain fairly stable over the 2000 Given the very rough projections for fish demand and the
2030 period (see above), the global fish supply is projected to more elaborate projections for fish supply just reviewed
rise to 187 Mt by 2030 (World Bank 2014). These figures are above, will fisheries and aquaculture be able to maintain their
consistent with the projections proposed by OECD-FAO in current contribution to food security in the future? The answer
which the global fish production will reach 181 Mt in 2022, that emerges from the literature is that as far as food availabil-
of which 161 Mt would be destined for direct human con- ity and demands are concerned, it is a conditional yes. All the
sumption (OECD-FAO 2013). projection models currently available seem to agree that the
A key element in this discussion is the importance of overall fish consumption rate could be maintained, in other
fishmeal and fish oil and how markets and technological in- words, that the fisheries and aquaculture aggregated growth
novations will respond to price signals (Merino et al. 2010, will keep up with population growth rates. In fact, the OECD-
2012; FAO 2012; World Bank 2014; Troell et al. 2014).4 The FAO even estimates that the world fish consumption will in-
consensus is that the use of fishmeal in aquaculture feeds is crease by another 10 % and reach 20.9 kg per capita per year
expected to decrease in percentage with time, thanks to in- by 2023. The World Bank-FAO-IFPRI report is slightly more
creasingly effective replacements, including plant proteins,
waste products from fish and terrestrial animals and use of
better/improved breeds of aquatic animals with better feed
conversion (Tacon et al. 2011). Formulated feeds are a signif-
icant factor in production costs, and this is a strong incentive
to develop technology that will make feeds more affordable
and sustainable. Overall, the proportion of fish used for non-
direct human consumption has decreased from 30 % in the
early 1990s to 15 % in 2010 and the World Bank analysis
concluded that the projected expansion of aquaculture will
be achieved with only an 8 % increase in the global fishmeal
supply during the 20102030 period (World Bank 2014).

4
For instance, one of the 6 selected scenarios of the World Bank-FAO-
IFPRIs analysis is specifically focusing on these issues (Scenario 2
BExpanded Use of Fish Processing Waste in Fishmeal and Fish Oil
Data source: FAO FishStat and IMPACT projection model.
Production^) (World Bank 2014) and one of the selected issues of the
2012 FAO SOFiA report was on BDemand and supply of aquafeed and Fig. 5 Global fish production: data and projection 19842030 from the
feed ingredients for farmed fish and crustaceans: trends and future IMPACT model (World Bank 2014). Data source: FAO FishStat and
prospects^ (FAO 2012: 171182). IMPACT projection model
Feeding 9 billion by 2050 Putting fish back on the menu

conservative and estimates that the per capita fish growth will increase fish consumption while low-income peo-
consumption will remain at around 18 kg per year in 2030. ple will experience reductions in their fish consumption. This
Merino et al. (2012) and Barange et al. (2014) reached the is an important distributional and food security consequence.
same conclusion for 2050, but this is essentially due to the To avoid that, fish production needs to expand faster than
underlying assumptions of their models. population growth.
All these analyses, however, stress the fact that for this
outcome to occur, several specific conditions must be satis-
fied: capture fisheries will need to be exploited according to
sustainable principles; very significant technological develop- Can fish contribute further to the 2050 global food
ment will need to take place in aquaculture feeds to reduce security challenge?
fishmeal dependency and in farm management and germ-
plasm to improve the overall efficiency of aquaculture; and The arguments presented above are mainly constructed
discards, waste and losses will need to be reduced. Some of around a sectorial perspective. At least three other trans-
these conditions could be particularly challenging and a busi- sectorial (or systemic) arguments should be considered in re-
ness-as-usual approach is expected to fail.5 lation to the wider debate of feeding 9 billion by 2050.
In addition to whether or not the world (and the markets) First, in terms of animal protein availability, with 18.2 kg
will be responsible enough to ensure that these conditions are per capita per year, fish is providing 115, 133, and 189 % more
satisfied, the question of who will be the winners and losers is protein per capita than pig, poultry and beef respectively. In
vital to the other key FSN issues, which are: access to liveli- fact, fish (combining capture fisheries and aquaculture) has
hoods in fish value chains and affordability of fish. been the main contributor to the 61 % increase in the world
Unfortunately, the consensus is that in the coming decades, per capita consumption of animal protein for the period 1969
the current situation regarding the imbalance between con- 2009 (Table 1). As economic development is expected to con-
sumers in developed and developing countries is unlikely to tinue driving an increasing trend in animal protein demand,
change. While the present figures indicate the lowest fish (OECD-FAO 2013) and aquaculture is projected to remain the
consumption per capita in Africa (9.1 kg in 2009), the fastest growing food commodity sector, this sector will soon
different models project that this imbalance will deteriorate become even more central in the future food security of the
further. The OECD expects that per capita consumption in world population.
Africa will decrease by a further 10 % by 2024, while that
of Asia will show the highest growth rate (+14 %, OECD- Table 1 World per capita meat and fish food supply (kg per capita per
FAO 2013). The World Bank-FAO-IFPRI projection year) 19692009
(World Bank 2014) is even more alarming with per capita
1969 1979 1989 1999 2009
fish consumption expected to decline in sub-Saharan
Africa by 1 % annually to 5.6 kg in 2030.6 Yet at the same Pig meat 9.4 11.5 13.1 15.1 15.8
time developing countries will account for more than 91 % 25.2 % 27.7 % 28.2 % 28.1 % 26.3 %
of the total increase in fish consumption. Even so, their Poultry meat 3.8 5.6 7.3 10.7 13.6
annual per capita fish consumption will still remain below 10.2 % 13.5 % 15.7 % 19.9 % 22.7 %
that of more developed regions (19.8 vs. 24.2 kg) (World Bovine meat 10.8 10.7 10.3 9.7 9.6
Bank 2014). 29.0 % 25.8 % 22.2 % 18.1 % 16.0 %
A last important point needs to be mentioned. In view of Mutton and goat meat 1.8 1.5 1.7 1.8 1.9
the importance of income growth and urbanization as drivers 4.8 % 3.6 % 3.7 % 3.4 % 3.2 %
of fish demand, the objective of merely maintaining fish pro- Meat, other 0.8 0.8 0.7 0.8 0.9
duction growth rate at the same level as the growth in world 2.1 % 1.9 % 1.5 % 1.5 % 1.5 %
population would not be enough to prevent fish price from
Capture 10 10.4 11 10.5 10
increasing. The urban population with the highest income
26.8 % 25.1 % 23.7 % 19.6 % 16.7 %
5 Aquaculture 0.7 1 2.4 5.1 8.2
Merino et al. (2012) estimate that the fish used in aquaculture feed to
produce one unit of output would have to be reduced by at least 50 % 1.9 % 2.4 % 5.2 % 9.5 % 13.7 %
from current levels to secure sustainability. If not, demand will push the Capture and Fisheries 10.7 11.4 13.4 15.6 18.2
price of fishmeal products up to levels where the short term economic 28.7 % 27.5 % 28.8 % 29.1 % 30.3 %
incentive to exploit small pelagic (the main source of this fishmeal) be-
yond their maximum sustainable yield would be high, potentially leading Total 37.3 41.5 46.5 53.7 60
to increases in fishing capacity and rapid depletion of resources.
6
Some other regions of the world are also expected to face lower fish Source: FAOStat
consumption rate per capita: Japan, Latin America, Europe, and Central Figures in percent are the respective contribution of each sector to the
Asia. total figure

Second, in term of efficiency, fish in aquaculture systems
are very efficient converters of feed into protein in fact far
more efficient than most terrestrial livestock system (Fig. 6a).
For instance, poultry converts about 18 % of their consumed
food and pigs about 13 %, as compared with 30 % in the case
of fish (Hasan and Halwart 2009). Production of 1 kg of beef
protein requires 61.1 kg of grain, production of 1 kg of pork
protein requires 38 kg of grain, while fish only requires
13.5 kg (Hall et al. 2011). Most of this difference comes from
two biological characteristics of fish which give them great
advantages over land-based livestock in growth performance:
(i) the fact that fish are poikilotherms and therefore do not
expend energy maintaining a constant body temperature; and
(ii) the fact that, because finfish are physically supported by
the aquatic medium, fewer resources are used on bony skeletal
tissues, and a larger part of the food they eat is effectively
allocated to body growth.
Third, in terms of carbon footprint, aquatic animal produc-
tion systems have a lower carbon footprint per kilogram of
output compared with other terrestrial animal production sys-
tems (Hall et al. 2011). As a consequence, nitrogen and phos-
phorous emissions (kg of nitrogen and phosphorus produced
per tonne of protein produced) from aquaculture systems are
much lower than those in beef and pork production systems
and slightly higher than that of poultry (Fig. 6b). In fact, some
aquaculture production systems such as farming of bivalves
absorb nitrogen and phosphorous emissions from other systems.
All these reasons are important arguments for giving fish
far greater attention in the food security debate and in the
current discussion about how to feed 9 billion by 2050. The
CFS has started to recognize this and the report that was
commissioned in 2012 indeed recommends that fish (a) be-
comes an integral element in inter-sectoral national FSN
70
kg / tonne of protein produced
60
50
40
30
20
10
0
beef pork fish
Source: Data for fish are derived from Hall et al. (2011). Data for beef, pork and chicken are derived from Flachowsky (2002) in Potrk
(2003).
Fig. 6 a Feed and protein conversion efficiency of beef, pork and fish; b Nitrogen and phosphorous emissions for animal production systems. Source:
Data for fish are derived from Hall et al. (2011). Data for beef, pork and chicken are derived from Flachowsky (2002) in Potrk (2003)
C. Bn et al.
policies and programmes and (b) should be more systemati-
cally included in countries nutritional programmes and inter-
ventions aiming at tackling micronutrient deficiencies espe-
cially among children and women (HLPE 2014: 18
Recommendation 1a and 1b).
Conclusion
Fish is already making a major contribution to human food
supply and to the support of FSN for more than 660 million
fish-workers and their families. It also provides more than 4.5
billion consumers with at least 15 % of their average per capita
intake of animal protein. In addition, because fish is more nu-
tritious than staple foods such as cereals, providing in particular
essential fatty acids and micronutrients, it can play an extreme-
ly important role in improving the nutritional status of individ-
uals, in particular those at risk such as children and women.
Yet limited attention has been given so far to fish as a key
element in FSN strategies at national level and in wider de-
velopment discussions and interventions. Part of the problem
might have been that specialists in fisheries debates have con-
centrated predominantly on questions of biological sustain-
ability and on the economic efficiency of fisheries, neglecting
issues linked to its contribution to reducing hunger and mal-
nutrition and to supporting livelihoods (Kawarazuka and Bn
2010, 2011). On the other end of the spectrum, and with too
few exceptions, most (non-fishery) food security experts and
decision-makers seem unfamiliar with these facts and, there-
fore, unaware of the critical role that aquaculture is likely to
play in the future. The problem is particularly pronounced in
the current debate about how to make the food system more
nutrition sensitive, i.e., how to change and improve the food
1400
1200
1000
800
600 Nitrogen
Phosphorus
400
200
0
beef pork chicken fish
-200 bivalves
(average)
Feeding 9 billion by 2050 Putting fish back on the menu
systems in order to advance nutrition. As a consequence, fish
has so far been only marginally included in the international
FSN debate. Many nutritional programmes are still not aware
of, or not recognizing and building on the potential of fish for
the reduction of micronutrient deficiency.
In this paper, we make the case that fish deserves more
attention in food policies due to its importance in the food
basket, its unique nutritional properties, its higher efficiency
of production and carbon footprint compared to other forms of
animal production systems. We acknowledge some chal-
lenges, however, especially in making fish more affordable
for the poor or in maintaining or restoring the environmen-
tal sustainability of the sectors. We also stress that the ability to
meet the potential average fish demands of 9 billion people
masks inequalities and inequities in who eats the fish and who
benefits from the value chains. At present, people in Africa,
the poor in many societies, and women and minority groups,
including small-scale fishing and aquaculture communities
are in tension with large corporations and production units
over access to fish and fish-related employment (HLPE
2014). We have shown that the best available projections for
fish supply and demand are relatively positive in terms of the
capacity to meet future demands, although more solid model-
ling is still needed to better incorporate demand projections.
Climate change impacts on fish production also create uncer-
tainty in the projections.
In conclusion, fish should certainly be on the menu.
Acknowledgments This paper was part of a workshop sponsored by
the OECD Co-operative Research Programme on Biological Resource
Management for Sustainable Agricultural Systems.
Open Access This article is distributed under the terms of the Creative
Commons Attribution License which permits any use, distribution, and
reproduction in any medium, provided the original author(s) and the
source are credited.
References
Allison, E. H. (2011). Aquaculture, fisheries, poverty and food security.
Working Paper 201165. Penang: WorldFish Center.
Allison, E. H., Bn, C., & Andrew, N. L. (2011). Poverty reduction as a
means to enhance resilience in small-scale fisheries. In R. S.
Pomeroy & N. L. Andrew (Eds.), Small-scale fisheries management
frameworks and approaches for the developing world (pp. 216
238). Wallingford: CABI.
Allison, E. H., Delaporte, A., & Hellebrandt de Silva, D. (2013).
Integrating fisheries management and aquaculture development
with food security and livelihoods for the poor. Report submitted
to the Rockefeller Foundation. Norwich: School of International
Development, University of East Anglia.
Arthur, R., Bn, C., Leschen, W., & Little, D. (2013). Fisheries and
aquaculture and their potential roles in development: an assessment
of the current evidence. London, UK: Marine Resources Assessment
Group Limited (MRAG). (http://r4d.dfid.gov.uk/pdf/outputs/
fisheries/61091-Fisheries_and_Aqua_Evidence_Review.pdf).
Barange, M., Merino, G., Blanchard, J. L., Scholtens, J., Harle, J.,
Allison, E. H., et al. (2014). Impacts of climate change on marine
ecosystem production in fisheries-dependent societies. Nature
Climate Change, 4, 211216.
Bn, C. (2006). Small-scale fisheries: assessing their contribution to
rural livelihoods in developing countries. FAO Fisheries Circular,
No. 1008. Rome: Food and Agriculture Organization (FAO).
Bn, C., & Friend, R. (2011). Poverty in small-scale inland fisheries: old
issues, new analysis. Progress in Development Studies, 11(2), 119
144.
Bn, C., Macfadyen, G., & Allison, E. H. (2007). Increasing the contri-
bution of small-scale fisheries to poverty alleviation and food secu-
rity. FAO Fisheries Technical Paper, No. 481. Rome: FAO.
Bn, C., Steel, E., Kambala Luadia, B., & Gordon, A. (2009). Fish as the
Bbank in the water^ - evidence from chronic-poor communities in
Congo. Food Policy, 34, 104118.
Bn, C., Hersoug, B., & Allison, E. H. (2010). BNot by rent alone^:
analysing the pro-poor functions of small-scale fisheries in develop-
ing countries. Development Policy Review, 28(3), 325358.
Betru, S., & Kawashima, H. (2009). Patterns and determinants of meat
consumption in urban and rural Ethiopia. Livestock Research for
Rural Development, 21(9/143).
Beveridge, M. C. M., Thilsted, S. H., Phillips, M. J., Metian, M., Troell,
M., & Hall, S. J. (2013). Meeting the food and nutrition needs of the
poor: the role of fish and the opportunities and challenges emerging
from the rise of aquaculture. Journal of Fish Biology, 83, 10671084.
Blanchard, J., Jennings, S., Holmes, R., Harle, J., Merino, G., Allen, I.,
et al. (2012). Potential consequences of climate change on primary
production and fish production in 28 large marine ecosystems.
Philosophical Transactions of the Royal Society, B: Biological
Sciences, 367(1605), 29792989.
Bonham, M. P., Duffy, E. M., Robson, P. J., Wallace, J. M., Myers, G. J.,
Davidson, P. W., et al. (2009). Contribution of fish to intakes of
micronutrients important for foetal development: a dietary survey
of pregnant women in the Republic of Seychelles. Public Health
Nutrition, 12(09), 13121320.
Boyd, C. E., Tucker, C., McNevin, A., Bostock, K., & Clay, J. (2007).
Indicators of resource use efficiency and environmental perfor-
mance in fish and crustacean aquaculture. Reviews in Fisheries
Science, 15, 327360.
Caddy, J. F., & Seijo, J. C. (2005). This is more difficult than we thought!
The responsibility of scientists, managers and stakeholders to miti-
gate the unsustainability of marine fisheries. Philosophical
Transactions of the Royal Society London B: Biological Sciences,
360(1453), 5975.
Cheung, W. W. L., Lam, V., Sarmiento, J., Kearney, K., Watson, R.,
Zeller, D., & Pauly, D. (2009). Large-scale redistribution of maxi-
mum fisheries catch potential in the global ocean under climate
change. Global Change Biology, 16, 2435.
Cheung, W. W. L., Dunne, J., & Sarmiento, J. L. P. D. (2011). Integrating
ecophysiology and plankton dynamics into projected maximum
fisheries catch potential under climate change in the Northeast
Atlantic. ICES Journal of Marine Science, 68, 10081018.
de Schutter, O. (2012). The right to food - Note to the General-Secretary
from the Special Rapporteur on the right to food. New York: United
Nation, Sixty-seventh session General Assembly.
Delgado, C. L., Crosson, P., & Courbois, C. (1997). The impact of live-
stock and fisheries on food availability and demand in 2020. MSSD
Discussion Paper, No.19. Washington, DC: IFPRI.
Eide, A., Bavinck, M., & Raakjr, J. (2011). Avoiding poverty:
Distributing wealth in fisheries. In J. Svein & A. Eide (Eds.),
Poverty mosaics: Realities and prospects in small-scale fisheries
(pp. 1325). Dordrecht: Springer.
FAO. (2012). The state of world fisheries and aquaculture 2012. Rome:
FAO.
Editorial
Nutrigenomics and the future of dietetics
The role of the genome and the interaction of the environ-
ment on the genome and its products are of increasing
importance to health care. Genetic principles are now rec-
ognised as being fundamental to health and disease, a real-
isation that is leading to a change in focus from disease
management to disease prevention and health promotion.1
Not surprisingly, it is the integration of genetic principles
and technologies into the food and nutrition sciences, the
emerging field called nutrigenomics, that will be critical to
realising the full potential of applying genomics to health
care. Nutrigenomics is concerned with how dietary com-
ponents interact with genetic sequences and their products
to impact the functioning of the organism, particularly with
respect to health and disease.25
Dietetics professionals will play a central role in dis-
ease management and prevention in the years ahead in
ways never before possible.6,7 Clinicians will possess
knowledge and skills unique within health care. Those in
the community health and public policy arenas will under-
stand the molecular basis for ethnic differences in nutrient
requirements and will be in a position to develop effective
dietary approaches tailored to each ethnic group. Dietetics
professionals in the food sciences will use nutrigenomics
to identify and isolate the bioactive components in food
and will develop health-promoting functional foods. The
need for education of consumers and health care profes-
sionals will be extensive and ongoing. Clearly, nutrige-
nomics will have a significant impact on the field of
dietetics.
The concept of nutrigenomics is not new. What is new
is the appreciation of its integral role in health and disease.
The biological world has long benefited from the practical
applications of this concept. From primitive unicellular
microorganisms to the most highly evolved mammals,
genetic mechanisms exist that allow information from the
environment surrounding a cell to be transmitted to the
genetic material deep within that cell. In elaborate signal-
ling cascades that impact gene expression, bioactive com-
ponents from food serve to inform the organism about the
nutritional sufficiency of its environment. These same bio-
active dietary components supply an organism with essen-
tial nutrients that, because of genetic mutations, the
organism cannot synthesise and depends upon the environ-
ment to supply. The evolutionary process itself has been
significantly impacted by the goodness-of-fit between the
genetic make-up of an organism and the food present in its
environment.
Contribution of the Human Genome Project
The Human Genome Project has been the impetus for the
development of nutrigenomics as it is emerging today.8
This ambitious project, which began in 1990, was initially
conceived as a 15-year multinational effort to identify the
sequence of three billion nucleotides that comprise human
genetic material. The project has been highly successful,
has grown significantly in scope and is ongoing. Today
the human genome sequence is virtually complete, the
genomes of numerous other organisms have been
sequenced and the focus has shifted to determining the
products of these sequences and their roles in the function
of organisms.
One of the earliest practical applications derived from
the Human Genome Project has been in the field of phar-
macogenomics, the study of genetic influences on drug
metabolism.9 For any given pharmaceutical, there are indi-
viduals within a population who benefit from the drugs
action, those who derive little benefit and those who expe-
rience harmful side-effects. The underlying basis for this
heterogeneous response within a population is the genetic
heterogeneity with respect to key drug-metabolising
enzymes. Specific genetic variants of these enzymes have
been associated with particular drug responses. Diagnostic
tests are now available that allow clinicians to detect which
genetic variants an individual carries and predict that indi-
viduals response to a particular drug. Pharmacogenomics
points up the advisability of matching pharmaceutical
therapy with the drug-metabolising capabilities of the indi-
vidual, abilities that derive from the individuals genetic
make-up.
In many respects, nutrigenomics parallels the develop-
ment of pharmacogenomics. As with pharmaceuticals, a
number of key enzymes and metabolic reactions are
involved with the digestion, absorption and metabolism of
food. Genetic variations in these key processes lead to a
populations heterogeneous response to diets in general and
specific food components in particular. The ability to
match foods to the genetically determined capability of
using that food for positive benefit is crucial to an
individuals optimal health.10,11
However, nutrigenomics differs from pharmacogenom-
ics in several important aspects. Many decades of pharma-
cological research have led to a strong foundation of
scientific knowledge, which has given rise to the practical
applications of pharmacogenomics. In contrast, nutrition
research is in its infancy. Unlike drugs, food is a complex
mixture of numerous components. Furthermore, an indi-
viduals genetic material has been interacting with food
components for its entire lifetime, which may have lasting
physiological consequences as compared with the limited
druggenome interactions. These important differences
will likely slow the entry of nutrigenomics into wide-
spread application compared with the rapid entry of
pharmacogenomics.
Impact of nutrigenomics on dietetics
Nutrigenomics will ultimately impact all of dietetics.
Research in the food and nutrition sciences is central to
understanding underlying mechanisms that form the basis
for practical applications. From food science research will
come insight into which bioactive food components effect
genetic changes and how. These components will be iso-
lated and added to commonly eaten foods or made avail-
able as dietary supplements. Food scientists will be
Nutrition & Dietetics (2005) 62: 2/3 63
instrumental in developing new functional foods that con-
tain bioactive dietary components known to promote health
and reduce disease burden. From nutrition research will
come strong associations between genes and diseases,
which in turn will lead to diagnostic tests for the detection
of particular genes. Nutrition research will further identify
the influence of the environment on the outcome of those
associations and be the source of targeted, effective thera-
peutic interventions. There will likely evolve a universal
set of dietary recommendations that are based on genetic
considerations, with dietary guidelines for specific popu-
lations reflecting the predominant genetic composition of
that population.
Research into consumer acceptance and concerns is
another important component of this emerging field. It is
not enough to develop foods that promote health. In order
for nutrigenomics to be maximally useful for health pro-
motion purposes, foods must be tasty, eye-appealing and
convenient. Further, consumers must come to view their
genetic make-up as simply another piece of important
health information, similar to height, weight and blood
pressure. Additionally, the ethical, legal and social impli-
cations of nutrigenomics are critical issues to address.
The application of nutrigenomics will not wait until the
research base has been built. Potential applications of food
and nutrition research vary widely, from maximising the
nutritional composition of foods, to guiding the dietary
recommendations of populations and individuals, to
designing nutritional interventions targeted to particular
genetic constitutions. In the clinical realm, dietetics prac-
titioners will have the opportunity to expand from nutrition
counselling to genetic counselling to a combination of the
two as diagnostic tests become available that pinpoint the
genetic basis for a disorder and help determine the appro-
priate lifestyle interventions. Education and communica-
tion skills are integral to all of these anticipated
applications.
Preparing for the nutrigenomics era
The scope of dietetics is expanding as its central role in
health promotion and disease prevention becomes estab-
lished.6,7 Preparing for this expanding role is challenging,
yet essential. The initial impact of nutrigenomics on die-
tetics will be felt increasingly over the next decade and then
will escalate as the research foundation clarifies the inter-
actions among genes, environment and bioactive food
components and provides the basis for using these food
components to effect desirable outcomes. A thorough
understanding of food and nutrition at the mechanistic
level will be required. Dietetics professionals must develop
a solid base of knowledge from the core sciences: biology,
chemistry, biochemistry, physiology, food science and
nutrition, with genetic principles and technologies inte-
grated throughout. Critical thinking skills are essential in
an age of science-based practice. A lifelong approach to
learning is necessary in order to stay abreast of dietetics as
the pace of discovery escalates. Communication skills of
all types will continue to be important regardless of prac-
tice focus. An understanding of the psychosocial dimen-
sion of human behaviour will be needed in many areas of
dietetics. The dietetics professional of tomorrow is more
likely to be an entrepreneur than has been true in the past
and will benefit from attaining business development and
management skills.
64 Nutrition & Dietetics (2005) 62: 2/3
It is unlikely that all this knowledge and expertise can
be acquired in the undergraduate years alone. Nutrigenom-
ics provides an opportunity to reconsider the education and
training of dietetics professionals. To develop the level of
competency required, graduate level training is highly
desirable if not essential. Furthermore, the undergraduate
program will likely offer tracks that steer students into one
of several possible practice directions and will focus on the
academic preparation needed for successful practice.
The masters level graduate program would then focus on
the practical training needed for specialty certification
along with any additional educational needs. Graduates
will then function as entry-level practitioners, laboratory
associates, educators and entrepreneurs. Doctoral level
programs would include extensive practical experience and
prepare students to conduct independent research, develop
nutrition policy, or practise as relatively autonomous health
care professionals specialising in nutrigenomics.
The dietetics professional as thought leader
A number of issues will emerge that will need to be
addressed by thought leaders within the dietetics field.
Among these issues are certainly the ethical, social and
legal issues alluded to earlier. Thought leaders will also
grapple with the best diet for the human genome. Given
that human genes have changed little over the past
10 000 years since the agricultural revolution,12 would that
diet more resemble humans ancient diet rather than the
highly processed diet of today? If the cultivation of grains
occurred late in human development, is it sound from a
genetic perspective for todays diet to be heavily grain-
based? To what extent can the diet be a universal one fine-
tuned for genetically distinct populations?
Issues relating to nutrient requirements, recommenda-
tions for nutrient intake, and how best to, and even whether
to, fortify food with nutrients and other bioactive food
components must be addressed. The amounts of each nutri-
ent will vary based on the genetic variation of an individ-
ual, with some requiring far greater levels of particular
nutrients than what is currently thought to be the normal
range. How will the era of genomics impact the develop-
ment of global dietary recommendations? The impact of
food fortification on the human species from an evolution-
ary perspective will need to be debated. Will food fortifi-
cation lead to selection for genotypes that would not
normally survive and are ill-suited for todays environ-
ment? These and many other issues with far-reaching con-
sequences will be debated.
Nutrigenomics will be a major force in dietetics.
Dietetics professionals have an unprecedented opportunity
to carve out unique, valuable niches within the research
and application sectors of dietetics and to serve as thought
leaders as nutrigenomics is integrated into health care, the
food industry and public policy.
Ruth M. DeBusk
Geneticist and Clinical Dietitian,
Tallahassee, Florida, USA
References
1. Guttmacher AE, Collins FS. Genomic medicinea primer. N Engl
J Med 2002; 347: 151220.
2. Kaput J, Rodriguez RL. Nutritional genomics: the next frontier in
the postgenomic era. Physiol Genomics 2004; 16: 16677.
24 June 2014, 3.49am AEST

Setting the standards: who regulates

Australian GM food?
Viewed at http://theconversation.com/setting-the-standards-who-regulates-australian-gm-
food-25533

By David Tribe

Senior Lecturer in Food Biotechnology and Microbiology, Agriculture and Food Systems at
University of Melbourne

More Events

Field testing genetically modified crops is a necessary step of the process, but who makes
sure its all safe? Andreas Beer/Flickr, CC BY-NC-SA

In this second instalment of GM in Australia a series looking at the facts, ethics,

regulations and research into genetically modified crops David Tribe walks us through the
bodies responsible for GM policy and oversight.
In the past, new crops were introduced into the food supply without any formal scientific
evaluation. Humans learned by trial and error how to safely prepare foods such as cassava
and potato, even though they can be toxic.

With the advent of crop genetic engineering in the 1980s, public controversy and intense
public scrutiny over genetically modified (GM) foods meant that the trial and error method of
discovering whether new GM foods were safe became unacceptable.

Scientific safety assessment of new GM foods as well as government regulation of their

introduction was introduced in many countries.

In Australia, this occurred during the major policy reform of all Australian food safety
regulation, through the revision of the national Food Standards Code (1994-2002).

During this period, safety assessment of GM foods became the responsibility of Food
Standards Australia New Zealand (FSANZ). Section 1.5.2 of the Food Standards Code
defines the requirements for the compulsory pre-market assessment and labelling of foods
produced by gene technology.

Comparative safety assessment for new whole foods

Food scientists carrying out GM food safety assessments realised that unintended variation in
potentially toxic non-nutrient plant chemicals was a possible source of hazard in new GM
crop varieties.

The presence of thousands of potentially toxic natural substances in plant foods meant that
the safety assessment of GM foods had to be done by different procedures than those that had
worked well with food additives and synthetic pesticides.

Regulators realised that if the hundreds of thousands of natural substances present in plant
foods such as soybeans or maize were subjected to the same standards used with food
additives and synthetic pesticides, all food GM or non-GM would fail regulatory
standards.

A solution to this quandary was developed. Relative safety of GM foods is assessed by

systematically looking for all the chemical differences that can be found between:

a GM food and
an otherwise comparable non-GM food that can be presumed to be already safe
because of its history of safe use.

If no meaningful differences are detectable in a new crop variety, the GM food can be
assumed to be at least as safe as its non-GM counterpart.

The standards agency FSANZ provides many further details of how safety assessment of GM
foods is carried out, before they are allowed to enter the commercial marketplace.
Box 4.1 from their introductory booklet on the topic gives the main features of the process.

Whos involved in setting standards?

Most GM crops are commodities that are extensively traded on world markets, such as maize,
soybeans and canola. These have to meet food and regulatory standards in several different
jurisdictions (such as in the US, the EU, Japan and China).

In practice, traded GM commodities have to pass regulatory scrutiny in counties to which

they are exported. The safety assessment frameworks in these different trading countries have
a substantial degree of consistency.

The UN forum Codex Alimentarius Commission is the major international intergovernmental

forum in which this safety framework was developed. (Learn more about Codex in the video
below.)

Codex has established international guidelines for the safety assessment of foods derived
from modern biotechnology. The Australian protocols for the pre-market safety assessment of
GM foods are congruous with Codex principles.

Codex emphasises the use of sound scientific evidence for food risk assessment, and the
separation of risk policy and risk management from scientific risk assessment. In Australia,
the management of GM food safety follows these principles.

Food regulatory policy and political oversight are decided by an Australian

intergovernmental board called the Legislative and Governance Forum on Food Regulation.

GM safety assessment and food standards development are carried out by FSANZ.

In establishing standards and evaluating pre-market applications for the commercialisation of

GM foods, FSANZ interacts with other agencies in particular the Gene Technology
Regulator, whose responsibilities include licensing the deliberate environmental release of
genetically modified organisms.

Genetically modified foods have now been in the marketplace for nearly two decades without
any harmful effects identified.
The absence of any evidence pointing to the lack of food safety in the recent farmer court
case in Western Australia (fought over the unintended presence of genetically modified
canola plants on another farmers property), underlines the overall conclusion that there is no
credible evidence of any food safety risk with the GM foods that have been approved so far.

More tangible problems of food safety such as the mould toxins that can ruin staple grains
and cause cancer in developing countries should now get more attention
GM oilseed crops and the Australian oilseed industry
Ruth Holtzapffel, Hilary Johonson and Osman Mewett

Executive Summary
Oilseed crops are The worlds major oilseed crops are soybean, rapeseed (including
important to Australian canola), peanut, oil palm and sunflower. Australia produces
agriculture. between 2 and 3 million tonnes of oilseeds each year, with canola
and cottonseed being the major crops. These two crops account for
92% of Australias total oilseed production with peanuts, soybeans
and sunflower seeds accounting for the remaining 8%.
The gross value of oilseed production averaged around 7% of the
total gross value of Australian broadacre crop production over the
three years to 2005/06. Over the same period, Australian exports of
oilseeds averaged around 8.5% of the total value of Australian
exports of grain and oilseeds.
Some oilseeds also provide benefits to wheat and barley crops
through their use in rotation cropping cycles.
Oilseeds are widely used Vegetable oils are widely consumed in our diets, including in
in food, feed and margarines and processed and fried foods. Meal derived from
industrial applications oilseeds is an important source of protein in the stockfeed industry
and vegetable oils are also used in the manufacture of soaps,
industrial lubricants and paints.
and have been Oilseeds have been selected over many generations to contain
extensively modified to desirable fatty acids to suit their different uses. For example, canola
meet these needs. was developed by selective breeding to contain low levels of anti-
nutritional compounds. Other modifications have included breeding
canola and sunflower varieties with high levels of monounsaturated
fats to improve the frying quality of their oils. Chemical
modifications can also be made to purified vegetable oils to make
margarines or to improve their stability for cooking and industrial
purposes.
Modifications are The 2005 BRS report Whats in the Pipeline? highlighted the
continuing to produce significant amount of research currently being conducted on
oilseeds that are easier genetically modified (GM) oilseed crops. This includes a number of
and cheaper to grow traits designed for easier, cheaper and more sustainable production
(e.g. herbicide tolerance, insect resistance, increased yields). GM
cotton varieties with herbicide tolerance and insect resistance have
been widely adopted in Australia and provide benefits to the
environment and farmers.

GM oilseed crops 5
 produce healthier A significant amount of research is also being directed into
oils developing healthier oilseeds, particularly those which produce high
levels of polyunsaturated or long chain omega-3 fatty acids; higher
levels of essential amino acids and/or vitamins; or decreased
allergenicity. These improvements and modifications are expected
to provide health benefits for people throughout the world and to
add value to oilseed crops. There are commercial drivers for this
research due to public health policies recommending consumption
of long chain omega-3 fatty acids, industry pressures to reduce
processing costs and the increasing appeal of healthy foods
for consumers.
and may produce Research is also underway to produce oilseed crops designed to act
pharmaceutical and as biofactories, producing pharmaceutical or industrial compounds

GM oilseed crops 6
 industrial products in instead of food, feed or fibre. Examples include pharmaceuticals
the future. (antibodies, vaccines or enzymes) and industrial compounds
(biofuels, bioplastics, lubricant oils or enzymes). These
developments are reviewed in more detail in the 2007 BRS report
Plant molecular farming in Australia and overseas.
GM oilseed crops that Combining a literature review and extensive consultations with
have been developed, or representatives of the oilseed industry and State Government
are in development, in Agencies, BRS has identified a range of GM oilseed crops that are
Australia include already in commercial production overseas or are under
herbicide tolerant and development in Australia and overseas.
insect resistant cotton, In Australia, GM herbicide tolerant and insect resistant cotton
and herbicide tolerant varieties and GM herbicide tolerant canola have been approved for
canola commercial release. Research is continuing into improving the fatty
acid profiles (e.g. high oleic and/or low linolenic acid levels) and
developing novel fatty acid compositions (e.g. long chain omega-3
fatty acids) for Australian oilseed crops.
while overseas GM Overseas countries have approved the commercial release of a
cotton, canola and number of GM oilseed crops, namely cotton, canola and soybean
soybean have been varieties incorporating herbicide tolerance, insect resistance or a
commercialised. combination of the two. Overseas researchers are developing
oilseed crops with improved environmental stress tolerance,
improved disease resistance, increased nitrogen use efficiency,
increased seed size, reduced pod shatter, increased yield and
increased oil content. Further research overseas is investigating
ways in which to produce oilseed crops with altered fatty acid
profiles or novel fatty acid composition, increased essential amino
acid content (e.g. methionine and lysine), production of essential
vitamin precursors (e.g. Vitamin A and E), increasing the
nutritional value of stockfeed (e.g. by reducing anti-nutritional
compounds such as phytic acid), and as expression platforms for
pharmaceutical and industrial products (e.g. plant-made vaccines or
bioplastics).
Oilseeds play a vital role The stakeholder consultations highlighted the important place that
as rotational break oilseeds have in Australian agriculture. In particular, canola
crops for wheat and provides value as rotational break crops to our major crop exports:
barley wheat and barley. Many people consulted were interested in the
potential for lower production costs (first generation traits), higher
value uses for oilseed crops such as specialty oils (second
generation traits), or crops as biofactories (third generation traits) as
potential ways to increase the value of oilseed crops. There was also
a desire for public breeding programmes to continue to ensure that
oilseed traits relevant to Australia are developed in the future.
and there are Canola and other oilseeds provide significant benefits to the
opportunities for Australian agricultural sector. A number of opportunities for
expansion and expansion and diversification of the Australian oilseed industry have
diversification. been identified, including: supplying increased quantities of local
oilseed meal to the stockfeed industry; developing stable oils for
frying that compete with palm and soybean oil imports; developing
value-added oils such as long chain omega-3 fatty acids to supply a
niche market for omega-3-containing oilseed crops; producing high
volumes of oil for use in biodiesel production; and

GM oilseed crops 7
 meeting the needs of niche markets for GM and non-GM oilseeds as
they develop.
Experience with GM Experience with GM cotton crops in Australia and GM cotton and
oilseed crops in canola crops overseas has shown that there are benefits for farmers
Australia and overseas and the environment from growing these crops. For instance, in
show economic, Australia the GM insect resistant cotton Bollgard II led to an 85%
environmental and reduction in insecticide use in comparison to conventional cotton
agronomic benefits. over the first three seasons following its introduction in the 2002/03
season. The implementation of other Integrated Pest Management
practices has also contributed to this reduction.
Results from studies overseas show net economic benefits for many
farmers growing GM crops. For GM cotton, the level of this benefit
varies between countries and regions within countries due to
differences in environmental and climatic conditions and in some
countries, the way that GM crops are developed and sold.
In Canada, a number of agronomic benefits have been associated
with the adoption of GM canola, including: improved yields;
decreased herbicide use and increased weed management options;
and increased ease in adopting minimum and no-till cultivation
practices.
A study by the Canola Council of Canada released in 2005 indicated
that Canadian canola farmers who chose to grow GM varieties were
better off when compared to those who continued to cultivate
conventional canola varieties. Canadian growers have not lost
market share in their main export markets despite the majority of
their canola crop being comprised of GM varieties. There is no
evidence that GM canola is having difficulty finding ready markets
throughout the world. No price premium on bulk non-GM canola
shipments has been identified.
Current challenges faced Consultations with the Australian oilseed industry identified a
by the oilseed industry in number of current issues, perhaps the most immediate of which is
Australia include that posed by the current drought. Canola production in the 2006/07
drought, competition season is estimated to be the lowest in a decade with New South
from soybean and palm Wales the worst affected state.
oils The development of soybean and palm oil varieties that produce oil
of similar quality to canola is of concern to the oilseed industry as
both these oils are available in large volumes and at low prices on
the world market. Developments such as these may affect the
competitiveness of the Australian oilseed industry.
and the inability to The stakeholder consultations also identified a number of important
adopt GM food crops challenges that will need to be addressed if GM oilseeds are to be
such as canola due to widely grown in Australia. The most important of these are the
moratoria. current State and Territory Government moratoria that restrict the
commercial growing of GM crops in many parts of Australia.
Other challenges identified by stakeholders were the need to
address the perceived lack of public acceptance of GM
technologies; and supply chain management issues such as
segregation and coexistence. Many stakeholders believed that the
introduction of GM oilseeds was necessary for the future viability

GM oilseed crops vii

 of the industry, while a few stakeholders strongly disagreed.
Further investment in developing Australian GM oilseeds will be
encouraged when there is a transparent and predictable pathway to
market. There is also a need for communication of independent,
credible, factual and practical information to producers and
consumers, to inform discussion of these issues.
Following a decade of growing GM cotton in Australia, it is clear
that these crops have provided economic, environmental and
agronomic benefits to Australia. Similar benefits may result from
growing other GM oilseed crops, if the issues identified in this
report can be addressed across the whole oilseed industry. The
widespread introduction of GM oilseed crops overseas is likely to
continue in the future. Many of the countries adopting GM crops
are competing with Australia in the world oilseed market.
In conclusion, available information strongly suggests that there will
be economic, environmental and agronomic benefits from growing
GM oilseed crops in Australia and opportunity costs if they are not
adopted

GM oilseed crops 8
Chapter 1 Introduction
This report is the result of a study funded by the Department of Agriculture, Fisheries and Forestry
(DAFF) using funds provided under the National Biotechnology Strategy to investigate the potential
impacts of genetically modified (GM) oilseed crops on the Australian oilseed industry. The aim of
this work is to enhance the understanding of applications of biotechnology in the agriculture and
food sectors. This report identifies potential effects of GM oilseeds on Australian agriculture.
Research for this report was based on a combination of a desktop review of published literature,
review of web-based information on Australian and international trials of GM crops and
consultation with the Australian oilseed industry. These consultations focused on farmers,
researchers, seed suppliers, food and feed manufacturers.
Definitions of the term GM vary, but for this report, the term refers to plants that have acquired
new genes by laboratory gene technology methods, as defined in the Gene Technology Act 2000
(Cth).

Section 1.1 Why focus on the oilseed industry?

The Bureau of Rural Sciences (BRS) report Whats in the Pipeline? (Glover et al., 2005)
highlighted the significant amount of research that is being conducted internationally into GM
oilseed crops. This included a large amount of work on second generation traits, which alter the
nutritional value of crops. For example, oilseed plants are being modified for enrichment of
specific desirable oils, minimising the purification steps in the oil extraction process; modified to
produce oils of consistent high quality and composition; and modified to produce desired ratios of
different fatty acids. Oilseed crops are also being modified to increase the quality of the meal for
animal feed.
First generation (input trait) GM cotton and canola varieties have already been approved for
commercial production in Australia and GM cotton, canola and soybean crops are widely grown
overseas. GM cotton was the first GM food crop grown commercially in Australia (e.g. cottonseed
oil is used for deep frying applications by the food industry and cotton linters (short cellulose
fibres) are used to manufacture a variety of food thickening agents), and GM canola is likely to be
the next food crop to be grown. In addition, third generation traits are being developed to allow
production of pharmaceutical or industrial products in plants.
Australia produces between 2 and 3 million tonnes of oilseeds each year; the major crops being
canola and cottonseed. Together these two crops account for 92% of total oilseed production in
Australia, with peanuts, soybeans and sunflower seeds accounting for the remaining 8%. The
impact of GM oilseed traits on Australian agriculture could result from their direct uptake in these
main crops, or via the uptake of GM traits in these or other oilseeds by key international
competitors. Both of these alternatives need to be considered to allow informed decision making on
the uptake of GM technology by the oilseed industry.

Section 1.2 International oilseed production, supply and

demand
The worlds major oilseed crops are soybean, rapeseed (including canola, Box 1), peanut, oil palm
and sunflower (Friedt and Luhs, 1998) (Table 1.1). Of these, palm oil and meal, soybean seed, oil
and meal and sunflower oil and meal are the most traded commodities (in terms of both export
volume and percentage of production that is exported). Approximately one third of the worlds
soybean seed, meal and oil products are exported. Similarly, sunflower meal and oil and palm
products have significant export markets. In contrast, the proportion of production that is exported

GM oilseed crops 3
is relatively low for peanut, cottonseed and rapeseed. These are mainly produced for Australian
consumption (Table 1.1).
BOX 1 Canola versus Rapeseed whats the difference?
Canola is the name for Brassica oilseed crops that have less than 2% erucic acid (a fatty acid) and
less than 30 micromoles of glucosinolates per gram of seed solids. Canola varieties were developed
in Canada in the 1950s from Brassica napus and B. campestris (Canola Council of Canada, 2005a;
Tribe and Kalla, 2005). It is also known as double zero or double low rapeseed (Gunstone,
2004; Canola Council of Canada, 2005a).
Prior to the development of canola-type varieties, rapeseed oil was used both as a food oil and in a
variety of industrial purposes, with a long history of use as lamp oil (Tribe and Kalla, 2005).
Following feeding experiments with rats that suggested that high dietary exposure to erucic acid
could lead to fatty deposits and lesions in heart muscle, it was decided to decrease the levels of
erucic acid in rapeseed oil (FSANZ, 2003a). Glucosinolate is an anti-nutrient that can reduce
uptake of iodine in rats (Duncan, 1991).
The term rapeseed is used in many countries to describe oilseeds produced from various Brassica
species, including B. napus, B. rapa and B. campestris. Where the oil produced from these species
has low erucic acid and the meal has low glucosinolate levels, it can be called canola. Canada and
Australia are the main users of this term. When countries report rapeseed production levels, the
figures will often also include Indian mustard (B. juncea).
Australia mainly grows canola varieties of rapeseed and the majority of GM research is occurring
in these varieties. Therefore, the term canola is used frequently throughout this report. However,
when describing world production figures, rapeseed is used as this is the most common term.
Soybean is produced mainly in the United States, South America (predominantly Argentina and
Brazil) and China (Appendix A). From Table 1.1 and Appendix D, it is clear that Australia grows
relatively small quantities of soybean and is a net importer of this oilseed.
Although Australia is also a relatively minor rapeseed producer, in the three years to 2005-06 it
accounted for 19% of world canola exports (if intra-European Union trade is excluded) (Foster and
French, 2007). The countries that produce the largest amounts of rapeseed (the European Union,
China and India) are usually net importers as they have large domestic markets. Production, supply
and demand figures for world rapeseed (Appendix B) indicate that Canada is Australias main
competitor for rapeseed markets. The major global import markets for rapeseed are shown in
Figure 1.1.
Although cotton fibre is an important export commodity for Australia, the export market for
cottonseed is small (Tribe and Kalla, 2005). Australia supplies approximately 28% of this market
(Appendix C). The United States and Australia are the major exporters of cottonseed. The major
import markets for cottonseed are shown in Figure 1.2. Australia is not a significant exporter of
oilseeds other than canola and cottonseed.

GM oilseed crops 4
 Japan (41%)
Mexico (20%)
Pakistan (13%)
China, Peoples Republic of, (8%)
United States, (7%)
EU-25 (4%)
Canada (3%)
Bangladesh (2%)

Figure 1.1 Relative sizes of global import markets for rapeseed (based on production
20032006).

Source: data in Appendix B

Mexico (30%)
Japan (17%)
EU-25 (17%)
Korea, Republic of (12%)
Turkey (10%)
South Africa, Republic of (7%)
Saudi Arabia (4%)
Others (2%)

Figure 1.2. Relative sizes of global import markets for cottonseed (based on production
20032006).

Source: data in Appendix C

GM oilseed crops 5
Section 1.3 Oilseeds are important to the Australian economy
The gross value of oilseed production averaged $766 million over the three years to 2005/06;
around 7% of the total gross value of Australian broadacre crop production (ABARE pers. comm.,
2006). Over the same period, Australian exports of oilseeds averaged $478 million, around 8.5% of
the total value of Australian exports of grain and oilseeds (ABARE, 2006). Oilseed production also
provided benefits to wheat and barley crops through their use in rotation cropping cycles.
Appendix D presents the amounts of oilseeds, wheat and barley produced in Australia,
demonstrating the much larger volume of cereal production.
Meal derived from oilseeds is an important source of protein in the stockfeed industry. Canola and
soybean meals are used extensively and cottonseed meals are also important. Sunflower meal is
used to a minor extent (see Table 3.1 Chapter 3). The stockfeed industry currently imports
significant volumes of soybean meal, as there is insufficient production in Australia. For instance,
imports of soybean meal can be up to 300 kt per year. There could be opportunities within the
oilseed industry in Australia to expand to meet some of this demand with meal from other oilseeds,
provided that there was also a market for the oil (Stakeholder Consultations).
Although oilseed crops make a relatively small contribution to the Australian economy as a
domestic and export commodity, they have a high value as break crops for cereals, such as wheat
and barley, that are of greater value to the Australian economy. A break crop is used in a crop
rotation system in order to help control weeds and plant diseases. When planted as part of a crop
rotation, they can increase the yield of subsequent cereal crops (Angus et al., 1991; Angus et al.,
1999; Evans et al., 2003). The size of this impact by canola on the following wheat crop is variable.
For instance, over a four year period (19901994) the two-year gross margin of canola-wheat was
approximately 50% higher than for wheat-wheat, however, in 1995-1996, which were both good
cropping seasons, the two-year gross margin was only about 25% higher for canola-wheat than for
wheat-wheat (Evans et al., 2003). This benefit for the following cereal crop is a good reason to
encourage the continuation of oilseed crops in Australian crop rotations where they contribute to
the competitiveness and sustainability of Australian agriculture (Stakeholder Consultations).

GM oilseed crops 6
Table 1.1. World oilseed production and exports (2005/06)
Rapeseed Cottonseed Soybean Sunflower Palm fruit Peanuts
a
(MMT) (MMT) (MMT) (MMT) (MMT) (MMT)
b
Total oilseed 48.42 42.31 224.12 29.98 9.71 33.09
production c
130
Total oilseed 6.35 1.05 66.35 2.01 0.10b 1.83
exports
(13%) (2%) (30%) (7%) (1%) (6%)
(% production)
Total meal 25.37 14.24 143.90 11.10 5.06b 5.72
production
Total meal 2.23 0.45 48.48 3.43 3.76b 0.15
exports
(9%) (3%) (34%) (31%) (74%) (3%)
(% production)
Total oil 16.59 4.67 33.87 10.46 34.80d 4.93
production 4.20 b

Total oil exports 1.29 0.11 9.41 3.09 25.89d 0.17

(% production) (8%) (2%) (28%) (30%) (74%) (3%)
b
1.93
(46%)
Note: India, China, Pakistan, Russia and Canada are the major producers of Indian mustard seed. Indian mustard
production figures are not available as they are usually combined with figures for rapeseed production.
a. Oil is produced from both the mesocarp (flesh) and kernel of the palm fruit. Palm oil (from the mesocarp) is high in
palmitic and stearic acid, while palm kernel oil is high in lauric acid (Gunstone, 2004). Meal is produced from the
kernels.
b. Figures for palm kernel.
c. Estimate of total palm fruit production based on 30% oil (total of mesocarp and kernel, Gunstone, 2004) and
39 MMT oil production.
d. Figures for palm mesocarp.
Source: (USDA-FAS, 2006b)

Section 1.4 Manipulation of oils through conventional

breeding or chemical modification
The most significant manipulation of an oilseed crop occurred in Canada in the 1950s when
rapeseed was made more suitable for use as a food crop through the development of new varieties
containing low levels of anti-nutritional factors (substances in foods which cause detrimental
effects to human and animal growth or performance). As discussed in Box 1, canola is now a
widely grown crop used in a variety of food applications, including cooking oils and margarines.
This genetic manipulation of oilseed crops continues today as each crop sector of the oilseed
industry strives to produce oils better suited to food or industrial applications at prices that are
competitive with alternative supplies, and which contain lower quantities of trans-fatty acids.
There is now a strong trend to breed crops that produce oils for specific end uses, rather than put
naturally occurring oils through a range of industrial processes to prepare them for the food
industrys various requirements for stability, solidity etc. There is great potential for the production
of designer oils for specific end uses. Such oils could potentially replace petrochemical-based

GM oilseed crops 7
products and may also lead to the development of new applications due to the unique properties
present in some of the more unusual fatty acids (EPOBIO, 2006).

1.4.1 Nutritional modifications

Fatty acids are essential to human and animal health and nutrition. However, high levels of
consumption of some fatty acids can be detrimental to health. Consumption of high levels of
saturated fatty acids (with the exception of stearic acid) can lead to an increase in total cholesterol,
which in turn has been linked to coronary heart disease. Unsaturated fatty acids, on the other hand,
reduce levels of bad cholesterol (low-density lipoproteins or LDL), increase good cholesterol
(high-density lipoproteins or HDL) and improve coronary health. Stearic acid has a neutral effect
on coronary health (Fernandez and West, 2005).
Fatty acids also vary in their stability and hence their suitability for a variety of cooking
applications. Saturated and monounsaturated oils are very stable and are used for high-temperature
deep frying, as well as applications that require solidity at room temperature. In contrast,
polyunsaturated fatty acids are unstable and readily oxidise at high temperatures resulting in
unpleasant odours, flavours and discolouration. Such properties have meant that cooking oils with
high levels of polyunsaturated fatty acids are routinely partly or wholly hydrogenated (chemically
modified) to increase their saturated fatty acid content. Hydrogenated oils are more stable during
cooking and are also solid at room temperature, allowing them to be used in margarines. However,
partial hydrogenation of polyunsaturated fatty acids leads to the production of trans-fatty acids,
which can cause health problems by raising total cholesterol and reducing HDL (good
cholesterol), similar to the effects of a diet high in saturated fatty acids (Australian Oilseeds
Federation, 2004).
In Australia, margarines are produced through a process of transesterification, where a soft oil and
hard fat are mixed together and modified to produce a soft, spreadable product. This process does
not produce trans-fatty acids (Stakeholder Consultations).
Figure 1.3 summarises the relative stability and healthiness of different fatty acids and specifies
the relative levels of each of the main fatty acids in canola and soybean oils. As indicated in this
figure, oils containing higher levels of oleic or stearic acids result in more stable and healthy
cooking oils for the food service industry.
Given this relationship between the nutritional and functional value of the different fatty acids,
oilseeds that are high in oleic and/or stearic acid and low in palmitic and polyunsaturated fatty
acids, produce healthy and stable cooking oils. Altering the oil profile of current oilseed crops to
meet such a profile and thus improve their health benefits, has been a major focus of non-GM
breeding programmes and GM oilseed crop research both domestically and overseas.
Developments in GM oilseed crops are discussed further in Chapter 2.
High oleic canola (Monola) has been conventionally bred in Australia and is of value to the frying
sector of the food industry, having 50% longer frying life than canola oil. It is produced in
segregated supply chains on a relatively small scale (White, 2004). Similar varieties have been bred
in Canada, where canola varieties with reduced linolenic acid and with high stearic acid have also
been bred (Carr, 2005). Dow AgroSciences has also developed canola varieties under the
trademark NexeraTM that produce high oleic acid oil called NatreonTM. As indicated in Figure 1.3,
these combinations of fatty acids produce healthier and more stable oils.
Conventional breeding of soybean has also resulted in the development of varieties with increased
oleic acid levels and/or reduced saturated fatty acids. Soybeans with low linolenic acid have been
produced in the United States and are marketed as Vistive by Monsanto. Although these traits are
not the result of gene technology, they have been incorporated by the seed companies into GM
herbicide tolerant soybeans because of their wide acceptance in North America and are not
available in non-GM backgrounds.

GM oilseed crops 8
Other oilseeds have been similarly modified, with a GM high oleic acid cottonseed variety
developed and trialled in Australia by the Commonwealth Scientific and Industrial Research
Organisation (CSIRO) (Liu et al., 2002a; Liu et al., 2002b; OGTR, 2003f). Conventionally bred,
high oleic sunflowers (e.g. Monosun and Sunola) have also been developed and are available
commercially; they are produced in a segregated supply chain in Australia.
Some oilseed modifications can be made using conventional breeding techniques, however, some
cannot and for these, GM techniques are required if these desirable traits are incorporated into the
plants. An example of this is the production of omega-3 fatty acids in oilseeds discussed in Chapter
2.

Saturated oils Unsaturated oils Polyunsaturated oils

palmitic stearic oleic linoleic linolenic

Stable Unstable

LDL = LDL LDL

Stable and healthy

cooking oils
Canola
4% 2% 22% 10%
62%
Soybean
11% 4% 54% 7%
24%
High-oleic soy
6% 3% 3% 4%
84%
Low linolenic soy
11% 5% 55% 3%
26%

Figure 1.3 Relationship between selected oils and their nutritional and functional
attributes.
Source: Figure modified from Carr (2005).

As shown in Figure 1.4, the changes being made to the fatty acid profiles of oilseed crops result in
fatty acid compositions that are increasingly similar to one another, which is likely to lead to
increased substitutability between these oils, dependant on relative availability and price, as
discussed below in Section 1.5.

GM oilseed crops 9
 Canola

Cottonseed

Soybean

HO Canola

HO Cottonseed

HO Soybean

Saturate Monounsaturate Polyunsaturate

Figure 1.4 Relative fatty acid compositions of conventional and high-oleic (HO)
cultivars of major oilseed crops.
Source: Singh et al (2005)

1.4.2 Industrial modifications

A major non-food use for vegetable oils is in the area of manufacturing surfactants, for example
detergent and soap production. Currently, the main sources for production of detergents are
coconut and palm kernel oil (EPOBIO, 2006). Vegetable oils are also manipulated for use as non-
food industrial lubricants, for instance for use as machinery oils. Although vegetable oils are a
renewable resource with lower pollution risks, they have some disadvantages in terms of stability
and purity (Cahoon, 2003; Castro et al., 2006). For example, the United States Department of
Agricultures (USDA) National Center for Agricultural Utilization Research has research
programmes that aim to increase the use of vegetable oils (particularly soybean oil) in industrial
applications (e.g. USDA Food and Industrial Oils program (USDA-FIO), (2006). Vegetable oils
may also be a sustainable replacement for petrochemicals in the production of paints, surface
coatings, inks, plasticisers, solvents and high utility polymers (e.g. both flexible and rigid foams,
and high performance adhesives, seals, gaskets and hard plastic parts) (EPOBIO, 2006).
The major advantages of vegetable oils are that they are biodegradable and have low toxicity.
Although vegetable oils can sometimes outperform mineral oil based products, there are some
drawbacks which currently restrict their use (EPOBIO, 2006). For example, one problem
encountered when using vegetable oils is their tendency to oxidise more rapidly than mineral oils.
They are also less stable at high temperatures, decreasing their usefulness for many industrial
applications. These problems can be reduced through a chemical modification known as
epoxidisation. Epoxidised soybean oil has been tested for suitability as a high temperature lubricant
oil and shown to have increased performance over unmodified soybean oil (Castro et al., 2006).

Section 1.5 Oils are readily substitutable

Most oilseeds contain combinations of five main fatty acids: stearic, palmitic (both saturated fats),
oleic (a monounsaturated fat), linoleic and linolenic (both polyunsaturated fats) with a number of
minor fatty acids (Table 1.2). The fatty acid profile of each oilseed crop is used to determine the
applications for which it can be used (Tribe and Kalla, 2005). However, most oils have a variety of
possible uses and substitution may occur between oils with similar properties. For example, linseed
oil (primarily an industrial oil) has been used in margarine and cottonseed oil (an edible oil) has

GM oilseed crops 10
been used in the past to manufacture soap (Nyberg, 1970). The degree to which substitution occurs
depends on both the composition and price of the oils in question.
Soybean and canola oils are readily interchangeable and this substitutability will increase with the
growing availability of newly developed soybean varieties with oil profiles closer to that of canola,
particularly high oleic varieties. As demonstrated in Figure 1.4, the current trend towards
developing high oleic acid cultivars is resulting in increased similarity of the oil profiles for
cottonseed, soybean and canola. Soybean dominates the world oilseed market due to its lower
production costs, higher quality meal and increases in oil quality, resulting from both conventional
breeding and GM technology, which will increase the value of its oil too. This could be a significant
threat to the Australian oilseed industry in the future (Stakeholder Consultations). This
is discussed further in Chapter 3 of this report.
Palm oil is widely used in the food services industry as it is very cheap; there are large volumes
available; and it has good properties for cooking applications. Its high level of saturated fatty acids
means that it has a high level of thermal stability and a long shelf life. As a result, it does not
require hydrogenation and there are no trans-fatty acids associated with it. However, palm oil
contains higher levels of saturated fats than many other vegetable oils (Gunstone, 2004). Health
professionals recommend minimising the consumption of saturated fatty acids to decrease the risk
of developing coronary heart disease (European Commission, 2001; NHMRC, 2003; USDHHS and
USDA, 2005).
Substitution of palm oil with other sources of oil such as high oleic oils would allow Australian
oilseeds to gain a larger domestic market (Stakeholder Consultations). The main impediment to this
strategy would be the very low cost of palm oil; Australian production of alternatives is unlikely to
be price competitive. There are also limits to such substitution. As an example, there are some
popular biscuits manufactured in Australia for which there is no substitute for palm oil because the
texture of these biscuits is not acceptable when another oil is substituted (Stakeholder
Consultations).

GM oilseed crops 11
Section 1.6 Current opportunities and challenges for the
Australian oilseed industry
A range of people from the oilseed industry were interviewed using the questions presented in
Appendix E. These people were farmers, seed producers, stockfeed manufacturers, food
manufacturers and people working for relevant State Government agencies and had expertise in
different aspects of the Australian oilseed industry. In total, 21 people, representing 17
organisations were interviewed. Six other organisations were approached but declined to
participate. The organisations and people targeted for consultation were selected with the aim of
gaining a balanced view of issues facing the oilseed industry and the effects that GM oilseeds may
have on it. Thus people expressed both positive and negative opinions about GM oilseeds and the
range of views has been reflected in this report. The issues raised are summarised below and are
explored in more detail in Part 2 of this report.
opportunities for expansion of the oilseed industry in Australia
regulatory issues
potential for developing niche markets
potential for health benefits for consumers
agronomic, environmental and economic benefits from GM cotton
potential benefits from adopting GM canola
effects of seasonal variability on the oilseed industry
the necessity for canola to compete with improved soybean oil traits
consumer attitudes towards GM foods
segregation and coexistence strategies
effects of the State and Territory Government moratoria on growing GM food crops
liability issues.

GM oilseed crops 12
Full article go to http://www.ipsnews.net/2013/06/biofortification-may-hold-keys-to-hidden-hunger/
IPS News Agency

Biofortification May Hold Keys to Hidden Hunger by Thalif Deen

ROME, Jun 21 2013 (IPS) - The U.N. Food and Agriculture Organisation (FAO), which works to end
malnutrition among more than two billion people worldwide, is expressing strong support for enriching the
micronutrient content of plants.
In technical terms, it is called biofortification: a nutrition-specific intervention designed to enhance the
micronutrient content of foods through the use of agronomic practices and plant breeding.
The breeding is taking place at HarvestPlus, an international programme supported by the Consultative
Group on International Agricultural Research (CGIAR) and at national agricultural research centres, mostly
in developing countries.
The first nutritious crop, developed by African scientists and released in partnership with the Internal Potato
Center (CIP), was the orange sweet potato, which has been effective in providing up to 100 percent of daily
vitamin A needs for young children, according to CGIAR.
Six additional nutritious crops are now being developed through the use of conventional breeding methods:
vitamin A-rich cassava and maize, iron-rich beans and pearl millet, and zinc-rich wheat and rice.
The first three crops are targeted to Africa and the rest to South Asia.
New varieties of the first four crops were launched in 2012, says CGIAR, with wheat and rice expected to
follow later this year.
While it takes time to produce the amount of seed necessary to meet demand, up to half a million farmers
will be growing these nutritious crops by year end, it predicts.
Asked how far plant breeding can go in resolving hunger and nutrition problems worldwide, Dr. Erick Boy,
head of nutrition at HarvestPlus, told IPS, Our focus is on hidden hunger, caused by not getting enough
minerals and vitamins in the diet that is the major hunger problem the world faces today.
The six new varieties of staple crops we are developing are more nutritiousthey contain higher amounts
of vitamin A, zinc, or iron, he added.
Lack of these nutrients is what causes widespread suffering and health problems, especially for women and
children.
Boy said these crops will be distributed to more than three million farming households in seven countries in
Africa and Asia by 2015.
Not bad for a programme that started from scratch to develop these crops beginning only in 2003, he
noted.
When eaten regularly, these nutritious crops could provide on average 50 percent of vitamin A, zinc, or iron
requirements. According to CGIAR, more than two billion people worldwide do not get enough of these
crucial nutrients in their diets.
Deficiencies can lead to lower IQ, stunting, and blindness in children; increased susceptibility to disease for
both children and adults; and higher health risks to mothers and their infants during childbirth.
According to the World Bank, malnourished children are more likely to drop out of school and have lower
incomes as adults, thus reducing overall economic growth.
In its latest annual flagship publication The State of Food and Agriculture (SOFA) released here, FAO
explains that unlike food fortification, which occurs during food processing, biofortification involves
enriching the micronutrient content of plants.
Questions remain about the readiness of consumers to purchase biofortified foods, especially when they look
or taste different from traditional varieties. But, FAO says, early evidence suggests that consumers are
willing to buy them and may even pay a premium.
In Uganda, FAO discovered consumers were willing to pay as much for the orange-fleshed varieties of
sweet potato as for the white varieties, even in the absence of a promotional campaign.
Similar results were found for nutritionally-enhanced orange maize in Zambia, where consumers did not
confuse it with ordinary yellow or white maize. They were also willing to pay a premium when its
introduction was accompanied by nutrition information.
Asked why the project targets Asia and Africa and not Latin America, CGIARs Dr. Boy said, Our focus is
on subSaharan Africa and South Asia because if you look at any map of hidden hunger, these are the regions
marked in red.
Latin American countries have done a better job of improving nutrition over the past two decades, he added.
There are still, however, pockets where hidden hunger is a problem.
So we are also working in this region. In fact, I am in Guatemala now to work with stakeholders to buy in
to our high-iron beans and high zinc-maize initiative there. We anticipate that we could have varieties of two
to three crops that are rich in iron and zinc to LAC farmers by 2015, Boy added.
Meanwhile, in early June, the UK government granted 30 million [46.4 million dollars] to HarvestPlus to
develop and deliver six nutritious crops to several million farming households in Africa and Asia.
The grant was announced at a high-level international meeting in London that brought together a range of
partners to make strong political and financial commitments to improve nutrition globally.
In his opening remarks, British Prime Minister David Cameron said, It has to be about doing things
differentlyFor science, its about harnessing the power of innovation to develop better seeds, [and] more
productive and nutritious crops.
 European Journal of Nutrition & Food Safety
5(1): 59-71, 2015, Article no.EJNFS.2015.003
ISSN: 2347-5641

SCIENCEDOMAIN international
www.sciencedomain.org

Customers Purchasing Organic Food - Do They Live

Healthier? Results of the German National
Nutrition Survey II
Marianne Eisinger-Watzl1*, Friederike Wittig1, Thorsten Heuer1
and Ingrid Hoffmann1
1
Department of Nutritional Behaviour, Max Rubner-Institute, Federal Research Institute of Nutrition
and Food, Haid-und-Neu-Str. 9, 76131 Karlsruhe, Germany.

Authors contribution

This work was carried out in collaboration between all authors. Author IH headed the project. Author
MEW conducted the analysis and interpretation and drafted the manuscript. Author FW was involved
in the early work and assisted with the interpretation of the results. Author TH was in charge of data
management. All authors were involved in the critical revision of the manuscript for important
intellectual content. All authors read and approved the final manuscript.

Article Information

DOI: 10.9734/EJNFS/2015/12734
Editor(s):
(1) Prof. Hans Verhagen, Senior Scientific Advisor Nutrition and Food Safety, National Institute for Public Health and the
Environment (RIVM), P.O. Box 1, 3720 BA, Bilthoven, The Netherlands.

Received 16th July 2014

nd
Original Research Article Accepted 22 October 2014
Published 17th November 2014

ABSTRACT

Aims: Using national food consumption data this paper addresses issues whether buyers of
organic food make healthier food choices and pursue a healthy lifestyle concerning smoking,
physical exercise and body weight compared to non-buyers.
Study Design: The German National Nutrition Survey II (NVS II) is a nationwide food consumption
study providing additional information on social demographics, health, and lifestyle aspects as well
as anthropometric measurements. Using data of several assessment tools, a comparison was
conducted between buyers and non-buyers of organic food.
Place and Duration of the Study: From November 2005 to November 2006, data collection took
place in about 500 randomly chosen sample points across Germany.
Methodology: 13,074 participants aged 18-80 years were divided into groups of buyers (44.9%)
and non-buyers (55.1%) of organic food. According to their purchase frequency, the organic food
buyers were further differentiated into intensive, occasional or infrequent purchase groups. A diet
history method was applied to assess food consumption, trained staff measured BMI while
questionnaires were used for social demographic description and healthy lifestyle factors.
____________________________________________________________________________________________________

*Corresponding author: Email: marianne.eisinger-watzl@mri.bund.de;

 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003

Results: Buyers of organic food consumed 17% more fruit and 23% more vegetables as well as
less meat/sausages (25%) and soft drinks (58%) than non-buyers did (P< .001, resp.). These
results are more pronounced for women and for intensive buyers. Additionally, buyers of organic
food exhibit healthier lifestyle characteristics in respect to smoking behaviour, physical activity, and
body weight compared to non-buyers.
Conclusion: German buyers of organic food demonstrate health behaviours complying better with
the recommendations for a healthy lifestyle compared with non-buyers. Independent of the
discussion whether organically produced food exerts additional health effects, buyers of organic
food make healthier food choices than non-buyers, thereby benefiting for their overall health.

Keywords: Food consumption survey; organic food; food choice; healthy lifestyle.

1. INTRODUCTION more vegetables, fruit and cereals as well as less

Data on market statistics exhibit raising growth of
the global organic market although the numbers
vary a lot between countries and regions [1]. The
enormous increase of organic markets in many
countries prompted numerous studies to gain
more insights in the supply as well as the
demand sector. In the latter case, a multitude of
studies concentrate on customers perception
toward organic food and the factors explaining
customers decision-making process [2-13].
Summarizing these studies reveals that customer
motivation for buying organic food is not
homogenous. However, some motives are
consistent, such as a high concern regarding
personal and environmental health, including the
aspect of food safety, and animal welfare [6,8,14-
16].
In the focus of marketing research, these
customer attitudes and motivations are relevant
in order to understand and manage
developments in the organic markets. From a
public health point of view, interests concentrate
on the connection of this health driven customer
decision and other healthy lifestyle aspects.
Emerging questions are whether organic food
buyers compared to non-buyers make healthier
food choices and whether health consciousness
of organic buyers stated as purchase motivation
is related to smoking, physical activity and body
weight. These lifestyle factors are fundamental
behaviour connected to risk management of
major chronic diseases and therefore important
indicators of overall health behaviour [17,18].
So far, there are only few studies investigating
whether and how the stated interest in health of
organic food buyers reflects in lifestyle
behaviour. Regarding food choice, a study in
Great Britain showed that, compared to official
census data, 274 regular customers of organic
food, who completed a diet diary, consumed
animal products with the exception of fish [19]. A
Norwegian study obtained similar results
studying 63,808 pregnant women by means of a
food frequency questionnaire. Women with
frequent organic food consumption were more in
line with public recommendations for healthy
eating [20].
Two other studies, conducted in Germany
respectively Denmark, investigated eating habits
using household panel data and expenditures on
organic or conventional food. Households with
heavy expenditure on organic food complied
closer with the recommendations for a healthy
diet [21,22]. However, the data resulted from
food purchase of entire households rather than
food consumption of individuals. The terms
consumer and purchaser are often used
synonymously in literature. Hughner et al. [10]
emphasise the incorrect terminology in their
overview on studies regarding the decision-
making process for buying organic food. They
mention that in all studies the term consumer is
used relying on purchase data. In consequence,
they demand the accurate differentiation
between purchasers (or customers) and
consumers of organic food as well as additional
research on food consumption.
The French study of the Nutrinet-Sant cohort is
the only study published so far comparing
organic food consumer and non-consumer based
on the assessment of food consumption of a
large general population (n=54,311) [23]. For
Germany, the National Nutrition Survey II
(NVS II), a representative food consumption
study, offers the required data for a sample of
more than 13.000 participants. Besides the data
on food consumption, the survey provides
information on social demographics, health and
lifestyle aspects as well as anthropometric
measurements [24,25] and therefore allows the
examination of differences in food choice as well
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003
as lifestyle characteristics between buyers and
non-buyers of organic food in Germany.
2. METHODOLOGY
2.1 Data Collection and Description of
Variables
The baseline study NVS II was commissioned by
the Ministry of Food, Agriculture and Consumer
Protection and the dietary assessment referred
to in the present study was undertaken from
2005 until 2006 [24]. At about 500 randomly
chosen sample points across Germany, the local
register offices randomly selected 14-80 year old
participants of which 18-80 year old participants
were integrated in this evaluation. Computer
assisted personal interviews were conducted at
study centers to obtain basic socio-demographic
information followed by a computer assisted diet
history interview. The diet history method is a
dietary assessment instrument asking directly for
the habitual food consumption [26,27]. As
software DISHES (Diet Interview Software for
Health Examination Studies) was used, a
program developed by the Robert Koch-Institute,
Berlin [28,29]. The standardized interviews lead
participants through the sequence of meals of a
day followed by questions on the frequencies of
consumed food. Participants also completed a
questionnaire on health-related aspects and
leisure time activities. Additionally, anthropo-
metric measurements were conducted at the
study centers [24]. During data assessment,
participants and interviewers were not aware of
the future research question regarding organic
purchase. This diminishes a possible response
bias.
There was no specific differentiation for the
consumption of organically or conventionally
produced food. By means of a questionnaire,
participants were classified into buyers and non-
buyers based on the question: Do you buy
organic food? (yes/no). Furthermore, buyers
were asked to give detailed information on the
frequency of their purchase of twelve organic
food items. The frequencies were coded with
points and a ratio was achieved by dividing the
sums of points by the number of given answers
to correct for the possibility that no answer was
given or a person did not eat/drink a food item.
Based on the ratios buyers were divided in either
the intensive, occasional or infrequent purchase
group.
Social class was defined as an interrelation of
several factors, so an index was generated
based on the degree of education, the
employment status of the households principle
earner, and the monthly net income of the house-
hold. Individuals were classified according to
these characteristics by a point system [24]. The
answers to four questions, (1) arrangement of
food items perceived as important for a healthy
diet, (2) classification of probiotic yoghurt, (3)
classification of ACE-beverages and (4)
comprehension of the 5-a-day campaign, from
the questionnaire were combined to a NVS II
nutrition knowledge index. Information about
smoking (yes/no) and physical activity (yes/no)
was retrieved according to statements from a
questionnaire. Body Mass Index (BMI) was
calculated dividing weight into kilograms by
height in square meters. In order to classify the
BMI, the WHO definition was employed [30].
The consumption of fruit, vegetables, meat and
sausages as well as sweets and soft drinks
serves as indicator for a healthy food pattern.
Food consumption data were evaluated and
compared with the dietary guidelines of the
German Nutrition Society [31,32].
2.2 Data Analysis
For descriptive analysis, Pearsons Chi-square
test was applied to 2x2 contingency tables,
testing for independence of two criteria. For
multiple comparisons, the Bonferroni correction
was subsequently used. Food consumption was
calculated as arithmetic mean ( standard error
of mean). The presentation of the means instead
of medians was chosen because several food
groups were consumed by less than 50% of the
participants. Consumption data did not show
normal distribution and could not be normalized
by log-transformation, thus non-parametric tests
were carried out. The Mann-Whitney-Wilcoxon
test was applied to compare food consumption of
buyers and non-buyers of organic food. Kruskal-
Wallis test was used to compare more than two
groups.
Statistically significant differences were defined
at the p-level of 0.05. Calculations were
performed with SAS program version 9.2. (SAS
Institute Inc., Cary, North Carolina, USA).
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003

3. RESULTS buyers reached the highest average age of 52.4

Of the 13,074 participants 44.9% stated to
purchase organic food whereas, 55.1% did state
not to do so. The frequency of purchasing
organic food revealed a proportion of 5.1%
intensive, 26.2% occasional and 13.6%
infrequent buyers summing up to the 44.9% of
total buyers.
3.1 Social Demographic Description
More women (54%) than men participated in the
NVS II; sex difference was even more
pronounced regarding buyers of organic food
with 61% being female. The more intense the
organic purchase the higher the proportion of
women: 67% intensive, 61% occasional and 59%
infrequent buyers of organic food were female. In
the non-buyer group, sex difference was almost
balanced (49% female vs. 51% male).
The comparison of age categories revealed that
the proportions of the youngest (age 18-24) and
the senior participants (age 65-80) were bigger in
the non-buyer-group than in the buyer-group and
vice versa in the age groups of 35-50 and 51-64
years. There was no difference in the age group
of 25-34 years (Fig. 1). The group of intensive
years compared to 49.5 years for the occasional
buyers, 47.9 years for the infrequent buyers and
48.4 years for the non-buyers (P< .001, resp.).
Regarding social class affiliation more buyers of
organic food belong to the two upper classes
(60%) compared to non-buyers (44%) (Fig. 2).
The proportions of the two lowest classes sum
up to 14% for buyers and to 24% for non-buyers.
Every social class is represented in the three
frequency groups of organic food purchase
(intensive, occasional, infrequent).
3.2 Food Consumption
Analyses of the diet history interviews revealed
that mean usual consumption of fruit of
participants buying organic food was 17% higher
than the consumption of non-buyers (P < .001)
(Fig. 3, by sex). Half of the organic food buyers
(50%) met the dietary guideline of the German
Nutrition Society of 250 g fruit per day [31,32],
while 40% of the non-buyers met the guideline
(P < .001). Regarding the frequency of the
organic purchase 65% of the intensive, 52% of
the occasional and 42% of the infrequent buyers
met the recommendation (P < .001, resp.).
Differentiated by sex 54% of all female organic

18-24 years 25-34 years 35-50 years 51-64 years 65-80 years
100%
20.3 *
22.0
80%

26.0 *** 23.6

60%

40% *** 31.7

36.4

20%
P=.851
11.4
11.5
5.8 11.2
0% ***
Buyer Non-buyer
Fig. 1. Buyer (n= 5875) and non-buyer (n= 7199) of organic food by age group
*** P < .001, * P <.05
Comparisons are based on Chi square test, Bonferroni correction
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003
buyers and 44% of all male organic buyers
consumed at least 250 g/d fruit (non-buyers: 45%
for women and 36% for men). There were
significant differences between female organic
buyers and non-buyers as well as between the
male organic buyers and non-buyers (P < .001,
resp.).
For women and men, mean consumption of
vegetables was 23% higher for the organic food
buyer group than for the non-buyer group
(P < .001) (Fig. 3, by sex). The German dietary
guideline for vegetable consumption of at least
400 g per day [31,32] was reached by 18% of all
organic food buyers and by 10% of all non-
buyers (P < .001). The proportion of meeting the
recommendation within the three frequency
groups of organic food purchase was 27% for
intensive, 19% for occasional and 14% for
infrequent buyers (P < .001, resp.). Differentiated
by sex, 20% of the women and 16% of the men
of the organic food buyers met the
recommendation compared to 11% female and
10% male non-buyers (differences between
female buyers and non-buyers as well as male
buyers and non-buyers: P < .001, resp.).
Organic food buyers consumed 25% less meat
and sausages than non-buyers of organic food
(P < .001) (Fig. 3, by sex). The German dietary
lower class lower middle class
100%
*** 14.9
27.7
80%
29.5
60% ***
32.3
40% 32.0
***
25.8
20%
15.6
9.9 ***
4.3 *** 8.1
0%
Buyer Non-buyer
Fig. 2. Buyer (incl. sub-groups) and non-buyer of organic food by social class, intensive
(n= 661), occasional (n= 3428), infrequent (n=1786)
*** P < .001; Comparisons are based on Chi square test, Bonferroni correction.
guideline for meat and meat products is 300 to
600 g per week [31,32]. Of all organic food
buyers, 48% exceed the recommendation
compared to 64% of the non-buyers (P < .001).
This was the case for 33% of the intensive
buyers compared to 48% of the occasional and
53% of the infrequent buyers (P < .001, resp.).
The proportion of women exceeding the
recommendation was 35% for organic food
buyers and 45% for non-buyers (P < .001) while
the proportion of men exceeding the
recommendation was 68% for buyers of organic
food and 81% for non-buyers (P <.001).
All men regardless of buying organic food and in
which frequencies, consumed comparable
amounts of sweets while organic food buying
women ate fewer sweets (45 g/d) than non-
buying women (48 g/d) (P < .05) (Fig. 4). Female
intensive buyers consume the least sweets
(41 g/d), followed by female occasional buyers
(45 g/d) and female infrequent buyers (48 g/d)
(P < .001, resp.). Buyers of organic food
consumed 58% less soft drinks (like soda and ice
tea) than non-buyers (P < .001) (Fig. 4, by sex).
However, a high proportion of all participants did
not drink soft drinks at all: 77% of organic food
buyers and 65% of non-buyers (P <.001).
Considering only the consumers of soft drinks,
middle class
28.8 24.4
30.6
31.5
31.0 33.0
27.6
23.6 25.3
10.3 11.6
9.0
4.5 3.9 5.0
Intensive buyer Occasionally Infrequent
buyer buyer
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003

the recorded quantities of the buyers were still
37% less than the quantities of the non-buyers of
organic food (P < .001). The average soft drink
consumption of women buying intensive,
occasional and infrequent organic food were
163 ml/d, 226 ml/d and 260 ml/d, while women
belonging to the non-buyer group consumed
313 ml/d (P< .001, resp.). The corresponding
consumption in men was 345 ml/d (intensive),
310 ml/d (occasional), 329 ml/d (infrequent) and
505 ml/d for non-buyers (P < .001, resp.).
Altogether, organic food buyers more often met
the food-based dietary guidelines of the German
Nutrition Society [31]. Consequently, the
absolute nutrient intake as well as nutrient
densities of buyers of organic food corresponded
better to the recommendations than those of the
non-buyers [33] (data not presented).
3.3 Health Aspects and Anthropometric
Measurements
Buyers of organic food more often were non-
smokers and vegetarians and indicated more
often to be physically active (Table 1). In
addition, organic food buyers more often were
interested in nutrition information and achieved a
higher score in the NVS II nutrition knowledge
index. More frequently, they stated to be satisfied
with their personal health status. Within the buyer
group, the proportion of normal body weight was
higher and the proportions of overweight and of
obese participants lower than in the non-buyer
group. This was also the case for women and
men respectively (with the exception of the male
overweight participants) (Fig. 5).

450
dietary guidelines
400 (DGE 2013)
350
300
***
250 *** *
g/day

200 ***
150
100
***
50 ***
0
Buyer Non- Buyer Non- Buyer Non- Buyer Non- Buyer Non- Buyer Non-
buyer buyer buyer buyer buyer buyer

women men women men women men

Fig. 3. Mean consumption ( SE) of fruit, vegetables and meat/sausages of buyer and non-
buyer of organic food, by sex Buyer: women (n=3577), men (n= 2298); Non-buyer: women
(n= 3536), men (n= 5961)
*** P < .001; Comparisons are based on Mann-Whitney-Wilcoxon test.

Table 1. Lifestyle factors of buyers and non-buyers of organic food

Percentage of purchase group1 Buyers Non-buyers

n=5875 n=7199
Non smoker 80,0 71,0
Being physical active (yes/no) 66,0 52,0
Vegetarian 1,8 0,3
Interest in nutrition information 84,0 64,0
High score in the NVS II nutrition knowledge index 39,0 23,0
High satisfaction with personal health status 78,0 74,0
1
differences between the purchase groups in every case: P < .001; Comparisons are based on Chi square test
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003

250

200

150

***
100

50

P=.790
0 * ***
Buyer Non- Buyer Non- Buyer Non- Buyer Non-
buyer buyer buyer buyer

women men women men

sweets g/d soft drinks ml/d

Fig. 4. Mean consumption ( SE) of sweets and soft drinks of buyer and non-buyer of organic
food, by sex buyer: women (n=3577), men (n= 2298); non-buyer: women (n= 3536),
men (n= 5961)
*** P < .001, * P <.05; Comparisons are based on Mann-Whitney Wilcoxon test.

underweight (BMI<18.5) normal weight (BMI: 18.5-24.9)

overweight (BMI: 25.0-29.9) obese (BMI30.0)

100%
17.9 *** 22.5 17.7 *** 22.2 18.2 *** 22.6
80%
27.7 ***
35.5 *** 31.5
60% 39.2 47.6 d 46.6

40%
52.6 ***
45.1 *** 44.8
20% 37.2 33.5 *** 30.0
c
1.4 a 1.1 1.9 b 1.6 0.6 0.7
0%
Buyer Non-buyer Buyer Non-buyer Buyer Non-buyer

total women men

Fig. 5. Organic food buyer and non-buyer by body weight (BMI) and sex organic food buyer:
women (n=3577), men (n= 2298);Non-buyer: women (n= 3536), men (n= 5961)
*** P < .001, aP=.178, b P=.277, cP=.675, dP=.473;
Comparisons are based on Chi square test, Bonferroni correction
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003
4. DISCUSSION AND CONCLUSION
With the data of the NVS II, the nationwide
representative nutrition survey for Germany, a
comparison between buyers and non-buyers of
organic food was conducted focusing on food
consumption and lifestyle factors. The results
show that, compared to non-buyers, German
buyers of organic food eat more fruit and
vegetables, less meat and sausages and
therefore demonstrate a more favorable food
choice. This food pattern is more in line with
national and international recommendations
[31,34-38]. These results apply to both, women
and men, while women do even better. The more
intense the purchase of organic food the more
favorable the food choice turns out to be.
Additionally, buyers of organic food decide to
lead a healthier lifestyle as indicated by lower
body weight, higher level of physical activity and
less smokers than non-buyers.
The French web-based cohort study from a large
sample of volunteers identified five clusters, two
of them comprising occasional and regular
organic food consumers [23]. The comparison of
food consumption, assessed with three 24-h
recalls, between non-consumers and consumers
of organic food revealed that food patterns of
organic food consumers show a higher
agreement with dietary guidelines, which gets
even better with increasing frequency of
consumption. In addition, the presented results
regarding body weight were confirmed by the
French study.
A higher percentage of normal body weight in the
organic food group was also found in the
Norwegian study with pregnant women [39]. In
1998, a comparison in the Netherlands between
a sample of 271 customers of health food stores
and a nationwide household sample (n=581)
showed that organic food buyers displayed the
lowest BMI [40].
Healthy food choices, normal body weight, being
physically active, as well as non-smoking
contribute substantially to a healthy lifestyle [18].
There is a robust understanding of links between
healthy lifestyle factors and a lower risk of
developing diabetes, heart disease, and cancer.
In the EPIC-study, the lifestyle factors healthy
diet, non-smoking, high level of physical activity
and normal body weight are associated with a
reduced risk of developing diabetes (93%),
myocardial infarction (81%), stroke (50%) and
cancer (36%) compared to participants who do
not indicate any health factors [41]. Data of the
NHANES study in the United States show the
impact of three lifestyle factors (healthy diet, non-
smoking and being physically active) with a risk
reduction of 65% for cardiovascular disease and
83% for cancer [17]. Other studies also confirm
these findings but are even more specific for
special diseases like heart disease [42] and
diabetes mellitus [43]. The four lifestyle factors
(healthy diet, non-smoking, high level of physical
activity and normal body weight) also proved
responsible for a risk reduction of 66% regarding
the endpoint all-cause mortality [44]. Ford et al.
[17], found strong relationships for three lifestyle
factors (healthy diet, non-smoking and being
physically active) and all-cause mortality with an
82% risk reduction compared to participants
without an indication of health factors.
The outcome is not to be confused with the
question whether health effects of organically
produced food compared to conventionally
produced food are evident. Related research
concentrates on differences in health effects of
foods from the two farming systems and is
discussed elsewhere [45-51]. Instead, the data
set of the NVS II allows the comparison of food
consumption and lifestyle factors of buyers and
non-buyers of organic food. The results show
overall better health behaviour of buyers of
organic food.
The presented finding that for the German
population from 2005-2006 45% are buyers and
55% were non-buyers differs from results of the
kobarometer. The kobarometer is an
annual assessment initiated by the Federal
Ministry of Food, Agriculture and Consumer
Protection of Germany. For the years 2005 and
2007 the data collection indicated about 75% of
the interviewed participants (about 1.000
participants starting at age 14) buying organic
food in different frequencies [52,53]. In contrast,
the German household study with a sample
of 13,000 totaled 50% as buyers for 2008 [21]
showing an increase from 36% of the same
household panel in 2005 [54]. Thus, the finding
of the present study that 45% of for the German
population bought organic food in the year 2006
is in a comparable range with the household
study 2008 regarding sample size and
percentage of organic purchase.
Socio-demographics are used to test how the
profile of the organic buyers of the NVS II
compares to other international and German
studies. The finding that women are generally
 Eisinger-Watzl et al.; EJNFS, 5(1): 59-71, 2015; Article no.EJNFS.2015.003
more engaged in health and nutrition has often
been demonstrated [10,14,23,55,56].
Participants with a higher social status, i.e.
higher level of education and income, have also
more often been linked to intensive organic
purchase behavior [2,23,54,57,58], while age is a
factor with no clear association [10,14,39,54]. In
summary, concerning socio-demographic
variables the findings of the present study are in
concordance with other studies on organic
buyers.
All 13,074 participants underwent a personal
interview regarding their usual food consumption
of the last four weeks using the diet history
method. The advantage of this dietary
assessment method is the focus on long-term
dietary habits while limitations comprise the
cognitive effort to remember all consumed food
and beverages of the last four weeks [26,59].
Especially the answers for inhomogeneous food
groups, like vegetables and pastries, may be
affected by either social desirability or
incapability of memorizing complexity [60]. As the
limitation of this assessment method applies to
both study groups, it should not affect the
comparison between buyers and non-buyers of
organic food.
As strengths of the presented study has to be
seen the sample size of 13,074 participants and
the representative database for Germany.
Together with the French cohort [23], two studies
with the general population are available being in
accordance regarding the main results that
consumers buying organic food display dietary
patterns more in line with dietary guidelines.
Additionally, with both studies body weight was
assessed demonstrating that buyers of organic
food display lower body weight than non-buyers.
A limitation of both studies is the cross-sectional
design. The data only represent a situation for a
specific moment. Therefore, causes and effects
cannot be differentiated. Whether the effort for a
healthier nutrition leads to the purchase of
organic food or buying organic food leads to
healthy choices cannot be determined and
causalities cannot be deduced.
The results of the presented study demonstrate
for the German population that buyers of organic
food make healthier food choices and their
lifestyle is closer in line with recommendations
for a healthy living. In doing so, subjects probably
reduce their personal risk regarding several
major diseases. Whether consumption of organic
food has an additional positive influence on
health is still under scientific debate. From a
public health point of view, organic food buyers
lead a way of life closer to the recommendations
for a healthy lifestyle and thereby profit by
scientifically well-established health benefits and
risk reductions.
ACKNOWLEDGMENTS
Franziska Lindner (KIT, Dept. Mathematics,
Karlsruhe) was a consultant for methods of
analysis.
The authors acknowledge the work of many
colleagues involved in the NVS II.
FINANCIAL SUPPORT
The Federal Ministry of Food, Agriculture and
Customer Protection of Germany funded the
German National Nutrition Survey II.
The Federal Ministry of Food, Agriculture and
Customer Protection of Germany financed the
present study in the framework of the Federal
Scheme of Organic Farming (BL, now BLN;
grant number: 28080E056). The project was
granted in collaboration with Prof. A. Spiller,
Marketing for Food and Agricultural Products,
University of Gttingen.
COMPETING INTERESTS
The authors have declared that no competing
interests exist.
REFERENCES
1. Willer H, Lernoud J. Organic Agriculture
Worldwide: Current Statistics. BioFach,
12.2.2014, Messezentrum, Nrnberg
2014.Accessed 4 March 2014.
Available:http://www.fibl.org/fileadmin/docu
ments/de/news/2014/willer-2014-global-
data.pdf
2. Dimitri C, Dettmann RL. Organic food
consumers: What do we really know about
them? British Food Journal.
2012;114(8):115783.
3. Aertsens J, Mondelaers K, Verbeke W,
Buysse J, Van Huylenbroeck G. The
influence of subjective and objective
knowledge on attitude, motivations and
consumption of organic food. British Food
Journal. 2011;113(10-11):135378.
4. Hjelmar U. Consumers' purchase of
organic food products. A matter of
convenience and reflexive practices.
Appetite. 2011;56(2):33644.
 REVIEW ARTICLE
PUBLISHED: 3 FEBRUARY 2016!| ARTICLE NUMBER: 15221 | DOI: 10.1038/NPLANTS.2015.221

Organic agriculture in the twenty-first century

John P. Reganold* and Jonathan M. Wachter

Organic agriculture has a history of being contentious and is considered by some as an inefficient approach to food production.
Yet organic foods and beverages are a rapidly growing market segment in the global food industry. Here, we examine the per-
formance of organic farming in light of four key sustainability metrics: productivity, environmental impact, economic viability
and social wellbeing. Organic farming systems produce lower yields compared with conventional agriculture. However, they are
more profitable and environmentally friendly, and deliver equally or more nutritious foods that contain less (or no) pesticide
residues, compared with conventional farming. Moreover, initial evidence indicates that organic agricultural systems deliver
greater ecosystem services and social benefits. Although organic agriculture has an untapped role to play when it comes to the
establishment of sustainable farming systems, no single approach will safely feed the planet. Rather, a blend of organic and
other innovative farming systems is needed. Significant barriers exist to adopting these systems, however, and a diversity of
policy instruments will be required to facilitate their development and implementation.

O
rganic agriculture has a history of being contentious.
Emblematic of this, and representing the prevailing attitudes
of many farmers and scientists in the 1970s and 1980s, are
the unsympathetic words uttered in 1971by then US Secretary of
Agriculture Earl Butz: Before we go back to organic agriculture in
this country, somebody must decide which 50 million Americans
we are going to let starve or go hungry1. At the turn of the twenty-
first century, sceptics considered organic agriculture to be ideologi-
cally driven and inefficient 2,3. They argued that organic agriculture
relies on more land to produce the same amount of food as con-
ventional agriculture and that adopting organic agriculture on too
large a scale could potentially threaten the worlds forests, wetlands
and grasslands2,3. They also asserted that organic agriculture has too
many shortcomings and poor solutions to agricultural problems2,4.
Organic agriculture is still considered by some critics as being an
inefficient approach to food security 5,6 and a farming system that
will become less relevant in the future6.
Yet the number of organic farms, the extent of organically farmed
land, the amount of research funding devoted to organic farming and
the market size for organic foods have steadily increased7. Sales of
organic foods and beverages are rapidly growing, increasing almost
fivefold between 1999and 2013to US$72 billion (ref.7; Fig.1); this
2013 figure is projected to double by 2018. Moreover, recent interna-
tional reports recognize organic agriculture as an innovative farm-
ing system that balances multiple sustainability goals and will be of
increasing importance in global food and ecosystem security 810.
Here, we review the performance of organic farming systems
in the context of sustainability metrics and global challenges, and
examine some of the barriers to the adoption of organic farming
systems and the policies needed to overcome them.
Organic practices and certification
Organic agriculture, sometimes called biological or ecological
agriculture, combines traditional conservation-minded farming
methods with modern farming technologies. It emphasizes rotating
crops, managing pests naturally, diversifying crops and livestock,
and improving the soil with compost additions and animal and
green manures (Fig. 2). Organic farmers use modern equipment,
improved crop varieties, soil and water conservation practices, and
the latest innovations in feeding and handling livestock. Organic
farming systems range from strict closed-cycle systems that go
beyond organic certification guidelines by limiting external inputs
as much as possible to more standard systems that simply follow
organic certification guidelines.
Rudolf Steiners 1924 course on biodynamic agriculture sparked
the evolution of organic agriculture in Europe1. Organic agricul-
ture was established in its own right in the 1930s and 1940s, being
developed in Britain by Lady Eve Balfour and Sir Albert Howard,
in Switzerland by Hans Mueller, in the United States by J. I. Rodale
and in Japan by Masanobu Fukuoka1. By the 1970s, organic foods
had grown in popularity, prompting the first organic certification
standards to be drafted in Europe and the United States, and com-
mencing an ongoing evolution of certifiers that now includes 283
organic certification bodies worldwide operating in 170 countries7.
This proliferation of certifiers reflects both a complex history of
sometimes competing independent standards and the demand for
access to certifiers around the world.
Many farms in both developed and less-developed countries
implement organic practices but are not certified organic. However,
growers are increasingly turning to certified organic farming sys-
tems as a way to provide verification of production methods,
decrease reliance on non-renewable resources, capture high-value
markets and premium prices, and boost farm income. Although
requirements vary slightly between certifying agencies, they pro-
mote soil quality, crop rotations, animal and plant diversity, bio-
logical processes, and animal welfare, while generally prohibiting
irradiation, sewage sludge, genetic engineering, the prophylactic
use of antibiotics, and virtually all synthetic pesticides and fertiliz-
ers. Standards continue to evolve with changing technologies and
socioecological conditions; some requirements are based on scien-
tific evidence, whereas others are driven by ideology.
As most certification standards originated in temperate devel-
oped countries, they are not always applicable in other regions, espe-
cially in less-developed countries. High demand for organic foods
in Europe and North America has resulted in the import of organic
foods from large farms in less-developed countries7. Although pre-
mium prices for exported foods may be beneficial to farmers, the
inaccessibility of many of these foods to local consumers raises

Department of Crop and Soil Sciences, Washington State University, Pullman, Washington 99164, USA.
*e-mail: reganold@wsu.edu

NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants 1

2016 Macmillan Publishers Limited. All rights reserved

REVIEW ARTICLE NATURE PLANTS DOI: 10.1038/NPLANTS.2015.221

80 45
Sales, other Food, fibre and fuel for
Sales of organic food (US$billion)

70 40 human and livestock use

Sales, Europe
and consumption

Organic land area (ha 106)

60 Sales, North America 35
Global sales, no breakdown Physical methods
30
50 Organic land area
25 Crop and
40 animal Biological
20 Biological Disease health Weed controls
30 controls management management
15
20 10
Diverse
10 5 Balanced Pest
crop and
nutrient management
0 0 livestock
supply Temporal and
rotation
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
spatial patterns
Year Species, variety Diverse Hedges,
Supplementary and breed species margins
Biologically
Figure 1 | Annual global market for organic foods and land area of organic lime, organic
active soil
selection balance and other
production. Increasing sales of organic food in nominal billions of US fertilizers and habitat areas
compost
dollars (bars), broken down by contributions from North America (red),
Europe (green) and the rest of the world (orange), and increasing total
global land area under organic production in millions of hectares (blue Figure 2 | Organic management practices. The complex interactions
line)7,9698. European and North American contributions were not available among structural factors and tactical management strategies on a
for the years 1998 through 2000, shown in grey. diversified organic farm producing food, fibre and fuel for human and
livestock use and consumption. Structural factors, represented by
questions about food security and social equity. Participatory guar- circles, are the foundation of organic management, with diverse crop and
antee systems, which rely on local stakeholder verification, have livestock rotations at the centre. Tactical management decisions are used
emerged as a more locally focused alternative to traditional certifi- to supplement the structural factors and include the use of: biological
cation7, and could lead to the development of more locally relevant controls; supplementary lime, organic fertilizers and compost; hedges,
visions for the production and consumption of organic foods. margins and other habitat areas; species, variety and breed selection;
temporal and spatial patterns; and physical weed management. Figure
Sustainability of organic agriculture adapted from ref.99, 2001 Elsevier.
About 38% of Earths land cover is occupied by agriculture11.
Although agriculture provides growing supplies of food and other and wheat, which yield 28 and 27% less, respectively 17. Another
products, it is a major contributor to greenhouse gases, biodiversity meta-analysis found fruits, soybeans and oilseed to be the highest
loss, agrochemical pollution and soil degradation1214. Most of these yielding organic crops, and wheat and vegetables the lowest, yield-
environmental consequences come from arable land, which com- ing 37 and 33% less than conventional systems respectively 18. In
prises around 12% of the land cover 11. The challenge of feeding a cases where organic crop rotations depend on green manure crops,
growing population expected to reach 9to 10 billion people by 2050 food production over the whole rotation may be lower than one-to-
while protecting the environment is daunting. Adopting truly sus- one crop yield comparisons suggest 17.
tainable farming systems on a wide scale is our best opportunity for Although meta-analysis is a great tool that can describe broad
meeting this grand challenge and ensuring future food and ecosys- patterns not immediately visible in primary field research19,20, it
tem security. Concerns about the unsustainability of conventional must also be treated with caution, because no single farming sys-
agriculture have promoted interest in other farming systems, such tem or practice works best everywhere. Still, these studies1519 give
as organic, integrated and conservation agriculture810. strength to the argument that adoption of organic agriculture
According to a US National Academy of Sciences report 10, any under agroecological conditions where it performs best may close
farm, be it organic or conventional, can only be deemed sustain- the yield gap between organic and conventional systems. Under
able if it produces adequate amounts of high-quality food, enhances severe drought conditions, which are expected to increase with
the natural-resource base and environment, is financially viable, climate change in many areas, organically managed farms have
and contributes to the wellbeing of farmers and their communi- frequently been shown to produce higher yields than their conven-
ties. With the rise of organic farming in the past two decades, hun- tional counterparts21,22, due to the higher water-holding capacity of
dreds of research studies comparing different aspects of organic and organically farmed soils23. In addition, improvements in manage-
conventional farming systems have been published. This section ment techniques and crop varieties for organic systems may also
focuses on assessing such comparison studies across these four sus- close this yield gap. For example, direct selection of wheat cultivars
tainability areas. in organic systems has resulted in improved yields in organic sys-
tems when compared with indirect selection of wheat cultivars in
Production. Production includes crop and animal yield and their conventional systems24.
quality. Numerous individual studies have compared yield differ- Whereas organic systems yield less food, organic foods have
ences between organic and conventional systems. These data have significantly less to no synthetic pesticide residues compared with
been synthesized in several meta-analyses or reviews; accord- conventionally produced foods2528. Studies have also found that
ing to these studies, yield averages are 8 to 25% lower in organic children who eat conventionally produced foods have significantly
systems1519. However, with certain crops, growing conditions and higher levels of organophosphate pesticide metabolites in their
management practices, organic systems come closer to matching urine than children who eat organically produced foods29,30. In 2012,
conventional systems in terms of yields. According to one such the American Academy of Pediatrics reported that an organic diet
synthesis study, the best yielding organically grown crops or crop reduces childrens exposure to pesticides, and provided resources
groups are rice, soybeans, corn and grass-clover, which yield 6 to for parents seeking guidance on which foods tend to have the high-
11% less than conventional systems; the lowest yielding are fruits est pesticide residues31. Although these data show that organic foods

2 NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants

2016 Macmillan Publishers Limited. All rights reserved

NATURE PLANTS DOI: 10.1038/NPLANTS.2015.221 REVIEW ARTICLE
a b
N added in N fixed N loss through Volatile N loss N added in N fixed N loss through Volatile N loss
manure and by lightning denitrification during senescence manure and by lightning denitrification during senescence
amendments Atmospheric amendments Atmospheric
Fertilizer N Fertilizer N
addition N deposition Biological addition N deposition Biological
Gaseous N fixation Gaseous N fixation
N loss N loss

Mineralization Mineralization
Soil Soil Soil Soil
Crop N Crop N
organic N mineral N organic N mineral N
Immobilization Crop N N export Immobilization Crop N N export
uptake in harvest uptake in harvest

Crop residue N return

Crop residue N return

Erosion N loss N runoff N leaching Erosion N loss N runoff N leaching

Figure 3 | Hypothetical nitrogen stocks and flows of two contrasting cropping systems. a,b, Cropping systems relying mainly on mineral nitrogen inputs
(a) have relatively higher nitrogen losses to air and water than cropping systems with emphasis on biological N fixation, manure and other organic matter
amendments, cover crops and perennial crops, and low reliance on mineral N fertilizer, such as organic and integrated systems (b). The width of the
arrows is relative to the size of the nitrogen flux; boxes representing nitrogen stocks are not scaled to the pool size. Figure adapted from ref.100, 2015
The National Academies100. Arrows represent nitrogen inputs (green), losses (orange) and transformations (blue).

may present some clear advantages when it comes to synthetic pes-
ticide residues, the human health impacts of pesticide exposure
from food are not clear 26, and organically certified pesticides need
to be better identified and taken into account 28.
At least 15 reviews or meta-analyses26,27,3244 of the scientific
literature comparing the nutrition of organic and conventional
foods have been published in the past 15 years. Twelve of these
studies27,3234,3639,4144 found some evidence of organic food being
more nutritious (for instance, having higher concentrations of
vitamin C, total antioxidants and total omega-3 fatty acids, and
higher omega-3to -6 ratios). Whether or not these are nutritionally
meaningful differences continues to be debated26,43. The other three
studies26,35,40 concluded that there were no consistent nutritional dif-
ferences between organic and conventional foods. However, one of
the three studies found that conventional chicken and pork had a
33% higher risk for contamination with antibiotic-resistant bacteria
compared with organic alternatives26.
Environment. Reviews and meta-analyses generally support the
perception that organic farming systems are more environmentally
friendly than conventional farming systems4558. For example, such
aggregate studies have found that organic farming systems con-
sistently have greater soil carbon levels, better soil quality and less
soil erosion compared with conventional systems4551. In addition,
organic farms generally have more plant diversity, greater faunal
diversity (insects, soil fauna and microbes, birds) and often more
habitat and landscape diversity 4655. Most functional groups, such as
herbivores, pollinators, predators and producers (plants), are more
diverse in organic farming systems5153. Moreover, in a study cov-
ering eight western and eastern European countries, insecticides
and fungicides had consistently negative effects on biodiversity,
with insecticides also reducing the biological control potential in
farming systems56.
As organic agriculture uses virtually no synthetic pesticides,
there is little to no risk of synthetic pesticide pollution of ground
and surface waters46. With respect to nitrate and phosphorous leach-
ing and greenhouse gas emissions, organic farming systems score
better than conventional farming when expressed per unit produc-
tion area46,49,51,57,58; however, given the lower land-use efficiency of
organic farming in developed countries, this positive effect is less
pronounced and in some cases reversed when expressed per unit
product 49,57,58. In a meta-analysis of environmental quality param-
eters, organic farms were found to have lower nitrate leaching,
nitrous oxide emissions and ammonia emissions per unit of field
area, but higher leaching and emissions per unit product 48. Severe
degradation of freshwater and marine ecosystems around the world
is linked to excessive use of nitrogen and phosphorous fertiliz-
ers12,59, leading to eutrophication of freshwater and the production
of hypoxic zones in coastal waters. Lower nutrient pollution from
organic compared with conventional systems can be illustrated by
differences in their nitrogen cycling and losses (Fig.3).
Organic systems are usually more energy efficient than their
conventional counterparts4648,51,54,58. For example, in Germany, Italy,
Sweden and Switzerland, organic farms were found to use signifi-
cantly less energy on a per-hectare basis than their conventional
counterparts, and 70% of organic farms and 30% of conventional
farms had significantly lower energy consumption per unit of out-
put 45. The generally lower energy use4648,54 and higher soil organic
matter 4549 of organic systems make them ideal blueprints for devel-
oping methods to limit fossil fuel emissions and build soil carbon
stores, important tools in addressing climate change.
Economics. Whether organic agriculture can continue to expand
globally will primarily be determined by its financial performance
compared with conventional agriculture17,60. The main factors that
determine the profitability of organic agriculture include crop yields,
labour and total costs, price premiums for organic products, the
potential for reduced income during the organic transition period
(usually three years), and potential cost savings from the reduced
reliance on non-renewable resources and purchased inputs61.
To the best of our knowledge, only one meta-analysis has ana-
lysed the financial performance of organic and conventional agri-
culture20. The analysis combines findings from 40years of studies
covering 55 crops grown on five continents. When actual price
premiums (higher prices awarded to organic foods) were included,
organic agriculture proved significantly more profitable (22to 35%
greater net present values) and had higher benefit/cost ratios (20to
24%) than conventional agriculture. When organic premiums
were taken away, net present values (27to 23%) net returns
accounting for the time value of money and benefit/cost ratios
(8 to 7%) of organic agriculture were significantly lower than
conventional agriculture20.

NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants 3

2016 Macmillan Publishers Limited. All rights reserved

REVIEW ARTICLE NATURE PLANTS DOI: 10.1038/NPLANTS.2015.221

Conventional Organic

Yield Soil quality Yield Soil quality

Nutritional Minimize Nutritional Minimize

quality energy quality energy
use use

Minimize Minimize
pesticide Biodiversity pesticide Biodiversity
residues residues

Reduce Reduce
worker Minimize worker Minimize
exposure to water exposure to water
pesticides pollution pesticides pollution

Employment Employment
Profitability Profitability
of workers of workers

Ecosystem Total costs Ecosystem Total costs

services services

Figure 4 | Assessment of organic farming relative to conventional farming in the four major areas of sustainability. Lengths of the 12 flower petals
are qualitatively based on the studies discussed in this Review1523,2529,3256,58,6274 and indicate the level of performance of specific sustainability metrics
relative to the four circles representing 25, 50, 75 and 100%. Orange petals represent areas of production; blue petals represent areas of environmental
sustainability; red petals represent areas of economic sustainability; green petals represent areas of wellbeing. The lengths of the petals illustrate that
organic farming systems better balance the four areas of sustainability.

Although price premiums were 29 to 32%, breakeven premi-
ums necessary for organic profits to match conventional profits
were only 5to 7%, even with organic yields being 10to 18% lower.
The size of organic premiums awarded, and the difference between
organic premiums and breakeven premiums, were consistent dur-
ing the 40-year study period. The fact that organic premiums were
significantly higher than breakeven premiums suggests that organic
agriculture can continue to expand even if premiums decline. The
study also found that total costs were not significantly different, but
labour costs were significantly (7to 13%) higher with organic farm-
ing practices20. Although one of the successes of conventional agri-
culture has been its ability to produce more with less labour, some
have found the extra labour of organic agriculture to be beneficial
in providing rural employment and development opportunities62,63.
Few economic studies have accounted for negative externalities
(such as environmental costs) or positive externalities (such as eco-
system services), with associated monetary values, in organic and
conventional comparison studies. Putting a price on the negative
externalities caused by farming, such as soil erosion or nitrate leach-
ing into groundwater, would make organic agriculture even more
profitable, given that its environmental impact is less than that of
conventional agriculture4558. Indeed, it has been estimated that a
switch to organic production would lower the external costs of agri-
cultural production in the United Kingdom by 75%, from 1,514
million yr1 to 385 million yr1 (ref.64).
A number of studies (for example, refs 65,66) have compared
ecosystem services in organic and conventional farming systems.
A few of these studies have accounted for the monetary value of
ecosystem services; these studies generally show that conventional
practices decrease the ability of farms to provide some economically
significant ecosystem services relative to organic practices6769. For
example, in a study comparing 14 organic arable fields with 15 con-
ventional ones in New Zealand70, the total economic value of three
ecosystem services (biological pest control, soil formation and the
mineralization of plant nutrients) in the organic fields was signifi-
cantly greater at US$232ha1yr1 compared with the conventional
fields at US$146ha1yr1. Factoring in such differences in economic
comparison studies would probably make up for price premiums
awarded to organic products. Price premiums and European sub-
sidies for organic farms are often justified on the grounds that they
compensate farmers for providing ecosystem services or avoiding
damage to the environment.
Wellbeing. How well organic, conventional and other farming
systems are performing in areas such as social equity (for instance,
issues of gender, race, ethnicity and class) and quality of life for farm
families and communities remains unclear due to limited research.
Available data indicate that both organic and conventional farming
systems need to make significant progress to meet social sustain-
ability goals10. However, organic farming has been shown to have
some sociocultural strengths, such as positive shifts in community
economic development, increased social interactions between farm-
ers and consumers71,72, and greater employment of farm workers and
cooperation among farmers62,63.
Although organic farming often requires additional manual
work on the farm, it reduces the exposure of farm workers to pesti-
cides and other chemicals. Such exposure can be particularly prob-
lematic in less-developed countries, where illnesses and death have
resulted from occupational and accidental exposure (due in part
to the fact that it is impractical and expensive for workers to use
safety equipment)73,74.
Organic certification programmes have adopted social wellbe-
ing goals. Guidelines of the International Federation of Organic
Agriculture Movements (IFOAM) stipulate that organic farmers
should be able to support themselves and other workers with fair
incomes, while maintaining safe and dignified working conditions75.
Furthermore, organically certified animals must be raised humanely
under conditions that allow for the expression of their natural behav-
iours and needs75. For example, European Union, US and Japanese
rules on organic production require livestock to have access to open
air or grazing whenever possible, and that sick animals be treated as
needed, even with the use of antibiotics if required7678.
Organic farming can improve food security by diversifying
on-farm crop and livestock operations, which diversifies income
sources and improves variety in diets79. Organic farming necessi-
tates diverse crop and livestock rotations, encourages the integration

4 NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants

2016 Macmillan Publishers Limited. All rights reserved

NATURE PLANTS DOI: 10.1038/NPLANTS.2015.221
of multiple farm enterprises and encourages the use of leguminous
crops for biological nitrogen fixation. By growing a higher diversity
of more nutrient-rich (such as vegetables) and more protein-rich
(such as legumes and meats) foods, whether for export or subsist-
ence, a farmer has access to at least a portion of these foods. For
example, following 840 small organic and non-organic farms in the
Philippines, researchers found the increase in vegetable and protein
consumption from 2000to 2007to be two to three times greater for
the more diversified organic farmers than conventional farmers80.
Balancing sustainability metrics. Some argue that significantly
scaling-up organic land area may increase nitrogen and other nutri-
ent limitations on yields17, and question whether the greater land
area required by organic agriculture to maintain yields counteracts
its environmental gains2,3. Probably the biggest criticism of organic
agriculture is its lower yields compared with conventional agricul-
ture4,5, a particularly salient challenge given the task of feeding a
growing world population without further agricultural expansion13.
Conversely, some contend that the environmental advantages of
organic agriculture far outweigh the lower yields, and that increasing
research and breeding resources for organic systems would reduce
the yield gap16,17,24,81. Others suggest that multifunctional farming
systems, such as organic, coupled with more plant-based diets and
reduced food waste, are necessary elements of a more sustainable
food system16,54,65. Sometimes excluded from these arguments is the
fact that we already produce adequate kilocalories of food to more
than feed the world but do not provide adequate access to all indi-
viduals82. Globally, 1.9 billion adults are overweight and of these 600
million are obese83, while 793 million people are undernourished
and more than 28% of children under the age of fiveare stunted due
to malnourishment 82,84.
Debates aside, although yield is an important sustainability
metric, the issue is more complicated than kilograms of food per
hectare. Mainstream conventional farming systems have provided
growing supplies of food and other products but often at the expense
of the other three sustainability goals. Environmental degradation,
public health problems, loss of crop variety and genetic biodiversity,
and severe impacts on ecosystem services have not only accompa-
nied conventional farming systems but have often extended well
beyond their field boundaries. Such negative externalities are not
accounted for.
The performance of organic farming systems in the context of
sustainability metrics indicates that they better balance multiple
sustainability goals than their conventional counterparts (Fig. 4).
Based on present evidence, we argue that although organic farm-
ing systems produce lower yields compared with conventional agri-
culture, they are more profitable and environmentally friendly, and
deliver equal or more nutritious foods with less to no pesticide resi-
dues. In addition, initial evidence indicates that organic agriculture
is better at enhancing the delivery of ecosystem services, other than
yield, as well as some social sustainability benefits. Importantly, the
body of research studies has been heavily biased towards developed
countries, whereas studies in the less-developed world, especially in
tropical and subtropical climates, need to be greatly increased.
With only 1% of global agricultural land in organic production7,
and with its multiple sustainability benefits, organic agriculture can
contribute a larger share in feeding the world. Yet, significant barri-
ers to farmers adopting organic practices remain in both developed
and less-developed countries.
Barriers and policies
Obstacles to farmers adopting organic agriculture include power-
ful vested interests and existing policies, a lack of information and
knowledge, weak infrastructure and other economic challenges,
and misperceptions and cultural biases (Fig.5). Global and national
agribusiness corporations, agrochemical industries, commodity
NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants
2016 Macmillan Publishers Limited. All rights reserved
REVIEW ARTICLE
groups and food companies have a strong vested interest in the
conventional agroindustrial model, command ever-greater market
power in the food system and have heavily influenced public policy
to favour this model10,85. The consolidation of industries, the con-
centration of market power, and many past and current agricultural
policies have led to decreased agricultural diversity 10 and have dis-
incentivized agricultural innovation81.
Considerably less public and private funding has been put
towards research and development for organic systems than towards
conventional systems worldwide; this has resulted in a lack of crop
and livestock breeding for organic farming conditions and a dearth
of knowledge and information resources supporting organic farm-
ers17,19. Historically, public funding for research on organic systems
has been higher in Europe than in the United States7. Moreover,
research on organic agriculture in less-developed countries rep-
resents only a small fraction of the overall scientific literature on
the topic17,19,52.
Some farmers face infrastructure and economic barriers, which
include certification costs and access to markets, loans and insur-
ance. Many areas, especially rural regions and less-developed coun-
tries, lack access to additional labour, markets for organic foods,
infrastructure for storage and distribution, or appropriate certifi-
cation requirements86,87. Finally, strong cultural biases against the
connotations of organic agriculture, and conventional mindsets
held by some individuals and organizations, limit the spread of
organic practices86,87.
With these obstacles in mind, governments should focus on
creating an enabling environment for the development and adop-
tion of not just organic but also other innovative and more sustain-
able farming systems88. These efforts must be targeted at improving
agricultural performance in all four areas of sustainability and
will require a diversity of knowledge-based, legal and financial
policy instruments89.
Knowledge-based policy instruments are needed to create an
enabling environment for agricultural innovation, education and
outreach. Specifically, policy instruments must: ensure farmer
and scientist engagement in research and development decision-
making; improve farmer knowledge and capacity through effective
extension and outreach infrastructure, such as the use of farmer
field schools and communication technologies; and enhance wom-
ens educational and leadership opportunities90.
Legal instruments must play a stronger role in ensuring open and
competitive markets, limiting commercial influence in government
and increasing transparency in the food production system. In addi-
tion, they are needed to reduce food waste, to improve the security
of land tenure for farmers, and to develop national targets for trade
policies that promote food and ecosystem security.
Financial instruments are needed to give monetary value to the
externalities that arise from agricultural practices and to empower
farmers through access to capital, infrastructure and competitive
markets88. In developed countries, direct and indirect crop subsidies
and biofuel incentives should be replaced by targeted agro-environ-
mental incentives, such as payments for biodiversity protection
and soil conservation. Some policy organizations have found that
raising the costs of fossil fuels, irrigation water and other limited
resources strongly encourages more efficient farming systems91. In
less-developed countries, targeted input subsidies and investment in
rural infrastructure are key financial instruments. For example, sub-
sidizing organic nutrient inputs alongside mineral fertilizer inputs
for the poorest farmers can be an effective strategy for increasing
yields and building soils92.
Beyond organic
More than 40years after Earl Butzs comment, we are in a new era
of agriculture, as reflected in the words of current US Secretary
of Agriculture Tom Vilsack: Organic agriculture is one of the
5
REVIEW ARTICLE NATURE PLANTS DOI: 10.1038/NPLANTS.2015.221

to sustainable agricult
riers ure
Bar
stin g policies Lack
exi of i
and nfo
rm
ests at
er io
nt

n
i

an
d
ste

dk
l ve

no
Production Environment

wle
erfu

dge
Pow

Legal instruments
Measures to ensure open and
Financial instruments Sustainable competitive markets

n ge s
Increased farmer access to capital agriculture Reduction of commercial influence

c c h a ll e
Elimination of crop subsidies and in government
biofuel incentives Development of national targets and trade
M is p

Use of agro-environmental incentives policies for food and ecosystem security

omi
Increased costs of scarce resources Wellbeing Economics Enhanced security of land tenure
erc

co n
Targeted input subsidies for
ep

resource-poor farmers

re
ti o

he
Investment in rural infrastructure
ns

ot
d
an

cu

nd
a
l tu re
ral tu
b uc
ias e
s fra str
in
Weak

Knowledge-based instruments
Public investment in agroecological research
Farmer engagement in research and development
Outreach to improve farmer knowledge and capacity
Enhancement of women's educational and leadership opportunities
Increased transparency throughout the food chain

Figure 5 | Policy instruments for overcoming barriers to farmers adopting more sustainable farming systems. For any farm to be sustainable, it must
meet four goals, shown in the centre: (1) produce adequate amounts of high-quality food (production); (2) enhance the natural-resource base and
environment (environment); (3) be financially viable (economics); and (4) contribute to the wellbeing of farmers and their communities (wellbeing)10.
Despite the appeal of a sustainable agriculture philosophy, the task for farmers to achieve agricultural sustainability is challenging. Just because a farm
is organic does not mean that it is sustainable. However, research shows that organic farming systems better balance the four sustainability goals than
their conventional counterparts and are more likely to achieve agricultural sustainability (overlapping area). Yet, significant barriers to adopting organic
agriculture exist, including powerful vested interests and existing policies, lack of information and knowledge, weak infrastructure and other economic
challenges, and misperceptions and cultural biases. In fact, many of these same barriers exist for other innovative systems, such as agroforestry,
conservation agriculture, integrated farming and mixed croplivestock systems. A diversity of policy instruments is needed to overcome these barriers, and
can be categorized as financial, legal and knowledge-based instruments. Examples of these instruments are shown in the figure.

fastest growing segments of American agriculture and helps farmers
receive a higher price for their product as they strive to meet grow-
ing consumer demand93. Moreover, organic agriculture has been
able to provide jobs, be profitable, benefit the soil and environment,
and support social interactions between farmers and consumers.
Although organic agriculture has an untapped potential role in
global food and ecosystem security, no one farming system alone
will safely feed the planet. Rather, a blend of organic and other inno-
vative farming systems, including agroforestry, integrated farming,
conservation agriculture, mixed crop and livestock, and still undis-
covered systems, will be needed for future global food and ecosys-
tem security. For example, integrated farming systems that blend
mostly organic with some conventional practices have been shown
to be more sustainable than conventional farming systems94,95 and
are likely to play a central role. Achieving global food and ecosys-
tem security requires more than just achieving sustainable farming
systems worldwide. We need to reduce food waste, improve food
distribution and access, stabilize the human population, eliminate
the conversion of food into fuel, and change consumption patterns
towards a more plant-based diet.
Equal adherence to all four sustainability goals of production,
environment, economics and social wellbeing does not limit but
encourages farmers and researchers to innovate. The challenge fac-
ing policymakers is to create an enabling environment for scaling-up
organic and other innovative farming systems to move towards truly
sustainable production systems. This is no small task, but the con-
sequences for food and ecosystem security could not be bigger. To
make this happen will require mobilizing the full arsenal of effective
policies, scientific and socioeconomic advances, farmer ingenuity
and public engagement.
Received 7 September 2015; accepted 17 December 2015;
published online 3 February 2016
References
1. Lockeretz, W. in Organic Farming: An International History (ed. Lockeretz, W.)
18 (CABI, 2007).
2. Trewavas, A. Urban myths of organic farming. Nature 410, 409410 (2001).
3. Emsley, J. Going one better than nature? Nature 410, 633634 (2001).
4. Kirchmann, H. & Thorvaldsson, G. Challenging targets for future agriculture.
Eur. J.Agron. 12, 145161 (2000).
5. Connor, D.J. & Mnguez, M.I. Evolution not revolution of farming systems
will best feed and green the world. Glob. Food Secur. 1, 106113 (2012).
6. Pickett, J.A. Food security: intensification of agriculture is essential, for which
current tools must be defended and new sustainable technologies invented.
Food Energ. Secur. 2, 167173 (2013).
7. Willer, H. & Lernoud, J. (eds) The World of Organic Agriculture: Statistics and
Emerging Trends 2015 (FiBL-IFOAM, 2015).
8. International Assessment of Agricultural Science and Technology for
Development Agriculture at a Crossroads: Global Report (Island, 2009).

6 NATURE PLANTS | VOL 2 | FEBRUARY 2016 | www.nature.com/natureplants

2016 Macmillan Publishers Limited. All rights reserved

 The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0007-070X.htm

FJ
18,2
Customer value perception of
organic food: cultural differences
and cross-national segments
Barbara Seegebarth
eived 2 July 2015
Institute of Marketing, Technische Universitt Braunschweig,
ised 4 November 2015 Braunschweig, Germany
epted 4 November 2015
Stefan Henrik Behrens, Christiane Klarmann and Nadine Hennigs
Marketing and Management, Leibniz University of Hannover,
Hannover, Germany, and
Lisa Luebbehusen Scribner
Cameron School of Business, University of North Carolina Wilmington,
Wilmington, North Carolina, USA

Abstract
Purpose Due to consumer concerns about food-related diseases and an increase in the use of
genetically modified food, more and more green consumers integrate environmental considerations
into daily purchases, asking for healthier, safer and higher quality food. Marketing managers still face
the challenge of broadening the understanding of how and why consumers purchase organic food.
Specifically, a deeper understanding of the value dimensions consumers perceive in the context of
organic food products is required to develop and implement successful management strategies which
might transfer positive consumer perceptions to actual buying behavior and satisfaction. The paper
aims to discuss these issues.
Design/methodology/approach Drawing on theoretical and empirical insights on organic food
consumption in different markets, this research investigates antecedents of organic food products and
differences regarding the relative importance of the value-based drivers across two Western nations.
Findings The results from survey data indicate significant differences in the value perceptions,
especially the functional and individual value perceptions, and recommendation behavior related to
organic food for consumers from the USA and Germany. In addition, the segmentation approach
provides evidence for consumer segments that cross-national borders: the convinced opponents, the
silent/private consumers, the prestige-seekers and the passionate evangelists.
Originality/value Consequently, instead of a country-based segmentation approach, marketers
should emphasize the different types of consumers across national borders in order to address the
differences in customer value perception in the organic food market.
Keywords Food marketing, Customer perceived value, Cross-national market segmentation,
Organic food consumption
Paper type Research paper

Introduction
The organic food market has become a rapidly growing sector in developed
agricultural economies (Lockie et al., 2002; Chen, 2007). Particularly in Europe, America,
Latin America and Australia, more and more green consumers integrate
environmental considerations into daily purchases, asking for healthier, safer and
sh Food Journal higher quality foods (Lockie et al., 2002; Chen, 2007; Siderera et al., 2005). This
118 No. 2, 2016
96-411
phenomenon is largely due to consumer concerns about food-related diseases and an
merald Group Publishing Limited
070X
increase in the use of genetically modified food (Chen, 2009; Hamzaoui Essoussi and
10.1108/BFJ-07-2015-0235 Zahaf, 2008; Vindigni et al., 2002). Being suspicious of conventionally produced
products, the new consumer orientation leads to a high demand for organic products so Customer
that even discounters offer organic products in large quantities at exceptionally low value
prices (Stolz et al., 2010). Consumers often associate organic food with an environment
friendly agricultural production as well as certain intrinsic quality and safety
perception of
characteristics (Brennan and Kuri, 2002; Vindigni et al., 2002). Reasoning that the term organic food
organic refers to the production process and not only to the product itself (Zakowska-
Biemans, 2011), organic food is a produce of organic farming, which is a type of
farming that sets very strict limits on the amount of artificial synthetic inputs allowed
[]. These inputs include the ones used in production (fertilisers, pesticides, herbicides,
antibiotics, etc.) and processing (food additives, including artificial flavourings,
preservatives and colourings) (First and Brozina, 2009, p. 186). However, from a
rational point of view, organic food may not necessarily be healthier
or tastier than its conventionally produced counterpart (Torjusen et al. 1999 quoted
by Chen, 2009).
During the last decades and with regard to the immense future potential of the
market for environmental and ethical goods, the interest of both marketing researchers
and practitioners in ethical consumerism in general and organic food consumption in
particular has increased significantly. With special focus on consumer demand for
organic food, existing studies have investigated the role of risk and value perceptions,
personal, health and environmentally friendly behavior in organic food purchases, as
well as demographic characteristics of organic food consumers across nations
(e.g. Chen, 2009; Lockie et al., 2002; Magnusson et al., 2003; Saba and Messina, 2003).
The consumption of organic food has been shown to provide the consumer with deeper
values and has become one of the fastest growing food trends not only in the EU but
also in the USA (Nie and Zepeda, 2011; Thgersen, 2010). As a result, marketing
managers have realized the importance of customer ethics and values and how meeting
ethical demands is critical if they wish to gain a competitive advantage (Browne et al.,
2000). However, in the context of environmental and ethical purchases, empirical
studies often reveal a substantial gap between positive consumer attitudes, buying
intentions and effective buying behavior (e.g. Cowe and Williams, 2000; De Pelsmacker
et al., 2005; King and Bruner, 2000; Strong, 1996). Therefore, the question remains of
how to meet the consumers expectations toward organic food products and support
the transfer of positive perception to actual buying behavior: what motivates
consumers to buy organic food products? What are the perceived value dimensions
related to organic food consumption? Does the organic character of a product increase
the willingness to pay a price premium or help to build product loyalty?
With regard to consumption values that directly explain why consumers choose to
buy or avoid particular products (Sheth et al., 1991), the aim of the present study is to
examine the value of organic food consumption as perceived by consumers in
important markets such as the USA and Germany. Specifically, a deeper understanding
of the value dimensions consumers perceive in the context of organic food products is
provided as a general basis to develop and implement successful management
strategies which might transfer positive consumer perceptions to satisfaction and
actual buying behavior.
Incorporating relevant theoretical and empirical findings, this study contributes to
the field of inquiry on organic food consumption by:

(1) investigating the antecedents leading to the consumption of organic food

products;
FJ (2) examining significant differences regarding the relative importance of the
18,2 value-based drivers across US and German consumers; and
(3) identifying segments of organic food consumers that cross-national borders.
Thus, the present research attempts to broaden the understanding and findings of
former studies in the domain of organic food with an empirical investigation into
cross-national customer value comparisons. The paper is structured as follows: first,
the conceptual model and related hypotheses are presented based on existing research
in organic food marketing and customer perceived value. Second, the methodology and
results of our empirical study, including principal component analysis, analysis of
variance (ANOVA) and cluster analysis based on US and German samples are
described. Third, the results are discussed with reference to managerial implications
and further research steps.

Conceptualization and hypotheses development

In an attempt to examine a customers perceived preference for and evaluation of a
certain organic food product, the construct of customer perceived value is of particular
importance. Understood as a consumers overall assessment of the utility of a product
based on perceptions of what is received and what is given (Zeithaml, 1988, p. 14) and
based on an interactive relativistic consumption preference experience (Holbrook, 1994,
p. 27), customer perceived value creates a trade-off between product-related benefits and
sacrifices in the perspective of current as well as potential customers in different phases
of the purchase process (Woodruff, 1997; Sweeney and Soutar, 2001). Referring to our
conceptual model, following the value dimensions conducted by Sweeney and Soutar
(2001), we assume that the consumers individual perceived value addresses the
experiences, feelings and emotions a certain brand or product provides to the consumer
(Sheth et al., 1991; Smith and Colgate, 2007; Sweeney and Soutar, 2001). We relate to this
concept, as customer value is well established in the marketing literature and represents
a key determinant for consumer behavior (Papista and Krystallis, 2013). As shown in
Figure 1 and discussed in the following paragraphs, customer perceived value of organic
food is purported to have a positive impact on behavioral outcomes such as purchase
intention, willingness to recommend and to pay a premium price.

Customer Perceived Value

Financial Functional Individual Social

Value Value Value Value

Purchase Willingness Recommendation

Intention To Pay Behavior
gure 1.
e conceptual
del Related Outcomes
Antecedents: customer value perception of organic food Customer
The question of what really adds value in consumers perceptions of organic food value
products is conceptualized in this paper in accordance with the insights of Sweeney and
Soutar (2001), Smith and Colgate (2007) and Wiedmann et al. (2007, 2009) along the
perception of
following four dimensions: the financial value, the functional value, the individual value organic food
and the social value.
Financial value perception. Reasoning that the acceptance of food is significantly
influenced by the perceived value of money (Brennan and Kuri, 2002), theoretical and
empirical evidence suggests that the price of a certain product primarily depends on
production costs, quality, reputation and objective, easily observable product
characteristics (Chang and Wildt, 1994; Edwards and Spawton, 1990; Zeithaml,
1988). Taking consumers associations with organic food products into account
(i.e. chemical free, healthy and environmentally friendly production) it can be assumed
that the price functions as an indicator for high quality (Cicia et al., 2002) and reflects
consumers desires for doing something good for themselves (being worth it).
Moreover, some consumers even highlight a change in their consumption behavior:
as an example, due to the expensive nature of organic food products, some consumers
only buy what they are going to eat (VegaZamora et al., 2014).
Functional value perception. As aforementioned, organic food products are often
associated with an environmentally friendly agricultural production process as well as
certain intrinsic quality and safety characteristics (Brennan and Kuri, 2002; Vindigni
et al., 2002; Zakowska-Biemans, 2011). Consequently, studies on food consumption
often define quality as a superior but difficult to recognize reason for purchasing a
certain food product (Grunert, 2005).
Individual value perception. With special focus on the individual dimension of value
perception, a consumers subjective understanding of the organic food concept focusses
on the assumption that food has a direct influence on consumers well-being and
combines the organic-related associations with consumers increasing environmental
concerns. Representing a certain lifestyle (Davies et al., 1995; Grunert et al., 2001), and
influencing an individuals value system (Schifferstein and Oude Ophuis, 1998), organic
food products are being chosen more regularly by more consumers in recent decades
(Pino et al., 2012; Wier and Calverley, 2002).
Social value perception. In our context, the social benefits of behaving according to
prevailing social norms might encourage consumers to engage in environmentally
friendly behavior (Papista and Krystallis, 2013). Thus, as food is frequently consumed
in public situations, consumers intend to consume food that is positively recognized
within their own social groups and emphasize their social identity (Vindigni et al., 2002).

Related outcomes
Apart from the empirical investigation of the dimensions of customer value perception
in the context of organic food, the analysis of behavioral outcomes are of particular
importance in order to determine if and to what extent positive value perceptions are
translated into actual organic food consumption. Typically, when dealing with ethical
purchases such as fair trade or organic food, results reveal a substantial gap between
consumer attitudes, buying intentions and effective behavior (e.g. De Pelsmacker et al.,
2005; Greenwald and Banaji, 1995; King and Bruner, 2000; Maison, 2002; Shaw
and Clarke, 1999; Strong, 1996; Ulrich and Sarasin, 1995). Even if studies suggest
that consumers prefer to buy ethical products (e.g. Cryer and Ross, 1997;
FJ Fernandez-Kranz and Merino-Castello, 2005; Mohr et al., 2001), only a niche market
18,2 actually buys them (Cowe and Williams, 2000). To assess if and to what extent positive
attitudes toward organic food products result in actual behavior, for the purposes of
this study and in accordance to existing studies on behavioral loyalty dimensions
(e.g. Zeithaml et al., 1996), three basic components of actual consumer behavior are of
particular interest: consumers purchase intention, recommendation behavior and
willingness to pay a higher price.
Indicating positive customer experiences and signaling that customers perceive
products as favorable (Gupta and Zeithaml, 2006), purchase intentions as an antecedent
of behavioral loyalty creates sales and profits (Chaudhuri and Holbrook, 2001) enhances
cash flows, and secures long-term survival ( Jacoby and Chestnut, 1978). Against the
backdrop of our research context, we suggest that the individuals perceived value is an
important driver for the consumers intention to (re)buy organic food products. Likewise,
loyal customers are willing to engage in positive word-of-mouth communication and
recommend a given brand or product to relevant others (Reichheld and Sasser, 1990;
Zeithaml et al., 1996). Furthermore, the perceived value of a brand or product has been
shown to positively affect a consumers willingness to pay premium prices (Keller, 1993).
Therefore, we suggest that in the context of organic food consumers are willing to
pay a considerably higher price and continue to purchase organic food even with a
price increase.

Country comparison
As it is widely discussed that the market for organic products continues to grow
globally (Chen, 2007; Nie and Zepeda, 2011; Thgersen, 2010), it becomes not only more
important to understand why consumers buy organic foods at an individual level, but
also why consumers across nations buy organic food products for different reasons
(Bartels and Reinders, 2010; First and Brozina, 2009). Drawing from prior findings in
cultural studies and based on our conceptual framework, one main research question
will be analyzed by using a cross-national data set that allows comparisons between US
and German consumers perceptions:
RQ1. Do similar perceptions of organic food exist across countries, or are there
significant differences in perceptions of organic food products across
countries?
Based on this, with reference to Clarks (1990) general hypothesis that is often used in
cross-national studies, we formulate:
H0. Consumers in different countries exhibit the same perceptions of organic food.
H1. Consumers in different countries exhibit differing perceptions of organic food.

Research methodology
To measure the underlying value dimensions of consumers intentions to buy organic
food (Davies et al., 1995) in the USA and Germany, this study employed existing
multi-item measures to draw on new conclusions from an online survey. Related to
organic food, respondents were pleased to evaluate their financial, functional, social
and individual values adapted from Sweeney and Soutar (2001), their willingness to
pay a premium price (Lockie et al., 2002, 2004), their recommendation intention
(Wangenheim and Baon, 2004; Yu and Dean, 2001) and the purchase intention
(Michaelidou and Hassan, 2008; Tarkiainen and Sundqvist, 2005). Besides, some Customer
general statements related to the organic food market have been included with value
reference to health orientation (Tarkiainen and Sundqvist, 2009; Lockie et al., 2002) and
general attitudes to organic products (Chen, 2007).
perception of
A German version of the latent variables was created using back translation organic food
according to equivalent meaning. All items were rated on a five-point Likert scale
(1 strongly disagree; 5 strongly agree) and we operationalized all latent variables
via reflective measurement models. Data were collected at a US and two German
universities. The study sample included 206 US and 240 German respondents. Sample
characteristics are shown in Table I.
Even if the sample used is not representative for the US and Germany, it offers a
balanced set of data for the investigation of customer value perception and related
behavior in the organic food market.

Results and discussion

To examine possible differences and/or similarities across countries within the data
analysis using SPSS 20.0, we first identified the various antecedents and behavioral
outcomes of customer value perception through a factor analysis using the principal
component method with varimax rotation. Based on the factor scores for each country,
we used ANOVA to determine whether there are significant differences across
countries in the psychological and context-related antecedents as well as behavioral
outcomes. In order to assess common method variance (Podsakoff et al., 2003), we used
Harmans (1976) one-factor test. Principal component factor analysis with varimax

Germany USA
Variable n % n %

Age o20 years 6 2.5 2 1.0

20-29 years 209 87.08 185 92.5
30-39 years 25 10.42 9 4.5
40-49 years 2 1.0
50-59 years 2 1.0
Mean age 25.15 years 22.36 years
Gender Male 123 51.25 118 57.28
Female 117 48.75 88 42.72
Living environment Urban 193 80.42 165 80.49
Rural 47 19.58 40 19.51
Education Secondary modern school 1 0.42 1 0.49
Intermediate school 7 2.92 114 55.88
University entrance diploma 122 50.84 57 27.94
University degree 110 45.83 32 15.69
Occupation Full time 47 19.58 15 7.32
Part time 15 6.25 60 29.27
Job training 1 0.42 1 0.49
Student 177 73.75 129 62.93
Income Very low income 47 19.58 61 29.90
Low income 72 30.00 71 34.80
Medium income 78 32.50 58 28.43 Table I.
High income 41 17.08 10 4.90 Demographic profile
Very high income 2 0.83 4 1.96 of the sample
FJ rotation revealed that common method bias presents no problem in our study; the first
18,2 factor accounted for 17 percent of the variance. The results of the measurement of the
constructs, the ANOVA and cluster results are described below.
Measurement of constructs (PCA)
A cross culturally applicable measurement instrument should lead to similar patterns
of reliability, factor loadings and factor structure in samples from different countries.
As Table II shows, the factor structure of our cross-national database largely confirms
our proposed conceptualization.

Cross-national comparison (ANOVA)

ANOVA was used to explore possible cross-national differences between US and
German consumers perceptions. For this purpose, the country was the independent

Items Factor loadings

Value perception
Financial value Cronbachs 0.825
OF is reasonably priced 0.795
OF offers good value for money 0.874
OF is a good product for the price 0.849
OF is economically 0.726
Functional value Cronbachs 0.822
OF has consistent quality 0.749
OF is well made 0.853
OF has an acceptable standard of quality 0.813
OF is of consistent quality 0.822
Individual value Cronbachs 0.890
OF is something that I would enjoy 0.864
OF is enticing to me 0.819
OF is something that I would feel comfortable using 0.768
OF makes me feel good 0.862
OF gives me pleasure 0.876
Social value Cronbachs 0.897
OF helps me to feel accepted by others 0.865
OF improves the way I am perceived 0.914
OF makes a good impression on other people 0.857
OF gives me social approval 0.863
Related outcomes
Purchase intention Cronbachs 0.934
I intend to buy organic food in the near future 0.923
I am going to buy organic food in the future again 0.925
I will continue to buy organic food 0.923
The next time I purchase healthy food I intend for it to be organic food 0.883
Willingness to pay Cronbachs 0.756
I am willing to pay a higher price for organic food than for
non-organic food 0.897
I am willing to pay a lot more for organic food than for non-organic food 0.897
Recommendation behavior Cronbachs 0.912
ble II. I talk positively about organic food 0.896
asurement I recommend organic food to friends 0.946
constructs I encourage friends to buy organic food 0.925
(grouping) variable and the factors that were identified against the backdrop of our Customer
conceptual model were the dependent variables (factor mean scores). As presented in value
Table III our research hypothesis H1 can be confirmed because the participants
differed significantly ( p o 0.01) in the degree to which they agreed with the statements
perception of
related to the value-based dimensions as well as behavioral outcomes. organic food
With reference to the customer value perception of organic food, the results show that
American consumers rate the functional and the individual value components related to
organic food products significantly higher than the respondents from Germany.
More than Germans, consumers in the USA state that they buy organic food because it is
healthier and has better quality than conventional food products. Additionally, American
consumers associate organic food consumption with pleasure, relaxation and fun.
In contrast to this, even if not significant, German consumers place emphasis on the
financial value of organic food. In their opinion, organic food products offer excellent
value for money. In terms of behavioral outcomes, the results reveal significant
differences in the perception of the willingness to pay and recommendation behavior.
While German consumers are more willing to pay a premium price for organic food than
Americans are, US consumers are more active in positive word-of-mouth activities and
recommend organic food products to relevant others.
In sum, with regard to our initial research question and hypotheses, supportive of
H1, the ANOVA results provide evidence to confirm significant cross-national
differences between the antecedents and behavioral outcomes of organic food value as
perceived by consumers from Germany and the USA Hence, the perception and
consumption of organic food products are multifaceted and comprise a combination of
aspects whose importance differs across countries. Nevertheless, what is the optimal
basis on which appropriate marketing strategies should be developed in the organic
food market? Would a country-based segmentation approach cover the differences in
customer value perception or would an approach aimed at different types of consumers
across national borders be preferable? Hence, in the next step, a cluster analysis was
chosen to identify different types of organic food consumers based on the assumption
that common structures in consumer perception and behavior exist cross-nationally.

Common structures across countries (cluster analysis)

To identify different groups of organic food consumers across countries, we followed
the suggestions of Mooi and Sarstedt (2011) by combining hierarchical and k-mean
clustering. Thus, factor means for each respondent were saved and consequently these
scores were used in clustering the respondents into market segments. An initial

USA Germany F Sig

Value perception
Financial value 0.117 0.100 5.257 0.022
Functional value 0.262 0.223 27.451 0.000
Individual value 0.331 0.283 45.933 0.000
Social value 0.043 0.036 0.689 0.407
Related outcomes
Purchase intention 0.006 0.005 0.014 0.907
Willingness to pay 0.100 0.085 3.816 0.051 Table III.
Recommendation behavior 0.375 0.320 60.630 0.000 ANOVA results
FJ hierarchical clustering procedure was employed to obtain the number of possible
18,2 clusters and seed points for a k-means cluster analysis. The respondents were
partitioned using a hierarchical procedure and the variance ration criterion was used to
identify the correct number of clusters. Wards method of minimum variance was
chosen to determine the cluster differences in each stage of combinations and to
maximize homogeneity and heterogeneity within and between clusters. The results
strongly suggested the presence of four clusters which were validated using non-
hierarchical k-means clustering. After the clusters were identified, a discriminant
analysis was used to verify the cluster groupings. Overall, 94.1 percent of the cases
were assigned to the correct groups; thus, the results of the cluster analysis were valid
for the useful classification of consumer subgroups based on their organic food value
perceptions that differed significantly from one another. Thus, as shown in Table IV,
this four-cluster solution produced the most interpretable and stable results.
For the development of a profile for each market segment, more detailed information
was obtained by examining the questionnaire variables that were cross-tabulated by
cluster segment. Comparisons among the five clusters were conducted with respect to a
variety of descriptive variables, including demographic and socioeconomic
characteristics. Based on the variables from which they were derived, the four
clusters can be described as follows:
Cluster 1: the convinced opponents. This cluster constitutes 24.3 percent of the
sample, has a mean age of 24.3 years, and consists of 65.4 percent male and 34.6 percent
female respondents. With regard to nationalities, this group primarily consists of
German consumers (66.4 percent). Compared to the rest of the clusters, the members of
this group showed the lowest ratings for all factors included in our analysis. The
convinced opponents are significantly more likely than the members of other clusters to
state that organic food products are too expensive and of worse quality than
conventional food products. Furthermore, they are convinced that organic fraud is
common. These consumers are not willing to pay a premium price for organic food and
do not consider purchases of organic food products as an alternative. They have no
intention to buy organic food and are absolutely not willing to recommend organic food
products to relevant others. In sum, consumers in this group can be regarded as
unquestionable antagonists of the organic food movement.
Cluster 2: the silent/private consumers. The second cluster forms 26.6 percent of the
sample, has a mean age of 24.1 years, and is comprised of 46.2 percent male and
53.9 percent female respondents. With regard to country composition, this cluster is
composed of 57.3 percent German and 42.7 percent US consumers. The consumers in

Cluster 1 Cluster 2 Cluster 3 Cluster 4 F Sig

Value perception
Financial value 0.876 0.178 0.077 0.983 79.409 0.000
Functional value 0.842 0.206 0.132 0.911 93.044 0.000
Individual value 1.081 0.203 0.048 1.079 179.466 0.000
Social value 0.566 0.843 0.624 0.870 57.466 0.000
Related outcomes
ble IV. Purchase intention 1.199 0.460 0.095 0.993 214.768 0.000
means Willingness to pay 0.988 0.238 0.053 0.984 117.102 0.000
ster results Recommendation behavior 1.103 0.194 0.129 1.261 285.473 0.000
this cluster are more likely than members of Cluster 1 to perceive organic food as a Customer
valuable alternative to conventional food products. In their opinion, health-oriented value
aspects are particularly important; they perceive organic food to be healthier and less
contaminated with chemical additives. They state that they are satisfied consumers of
perception of
organic food and intend to buy organic food products again in the future even if they organic food
have to pay a premium price. Nevertheless, consumers in this group are not likely to
recommend organic food products to others; statements related to the social value
dimension reveal the lowest ratings of all groups. In sum, this group consists of people
who consume organic food for personal reasons and have absolutely no interest in
aspects related to buying to impress others.
Cluster 3: the prestige-seekers. Constituting 28.6 percent of the sample, this cluster
consists of 55.6 percent male and 44.4 percent female respondents with a mean age of
23.7 years. With regard to nationality, this group comprises 47.6 percent US and
52.4 percent German consumers. The consumers in this cluster state that they perceive
organic food to be expensive, but in fashion. Therefore, even if they show only medium
ratings for most of the value components as well as the behavioral outcomes, they
reveal particular interest in the social value dimension. The main reason for prestige-
seekers to buy organic food products is to make a good impression on relevant others.
Consequently, in contrast to Cluster 2, this group is very interested in conspicuous
consumption and wants to be regarded as responsible, progressive consumers.
Cluster 4: the passionate evangelists. The cluster represents 20.5 percent of the sample,
has a mean age of 23.3 years, and comprises 47.8 percent male and 52.2 percent female
respondents. Concerning the nationalities of this group, this group primarily consists of US
consumers (60 percent). Taken as a whole, the members of this cluster place significantly
more emphasis than other consumers on all aspects included in our analysis and exhibit
the highest ratings for all factors. They value the healthiness and quality of organic food
products and are absolutely willing to pay a premium price. Additionally, they state that
organic food consumption makes them happy and they consider themselves as satisfied
and loyal consumers. Moreover, they can be regarded as loyal advocates as evidenced by
the highest ratings for the willingness to recommend organic food to relevant others. All in
all, the passionate evangelists represent an enthusiastic group of loyal consumers who are
absolutely convinced that organic food is the best alternative in the food industry.
In sum, the cluster analysis results provide evidence that even with the ANOVA
showing significant cross-country differences, there are similarities in consumer
perceptions that cross-national borders.

Conclusion
Main contribution
In times when consumers are increasingly concerned about certain risks associated
with conventional or genetically modified food products, environmental and ethical
aspects provide consumers with a feeling of security and well-being. Widely perceived
to be reliable, fair, sustainable and healthy, organic food consumption reflects the
association of responsibility and environmental friendliness to the individual decision-
making process and related buying behavior. As a consequence of the increasing
importance of organic products, a deeper understanding of the key drivers of
consumers organic food perception and consumption behavior is a major challenge for
practitioners and researchers in the domain of food marketing. Following a holistic
perspective, the aim of the present study was to examine the various dimensions of
FJ customer perceived value that lead the customer to choose and consume organic food
18,2 products and to identify differences and similarities regarding the relative importance
of the value-based drivers across US and German consumers. Our empirical results are
supportive of the hypothesized significant differences in the value perceptions of
consumers from the USA and Germany, particularly in terms of the functional and
individual value perception as well as the recommendation behavior related to organic
food. Nevertheless, the cluster analysis provides evidence for consumer segments that
cross-national borders: the convinced opponents, the silent/private consumers, the
prestige-seekers and the passionate evangelists. Consequently, instead of a country-
based segmentation approach, the segmentation based on different types of consumers
across national borders would be preferable to address the differences in customer
value perception in the organic food market.

Next research steps

Nevertheless, the results of our study clearly need to be considered in the light of its
research limitations. Apart from the size of the sample group of respondents, solely
investigating German and US consumers (students) might have an influence on the
results. We recommend examining our concept on the background of representative
samples from each country. At a minimum, a larger international sample may help
discover different customer perceptions within national borders as well as across
borders. Moreover, the study focussed on organic food in general. Differentiating
various product groups might enhance the understanding of organic food consumption
related to the diverse value dimensions. Thus, further research needs to be carried out
in order to extend insights on organic food consumption.

Managerial implications
From a managerial perspective, our study may form the basis of a deeper
understanding of perceived value and related consumer behavior in the context of the
rapidly growing and highly competitive market of organic food. Managers have to be
aware of what influences consumers buying behavior apart from safety issues and
moral judgment. Our results show that organic food producers should include the value
dimensions in marketing and communication strategies to increase sales and overall
profits. Nevertheless, our results show differences in the importance of these
dimensions in a cross-national context and with regard to different consumer segments.
Thus, in view of the US market, organic food producers should focus on the quality
and health-related aspects. We recommend that organic food marketers integrate
information on the principles and the benefits of organic food in their communication
strategies. Regarding signals of high quality of organic food, communication about
trustworthy labels, food certification standards or details about production conditions, the
origin of organic food products or additional information on animal welfare represent
initial communication strategies. When addressing German consumers, the
communication should focus on the financial aspects, as being worth it. However, the
promoted price for high-quality organic food has to be carefully calculated regarding
consumers perceptions that organic food is worth the price in comparison to
conventionally produced food, but it is not perceived as a luxury good (Yiridoe et al., 2005).
With regard to the rising demand for organic food and referring to our results, we suggest
marketers increase the availability of organic foods by setting up new stores or by
increasing the availability of organic food products in supermarkets and/or discounters.