British Journal of Nutrition (2002), 88, Suppl. 2, S123
1079/BJN2002674
q International Life Sciences Institute 2002
Foreword
Increasing knowledge of the relationship between diet
and health leads to new insights into the effects of
food ingredients on physiological functions and health.
These insights generate interest among scientists, health
care providers and consumers, and in turn stimulate the
food industry to match consumers desire for short- and
long-term health benefits through food products that pro-
mote health and well-being and reduce the risk of
chronic diseases. These foods are called functional
foods. The term functional food emerged about a
decade ago in a number of countries, e.g. Japan and
the United States, and was stimulated and supported by
research on the physiological effects of food components
and their consequent health benefits.
Despite the absence of a universally accepted definition
for functional foods, there is growing general agreement
that some foods and beverages have beneficial effects
beyond normal nutrition. Functional foods have a future.
However, the success of functional foods among con-
sumers will depend on their interest and confidence in
these products and will require the accurate communication
of their health benefits as well as a favourable and
DOI: 10.
transparent regulatory framework for approval of the new
products and their associated health claims.
In 1995, the International Life Sciences Institute (ILSI)
took the initiative in organising the first international
symposium on functional foods, East West Perspectives
on Functional Foods, held in Singapore. The 2001 ILSI
symposium discussed in this supplement, Functional
Foods: Scientific and Global Perspectives, built on this
by reviewing the current global status of functional foods
and the scientific basis for biomarkers related to the
enhancement of function and the reduction of disease
risk. The symposium, which took place in Paris, also
discussed the issues of human genetic variability and the
safety of functional foods, as well as communication
requirements from the scientific, consumer and regulatory
points of view. Finally, the conference provided a future
outlook on new trends in functional food science.
Paulus M. Verschuren
Unilever Health Institute
Wim Saris
Maastricht University
British Journal of Nutrition (2002), 88, Suppl. 2, S125S130 DOI: 10.1079/BJN2002675
q International Life Sciences Institute 2002

Summary Report

Functional Foods: Scientific and Global Perspectives

P. M. Verschuren*
Unilever Health Institute, Olivier van Noortlaan 120, NL-3133 AT Vlaardingen, The Netherlands

Key symposium messages
Session I. Global View on Functional Foods
The symposium opened with a review of the functional food
concepts in place in various regions around the world. This
global overview included Asia, Europe, Latin America and
North America. The session touched on differences in culture
and beliefs with respect to foods, functional foods and medi-
cines, and their impact on health and disease.
In Asia, where functional foods have been regarded as
an integral part of the culture for many years, there is a
firm belief that foods and medicine come from the same
source and serve the same purpose. However, scientists
and regulators have only recently begun to agree that the
functionality of functional foods should be found in
whole foods rather than in their individual components.
In Japan, research on functional foods started in the early
1980s. In 1991, a specific regulatory framework concern-
ing Food for Specific Health Use (called FOSHU) was
introduced, which made it possible to make limited
health claims after receiving approval from the Ministry
of Health. The key to success for food manufacturers is
to develop products that are accepted by consumers and
are consistent with the consumers understanding and
appreciation of functional foods within the existing culture.
Because the state of a persons health may range from
optimal to a state of disease, it is believed that functional
foods have a major role to play in all states of health,
including maintaining health and preventing disease.
In contrast to Asia, the concept of functional foods in
Europe is relatively new, and at present there are no
Europe-wide regulations in place. In developing health
claims legislation, the EU has adopted a working definition
resulting from a European consensus published in 1999: a
food can be regarded as functional if it is satisfactorily
demonstrated to beneficially affect one or more target
functions in the body, beyond adequate nutritional effects,
in a way which is relevant to either an improved state of
health and well-being, or reduction of risk of disease.
The target of functional foods is seen as clearly different
from that of drugs, which are aimed at preventing or
curing diseases.
The scientific underpinning of health claims may derive
from experimental studies and/or epidemiological and inter-
vention studies where the use of biomarkers is regarded as a
key element. The EU PASSCLAIM project (Process for the
Assessment of Scientific Support for Claims on Foods) is
intended to provide industry, academics, consumer groups
and regulators with the means to evaluate the scientific basis
for health claims. Although there is increasing demand by
the European consumer for healthier food products, so far
few functional food products have actually reached the
market place. This is due partly to the fact that developing ade-
quate scientific support for health claims can be relatively
expensive and time-consuming. Moreover, some food manu-
facturers struggle with how best to communicate diet- and
health-related information to consumers.
Research on food and nutrition has been an important
topic in the EU Framework Programmes for Research
and Technology Development of the European Commis-
sion. In the 1990s a significant number of EU projects
addressed issues such as fibres and pro- and prebiotics,
whereas more recent EU programmes focus on areas
such as antioxidants, vitamins and phyto-oestrogens, as
well as the socio-economic aspects of nutrition and health.
Although there is no specific definition of functional
foods in Latin America, the concept of functional foods
is presently under discussion in several Latin American
countries. Brazil is in the process of developing health
claims, and although Argentina does not have health
claims regulations in place, it does approve claims on a
case-by-case basis. Provided there is adequate scientific
validation, both countries have approved nutrient function
and disease risk reduction claims for specific food
products. Usually, locally conducted research to support
product efficacy for health claims approval is requested.
The Latin American population suffers from diseases of
nutrient excess as well as nutrient deficiency. Micronutrient

Abbreviations: COX-2, cyclo-oxygenase-2; CVD, cardiovascular disease; FOSHU, food for specific health use; GI, gastrointestinal; ILSI, International Life
Sciences Institute; NOAEL, no observed adverse effect level; PASSCLAIM, Process for the Assessment of Scientific Support for Claims on Foods;
SYNCAN, Synbiotics and Cancer Prevention in Humans.
* Corresponding author: Drs P. M. Verschuren, fax +31 10 460 57 26, email paulus.verschuren@unilever.com
S126 P. M. Verschuren
deficiencies exist for I, Fe and vitamin A, and sub-optimal
intakes of Zn, folate, Ca and Se are reported. Nutritional
excess includes energy, fats, carbohydrates, cholesterol,
salt and sugar, in addition to decreased physical activity.
There is thus a need in Latin America for functional
foods that contain specific micronutrients through fortifica-
tion, processing and genetic engineering. Although most
consumers in the region are not aware of functional
foods, they are concerned with their diet and health. In
this respect, natural foods are perceived as healthier
than processed foods. Although the consumer associates
food with healing and curing diseases, there is a clear dis-
tinction between the roles of foods and drugs.
It was stressed that basic nutrition research on the health
impact of locally produced products such as grains and
fruits in Latin America is inadequate. Epidemiological and
clinical research is needed to correlate regional foods/diets
and diseases. Also needed is a better understanding of the
genetic background of the populations and what influence
this may have on health and disease states.
In the United States and Canada there is also no univer-
sally accepted definition of functional foods, but 60 % of
the people select foods they believe are functional. The
present understanding of the US population appears to be
that functional foods include both nutrients and non-
nutrients. Foods and drugs are clearly distinguished by
their intended use and presentation, as well as by their
targeted population. In the United States, functional
attributes can be communicated through health claims,
structure function claims, and nutrient content claims.
The Food and Drug Administration must approve health
claims, and fifteen claims have been approved to date.
These claims describe the relationship between a food
component and a disease or health-related condition. The
approval of claims has been based on an extensive
review of existing scientific literature, in the form of an
authoritative statement of a scientific body of the US
government or the National Academy of Sciences. Nutrient
content and structure function claims are clearly defined
in the regulations and do not need to be approved by the
Food and Drug Administration. Canada also has under-
taken initiatives to establish nutrient and health claims
regulations similar to those in the United States.
Although both scientists and regulators are gaining more
experience and scientific understanding of health and
nutrition, the complexity of the issues continues to grow.
Recognising the role of food consumption patterns in
growth, development and disease risk is important to
understanding how diet influences these areas. There is a
myriad of possible interactions between specific dietary
components and specific genetic pathways involved in
health. Understanding these interactions will provide
guidance in establishing dietary recommendations and
communicating benefits to targeted populations.
Session II. Scientific Basis of Biomarkers and
Benefits of Functional Foods: Enhancement of
Function and Reduction of Disease Risk
The second session reviewed the scientific support cur-
rently available for biomarkers that indicate enhancement
of physiological function and disease risk reduction.
Regarding enhancement of physiological function, the
session included gastrointestinal (GI) physiology, the
immune system, physical performance, and behavioural
and psychological functions. Regarding disease risk
reduction, the session focused on biomarkers for evaluating
the effects of functional foods on the risk of cardiovascular
disease (CVD), obesity/diabetes, cancer and osteoporosis,
all elucidated with examples. The need for well-designed
placebo-controlled studies in target populations was
discussed, as were the identification and validation of
appropriate surrogate biomarkers. Session highlights
follow.
GI tract physiology and functions are related to meal-
induced responses and longer-term adaptive changes, to
the mechanisms of nutrient absorption and digestion, and
to the impact on metabolic effects in other organs. The
digestion and absorption of food are regulated by several
factors such as the hormone cholecystokinin, which is
released in response to a meal, thereby preparing the GI
tract to digest food and ensuring a steady rate of nutrient
absorption. Cholecystokinin also signals short-term satiety
and can be regarded as a mediator of satiety and feelings of
hunger. A second example focused on the digestion and
absorption of carbohydrates and their impact on glucose,
insulin and energy metabolism. The role of non-digestible
fermentable carbohydrates for GI function, including the
microflora of the large intestine, and potential systemic
consequences (e.g. on the immune system) highlighted
the importance of this organ system. Also addressed was
the role of viscosity in relation to the metabolic effects
of soluble dietary fibres on nutrient absorption rate,
plasma cholesterol and glycaemic response. Research was
called for to elucidate the bioavailability of bioactive
compounds, the role of the food matrix, and health-related
outcomes originating from modulation of the gut
microflora.
The functioning of the immune system and its role in
protecting the host were reviewed, as were factors such
as the balance between pro- and anti-inflammatory effects,
inter-individual variations and available biomarkers.
Although it is clear that individuals with sub-optimal
immune responses are more susceptible to infectious
agents, it is not clear how modulation of the immune func-
tion in healthy individuals relates to susceptibility to
infections. Food components can play a key role in main-
taining optimal immune function; undernutrition appears to
impair this function and, consequently, susceptibility to
infections. Animal and human studies have demonstrated
that optimising food intake through nutrient fortification
can restore resistance to infections but that excesses of
particular nutrients can lead to impaired immune function.
Moreover, specific dietary components can increase a
particular immunological parameter while decreasing
another. Among the most studied nutrients in this regard
are vitamin A, vitamin E and Zn. There is emerging
evidence that probiotic bacteria may improve host
immune function, although, at present, human data are
still controversial and the relevance of such immune modu-
lations for human health is unclear. To correlate dietary
intervention, immune function and consequent health
 Summary Report
impact, research should also focus on whole-body
parameters, such as the incidence of infections, in addition
to individual immune parameters.
During the last century there were significant changes in
elucidating the role of diet in exercise and physical per-
formance. It was shown that exercise-induced reductions
in muscle glycogen correlate well with the development
of fatigue and that the intake of carbohydrates (versus
protein or fat) improves performance. Nutrition research
to identify nutrients that support energy metabolism,
promote fluid balance, increase muscle mass and improve
overall performance (e.g. caffeine, creatine, ribose, certain
amino acids and L -carnitine) led to the development of
tailored foods and drinks that are easily digestible, help
sustain the exercise load and improve the recovery of
athletes from intense exercise. So far it is recognised that
sports nutrition research is, in scientific terms, the best-
established functional food area. Accordingly, measure-
ment techniques (for gastric emptying, rate of intestinal
absorption, appearance of substrates in blood and sub-
sequent storage) and laboratory exercise protocols that
measure performance accurately have been developed
and validated. As in other functional food areas, solid
scientific underpinning of product claims is a prerequisite.
Interest in functional foods for the enhancement of
behavioural and psychological functions is targeted to
foods that can influence appetite and satiety, vitality,
stress and other subjective states of mood and well-being,
cognitive and mental performance, and sleep. In principle,
biological markers for behavioural and psychological
responses could be identified, but the complexity of these
responses would make their identification far from
simple. Because of progress in, for example, the field of
appetite and satiety control via specific GI and metabolic
responses, and in mood research via changes in plasma
profiles of metabolites, these areas would present good
opportunities for further research. Food intake and
hunger rating profiles are often used as biomarkers for
appetite and satiety. In this respect, it has been shown
that proteins increase satiety more than carbohydrates
and fats. There also is a clear need to distinguish between
short-term postprandial responses to food ingestion and
long-term effects of dietary adjustment, because they
require different research methodologies. Research related
to cognitive performance needs psychobiological assays.
Several types of trials have measured different abilities,
but no unambiguous general biomarkers are available. It
is necessary to ensure that foods that provide improve-
ments for one phenomenon do not exert deleterious effects
on other bodily functions. On balance, there is no clear
conceptual framework to develop and assess the efficacy
of foods to improve behavioural and psychological
functions, and there is a need for markers that represent
real-world situations.
There appear to be many potential biomarkers for
chronic diseases such as CVD, diabetes, cancer and
osteoporosis, but both their ease of use and level of valid-
ation may vary significantly. It was proposed that four criti-
cal features should characterise effective biomarkers for
disease risk: appropriate response in clinical or dietary
trials, consistency with epidemiology, appropriate response
S127
in hereditary disease states, and plausibility of mechanism.
When these criteria are applied to the plethora of bio-
markers that have been proposed for CVD, only blood
pressure and blood cholesterol levels stand out as valid
markers. However, arterial intima-media thickness also
appears to be an appropriate non-invasive marker for
CVD with important potential use in dietary interventions,
e.g. its positive association with blood homocysteine
levels. Various indicators of endothelial function have
been identified (e.g. vascular cell adhesion molecule 1,
intercellular adhesion molecule 1 and p-selectin), but
there is little evidence that changing these indicators will
affect the disease risk. Markers are also available for
CVD other than atherosclerosis, and in this context the
measurement of thrombotic components, inflammatory
markers and arrhythmia were highlighted.
There is an urgent need for efficacious functional foods
for the control of obesity in view of the scope of the obes-
ity epidemic, which is paralleled by a corresponding
growth in type 2 diabetes. There are various non-invasive
markers of obesity (e.g. BMI, dual-energy X-ray absorptio-
metry scanning, bioelectrical impedance, and computerised
tomography and magnetic resonance imaging scanning for
visceral obesity). Fat replacers, fat-binding agents and
ingredients that increase energy expenditure provide clear
opportunities for developing functional foods in this area.
It has been suggested that specific fatty acids such as
medium-chain fatty acids may have a role in weight
reduction. However, there is no conclusive evidence of
their efficacy, and GI implications may limit their incorpor-
ation into diets at higher levels. Results at the cellular level
indicate that high Ca concentrations may increase fat oxi-
dation, but as yet there are few supportive in vivo data.
A potential promising area may be functional foods that
increase satiety. Foods also can be used to reduce the
risk of diabetes, and a variety of biomarkers can be applied
for their evaluation (e.g. oral glucose tolerance tests,
fasting blood glucose, insulin levels and insulin sensi-
tivity). At present, however, there are few well-executed
nutritional studies in this field.
In the area of bone health and osteoporosis, many of the
valid biomarkers, e.g. fracture frequency and bone mineral
content/density, require long-term intervention. Although
bone mineral content measurements, for example, are
relatively simple and non-invasive, two to four years of
intervention are needed to provide meaningful data.
Shorter-term markers are based on the measurement of
Ca metabolism/excretion or on urinary markers of bone
formation and resorption. A new technique based on 41Ca
tracer technology appears to provide a highly sensitive
and direct measure of calcium release/bone loss and
requires relatively small numbers of subjects to achieve
statistical power. Opportunities for developing functional
foods in this area include in particular those that enhance
Ca resorption, such as casein phosphopeptides, whey
protein concentrates and non-digestible oligosaccharides.
In the field of cancer risk a number of potential bio-
markers have been identified, especially for colorectal
cancer, including mucosal markers (e.g. cyclo-oxygenase-
2 (COX-2), DNA repair capacity, microsatellites), faecal
markers (e.g. calprotectin, AP-1 activation, cytotoxicity,
S128 P. M. Verschuren

genotoxicity) and immunological/inflammatory markers
(e.g. natural killer cells, COX-2 suppression). Many dietary
intervention studies on prevention of colorectal cancer
have been carried out on adenoma patients with polyp
recurrence as the endpoint, but the relevance of such bio-
markers for the general population is not yet established.
For most of the markers of colorectal cancer in use, there
is clearly a need for further and more systematic evalu-
ation. These markers are currently being applied in a
major EU-funded intervention trial (Synbiotics and
Cancer Prevention in Humans, SYNCAN), and this study
is expected to provide valuable information on their utility
and suitability. Given the epidemiological associations of
dietary components and cancer risk, there are good
opportunities for developing functional foods in this area,
particularly from plant sources.
Genomic and post-genomic technologies were high-
lighted as having great potential for the identification of
new biomarkers for disease risk and for intervention
trials, with proteomics being potentially most valuable.
However, there was general consensus that the scientific
community needs to arrive at well-validated biomarkers.
If these are not available, it may be more appropriate to
invest in a small number of well-conducted trials with
clear disease endpoints or other hard markers than in a
larger number of shorter-term studies based on poorly
validated biomarkers.
A warning note was sounded that dietary regimens that
reduce the risk of one disease may increase the risk for
another, e.g. an increased Ca intake may reduce the risk
for osteoporosis but also may increase the risk for prostate
cancer.
Table 1 lists the biomarkers for well-being and disease
risk reduction discussed at the meeting.
Session III. Human Genetic Variability and
Safety of Functional Foods
Emerging information about genetic variation and its
impact on human health makes this an exciting new area
related to understanding the effects of functional foods.
Single nucleotide polymorphisms occur frequently, but
there is little evidence that such variation in genetic
expression explains a significant portion of the variability
seen in the response to disease risk factors, especially
those linked to dietary components. Nevertheless, the
emerging information on the impact of genetic variability
signals a significant potential for research. The difficulty
of showing actual health effects from nutrients was
acknowledged. For example, a recent trial with vitamin E
and a multimicronutrient supplement in elderly individuals
did not show any improvement in infection rates. In
addition, elevated homocysteine levels are associated
with increased risk for CVD, but evidence of direct
causality is lacking. Variation in the gene coding for
methylenetetrahydrofolate reductase, a folate-dependent
enzyme responsible for remethylation of homocysteine to
methionine, has been examined as a way of addressing
the causality issue. The majority of epidemiological studies
examining individuals with this mutation and the resulting
elevation in homocysteine, however, did not find an associ-
ation with CVD risk. Nevertheless, evidence is emerging
that elevated homocysteine levels are positively associated
with carotid artery intima-media thickness, which could be
used as an intermediate marker for the progression of
atherosclerosis. The benefit of folate ingestion in reducing
homocysteine levels is complicated by variations in the
bioavailability of different folate isomers (variations in
the length of the glutamate chain). Pooled analysis of all
available observational data responding to well-defined
criteria may be the best way to establish true relationships.
The traditional food additive safety assessment approach
of establishing a no observed adverse effect level
(NOAEL) may not be appropriate for assessing the safety
of whole foods, including functional foods. Whole foods
are complex mixtures, and it is difficult to generate acute
effects, as is expected with the NOAEL approach. More-
over, this approach is based on the assumption that no
effect is the best outcome, but with functional foods, an
actual effect a beneficial one is the expected
outcome. Therefore, with functional foods the standard
should be wholesomeness rather than safety.
The EU Novel Food Directive sets out the principle of
substantial equivalence based on a history of safe use,
the standard that has been applied to foods historically.
New foods should be evaluated against a reference food
for composition, both nutritive and non-nutritive. It is not
likely that routine safety assessment protocols can be
developed for whole foods and food components.
Consequently, the safety assessment will need to proceed
on a case-by-case basis. Better understanding of the
mechanisms underlying the action of foods and food com-
ponents will make it possible to move away from the use of
conventional safety factors.
Consumer issues. Sessions addressing consumer issues
identified communication and consumer understanding as

Table 1. Biomarkers for well-being and disease risk reduction discussed at the meeting

Physical performance Muscle glycogen, endurance time trial

Gut function Gastrointestinal hormones, e.g. cholecystokinin; physical/chemical parameters, e.g. viscosity; biological
responses, physiological responses, e.g. transit time
Immune function Whole-body measures, e.g. delayed hyperactivity, vaccine response
Appetite control Reduction in food intake, reduction in energy intake, hunger rating profiles
Cognitive function No conceptual framework need for markers that represent complexity of real-world decision making
Atherosclerosis Blood pressure, LDL cholesterol, HDL cholesterol, intima-media thickness
Obesity BMI, measures of fatness
Diabetes Glucose tolerance, fasting blood glucose, insulin levels
Cancer Recurrence of colon polyps, aberrant crypt foci
Bone health Bone density, Ca kinetics
 Summary Report
key factors for the success of functional foods. A first issue
concerns the health claims that are justified and the level of
scientific evidence needed to support such claims, and how
can they be communicated effectively to the consumer.
Second, the extent to which nutritionists understand the
barriers to changing eating patterns has to be addressed.
Pan-European surveys (e.g. the EU-funded HealthSense
project) have shown that consumers have a good under-
standing of the link between diet and health and take the
need for healthy eating seriously. Nevertheless, most con-
sumers believe they do not need to change their diet
because they feel it is healthy enough. In other words, con-
sumers admit that a problem exists for society but not
necessarily for them personally. Research reflects this by
the fact that consumers rank the non-communicable
chronic diseases as high concerns for society but as low
concerns for themselves. Consumer perceptions of food
safety issues constitute a third area of interest. Their
perception of risk is a function of their own control, their
possible benefit and the familiarity of the issue. Conse-
quently, in terms of health impact, consumers tend to
rank relatively unimportant or low-risk issues much
higher than important or high-risk ones. For the consumer,
risk is an emotional rather than a scientific issue, which
suggests a mistrust of science and scientists. To improve
the consumers understanding of risk, the scientific
community needs to better understand the consumers
perspectives in this respect.
Session IV. The Future of Functional Foods
Functional foods are a reality today and are likely to be so
in the future. The key drivers behind functional food
research and development are the food industry, consumers
and governments. The progress of the science, particularly
in the area of nutrition, is fundamental to the development
of innovative food solutions for the improvement of key
body functions and consequently consumer health. Past
research has been strong in this field and has been able
to develop sound and effective research programmes. In
addition, the rapid development of the genomic sciences
will allow us to improve our understanding of the effects
of nutrients on gene function and health outcome and to
predict individual nutritional needs. The fact that these
new tools will have an impact on functional science is
given; only the time frame is unpredictable. Overall,
scientific research is focusing on the validation and use
of biomarkers of functional improvement, the evaluation
of the safety of foods and ingredients, the solid under-
standing of the mechanisms of action, and, of course, the
discovery of bioactive ingredients.
Clearly, the success of functional foods is also
dependent on the ability of the food industry to develop
efficacious products that meet consumer needs. The
global potential for functional foods is significant and
growing because of increasing health consciousness and
the self-care trends associated with ageing, knowledgeable
and wealthier consumers. Todays consumer expects foods
to be convenient, safe, healthy, and above all tasty.
Functional foods in particular are expected to provide a
credible health benefit beyond basic attributes to ensure
S129
daily and future health. These health messages need to
be communicated in a transparent, credible and under-
standable manner to the different stakeholders, including
expert scientists with a consumer focus and the various
opinion leaders, including health care providers, consumer
organisations and the media.
In all of this, governments have a major role to play in
the future of functional foods. They can create a favourable
environment for basic and applied research programmes,
promote continuous and truthful consumer education on
nutrition science, facilitate integration of public health
issues, competitive and innovative economic development
with ethical and ecological perspectives, ensure the protec-
tion of consumers in the short and long term, and enforce
regulatory systems for flexible and credible science-based
claims on nutrition and health.
Finally, functional food science will create many
opportunities, but its ultimate success and impact on public
health will depend on the consumers appreciation of pro-
ducts based on objective criteria like taste and convenience
and subjective criteria like trust and credibility.
Conclusions
The International Life Sciences Institutes (ILSI) inter-
national symposium on Functional Foods: Scientific
and Global Perspectives provided an opportunity to
address progress in this field since the first such inter-
national conference, which ILSI held in Singapore in
1995. Despite significant areas of controversy on func-
tional foods, participants in the 2001 symposium
agreed with Dr John Milners statement that functional
foods provide: an unprecedented opportunity to expand
the use of food to improve health, decrease the risk of
disease, and increase productivity. The meeting partici-
pants also concurred that the ability to use claims was
essential to the further development and success of func-
tional foods.
The areas of disagreement are associated with the
definition of the term functional foods and with the
standards of scientific evidence required to support
health claims. Differences in definition include whether
nutrients and non-nutrients are included, whether dietary
supplements and even drugs are included, and whether
functional foods should include only unprocessed foods
or should also include enhanced or fortified foods.
Across the world there are two distinctly different
approaches to standards of evidence. Some countries are
content with a culture or belief system approach based
on history of use. Other countries rely on the paradigm
of scientific substantiation that includes validated bio-
markers. Such markers allow physiological responses to
be measured in terms of improved well-being, enhanced
function or reduction in disease risk.
Several important underlying concepts were addressed.
First, a large amount of information is available, especially
with the coming explosion of genetic marker information,
but the translation of these individual pieces of information
into knowledge is moving at a much slower pace. Second,
there is a critical need to demonstrate valid and relevant
health effects of functional foods and food components
S130 P. M. Verschuren
in man. In this respect, a lack of validated biomarkers for
specific health outcomes is a significant barrier to progress.
Third, there is a need to better understand what normal
physiological function is and what the range of normal
variability is. Current work with single nucleotide poly-
morphisms is providing ever-expanding information
about normal human genetic variability. However, this
variability explains only a small amount of the risk asso-
ciated with a specific change and, consequently, the diffi-
culty of linking a specific health outcome to intake of a
specific food or food component.
There are significant communication issues regarding
functional foods that go beyond scientific understanding.
Taste and convenience are still the primary reasons why
consumers choose specific foods, but there is a growing
interest in self-administered health care. To develop
credible and effective messages for consumers, there
must be in-depth understanding of consumers needs
and wants and how they view the concept of functional
foods and their consequent health benefits. This under-
standing is a prerequisite to generating behavioural
change. The benefits of functional foods must be
made accessible to both rich and poor. Most impor-
tantly, the food industry must do no harm in providing
functional foods. To communicate the emerging scienti-
fic information to the consumer effectively, flexible
science-based regulatory systems must be adopted by
governmental institutions.
British Journal of Nutrition (2002), 88, Suppl. 2, S131S132 DOI: 10.1079/BJN2002676
q International Life Sciences Institute 2002

EU-funded research on functional foods

Jurgen Lucas*
Health, Food and Environment (E.2), Research DG, European Commission, SDME 8-32,
B-1049 Brussels, Belgium

Research on food and nutrition has been an important topic in all Framework Programmes for
Research and Technological Development of the European Commission. From the Second
Framework Programme (ECU 2 million for four projects on functional foods) to the Fifth
Framework Programme (e51 million for thirty-three projects on functional foods), the invest-
ment in research projects on functional foods has been increasing by quite an extent. In the
early 1990s, the topics were fibres, pro-, pre- and synbiotics. Nowadays, the range of subjects
has been broadened to antioxidative effects, vitamins, phyto-oestrogens and the socio-economic
area.

Functional foods: Framework Programmes: European Commission

Research on food and nutrition has been an important topic of ECU 12 million were supported. Two projects were
in all Framework Programmes for Research and Tech- again in the field of gastrointestinal health: analysis of
nological Development of the European Commission. the intestinal flora, and demonstration of nutritional
Within this field of research, functional foods are of functionality of probiotics. Four projects were about the
ever-increasing importance (see Fig. 1). The main empha- antioxidative effects of food components: antioxidative
sis here has been on the elucidation of mechanisms of additives, phenols and tannins (two projects), antioxidants
action and on the proof of effectiveness in prevention in tomato processing. The other projects covered the
and therapy. subjects of conjugated linoleic acid, phyto-oestrogens,
The Second Framework Programme (1989 1994) was n 2 3 polyunsaturated fatty acids and caseinophospho-
the first under which research on food science and nutrition peptides. The socio-economic area was covered by projects
was funded. In the specific programme FLAIR (Food-
Linked Agro-Industrial Research), four projects on
functional foods were funded with a contribution of ECU
2 million. The projects were about production and nutri-
tional properties of fibres, the role of lectins in intestinal
microbial ecology, the consumption of resistant starch,
and human probiotic strains.
The subjects and number of projects within the Third
Framework Programme (1991 1994) were comparable.
Within the specific programme AIR (Agro-Industrial
Research), five projects on functional foods were funded
with a contribution of ECU 5 million. The subjects were
casein complexes in reduced-fat foods, probiotics,
carbohydrates/fibres and colon function (three projects).
In the Fourth Framework Programme (1994 1998), the
area of functional foods was dealt with under the scope of Fig. 1. Research on functional foods in the Framework Pro-
the specific programme FAIR (Agriculture and Fisheries grammes (FP) of the European Commission: (B), available budget;
including agro-industry, food-technologies, forestry, (B), amount thereof for functional foods. FLAIR, Food-Linked Agro-
aquaculture and rural development). The number of pro- Industrial Research (FP2); AIR, Agro-Industrial Research (FP3);
FAIR, Agriculture and Fisheries including agro-industry, food-tech-
jects and the total contribution doubled, and the range of nologies, forestry, aquaculture and rural development (FP4); KA1,
subjects broadened in comparison with the Second Frame- Key Action 1 Food, Nutrition and Health of Quality of Life
work Programme. Twelve projects with a total contribution and Management of Living Resources (FP5).

* Corresponding author: Dr J. Lucas, fax +32 2 296 43 22, email jurgen.lucas@cec.eu.int

British Journal of Nutrition (2002), 88, Suppl. 2, S133S138 DOI: 10.1079/BJN2002677
q International Life Sciences Institute 2002

Global view on functional foods: European perspectives

M. B. Roberfroid*
Universite Catholique de Louvain, 7A rue du Rondia, B-1348 Louvain-la-Neuve, Belgium

In the context of this presentation, European means the European scientific community, which,
over the last decade, has been working on the concepts for functional foods, producing a
number of documents including a consensus paper, guidelines and scientific publications.
The Functional Food Science in Europe (FUFOSE) project has been quite a unique attempt
to establish a strong scientific framework to justify the functional food concept, to discover
and develop new functional foods that are primarily function-driven, and to substantiate
claims scientifically. Being clearly positioned as part of nutrition, the functional food concept
is, however, quite distinct from other approaches like food supplementation or food fortifica-
tion, and functional foods are different from nutraceuticals, pharmafoods, vitafoods and alica-
ments, all terms that are not defined conceptually. Functional foods are food products to be
taken as part of the usual diet in order to have beneficial effects that go beyond what are
known as traditional nutritional effects. Moreover, these beneficial effects have to be demon-
strated scientifically to justify two specific types of claim: the enhanced function claim or
the reduction of disease risk claim.
Functional food is a key concept for the future of nutrition as a science because it results
from the implementation in nutrition of all the basic scientific knowledge that has accumulated
over the past two or three decades. To the benefit of public health this progress cannot be
ignored, it needs to be recognized fully and used. But, today, functional food is still mainly
a scientific concept that serves to stimulate research and the development of new products.

Functional foods: Markers: Claims: FUFOSE

Nutrition: a science for the twenty-first century
At the beginning of the twenty-first century, the society of
abundance, which characterizes most of the occidental/
industrialized world, faces new challenges from increased
costs of health care, increased life expectancy, new scien-
tific knowledge and the development of new technologies
leading to major changes in life-styles (Table 1). Nutrition,
as a science of the twenty-first century, will, in addition to
maintaining the emphasis on a balanced diet, target opti-
mum (optimized) nutrition (Milner, 2000), with the object-
ive of maximizing physiological functions in order to
ensure both maximum well-being and health and simul-
taneously minimize the risk of disease throughout the
life-span. At the same time, and because of a better under-
standing of the interactions between genes and nutrition
(Kok, 1999), it will have to match the individuals
unique biochemical needs with a tailored selection of nutri-
ent intakes for that individual. These interactions include
polymorphism and inter-individual variations in response
to diet, dietary alteration and modulation of gene
expression, and dietary effects on disease risk.
Thus, at the turn of the new century, the major challenge
of the science of nutrition is thus to progress from improv-
ing life expectancy to improving life quality/wellness.
On the road to optimum (optimized) nutrition, which is
an ambitious and long-term objective, functional food
is, amongst others, a new, interesting and stimulating con-
cept. It is supported by sound and consensual scientific data
generated by the recently developed functional food
science, which aims to improve dietary guidelines by inte-
grating new knowledge on the interactions between food
components and body functions and/or pathological
processes.
Functional food: defining the concept
Functional food cannot be a single well-defined/well-
characterized entity. Indeed, a wide variety of food products
are, or will be in the future, characterized as functional
foods, with a variety of components both classified
and not classified as nutrients affecting a variety of
body functions relevant to either a state of well-being
and health and/or to the reduction of the risk of a disease.
Thus no simple, universally accepted definition of func-
tional food exists or will (ever) exist. Functional food has

Abbreviations: FUFOSE, Functional Food Science in Europe; ILSI, International Life Sciences Institute.
* Corresponding author: Professor M. B. Roberfroid, fax +32 10 45 93 01, email roberfroid@orafti.com
British Journal of Nutrition (2002), 88, Suppl. 2, S139S143 DOI: 10.1079/BJN2002678
q International Life Sciences Institute 2002

Global view on functional foods: Asian perspectives

S. Arai*
Department of Nutritional Science, Tokyo University of Agriculture, 1-1-1 Sakuragaoka,
Setagaya-ku, Tokyo 156-8502, Japan

There is an old Chinese thought meaning that medicine and food are isogenic. In Japan, a
national research project started to endorse the thought scientifically, with the birth of func-
tional foods that are now known to function to reduce the risk of life-style related diseases.
The science gave rise to a national policy in which some functional foods were legally approved
in terms of Foods for Specified Health Use. China has also constructed a unique system for
approving some functional foods. This paper pinpoints a recent trend of functional foods
characteristic to Asia, with special reference to relevant topics in Japan.

Functional foods: Japan: Health: FOSHU

Introduction foods. Epoch-making advances in instrumental analysis

The term physiologically functional food, which first
appeared in Nature news in 1993 with the headline
Japan explores the boundary between food and medicine
(Swinbanks & OBrien, 1993), gave a strong international
impact. Neither the terminology nor the concept of func-
tional food had existed until nine years earlier. In 1984,
an ad hoc research group started a systematic, large-scale
national project under the sponsorship of the Ministry of
Education, Science and Culture (MESC) to explore the
interface between food and medical sciences (Arai,
1996). It seems that the realization of this research event
first in Japan reflects an underlying thought in the minds
of Japanese people throughout history, which originates
in the ancient Chinese saying: Medicine and food are
isogenic.
However, the advent of modern food science in Japan,
which took place about 100 years ago, made its subsequent
research proceed towards a purely nutritional rather than a
medical science. Especially in the early twentieth century,
the nutritive value of foods was of great academic concern.
A number of scientists in the field of chemistry as well as
food science made particular efforts to discover new
nutrients.
More than 10 years after World War II, nutritional
problems due to food shortage were almost solved in
many countries and a new period characterized by high
economic growth began. It prevailed in the time encom-
passing the 1960s, when the social climate focused on
food preference and aversion. Such a trend of hedonism
led to academic studies on the sensory properties of
assisted the studies greatly. Japan, among many developed
countries, thus contributed to two mainstays of food
science: nutrition and hedonics. In accordance with this,
food industries were activated to supply a great deal of
nutritionally and sensorily acceptable products to the
market. We enjoyed thus our rich dietary life for a long
time until almost 20 years ago, when a new, antithetic
situation arose.
Functional food science and policy born in Japan
In the 1980s, as the ageing society began to manifest itself
in many countries of the world, prompt increases in the
so-called life-style related diseases became a matter of
public concern. Growing awareness was then observed of
the need for eating to beat the odds. The purpose was to
prevent life-style related diseases such as diabetes, arterio-
sclerosis, osteoporosis, allergies, cancer, and even some
kinds of infectious diseases, through improved dietary
practices in daily life. This gave a strong impetus to food
science in Japan. In 1984, the above-mentioned MESC
project entitled Systematic Analysis and Development of
Food Function commenced, where the terminology and
the concept of functional food were first proposed. It
was followed by a second project (1988 1991) entitled
Analysis of Body-modulating Functions of Foods. In
1992, the last in the series of MESC functional foods
projects was realized, focusing on Analysis and Molecular
Design of Functional Foods. A total of 60 academic
professionals participated, most from medical as well as
food science fields.

Abbreviations: FOSHU, food for specified health use; ILSI, International Life Sciences Institute; MESC, Ministry of Education, Science and Culture.
* Corresponding author: Dr S. Arai, fax +81 3 5815 7965, email ara@kiwi.ne.jp
British Journal of Nutrition (2002), 88, Suppl. 2, S145S150 DOI: 10.1079/BJN2002679
q International Life Sciences Institute 2002

Functional foods: Latin American perspectives

Franco M. Lajolo*
Departamento de Alimentos e Nutricao Experimental, Faculdade de Ciencias Farmaceuticas,
Universidade de Sao Paulo, Av. Prof. Lineu Prestes 580, Bl. 14, 05508-900 Sao Paulo, Brazil

The perspectives of Latin America as a potential producer and consumer of functional foods
will depend largely on the level of information and income of the population, credibility of
the products, research investments and regulatory practices. The characteristics of Latin
America are diverse at the regional and sub-regional levels. However, as part of the
demographic and epidemiological transition currently underway, common trends can be
identified such as increasing urbanization and life expectancy, the occurrence of obesity and
malnutrition, increasing incidences of chronic diseases and causes of mortality rates, all of
which suggests the importance of diet and functional foods in public health policies. The
Latin American population in general has no knowledge of functional foods, but in the more
urbanized areas there is an increasingly health-conscious consumer, aware of the importance
of food for health, due to the media and local traditions. More investment in research is
important to explore the existing plant biodiversity that is a rich source of new foods and
bioactive compounds, some of which are already used for health improvement and well-
being. Clinical validation of functional foods should consider functional food science concepts
and also the diverse cultural and genetic background of the local population. In Latin American
scientific and regulatory communities, the functional foods concept has been associated with
foods having health benefits beyond those of basic nutrition but is not defined officially in
the emerging regulatory codes. Regulation existing in some countries is focused on safety
and efficacy; both functional and health claims are allowed (risk reduction) provided they
have scientific validation. This allowed introduction on the market of several products with
health claims, some of them submitted to post-marketing surveillance.

Functional food: Latin America: Needs and challenges

Demographic and urbanization trends
in Latin America
During the last 30 years, the declining mortality and fertil-
ity rates in Latin America have had an impact on the
populations growth and structure. In 1950 the population
in the Americas was 331 million and in the year 2000 it
was estimated at 823 million. About one-third resides in
the USA, one-third is found in Brazil (170 million) and
Mexico (99 million), and the remaining third in the other
countries of the region.
This demographic transition began with a decline in
infant and child mortality, with most countries having
halved their mortality rates among children under 1 year
old in the last twenty years. Life expectancy during the
last five years has stood at 70 years, with a range through-
out the region from 54.1 to 79.2 years, and this change has
resulted in ageing of the population, which increases
chronic and degenerative diseases and disabilities. At the
same time there has been an increase in the urban
population, estimated now at 76 %, although it varies
from 853 % for the southern cone to 483 % for Central
America (Pan American Health Organization (PAHO),
1999a). In past decades mortality rates in Latin America
have declined, with an increase of 18 years in life span.
Although major advances have been made in controlling
infectious diseases rates are still high, similar to those of
non-communicable diseases, due to differences in public
health care in different social groups.
Nutrition in transition: globalization and its impact on
nutrition patterns and diet-related diseases
As has happened all over the world, industrialization,
urbanization and market globalization have had a strong
impact on Latin Americans life-styles and diets, and the
nutritional status of the population. The contemporary
urbanization has resulted in a decline of undernutrition in

Abbreviations: PAHO, Pan American Health Organization.

* Corresponding author: Dr F. M. Lajolo, fax 55 11 3815 4410, email fmlajolo@usp.br
British Journal of Nutrition (2002), 88, Suppl. 2, S151S158 DOI: 10.1079/BJN2002680
q International Life Sciences Institute 2002

Functional foods and health: a US perspective

J. A. Milner*
Nutritional Science Research Group, Division of Cancer Prevention, National Cancer Institute,
6130 Executive Boulevard, Suite 3164, Rockville, MD 20892, USA

Linkages between diet habits and the quality of life continue to surface on numerous fronts.
Collectively these epidemiological, pre-clinical and clinical studies provide rather compelling
evidence that numerous essential and non-essential dietary components are capable of influen-
cing growth, development and performance and disease prevention. Scientific discoveries and
widespread interest in the potential medicinal benefits of foods and food components have
fostered a variety of content, health and structure function claims. Unfortunately, defining
the ideal diet is complicated by the numerous and diverse components that may influence bio-
logical processes. Inconsistencies in the literature may reflect the multi-factorial nature of these
processes and the specificity that individual dietary constituents have in modifying genetic and
epigenetic events. New and emerging genomic and proteonomic approaches and technologies
offer exciting opportunities for identifying molecular targets for dietary components and thus
determining mechanisms by which they influence the quality of life. All cells have unique
signatures that are characterized by active and inactive genes and cellular products. It is
plausible that bridging knowledge about unique cellular characteristics with molecular targets
for nutrients can be used to develop strategies to optimize nutrition and minimize disease risk.

Diet: Cancer: Polymorphism: Genes: Phytochemicals

Introduction and increased risk of some diseases. Unquestionably,

strategies that optimize nutrition by the use of foods or sup-
There is little doubt that nutrition and health are intimately
plements are highly commendable and considered by many
linked. For generations, people have believed that foods
to be appropriate for improving the overall quality of life.
could do more than merely provide energy. Beliefs in the
The importance of such strategies is emphasized by the
medicinal properties of foods were highlighted in a
recognition that about one-third of all cancers are related
number of the early writings of mankind. Hippocrates is
to dietary habits. Actually, more than half of the deaths
frequently quoted as saying: Let food be thy medicine
occurring in the United States correlate with food patterns
and medicine be thy food. Epidemiological, pre-clinical
(National Center for Health Statistics, 1997). Many in
and clinical studies continue to provide fundamental
North America believe that modifying their diet and the
insights into the dynamic relationships between nutrients
use of supplements are two of the most important options
defined here as any substance in the diet that brings
available to reduce health care costs and improve personal
about a physiological effect and health. Today, claims
health (Eisenberg et al. 1998; Gilbert, 1999; Aldana, 2001).
about the ability of foods and food components to reduce
In spite of a growing number of studies, workshops and
disease risks or enhance the quality of life continue to cap-
conferences devoted to understanding the dynamic
tivate our lives. In North America, passionate discussions
relationship existing between nutrition and disease, it
about foods and beverages are commonplace between
remains virtually impossible to determine who will and
friends, relatives and even complete strangers.
will not benefit most, if at all, from dietary intervention.
What is recognized is that the effect of food on health
varies considerably among people, both in terms of its
Unprecedented opportunities direction and magnitude (Clydesdale, 1998; Milner,
Inappropriate nutrition is a primary factor in unattained gen- 2000). Part of the difficulty in determining those who
etic potential, reduced physical and cognitive performance, might benefit arises from the incredibly complex chemical

Abbreviations: ATBC, Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study; CGAP, Cancer Genome Anatomy Project; FDA, Food and Drug
Administration; ILSI, International Life Sciences Institute.
* Corresponding author: Dr J. A. Milner, fax +1 301 480 3925, email milnerj@mail.nih.gov
British Journal of Nutrition (2002), 88, Suppl. 2, S159S163 DOI: 10.1079/BJN2002681
q International Life Sciences Institute 2002

Gastrointestinal physiology and functions

Barbara O. Schneeman*
Department of Nutrition, University of California, Davis, CA 95616, USA

While the health benefit of a functional food may be a metabolic response that lowers risk for
disease, the actual target for the food or food component may be on the functioning of the
gastrointestinal tract (GIT). For example, slowing absorption from the intestine, as measured
by examining the appearance of the nutrient or food component in the blood, the hormone
response associated with absorption of the compound or excretion of the compound, may pro-
vide a health benefit. However, the food component may slow absorption by delaying gastric
emptying, altering the mixing within the intestinal contents or decreasing the availability of
digestive enzymes in the intestine. These measures of GIT function provide validation of the
mechanisms by which the functional food or food components affect metabolism. Bioavailabil-
ity of physiologically active compounds from foods will be determined by the digestibility of
foods that contain these compounds, their subsequent absorption and utilization by tissues. The
physical structure of foods contributes to the functional effects of foods as well as to the avail-
ability of compounds from foods. For example, recent studies have demonstrated that changing
the viscosity of the gut contents alters absorption and GIT response. Additionally, food struc-
tures such as the plant cell wall change the availability of absorbable compounds along the gas-
trointestinal contents. The areas of probiotics and prebiotics have highlighted the potential
importance of gut microflora in health. While evidence suggests biological activity relevant
to disease risk reduction, the long-term implications of the microbial activity have yet to be
established.

Gastrointestinal tract: Health benefit: Absorption: Digestibility: Pro-/prebiotics

Introduction that are consumed for a specific health benefit or as part of

The gastrointestinal tract (GIT) serves as an interface
between the body and the external environment. The GIT
is a highly specialized organ system that allows man to
consume food in discrete meals as well as a very diverse
array of foodstuffs to meet nutrient needs. In the GIT
food is converted to compounds that can be absorbed
into the body. The organs of the GIT include the mouth,
oesophagus, stomach, small intestine and large intestine;
in addition, the pancreas and liver secrete into the small
intestine. The system is connected to the vascular,
lymphatic and nervous systems to facilitate regulation of
the digestive response, delivery of absorbed compounds
to organs of the body and the regulation of food intake.
Johnson (1997) provides an excellent overview of the
gastrointestinal system. A primary function of the GIT is
to extract nutrients from the complex mixture of foods as
consumed. Foods contain more than essential nutrients,
and the GIT has a role in metabolizing and eliminating
non-nutrient and toxic compounds as well.
Functional foods include the categories of food products
a dietary plan to lower risk for chronic disease. As such,
the components of these foods may exert their effects
through a direct action on the GIT or become available
to other target organs in the body after absorption from
the GIT. In this context, three perspectives are useful to
examine the importance of GIT physiology and functions
in mediating the effects of functional foods: (1) meal-
induced responses in the GIT caused by factors in foods,
which may result in longer-term adaptive changes; (2)
the ability of foods or mixtures of foods to alter the
digestive and absorptive functions of the GIT in a
manner that influences metabolism; and (3) the impact
that the GIT, through its adaptation to diet, has on risk
factors for disease. Examples are used below to illustrate
each of these perspectives regarding the physiology and
function of the GIT.
Meal-induced responses
Regulation of the gastrointestinal response to a meal
involves a complex set of hormone and neural interactions.

Abbreviations: CCK, cholecystokinin; GIT, gastrointestinal tract; SCFA, short-chain fatty acids.
* Corresponding author: Dr B. O. Schneeman, fax +1 530 752 8699, email boschneeman@ucdavis.edu
British Journal of Nutrition (2002), 88, S165S176 DOI: 10.1079/BJN2002682
q International Life Sciences Institute 2002

The immune system: a target for functional foods?

Philip C. Calder* and Samantha Kew

Institute of Human Nutrition, School of Medicine, University of Southampton, Bassett Crescent East,
Southampton SO16 7PX, UK

The immune system acts to protect the host from infectious agents that exist in the environment
(bacteria, viruses, fungi, parasites) and from other noxious insults. The immune system is
constantly active, acting to discriminate non-self from self. The immune system has two
functional divisions: the innate and the acquired. Both components involve various blood-
borne factors (complement, antibodies, cytokines) and cells. A number of methodologies
exist to assess aspects of immune function; many of these rely upon studying cells in culture
ex vivo. There are large inter-individual variations in many immune functions even among
the healthy. Genetics, age, gender, smoking habits, habitual levels of exercise, alcohol
consumption, diet, stage in the female menstrual cycle, stress, history of infections and
vaccinations, and early life experiences are likely to be important contributors to the observed
variation. While it is clear that individuals with immune responses significantly below normal
are more susceptible to infectious agents and exhibit increased infectious morbidity and
mortality, it is not clear how the variation in immune function among healthy individuals relates
to variation in susceptibility to infection. Nutrient status is an important factor contributing to
immune competence: undernutrition impairs the immune system, suppressing immune functions
that are fundamental to host protection. Undernutrition leading to impairment of immune func-
tion can be due to insufficient intake of energy and macronutrients and/or due to deficiencies in
specific micronutrients. Often these occur in combination. Nutrients that have been demon-
strated (in either animal or human studies) to be required for the immune system to function
efficiently include essential amino acids, the essential fatty acid linoleic acid, vitamin A,
folic acid, vitamin B6, vitamin B12, vitamin C, vitamin E, Zn, Cu, Fe and Se. Practically all
forms of immunity may be affected by deficiencies in one or more of these nutrients.
Animal and human studies have demonstrated that adding the deficient nutrient back to the
diet can restore immune function and resistance to infection. Among the nutrients studied
most in this regard are vitamin E and Zn. Increasing intakes of some nutrients above habitual
and recommended levels can enhance some aspects of immune function. However, excess
amounts of some nutrients also impair immune function. There is increasing evidence that
probiotic bacteria improve host immune function. The effect of enhancing immune function
on host resistance to infection in healthy individuals is not clear.

Nutrition: Immune function: Lymphocyte: Macrophage: Monocyte: Cytokine: Infection

The immune system immune system. Both components of immunity involve

Introduction
The immune system acts to protect the host from infectious
agents that exist in the environment (bacteria, viruses,
fungi, parasites) and from other noxious insults. The
immune system is constantly active, acting to discriminate
non-self from self. The immune system has two func-
tional divisions: the innate (or natural) immune system
and the acquired (also termed specific or adaptive)
various blood-borne factors (complement, antibodies,
cytokines) and cells. These cells are generally termed
leucocytes (or white blood cells). Leucocytes fall into
two broad categories: phagocytes (which include granulo-
cytes (neutrophils, basophils, eosinophils), monocytes and
macrophages) and lymphocytes. Lymphocytes are classi-
fied as T lymphocytes, B lymphocytes and natural killer
cells. T lymphocytes are further divided into helper T
cells (these are distinguished by the presence of the

Abbreviations: DTH, delayed-type hypersensitivity; IFN-g, interferon-g; Ig, immunoglobulin; IL, interleukin; MHC, histocompatibility complex; TNF,
tumour necrosis factor.
* Corresponding author: Dr P. C. Calder, fax +44 23 8059 4383, email pcc@soton.ac.uk
British Journal of Nutrition (2002), 88, Suppl. 2, S177S186 DOI: 10.1079/BJN2002683
q International Life Sciences Institute 2002

Functional foods and food supplements for athletes: from myths to benefit
claims substantiation through the study of selected biomarkers

Fred Brouns1,2*, Michiel van Nieuwenhoven1, Asker Jeukendrup3 and Wouter van Marken Lichtenbelt1
1
Nutrition and Toxicology Research Institute Maastricht, Maastricht University, Maastricht, The Netherlands
2
Cerestar-Cargill Vilvoorde R&D Centre, Havenstraat 84, B-1800 Vilvoorde, Belgium
3
School of Sport and Exercise Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

The development of the sports food market and industrial involvement have led to numerous
nutritional studies to define the type of nutrients that are most suited to support energy metabol-
ism, fluid balance and muscle function. The key question in many of these studies was: Does
the product lead to a significant product/consumer benefit that can be used as a claim on the
package? New methods and techniques have been developed, partly with sponsorship of the
food industry, with the goal of measuring the effects of specific nutrients and supplements
on athletic performance and metabolism. In line with this development, a wide variety of
supplements and sports foods/drinks labelled with various performance or health benefit state-
ments have been launched on the sports nutrition market. Although a variety of products have
been tested clinically, there are also many products on the market with benefit claims that
cannot be supported by sound nutritional and sports physiological science. The current short
review highlights some of the methods and biomarkers that are used to substantiate product/
consumer benefit claims for foods and drinks that are marketed as functional foods for athletes.

Sports food: Energy metabolism: Athletic performance: Benefit claims: Biomarkers

Introduction prior to as well as during competition. To improve the

During the last century there have been enormous changes
in the understanding of the role of diet in exercise and
physical performance. Almost a century ago it was con-
sidered that protein was the most important energy
source for muscle. However, since the classical Scandina-
vian studies on the role of diet on physical performance
in the 1960s, the focus has changed completely to carbo-
hydrate. It was shown that exercise-induced reductions in
muscle glycogen correlated well with the development of
fatigue and that optimizing carbohydrate stores and ingest-
ing carbohydrate during prolonged exercise improved
performance. Since then, carbohydrate loading and the
consumption of carbohydrate electrolyte drinks during
exercise have become common practice among endurance
athletes.
This was also the beginning of a new challenge to the
food industry: the development of sports-specific food
products and drinks. Sports foods/drinks should provide
fluid and energy (i.e. carbohydrate) rapidly and not cause
any gastrointestinal discomfort, to allow consumption
athletes recovery from intense and exhausting exercise,
other products were created to target the physiological
functions that are involved in this recovery process.
Specific blends of carbohydrate were studied to obtain evi-
dence about effective energy fuels during exercise, with the
aim of improving performance and delaying fatigue devel-
opment. In line with these developments many products
have been launched on the market to target the fitness
and physical performance sector, mostly with attractive
benefit claims.
Examples of such benefits are performance enhancement,
more power, less gastrointestinal problems, improved
recovery, less muscle cramps/pain and reduction of
body fat/increased muscle mass. To obtain clinical evidence
for a benefit, the development and use of measurement tech-
niques that allowed measurement of real treatment effects
was essential. Accordingly, techniques to measure the
rate of gastric emptying, intestinal absorption, appearance
of substrates in blood and their subsequent oxidation or
storage, as well as laboratory exercise protocols to measure
performance accurately, were developed and validated.

Abbreviations: BIA, bio-impedance analysis; CT, computed tomography; DXA, dual X-ray absorptiometry; ECW, extracellular water; FC, food
components; H, hypothetical; ICW, intracellular water; M, mixed; MRI, magnetic resonance imaging; N, negative; P, positive; SF, skin folds; TAN,
total adenine nucleotides; TBW, total body water.
* Corresponding author: Dr F. Brouns, fax +32 2 2570740, email fbrouns@be.cerestar.com
British Journal of Nutrition (2002), 88, Suppl. 2, S187S211 DOI: 10.1079/BJN2002684
q International Life Sciences Institute 2002

Functional foods: psychological and behavioural functions

Louise Dye* and John Blundell

School of Psychology, University of Leeds, Leeds LS2 9JT, UK

It is easier to demonstrate the consistent effects of foods on satiety than on cognitive perform-
ance. This is understandable since the satiety system incorporates physiological signalling sys-
tems that mediate the effects of foods on function. Specific manipulations of proteins,
carbohydrates and fats have the potential to act as functional foods for appetite control. Because
of the importance of the optimal functioning of cognitions for survival, these functions are quite
strongly protected against short-term dietary and physiological perturbances. Therefore, food
manipulations may be better detected through the degree of effort exerted to maintain perform-
ance rather than via changes in the actual performance itself. This procedure has not been
widely used hitherto. The concept of biomarkers may have to be interpreted differently from
research on physiological systems or clinical endpoints. For satiety, adjustments in the profile
of hunger could serve as a biomarker or surrogate endpoint. For cognitions, correlated physio-
logical variables may be more difficult to measure than the functional endpoint itself. Changes
related to unitary functions (such as tracking) could serve as biomarkers for more complex,
integrated skills (such as car driving). Since food manipulations may affect multiple functions,
the challenge is to design foods with good satiety control that do not impair mental perform-
ance; or alternatively to engineer foods that optimise cognitive performance without compro-
mising satiety. This rapidly developing field has great potential for close collaboration
between academia and industry in the production of commercially successful products that
show clear improvements in human functioning with the capacity to protect against disease
or impairment.

Satiety: Biomarker: Cognitive performance: Functional foods

Introduction In these areas of research it is not clear that the

In the First International Conference on East West
Perspectives on Functional Foods, the major emphasis
was on physiological and metabolic functions. Although
much of this material was relevant for well-being and
quality of life, there was almost no mention of the direct
effects of foods on psychological functions involving the
expression of behaviour (such as food intake or appetite)
or the articulation of mental state (mood) or mental
abilities (cognitive performance).
In the intervening period the world-wide epidemic of
obesity has been recognised, thus bringing appetite control
to the forefront of functions susceptible to modulation by
foods and food ingredients. Consequently, functional
foods for the control of appetite (and ultimately body
weight) now constitute a major health goal and research
objective. In addition, the principle that foods can reliably
affect cognitive performance is receiving validation and
experimental support. This domain of research is still in
a relative state of infancy but also rapid development.
concept of biomarkers has the same meaning as in other
domains which relate physiological functions to some
health endpoint or disease protection. However, it is poss-
ible to suggest ways in which biomarkers for complex
psychological and behavioural functions could be
identified.
This interim report has evolved from the International
Life Science Institutes Functional Food Science in
Europe (FUFOSE) Project and is designed to provide a
perspective on the field and to indicate possibilities for
future developments. The FUFOSE programme concen-
trated on the behavioural functions of appetite control, cog-
nitive performance and mood, and this approach has been
followed here.
Functional foods for satiety
The control of appetite
The ultimate objective of research in this area is to develop

Abbreviations: AUC, area under curve; FUFOSE, Functional Food Science in Europe; 5-HT, 5-hydroxytryptamine.
* Corresponding author: Dr. L. Dye, fax +44 113 233 5749, email louised@psychology.leeds.ac.uk
British Journal of Nutrition (2002), 88, Suppl. 2, S213S218 DOI: 10.1079/BJN2002685
q International Life Sciences Institute 2002

Biomarkers and functional foods for obesity and diabetes

James O. Hill1* and John C. Peters2

1
Center for Human Nutrition, University of Colorado, Health Sciences Center, Box C225, 4200 East Ninth Avenue,
Denver, CO 80262, USA
2
Procter & Gamble Company, 6071 Center Hill Road, Cincinnati, OH 45224, USA

Obesity has reached epidemic proportions in many countries around the world. Because of the
close relationship between obesity and type 2 diabetes, an epidemic of diabetes is close behind
the obesity epidemic. Preventing and treating obesity is becoming an increasing priority. In the
United States, over 60 % of the adult population is overweight or obese and thus at increased
risk of developing diabetes and cardiovascular disease. While the aetiology of obesity and dia-
betes is complex, diet clearly plays an important role both in the development and management
of these diseases. There is interest in functional foods that could help in prevention and/or
management of obesity and type 2 diabetes. This could involve food products that help manage-
ment of hunger or that increase satiety. It could also involve foods that contribute to more
inefficient use of ingested energy (i.e. foods that stimulate energy expenditure more than would
be expected from their energy content). As the concept of insulin sensitivity becomes generally
more accepted by health care professionals and the public, foods may be targeted towards
maximizing insulin sensitivity and towards prevention of diabetes. In addition to foods that
impact upon body weight, these may include foods that affect the glucose and/or insulin
levels that are seen either following the ingestion of food or later in the day. The present
paper reviews the complex aetiology of obesity and diabetes and considers a potential role
for functional foods in prevention and treatment of obesity and diabetes.

Obesity: Type 2 diabetes: Biomarkers: Functional foods

The need for tools to manage obesity
The rate of increase in obesity in the US population has
led many to label this public health threat as an epidemic.
This is supported by data from the National Health and
Nutrition Examination Surveys (National Center for
Health Statistics, 1999) and by data from the Behavioral
Risk Factor Surveillance System (Mokdad et al. 2001).
All indications are that the prevalence of obesity will
continue to increase over the next several years.
The obesity epidemic presents a serious threat to public
health given the known relationship between obesity and
other serious chronic diseases such as type 2 diabetes
and cardiovascular disease (Pi-Sunyer, 1993; National
Institutes of Health, 1998; Fagot-Campagna et al. 2000).
In fact, there is convincing evidence that the obesity epi-
demic is being followed by an epidemic of type 2 diabetes
(Harris et al. 1998). The metabolic syndrome (syndrome X
or insulin-resistance syndrome) has been defined as a
cluster of conditions including obesity (particularly visceral
obesity) and insulin resistance that are frequently seen
together and which impart increased risk of development
of type 2 diabetes (Reaven, 1998). Thus, the vast majority
of persons with type 2 diabetes are obese and weight loss
may be one of the best treatments both to prevent and
treat type 2 diabetes (Maggio & Pi-Sunyer, 1997).
In general, efforts to treat obesity have not met with
great success, but we may be more successful than popular
belief. Obesity treatment guidelines issued by the National
Institutes of Health (1998) suggest an initial goal of 10 %
weight loss for obesity treatment. Wing & Hill (2001)
have estimated that about 20 % of those obese individuals
who attempt weight loss are successful in achieving a 10 %
weight loss and maintaining that weight loss for a year.
With over 60 % of the US adult population being over-
weight or obese, there is great demand for better tools to
help people achieve weight loss and maintenance of
weight loss.
Treatment strategies for type 2 diabetes usually involve
pharmacological treatment aimed at stimulating insulin
secretion or increasing insulin sensitivity. The disease is
progressive (leading to serious negative consequences
such as blindness and kidney disease) and treatment aims
to slow this progression (Lebovitz, 1994). Interestingly,

* Corresponding author: Dr. J. O. Hill, fax +1 303 315 3273, email james.hill@uchsc.edu
British Journal of Nutrition (2002), 88, Suppl. 2, S219S224 DOI: 10.1079/BJN2002686
q International Life Sciences Institute 2002

Scientific basis of biomarkers and benefits of functional

foods for reduction of disease risk: cancer

Joseph J. Rafter*
Department of Medical Nutrition, Karolinska Institutet, NOVUM, S-141 86 Huddinge, Sweden

One of the most promising areas for the development of functional foods lies in modification of
the activity of the gastrointestinal tract by use of probiotics, prebiotics and synbiotics. While a
myriad of healthful effects have been attributed to the probiotic lactic acid bacteria, perhaps the
most controversial remains that of anticancer activity. However, it must be emphasised that, to
date, there is no direct experimental evidence for cancer suppression in man as a result of con-
sumption of lactic cultures in fermented or unfermented dairy products, although there is a
wealth of indirect evidence, based largely on laboratory studies. Presently, there are a large
number of biomarkers available for assessing colon cancer risk in dietary intervention studies,
which are validated to varying degrees. These include colonic mucosal markers, faecal water
markers and immunological markers. Overwhelming evidence from epidemiological, in vivo,
in vitro and clinical trial data indicates that a plant-based diet can reduce the risk of chronic
disease, particularly cancer. It is now clear that there are components in a plant-based diet
other than traditional nutrients that can reduce cancer risk. More than a dozen classes of
these biologically active plant chemicals, now known as phytochemicals, have been identified.
Although the vast number of naturally occurring health-enhancing substances appear to be of
plant origin, there are a number of physiologically active components in animal products
(such as the probiotics referred to above) that deserve attention for their potential role in
cancer prevention.

Probiotics: Cancer: Biomarkers: Phytochemicals: Animal products

Introduction the focus of increasing research activity in recent years, are

In recent years, there has been a growing interest in the
concept of functional foods by both the food industry
and the consumer. Functional foods are defined as food-
stuffs that improve overall health and/or reduce the risk
of disease. Thus, with the consumer becoming more inter-
ested in foods which benefit health and the food industry
beginning to understand the market potential of functional
foods, solid research activity in the area has been initiated
world-wide. In addition, when one considers the costs to
society of diseases such as obesity, cardiovascular disease,
diabetes, food allergies, osteoporosis and cancer, all of
which may be influenced by diet, the potential of well-
characterised functional foods becomes even more evident.
Although a large number of naturally occurring health-
enhancing substances are of plant origin, there are a
number of physiologically active components in animal
products that deserve attention for their potential role in
disease prevention.
One such example of a functional food, which has been
probiotics: live microbial feed supplements that benefi-
cially affect the host animal by improving its intestinal
microbial balance (Fuller, 1989). Probiotics usually refer
to highly selected lactic acid bacteria, e.g. Lactobacillus
spp., Bifidobacterium spp. and Streptococcus spp., with
defined gut survival properties and associated biological
activities, that can be ingested in fermented milk products
or as a supplement. The list of healthful effects attributed
to probiotic bacteria is extensive and includes alleviation
of lactose intolerance symptoms, serum cholesterol
reduction, anticancer effects, alleviating constipation and
relieving vaginitis, to name but a few. The vast majority
of studies on the anticancer effects deal with colorectal
cancer (Hirayama & Rafter, 2000), although there are
some on breast and bladder cancer.
There is also a large amount of evidence, from epi-
demiological, in vivo, in vitro and clinical trial data,
indicating that a plant-based diet can reduce the risk of
chronic disease, particularly cancer. In 1992, a review of
200 epidemiological studies showed that cancer risk in

Abbreviations: COX-2, cyclo-oxygenase-2; SYNCAN, Synbiotics and Cancer Prevention in Humans.

* Corresponding author: Dr J. J. Rafter, fax + 46 8 711 66 59, email joseph.rafter@mednut.ki.se
British Journal of Nutrition (2002), 88, Suppl. 2, S225S232 DOI: 10.1079/BJN2002687
q International Life Sciences Institute 2002

Biomarkers of bone health appropriate for evaluating functional

foods designed to reduce risk of osteoporosis

Connie M. Weaver1* and Michael Liebman2

1
Department of Foods and Nutrition, Purdue University, West Lafayette, IN 47907-1264, USA
2
Department of Family and Consumer Sciences (Human Nutrition), University of Wyoming, Laramie, WY 82071-3354, USA

Osteoporosis is a growing global problem. The health care costs and decreased productivity and
quality of life are staggering. Much research is invested in life-style approaches to build peak
bone mass during growth to prevent osteoporosis as well as to treat the disease in later life.
Functional foods have enjoyed a niche in bone health. Foods fortified with Ca are most popular.
Other bone nutrients such as vitamin D, Mg and vitamin K are sometimes added. Future
products are likely to include enhancers of Ca absorption such as inulin or whey proteins.
Dietary factors that reduce urinary Ca loss (plant proteins) or suppress bone resorption (possibly
phyto-oestrogens) are also gaining attention. Methodologies for evaluating the effectiveness of
functional foods on bone health include measures of bone quality such as bone densitometry or
measures of Ca metabolism, particularly absorption. Biochemical markers for bone turnover are
less satisfactory for diet-related effects. Use of a rare isotope, 41Ca, and accelerator mass
spectrometry offers a new approach for assessing the ability of functional foods to suppress
bone resorption.

Osteoporosis: Bone health: Food fortification: Biochemical biomarkers: Calcium

Introduction intakes adequate for bone accretion and maintenance.

There are many other health reasons to achieve adequate
Biomarkers to evaluate the effectiveness of functional
Ca intake. The role of Ca and dairy product consumption
foods for bone health are quite advanced compared with
in vascular tone and blood pressure control was reviewed
many other diseases. Yet, simple, inexpensive and rapid
recently (McCarron & Reusser, 1999). Observational
methods for evaluating large numbers of people are still
studies suggest adequate Ca decreases risk of kidney
lacking. The appropriate choice of an outcome measure
stones, probably by decreasing urinary oxalate level
depends on the mechanism of action of the functional
(Heller, 1999). Increasing Ca and dairy food intakes
food and the targeted population. Functional foods
appear to reduce risk of colon cancer, probably through
designed to prevent osteoporosis may work by providing
lowering faecal bile acid and fatty acid concentrations,
a key nutrient important to bone development and mainten-
which lowers cytotoxity (Holt, 1999). The reversal of
ance, by enhancing Ca absorption or retention, by building
dietary fat-induced hyperplasia and hyperproliferation in
peak bone mass or by suppressing bone turnover and,
mammary and colonic tissues in mice when Ca and vitamin
therefore, bone loss. Because functional foods for bone
D were supplemented led to the recent suggestion that Ca
health are largely targeted towards increasing Ca intake
is also important in preventing breast cancer (Lipkin &
or utilization of Ca, the focus of this review will be on
Newmark, 1999). A multi-centre trial showed that Ca
strategies to improve Ca nutrition and methods to evaluate
supplementation significantly (P , 0.05) decreased pre-
their effectiveness.
menstrual symptoms relative to those observed in a
placebo-controlled group (Thys-Jacobs et al. 1998). More
recently, adequate Ca intake has been associated with
Importance of adequate dietary calcium
maintenance of body fat and body weight (Lin et al.
There is a renewed interest in the importance of adequate 2000; Zemel et al. 2000).
Ca intake, because of a greater appreciation of its role in Ca intakes for most groups over the age of 11 years fall
health and disease prevention. The role of Ca in bone short of the recommended intakes in North America (Food
health is best understood as the primary cation in bone and Nutrition Board, 1997) and many other parts of the
(Bryant et al. 1999). Requirements of Ca are based on world. Ca absorption and retention are inefficient. Ca

* Corresponding author: Dr C. M. Weaver, fax +1 765 494 0674, email weavercm@cfs.purdue.edu

British Journal of Nutrition (2002), 88, Suppl. 2, S233S235
q International Life Sciences Institute 2002
Functional foods, trends and future
J. A. Weststrate*, G. van Poppel and P. M. Verschuren
Unilever Health Institute, Olivier van Noortlaan 120, NL-3133 Vlaardingen, The Netherlands
The concept of functional foods is often considered to have
emerged in Japan in the late 1980s. However, functional
foods actually have a quite long history. In China, Japan
and other Asian countries, many types of foods have trad-
itionally been associated with specific health benefits. In
Western societies, a well-known global brand, namely
Coca-Cola, actually started as a functional food at the
beginning of the twentieth century. What is probably of
more recent origin is the development of nutritional
science backing functional foods. During the second half
of the twentieth century, exciting new nutritional insights
emerged that allowed for the development of foods and
beverages with a claimed health benefit, based on scientific
evidence.
From a market perspective, functional foods are difficult
to quantify because different definitions are used. If we
view functional foods as foods that make specific health
claims, then the market in the USA, Japan and Europe is
estimated to be worth about e7 billion. The functional
foods market can, however, be seen as part of a broader
health-based/driven foods market. This includes natural
and organic foods, low and lite, weight management
and vitamin- and mineral-fortified products as well as func-
tional foods. This latter market is very large with a global
estimate of around e95 billion for the year 2000. The most
popular foods are energy/sports drinks, probiotic dairy prod-
ucts, heart health spreads and ready-to-eat cereals. Major
functionality claims are for gut health (especially in
Japan and Europe), heart health (especially in the USA
and Europe), promoting natural defences, and boosting
energy levels. The outlook for the future for health-
based/driven foods is bright. An overall growth rate of
10 % per annum for the next five years is possible, signifi-
cantly outperforming the overall foods and beverage mar-
kets growth of about 2 % per annum.
To get back to the history of functional foods, let us
elaborate one of the first examples of a functional food
with a specific health claim. In the late 1960s, Unilever
developed spreads high in polyunsaturated fatty acids
aimed at reducing blood cholesterol level. Spreads under
the Becel and Flora brands were developed based on
mechanistic understanding and clinical efficacy data
(Keys et al. 1965) and were targeted initially at hypercho-
lesterolaemic patients. These products developed gradually
Abbreviations: FUFOSE, Functional Food Science in Europe.
* Corresponding author: Dr J. A. Weststrate, fax +31 10 460 5993, email jan.weststrate@unilever.com
DOI: 10.1079/BJN2002688
into successful mainstream consumer products. This
example shows that functional foods can indeed help in
addressing specific consumer needs and can actually con-
tribute to improvement of public health. In many European
countries, increasing sales of products enriched with poly-
unsaturated fatty acids were paralleled by more favourable
ratios of polyunsaturated to saturated fat in national diets,
as advocated by many nutritional authorities around the
globe (National Heart Foundation of Australia, 1999;
Health Council of the Netherlands, 2001). An example of
the impact of health-driven foods without a specific
health claim is in the low and lite area. In The Nether-
lands, there has been a decrease in fat consumption in
recent years. Although the decrease is moderate compared
with the recommendations of health authorities, data from
food consumption surveys (Anonymous, 1988; Anon-
ymous, 1998) show that the decease is due to changes in
the composition of industrial products, rather than to
changes in food choices or preferences of consumers.
More recently, the arena of cholesterol-lowering spreads
was extended further by products with increased efficacy
through the addition of plant sterols or stanols (Law,
2000). This again demonstrates that developments in nutri-
tional science can be incorporated successfully into func-
tional foods, if there is sufficient consumer awareness
and need. However, sound science and perceived consumer
need from a public health perspective are no guarantee for
the success of functional food products, as unfortunate
examples of unsuccessful market introductions show. It is
hard to predict the prerequisites for success of a functional
food, but factors undoubtedly include consumer need and
awareness; consumer acceptance of a food solution; power-
ful communication of health benefits to the consumer;
uncompromised taste; optimal convenience; adequate
retail or out-of-home availability; proven safety and effi-
cacy; acceptable price level; assurance and support from
different sources, including scientific opinion leaders; and
a clear regulatory framework for making claims, providing
a level playing ground for all companies. The reader may
be disappointed that nutritional science does not feature
prominently in this list of prerequisites, but it is a condition
sine qua non. The list does demonstrate that, from the per-
spective of industrial research and development but prob-
ably also from the public health perspective, research in
S234 J. A. Weststrate et al.
functional foods will only pay off if it is closely integrated
with the other prerequisites for success of a functional
food.
The functional foods that we find on the market today
are often based on general discoveries in nutritional
science, and less on a deliberate research strategy to
develop functional foods. Often, research was product-
driven, i.e. aimed at identifying positive aspects of existing
products and functional foods were then developed by opti-
mising these positive aspects. For the future, an approach
that integrates insights into consumer needs and demands
(market pull) and a structured scientific research process
(science push) will give the largest chance of real inno-
vations. The scientific research process into functional
foods will be powered by technology and insights available
from other disciplines, such as informatics, pharmacology,
engineering, proteomics and genomics. Insights into gen-
etic susceptibility arising from the human genome project
and possibilities for individualised health monitoring
using micro-engineering and sensor technology may alter
dramatically the way in which we deal with food and
health. For this, however, consumer acceptance is a first
prerequisite. This is also true for the attainments of
modern biotechnology that may also have a major impact
on possibilities to optimise functional foods.
Schematically speaking, the combination of market
pull and science push in functional foods research will
result in a research funnel starting from consumer needs
and narrowing down to the final functional foods products
by the following stepwise approach:
1. Consumer understanding: what kind of health benefits
in foods or technology solutions do consumers really
want?
2. Bio-informatics: what molecules could do the job?
3. In vitro screening and in vivo testing: which molecules
work best in model systems?
4. Bioavailability: is the bioactive compound digested
and absorbed?
5. Functional food technology: can we source the
ingredient and make an attractive food?
6. Biomarkers: can we measure relevant effects in man?
7. Human intervention studies: does it really work?
8. Communication: how do we explain the benefits?
With regard to biological benefits in functional foods,
the International Life Sciences Institutes concerted
action on Functional Food Science in Europe (FUFOSE;
Diplock et al. 1999) has proposed six broad groups that
are considered relevant from a scientific perspective.
These are (1) growth, development and differentiation;
(2) substrate metabolism; (3) defence against reactive oxi-
dative species; (4) the cardiovascular system; (5) gastroin-
testinal physiology and function; and (6) behaviour and
psychological functions. These benefit categories indicate
that functional foods should primarily be aimed at function
improvement or (longer-term) disease risk reduction for
healthy people, and not at disease treatment for sick
people.
The broad categories identified by FUFOSE are still
relevant, but the products that have been introduced on
the market in recent years seem to indicate that although
there is a place for products specifically aimed at disease
reduction, there also is a trend towards products providing
daily health benefits. Such functional benefits like healthy
attractive skin, the ability to deal with stress and mental
and physical performance may not always be exciting
from a public health or medical science perspective, but
they are very relevant for the consumer. In addition,
providing a daily benefit can indeed contribute to longer-
term health objectives. For instance, a daily benefit
like satiety and all necessary nutrients from a low-cal-
orie meal replacement has been shown to contribute to
ability to lose weight and maintain a healthy dietary
pattern. If products claim daily health benefits, then
these will probably be in the area of functional ben-
efits. It should be clear that the same stringent criteria
for claim substantiation should be applied for functional
benefit claims as for disease reduction claims. As a
result, there is a challenge in further developing the
science to underpin functional food development in
these areas.
Returning to the market for functional foods: what
sectors will grow and what trends will rule in the next
five years? We predict further growth of three distinct
sectors. A first sector will be products making claims
backed by extensive scientific research for which endorse-
ment is key, especially in the area of heart health and
weight management. Examples have been given pre-
viously. A second sector, also with scientific backing
but focusing more on daily benefit claims, will target
enhancement of physical and mental performance, with
sports and energy drinks becoming even more popular.
A third sector will be focused around general good for
you products that will make more general claims,
based on scientific evidence around ingredients being
used such as probiotics and antioxidants, but with less
extensive research on specific product claims. For all of
these sectors, the range of success factors mentioned pre-
viously applies. Three of these success factors seem to be
paramount: taste, convenience and trust. First, consumers,
except for a small minority, will not be willing to sacrifice
taste for health. Secondly, products must fit consumers
busy life-styles: ready-to-eat or ready-to-heat and on-
the-go formats will become more popular. Third, we
expect a more stringent regime for claims; such that
industry is only allowed to communicate what it has evi-
dence for. Consumers want to be able to trust claims. In a
world with an abundance of choices we increasingly
expect brands to become beacons of trust and guidance
for consumers. Around these brands marketers will aim
to build better relationships with consumers by providing
on-line information and advice, and perhaps service
options. Although there has been speculation about phar-
maceutical companies taking an interest in functional
foods, we foresee that for the next five years the sector
will be dominated by, in particular, diversified food con-
glomerates and the dairy sector. Marketed products will
be based on the findings of nutrition science of the last
decade with a few new options added. Only when we
take an outlook of, say, ten years may we expect that
some functional foods will be marketed that take into
account gene polymorphisms that may be relevant for
 Trends and future
the development of chronic diseases. Also in that time
frame we may see a comeback in Europe of ingredients
for functional foods produced by modern biotechnology.
Apart from the developed market, we expect that in devel-
oping markets the sector of health-based/driven foods will
grow. This is already occurring in the form of vitamin and
mineral fortification of popular foods for the general
population, but quite soon we may also see the emergence
of more advanced functional foods, e.g. targeting heart
health or the increasing prevalence of overweight in
developing countries.
The area of functional foods is beginning to come of
age. Its adolescence will be driven by exciting new
scientific developments. However, the area of functional
foods will only grow successfully if we are able to
integrate credible science with thorough consumer under-
standing, uncompromised taste and convenience, and
effective communication. A key challenge to ensure the
bright future of functional foods is to provide solid
guarantees to consumers that they can trust the safety of
functional foods and their promises about better health,
performance, development or growth.
S235
References
Anonymous (1988) Wat eet Nederland. Resultaten van de voed-
selconsumptiepeiling 1987 1988. Rijswijk: Ministerie van
Welzijn, Volksgezondheid en Cultuur.
Anonymous (1998) Zo eet Nederland 1998. Resultaten van de voed-
selconsumptiepeiling 19971998. Den Haag: Voedingscentrum.
Diplock AT, Aggett PJ, Ashwell M, Bornet F, Fern EB &
Roberfroid MB (1999) Scientific concepts of functional foods
in Europe: consensus document. British Journal of Nutrition
81, Supp. 1, S1 S27.
Health Council of the Netherlands (2001) Dietary Reference
Intakes: Energy, Proteins, Fats and Digestible Carbohydrates.
Publication 2001/19. The Hague: Health Council of the
Netherlands.
Keys A, Anderson JT & Grande F (1965) Serum cholesterol
response to changes in the diet: IV. Particular saturated fatty
acids in the diet. Metabolism 14, 776 786.
Law M (2000) Plant sterol and stanol margarines and health. Brit-
ish Medical Journal 320, 861 864.
National Heart Foundation of Australia (1999) A review of the
relationship between dietary fat and cardiovascular disease.
Australian Journal of Nutrition and Dietetics 56, Suppl. 4,
S5 S22.
S226 C. M. Weaver and M. Liebman

absorption in adults averages only 30 % (Heaney et al.
1988) and losses through endogenous secretions approxi-
mate 120 mg/d regardless of intake (Wastney et al.
1996). Early man is thought to have consumed liberal
amounts of Ca compared with the intake of modern man
(Eaton & Konner, 1985). Because osteoporosis and the
other diseases described above typically occur post repro-
duction, there is little evolutionary pressure to adapt to
the current low intake of Ca. Increasing Ca intake is a
prudent solution to the Ca deficit, but improving absorption
and retention could also improve Ca nutriture.
Absorption: mechanisms and methodology
Ca bioavailability is frequently equated to Ca absorption.
Absorption is the first barrier to achieving Ca homeostasis.
Dietary factors that affect excretion or bone resorption, i.e.
net retention, will be discussed in later sections. Ca is
absorbed both by active, transcellular and by passive,
paracellular processes (Fig. 1). Active absorption domin-
ates at low Ca intake, but owing to its saturable nature
and subsequent down-regulation at adequate intake, this
pathway becomes less important with increasing Ca
intake. Ca absorption efficiency is inversely related to Ca
load over a wide range of intake, although the absolute
quantity of Ca absorbed increases with increased load
(Heaney et al. 1990b).
Methods to assess Ca bioavailability include Ca balance,
determination of bone (or whole-body) Ca retention, and
the use of intestinal loops, Caco-2 cells and isotopic
tracers. These span studies in man, animal models and in
vitro techniques. Ca balance studies are expensive and
can lead to erroneous results unless sufficient attention is
paid to ensuring the completeness of food intake and
urine and faecal collections. An additional challenge is
the demarcation of faeces to time intervals that can be

Fig. 1. Calcium absorption from intestinal lumen through intestinal epithelium to blood. Calcium can be absorbed (1) transcellularly, by a
process mediated by 1,25(OH)2 vitamin D-induced carrier, calbindin D9, or by endocytosis of acidic lysosomes; or (2) paracellularly, a passive
process.
 Bone health
related to specific Ca intake; validity is improved markedly
in this regard when faecal results are corrected for recovery
of a quantitative faecal marker (Hargreaves & Rose, 1965).
Appropriate uses of balance studies include comparisons of
the bioavailability of different sources of Ca and to help
quantify complex nutrient interactions which affect bio-
availability (Mertz, 1987). Ca balance studies have contrib-
uted to the body of literature that suggests that both oxalate
(Kelsay & Prather, 1983; Liebman & Doane, 1989) and
phytate (Morris & Ellis, 1985; Liebman & Landis, 1989)
are important inhibitors of Ca absorption. Balance studies
can determine net Ca absorption, but not true absorption.
In animal models, the bone (femur) uptake method
compares femur accumulation of an oral dose of an iso-
tope tracer with a dose injected intraperitoneally. This
method is based on the assumption that the intraperito-
neal injection behaves as an oral dose with 100 %
absorption. With use of the whole-body counting tech-
nique, absorption of Ca from an oral dose of an isotopic
tracer is based on extrapolating the linear portion of a
retention curve to time zero (Koo et al. 1993). Koo
et al. (1993) demonstrated that assessments of radiotracer
activity of Ca in femur or whole body are equally accu-
rate for comparing bioavailability of Ca among sources.
Not having to collect excreta or multiple blood samples
is a clear advantage of these methods.
Ca absorption can also be assessed with use of the in
situ loop and everted sac methods. Studies based on
these methods clearly established the existence of two
Ca absorptive processes, a vitamin D-dependent, satur-
able, transcellular pathway that predominates in the duo-
denum and a non-saturable, paracellular pathway that
occurs throughout the entire length of the small intestine
(Pansu et al. 1983; Bronner et al. 1986). However, they
are insensitive to the important factor of exposure time
of the chyme to the mucosa.
Pinto et al. (1983) demonstrated that cultured Caco-2
cells, a human colon adenocarcinoma cell line, exhibit
structural and functional differentiation patterns character-
istic of mature enterocytes. Yee (1997) reported a strong
correlation between in vitro permeability across Caco-2
cells and in vivo (small intestinal) absorption for a variety
of compounds. It has also been confirmed recently that
vitamin D-mediated Ca transport in these cells is a specific,
transcellular process that requires transcriptional events
normally mediated through the vitamin D receptor (Fleet
& Wood, 1999). However, in culture, these cells form a
membrane with tight junctions that interfere with
measurement of transport.
Some have tried to use rise in urinary Ca to determine
relative bioavailability, but this is a weak approach
especially for Ca loads less than 500 mg. For example,
variability in rise in urinary Ca was 77 99 % compared
with 38 60 % for serum Ca following ingestion of three
Ca salts (Heaney et al. 2001). Increments in serum Ca
are also less sensitive than following the fate of isotopic
tracers of Ca. Isotopic tracers have been used to label
dietary sources of Ca for the measurement of Ca bioavail-
ability in both animal models and man. Ca has two useful
radioisotopes (45Ca and 47Ca), several useful stable iso-
topes (42Ca, 43Ca, 44Ca, 46Ca and 48Ca), as well as one
S227
long-lived radioisotope (41Ca). Tracers can be followed
by appearance in plasma, excretion or by whole-body
retention in the case of 47Ca. Use of tracers for determining
Ca absorption gives true absorption, in contrast to net
absorption determined by balance. Intrinsic labelling of a
broad array of hydroponically grown plants, milk made
into a variety of dairy products and salts has enabled deter-
mination of Ca absorption in man from most Ca-rich food-
stuffs in the Western diet (Weaver et al. 1999). Isotopic
tracers can also be used to determine transfer rates, sites
of absorption and pathway of absorption. When absorption
is plotted v. various Ca loads, a curvilinear relationship
suggests active transport and a linear relationship suggests
paracellular absorption (see Fig. 2).
Modifying calcium absorption
In a recent review, Bronner & Pansu (1999) stated: It is
obvious that calcium must be ionized and in solution to
be absorbed. Typically, Ca must be dissociated from its
ligands in a foodstuff prior to absorption. However, there
is recent evidence that Ca from a small-molecular-weight
compound did not require dissociation prior to absorption
(Hanes et al. 1999a,b). This was apparent because Ca
and oxalic acid have very different serum appearance pro-
files, yet doubly labelled 45Ca-[14C]oxalate showed parallel
serum profiles of the two labels until the unabsorbed salt
reached the colon, where bacterial hydrolysis presumably
occurs. It is tempting to conclude that absorption of the
salt occurred paracellularly, but calcium oxalate could con-
ceivably transverse the epithelial membrane because it is
apolar. The implications of this finding might be greatest
for individuals with defective active Ca absorption
capacity.
Solubilization is a reasonable assumption prior to Ca
absorption, at least at the absorptive surfaces. Yet, in
vitro solubility of Ca salts at neutral pH over a wide
range had little impact on Ca absorption (Heaney et al.
1990a). For example, Ca absorption from CaCO3 with a
solubility of 014 mmol/l is as good as from tricalcium
phosphate, which is nearly an order of magnitude more
soluble at 097 mmol/l. Only salts that are at the extreme
Fig. 2. Illustration of (A) saturable and (B) non-saturable
components of calcium kinetics curves. At large calcium loads,
active transport approaches a small value. The linear portion of the
curve reflects constant and paracellular absorption.
S228 C. M. Weaver and M. Liebman

ends of solubility have appreciably different Ca absorption
efficiencies.
Other factors that influence Ca bioavailability include
the presence of inhibitors and enhancers in the diet.
Functional foods for bone health may decrease the con-
tent of inhibitors and increase the content of enhancers.
The most potent inhibitor of Ca absorption is oxalate
(Heaney & Weaver, 1989). Fibre has been labelled as
an inhibitor of Ca absorption since the early balance
studies of McCance & Widdowson (1942). However,
research with purified fibres has suggested that fibre
does not influence Ca absorption appreciably (Heaney
& Weaver, 1995). This is supported by the finding that
Ca absorption efficiency from fibre-rich brassica
vegetables is higher than from other foodstuffs (Weaver
et al. 1999). The presence of an enhancer of Ca absorp-
tion in brassica vegetables is suspected, but not yet
identified. In contrast to the lack of fibre effect, phytates
associated with cereal and legume fibres can decrease Ca
absorption (Heaney et al. 1991; Weaver et al. 1991). The
capacity of phytate-rich extruded wheat bran cereal to
bind Ca is linear over a wide range of Ca intake
(Weaver et al. 1996).
Few Ca absorption enhancers have been identified.
Most research has concentrated on searching for highly
absorbable Ca salts. Only a few have been identified,
including calcium citrate malate with a solubility of
80 mmol/l (Miller et al. 1988), calcium gluconate glycero-
phosphate (Schanler & Abrams, 1995) and calcium
ascorbate (Tsugawa et al. 1999). Additionally, a few Ca
absorption enhancers have been identified. Consumption
of 15 g oligofructose/d increased stable isotopic tracer Ca
absorption from 478 % during a placebo period to
601 % (van den Heuvel et al. 1999). Feeding of 40 g
inulin/d increased apparent Ca absorption in adults
participating in a balance study from 213 to 337 %
(Coudray et al. 1997). Certain amino acids, notably
lysine, and casein phosphopeptides, digestive products
from milk proteins, have Ca absorption-enhancing effects
under some conditions such as in women with low absorp-
tion efficiency (Heaney et al. 1994). Yet another type of Ca
absorption enhancer is the hydrolysis product of phytic
acid, 1,2,3,6-inositol tetrakisphosphate, which improved
45
Ca absorption efficiency from 262 to 307 % in rats
using calcium ascorbate as the Ca source (Shen et al.
1998).
Enhancement of Ca absorption has typically been
attributed to the formation of soluble complexes with Ca
which prevent precipitation by P in the gut. Other mechan-
isms of enhancing Ca absorption deserve to be explored.
Increasing paracellular absorption is promising because it
is not limited by becoming saturated, it is vitamin D-
independent, and it occurs throughout the length of the
intestine in contrast to active absorption which is dominant
in the duodenum. If the intercellular junction spaces
illustrated in Fig. 1 can be widened, more Ca could be
absorbed. Or, if solvent drag could be increased, even
though water would flow bidirectionally, given the large
volume of blood that would serve to dilute ions extruded
from the basolateral membrane, ions would have a net
movement from lumen to blood. Ideal compounds would
be those that could be incorporated into Ca-containing
food to enhance absorption of Ca but would have only a
transient effect, so that transfer of undesirable organisms
and ions would be minimized. Pappenheimer & Reiss
(1987) provided evidence that glucose and amino acids
in the lumen of the small intestine increase solvent drag
through paracellular channels by fuelling the contraction
of epithelial cytoskeletal elements, thereby opening tight
junctions to allow mass transport of nutrients.

Fig. 3. Relationship between calcium retention and calcium intake in adolescent

Caucasian girls (mean and 95 % confidence intervals). Calcium intakes required to
achieve: (A) 70 % of maximal retention at the lower edge of the confidence interval; (B)
100 % of maximal retention at the upper edge of the confidence interval; and (C)
100 % of maximal retention at the mean.
 Bone health
Modifying calcium retention
Dietary factors that alter Ca absorption or Ca excretion can
modify Ca retention. Nutrients with the greatest impact on
Ca loss include salt and protein, both of which increase
urinary Ca output. Ironically, Ca intake influences urinary
Ca output only modestly; only 6 % of the variance in
urinary Ca in adolescents was explained by Ca intake
(Jackman et al. 1997). Data on the effect of dietary
variables on Ca retention are collected largely in adults.
We know little of these relationships across the life span
or of the impact of race or other genetic determinants.
Perhaps a more useful way of determining the impact of
dietary variables on Ca retention is to determine how Ca
requirements might be affected by various dietary patterns
and for different populations. The relationship between Ca
retention and Ca intake has been examined for various
populations (Matkovic & Heaney, 1992). For all popu-
lations, there is a linear relationship up to a certain
intake after which there is a levelling off of the curve so
that further increases in Ca intake produce little further
gain in retention. The intake at which this plateau occurs
is a reference value for setting requirements to maximize
development of peak bone mass during growth and minim-
ize bone loss during ageing. There is no consensus on
whether requirements should be set at 100 % of maximum
retention or at some lower target. Nor is it clear how to
determine intake for a given percentage retention taking
into account variability. Figure 3 shows the mean and
95 % confidence interval of the relationship between Ca
retention and Ca intake in adolescent girls using the data
of Jackman et al. (1997). Should the Ca requirement be
set at extrapolated value A (70 % of maximal retention
taken at the lower limit of the 95 % confidence interval),
value B (100 % maximal retention taken at the lower
limit of the 95 % confidence interval), value C (the mean
Fig. 4. The curve of calcium retention v. calcium intake shifts to the left with high
bioavailability, low-salt and low-protein diets. A is the original intake to achieve maximal
retention at the upper edge of the confidence interval and B is the new intake when the
curve is shifted to the left.
S229
100 % maximal retention), or some other value? The first
value is arbitrary. The second value is dependent on
sample size and inter-subject variability, which influences
the confidence interval. The third value may be
impractically high.
Regardless of how the relationship of Ca retention to Ca
intake is used to set requirements, knowing the relationship
is the first step. Understanding how the relationship can be
influenced by dietary life-style and genetic factors is the
next area of needed research. Theoretically, dietary pat-
terns which include high bioavailable Ca, low salt and
low protein can shift the curve to the left (Fig. 4), whereas
diets characterized by low Ca bioavailability, high salt and
high protein can shift the curve to the right (Fig. 5). Simi-
larly, the curve might be shifted according to race, ethnic
group, physical activity level, smoking and other factors.
Modifying bone turnover
Functional foods that promote health by modifying bone
turnover may work by enhancing bone formation or
suppressing bone resorption. Increasing Ca intake during
adolescence results in increased Ca absorption, which
suppresses bone resorption by the equivalent amount
(Wastney et al. 2000). Phyto-oestrogens may provide
some protection against bone resorption similar to oestro-
gen, although this is not yet well documented (Weaver
et al. 2001). Suppression of bone resorption can lead to
increased Ca retention.
Use of Ca isotopic tracers and kinetic analysis allows
quantification of bone formation and bone resorption in
units of Ca such as mg of Ca per day. Use of just one
dose of the rare isotope, 41Ca, and accelerator mass spec-
trometry opens the possibility of determining types of
diet changes that might suppress bone resorption in
S230 C. M. Weaver and M. Liebman

Fig. 5. The curve of calcium retention v. calcium intake shifts to the right with low
bioavailability, high-salt and high-protein diets. A is the original intake to achieve maxi-
mal retention and B is the new intake when the curve is shifted to the right.

individuals followed longitudinally. Approximately two
months after dosing, the appearance of 41Ca in the urine
directly reflects bone resorption. The sensitivity of accel-
erator mass spectrometry allows the tracer to be followed
for years, thereby allowing assessments of changes in
diet or other life-style factors.
Biochemical markers of bone turnover have also been
used to determine qualitative changes in bone turnover
(Weaver, 1998). Some common biochemical markers of
bone formation include serum osteocalcin and bone alka-
line phosphatase. Biochemical markers of bone resorption
are typically urinary crosslinks of collagen. These assays
are not in units of bone or Ca and results are highly vari-
able. Larger sample sizes are required to find significant
treatment effects. Frequently, the subtle effects of diet
cannot be detected.
Bone mineral density
Bone mineral density is a strong biomarker for fracture. To
evaluate the effectiveness of functional foods using this
method, the intervention period needs to be much longer
than for evaluating parameters of Ca metabolism. Ideally,
interventions would have the duration of three to four
sigmas. Each sigma, the period for a complete cycle of
bone resorption and formation, is about four months in
man. Trials of short duration fail to show the long-term
impact on bone health. One study showed the benefit of
one year of Ca-fortified foods on bone measures in growing
children (Bonjour et al. 1997).
availability of an array of Ca-fortified foods and
supplements, intake of Ca remains inadequate. The role
of Ca bioavailability and the ability to maximize absorp-
tion and retention are less important under conditions of
adequate intake. A variety of methods are available to
evaluate the effectiveness of functional foods for bone
health. Their effect on Ca absorption can be determined
by in vitro techniques, use of animal models or in man.
Use of Ca isotopic tracers offers a specific way to
determine the point of Ca metabolism affected. Under-
standing mechanisms of Ca absorption and how to increase
absorption and retention efficiency are also important.
In the next decade, we will understand better the dietary
and life-style factors that influence Ca absorption by both
transcellular and paracellular routes. We will understand
better how dietary factors influence the relationship
between Ca retention and Ca intake. We will understand
better the genetic factors that influence Ca absorption and
retention. That Ca retention has a large genetic component
was demonstrated by the greater response to Ca of three
generations of women who were from osteoporotic
families compared with women from healthy families
(OBrien et al. 1998). A beginning in identifying a specific
gene that may influence Ca absorption is the significant
association of vitamin D receptor gene Fok1 polymorphism
with Ca absorption and bone mineral density in children
aged 75 12 years (Ames et al. 1999). Perhaps one day
we will be able to tailor the food supply for identifiably
vulnerable segments of the population.

Conclusions
Acknowledgements
Increasing the Ca intake of the general population is the
most effective strategy for using functional foods for This work was support by grants PHS AR 40553 and HD
bone health. Despite education efforts and the increased 36609.
 Bone health
References
Ames SK, Ellis KJ, Gunn SK, Copeland K & Abrams SA (1999)
Vitamin D receptor gene Fok1 polymorphism predicts calcium
absorption and bone mineral density in children. Journal of
Bone and Mineral Research 14, 740746.
Bonjour JP, Carrie AL, Ferriri S, Clavien H, Slosman D, Thientz
G & Rizzoli R (1997) Calcium-enriched foods and bone mass
growth in prepubertal girls a randomized, double-blind,
placebo-controlled trial. Journal of Clinical Investigation 99,
12871298.
Bronner F & Pansu D (1999) Nutritional aspects of calcium
absorption. Journal of Nutrition 129, 9 12.
Bronner F, Pansu D & Stein WD (1986) An analysis of intestinal
calcium transport across the rat intestine. American Journal of
Physiology 250, G561 G569.
Bryant RJ, Cadogan J & Weaver CM (1999) The new Dietary
Reference Intakes for calcium: implications for osteoporosis.
Journal of the American College of Nutrition 18, 406S 412S.
Coudray C, Bellanger J, Castiglia-Delavaud C, Remesy C,
Vermorel M & Rayssignuier Y (1997) Effect of soluble or
partly soluble dietary fibres supplementation on absorption
and balance of calcium, magnesium, iron and zinc in healthy
young men. European Journal of Clinical Nutrition 51,
375380.
Eaton SB & Konner M (1985) Paleolithic nutrition: a con-
sideration of its nature and current implications. New England
Journal of Medicine 312, 283 289.
Fleet JC & Wood RJ (1999) Specific 1,25(OH)2D3-mediated
regulation of transcellular calcium transport in Caco-2 cells.
American Journal of Physiology 276, G958 G964.
Food and Nutrition Board (1997) Dietary Reference Intakes for
Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride.
Standing Committee on the Scientific Evaluation of Dietary
Reference Intakes, Food and Nutrition Board, Institute of
Medicine. Washington, DC: National Academy Press.
Hanes DA, Weaver CM, Heaney RP & Wastney ME (1999a)
Absorption of calcium oxalate does not require dissociation
in rats. Journal of Nutrition 129, 170 173.
Hanes DA, Weaver CM & Wastney ME (1999b) Calcium and
oxalic acid kinetics are different in rats. Journal of Nutrition
129, 165169.
Hargreaves T & Rose GA (1965) The reproducibility of the
balance method in man as applied to calcium and phosphorus.
Clinical Science 28, 537 542.
Heaney RP, Dowell S, Bierman J, Hale C & Bendich A (2001)
Absorbability and cost effectiveness in calcium supple-
mentation. Journal of the American College of Nutrition 20,
239246.
Heaney RP, Recker RR & Hinders SM (1988) Variability of
calcium absorption. American Journal of Clinical Nutrition
47, 262264.
Heaney RP, Recker RR & Weaver CM (1990a) Absorbability of
calcium sources. The limited role of solubility. Calcified Tissue
International 46, 300 304.
Heaney RP, Saito Y & Orimo H (1994) Effect of caseinphos-
phopeptide on absorbability of co-ingested calcium in normal
postmenopausal women. Journal of Bone and Mineral
Metabolism 12, 77 81.
Heaney RP & Weaver CM (1989) Oxalate: effect on calcium
absorbability. American Journal of Clinical Nutrition 50,
830832.
Heaney RP & Weaver CM (1995) Effect of psyllium on absorp-
tion of co-ingested calcium. Journal of American Geriatric
Society 43, 1 3.
Heaney RP, Weaver CM & Fitzsimmsons ML (1990b) The
S231
influence of calcium load on absorption fraction. American
Journal of Clinical Nutrition 5, 1135 1138.
Heaney RP, Weaver CM & Fitzsimmons ML (1991) Soybean
phytate content: effect on calcium absorption. American
Journal of Clinical Nutrition 53, 745 747.
Heller JH (1999) The role of calcium in the prevention of kidney
stones. Journal of American College of Nutrition 18,
373S 378S.
Holt PR (1999) Dairy foods and prevention of colon cancer:
human studies. Journal of the American College of Nutrition
18, 379S 391S.
Jackman LA, Millane SS, Martin BR, Wood OB, McCabe GP,
Peacock M & Weaver CM (1997) Calcium retention in
relation to calcium intake and postmenarcheal age in adoles-
cent females. American Journal of Clinical Nutrition 66,
327 333.
Kelsay JL & Prather ES (1983) Mineral balances of human
subjects consuming spinach in a low-fiber diet and in a diet
containing fruits and vegetables. American Journal of Clinical
Nutrition 38, 12 19.
Koo J, Weaver CM, Neylan MJ & Miller GD (1993) Isotopic
tracer techniques for assessing calcium absorption in rats.
Journal of Nutritional Biochemistry 4, 72 76.
Liebman M & Doane L (1989) Calcium and zinc balances during
consumption of high and low oxalate-containing vegetables.
Nutrition Research 9, 947955.
Liebman M & Landis W (1989) Calcium and zinc balances of
premenopausal women consuming tofu- compared to cheese-
containing diets. Nutrition Research 9, 5 14.
Lin Y-C, Lyle RM, McCabe LD, McCabe GP, Weaver CM &
Teegarden D (2000) Calcium intake relates to change in body
weight in young women. Journal of the American College of
Nutrition 19, 754760.
Lipkin M & Newmark HL (1999) Vitamin D, calcium and pre-
vention of breast cancer: a review. Journal of the American
College of Nutrition 18, 392S 397S.
McCance RA & Widdowson EM (1942) Mineral metabolism of
healthy adults on white and brown bread dietaries. Journal of
Physiology 101, 44 85.
McCarron DA & Reusser ME (1999) Finding consensus in the
dietary calcium blood pressure debate. Journal of the Ameri-
can College of Nutrition 18, 398S 405S.
Matkovic V & Heaney RP (1992) Calcium balance during human
growth: evidence for threshold behavior. American Journal of
Clinical Nutrition 55, 992 996.
Mertz W (1987) Use and misuse of balance studies. Journal of
Nutrition 117, 1811 1813.
Miller JZ, Smith DL, Flora L, Slemenda C, Juang X & Johnston
CC (1988) Calcium absorption from calcium carbonate and a
new form of calcium (CCM) in healthy male and female
adolescents. American Journal of Clinical Nutrition 48,
1291 1294.
Morris ER & Ellis R (1985) Bioavailability of dietary calcium:
effect of phytate on adult men consuming nonvegetarian
diets. In Nutritional Bioavailability of Calcium, pp. 63 72
[C Kies, editor]. Washington, DC: American Chemical
Society.
OBrien KO, Abrams SA, Liang LK, Ellis KJ & Gagel RF
(1998) Bone turnover response to changes in calcium
intake is altered in girls and adult women in families with
histories of osteoporosis. Journal of Bone and Mineral
Research 13, 491 499.
Pansu D, Bellaton C, Roche C & Bronner F (1983) Duodenal and
ileal calcium absorption in the rat and effects of vitamin D.
American Journal of Physiology 244, G695 G700.
Pappenheimer JR & Reiss KZ (1987) Contribution of solvent drag
through intracellular to absorption of nutrients by the small
S232 C. M. Weaver and M. Liebman
intestine of the rat. Journal of Membrane Biology 100,
123 136.
Pinto M, Robine-Leon S, Appay M, Kedinger M, Triadou N, Dus-
saulx E, Lacroix B, Simon-Assmann P, Haffen K, Fogh J &
Zweibaum A (1983) Enterocyte-like differentiation and
polarization of the human colon carcinoma cell line Caco-2
in culture. Biology of the Cell 47, 323330.
Schanler R & Abrams SA (1995) Postnatal attainment of
intrauterine mineral accretion rates in low birth weight
infants fed fortified human milk. Journal of Pediatrics 126,
441 447.
Shen X, Weaver CM, Kempa-Steczko A, Martin BR, Phillippy
BQ & Heaney RP (1998) An inositol phosphate as a calcium
absorption enhancer in rats. Journal of Nutritional
Biochemistry 9, 298 301.
Thys-Jacobs S, Starkey P, Bernstein D & Tian J (1998)
Premenstrual Syndrome Study Group. Calcium carbonate and
premenstrual syndrome: effects on premenstrual and menstrual
symptoms. American Journal of Obstetrics and Gynecology
179, 444 452.
Tsugawa N, Yamabe T, Takeuchi A, Kamao M, Nakagawa K,
Nishijima K & Okano T (1999) Intestinal absorption of calcium
from calcium ascorbate in rats. Journal of Bone and Mineral
Metabolism 17, 30 36.
van den Heuvel E, Muys T, van Dokskum W & Schaafsma G
(1999) Oligofructose stimulates calcium absorption in
adolescents. American Journal of Clinical Nutrition 69,
544 548.
Wastney ME, Martin BR, Peacock M, Smith D, Jiang X-Y,
Jackman LA & Weaver CM (2000) Changes in calcium
kinetics in adolescent girls induced by high calcium intake.
Journal of Clinical Endocrinology and Metabolism 85,
4470 4475.
Wastney ME, Ng J, Smith D, Martin BR, Peacock M & Weaver
CM (1996) Differences in calcium kinetics between adolescent
girls and young women. American Journal of Physiology 271,
R208 R216.
Weaver CM (1998) Use of calcium tracers and biomarkers to
determine calcium kinetics and bone turnover. Bone 22,
103S 104S.
Weaver CM, Heaney RP, Martin BR & Fitzsimmons ML (1991)
Human calcium absorption from whole wheat products.
Journal of Nutrition 121, 1769 1775.
Weaver CM, Heaney RP, Teegarden D & Hinders SM (1996)
Wheat bran abolishes the inverse relationship between calcium
load size and absorption fraction in women. Journal of
Nutrition 126, 303 307.
Weaver CM, Proulx WR & Heaney RP (1999) Choices for
achieving dietary calcium within a vegetarian diet. American
Journal of Clinical Nutrition 70, 543S 548S.
Weaver CM, Spence LA & Lipscomb ER (2001) Phytoestrogens
and bone health. In Nutritional Aspects of Osteoporosis,
pp. 315 324 [P Burckhardt, B Dawson-Hughes and RP
Heaney, editors]. New York: Academic Press.
Yee S (1997) In vitro permeability across Caco-2 cells (colonic)
can predict in vivo (small intestinal) absorption in man
fact or myth. Pharmaceutical Research 14, 763 766.
Zemel MB, Shi H, Greer B, Direnzo D & Zemel PC (2000)
Regulation of adiposity by dietary calcium. FASEB Journal
14, 1132 1138.
S220 J. J. Rafter
people consuming diets high in fruits and vegetables was
only one-half that in those consuming few of these foods
(Block et al. 1992). It is now clear that there are com-
ponents in a plant-based diet other than traditional nutrients
that can reduce cancer risk. More than a dozen classes of
these biologically active plant chemicals, now known as
phytochemicals, have been identified. Several examples
are given below.
Functional foods from animal sources
Dairy products
Epidemiological studies regarding the effect of milk and
dairy products on colorectal cancer incidence provide
conflicting results. Some studies report a negative associ-
ation between consumption of dairy products and disease
incidence, whereas others report no or even a positive
association. One reason for this discrepancy in the epi-
demiological observations may lie in the fact that some
dairy products, such as cheese and cream, contain large
amounts of animal fat, which is considered a risk factor
by some for colorectal cancer. In line with this assumption,
it appeared that the positive correlation between dairy
products and colorectal cancer found in one study was
due to cheese consumption (Iscovich et al. 1992). In
addition, an international correlation study observed a
negative correlation between consumption of dairy pro-
ducts and colorectal cancer only after adjustment for
animal fat intake (McKeown-Eyssen & Bright-See,
1984). Most of the studies that reported no association
did not adjust for fat intake. Hence, epidemiological
studies provide weak evidence that consumption of milk
and dairy products could protect against colorectal cancer.
The main hypothesis to explain this effect has been the
high Ca content of dairy products. Thus, many previous
studies have focused on Ca and consequently examined
the effects of dietary Ca supplements (Lapre et al. 1993)
or Ca-deprived milk (Govers et al. 1996) on risk markers
for the disease, in controlled dietary intervention studies.
While the results of these studies generally do show a
protective effect of Ca on the biomarkers, this is not
always the case. However, recently, a great deal of
research has been dealing with other components in dairy
products, particularly fermented dairy products, i.e. the
probiotic bacteria.
Many studies confirm the involvement of the endo-
genous microflora in the onset of colon cancer. This
makes it reasonable to think that changing the intestinal
microflora could influence tumour development. Thus,
considerable attention has focused on dietary components
that can influence the gut microflora as a strategy for
colon cancer prevention. In particular, evidence has been
accumulating from animal studies for a protective role
for probiotics, prebiotics (non-digestible carbohydrates
that stimulate lactic acid bacteria numbers in the gut) and
synbiotics (combinations of pro- and prebiotics).
Several studies have indicated that addition of probiotic
cultures (e.g. Lactobacillus rhamnosus GG, Lactobacillus
salivarius or Bifidobacterium longum ) to the diet of rats
treated with colon carcinogens reduces colon tumour
incidence and multiplicity (Hirayama & Rafter, 2000).
The precise mechanisms by which lactic acid bacteria
may inhibit colon cancer are presently unknown. However,
such mechanisms might include: enhancing the hosts
immune response; binding and degrading potential
carcinogens; quantitative and/or qualitative alterations in
the intestinal microflora incriminated in producing putative
carcinogen(s) and promoters (e.g. bile acid-degrading
bacteria); producing anti-tumourigenic or anti-mutagenic
compounds in the colon; alteration of the metabolic
activities of intestinal microflora; alteration of physico-
chemical conditions in the colon; and effects on the
physiology of the host.
A wide range of studies have also been conducted in
rodent models, including transgenic mice, on the cancer-
modulating effect of dietary prebiotics such as fructo-
oligosaccharides, inulins and lactulose. These have
consistently demonstrated a reduction in the endpoints
measured colonocyte DNA damage, aberrant crypt
foci in the colon (believed to be pre-neoplastic lesions),
number of tumours per animal, size and growth rate of
the tumours, mean survival time, increase in life span
(Pierre, 1997). In all animal experiments where mixtures
of probiotics (Bifidobacterium spp., Lactobacillus casei,
L. rhamnosus ) and prebiotics (lactulose, long-chain
inulin, short-chain fructo-oligosaccharides) were tested,
the effect of the mixture on the measured biomarkers
was greater than the sum of the two separately, suggesting
a combinational advantage of synbiotics.
However, to date, there is no definite evidence that
probiotics actually reduce the risk of human colon
cancer. The epidemiological studies are inconsistent
(Hirayama & Rafter, 2000). Consumption of large quan-
tities of dairy products such as yoghurt and fermented
milk containing Lactobacillus or Bifidobacterium may be
related to a lower incidence of colon cancer. An epidemio-
logical study performed in Finland demonstrated that,
despite the high fat intake, colon cancer incidence was
lower than in other countries because of the high consump-
tion of milk, yoghurt and other dairy products (Intestinal
Microbiology Group, 1977). In two population-based
case control studies of colon cancer, an inverse associ-
ation was observed for yoghurt (Peters et al. 1992) and cul-
tured milk (Young & Wolf, 1988) consumption, adjusted
for potential confounding variables. It can also be men-
tioned that an inverse relationship has been demonstrated
between the frequency of consumption of yoghurt and
other fermented milk products and breast cancer in
women (vant Veer et al. 1989). On the other hand, two
companion American prospective studies, the 1980 1988
follow-up of the Nurses Health Study and the 1986
1990 follow-up of the Health Professionals Follow-up
Study, did not provide evidence that intake of dairy pro-
ducts is associated with a decreased risk of colon cancer
(Kampman et al. 1994a). In a cohort study in The Nether-
lands, it was shown that the intake of fermented dairy pro-
ducts was not significantly associated with colorectal
cancer risk in an elderly population with a relatively
wide variation in dairy product consumption, although a
weak non-significant inverse association with colon
cancer was observed (Kampman et al. 1994b).
 Reduction of disease risk: cancer
In addition, there are some studies examining the effect
of probiotics on the biomarkers for colon cancer risk (see
below) in healthy volunteers and patients. Consumption
of lactic acid bacteria by volunteers has been shown to
reduce the mutagenicity of urine and faeces associated
with the ingestion of carcinogens in cooked meat (Lidbeck
et al. 1992). Mucosal cell proliferative activity in upper
colonic crypts of patients with colon adenomas (believed
to be a risk factor for tumour development) decreased
significantly after the administration of Lactobacillus
acidophilus and Bifidobacterium bifidus cultures (Biasco
et al. 1991).
Beef
An anticarcinogenic fatty acid known as conjugated
linoleic acid was first isolated from grilled beef in 1987
(Ha et al. 1987). Nine different isomers of conjugated
linoleic acid have been reported as occurring naturally in
food. Conjugated linoleic acid is unique in that it is
found in highest concentrations in fat from ruminant
animals (e.g. beef, dairy, lamb). In recent years, conjugated
linoleic acid has been shown to be effective in suppressing
forestomach tumours in mice, aberrant colonic crypt foci in
rats and mammary tumours in rats (Ip & Scimeca, 1997).
Functional foods from plant sources
Soyabeans
Several classes of anticarcinogens have been identified in
soyabeans, including protease inhibitors, phytosterols,
saponins, phenolic acids, phytic acid and isoflavones. Of
these, isoflavones (genistein and daidzein) are particularly
noteworthy because soyabeans are the only significant
dietary source of these compounds. Isoflavones are
heterocyclic phenols structurally similar to the oestrogenic
steroids. Because they are weak oestrogens, isoflavones
may act as anti-oestrogens by competing with the naturally
occurring endogenous oestrogens. This may explain why
populations that consume significant amounts of soya
have reduced risk of oestrogen-dependent cancer. How-
ever, more epidemiological data and clinical intervention
trials are needed to investigate the role of soya in reducing
cancer risk.
Flaxseed
There has been an increasing interest in fibre-associated
compounds known as lignans. The two primary mam-
malian lignans, enterodiol and its oxidation product,
enterolactone, are formed in the intestinal tract by bacterial
action on plant lignan precursors. Flaxseed is one of the
richest sources of mammalian lignan precursors. Because
enterodiol and enterolactone are structurally similar to
both naturally occurring and synthetic oestrogens, and
have been shown to possess weakly oestrogenic and
anti-oestrogenic activities, they may also play a role in
the prevention of oestrogen-dependent cancers. In rodents,
flaxseed has been shown to decrease tumours of the colon,
mammary gland and lung. Phipps et al. (1993) demon-
S221
strated that the ingestion of 10 g of flaxseed per day elicited
several hormonal changes associated with reduced breast
cancer risk. However, as is the case with soya, epidemio-
logical data are required to support the hypothesis that
enterodiol and enterolactone have anticarcinogenic
properties in man.
Tomatoes
Tomatoes have received much attention in recent years
because of interest in lycopene, the primary carotenoid in
this fruit, and its potential role in cancer risk reduction
(Weisburger, 1998). In a prospective cohort study of
more than 47 000 men, those who consumed tomato
products ten or more times per week had less than one-
half the risk of developing advanced prostate cancer
(Giovannucci et al. 1995). Interestingly, lycopene is the
most abundant carotenoid in the prostate gland. Other
cancers whose risk have been inversely associated with
serum or tissue levels of lycopene include breast, digestive
tract, cervix, bladder and skin (Clinton, 1998). Proposed
mechanisms by which lycopene could influence cancer
risk are related to its antioxidant function. Lycopene is
the most efficient quencher of singlet oxygen in biological
systems (Di Mascio et al. 1989).
Garlic
Garlic (Allium sativum ) is probably the herb most widely
quoted in the literature for medicinal purposes. The intact
garlic bulb contains an odourless amino acid, which is
converted enzymatically by allinase into allicin when the
garlic cloves are crushed (Block, 1992). Allicin then
decomposes spontaneously to form numerous sulfur-
containing compounds, some of which have been investi-
gated for their chemopreventive activity. Garlic com-
ponents have been shown to inhibit tumourigenesis in
several experimental models. However, additional reports
have shown garlic to be ineffective. Inconclusive results
are likely to be due to differences in the type of garlic
compounds or preparations used by various investigators.
Several epidemiological studies show that garlic may be
effective in reducing human cancer risk (Dorant et al.
1993). However, it should be mentioned that not all such
studies have shown garlic to be protective. A review of
twenty epidemiological studies (Ernst, 1997) suggested
that allium vegetables, including onions, may confer a
protective effect on cancers of the gastrointestinal tract.
Tea
Much attention has focused on the polyphenolic constitu-
ents of tea, particularly green tea. Polyphenols comprise
up to 30 % of the total dry weight of fresh tea leaves.
Catechins are the predominant and most significant of all
tea polyphenols (Graham, 1992). The four major green
tea catechins are epigallocatechin-3-gallate, epigallo-
catechin, epicatechin-3-gallate and epicatechin. Much of
the work on the health effects of tea has focused on its
cancer chemopreventive effects. Research results from labora-
tory animals tend to support a cancer chemopreventive
S222 J. J. Rafter

effect of tea components (Dreosti et al. 1997). However, the Biomarkers available for assessing diet-related
epidemiological studies are still somewhat inconclusive. It changes in colon cancer risk
has been suggested that benefits from tea consumption are
Since colon cancer has been a major target for the cancer
restricted to high intakes in high-risk populations (Kohl-
chemopreventive effects of probiotics and functional food
meier et al. 1997). The consumption of five or more
components, I shall limit my examples of biomarkers to
cups of green tea per day was shown to be associated
this cancer form. Presently, there are a large number of
with decreased recurrence of stage I and stage II breast
biomarkers available for assessing colon cancer risk in
cancer in Japanese woman (Nakachi et al. 1998).
dietary intervention studies, which are validated to varying
degrees. These include colonic mucosal markers, faecal
water markers and immunological markers.
Broccoli and other cruciferous vegetables
Epidemiological evidence has also associated the frequent Colon mucosa biomarkers
consumption of cruciferous vegetables with decreased
1. Adhesion of Gram-negative bacteria
cancer risk. In a review of eighty-seven case control
2. Modulation of cyclo-oxygenase-2 (COX-2)
studies, Verhoeven et al. (1996) demonstrated an inverse
3. Proliferation
association between consumption of total brassica vege-
4. K-ras
tables and cancer risk. The percentages of case control
5. Genetic instability
studies showing an inverse association between consump-
6. Apoptosis
tion of cabbage, broccoli, cauliflower and Brussels sprouts
7. DNA-repair integrity
and cancer risk were 70, 56, 67 and 29 %, respectively. The
8. Metastasis markers
anticarcinogenic properties of cruciferous vegetables have
9. Microsatellite instability
been attributed to their relatively high content of gluco-
10. Oxidative DNA damage
sinolates (Verhoeven et al. 1997). Glucosinolates are a
11. Gene-specific damage
group of glycosides stored within cell vacuoles of all
cruciferous vegetables. Myrosinase, an enzyme found in
plant cells, converts these compounds to a variety of
Faecal water markers
hydrolysis products, including isothiocyanates and indoles.
While a wide variety of naturally occurring and synthetic In recent years, there has been considerable interest in the
isothiocyanates have been shown to prevent cancer in role of the aqueous phase of human faeces (faecal water) in
animals (Hecht, 1995), attention has been focused on a studies examining the mechanisms underlying the dietary
particular isothiocyanate isolated from broccoli, i.e. aetiology of colon cancer. The motivation is that com-
sulforaphane. Sulforaphane has been shown to be a good ponents of this faecal fraction are more likely to be able
inducer of a particular phase II enzyme, quinone reductase. to exert untoward effects on the cells of the colonic epi-
Indole-3-carbinol has received attention for its cancer thelium than components bound to food residues and the
chemopreventive properties, particularly of the mammary bacterial mass.
gland. In addition to the induction of phase I and phase
1. Cytotoxicity
II detoxification reactions, indole-3-carbinol may reduce
2. COX-2 induction
cancer risk by modulating oestrogen metabolism.
3. Caspase induction
4. Calprotectin levels
5. Activator protein-1 activation
Citrus fruits 6. Bile acid levels
7. Effects on metastasis
Several epidemiological studies have shown that citrus
8. Genotoxicity
fruits are protective against a variety of human cancers.
9. Effects on cell metabolism
Although oranges, lemons, limes and grapefruits are a
10. Gene induction
principal source of such important nutrients as vitamin
C, folate and fibre, Elegbede et al. (1993) have suggested
that another component is responsible for the anticancer
Immunological and inflammatory response markers
activity. Citrus fruits are particularly high in a class of
phytochemicals known as the limonoids (Hasegawa & In the colon:
Miyake, 1996). In recent years, evidence has been
1. Suppression of COX-2 induction by pro-inflammatory
accumulating in support of the cancer-preventative
cytokines
effect of limonene (Gould, 1997). Crowell (1997)
showed this compound to be effective against a variety In blood:
of both spontaneous and chemically induced rodent
1. Natural killer cells
tumours.
2. Lymphocyte proliferation
However, mindful of the importance of the overall
3. Cytokines (interleukin-2, interleukin-b, tumour necro-
dietary pattern in cancer risk reduction, one must question
sis factor)
the clinical implications of a single phytochemical in
isolation. Thus, in designing a dietary intervention study/clinical trial
 Reduction of disease risk: cancer
to study anticancer effects (colon cancer) of functional
foods, it is recommended to use as many of the above
state-of-the-art biomarkers as is feasible.
Examples of studies addressing changes in
biomarker response to a functional food
Shift from dairy product-rich to dairy product-free diet
Cytotoxicity of faecal water is now an accepted risk marker
for colon cancer and several studies have correlated
toxicity of this faecal fraction with a higher colonic cell
proliferation and increased colon cancer risk (Lapre &
Van der Meer, 1992). Genotoxicity per se in human
faeces is also generally accepted as a risk marker for
colon cancer. Recently, Pool-Zobel et al. (1996) have
demonstrated, by employing the single cell gel electrophor-
esis (COMET) assay, that colon carcinogens induced gen-
otoxicity in the colon of rats and that this genotoxicity was
altered by dietary manipulations.
To determine whether the cytotoxicity and genotoxicity
of the aqueous phase of human stools (faecal water) were
affected by a change in dairy product intake, eighteen
healthy male and female volunteers were randomly divided
into two groups (Glinghammar et al. 1997). In a cross-over
design, the volunteers shifted from their normal dairy
product-rich to a dairy product-free diet. Nutritional
analysis of the food consumed, during the study period,
showed a significant decrease in energy intake from
9000 kJ/d to 7866 kJ/d (P 0.004), due to decreased
intakes of protein and fat. Carbohydrate and fibre intakes
remained unchanged during the intervention. Ca intake
decreased significantly from 1488 mg/d to 372 mg/d
(P , 0.001), with phosphate and vitamin D displaying
similar significant decreases (P , 0.001). Cytotoxicity of
faecal water, analysed by the HT-29 cytotoxicity assay,
indicated a significant decrease (P 0.025) in cell survival
from 34 % to 20 % when dairy products were excluded
from the participants diet. This effect is most likely due
to the decreased intake of dairy Ca and possibly phosphate.
The COMET assay, used to analyse genotoxicity of faecal
waters, indicated no differences brought about by the diet-
ary intervention. These findings indicate that a shift from a
dairy product-rich to a dairy product-free diet results in a
significant effect (P 0.025) on an accepted risk marker
for colon cancer, and may suggest that the mechanism by
which dairy products are protective is at the level of
tumour promotion rather than initiation.
Clinical trial to examine the effect of a synbiotic
preparation on colon cancer risk biomarkers
in adenoma patients
Another example of such a study is the ongoing Symbiotics
and Cancer Prevention in Humans (SYNCAN) project,
funded by the EU, and involving eight research centres
in Europe. It involves a twelve-week, randomised,
double-blind placebo-controlled trial of a food supplement
containing L. rhamnosus GG, Bifidobacterium Bb-12 and
Raftilose Synergyl in adenoma patients. In this study, all
of the colon cancer risk biomarkers, listed above, will be
S223
measured. In parallel, a long-term tumourigenesis study
in rats, using the same synbiotic combination and assaying
for the same biomarkers, is being carried out. It is hoped
that the results of this study will provide much needed
information on the cancer-protective effects of synbiotics
in man and on the underlying mechanisms.
References
Biasco G, Paganelli GM, Brandi G, Brillanti S, Lami F, Callegari
C & Gizzi G (1991) Effect of Lactobacillus acidophilus and
Bifidobacterium bifidum on rectal cell kinetics and fecal pH.
Italalian Journal of Gastroenterology 23, 142.
Block E (1992) The organosulfur chemistry of the genus Allium
implications for the organic chemistry of sulfur. Ange-
wandte Chemie, International Edition Engl. 31, 1135 1178.
Block G, Patterson B & Subar A (1992) Fruit, vegetables, and
cancer prevention: a review of the epidemiological evidence.
Nutrition and Cancer 18, 1 29.
Clinton SK (1998) Lycopene: chemistry, biology and implications
for human health and disease. Nutrition Reviews 56, 35 51.
Crowell PL (1997) Monoterpenes in breast cancer chemo-
prevention. Breast Cancer Research and Treatment 46,
191 197.
Di Mascio P, Kaiser S & Sies H (1989) Lycopene as the most
efficient biological carotenoid singlet oxygen quencher.
Archives Biochemistry and Biophysics 274, 532 538.
Dorant E, van den Brandt PA, Goldbohm RA, Hermus RJJ &
Sturmans F (1993) Garlic and its significance for the prevention
of cancer in humans: a critical review. British Journal of
Cancer 67, 424 429.
Dreosti IE, Wargovich MJ & Yang CS (1997) Inhibition of
carcinogenesis by tea: the evidence from experimental studies.
Critical Reviews in Food Science and Nutrition 37, 761 770.
Elegbede JA, Maltzman TH, Elson CE & Gould MN (1993)
Effects of anticarcinogenic monoterpenes on phase II hepatic
metabolizing enzymes. Carcinogenesis 14, 1221 1223.
Ernst E (1997) Can allium vegetables prevent cancer?
Phytomedicine 4, 79 83.
Fuller R (1989) Probiotics in man and animals. Journal of Applied
Bacteriology 66, 365 378.
Giovannucci E, Ascherio A, Rimm EB, Stampfer MJ, Colditz GA
& Willett WC (1995) Intake of carotenoids and retinol in
relation to risk of prostate cancer. Journal of the National
Cancer Institute 87, 1767 1776.
Glinghammar B, Venturi M, Rowland IR & Rafter J (1997) Shift
from a dairy product-rich to a dairy product-free diet: influence
on cytotoxicity and genotoxicity of fecal water potential risk
factors for colon cancer. American Journal of Clinical
Nutrition 66, 1277 1282.
Gould MN (1997) Cancer chemoprevention and therapy by
monoterpenes. Environmental Health Perspectives 105,
977 979.
Govers MJAP, Termont DSML, Lapre JA, Kleibeuker JH, Vonk
RJ & Van der Meer R (1996) Calcium in milk products precipi-
tates intestinal fatty acids and secondary bile acids and thus
inhibits colonic cytotoxicity in humans. Cancer Research 56,
3270 3275.
Graham HN (1992) Green tea composition, consumption and
polyphenol chemistry. Preventive Medicine 21, 334 350.
Ha YL, Grimm NK & Pariza MW (1987) Anticarcinogens from
fried ground beef: health-altered derivatives of linoleic acid.
Carcinogenesis 8, 1881 1887.
Hasegawa S & Miyake M (1996) Biochemistry and biological
S224 J. J. Rafter
functions of citrus limonoids. Food Reviews International 12,
413 435.
Hecht SS (1995) Chemoprevention by isothiocyanates. Journal of
Cellular Biochemistry Supplement 22, 195 209.
Hirayama K & Rafter J (2000) The role of probiotic bacteria in
cancer prevention. Microbes and Infection 2, 681 686.
Intestinal Microbiology Group, International Agency for Research
on Cancer (1977) Dietary fiber, transit time, faecal bacteria,
steroids and colon cancer in two Scandinavian populations.
Lancet 2, 207211.
Ip C & Scimeca JA (1997) Conjugated linoleic acid and linoleic
acid are distinctive modulators of mammary carcinogenesis.
Nutrition and Cancer 27, 131135.
Iscovich JM, L A bbe KA, Castelleto R, Calzona A, Bernedo A,
Chopita NA, Jmelnitzsky AC & Kaldor J (1992) Colon
cancer in Argentina. Risk from dietary items. International
Journal of Cancer 51, 851 857.
Kampman E, Giovannucci E, vant Veer P, Rimm E, Stampfer
MJ, Colditz GA, Kok FJ & Willett WC (1994a) Calcium, vita-
min D, dairy foods, and the occurrence of colorectal adenomas
among men and women in two prospective studies. American
Journal of Epidemiology 139, 16 29.
Kampman E, Goldbohm RA, van den Brandt PA & vant Veer P
(1994b) Fermented dairy products, calcium, and colorectal
cancer in the Netherlands cohort study. Cancer Research 54,
3186 3190.
Kohlmeier L, Weerings KGC, Steck S & Kok FJ (1997) Tea and
cancer prevention an evaluation of the epidemiologic
literature. Nutrition and Cancer 27, 1 13.
Lapre J, De Vries T, Termont DS, Kleibeuker J, De Vries EG &
Van der Meer R (1993) Mechanism of the protective effect of
supplemental dietary calcium on cytolytic activity of fecal
water. Cancer Research 53, 248253.
Lapre JA & Van der Meer R (1992) Diet-induced increase of
colonic bile acids stimulates lytic activity of fecal water and
proliferation of colonic cells. Carcinogenesis 13, 41 44.
Lidbeck A, Overvik E, Rafter J, Nord CE & Gustafsson J-A
(1992) Effect of Lactobacillus acidophilus supplements on
mutagen excretion in feces and urine in humans. Microbial
Ecology in Health and Disease 5, 59 67.
McKeown-Eyssen GE & Bright-See E (1984) Dietary factors in
colon cancer: international relationships. Nutrition and
Cancer 6, 160 170.
Nakachi K, Suemasu K, Suga K, Takeo T, Imai K & Higashi Y
(1998) Influence of drinking green tea on breast cancer
malignancy among Japanese patients. Japanese Journal of
Cancer Research 89, 254 261.
Peters RK, Pike MC, Garabrant D & Mack TM (1992) Diet and
colon cancer in Los Angeles County, California. Cancer
Causes and Control 3, 457 473.
Phipps WR, Martini MC, Lampe JW, Slavin JL & Kurzer MS
(1993) Effect of flax seed ingestion on the menstrual cycle.
Journal of Clinical Endocrinology and Metabolism 77,
1215 1219.
Pierre F (1997) Short-chain fructo-oligosaccharides reduce the
occurrence of colon tumors and develop gut-associated
lymphoid tissue in MIN mice. Cancer Research 57, 225 228.
Pool-Zobel BL, Neudecker C, Domizlaff I, Ji S, Schillinger U,
Rumney CJ & Rowland IR (1996) Lactobacillus and
bifidobacterium-mediated antigenotoxicity in the colon of
rats. Nutrition and Cancer 26, 365 380.
vant Veer P, Dekker JM, Lamers JWJ, Kok FJ, Schouten EG,
Brants HAM, Sturmans F & Hermus RJJ (1989) Consumption
of fermented milk products and breast cancer: a case control
study in the Netherlands. Cancer Research 49, 4020 4023.
Verhoeven DT, Goldbohm RA, van Poppel G, Verhagen H & van
den Brandt PA (1996) Epidemiological studies on brassica
vegetables and cancer risk. Cancer Epidemiology, Biomarkers
and Prevention 5, 733748.
Verhoeven DT, Verhagen H, Goldbohm RA, van den Brandt PA
& van Poppel G (1997) A review of mechanisms under-
lying anticarcinogenicity by brassica vegetables. Chemical
Biological Interactions 28, 79 129.
Weisburger JH (1998) International Symposium on Lycopene and
Tomato Products in Disease Prevention: an introduction.
Proceedings of the Society of Experimental Biology and
Medicine 218, 93 94.
Young TB & Wolf D (1988) A case control study of proximal
and distal colon cancer and diet in Wisconsin. International
Journal of Cancer 42, 167 175.
214 J. O. Hill and J. C. Peters
many of the treatments for type 2 diabetes such as insulin
and sulfonylureas have been shown to produce weight gain
and potentially exacerbate the disease (UK Prospective
Diabetes Study Group, 1998a,b).
Prevention of type 2 diabetes has recently been shown to
be a promising strategy. The Finnish Diabetes Prevention
Study (Tuomilehto et al. 2001) and the larger Diabetes Pre-
vention Program (DPP Research Group, unpublished
results) both found that the risk of developing diabetes in
obese, insulin-resistant subjects could be cut by more
than half by life-style modification. The life-style modifi-
cation consisted of modest weight loss (about 5 %) and
an increase in physical activity (thirty minutes of physical
activity about three times per week).
Modifying energy balance: biomarkers for obesity
Obesity arises from an energy imbalance whereby energy
intake exceeds energy expenditure. Dealing with obesity
either prevention or treatment requires modification
of one or both components of energy balance. Approaches
to weight management (including a functional food
approach) therefore can target multiple aspects of the
energy balance systems: food intake, energy expenditure
and energy storage. All of these approaches are currently
being taken by pharmaceutical companies; however,
developing foods for weight management may be a more
attractive approach for dealing with the 61 % of the
population that is overweight or obese.
Regardless of the approach to weight management, a
major limitation in assessing and documenting success is
the lack of suitable biomarkers to assess the impact of
different treatments on components of energy balance.
Assessment of body energy stores
Body weight can easily be assessed using a bathroom scale,
but assessing the increased health risk of obesity is more
difficult. The majority of data available relates BMI to
morbidity and mortality (National Institutes of Health,
1998). While BMI (weight divided by the square of
height) is obtained relatively easily, it does not provide
information about body composition (i.e. body fat v.
fat-free mass). Body composition can be assessed accu-
rately in the laboratory, but these techniques are not readily
available to the general population (Lohman, 2001). The
techniques available to the general population for assessing
body composition (such as bioelectrical impedance and
skinfold thickness) are not considered to be highly accurate
or precise (Lohman, 2001).
Energy intake
Total energy intake. It is surprisingly difficult to measure
accurately the amount of energy consumed by a free-living
individual. There are accurate tables for converting food
eaten into energy (kJ or kg) consumed, but the difficulty
is in determining the amount and composition of food
eaten. The typical way of assessing food intake is to
have individuals keep records of food eaten. This is fre-
quently done using food diaries (for periods of three to
seven days), 24-hour recalls of all food eaten during the
previous twenty-four hours, or assessments of the
frequency with which various foods are usually consumed
(Westerterp, 1998). While these techniques allow esti-
mation of energy intake, they lack both accuracy and
precision (Heitman & Lissner, 1995; Westerterp, 1998).
For example, when comparing self-reported energy intake
with accurate measures of energy expenditure (e.g. using
doubly labelled water) for individuals in energy balance,
estimated energy intake is usually below measured
energy expenditure (Schoeller, 1995). Proponents of the
use of dietary self-reports argue that even if the absolute
energy intake is not accurate, the instruments provide an
accurate reflection of changes in energy intake (Willett,
1990; Kristal et al. 1994). Whether this is the case or not
is unclear.
Type of food eaten. There is great controversy about
the role of diet composition in the development and man-
agement of obesity. It is likely that some functional
foods aimed at reducing obesity will attempt to alter diet
composition. Determining diet composition accurately
may be even more difficult than determining total energy
intake. Perhaps changes in protein intake are more easily
assessed than changes in the other macronutrients. Protein
intake is correlated with urinary and blood levels of nitro-
gen, and these can provide a biomarker of changes in diet-
ary protein intake.
There is no universally accepted marker for carbo-
hydrate intake. One can obtain some sense of type of
carbohydrate intake from the glycaemic index, which is
the rise in blood glucose after carbohydrate intake.
Amount of carbohydrate intake is also correlated with
insulin level, but it is not clear that either glucose or insulin
level is a useful marker of changes in carbohydrate intake.
Additionally these measurements are not easily available to
the general population.
There is no clear useable biomarker for fat intake. The
type of fat consumed chronically can be determined from
biopsies of adipose tissue or from red blood cell membrane
fatty acid composition (Wardlaw et al. 1994), but these
techniques are invasive and do not provide information
about the amount of fat consumed. Blood levels of
cholecystokinin and enterostatin change with dietary
intake but it is not clear that these blood changes allow
accurate quantification of changes in dietary fat.
Assessment of energy expenditure
Functional foods that modify energy expenditure could be
useful in weight management. Assessment of energy
expenditure can be accomplished accurately under
laboratory conditions using indirect calorimetry or doubly
labelled water (Melby et al. 1998). There are currently
no biomarkers for energy expenditure that are available
to the general public. A new device called the Body
GEMe has just gone on the market in the USA (Nieman
et al. 2001) and appears to provide a measure of resting
energy expenditure of individuals. Like more expensive
hospital equipment, the Body GEMe determines energy
expenditure from measurement of the amount of oxygen
consumed by the body. Such measurements can be made
 Obesity and diabetes
relatively inexpensively in health clubs and by personal
dietitians and personal trainers. Devices like this may be
useful in assessing the efficacy of products to increase
energy expenditure.
Functional foods to reduce energy intake
One promising avenue to reduce energy intake using
functional foods is through increased satiety. The goal is
to provide foods that increase the sense of fullness and
encourage the individual to stop eating sooner, thereby
reducing total energy intake. In general, the three most
promising areas to increase satiety are to: (1) modify the
energy density of the diet; (2) modify the macronutrient
composition of the diet; and (3) modify the glycaemic
index of the diet.
Energy density
The energy density of the diet is the energy content per unit
of weight or volume and seems to be correlated with total
energy intake (Rolls & Bell, 1999, 2000; Stubbs et al.
1995a,b). Energy density is relatively easy to measure
for most foods and can be calculated by dividing the
energy content of the food in kJ by the weight or volume
of the food (Rolls & Barnett, 2000). In their book
Volumetrics, Rolls & Barnett (2000) explain the concept
of energy density and provide energy density values for a
range of foods.
There are substantial data to suggest that total energy
intake over the short term (a few days) varies directly
with the energy density of the diet (Rolls & Bell, 1999,
2000; Stubbs et al. 1995a,b). One physiological signal
for satiety may relate to the total weight or volume of
food ingested. This suggests that modifying the energy
density of the diet could be a way to reduce total energy
intake and reduce obesity. The main determinant of
energy density is the non-caloric content of the food,
primarily the water content (Grunwald et al., 2001).
Foods with a high water content have a low energy density.
Fibre also reduces energy density since it contributes sub-
stantially more to food weight than to caloric content.
However, energy density is also affected by the macro-
nutrient composition of the diet. Since fat is more
energy-dense (38 kJ/g) than either protein or carbohydrate
(17 kJ/g), reducing the proportion of fat in the diet can
have a major impact on reducing the energy density of
the diet. Since there are solid data showing that reducing
energy density reduces energy intake (at least in the short
term), functional foods aimed at modifying energy density
may be useful in managing obesity. It would be helpful to
have more long-term data substantiating the effect of
energy density and relating energy density to changes in
body weight.
Macronutrient composition and satiety
Researchers often refer to a hierarchy of satiety for macro-
nutrients with protein being the most satiating and fat the
least satiating, joule for joule (Hill & Prentice, 1995).
High-protein diets are currently popular for weight loss
215
and are based, in part, on the idea that high-protein diets
promote satiety. Popular diet books such as New Diet
Revolution (Atkins, 1992) and Protein Power (Eades &
Eades, 1996) are based on this premise. The data in support
of high-protein diets as facilitators of weight loss and
weight maintenance are still incomplete. Some studies
suggest that high-protein diets may be effective in produc-
ing short-term decreases in food intake (Rolls et al. 1988)
but whether or not these diets are helpful in long-term
weight maintenance or in prevention of weight gain is
not clear.
With regard to satiety and carbohydrate intake, it may
depend on the type of carbohydrate in the diet. While
intake of simple carbohydrate or sugar has been suggested
by the popular press to be associated with the development
of obesity, it is not clear that high-sugar diets lead to over-
eating or obesity compared with low-sugar diets (Lewis
et al. 1992; Hill & Prentice, 1995; Ludwig et al. 2001).
It has also been suggested that calories in liquid form
may affect satiety less strongly than solid foods with the
same calorie content (Mattes, 1996). While the available
data are not definitive, if substantiated they could implicate
caloric beverages as a facilitator of overeating and a causal
factor in obesity. Most calories in beverages are in the form
of carbohydrates, and there is little information about
whether different types of carbohydrates in liquid form
might affect energy intake differently.
Dietary fibre intake seems to best predict total energy
intake, with several reports of lower total energy intake
with high-fibre v. low-fibre diets (Pereira & Ludwig,
2001). There are probably several reasons why high-fibre
diets are associated with lower food intake. First, high-
fibre diets may trigger maximal sensory stimulation in
the mouth due to the increased need for chewing. High-
fibre diets also lead to slower gastric emptying and a
slower rate of nutrient absorption. Finally, a high fibre con-
tent reduces the energy density of the overall diet. Regard-
less of the reason, increasing dietary fibre is generally
thought to aid in weight management. Ludwig et al.
(1999) found that high dietary fibre seemed protective
against weight gain over a decade.
Modification of dietary fat type is not a commonly
accepted strategy for weight loss. There is some suggestion
that diets high in polyunsaturated fatty acids stimulate total
fat oxidation more than diets high in saturated fatty acids
(Jones & Schoeller, 1988) but this is somewhat controver-
sial. Other fats such as short- and medium-chain triacylgly-
cerols and n 3 fatty acids may have a greater impact on
energy metabolism but it is not clear that these would
play a major role in weight management.
Glycaemic index and satiety
The glycaemic index of a food is determined by the rise in
glucose that occurs after eating that food in relation to the
rise in glucose seen after eating a standardized food such as
white bread (Brand-Miller et al. 1999). This obviously
requires measuring glucose after consuming a food. High
glucose levels following eating would stimulate insulin
secretion which may increase appetite and facilitate other
disease processes linked to insulin action. In their book,
216 J. O. Hill and J. C. Peters
The Glucose Revolution, Brand-Miller et al. (1999) provide
the rationale for how glycaemic index may affect food
intake and obesity, and provide tables of glycaemic index
for many foods.
Whether or not the glycaemic index of the diet affects
energy intake and obesity remains controversial. There is
no convincing evidence that food intake is related directly
to glycaemic index, although there is some evidence that
high glycaemic diets are linked to weight gain (Ludwig,
2000). The glycaemic index of the total diet could be
modified by eating foods with a low glycaemic index. If
the glycaemic index were shown to affect food intake, a
good target would be to develop more good-tasting, low
glycaemic foods.
Functional foods to increase energy expenditure
Another way to reduce the likelihood of developing obesity
or to treat obesity would be to increase total energy
expenditure without increasing energy intake. While
some food supplements make the claim of increasing
energy expenditure, there are very few data available to
support the efficacy of these products.
One product on the market that has some demonstrated
efficacy is the combination of caffeine and ephedrine.
This combination has been shown to increase energy
expenditure modestly and is used for obesity treatment in
some countries (Astrup et al. 1992). There has been
recent concern about the long-term safety of ephedrine.
Recently it has been suggested that diets high in Ca may
be protective against weight gain and that part of the
mechanism may be an increase in energy expenditure
(Zemel et al. 2000). In several datasets, high Ca intake is
associated with a lower BMI (Davies et al. 2000), but
there has yet been no clear demonstration that this is a
causal relationship.
Oolong tea is another food that may have some impact
on increasing energy expenditure (Rumpler et al. 2001),
perhaps through its catechin content. Resting metabolic
rate was increased by 3 4 % during three days of oolong
tea consumption at five cups per day. Interestingly, most
of the rise in metabolic rate was from increased fat oxi-
dation, which should have the greatest impact upon
decreasing body fat stores.
Functional foods to alter nutrient partitioning
If some of the energy ingested is not absorbed completely,
this can reduce net energy available to meet metabolic
demands and can lead to weight loss. The approaches
that have been taken involve blocking absorption of carbo-
hydrates (starch blockers) or fats (fat blockers). The impact
of the currently available products (e.g. Acarbose) is not
clear.
Non-absorbable fats
Olestra is a non-absorbable fat substitute that has been
shown to reduce total energy intake (Hill et al. 1998),
and to be effective in producing weight loss (Bray et al.
2002). Olestra was approved by the Food and Drug
Administration in 1996 for use in savoury snack products.
Post market surveillance data suggest an association
between Olestra intake and weight maintenance. The
mechanism for a positive obesity effect may be its effects
on reducing energy density, although other mechanisms
cannot be ruled out.
Functional foods to prevent or manage diabetes
The results of the Finnish Diabetes Prevention Study
(Tuomilehto et al. 2001) and the larger Diabetes Preven-
tion Program (DPP Research Group, unpublished results)
show that modest weight loss (about 5 %) can reduce the
development of type 2 diabetes by half in individuals at
risk for this disease. There is a tremendous opportunity
to develop functional foods targeted at those at risk for
type 2 diabetes (i.e. insulin-resistant) to help them achieve
modest weight reduction. Theoretically, any functional
food that helped with weight loss could be marketed as
helping prevent diabetes in an insulin-resistant population.
Furthermore, it may be possible to target functional foods
towards management of diabetes in those who already
have type 2 diabetes. It is clear that weight control is an
effective diabetes management technique and, here again,
functional food aimed at weight loss could be targeted
towards those with type 2 diabetes. Moreover, it may be
possible to develop functional foods that impact insulin
action independently of weight loss. These would probably
be foods that increase the sensitivity of peripheral tissues to
insulin, or insulin sensitizers.
Insulin sensitizers
There are currently food supplements that claim to increase
insulin sensitivity, but there is very little evidence about
the effectiveness of these products. Perhaps the most
widely available of these is chromium picolinate, which
under some circumstances appears to affect insulin action
(Cefalu et al. 1999). Such products have not yet been
proved to be effective in treating diabetes.
The future: possibilities for new biomarkers
While we see a clear opportunity to develop and market
more functional foods for weight management, the lack
of accurate biomarkers to assess their effectiveness is a
barrier to this process. The future may hold promise for
developing better biomarkers and this would greatly
facilitate the development of functional foods.
It is almost certain that better and more accessible
methods of measuring body composition will be available
in the future. Accurate body composition, including body
fat distribution, is not available on a widespread basis out-
side laboratory settings. This is likely to change with
additional technological advances.
It is likely that we will develop better ways of assessing
total energy intake and intake of specific macronutrients by
adding non-metabolizable substances to the diet and moni-
toring their excretion. This will allow better assessment of
the effectiveness of functional food for changing total
energy intake or macronutrient content of the diet.
 Obesity and diabetes
As recognition of the need to assess and treat the
metabolic syndrome grows, physicians and other health
care professionals will be screening patients for insulin
resistance. It is likely that there will soon be consensus
on how to do this, thus providing a standardized way of
assessing insulin resistance that can be used to assess any
change in response to functional foods.
References
Astrup A, Buemann B, Christensen NJ, Toubro S, Thorbek G,
Victor OJ & Quaade F (1992) The effect of ephedrine/caffeine
mixture on energy expenditure and body composition in obese
women. Metabolism 41, 686 688.
Atkins RC (1992) Dr. Atkins New Diet Revolution. New York:
Avon Books.
Brand-Miller J, Wolever TMS, Colagiuri S & Foster-Powell K
(1999) The Glucose Revolution. New York: Marlowe &
Company.
Bray GA, Lovejoy JC, Most-Windhauser M, Smith SR,
Volaufova J, Denkins Y, DeJonge L, Rood J, Lefevre M,
Eldridge AL & Peters JC (2002) A nine-month randomized
clinical trial comparing a fat-substituted and fat-reduced diet
in healthy obese men: The Ole Study. American Journal of
Clinical Nutrition in press.
Cefalu WT, Bell-Farrow AD, Stegner J, Shong QW, King T,
Morgan T & Terry JG (1999) Effect of chromium picolinate
on insulin sensitivity in vivo. Journal of Trace Elements in
Experimental Medicine 12, 71 83.
Davies KM, Heaney RP, Recker RR, Lappe JM, Barger-Lux J,
Rafferty K & Hinders S (2000) Calcium intake and body
weight. Journal of Clinical Endocrinology and Metabolism
85, 46354638.
Eades MR & Eades MD (1996) Protein Power. New York:
Bantam Books.
Fagot-Campagna A, Pettitt DJ, Englegau MM, Burrows NR,
Geiss LS, Valdez R, Beckles GL, Saadine J, Gregg EW,
Williamson DF & Jarayan KMV (2000) Type 2 diabetes
among North American children and adolescents: an epidemio-
logic review and a public health perspective. Journal of
Pediatrics 136, 664 672.
Grunwald GK, Seagle HM, Peters JC & Hill JO (2001) Quantify-
ing and separating the effects of macronutrient composition and
non-energetic food components on energy density. British
Journal of Nutrition 86, 265 276.
Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE,
Little RR, Wiedmeyer HM & Byrd-Holt DD (1998) Prevalence
of diabetes, impaired fasting glucose, and impaired glucose
tolerance in US adults. The Third National Health and Nutri-
tion Examination Survey, 1988 1994. Diabetes Care 21,
518524.
Heitman BL & Lissner L (1995) Dietary underreporting by obese
individuals is it specific or non-specific? British Medical
Journal 311, 986 989.
Hill JO & Prentice AM (1995) Sugar and body weight regulation.
American Journal of Clinical Nutrition 62, 264S 274S.
Hill JO, Seagle HM, Johnson SL, Smith S, Reed GW, Tran ZV,
Cooper D, Stone M & Peters JC (1998) Effects of 14 days of
covert substitution of Olestra on spontaneous food intake.
American Journal of Clinical Nutrition 67, 1178 1185.
Jones PJH & Schoeller DA (1988) Polyunsaturated:saturated ratio
of diet fat influences energy substrate utilization in the human.
Metabolism 37, 145 151.
Kristal AR, Beresford SA & Lazovich D (1994) Assessing change
217
in diet-intervention research. American Journal of Clinical
Nutrition 59, 185S 189S.
Lebovitz HE (1994) Stepwise and combination drug therapy for
the treatment of NIDDM. Diabetes Care 17, 1542 1544.
Lewis CJ, Park YK, Dexter PB & Yetley EA (1992) Nutrient
intakes and body weight of persons consuming high and
moderate levels of sugars. Journal of the American Dietetic
Association 92, 703 713.
Lohman TG (2001) Body composition. In Eating Disorders and
Obesity: A Comprehensive Handbook, 2nd ed., pp. 62 66
[C Fairburn and K Brownell, editors]. New York: Guilford.
Ludwig DS (2000) Dietary glycemic index and body weight
regulation. Nutrition Society of Australia 24, 286 293.
Ludwig DS, Pereira MA, Kroenke CH, Hilner JE, Van Horn L,
Slattery ML & Jacobs DR Jr (1999) Dietary fiber, weight
gain and cardiovascular disease risk factors in young adults:
the CARDIA study. Journal of the American Medical
Association 282, 1539 1546.
Ludwig DS, Peterson KE & Gortmaker SL (2001) Relation
between consumption of sugar-sweetened drinks and childhood
obesity: a prospective, observational analysis. Lancet 357,
505 508.
Maggio CA & Pi-Sunyer FX (1997) The prevention and treatment
of obesity. Application to type 2 diabetes. Diabetes Care 20,
1744 1766.
Mattes RD (1996) Dietary compensation by humans for sup-
plemental energy provided as ethanol or carbohydrate in
fluids. Physiology and Behavior 59, 179 187.
Melby CR, Ho R & Hill JO (1998) Assessment of energy
expenditure. In Physical Activity and Obesity, pp. 204 233
[C Bouchard, editor]. Champaign, IL: Human Kinetics.
Mokdad H, Bowman BA, Ford ES, Vinicor F, Marks JS &
Koplan JP (2001) The continuing epidemics of obesity and dia-
betes in the United States. Journal of the American Medical
Association 286, 1195 1200.
National Center for Health Statistics (1999) Prevalence of over-
weight and obesity among adults: United States (electronic
citation). http://www.cdc.gov/nchs/products/pubs/pubd/hestats/
obese/obsc99.htm.
National Institutes of Health (1998) Clinical guidelines on the
identification, evaluation, and treatment of overweight and
obesity in adults the Evidence Report. Obesity Research
6, Suppl. 2, 51S 210S.
Nieman DC, Trone G & Austin M (2001) Validation of a new
handheld device for measuring resting metabolic rate. Obesity
Research 9, Suppl. 3, 202S.
Pereira MA & Ludwig DS (2001) Dietary fiber and body weight
regulation: observations and mechanisms. Pediatrics Clinics of
North America 48, 969 980.
Pi-Sunyer FX (1993) Medical hazards of obesity. Annals of
Internal Medicine 119, 655660.
Reaven GM (1998) Role of insulin resistance in human disease.
Diabetes 37, 1595 1607.
Rolls B & Barnett RA (2000) Volumetrics. New York: Harper-
Collins.
Rolls BJ & Bell EA (1999) Intake of fat and carbohydrate: role of
energy density. European Journal of Clinical Nutrition 52,
1 8.
Rolls BJ & Bell EA (2000) Dietary approaches to the treat-
ment of obesity. Medical Clinicians of North America 84,
401 418.
Rolls BJ, Hetherington M & Burley VJ (1988) The specificity of
satiety: the influence of foods of different macronutrient con-
tent on the development of satiety. Physiology and Behavior
43, 145 153.
Rumpler W, Seale J, Clevidence B, Judd J, Wiley E, Yamamoto
S, Komatsu T, Sawaki T, Ishikura Y & Hosoda K (2001)
218 J. O. Hill and J. C. Peters
Oolong tea increases metabolic rate and fat oxidation in men.
Journal of Nutrition 131, 2848 2852.
Schoeller DA (1995) Limitations in the assessment of dietary
energy intake by self-report. Metabolism 44, Suppl. 2, 18 22.
Stubbs RJ, Harbron CG, Murgatroyd PR & Prentice AM (1995a)
Covert manipulation of dietary fat and energy density: effect on
substrate flux and food intake in men eating ad libitum. Amer-
ican Journal of Clinical Nutrition 62, 316 329.
Stubbs RJ, Ritz P, Coward WA & Prentice AM (1995b) Covert
manipulation of the ratio of dietary fat to carbohydrate and
energy density: effect on food intake and energy balance in
free-living men eating ad libitum. American Journal of Clinical
Nutrition 62, 330 337.
Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen
H, Ilanne-Parikka P, Seinanen-Kiukaanniemi SK, Laakso M,
Louheranta A, Rastas M, Salminen V & Uusitupa M (2001)
Prevention of type 2 diabetes mellitus by changes in lifestyle
among subjects with impaired glucose tolerance. New England
Journal of Medicine 344, 1343 1350.
UK Prospective Diabetes Study Group (1998a) United Kingdom
Prospective Diabetes Study 24: a 6-year, randomized, con-
trolled trial comparing sulfonylurea, insulin, and metformin
therapy in patients with newly diagnosed type 2 diabetes that
could not be controlled with diet therapy. Annals of Internal
Medicine 128, 165 175.
UK Prospective Diabetes Study Group (1998b) Intensive
blood-glucose control with sulphonylureas or insulin com-
pared with conventional treatment and risk of complications
in patients with type 2 diabetes (UKPDS 33). Lancet 352,
837 853.
Wardlaw GM, Snook JT, Park S, Patel PK, Pendley FC, Lee M &
Jandacek RJ (1994) Relative effects on serum lipids and apoli-
poproteins of a caprenin-rich diet compared with diets rich in
palm oil/palm-kernel oil or butter. American Journal of Clini-
cal Nutrition 61, 535 542.
Westerterp KR (1998) The assessment of energy and nutrient
intake in humans. In Physical Activity and Obesity,
pp. 133 149 [C Bouchard, editor]. Champaign, IL: Human
Kinetics.
Willett W (1990) Nutritional Epidemiology. New York: Oxford
University Press.
Wing RR & Hill JO (2001) Successful weight loss maintenance.
Annual Reviews of Nutrition 21, 323 341.
Zemel MB, Shi H, Greer B, Dirienzo D & Zemel PC (2000)
Regulation of adiposity by calcium. FASEB Journal 14,
1132 1138.
S188 L. Dye and J. Blundell
foods that will protect individuals from weight gain or
bring about a weight loss. The first of these objectives
will be the most readily attainable. This means that foods
must prevent individuals from reaching a state of positive
energy balance this means ensuring that energy intake
does not exceed energy expenditure. Although it is more
readily recognised that foods should have the potential to
control energy intake, it should be kept in mind that
foods can also enhance energy expenditure either through
metabolic activity (e.g. thermogenesis) or by increasing
physical activity. It is important to recognise that behaviour
accounts for 100 % of energy intake and between 20 and
60 % of energy expenditure. Therefore the direct effect of
foods on behaviour is critical.
Considering the reduction of energy uptake, adjusting
the composition of the diet is a valid proposition. The
use of specific macronutrient manipulations can help to
suppress hunger and reduce the amount of food eaten.
Alternatively, specific materials can be incorporated into
food (such as yoghurts) so as to produce an intense
action on satiety-signalling systems. This strategy involves
the development of functional foods (nutraceuticals) for
appetite control. There is a huge opportunity for foods in
this sector of the food market.
Appetite control and the satiety cascade
The biological drive to eat is linked to the satiating power
of food. Satiating power, or satiating efficiency, describes
the capacity of a food to suppress and to inhibit further
eating. Food causes this effect by certain mediating pro-
cesses that can be roughly classified as sensory, cognitive,
post-ingestive (pre-absorptive) and post-absorptive. These
processes are operated by the impact of food on physio-
logical and biochemical mechanisms, and collectively
these processes have been referred to as the satiety cascade.
The way in which food is sensed and processed by the bio-
logical system generates neural and humoral signals, which
are used to control appetite. It follows that any self-
imposed or externally applied reduction in the food
supply, creating a calorific deficit, will weaken the satiating
power of food. One consequence of this will be the failure
of food to suppress hunger adequately (the biological
drive). The satiety cascade appears to operate as efficiently
in obese people as in lean individuals a normal appetite
response to reduced calorie intake is evident in obese
subjects.
Technically, satiety can be defined as inhibition of
hunger and eating that arises as a consequence of food con-
sumption. It can be distinguished from satiation, which is
the process that brings a period of eating to a halt. Conse-
quently, satiation and satiety act together to determine the
pattern of eating behaviour and the accompanying profile
of motivation. The conscious sensation of hunger is one
index of motivation and reflects the strength of satiation
and satiety. It is worth remembering that hunger is a bio-
logically useful sensation. It is a nagging, irritating feeling
that prompts thoughts of food and reminds us that the body
needs energy. The identification and management of
hunger are important factors underlying normal appetite
function and abnormalities of appetite and body weight.
Physiological satiety signals
A key feature in understanding the effect of foods on satiety
is the recognition of physiological satiety signals. Even
before food touches the mouth, physiological signals are
generated by the sight and smell of food. These events con-
stitute the cephalic phase of appetite. Cephalic phase
responses are generated in many parts of the gastrointest-
inal tract; their function is to anticipate the ingestion of
food. During and immediately after eating, afferent infor-
mation provides the major control over appetite. Afferent
information from food in the mouth provides primarily
positive feedback for eating while that from the stomach
and small intestine provides primarily negative feedback.
Initially, the brain is informed about the amount of food
ingested and its nutrient content via afferent input. The
gastrointestinal tract is equipped with specialised chemo-
and mechanoreceptors that monitor physiological activity
and pass information to the brain, mainly via the vagus
nerve. This afferent information constitutes one class of
satiety signals and forms part of the post-ingestive control
of appetite. A post-absorptive phase is also usually present
in response to digested nutrients crossing the intestinal wall
to enter the circulation. These products, which accurately
reflect the food consumed, may be metabolised in the peri-
pheral tissues or organs or may enter the brain directly via
the circulation. In either case, these products constitute a
further class of metabolic satiety signals. It has been
argued that the degree of oxidative metabolism of glucose
and free fatty acids in the liver constitutes a significant
source of information for the control of appetite. Addition-
ally, products of digestion and agents responsible for their
metabolism may reach the brain and bind to specific chemo-
receptors that influence that influence neurotransmitter syn-
thesis or alter some aspect of neuronal metabolism. In each
case the brain is informed about some aspects of metabolic
state resulting from food consumption.
Macronutrients and satiating power
A vulnerable point within the psychobiological system is
the interaction between the nature of the diet and the
bodys biological responses to food. The concept of the
satiety cascade implies that foods of varying nutritional
consumption will engage differently with the mediating
processes and will, therefore, exert differing effects on
satiation and satiety. Dietary variables could overcome
appetite control by a strong attractiveness mediated by
sensory mechanisms, by nutrient composition or by an
interaction between sensory and nutritional components.
Considering protein consumption, current surveys suggest
that most populations consume between 13 and 15 %
daily energy as protein and dietary recommendations
suggest the intake of 08 10 g of protein for every kg
of body weight. However, analysis of anthropological evi-
dence suggests the possibility that our huntergatherer
ancestors may have consumed diets containing 19 35 %
protein (Cordain et al. 2000). However, this view is not
universally accepted (Milton, 2000).
Do the macronutrients protein, carbohydrate and fat con-
tribute equally to satiating power? The answer is not
 Psychological and behavioural functions
obvious. One reason is that these macronutrients are not of
equal caloric density, and therefore comparisons among
them should be made on the basis of their individual caloric
contributions rather than upon weight. Investigations of the
action of macronutrients can be made by presenting preloads
in which one macronutrient is held constant and the other
two are varied systematically. For example, when fat is
held constant and protein and carbohydrate adjusted, it is
generally found that protein provides greater satiating
power than carbohydrate. Therefore, high-protein meals
would be expected to give rise to intense and prolonged
satiety.
Protein and satiety
Evidence for a specific effect of protein on satiety goes
back to the aminostatic concept of appetite control
(Mellinkoff et al. 1956). Early studies examined the effects
of breakfast nutritional composition and found that high-
protein breakfasts were followed by a consistent and pro-
longed sense of well-being, and that daily energy intake
computed from diaries was inversely related to the protein
content of breakfast. Unfortunately though, the outcomes
of both studies were compromised by methodological
shortfalls, such as not controlling for the energy content
or bulk of the breakfast meals. Similar tantalising but
problematic findings were reported for the effect of protein
on subjective hunger. Fryer et al. (1955) noted that a high-
protein weight-loss diet was associated with the least
reports of hunger, and Mellinkoff et al. (1956) found an
inverse correlational relationship between hunger and
serum amino acid levels.
Later studies enabled firmer conclusions to be drawn
regarding the satiating power of protein. Booth et al.
(1970) evaluated the effects of equicaloric high- and low-
protein composite meals on the intake of a nutritionally inter-
mediate cornflower pudding three hours later. Although the
palatability of the food offered must be questioned, the
authors nevertheless described a 26 % reduction in voluntary
intake after the high-protein meal. Using a completely differ-
ent methodology, Butler et al. (1981) administered a small
preload of an 8 g mixture of four amino acids or placebo
(134 kJ) half an hour before offering a cooked midday
meal. The amino acid preload led to a 10 % reduction in
energy intake from this meal compared with placebo.
Other studies have described the effects of high protein
loads or meals on carefully measured scales of subjective
motivation to eat. Spring et al. (1983) found that subjects
reported feeling significantly more full after eating 227 g
of turkey breast (high in protein) than after an equicaloric
amount of high-carbohydrate sherbet. Likewise, Hill &
Blundell (1986) found that eating a high-protein lunchtime
meal led subjects to rate themselves as having less desire to
eat and as feeling more full than after an equicaloric and
equivolume high-carbohydrate meal. A similar action of
protein was also seen in obese subjects (Hill & Blundell,
1986). Consequently, there is a good deal of evidence
that high-protein foods can exert a potent modulation of
human appetite. However, comparisons between different
types of protein, such as egg albumin, casein, gelatin,
soya protein, pea protein and wheat gluten, have failed to
S189
display differences arising from a manipulated meal
(Lang et al. 1998).
Proteins and satiety signalling
Considering the operations of the satiety cascade, proteins
in food could exert an action on satiety via the pre-absorp-
tive or post-absorptive mechanism. In fact there is evidence
for an action via both routes.
In general, proteins could intervene in the satiety cascade
by triggering, or maintaining, the release of the satiety hor-
mone cholecystokinin. There is good evidence in man that
cholecystokinin intensifies normal satiety (Greenough et al.
1998) and in animals the satiety effect of protein has been
shown to be mediated via cholecystokinin-A type receptors
(Trigazis et al. 1997).
It is clear that intact protein can have a potent satiety-
triggering effect, but in addition specific amino acids or
peptides could also exert a noticeable action. Individual
amino acids such as phenylanine or tryptophan (Hill &
Blundell, 1988) have been shown to increase satiety and
also adjust food preferences and selection. More recently,
the dipeptide phenylalanine aspartic acid, when delivered
in capsule form before a meal, significantly reduced food
consumption (Rogers & Blundell, 1989). Interestingly,
this effect was most prominent when the dipeptide was
administered an hour before the meal (Rogers et al. 1995).
There seems to be a clear potential for other functional
peptides such as caseinomacropeptide to exert a satiety-
enhancing effect. There is some evidence that caseino-
macropeptide can release cholecystokinin and may modulate
the release of other gastrointestinal hormones such as
glucagon-like peptide-1, which is known to be involved
in satiety (Blundell & Naslund, 1999).
Proteins and peptides may also exert effects via a post-
absorptive route. Important here is the concept of a
hierarchy of satiety power among the macronutrients
(Stubbs, 1996). This satiating concept depends upon the
limited store of protein in the body (lean body mass) and
the tight coupling between protein intake and oxidation.
According to this formulation the hierarchy of satiating
power is protein . carbohydrate . fats.
In addition, there is evidence for a strong effect of pro-
tein on diet-induced thermogenesis. A comparison between
two diets demonstrated that the diet containing 29 % pro-
tein produced a greater effect on 24 h diet-induced thermo-
genesis than a 9 % protein diet (Westerterp-Plantenga et al.
1999).
Another post-absorptive mechanism depends upon the
role of amino acids as precursors of brain neurotransmitters
for example, tryptophan for serotonin and phenylalanine
and tyrosine for dopamine and noradrenalin. The capacity
of these neurotransmitters to modulate the expression of
appetite through brain systems can be influenced by the
ratio of amino acids (e.g. tryptophan to large neutral
amino acids) in the plasma.
These mechanisms indicate how protein exerts effects on
satiety either as an entire food, a food component or as an
isolated entity. Potent effects have been demonstrated for
intact protein (Blundell & Hill, 1986; Hill & Blundell,
1990), dipeptides (Rogers et al. 1990, 1991) and single
S190 L. Dye and J. Blundell
amino acids (Hill & Blundell, 1988). This evidence indi-
cates the potential for proteins to influence the intensity
of satiety and to exert control over the pattern of food
consumption.
Carbohydrates and satiety
There are strong logical reasons why glucose should be
one of the most important nutrients to be monitored. As
Carlson (1991) has pointed out: because the brain controls
eating, it seems reasonable that hunger might be triggered
by a decrease in the brains primary fuel. The idea that the
metabolism of glucose in the body is related to the exist-
ence of hunger and eating is represented in the glucostatic
hypothesis formulated by Mayer (1953, 1955). It was pos-
tulated that the short-term articulation of energy intake
with energy needs is under glucostatic control (Van Itallie,
1990). In turn, this implies that dietary carbohydrates are
clearly involved in the short-term control of energy
intake. Similar views are embodied in the energostatic
hypothesis (Booth, 1972), metabolic control of food
intake (Friedman, 1991), the glycogenostatic hypothesis
(Astrup & Flatt, 1996) and the energetic concept of appe-
tite (Blundell & Rogers, 1991a). Although the most
obvious mechanism is that hunger is related directly to
fluctuations in the concentration of blood glucose, the
relationship between energy intake and glucose could be
mediated via arteriovenous glucose differences, the rate
of glucose utilisation in the liver, the activation of gluco-
receptive neurons in the periphery or the brain, or change
in the glycogen stores. Evidence for the role of glucose
in the initiation of eating comes from the detection of
transient declines in blood glucose being related to
meal initiation (Campfield & Smith, 1986) and the obser-
vation that increases in energy intake follow injections of
2-deoxy-D -glucose (Thompson & Campbell, 1977). Con-
versely, glucose infusions (particularly into the hepatic
portal system) produce an inhibition of the tendency to
eat (Novin et al. 1973). After the consumption of a carbo-
hydrate food, the glucose produced could be monitored
early by glucose-detecting interoreceptors in the upper
gastrointestinal tract linked to visceral afferent fibres in
the vagus nerve (Mei, 1985). In addition, Oomura (1988)
has described glucose-sensitive neurons (which decrease
their activity in response to applied glucose) and gluco-
receptor neurons (which are activated) in different parts
of the brain, which presumably provide an interface
between the presence of available glucose and activity in
aminergic or peptidergic neural pathways (Blundell,
1991). Apart from these direct links between the detection
of glucose and neural transmission, an indirect mechanism
has been postulated through which carbohydrate in the diet
could be related to neurochemical activity. This mech-
anism relates the proportion of carbohydrate to protein in
the diet to the ratio of tryptophan to other large neutral
amino acids in plasma, which determines the uptake of
tryptophan into the brain (Wurtman, 1982). In turn, it is
postulated that this leads to an activation of serotoninergic
neurons, which are functionally coupled to eating behav-
iour and food selection. The evidence for all stages in
this biobehavioural loop is equivocal (Blundell & Hill,
1987), but the ingestion of pure carbohydrate certainly
alters the ratios of plasma amino acids (Teff et al. 1989).
The mediator of this alteration is the glucose-induced
release of insulin.
It follows that many mechanisms exist to monitor the
activity of glucose released by ingested carbohydrates.
Therefore, it should be possible to demonstrate a relation-
ship between consumed carbohydrates and a modulation of
the expression of appetite. Although sweet carbohydrates
induce some positive feedback for eating through the
induction of oral afferent stimulation by sweet receptors,
this should be countered by the potent inhibitory action
via post-ingestive and post-absorptive mechanisms (see
below for effects of carbohydrates on cognitive perform-
ance and mood).
Carbohydrates and appetite: experimental evidence
On the basis of studies on rats, it was argued some years
ago that: if the cumulative inhibitory effects of carbo-
hydrate on feeding are indeed energostatic . . . then any
substance that can readily be used by the animal to provide
energy should produce an appropriate food intake compen-
sation over a period of several hours after loading (Booth,
1972). Studies have shown that this is also the case for
man. A variety of carbohydrates, including glucose, fruc-
tose, sucrose, maltodextrin and polysaccharides, exert
measurable effects when given in a preload or an experi-
mental meal. That is, they suppress later intake by an
amount roughly equivalent to their energy value, although
the time course of the suppressive action may vary accord-
ing to the rate at which the carbohydrates are metabolised.
The preload paradigm is a sensitive procedure for dis-
closing the effects of glucose and maltodextrin (glucose
polymer) on appetite when these carbohydrates are deliv-
ered in a natural food product. For example, in a study
using the uncoupling design (Blundell et al. 1988), the
actions of sweetness and energy content of a yoghurt pre-
load were separated experimentally by using judicious
combinations of glucose, maltodextrins, high-intensity
sweeteners and the yoghurt base (Rogers & Blundell,
1989). After consumption, feelings of hunger were sup-
pressed significantly by the high-energy preloads (glucose
and maltodextrins) to a greater extent than after the
lower-energy loads. Reduction of energy intake in the
test meal was proportional to the energy difference
(678 kJ) between the preloads. The accuracy of this adjust-
ment confirms the frequent, though not universal, finding
that human subjects adjust their voluntary intake in response
to covert manipulations of the carbohydrate content of a
food. In a more recent study, a comparison was made
between preloads containing maltodextrin or sucrose com-
pared with lower-energy loads with similar taste character-
istics. At one hour after consumption, both high-energy
preloads had suppressed intake compared with their low-
energy counterparts.
Taken together, these and other studies demonstrate that
the preload design is a sensitive procedure for assessing the
satiating power of carbohydrates. It appears that accurate
adjustment of subsequent energy intake (often called
energy compensation) is evident for glucose and
 Psychological and behavioural functions
maltodextrin. Measurement at one hour seems to be just about
perfect for detecting the effects of these carbohydrates on
satiety. However, this interval may not be appropriate for
other carbohydrates (or for other macronutrients).
Carbohydrate and the satiety index
Over twenty years ago it was suggested that: it could be of
great value to have tables showing the energy satiety ratio
of all the common foods to indicate their potential for caus-
ing over nutrition (Heaton, 1981). Subsequently, the term
satiating efficacy was proposed as an index to compare
the satiating potency of different foods (Kissileff, 1984).
A further measure explored by Holt et al. (1995, 1999)
relates the satiety response following a food to the satiety
response following a reference food (white bread). Iso-
energetic preloads (1000 kJ) of thirty-eight foods were con-
sumed, following which subjective ratings of motivation to
eat were completed over two hours. Significant differences
in the satiating effects of the foods were seen, indicating
that different foods differ in their satiating capacities.
This novel approach quantifies the effects of a food on
satiety and was based on the idea of developing a satiety
index analogous to the glycaemic index. These studies
showed that high-carbohydrate foods had a high satiety
index measured by the satiety area under curve (AUC).
The satiety index was positively correlated with the carbo-
hydrate content of the foods tested.
A further measure has been termed the satiety quotient,
which permits a measure of the effect of a food on satiation
(within an eating episode) and satiety (the effect following
eating). This allows determination of the degree of sup-
pression of hunger that can be brought about per unit of
food energy consumed (Green et al. 1997) and permits the
tracking of changes over time. Using data from a variety
of foods freely consumed, it was demonstrated that
hunger was reduced more per kJ of carbohydrate food con-
sumed than of high-fat food but the strength of this
varied with time. This finding confirms the proposal that,
joule for joule, carbohydrate is more satiating than fat
(Rolls et al. 1994).
Evidence from nutritional interventions
The results from short-term studies show clearly that
carbohydrates exert a marked influence over the expression
of appetite. Real-life intervention studies have demon-
strated that this degree of appetite control has practical
consequences. In one study, people identified as snackers
were encouraged to eat 25 % of their daily energy intake
from high-carbohydrate or high-fat snacks during separate
three-week periods. The subjects incorporated the snack
foods into their normal eating repertoires and maintained
a normal life-style. The study was monitored very carefully
to ensure compliance to the experimental requirements and
to assess food intake from the snack foods and from
the rest of the eating pattern. Measures of total energy
intake indicated that food consumption was physiologically
valid and quite in keeping with body size and the level of
physical activity. First, generous consumption of these
palatable high-carbohydrate foods did not generate abnor-
S191
mally high energy intakes; indeed daily energy intake
was an average of 364 kJ less than with the high-fat
snacks a value that could lead to a weight loss of
approximately 35 kg over the course of a year. Second,
the high-carbohydrate foods significantly lowered the total
daily fat intake down to a level recommended in the
National Dietary Guidelines (Lawton et al. 1998). Conse-
quently this high-carbohydrate intervention, which was
easily achieved, exerted a notable control over appetite
and the pattern of nutrient intake.
In a series of studies carried out in Scotland, subjects
were required to increase their consumption of high-
carbohydrate breakfast cereals by 60 g/d for twelve weeks
(Kirk et al. 1997). This intervention resulted in a 54 %
reduction in energy from fat and a 51 % increase in
energy from starch. A second study with overweight men
(body mass index 294 kg/m2), who were required to
increase their intake of high-carbohydrate breakfast
cereal by 90 g/d (approximately three bowls), led to an
increase in carbohydrate consumption from 40 to 47 % of
total energy (Crombie & Kirk, 1999). The positive changes
in carbohydrate intake were accompanied by small weight
losses.
Taken together, these studies show that a high intake of
carbohydrate foods can lower fat intake and does not com-
promise energy balance.
Fat and appetite control
The effect of dietary fat on appetite and energy balance has
been extensively researched and reviewed in the last ten
years (Blundell et al. 1996). Owing to the high palatability
and energy density of fatty foods, much of the evidence
indicates the potential of high-fat products to cause over-
consumption (Blundell & Macdiarmid, 1997). However,
there exists the fat paradox which arises from the conflict
between the capacity of fat to generate potent fat-induced
satiety signals and the presence of high-fat hyperphagia
(Blundell et al. 1995).
Because of the potency of fatty foods to stimulate a posi-
tive energy balance (partly through the process of passive
consumption), it is important to develop procedures to
mitigate this action. Three possibilities will be discussed
here: (1) fat substitutes or fat mimetics, (2) specific types
of fatty acids or triacylglycerols, and (3) modification of
fats through processes such as fractionation.
Fat substitutes: evidence from Olestra. Olestra is the
name of one of a class of compounds called sucrose polye-
sters. Olestra has the organoleptic properties of natural fat
but it is not hydrolysed by lipases. Therefore, following
consumption, it cannot produce metabolisable energy and
can be regarded as possessing no calories. However, Oles-
tra can be used in cooking and baking in similar ways to
natural fat and it can therefore be incorporated into a
wide range of foods. As a substitute for fat it therefore
effectively reduces the fat content of foods and the fat
energy absorbed by the body. Undigested Olestra passes
straight through the gastrointestinal tract. The key issue
is whether or not the consumption of Olestra-containing
foods with their reduced fat and energy contents
S192 L. Dye and J. Blundell
will lead to subsequent compensation, i.e. a later increase
in the intake of fat and/or energy.
In short-term studies, one initial experiment in healthy,
lean, young adult males found that the substitution of
20 or 36 g of fat by Olestra at a breakfast meal (Blundell
et al. 1991; Rolls et al. 1994) led to full compensation
during the course of the rest of the day. However, this
study is atypical (probably because of the highly selected,
well regulating subjects), and most studies have shown that
the replacement of dietary fat with Olestra during one day
does not cause full compensation during the next twenty-
four hours. The replacement of 55 g of fat by Olestra at
either lunch or dinner (Cotton et al. 1996) did not alter
hunger or food intake during the test day or the following
day. In a similar study, in which 55 g of Olestra was sub-
stituted in either three meals or five snacks across the
course of one day, no effect was detected on hunger or
food consumption. Therefore these studies indicate that
Olestra can lead to a short-term reduction in the consump-
tion of fat and energy.
Similar results have been obtained in longer-term
studies. For example, when lean and obese subjects substi-
tuted an average of 26 g of fat per day for two weeks (Hill
et al. 1998), total energy intake was reduced by 8 % and fat
intake by 11 %. This meant that subjects compensated for
only 20 % of the substituted energy.
However, the degree of compensation does appear to
vary with the absolute level to which fat is reduced.
When 55 g of Olestra reduced the percentage energy
from fat in the diet from 32 to 20 %, a compensation of
67 % was observed (Cotton et al. 1996). However, only
21 % compensation was seen with a reduction of fat
energy from 43 to 32 %. Taken together, several studies
suggest that a reduction of fat energy down to generally
recommended values (30 32 % of food energy per day)
will not generate a strong hunger drive or tendency towards
compensation. Interestingly, it does not seem to matter
whether or not subjects have knowledge about the presence
of a fat substitute in the foods being consumed (De Graaf
et al. 1996; Miller et al. 1998). Importantly, in one long-
term trial lasting for nine months, an Olestra-containing
diet (reducing fat from 32 to 25 % of energy) led to a
reduction of 63 kg in body weight. These outcomes
suggest that this strategy of fat substitution could be a
useful approach to functional foods for appetite control.
Enhancing satiety: fatty acids. It has been noted earlier
that foods containing a high percentage of fat have the
capacity to promote over-consumption in obese (Lawton
et al. 1993; Green et al. 1994a) subjects. This effect,
which occurs during the process of satiation, is almost
certainly due to the high energy density of high-fat
foods, but the somewhat weaker effect of fat on satiety
is probably due to the physiological action of fat in
generating post-ingestive inhibitory signals. Therefore the
demonstration of varying effects of fats on satiety signal-
ling could be useful in the development of fat-containing
foods that modulate satiety. The induction of physiological
satiety signals may well depend, at least in part, on the
composition of fatty acids in the particular fats used.
Two prominent structural features of fatty acids are their
chain length and degree of saturation. In principle it is
possible to manipulate both of these aspects. At the present
time only a few studies have investigated these effects.
In two studies in which chain length was manipulated,
medium-chain triacylglycerols have been shown to have
greater satiating power than long-chain triacylglycerols
(Rolls et al. 1988; Stubbs & Harbron, 1995). A further
study kept chain length constant and varied the degree of
saturation (Lawton et al. 2000). This short-term study
indicated that a meal containing largely monounsaturated
fat induced a weaker satiety than polyunsaturated fat or a
blend of saturated fat and monounsaturated fat. A second
study produced a similar but less decisive outcome. Taken
together, these and other studies suggest that monounsatu-
rated fatty acids may exert a relatively weak effect on
satiety. On the other hand, polyunsaturated fatty acids
such as linoleic acid are worthy of further investigation
for their relatively strong effect.
Enhancing satiety: novel treatments. It is known that
fat, introduced directly into the gastrointestinal tract, can
lead to an inhibition of eating that is mediated, at least in
part, by the slowing of gastric emptying (Welch et al.
1985). The importance of these gastrointestinal responses
means that if ingested fat (or fat-containing products) can
resist digestion and reach the intestine, they may be able
to induce a stimulation of satiety. One possible mechanism
involved here is the so-called ileal brake phenomenon.
This concept has been invoked to explain the action of a
specific treatment of fatty acids incorporated into yoghurt
as an emulsion.
This emulsion, known, as Olibra, is a food ingredient
containing fractionated palm oil and fractionated oat oil
in the proportion 95:5. This emulsion contains more
palmitic and linoleic acids, and less oleic acid, than milk
fat. A yoghurt containing 5 g of the Olibra fat was given
to subjects as a breakfast. In separate studies in lean
(Burns et al. 2000) and obese (Burns et al. 2001) subjects,
this single administration significantly reduced food intake
throughout the whole day. This effect, which was accom-
panied by an increased sense of fullness and decreased
hunger, was therefore maintained for up to eight hours.
This is an extremely powerful effect. If an action of this
intensity can be maintained with repeated administration,
then the product could well become a functional food for
appetite control. However, long-term weight-loss trials
are required before any specific obesity-controlling effect
can be claimed. Still, this novel manipulation suggests a
potential route for developing fats with an enhanced
action upon satiety.
Interim summary for satiety
Analysis of experimental studies indicates that all foods
share certain common features (weight, volume, texture)
which contribute to their effects on satiation and satiety.
However, the nutritional composition also plays a key
role and markedly modulates the intensity and duration
of satiety and the strength of satiation. This means that
the macronutrient composition of foods can favour con-
sumption (and over-consumption) and also limit the
amount of food willingly consumed and the motivational
feelings associated with eating. The macronutrients
 Psychological and behavioural functions S193

protein, carbohydrate and fats differ in their mechanisms of
inducing satiety and also vary in their impact on the size
and frequency of eating episodes. Therefore many oppor-
tunities exist for targeted manipulations of proteins, fats
and carbohydrates in the development of functional foods
for appetite control.
Biomarkers for appetite
A biomarker can be defined as a surrogate endpoint that
can be used instead of some clinical endpoint. How can
this be applied to a behavioural or psychological function
such as appetite control? If the ultimate objective of a func-
tional food for appetite control is the reduction or main-
tenance of food intake, then a surrogate or proxy for
this could be the profile of subjective hunger that oscillates
with the pattern of food intake and is usually well corre-
lated with the amount of food consumed (when hunger
fails to correlate with food eaten there are usually good
reasons). Consequently, it can be proposed that subject-
ively perceived hunger translated into an objective end-
point by means of the visual analogue rating scale can
be regarded as an objective biomarker for appetite control.
Functional foods for cognitive performance
Measurement of cognitive processes
In principle, a large number of mental or cognitive tasks
have the potential to disclose the effects of foods. How-
ever, in practice, a limited number of tests have been
used. Performance tasks reflect a number of cognitive func-
tions; for example, perception, memory, attention and
arousal, information processing, accuracy and speed of
movement (Bellisle et al. 1998). These tasks represent
single components of performance, which form part of
more complex skills and abilities (see Table 1); for example,
car-driving ability or operating machinery. Cognitive per-
formance encompasses not only measures of speed (reaction
time) but also of processing accuracy (measures of accu-
rate and inaccurate detection). Interventions may elicit

Table 1. Functions assessed by cognitive tests

Function Example of tests Common components of task

Reaction time (decision time and Simple (SRT) or choice (CRT) Stimulus appears (visual or auditory)
movement time) and S must make a single response,
usually by depressing a key. In CRT
one of a number of stimuli may
appear and S must make one of
two responses (e.g. left hand, right hand)
according to the type of stimulus
Vigilance (also known as attention), Search tests, e.g.
rapid information processing or categoric search Detection of stimulus items from particular
continuous performance categories
Digit symbol substitution S must replace digits with symbols
Stroop S must attend to certain features of stimuli
and ignore others
Bakan A series of numbers are presented visually
in rapid succession on a VDU screen.
S has to respond by depressing a key
when a sequence of three odd or three
even digits is detected
Visual information processing Critical flicker fusion threshold S must detect flicker and fusion of light
Frontal executive Immediate recall S shown a list of stimuli at a given rate
Working memory (e.g. one per second). At end of
(short-term memory) presentation S recalls the stimuli
Verbal memory S must recognise rather than recall
Spatial memory S must discriminate or recognise patterns
Associative memory
Word recognition
Pattern comparison
Immediate memory Digit span S must remember (recall) series of items
in forward or reverse order
Reasoning Arithmetic, logical, grammatical or S must process and indicate if stimulus
semantic (phrase, equation or expression) is true
or false
Psychomotor performance Driving simulation Tracking and deviation on a course are
(tracking ability) monitored continuously
Pursuit rotor S must trace a shape (maze) with a stylus
under time pressure. Error score
computed
Visuo-spatial motor task Simulator/driving task
Tapping task S must tap in rapid succession to a key

CRT, choice reaction time; S, subject; SRT, simple reaction time; VDU, visual display unit.
S194 L. Dye and J. Blundell
changes in function in some or all of these performance
components.
In choosing a task much depends on whether an immedi-
ate (short-term) effect of food is expected or whether a
chronic long-term adaptation to a diet is being examined.
The former and arguably more simple approach is the
focus of most current research. Measures that are fre-
quently utilised in assessing the effects of nutrients include
reaction time, attention and memory. Other tests have also
been used but selection is frequently based on accessibility
rather than likely effect of a nutrient on a particular cogni-
tive process. The proposed mechanism of action of a nutri-
ent in terms of its effects on blood glucose or other
neurotransmitter systems and the likely effects of changes
in these neurobiological parameters on cognitive processes
have not been considered. Thus a failure to detect an effect
of a particular macronutrient on a particular aspect of cog-
nitive function may be due to a true lack of effect, because
the neuropsychological processes needed to perform the
task are unaffected by the nutrient ingested. The effect of
compensatory effort in experimental situations the
Hawthorne effect (Roethlisberger, 1997) is well
known and can account for null results. Alternatively, the
intervention may affect only a small component of the
task (e.g. reaction time or decision time) such that an
effect on overall performance is not detected. Recent
advances in computer technology have meant that these
tasks can be administered in a structured, carefully con-
trolled fashion, with extremely accurate measurement of
processing and reaction time, correct responses and error
rates. Examination of the pattern of effects on sub-pro-
cesses within a task or on simple unitary measures of cog-
nition may be helpful in determining the mechanism of
action or neurotransmitter systems involved or affected
by the nutrient ingested.
Effects of foods on cognitive performance
Various studies have examined the effects of nutritional
interventions across a range of cognitive domains. The
effects of these manipulations on each of the major areas
of cognitive performance are summarised below.
Reaction time. Reaction time may be measured as a
response to a variety of stimuli. Visual or auditory stimuli
are presented and response time is measured. Simple or
choice reaction time can be employed, requiring the par-
ticipant to respond selectively to stimuli which have par-
ticular features, e.g. frequency, colour, spatial position,
tone, etc. Thus the reaction time task can comprise more
than one cognitive process. In addition, the reaction time
recorded can be broken down into movement and detection
components.
Studies that examine the effects of macronutrients on
reaction time are described in Table 2. The main con-
clusions of this body of research can be summarised as
follows.
1. Effects are more clearly demonstrable in the morning
(particularly breakfast studies).
2. Nutritional interventions, which facilitate a rise in
blood glucose, enhance performance of reaction time
tasks.
3. The findings of studies of macronutrients given in
combination suggest that fat may slow reaction time;
carbohydrate can impair peripheral processing and
reaction time, depending on the time of day and type
of carbohydrate ingested as well as ratio of carbo-
hydrate to protein.
4. Foods that facilitate speed of reaction should enhance
glucose availability, e.g. high glycaemic index. There
have been no studies that specifically compare the
effects of foods with different glycaemic indices.
One study (Kaplan et al. 2001) found similar improve-
ments in memory after high- and low-glycaemic-index
carbohydrates in the elderly.
Attention and vigilance. Vigilance or attention tasks
have been included in a large number of studies that aim
to assess the effects of nutrients on performance. The
nature of the task employed often differs and it is some-
times difficult to determine whether the cognitive processes
that may be involved in each task are similar or com-
parable. Digit symbol substitution and the Bakan test
have frequently been employed (see Table 3). However,
some studies do not adequately describe the specific test
used.
The effects of nutrients on attention and vigilance, as
described in Table 3, can be summarised as follows.
1. Attention can be difficult to disrupt in the first half of
the day but seems susceptible to fat ingestion at lunch
and later. It can be disrupted by high carbohydrate and
effects of this are most clear in terms of response time
measures.
2. Effects could be due to an impairment of either detec-
tion or movement, but work on reaction time measures
would suggest that movement is the most vulnerable
cognitive process.
3. Although participants appear to be slower, there is some
suggestion that they are also more accurate, which
could imply criterion shifts.
4. Protein appears to influence susceptibility to distraction.
5. The suggestion that attention can be enhanced by
energy requires verification.
Memory. Examination of functional food effects on
measures of memory has focused on short-term or working
memory (see Table 4). Tests employed usually require
immediate recall of a word list or similar stimuli presented
at a fixed rate, after which the subject is required to repeat
or write down as many of the stimuli as possible in a fixed
period of time. The number of items in the word list is
usually twenty to thirty in normal volunteer samples and
seven plus or minus two words should be the average
number recalled. Some samples, notably students, recall
almost double this rate, partly due to the effects of training
and intelligence. Two effects are common in memory
tasks. Primacy and recency effects relate to the increased
likelihood of recalling the earliest and latest stimuli from
a list, although few studies examine the order of words
recalled or request that the subject recalls the stimuli in
the order presented. The nature of the stimulus words are
 Psychological and behavioural functions
also important since efficiency of recall can be affected by
differences in the frequency, concreteness and imagery of
the words used as stimuli. It is therefore important that
memory tasks select and match stimuli on these features.
Other types of memory tasks sometimes employed include
delayed recall, where the subject is required to recall a list
presented some time earlier (usually ten to twenty minutes)
and has had to perform other tasks in the interim period.
Recognition recall requires the subject to identify stimuli
that were presented earlier from a list that includes
unseen, distractor stimuli. Recognition recall is generally
easier than free recall. Pattern comparison is another
form of memory task that examines spatial memory per-
formance. Paired associate learning involves remembering
two paired but semantically unrelated stimuli and recalling
one of the pair when the other is presented. Although there
are many variants of memory tasks, those used in the
assessment of nutrients have been limited.
A summary of the effects of nutrients on memory, out-
lined in Table 4, is given in the following.
1. Glucose appears reliably to improve memory, particu-
larly in vulnerable subjects such as the young and
elderly.
2. Pure carbohydrate in other forms improves memory
but can impair peripheral processing, attention and
reaction time, depending on the time of day, type of
carbohydrate ingested and ratio of carbohydrate to
protein. It is also associated with feelings of fatigue.
3. The carbohydrate:protein of a meal can produce
depletion or enhancement of tryptophan (Klassen
et al. 1999). Where tryptophan availability is
enhanced, faster memory scanning has been demon-
strated in stress-prone subjects who may have receptor
sensitisation of the serotoninergic system or 5-
hydroxytryptamine (5-HT) deficiency (Markus, 1999).
4. Tryptophan depletion selectively impairs memory
consolidation in normal volunteers (Riedel et al.
1999).
5. Thus memory tasks appear to be important measures
of cognitive function, which are susceptible to nutri-
tional manipulations.
Other aspects of performance. A large number of cog-
nitive tests can be used to examine the effects of nutrients
on performance and which measure functions other than
attention and memory.
Psychomotor performance involving hand eye coordi-
nation can be used to reflect the kind of cognitive pro-
cesses involved in operating machinery or driving.
These tasks may feature reaction time measures but
are generally more complex performance indicators
than the tests reviewed in the section on reaction time
above. The effect of macronutrients on these perform-
ance measures is shown in Table 5. The simplest
form of psychomotor task, which has been used reason-
ably widely in this area, is the tapping task. Dieters per-
formed better on a finger-tapping test in the morning
(Deijen et al. 1989). Lloyd et al. (1994, 1996) also
used a tapping task to assess breakfast and lunch
manipulations.
More complex psychomotor tasks involve tracking
S195
and control. Driving can be considered to be an ecolo-
gically valid psychomotor task and real or simulated
driving tasks are used to establish the effects of drugs
or other interventions on actual driving and driver
safety. There have been a small number of studies
examining nutrient effects on performance, which have
used tracking or driving simulator tasks. Performance
of a tracking task was better in a group of cadets
after repeated consumption of a tyrosine-rich drink
than in the group supplied with a carbohydrate-rich
drink (Deijen et al. 1999). Smith & Rich (1998) exam-
ined the effect of no snack, a chocolate bar or an equi-
caloric portion of cheese and biscuits on simulated
driving. Following consumption of the chocolate, sub-
jects drove more carefully and hit the side of the
track less often. It is difficult to determine from this
manipulation whether any specific ingredient in the cho-
colate was predominantly affecting the performance
although the caffeine content is not likely to be high
enough to have an alerting effect. An alternative expla-
nation is that mood may have been modified differen-
tially across the three conditions (foods were not
matched for taste, appearance and palatability) and this
could have mediated the performance effects observed.
Other tests assess perceptual processes such as vision.
Such tests have been used widely in assessing the import-
ance of the essential fatty acid, docosahexaenoic acid, in
neural maturation, visual acuity and brain function in
ageing (Horrocks & Yeo, 1999). The effect of docosahexa-
enoic acid supplementation on cognitive performance in
healthy young adults has not been examined specifically.
Another widely used test in psychopharmacology is Criti-
cal Flicker Fusion Threshold, which involves the dis-
crimination of flicker from fusion and is an index of
cortical arousal and visual information processing capacity.
Visual information processing (measured by Critical
Flicker Fusion Threshold) was increased significantly by
a decaffeinated Coca-Cola drink with added tryptophan,
compared with a decaffeinated Coca-Cola drink alone
(Cunliffe et al. 1998). This effect of amino acid manipu-
lation suggests impairment of performance follows protein
consumption. Fat ingestion has also been shown to
decrease flicker fusion frequency, indicating increased cen-
tral fatigue or decreased visual information processing
capacity (Cunliffe et al. 1997).
Other cognitive tasks such as reasoning tasks (involving
logical, semantic and grammatical reasoning), problem-
solving tasks and arithmetic tasks have also been employed
in studies of nutrient effects. Addition and sentence verifi-
cation tasks are not affected by missing breakfast (Dickie
& Bender, 1982; Smith & Kendrick, 1992; Smith et al.
1994a). Smith et al. (1994b ) showed that a high-fat
lunch did not affect logical reasoning. In comparison
with performance after a diet soft-drink (containing aspar-
tame), subjects solved more arithmetic problems in a
shorter time after consuming calorie-containing yoghurts
(Kanarek & Swinney, 1990).
Providing a supplementation of branched-chain amino
acids has also been shown to prevent impairment of per-
formance of shape rotation and identification tasks and to
improve performance of the Stroop task after exercise
 S196
Table 2. Effect of nutrients on reaction time

Reference Sample Study design Cognitive testing Effect on performance of task Comments

Spring 184 healthy Two meals for BK or lunch: Auditory RT No effect No pre-meal measure of
et al. adults HCHO (non-dairy sherbet) Assessment 2 h after performance
(1983) (129 M, HP (turkey breast) consuming the meal Between-subjects design
fifty-five F), Foods were not BK or lunch
18 65 years standard items
Lieberman Young men Two lunches: Auditory RT Slower RT to an auditory No baseline measures of
et al. HCHO (pita bread) Assessment after lunch stimulus after the HCHO meal cognitive performance
(1983) HP (turkey breast) than after the HP meal
Smith & Forty-eight Two conditions: Serial self-paced CRT Lunch consumption impaired
Miles subjects lunch (detection RT and detection time (RT) on CRT
(1986 b,c) no lunch movement time)
Assessment before and after Movement time slower in early
lunch afternoon than in morning,
independent of meal
consumption
Smith Five M and Three iso-energetic lunches RT-focused attention test Slower RT to peripheral targets
et al. six F (377 MJ in M, 272 MJ in F): after the HCHO meals than
(1988) HCHO, high in starch Assessment before and after the HP meal
(40 % starch, 15 % sugar, 75 min following lunch

L. Dye and J. Blundell

15 % protein, 30 % fat)
HCHO, high in sugar
(15 % starch, 40 % sugar,
15 % protein, 30 % fat)
HP (14 % starch, 1 % sugar,
55 % protein, 30 % fat)
Lloyd et al. Eighteen subjects Three iso-caloric lunches SRT Longer RT after LF and HF The MF/CHO meal was most similar
(1994) (fifteen F, (about 292 MJ): Assessment 30 min before lunches, compared with the MF in size and macronutrient content
three M), LF/HCHO (290 MJ, 29/54 % en) lunch, 30 min, 90 min and lunch, at 90 min post ingestion to the subjects habitual lunch.
mean age MF/MCHO (295 MJ, 45/42 % en) 150 min after finishing Results may reflect responses to
27 years MHF/LCHO (301 MJ, 62/24 % en) lunch changes in normal intake rather
than to the specific macronutrient
composition of the meals
Smith Experiment 1: Six conditions: SRT No effect of BK on performance Between-subject design
et al. forty-eight no BK decaffeinated coffee Five-choice serial response on SRT or CRT BK differs from lunch
(1994c ) university cereal toast BK (equivalent to time (CRT)
students LF) decaffeinated coffee Assessment 30 min before Caffeine enhanced performance of
(twenty-four cooked BK (equivalent to BK, 60 and 120 min after SRT but had no effect on CRT
M, twenty- HF) decaffeinated coffee the start of BK
four F) no BK caffeinated coffee
cereal toast BK (equivalent to
LF) caffeinated coffee
cooked BK (equivalent to
HF) caffeinated coffee
Smith Experiment 2: Four conditions: SRT No effects on tasks Effects are different from those
et al. forty-eight three-course meal (502 Five-choice serial response observed after lunch or BK
(1994c) university 628 MJ) decaffeinated drink task (CRT)
students three-course meal (502 Assessment before the meal
(twenty-four M, 628 MJ) caffeinated drink and 90 min and 180 min after
twenty-four F) no meal decaffeinated drink the start of the meal
no meal caffeinated drink
Wells & Eighteen males Two meals: SRT No effect on SRT
Read (groups A and B: HF/LCHO (318 MJ, 55/40 % en) Serial choice reaction Overall little change in
(1996) nine M in each) LF/HCHO (360 MJ, 88/7 % en) performance
Group A: ate meals at 10.30
(brunch)
Group B: BK (08.15) and test meal
at 12.30 (lunch)
Cunliffe Sixteen healthy 1672 kJ at BK of approx. 200 ml: Simple visual reaction time Slower after CHO
et al. volunteers pure CHO (maltodextrin) Hourly for 4 h post ingestion No effect of fat on RT
(1997) (nine M, pure fat (long-chain triacylglycerol
seven F), emulsion)
aged 20 47 years Control: mixed CHO, fat and protein
(mean, 29.7 years) (55 % CHO, 15 % protein, 30 % fat,
same sources of CHO and fat)
Lloyd et al. Sixteen healthy Four conditions: SRT No clear difference in BK provided in the study was higher
(1996) habitual BK eaters no BK Assessment beginning performance between the in en than usual BK
(fourteen F, two M), three iso-caloric BK (251 MJ): 30 min before, and 30 min, four conditions
age 26 years LF/HCHO (27/62 % en) 90 min and 150 min after BK
MF/MCHO (44/47 % en)

Psychological and behavioural functions

HF/LCHO (56/34 % en)
Habitual BK
(105 MJ, 25 % fat,
65 % CHO)
Fischer Fifteen young males 1670 kJ at breakfast, Simple and choice RT Fat improved RT performance
et al. spoonable creams of: Hourly for 3 h post ingestion compared with baseline at 1, 2
(2001) pure fat and 3 h, and reduced error rate
pure CHO
pure protein
No effect of CHO on SRT
CHO improved CRT at 1 h
CHO impaired CRT at 2 h
Protein improved CRT at 2 h
Lluch et al. Thirty-two Two iso-caloric BK (172 MJ) Simple and choice RT Slower RT after HF lunch than
(2000) volunteers and lunches (310 MJ for F, Assessment just before and after LF lunch
(sixteen M and 414 MJ for M) 30 min after finishing lunch No effect of habitual diet on
sixteen F; LF/HCHO (BK: 22/67 % en, performance
seventeen lunch: 20/61 % en)
LF and fifteen HF/LCHO (BK: 48/42 % en,
HF consumers) lunch: 54/32 % en)

BK, breakfast; CHO, carbohydrate; CRT, choice reaction time; en, energy; F, female; HCHO, high carbohydrate; HF, high fat; HP, high protein; LCHO, low carbohydrate; LF, low fat; M, male; MCHO, medium carbo-
hydrate; MF, medium fat; RT, reaction time; SRT, simple reaction time.

S197
 S198
Table 3. Effect of nutrients on attention tasks

Reference Sample Study design Cognitive testing Results Comments

Spring et al. 184 healthy adults Two meals for BK or lunch: Dichotic shadowing (sustained Impairment on the sustained No pre-meal measure of
(1983) (129 M, fifty-five HCHO (non-dairy sherbet) and selective aspects of selective attention task after performance
F), 1865 years HP (turkey breast) attention) HCHO lunch compared with HP Between-subjects design
Assessment 2 h after in subjects over 40 years of age Foods were not BK or
consuming the meal lunch standard items
Lieberman Young men Two lunches: DSST Poorer performance on a DSST No baseline measures of
et al. HCHO (pita bread) Assessment after lunch after the HCHO meal than after cognitive performance
(1983) HP (turkey breast) the HP meal
Smith et al. Five M and six F Three iso-energetic lunches Two tasks: Slower RT to peripheral targets
(1988) (377 MJ in M, 272 MJ in F): focused attention test after the HCHO meals than after
HCHO, high in starch search test the HP meal
(40 % starch, 15 % sugar, Susceptibility to distracting stimuli
15 % protein, 30 % fat) greater after the HP lunch than
HCHO, high in sugar after the HCHO meals
(15 % starch, 40 % sugar,
15 % protein, 30 % fat)
HP (14 % starch, 1 % sugar, Assessment before and 75 min
55 % protein, 30 % fat) following lunch
Deijen et al. Two groups: During three weeks, diet group Sustained visual attention task No effect on attention type tasks

L. Dye and J. Blundell

(1989) control (n=9) had a daily intake of 110 g Pattern comparison
diet (n=10) protein, 320 g CHO and 80 g fat DSST
21 27 years BK: 70 g protein (63 %), Assessment: baseline scores
25 g CHO (8 %) (last day of 3 weeks, after BK
lunch: 10 g protein (9 %), and dinner) and post-treatment
100 g CHO (31 %) scores (2 months later)
Benton Seventy young Glucose drink (50 g) after Reaction to Bakan test Improved decision time Between-subjects design,
et al. female baseline and 25 min later no dietary restrictions
(1994) students (25 g) or CHO-free placebo on test day
drinks (sweetened with
aspartame and acesulfame K)
Kelly et al. Experiment 1: Four lunches varying in energy and Choice of tasks: Performance on the task was Limited number of
(1994) two groups fat and carbohydrate content poorer after compared with before subjects and
of six subjects the meal dimensions measured
LF/low en (180 MJ) DSST No systematic variations in Subjects were allowed
performance as a consequence of to choose among the
either the energy or macronutrient tasks
content of the meals
HF/high en (353 MJ) Assessment before and after Unrestricted access to
LCHO/low en (181 MJ) consuming the lunch caffeine and nicotine
HCHO/high en (351 MJ)
Lloyd et al. Eighteen subjects Three iso-caloric lunches Bakan task (sustained attention) No effect on sustained
(1994) (fifteen F, (about 293 MJ): attention
three M),
mean age LF/HCHO (290 MJ, Assessment 30 min before
27 years 29/54 % en) lunch, 30 min, 90 min and
MF/MCHO (295 MJ, 150 min after finishing lunch
45/42 % en)
HF/LCHO (301 MJ,
62/24 % en)
Smith et al. Forty-six Four lunches: Repeated digit vigilance task Slower but more accurate Between-subjects design
(1994a ) university small HF: 530 g, 540 MJ, Focused attention task response after HF lunch than
students 55 % fat Categoric search task after LF lunch on selective
(twenty M, large HF: 840 g, 544 MJ, Assessment before lunch and attention tasks at 90 min
twenty-six F) 58 % fat 90 min after the start of the meal No effect of fat content or weight
small LF: 600 g, 351 MJ, of the meals on vigilance tasks
19 % fat
large LF: 840 g, 368 MJ,
23 % fat
Smith et al. Experiment 1: Six conditions: Repeated-digits vigilance task No effect of BK on performance Between-subject design
(1994c ) forty-eight no BK + decaffeinated coffee on sustained attention tasks
university cereal toast BK (equivalent to Assessment 30 min before BK,
students LF) + decaffeinated coffee 60 and 120 min after the start of
(twenty-four M, cooked BK (equivalent to BK
twenty-four F) HF) + decaffeinated coffee
no BK + caffeinated coffee
cereal toast BK (equivalent to
LF) + caffeinated coffee
cooked BK (equivalent to
HF) + caffeinated coffee

Psychological and behavioural functions

Wells et al. Experiment 1: Two testing days with two Sustained attention task Lipid affected the speed and
(1995) five males consecutive 3 h duodenal Assessment at seven time accuracy of performance in the
infusions of: points: one before the first sustained attention task
100 g intralipid/l followed by infusion and six during the two
isotonic saline (9 g NaCl/l) infusions
isotonic saline (9 g NaCl/l)
followed by 100 g intralipid/l
(cross-over balanced design)
Experiment 2: Fixed BK (318 MJ) and two iso- Sustained attention task Responses to stimuli in the
eight males energetic lunches (366 MJ): sustained attention task slower
LF/HCHO (7/76 % en) after the HF than after the LF
HF/LCHO (64/18 % en) meal
Wells & Eighteen males Two meals: Bakan task Increase in the frequency of
Read (groups A and B: HF/LCHO (318 MJ, false alarms in the Bakan test
(1996) nine M in each) 55/40 % en) after the LF/HCHO brunch
LF/HCHO (360 MJ,
88/7 % en)
Group A: ate meals at
10.30 (brunch)
Group B: BK (08.15) and
test meal at 12.30 (lunch)
Lloyd et al. Sixteen healthy habitual Four conditions: Bakan task No clear difference in BK provided in the study
(1996) BK eaters (fourteen F, no BK Assessment beginning 30 min performance between the was higher in energy than
two M), age 26 years three iso-caloric BK (251 MJ): before, and 30 min, 90 min and four conditions usual BK
Habitual BK (105 MJ, LF/HCHO (27/62 % en) 150 min after BK
25 % fat, 65 % CHO) MF/MCHO (44/47 % en)
HF/LCHO (56/34 % en)
Lluch et al. Thirty-two Two iso-caloric BK (172 MJ) and Attention No effect on attention
(2000) volunteers lunches (310 MJ for F, 414 MJ for Assessment just before and No effect of habitual diet on
(sixteen M and M): 30 min after finishing lunch performance
sixteen F; LF/HCHO (BK: 22/67 % en, lunch:
seventeen LF 20/61 % en)
and fifteen HF HF/LCHO (BK: 48/42 % en, lunch:

S199
consumers) 54/32 % en)
 S200
Table 3. Continued

Reference Sample Study design Cognitive testing Results Comments

Kanarek & Experiment 1: Four conditions: BK (126 MJ) Attention Subjects responded faster in the Practice effects observed;
Swinney sixteen males followed by: Assessment 15 min following attention task after the caloric therefore, a practice
(1990) (1922 years) lunch + caloric snack eating of the snack snack than after the non-caloric day introduced in the
(confectionary snack) Subjects blind to caloric content snack second experiment
no lunch + caloric snack of snacks Pre-lunch performance
lunch + non-caloric snack (drink) not assessed (in both
no lunch + non-caloric snack (drink) experiments)
Experiment 2: Same conditions, but caloric snack No effect on attention
eight males replaced by fruit-flavoured yoghurt
Smith & Forty-eight Two conditions: Detection of repeated numbers Deficit in performance following Caffeine also consumed
Miles university lunch (three-course meal Assessment before and after lunch consumption on number following lunch
(1986a ) students from university refectory, chosen lunch task
(eighteen M, by subjects)
thirty F) no lunch
Smith & Forty-eight Two conditions: Stroop colour and word task Stroop not influenced by lunch
Miles subjects lunch (selective attention)
(1986b ) no lunch Assessment before and after
lunch

L. Dye and J. Blundell

Smith & Twenty-four Four conditions: Detection of odd or even digits Hit rate positively correlated with
Miles university lunch (soup, two sandwiches, extraversion, negatively correlated
(1986c ) students fruit pie and fruit) with neuroticism
no lunch Post-lunch dip in hits detected and
with or without noise RT in the 40 dB (quiet) but not
(75 dB v. 40 dB) 75 dB (noisy) condition
Smith et al. Forty-eight Four conditions: Repeated-digits vigilance No effects on task Effects are different
(1994c ) university three-course meal task from those observed
students (502 628 MJ)+ Assessment before the meal after lunch or BK
(twenty-four M, decaffeinated drink and 90 min and 180 min after
twenty-four F) three-course meal the start of the meal
(502 628 MJ)+
caffeinated drink
no meal +decaffeinated
drink
no meal + caffeinated
drink
Fischer Fifteen young males 1670 kJ at breakfast, Combi test (peripheral attention) Best performance after fat
et al. spoonable creams of:
(2001) pure fat Hourly for 3 h post ingestion Protein better than CHO
pure CHO CHO more accurate
pure protein Protein more efficient

BK, breakfast; CHO, carbohydrate; DSST, digit symbol substitution test; en, energy; F, female; HCHO, high carbohydrate; HF, high fat; HP, high protein; LCHO, low carbohydrate; LF, low fat; M, male; MCHO, medium
carbohydrate; MF, medium fat; RT, reaction time.
 Table 4. Effect of nutrients on memory tasks

Reference Sample Study design Cognitive testing Results Comments

Deijen Two groups: During three weeks, Pattern comparison Diet group performed more
et al. control (n 9) diet group had a daily intake Memory scanning task slowly in the morning on
(1989) diet (n 10) of 110 g protein, 320 g CHO, Assessment: baseline scores memory scanning task than
2127 years 80 g fat (last day of 3 weeks, after the control group
BK: 70 g protein (63 %), BK and dinner) and post-
25 g CHO (8 %) treatment scores (2 months
lunch: 10 g protein (9 %), later)
100 g CHO (31 %)
Markus Forty-three subjects: Two iso-energetic diets: Memory scanning task after Performance in HS subjects
et al. twenty-two HS, CHO-rich, protein-poor diet controllable or was better with the CHO-rich
(1998) twenty-one LS (66 % CHO, 4 % protein, uncontrollable stress diet than with the protein-rich
30 % fat) diet, but only after controllable
protein-rich/CHO-poor diet stress
(40 % CHO, 26 % protein, 33 % fat)
Deijen Twenty-one cadets Five daily doses of an Memory task Group with the tyrosine-rich drink Between-subjects design
et al. (two groups, of ten iso-caloric drink (107 MJ): Assessment before the performed better on a memory
(1999) and eleven, receiving protein-rich drink (2 g tyrosine) combat course and on the task than the group consuming

Psychological and behavioural functions

two different drinks) CHO-rich drink sixth day of the course the CHO-rich drink
Benton Seventy young female Glucose drink (50 g) after Memory task Improved recall after glucose drink Between subjects design, no
et al. students baseline and 25 min later No differential effect on primacy or dietary restrictions on test
(1994) (25 g) or CHO-free placebo recency day
drinks (sweetened with
aspartame and acesulfame K)
Martin & Eighty female students Four groups: Brown Petersen memory Poorer performance after fasting Between subjects design
Benton BK 50 g glucose drink (n 28) task Glucose improved memory in
(1999) BK placebo drink (n 25) fasted subjects Nature of BK consumed not
fasted 50 g glucose drink BK consumption associated with known/controlled
(n 12) better memory
fasted placebo drink (n 15) Rapidly falling blood glucose
associated with better memory
Kelly et al. Experiment 1: two Four lunches varying in energy Number recognition task In both experiments: Limited number of subjects
(1994) groups of six subjects and fat and carbohydrate Repeated acquisition task performance on the tasks was and dimension measured
content: Assessment before and after poorer after than before the Subjects were allowed to
LF/low en (180 MJ) consuming the lunch meal choose among the tasks
HF/high en (353 MJ) (Experiment 1); after BK and no systematic variations in Unrestricted access
LCHO/low en (180 MJ) lunch (Experiment 2) performance as a consequence to caffeine and nicotine
HCHO/high en (351 MJ) of either the energy or
macronutrient content of the meal
Experiment 2: two BK, lunch, afternoon snack
groups of six subjects different in energy, fat and
CHO content:
LF (309 MJ)
MF (516 MJ)
HF (705 MJ)
LCHO (295 MJ)
MCHO (508 MJ)
HCHO (697 MJ)
Lloyd et al. Eighteen subjects Three iso-caloric lunches Free recall task (memory) No effect on memory
(1994) (fifteen F, three M), (about 293 MJ): Assessment 30 min before

S201
mean age 27 years LF/HCHO (290 MJ, 29/54 % en) lunch, 30 min, 90 min and
MF/MCHO (295 MJ, 45/42 % en) 150 min after finishing lunch
HF/LCHO (301 MJ, 62/24 % en)
 S202
Table 4. Continued

Reference Sample Study design Cognitive testing Results Comments

Smith Experiment 2: Four conditions: Four memory tasks: BK improved performance on Between-subjects design
et al. forty-eight no BK decaffeinated coffee free recall free recall and delayed
(1994) university cooked BK (equivalent to delayed recognition memory recognition memory tasks
students HF) decaffeinated coffee logical reasoning No effect on semantic memory Conclusions about the
(twenty-four M, no BK caffeinated coffee semantic reasoning task effects of BK on
twenty-four F) cooked BK (equivalent to Assessment 60 min before Impairment of the accuracy of performance
HF) caffeinated coffee BK, 45 and 105 min after the performing the logical depend on the type of
start of BK reasoning task task
Lloyd et al. Sixteen healthy habitual Four conditions: Free recall task No clear difference in BK provided in the study
(1996) BK eaters (fourteen F, no BK Assessment beginning 30 min performance between the was higher in energy than
two M), age 26 years three iso-caloric BK (251 MJ) before, and 30 min, 90 min four conditions usual BK
Habitual BK (105 MJ, LF/HCHO (27/62 % en) and 150 min after BK
25 % fat, 65 % CHO) MF/MCHO (44/47 % en)
HF/LCHO (56/34 % en)
Lluch et al. Thirty-two volunteers Two iso-caloric BK (172 MJ) and Associative memory (code No effect on memory task
(2000) (sixteen M and sixteen F; lunches (310 MJ for F, 414 MJ substitution)
seventeen LF and fifteen for M):
HF consumers) LF/HCHO (BK: 22/67 % en, Assessment just before and No effect of habitual diet on
lunch: 20/61 % en) 30 min after finishing lunch performance

L. Dye and J. Blundell

HF/LCHO (BK: 48/42 % en,
lunch: 54/32 % en)
Kanarek & Experiment 1: sixteen Four conditions: BK (126 MJ) Digit span recall (forwards Subjects recalled more digits Practice effects observed
Swinney males (1922 years) followed by: and backwards) in the backwards digit span Pre-lunch performance not
(1990) lunch caloric snack Assessment 15 min following test after the caloric snack assessed (in both
(confectionary snack) eating of the snack than after the non-caloric experiments)
no lunch caloric snack snack
lunch non-caloric snack (drink)
no lunch non-caloric snack (drink)
Parker & Thirty-three university Two conditions: Spatial memory Subjects who ate a BK took Between-subjects design
Benton students no BK Immediate recall less time to finish memory
(1995) (seventeen M, milk-based nutritional Assessment 2 h after BK tasks than subjects who did
sixteen F), mean age beverage not have BK
21 years No effect of BK on number of
errors
Smith Forty-eight students Four conditions: Free recall task No effects on memory tasks Effects are different from
et al. (twenty-four M, three-course meal (502 Delayed recognition memory those observed after
(1994c ) twenty-four F) 628 MJ) decaffeinated drink task lunch or BK
three-course meal (502 Assessment before the meal
628 MJ) caffeinated drink and 90 min and 180 min after
no meal decaffeinated drink the start of the meal
no meal caffeinated drink
Smith 144 volunteers Four conditions: Spatial memory task Subjects who consumed the Between-subjects design
et al. (seventy-two M, no BK decaffeinated coffee Working memory cereal BK performed better
(1999) seventy-two F), no BK caffeinated coffee Assessment before BK, on the spatial memory task
mean age 21 years cereal BK decaffeinated coffee 60 min and 120 min after than those in the no BK
cereal BK caffeinated coffee the beginning of BK condition

BK, breakfast; CHO, carbohydrate; en, energy; F, female; HCHO, high carbohydrate; HF, high fat; HP, high protein; HS, high-stress-prone; LCHO, low carbohydrate; LF, low fat; LS, low-stress-prone; M, male;
MCHO, medium carbohydrate; MF, medium fat.
 Psychological and behavioural functions
(Hassmen et al. 1994). Since exercise increases the plasma
concentration ratio of free tryptophan to other large neutral
amino acids leading to an elevation of 5-HT in the
brain, branched-chain amino acids may therefore minimise
central and mental fatigue during and after sustained
exercise.
Interim summary for cognitive performance
In summary, acute interventions with high carbohydrate
and low protein are sedating and anxiolytic. Protein-rich,
carbohydrate-poor interventions tend to produce arousal
and improve reaction time and vigilance. Nutritional inter-
ventions that facilitate a rise in blood glucose enhance per-
formance of memory and reaction time tasks. The decline
in glucose in hypoglycaemia impairs performance. Fat
appears to decrease alertness but this effect is often delayed
and the habitual diet is important in producing this
response. Acute effects vary with time of day. The effect
size for glucose is larger than for other macronutrient
manipulations and confounding variables are more easily
controlled. Chronic or longer-term intervention studies
are rarer but suggest that a protein-rich diet can increase
negative and decrease positive affect. Longer-term (habit-
ual diet) effects on performance are important because of
their relationship to certain disease states that carry with
them the risk of cognitive impairment. Poor glucose regu-
lation is associated with poor cognitive performance
although this has been examined only in healthy elderly
(Kaplan et al. 2000) and diabetic (Strachan et al. 1997)
populations.
Biomarkers for cognitive performance
Biomarkers for cognitive function are probably more diffi-
cult to measure than the actual function itself. We are cur-
rently lacking a strong theoretical base to relate
neurochemical or physiological activity to cognitive per-
formance. Biomarkers need to be identified in order to elu-
cidate the mechanisms by which food components may
affect performance. However, these mechanisms are un-
likely to be simple and therefore biomarkers are not an
easy route to identifying the effects of food on specific
functions.
It is also clear that cognitive performance is normally
well protected by a regulatory process that maintains a
stable output. This means that effectiveness of overt task
behaviour is difficult to disrupt or enhance but the cost
of maintaining efficient processing and performance
capacity may vary. Decrements are therefore difficult to
measure except in circumstances where large manipula-
tions or interventions have an effect on control mechan-
isms, e.g. drugs with strong effects on the central
nervous system. Any genuine effects will tend to be
subtle ones, and can only be detected reliably by measuring
both overt performance and the underlying maintenance
costs of performance protection. These should be detect-
able as latent decrements various markers of increased
regulatory strain seen, for example, in sympathetic
dominance in the autonomic nervous system, subjective
strain (increased effort, strain, fatigue as after-effect) and
S203
in impairment in secondary aspects of performance in com-
plex tasks where only the primary component is usually
protected. In this case a measure of increased selectivity
(narrowed attention) can be derived based on the relative
performance in the primary and secondary tasks. Decre-
ments should also be revealed by the use of tasks in
which upper control is a key factor, e.g. as used to assess
frontal lobe problems or fatigue after-effects (involving
effort, planning, sequential management). Thus any assess-
ment of cognitive performance should include measures of
costs as well as of overt performance, and assess both
upper and lower function (Hockey, 1997). One solution
might be to build a complex task from simpler com-
ponents, allowing one to test effects of cognitive com-
ponents alone (i.e. minimal control) and also combined
(with additional control requirement to manage the two
and plan shifts of focus between them). These behavioural
actions can be regarded as biological activities although
they clearly differ from a physiological index such as
blood glucose.
There are also some physiological indices that may be
good candidates as biomarkers of cognitive function.
These include measures of autonomic nervous system
activity, heart rate and electroencephalography, although
these are complex, variable and, currently, difficult to
work with in experimental settings where cognitive per-
formance is examined simultaneously. Blood glucose
level or rate of change has been used as a biomarker in a
number of studies. However, the action of glucose on the
cholinergic system may increase synthesis of acetylcholine
(Owens & Benton, 1994). Hence the performance-enhan-
cing effects of glucose on reaction time may not be
solely related to the increased availability of neural fuel.
Alternatively the effect may be produced by increased glu-
cose uptake in the frontal cortex in response to cognitive
load. The increase in heart rate associated with high cogni-
tive load supports this assertion. Other studies suggest that
stable performance is related to a balanced glucose metab-
olism and state of metabolic activation. Future markers
might include metabolites of neurotransmitters, e.g. 5-
HT, but the development of these will require an under-
standing of their involvement in simple and complex cog-
nitive activities. Moreover, the use of physiological
biomarkers relies on the assumption that peripheral
measures reflect central activity.
Functional foods for mood
Effect of food on mood and mood on performance
There are some well-documented effects of food on mood
and of mood on performance (for a review, see Rogers et al.
1995). Many studies have related the effects of food on
mood to the modulation of serotonin by carbohydrate
(e.g. Wurtman et al. 1981) but effects of consuming
other foods on mood have also been documented, e.g.
chocolates positive effects on mood have been attributed
to its orosensory properties (Macdiarmid & Hetherington,
1995). Carbohydrate produces a reduction in depression
in vulnerable samples such as premenstrually depressed
women and those with seasonal affective disorder (Wurtman
 S204
Table 5. Effect of nutrients on other tasks

Reference Sample Study design Cognitive testing Results Comments

Deijen Two groups: During three weeks, diet group Finger tapping Diet group performed better in the morning
et al. control (n 9) had a daily intake of 110 g protein, Assessment: baseline scores on finger tapping than the control group
(1989) diet (n 10) 320 g CHO, 80 g fat (last day of 3 weeks, after BK
2127 years BK: 70 g protein (63 %), 25 g and dinner) and post-treatment
CHO (8 %) scores (2 months later)
lunch: 10 g protein (9 %),
100 g CHO (31 %)
Deijen Twenty-one cadets Five daily doses of an Tracking task Group with the tyrosine-rich drink Between-subjects
et al. (two groups, of ten and iso-caloric drink (107 MJ): Assessment before the combat performed better on tracking task design
(1999) eleven, receiving two protein-rich drink (2 g tyrosine) course and on the sixth day than the group consuming the
different drinks) CHO-rich drink of the course CHO-rich drink
Kelly Experiment 1: two Four lunches varying in energy Choice of four psychomotor In both experiments: Limited number of
et al. groups of six subjects and fat and CHO content: tasks available: performance on the tasks was poorer subjects and
(1994) LF/low en (180 MJ) differential reinforcement of after than before the meal dimension
HF/high en (353 MJ) low-rate schedule of point no systematic variations in performance measured
LCHO/low en (181 MJ) presentation as a consequence of either the energy Subjects were
HCHO/high en (351 MJ) Assessment before and after or macronutrient content of the meals allowed to choose
consuming the lunch among the tasks
(Experiment 1); after BK Unrestricted access

L. Dye and J. Blundell

and lunch (Experiment 2) to caffeine and
Experiment 2: two BK, lunch, afternoon snack nicotine
groups of six subjects different in energy, fat and
CHO content:
LF (309 MJ)
MF (516 MJ)
HF (705 MJ)
LCHO (294 MJ)
MCHO (506 MJ)
HCHO (697 MJ)
Lloyd Eighteen subjects Three iso-caloric lunches Two-finger tapping task
et al. (fifteen F, three M), (about 293 MJ) Assessment 30 min before l
(1994) mean age 27 years LF/HCHO (290 MJ, 29/54 % en) unch, 30 min, 90 min and
MF/MCHO (295 MJ, 45/42 % en) 150 min after finishing lunch
HF/LCHO (310 MJ, 62/24 % en) CFFT Reduced after CHO and fat
Hourly for 4 h post ingestion
Cunliffe Sixteen healthy 1672 kJ at BK approx. 200 ml:
et al. volunteers (nine M, pure CHO (maltodextrin)
(1997) seven F), age pure fat (long chain triacylglycerol
2047 years emulsion)
(mean, 29.7 years) Control: mixed CHO, fat and protein
(55 % CHO, 15 % protein, 30 % fat
same sources of CHO and fat)
Lloyd Sixteen healthy habitual Four conditions: Two-finger tapping task No clear difference in BK provided in the
et al. BK eaters (fourteen F, no BK Assessment beginning 30 min performance between the study was higher
(1996) two M), age 26 years three iso-caloric BK (251 MJ) before, and 30 min, 90 min and four conditions in energy than
Habitual BK (105 MJ, LF/HCHO (27/62 % en) 150 min after BK usual BK
25 % fat, 65 % CHO) MF/MCHO (44/47 % en)
HF/LCHO (56/34 % en)
 Psychological and behavioural functions S205

BK, breakfast; CFFT, critical flicker fusion threshold; CHO, carbohydrate; en, energy; F, female; HCHO, high carbohydrate; HF, high fat; HP, high protein; LCHO, low carbohydrate; LF, low fat; M, male; MCHO, med-
et al. 1989). Carbohydrate has also been shown to increase

performance not

following lunch
feelings of sedation and fatigue (see Reid & Hammersley,

Caffeine also
1995 for a review of the effects of carbohydrate). These

consumed
assessed studies have not, however, considered the mediating effect
Pre-lunch

of mood on performance but rather have viewed a change

in mood as a result of food intake as separate from any
effects observed on performance. For instance, poorer
performance has been observed in samples of highly
restrained women or those currently dieting (Rogers &
problems and solved these problems

Green, 1993; Green et al. 1994b; Green & Rogers, 1995;

Reading times faster when subjects
had lunch than when they had no

Wing et al. 1995). Negative mood is also present in such

in less time after consuming the

afternoon, irrespective of lunch

Subjects solved more arithmetic

samples and it is difficult to untangle the effects of dieting

Deficit in estimation task in the
Slower RT after HF lunch than

yoghurt than after the drink

on performance from the effects of mood. Indeed the pre-

No effect of habitual diet on

sence of both may compound effects.

Most studies in this area have considered mood to be a
dependent variable rather than a predictor of level of per-
after LF lunch

formance. However, mood can affect performance in its

consumption
performance

own right and can modulate the motivation to perform

well, affecting compensatory effort and reducing the effect
lunch

of other interventions on overall level of performance. The

effect of mood on performance is modulated by time of
day, consumption of foods and stimulants such as caffeine,
and can be influenced by mood-induction techniques. Exer-
motor speed (e.g. tapping task)
grammatical reasoning spatial

cise can also have positive effects on mood, which may

Assessment before and after
30 min after finishing lunch

Assessment 15 min following

interact with food consumed. Thus it is important to exam-

Assessment just before and
information (e.g. Manikin)

Proportion estimation task

ine changes in subjective state after nutritional intervention

perceptual input (e.g.
simultaneous pattern

and with a degree of statistical complexity relate these

eating of the snack
Arithmetic reasoning

changes to effects on cognitive function, to determine

whether mood change is a correlate or a consequence of
comparison)

the food ingested and the level of cognitive function

Eight tasks:

demonstrated.
Reading

lunch
no lunch + non-caloric snack (drink)

Effect of mood on satiety

no lunch + fruit flavoured yoghurt
lunch + non-caloric snack (drink)
lunch + fruit flavoured yoghurt

Most of the studies examining the relationship between

LF/HCHO (BK: 22/67 % en,

HF/LCHO (BK: 48/42 % en,

and lunches (310 MJ for F,
Two iso-caloric BK (172 MJ)

mood and eating behaviour tend to focus on the effects of

lunch (three-course meal
BK (126 MJ) followed by

from university refectory,

food intake on mood (e.g. sleep and arousal) rather than

lunch: 20/61 % energy)

the converse (i.e. the effects of mood on food intake). In

chosen by subjects)
lunch: 54/32 % en)

fact, most of the work concerns the effects of dietary carbo-

377 MJ for M):

Four conditions:

hydrate on serotonin, which tends to modulate mood

Two conditions:

(Wurtman & Wurtman, 1989). The effects of manipu-

no lunch

lations on the macronutrient content of foods and diets

have been discussed elsewhere (for a review see Benton
& Donohoe, 1999). Some subjective feelings and mood
changes reported post ingestion in studies of performance
can be related to the satiating and thermogenic effects of
(sixteen M and sixteen

fifteen HF consumers)

students (eighteen M,

particular macronutrients, e.g. feelings of coldness or full-

F; seventeen LF and
Thirty-two volunteers

Forty-eight university

ness which decrease and increase, respectively, after pro-

tein ingestion (Cunliffe et al. 1997; Fischer et al. 2001).
ium carbohydrate; MF, medium fat.
Experiment 2:

When examining the effects of mood on food, most studies

eight males

have been involved with how mood initiates and stimulates

thirty F)

food intake (e.g. negative mood induces consumption of

comfort foods). Therefore, there is evidence to suggest
that mood (in particular negative mood) may act as a pre-
cursor for food intake (i.e. initiation of a single eating epi-
& Swinney

sode) and possibly over-consumption of highly palatable

(1986a )
& Miles
(2000)

(1990)
Kanarek

foods (e.g. chocolate). However, the effects of mood on

et al.

Smith
Lluch

satiation and satiety are still undetermined, but it is likely

that mood interacts with palatability to influence satiety.
S206 L. Dye and J. Blundell
Palatability
The precise relationship between palatability (often
regarded as the hedonic dimension of food) and appetite
has been debated for some time now (Ramirez, 1990).
The original work of Hill and colleagues (Hill et al.
1984) was instrumental in raising the issue by demonstrat-
ing that subjective hunger was higher after a preferred,
fixed preload compared with a non-preferred, fixed pre-
load. Therefore, it has been proposed that palatability has
an important influence on the initiation and termination
of eating events (Blundell & Rogers, 1991b). The palat-
ability of a food is as a result of the integration of orosen-
sory and post-ingestive stimuli. Like hunger, palatability is
a hypothetical construct; that is, an explanatory concept
that cannot be observed directly itself, but is inferred
from operationally defined and measurable events (Blun-
dell & Rogers, 1991a,b). Therefore, the term palatability,
like hunger, can be conceived of in two ways: either as an
intervening variable not measurable directly or, when con-
strued as the perceived pleasantness of food, as a subjective
experience that can be monitored objectively by means of a
rating. Consequently, in reference to human appetite, palat-
ability appears to be considered to be equivalent to per-
ceived pleasantness of food. Nonetheless, the relationship
of pleasantness (or palatability) to actual consumption
remains to be determined. With reference to the mediating
processes involved in the satiety cascade, it is important to
know how these mechanisms are related to fluctuations in
perceived palatability and hunger. For example, is high
palatability a sufficient or necessary condition for the
initiation of eating? It is suggested that the answer to
both situations must be no, since there are occasions
when even highly palatable food will not be consumed
and unpalatable food will be consumed. Another important
issue here is what is the relationship between palatability
and the termination of eating (satiation)? That is, what is
the relationship between palatability and hunger? It is
known that mood has a strong influence on cognitive
performance; therefore, the relationship amongst mood,
satiety, palatability and performance is important.
Effect of palatability on satiety
Work on the relationship between palatability and satiety
has demonstrated that hunger was increased at the end of
a preferred meal compared with a non-preferred meal
(Hill et al. 1984). This phenomenon has recently been con-
firmed (Yeomans, 1996; Yeomans et al. 1999). Moreover,
with respect to the profile of hunger during eating, there
was a tendency for hunger to increase during the early
stages of eating when the most palatable food was consumed
whereas fullness did not follow the same pattern. Therefore,
these data suggest that the manipulation of palatability leads
to stimulation of appetite, which in turn could increase food
intake through the process of satiation. What is less clear is
how palatability affects subsequent hunger and food intake
(i.e. satiety). From the data that exist there seems to be
some ambiguity about the effects of palatability on satiation
and satiety (Hill et al. 1984; Johnson & Vickers, 1992;
Rogers & Schutz, 1992; De Graaf et al. 1999).
Effect of palatability on performance
Whilst the relationship between palatability and food
intake has been examined in many studies, the relationship
between palatability and cognitive performance has not. In
fact, our search has not identified any studies that have
measured directly the effects of palatability per se on
cognitive performance. It is, however, logical that palat-
ability could influence cognitive performance via the med-
iating effects of mood (e.g. palatable food-induced increase
in endorphins). That is, if an increase in palatability causes
an increase in positive mood, then performance could also
be enhanced. Despite the lack of evidence on this topic,
some experiments intended to examine the effects of
foods (e.g. macronutrient content, energy value, energy den-
sity) on cognitive performance have inadvertently allowed
taste or palatability characteristics to vary.
Habitual diet
The composition of the habitual diet could have important
consequences for the physiological and behavioural patterns
of the participant. For example, phenotypes of habitual diet
composition (low-fat v. high-fat) have recently been identi-
fied using data from a national database (Macdiarmid et al.
1996). It has been suggested that individuals habitually
consuming different diets vary in certain physiological
and behavioural characteristics (Blundell & Cooling, 2000).
Therefore, habitual diet may have an important influence
on the effect of food on satiety, cognitive performance
and mood as well as sleep (LeNoury et al. 2000).
Habitual diet and satiety
In line with the habitual diet composition, the familiarity
with the preload could also have an important influence
on the outcome. The effects of the interventions in the
macronutrient (and energy) content of the food/diet on appe-
tite control and performance may be dependent on the habit-
ual diet of the participant. For example, a high-fat preload
may have a different effect on a habitual, high-fat consu-
mer compared with a habitual low-fat consumer (e.g. per-
ceptions of palatability). Tournier & Louis-Sylvestre
(1991) have hypothesised that a learned adjustment to the
calories in liquid food would occur with repeated exposure.
One study demonstrated that informed and uninformed
subjects ate similar weights of low- and high-energy
sweet preloads (Rolls et al. 1989). This suggests that,
when ingesting familiar foods, individuals may rely on pre-
vious experiences with these foods to determine how much
they should eat (Booth et al. 1982). Clearly participants
would not be able to do this with unfamiliar foods.
Habitual diet and performance
Habitual diet does not seem to influence performance in
short-term studies where nutritional interventions are
made (Lluch et al. 2000). However, performance seems
to be optimal when the food consumed most resembles
the usual diet, e.g. medium-fat, medium-carbohydrate
(Lloyd et al. 1994, 1996). Naturalistic studies do, however,
 Psychological and behavioural functions
suggest that the habitual diet, which influences parameters
such as cholesterol level, may influence levels of perform-
ance. Low plasma cholesterol was associated with slower
movement and decision time on a choice reaction time
task (Benton, 1995; Muldoon et al. 1997). However,
this linear relationship was found only in females; males
exhibited a non-linear relationship, suggesting that more
than one mechanism may be involved. The interaction of
habitual diet and cognitive performance in man requires
further investigation. There may be an interaction between
cholesterol, dieting and cognitive function because dieting
lowers cholesterol.
Evidence from animal studies suggests that there may
be differences between the chronic and acute impact of
diets varying in fat content. Studies in animals suggest
that the structure of the brain may be modified over time
by dietary fatty acids (Wainwright et al. 1994). A high-
fat diet generally disrupts cognitive function. In addition,
an enriched environment reduced the learning impairment
associated with frontal lobe function whereas glucose
enhanced memory performance by exerting beneficial
effects on hippocampus-related memory function (Green-
wood & Winocur, 1997).
Some physical illnesses are associated with impairment
of cognitive function and dietary intake is a risk factor for
the development of these conditions, e.g. atherosclerosis,
type 2 diabetes and hypertension. Obesity is a significant
risk factor in the development of all these conditions.
Rogers (2001) has reviewed the long-term impact of
diet on mood and cognitive function in relation to these
conditions. Antioxidants, vitamins and other nutritional
supplements have been reported to prevent or delay age-
related cognitive impairment (Kalmijn et al. 1997;
Lethem & Orrell, 1997). Epidemiological studies support
these findings (e.g. Gale et al. 1996) although some
controlled studies have shown little or no effect of supple-
mentation (Smith et al. 1999, 2000; Cockle et al. 2000).
Summary
This domain of research possesses great potential for
collaborations between academia and industry for the
development of such products with functional properties.
The current world-wide epidemic of obesity has generated
an urgent need for the development of functional foods
with the capacity to control appetite and influence the regu-
lation of body weight. Some foods already exist and others
are under development. There is a clear market need for
such products. At the same time, the requirement of
foods to produce vitality and influence dynamic life-
styles creates a drive to produce foods with the capacity
to empower people to cope well with an ever more
challenging technological environment.
The physiological system underlying the function of
satiety is better understood than that responsible for
competitive performance. Key sites in the brain and
the periphery have been identified, and both genetic
and pharmaceutical tools have contributed to understand-
ing the network. Consequently, some clear targets (and
processes) exist to guide the development of functional
S207
foods for satiety. In the field of cognitions, rather less is
known about the detail of the neural network although
developments in pharmaceuticals have identified certain
neurochemical systems and receptors. However, it is
likely to be more difficult to develop foods that can exert
specific and selective actions at such critically localised
targets. Moreover, because of the importance of maintain-
ing an adequate level of cognitive performance, people can
protect themselves from degradation in performance by
increasing effort. Consequently, measurement of both
applied effort and the performance endpoint are required
to assess the impact of nutritional manipulations.
A corollary of this is that, because of the requirement to
maintain a continuous high level of cognitive performance,
it is rather difficult for foods to enhance performance
even further (the ceiling effect). Therefore, an appropriate
strategy for foods may be to develop products that can pro-
tect people from an enforced performance decrement due
to fatigue or stress.
Another strategic issue concerns the unavoidable simul-
taneous effect of foods on different domains of psychologi-
cal behavioural functioning. In the field of pharmaceuticals
it is well known that drugs developed for the treatment of
one psychological condition for example, anxiety
can have collateral effects on appetite and satiety. The
capacity of some anxiolytic and neuroleptic drugs to
increase eating and body weight may contribute to a
form of iatrogenic-induced obesity. In the field of func-
tional foods it is therefore important to develop foods
that enhance satiety without exerting detrimental effects
on mood or cognitive performance; and to create more pro-
ducts for sustaining cognitive performance that do not
compromise satiety.
A further challenge in this area is also to go beyond
unitary cognitive functions to complete integrated perform-
ance relevant to modern life-styles such as vehicle driving,
technological monitoring and operating industrial equip-
ment. Only a very few studies (e.g. Moser et al. 1983)
have examined the effects of food materials on a compli-
cated skill such as car driving. This type of cognitive
task clearly requires the interpretation of a number of
cognitive functions such as reaction time, perceptual
ability, information processing, tracking ability and
memory.
The identification of biomarkers in the field of
psychological and behavioural functions requires more
thought and careful consideration. A biomarker can be
conceptualised as an indicator of normal biological pro-
cesses, a parameter associated with the psychological/
behavioural response or a surrogate endpoint. In other
areas of physiological functioning relevant to functional
foods, a convenient biomarker could be a measured
variable in blood or some other tissue that correlated
with a functional endpoint or disease state. A good
example could be a measure of intestinal flora for
bowel disease. However, in this particular field it may
be more difficult to measure some underlying physio-
logical or neural variable than it is to measure the func-
tional endpoint itself. For both satiety and cognitive
performance we have suggested how the issue of bio-
makers could be approached.
S208 L. Dye and J. Blundell
Acknowledgements
The authors thank Orkla Foods A.S. for support with the
scientific material for this review, and Susan Alexander
for assistance with the manuscript.
References
Astrup A & Flatt JP (1996) Metabolic determinants of body
weight regulation. In Regulation of Body Weight: Biological
and Behavioral Mechanisms, pp. 193 210 [C Bouchard and
GA Bray, editors]. Chichester: John Wiley.
Bellisle F, Blundell JE, Dye L, Fantino M, Fern E, Fletcher RJ,
Lambert J, Roberfroid M, Specter S, Westenhofer J & Wester-
terp-Plantenga MS (1998) Functional food science and beha-
viour and psychological functions. British Journal of
Nutrition 80, S173 S193.
Benton D (1995) Do low cholesterol levels slow mental pro-
cessing? Psychosomatic Medicine 57, 50 56.
Benton D & Donohoe RT (1999) The effects of nutrients on
mood. Public Health Nutrition 2, 403 409.
Benton D, Owens D & Parker P (1994) Blood glucose memory
and attention. Neuropsychologia 115, 129.
Blundell JE (1991) Pharmacological approaches to appetite
suppression. Trends in Pharmacological Sciences 12, 147 157.
Blundell JE, Burley VJ & Peters JC (1991) Dietary fat and human
appetite: effects of non-absorbable fat on energy and nutrient
intakes. In Obesity: Dietary Factors and Control, pp. 3 13
[DR Romsos, editor]. Tokyo: Japan Science Society Press.
Blundell JE & Cooling J (2000) Routes to obesity: phenotypes, food
choices and activity. British Journal of Nutrition 83, S33S38.
Blundell JE, Cotton JR, Delargy H, Green S, Greenough A, King
NA & Lawton CL (1995) The fat paradox: fat-induced satiety
signals versus high fat overconsumption. International Journal
of Obesity 19, 832 835.
Blundell JE & Hill AJ (1986) Influence of tryptophan on appetite
and food selection in man. In Amino Acids in Health and
Disease: New Perspectives, pp. 1 14 [S Kaufman, J Fernstrom,
R Roth and S Woo, editors]. UCLA Symposia on Molecular
and Cellular Biology, New Series vol. 55. New York: Alan R
Liss, Inc.
Blundell JE & Hill AJ (1987) Nutrition, serotonin and appetite:
case study in the evolution of a scientific idea. Appetite 8,
183 194.
Blundell JE, Lawton CL, Cotton JR & Macdiarmid JI (1996) Con-
trol of human appetite: implications for the intake of dietary
fat. Annual Review of Nutrition 16, 285 319.
Blundell JE & Macdiarmid JI (1997) Passive overconsumption fat
intake and short-term energy balance. Annals of the New York
Academy of Science 827, 392 407.
Blundell JE & Naslund E (1999) Glucagon-like peptide-1, satiety
and appetite control (invited commentary). British Journal of
Nutrition 81, 259 260.
Blundell JE & Rogers PJ (1991a) Satiating power of food. In
Encyclopedia of Human Biology, vol. 6, pp. 723 733. New
York: Academic Press, Inc.
Blundell JE & Rogers PJ (1991b) Hunger, hedonics, and the con-
trol of satiation and satiety. In Chemical Senses, pp. 127 148
[M Friedman, MG Tordoff and M Kare, editors]. Appetite
and Nutrition, vol. 4. New York: Marcel Dekker, Inc.
Blundell JE, Rogers PJ & Hill AJ (1988) Uncoupling sweetness
and calories: methodological aspects of laboratory studies on
appetite control. Appetite 11, 54 61.
Booth DA (1972) Postabsorptively induced suppression of appe-
tite and the energostatic control of feeding. Physiology and
Behavior 9, 199 202.
Booth DA, Chase A & Campbell AT (1970) Relative effective-
ness of protein in the late stages of appetite suppression in
man. Physiology and Behavior 5, 1299 1302.
Booth DA, Mather P & Fuller J (1982) Starch content of ordinary
foods associatively conditions human appetite and satiation
indexed by intake and eating pleasantness of starch-paired
flavours. Appetite 3, 163 184.
Burns AA, Livingstone MBE, Welch RW, Dunne A, Reid CA &
Rowland IR (2001) The effects of yoghurt containing a novel
fat emulsion on energy and macronutrient intakes in non-over-
weight, overweight and obese subjects. International Journal of
Obesity 25, 1487 1496.
Burns AA, Livingstone MBE, Welch RW, Dunne A, Robson PJ,
Lindmark L, Reid CA, Mullaney U & Rowland IR (2000)
Short-term effects of yoghurt containing a novel fat emulsion
on energy and macronutrient intakes in non-obese subjects.
International Journal of Obesity and Related Metabolic
Disorders 24, 1410 1425.
Butler RN, Davies M, Gehling NJ & Grant AK (1981) The effect
of preloads of amino acid on short-term satiety. American
Journal of Clinical Nutrition 34, 2045 2047.
Campfield LA & Smith FJ (1986) Functional coupling between
transient declines in blood glucose and feeding behaviour:
temporal relationships. Brain Research Bulletin 17, 427 433.
Carlson NR (1991) Physiology of Behaviour, 5th ed., p. 411.
Boston, MA: Allyn & Bacon.
Cockle SM, Haller J, Kimber S, Dawe RA & Hindmarch I (2000)
The influence of multivitamins on cognitive function and mood
in the elderly. Aging & Mens Health 4, 339 353.
Cordain L, Brand Miller J, Boyd Eaton S, Mann N, Holt SHA &
Speth JD (2000) Plant animal subsistence ratios and macro-
nutrient energy estimations in worldwide hunter gatherer
diets. American Journal of Clinical Nutrition 71, 682 692.
Cotton JR, Weststrate JA & Blundell JE (1996) Replacement of
dietary fat with sucrose polyester: effects on energy intake
and appetite control in nonobese males. American Journal of
Clinical Nutrition 63, 891 896.
Crombie N & Kirk TR (1999) Prevention of weight gain and blood
cholesterol reduction after consumption of a high carbohydrate
food in men. International Journal of Obesity 23, Suppl. 5,
657.
Cunliffe A, Obeid OA & Powell-Tuck J (1997) Post-prandial
changes in measures of fatigue: effect of a mixed or a pure
carbohydrate or pure fat meal. European Journal of Clinical
Nutrition 51, 831 838.
Cunliffe A, Obeid OA & Powell-Tuck J (1998) A placebo con-
trolled investigation of the effects of tryptophan or placebo
on subjective and objective measures of fatigue. European
Journal of Clinical Nutrition 52, 425 430.
De Graaf C, de Long LS & Lambers AC (1999) Palatability
affects satiation but not satiety. Physiology and Behavior 66,
681 689.
De Graaf C, Hulshof T, Weststrate JA & Hautvast JG (1996)
Nonabsorbable fat (sucrose polyester) and the regulation of
energy intake and body weight. American Journal of Physiol-
ogy 270, R1386 R1393.
Deijen JB, Heemstra ML & Orlebeke JF (1989) Dietary effects on
mood and performance. Journal of Psychiatric Research 23,
275 283.
Deijen JB, Wientjes CJE, Vullinghs HFM, Cloin PA & Langefeld
JJ (1999) Tyrosine improves cognitive performance and
reduces blood pressure in cadets after one week of a combat
training course. Brain Research Bulletin 48, 203 209.
Dickie NH & Bender AE (1982) Breakfast and performance in
schoolchildren. British Journal of Nutrition 48, 483 496.
Fischer K, Colombani PC, Langhans W & Wenk C (2001) Cog-
nitive performance and its relation to postprandial metabolic
 Psychological and behavioural functions
changes after different macronutrient ingestion in the morning.
British Journal of Nutrition 85, 393 405.
Friedman MI (1991) Metabolic control of calorie intake. In
Chemical Senses, pp. 19 38 [MI Friedman, MG Tordoff and
MR Kare, editors]. Appetite and Nutrition, vol. 4. New York:
Marcel Dekker, Inc.
Fryer JH, Moore NS, Williams HH & Young CM (1955) A study
of the interrelationships of the energy-yielding nutrients, blood
glucose levels and subjective appetite in man. Journal of
Laboratory and Clinical Medicine 45, 684696.
Gale CR, Martyn CN & Cooper C (1996) Cognitive impairment
and mortality in a cohort of elderly people. British Medical
Journal 312, 608 611.
Green MW & Rogers PJ (1995) Impaired cognitive functioning
during spontaneous dieting. Psychological Medicine 25,
10031010.
Green MW, Rogers PJ, Elliman NA & Gatenby SJ (1994b)
Impairment of cognitive function associated with dieting and
high levels of dietary restraint. Physiology and Behavior 55,
447452.
Green SM, Burley VJ & Blundell JE (1994a) Effect of fat- and
sucrose-containing foods on the size of eating episodes and
energy intake in lean males: potential for causing over-
consumption. European Journal of Clinical Nutrition 48,
547555.
Green SM, Delargy HJ, Joanes D & Blundell JE (1997) A satiety
quotient: a formulation to assess the satiating effect of food.
Appetite 29, 291 304.
Greenough A, Cole G, Lewis J, Lockton A & Blundell J (1998)
Untangling the effects of hunger, anxiety, and nausea on
energy intake during intravenous cholecystokinin octapeptide
(CCK-8) infusion. Physiology and Behavior 65, 303 310.
Greenwood CE & Winocur G (1997) An investigation of glucose
effects on learning and memory impairments induced by high
fat diets and environmental factors. Brain and Cognition 35,
415418.
Hassmen P, Blomstrand E, Ekblom B & Newsholme EA (1994)
Branched-chain amino acid supplementation during 30-km
competitive run: mood and cognitive performance. Nutrition
10, 405410.
Heaton KW (1981) Dietary fibre and energy intake. In Regulators
of Intestinal Absorption in Obesity, Diabetes and Nutrition,
pp. 283 294 [P Berchtold, A Cairella, A Jacobelli and
V Silano, editors]. Rome: Societa Editriee Universo.
Hill AJ & Blundell JE (1986) Macronutrients and satiety: the
effects of a high-protein or high-carbohydrate meal on subjec-
tive motivation to eat and food preferences. Nutrition and
Behaviour 3, 133 144.
Hill AJ & Blundell JE (1988) Role of amino acids in appetite con-
trol in man. In Amino Acid Availability and Brain Function in
Health and Disease, pp. 239 248 [G Huether, editor]. Berlin:
Springer-Verlag.
Hill AJ & Blundell JE (1990) Sensitivity of the appetite con-
trol system in obese subjects to nutritional and serotonin-
ergic challenges. International Journal of Obesity 14,
219233.
Hill AJ, Magson LD & Blundell JE (1984) Hunger and palatabil-
ity: tracking ratings of subjective experience before, during and
after the consumption of preferred and less preferred food.
Appetite 5, 361 371.
Hill JO, Seagle HM, Johnson SL, Smith S, Reed GW, Tran ZV,
Cooper D, Stone M & Peters JC (1998) Effects of 14 d of
covert substitution of Olestra for conventional fat on spont-
aneous food intake. American Journal of Clinical Nutrition
67, 11781185.
Hockey GRJ (1997) Compensatory control in the regulation
of human performance under stress and high workload: a
S209
cognitive energetical framework. Biological Psychology 45,
73 93.
Holt SHA, Brand Miller JC, Perocz P & Farmakalidis E (1995) A
satiety index of common foods. European Journal of Clinical
Nutrition 49, 675690.
Holt SHA, Delargy HJ, Lawton CL & Blundell JE (1999) The
effects of high-carbohydrate vs high-fat breakfasts on feelings
of fullness and alertness, and subsequent food intake. Inter-
national Journal of Food Science and Nutrition 50, 13 28.
Horrocks LA & Yeo YK (1999) Health benefits of docosahexa-
enoic acid (DHA). Pharmacological Research 40, 211 225.
Johnson J & Vickers Z (1992) Factors influencing sensory-
specific satiety. Appetite 19, 15 31.
Kalmijn S, Feskens EJM, Launer LJ & Kromhout D (1997) Poly-
unsaturated fatty acids, antioxidants, and cognitive function in
very old men. American Journal of Epidemiology 145, 33 41.
Kanarek RB & Swinney D (1990) Effects of food snacks on cog-
nitive performance in male college students. Appetite 14,
15 27.
Kaplan RJ, Greenwood CE, Winocur G & Wolever TMS (2000)
Cognitive performance is associated with glucose regulation in
healthy elderly persons and can be enhanced with glucose and
dietary carbohydrates. American Journal of Clinical Nutrition
72, 825 836.
Kaplan RJ, Greenwood CE, Winocur G & Wolever TMS (2001)
Dietary protein, carbohydrate, and fat enhance memory per-
formance in the healthy elderly. American Journal of Clinical
Nutrition 74, 687693.
Kelly TH, Foltin RW, Rolls BJ & Fischman MW (1994) Effect of
meal macronutrient and energy content on human performance.
Appetite 23, 97 111.
Kirk TR, Burkill S & Cursiter MC (1997) Dietary fat reduction
achieved by increasing consumption of a starchy food an
intervention study. European Journal of Clinical Nutrition
51, 455 461.
Kissileff HR (1984) Satiating efficiency and a strategy for con-
ducting food loading experiments. Neurological Biobehavioral
Reviews 8, 129135.
Klassen T, Riedel WJ, Deutz NEP, van Someren A & van Praag
HM (1999) Specificity of the tryptophan depletion method.
Psychopharmacology 141, 279 286.
Lang V, Bellisle F, Oppert J-M, Craplet C, Bornet FRJ, Slama G
& Guy-Grand B (1998) Satiating effect of proteins in healthy
subjects: a comparison of egg albumin, casein, gelatin, soy
protein, pea protein, and wheat glutin. American Journal of
Clinical Nutrition 67, 1107 1204.
Lawton CL, Burley VJ, Wales JK & Blundell JE (1993) Dietary
fat and appetite control in obese subjects: weak effects on
satiation and satiety. International Journal of Obesity 17,
409 416.
Lawton CL, Delargy HJ, Brockman J, Smith FC & Blundell JE
(2000) The degree of saturation of fatty acids influences post-
ingestive satiety. British Journal of Nutrition 83, 473 482.
Lawton CL, Delargy HJ, Smith FC, Hamilton V & Blundell JE
(1998) A medium-term intervention study on the impact of
high- and low-fat snacks varying in sweetness and fat content:
large shifts in daily fat intake but good compensation for daily
energy intake. British Journal of Nutrition 80, 149 161.
LeNoury JC, Cooling J & Blundell JE (2000) Different profiles
of appetite and sleeping in high-fat and low-fat (behavioural)
phenotypes. Appetite 35, 202.
Lethem R & Orrell M (1997) Antioxidants and dementia. Lancet
349, 1189 1190.
Lieberman HR, Corkin S, Spring BJ, Gowden JH & Wurtman RJ
(1983) Mood, performance and pain sensitivity: changes
induced by food constituents. Journal of Psychiatric Research
17, 135 145.
S210 L. Dye and J. Blundell
Lloyd HM, Green MW & Rogers PJ (1994) Mood and cognitive
performance effects of isocaloric lunches differing in fat and
carbohydrate content. Physiology and Behavior 56, 51 57.
Lloyd HM, Rogers PJ, Hedderley DI & Walker AF (1996)
Acute effects on mood and cognitive performance of breakfasts
differing in fat and carbohydrate content. Appetite 27, 151164.
Lluch A, Dye L & Blundell JE (2000) Effects of nutrient manipu-
lations on motivation to eat and performance in high fat and
low fat phenotypes. Appetite 35, 202.
Macdiarmid JI, Cade JE & Blundell JE (1996) High and low fat
consumers, their macronutrient intake and body mass index:
further analysis of the National Diet and Nutrition Survey of
British Adults. European Journal of Clinical Nutrition 50,
505 512.
Macdiarmid JI & Hetherington MM (1995) Mood modulation by
food: an exploration of affect and cravings in chocolate
addicts. British Journal of Clinical Psychiatry 34, 129 138.
Markus CR (1999) Relations between stress, food and mood: a
role for brain serotonin. PhD Thesis, University of Utrecht.
Markus CR, Panhuysen G, Tuiten A, Koppeschaar H, Fekkes D &
Peters ML (1998) Does carbohydrate-rich, protein-poor food
prevent a deterioration of mood and cognitive performance of
stress-prone subjects when subjected to a stressful task. Appe-
tite 31, 49 65.
Martin PY & Benton D (1999) The influence of a glucose drink
on a demanding working memory task. Physiology and
Behavior 67, 69 74.
Mayer J (1953) Gluostatic mechanism of the regulation of food
intake. New England Journal of Medicine 249, 13 16.
Mayer J (1955) Regulation of energy intake and body weight.
Annals of the New York Academy of Science 63, 15 43.
Mei N (1985) Intestinal chemosensitivity. Physiology Reviews 65,
211 237.
Mellinkoff SM, Frankland M, Boyle D & Greipel M (1956)
Relationship between serum amino acid concentration and fluc-
tuations in appetite. Journal of Applied Physiology 8, 535 538.
Miller DL, Catellanos VH, Shide DJ, Peters JC & Rolls BJ (1998)
Effects of fat-free potato chips with and without nutrition labels
on fat and energy intakes. American Journal of Clinical Nutri-
tion 68, 282 290.
Milton K (2000) Hunter gatherer diets a different perspective.
American Journal of Clinical Nutrition 71, 665 667.
Moser L, Plum H & Buckmann M (1983) Der Einflu von
Dextrose auf die psychophysische Leistungsfahigkeit des Auto-
fahrers. Aktuelle Ernahrungsmedizin 7, 7 14.
Muldoon MF, Ryan CM, Matthews KA & Manuck SB (1997)
Serum cholesterol and intellectual performance. Psychosomatic
Medicine 59, 382 387.
Novin D, Vanderweele DA & Rezek M (1973) Hepatic-portal
2-deoxy-D -glucose infusion causes eating: evidence for periph-
eral glucoreceptors. Science 181, 858 860.
Oomura Y (1988) Chemical and neuronal control of feeding
motivation. Physiology and Behavior 44, 555 560.
Owens DS & Benton D (1994) The impact of raising blood glu-
cose on reaction times. Neuropsychobiology 30, 106 113.
Parker PY & Benton D (1995) Blood-glucose levels selectively
influence memory for word lists dichotically presented to the
right ear. Neuropsychologia 33, 843 854.
Ramirez I (1990) What do we mean when we say palatable
food? Appetite 14, 159 161.
Reid M & Hammersley R (1995) Effects of carbohydrate intake
on subsequent food intake and mood state. Physiology and
Behavior 58, 421 427.
Riedel WJ, Klassen T, Deutz NEP, van Someren A & van Praag
HM (1999) Tryptophan depletion in normal volunteers pro-
duces selective impairment in memory consolidation. Psycho-
pharmacology 141, 362 369.
Roethlisberger FJ (1997) The Elusive Phenomena: An Autobio-
graphical Account of My Work in the Field of Organized Beha-
vior at the Harvard Business School, p. 47. Cambridge, MA:
Division of Research, Graduate School of Business Adminis-
tration, Harvard University Press.
Rogers PJ (2001) A healthy body, a healthy mind: long term
impact of diet on mood and cognitive function. Proceedings
of the Nutrition Society 60, 135 143.
Rogers PJ & Blundell JE (1989) Separating the actions of
sweetness and calories: effects of saccharin and carbohydrates
on hunger and food intake in human subjects. Physiology and
Behavior 45, 1093 1099.
Rogers PJ, Burley VJ, Alikhanizadeh LA & Blundell JE (1995)
Postingestive inhibition of food intake by aspartame: import-
ance of interval between aspartame administration and sub-
sequent eating. Physiology and Behavior 57, 489 493.
Rogers PJ & Green MW (1993) Dieting, dietary restraint and cog-
nitive performance. British Journal of Clinical Psychiatry 32,
113 116.
Rogers PJ, Keedwell P & Blundell JE (1991) Further analysis
of the short-term inhibition of food intake in humans by the
dipeptide L -aspartyl-L -phenylalanine methyl ester (aspartame).
Physiology and Behavior 49, 739 743.
Rogers PJ, Pleming HC & Blundell JE (1990) Aspartame ingested
without tasting inhibits hunger and food intake. Physiology and
Behavior 47, 1239 1243.
Rogers PJ & Schutz HG (1992) Influence of palatability on sub-
sequent hunger and food intake: a retrospective replication.
Appetite 19, 155 156.
Rolls BJ, Jacobs Laster L & Summerfelt A (1989) Hunger and
food intake following consumption of low-calorie foods.
Appetite 13, 115 127.
Rolls BJ, Kim-Harris S, Fischman MW, Foltin RW, Moran TH &
Stoner SA (1988) Food intake in dieters and nondieters after a
liquid meal containing medium-chain triglycerides. American
Journal of Clinical Nutrition 48, 66 71.
Rolls BJ, Kim-Harris S, Fischman MW, Foltin RW, Moran TH &
Stoner SA (1994) Satiety after preloads with different amounts
of fat and carbohydrate: implication for obesity. American
Journal of Clinical Nutrition 60, 476 487.
Smith AP, Clark R & Gallagher J (1999) Breakfast cereal and
caffeinated coffee: effects on working memory, attention,
mood, and cardiovascular function. Physiology and Behavior
67, 9 17.
Smith AP, Clark RE, Nutt DJ, Haller JG, Hayward S & Perry K
(2000) Vitamin C, mood and cognitive functioning in the
elderly. Nutrition and Neuroscience 2, 249 256.
Smith AP & Kendrick AM (1992) Meals and performance. In
Health and Performance, pp. 2 23 [AP Smith and DM
Jones, editors]. Handbook of Human Performance, vol. 2.
London: Academic Press.
Smith AP, Kendrick A, Maben A & Salmon J (1994a) Effects of
breakfast and caffeine on cognitive performance, mood and
cardiovascular functioning. Appetite 22, 39 55.
Smith AP, Kendrick A, Maben A & Salmon J (1994b) Effects of
fat content, weight, and acceptability of the meal on postlunch
changes in mood, performance, and cardiovascular function.
Physiology and Behavior 55, 417 422.
Smith AP, Leekham S, Ralph A & McNeill G (1988) The influ-
ence of meal composition on post-lunch performance efficiency
and mood. Appetite 10, 195203.
Smith AP, Maben A & Brockman P (1994c) Effects of evening
meals and caffeine on cognitive performance, mood and cardio-
vascular functioning. Appetite 22, 7 65.
Smith AP & Miles C (1986a) Acute effects of meals, noise
and nightwork. British Journal of Psychiatry 77,
377 387.
 Psychological and behavioural functions
Smith AP & Miles C (1986b) Effects of lunch on selective and
sustained attention. Neuropsychobiology 16, 117 120.
Smith AP & Miles C (1986c) The effects of lunch on cognitive
vigilance tasks. Ergonomics 29, 1251 1261.
Smith AP & Rich N (1998) Effects of consumption of snacks on
simulated driving. Perceptual and Motor Skills 87, 817 818.
Spring B, Maller O, Wurtman J, Digman L & Gozolino L (1983)
Effects of protein and carbohydrate meals on mood and per-
formance: interactions with sex and age. Journal of Psychiatric
Research 17, 155167.
Strachan MWJ, Deary IJ, Ewing FME & Frier BM (1997) Is type
II diabetes associated with an increased risk of cognitive dys-
function? A critical review of published studies. Diabetes
Care 20, 438 445.
Stubbs R (1996) Macronutrients effects on appetite. International
Journal of Obesity 19, S11 S19.
Stubbs RJ & Harbron CG (1995) Isoenergetic substitution of
MCT for LCT: effect on energy intake in ad libitum feeding
men. International Journal of Obesity 19, Suppl. 2, 28.
Teff KL, Young SN & Blundell JE (1989) The effect of protein or
carbohydrate breakfasts on subsequent plasma amino acid
levels, satiety and nutrient selection in normal males. Pharma-
cology, Biochemistry and Behavior 34, 829 837.
Thompson DA & Campbell RG (1977) Hunger in humans induced
by 2-deoxy-D -glucose: glucoprivic control of taste preference
and food intake. Science 198, 10651068.
Tournier A & Louis-Sylvestre J (1991) Effect of the physical state
of food on subsequent intake in human subjects. Appetite 16,
17 24.
Trigazis L, Orttmann A & Anderson GH (1997) Effect of a chole-
cystokinin-A receptor blocker on protein-induced food intake
suppression in rats. American Journal of Physiology 272,
R1826 R1833.
Van Itallie TB (1990) The gluostatic theory 19531988: roots and
branches. International Journal of Obesity 14, Suppl. 3, 110.
S211
Wainwright PE, Huang Y-S, Bulman-Fleming B, Levesque S &
McCutcheon D (1994) The effects of dietary fatty acid compo-
sition combined with environmental enrichment on brain and
behavior in mice. Behavioural Brain Research 60, 125 136.
Welch I, Saunders K & Read NW (1985) Effect of ileal and
intravenous infusions of fat emulsions on feeding and satiety
in human volunteers. Gastro 89, 1293 1297.
Wells AS & Read NW (1996) Influences of fat, energy and time
of day on mood and performance. Physiology and Behavior 59,
1069 1076.
Wells AS, Read NW & Craig A (1995) Influences of dietary and
intraduodenal lipid on alertness, mood and sustained concen-
tration. British Journal of Nutrition 74, 115 123.
Westerterp-Plantenga MS, Rolland V, Wilson SAJ & Westerterp
KR (1999) Satiety related to 24 h diet-induced thermogenesis
during high protein/carbohydrate vs high fat diets measured
in a respiration chamber. European Journal of Clinical Medi-
cine 53, 495 502.
Wing RR, Vazquez JA & Ryan CM (1995) Cognitive effects
of ketogenic weight reducing diets. International Journal of
Obesity 19, 811 816.
Wurtman JJ, Brzezinski A, Wurtman RJ & Laferrere B (1989)
Effect of nutrient intake on premenstrual depression. American
Journal of Obstetrics and Gynecology 161, 1228 1234.
Wurtman RJ (1982) Nutrients that modify brain function. Scien-
tific American 243, 42 51.
Wurtman RJ, Hefti F & Melamed E (1981) Precursor control
of neurotransmitter synthesis. Pharmacological Reviews 32,
315 335.
Wurtman RJ & Wurtman JJ (1989) Carbohydrates and
depression. Scientific American 260, 50 57.
Yeomans MR (1996) Palatability and the microstructure of eating
in humans: the appetiser effect. Appetite 27, 119 133.
Yeomans MR, Hails NJ & Nesic JS (1999) Alcohol and the
appetiser effect. Behavioral Pharmacology 10, 151161.
S178 F. Brouns et al.
Using such techniques it was observed that the rates of
gastric emptying, digestion and absorption are of crucial
importance with respect to digestive tolerance during exer-
cise. These factors were also found to determine the rate of
nutrient/substrate supply to the active muscles and the sub-
sequent conversion/oxidation for the generation of ATP.
This has led to a large number of studies on the effect of
various food applications on gastrointestinal function, toler-
ance and performance. Specific foods were shown to be
functional in terms of benefits mentioned above. More
recently the field of interest has shifted from macronutrients
and fluids to isolated nutritional or non-nutritional food
components (FC). Some examples are caffeine, creatine,
ribose, L -carnitine, certain amino acids, antioxidants, lactate,
pyruvate, glycerol, sodium bicarbonate and hydroxycitric
acid (Maughan, 1999; Spriet, 1999; Wagenmakers, 1999a;
Volek, 2000; Brouns, 2002). The present short review dis-
cusses some of the most favourable claims as well as
aspects of their validation.
Performance-limiting factors: a target for
supplementation
There are a number of sectors in sports nutrition that are
of great interest to the food industry. Basically these sec-
tors are related to performance-limiting factors in which
nutrition is thought to play a role. Research on the
effect of selected FC has focused on how to improve or
minimize the impact of these limitations. A number of
examples are listed below. The nutrients listed have
been used in studies and are given as examples only.
The outcome of these studies may have been positive
(P), negative (N), led to mixed results (M) or is still
hypothetical (H).
1. Muscle mass ) protein synthesis-stimulating FC
(e.g. branched-chain amino acids in combination
with a carbohydrate (P), arginine (N), creatine (H) or
b-hydroxy-b-methylbutyrate (M)).
2. Fat mass ) FC that induce fat loss, improving fat-
free mass (e.g. hydroxycitric acid (N), caffeine (M),
L -carnitine (N), chromium picolinate (N), chitosan
(H), pyruvate (N) and L -tyrosine (N)).
3. Bone and joint conditions ) FC to improve bone
mass, cartilage thickness, synovial fluid condition
(e.g. specific amino acids such as proline (H), lysine
(H), mineral combinations (H), phyto-oestrogens (H),
glucosamine (H), vitamin K (H) and cartilage prep-
arations (H)).
4. Dehydration ) drinks composed for rapid rehydra-
tion and fluid retention (e.g. hypotonic or isotonic
drink formulation containing 30 70 g carbohydrate/
litre plus 20 30 mmol Na/litre (P)).
5. Glycogen depletion ) carbohydrate types to maxi-
mize muscle and liver glycogen resynthesis (e.g. glu-
cose and glucose polymers for muscle (P), fructose
for liver (P), carbohydrate in combination with
amino acids (P)).
6. Carbohydrate availability and oxidation ) FC such
as selected carbohydrate types that are expected to
be digested rapidly and absorbed completely in
favour of complete oxidation (excellent energy
source) during exercise (P).
7. Low fat oxidation rate ) FC that influence the rate
of lipolysis, and improve fat uptake in muscle and
mitochondria and ultimately fatty acid oxidation rate
(e.g. caffeine (M), hydroxycitric acid (N), L -carnitine
(N) and oil containing medium-chain triacylglycerols
(N)).
8. Adenine nucleotides depletion ) FC that are thought
to enhance resynthesis as well as storage of ATP and
creatine phosphate (e.g. ribose (N) and creatine mono-
hydrate (P)).
9. Immunosuppression ) supply of FC that increase
resistance to disease and reduce exercise-induced
inflammations (e.g. specific vitamins such as vitamins
C and E (H), minerals such as Zn (H), trace elements
such as Se (H), colostrum extracts (H), glutamine (H),
achinacea (H), polyunsaturated fatty acids (H) and
probiotics (H)).
10. Neurostimulation ) supply of neuromodulating
FC that are thought to improve reaction times,
shorten neuromuscular impulse transmission, improve
the availability of precursors for hormones and
brain peptides, and improve cognitive function in
stress conditions (e.g. caffeine (P), specific amino
acids such as tyrosine (P), g-aminobutyric acid
(H) and tryptophan (H), branched-chain amino
acids (H), choline (H) and phosphatidylcholine
(lecithin; H)).
11. Suppressed hormone secretion ) FC thought to
enhance release of hormones or improve sensitivity
to hormones that are involved in protein synthesis,
substrate metabolism and recovery from exercise
(e.g. tryptophan (H, M), branched-chain amino acids
(H, M), tyrosine (H, M), arginine (H, M), ornithine
(H, M) and g-aminobutyric acid (H, M)).
12. Blood flow ) FC that are thought to enhance
blood flow by vasoactive effects (e.g. arginine (M)
and L -carnitine (H)).
13. Gastrointestinal distress ) supply of foods and
drinks that are optimally tolerated by the digestive
system during exercise (e.g. hypotonic food formulas
composed of rapidly digestible and completely absorb-
able nutrients (P)).
14. Poor nutritional status ) FC to ensure optimal nutri-
ent status of vitamins, trace elements and minerals
(e.g. specific micronutrient supplements (P)).
15. Poor performance ) FC that may increase endurance
or maximal oxygen uptake and thereby improve per-
formance (e.g. carbohydrate (P), coenzyme Q10 (N)
and branched-chain amino acids (N)).
16. Muscle damage ) FC that may reduce the occur-
rence of muscle damage during exercise and improve
recovery from it (e.g. antioxidants like vitamin E
(M) and b-carotene (M)).
17. Muscle cramps ) FC to support neuromuscular
events and reduce indices of cramp (e.g. Mg (H), Zn
(H) and ribose (H)).
18. Injuries ) FC to prevent the development of injuries
and speed up the recovery process from injury (e.g.
glucosamine (H) and chondroitin (H)).
 Food supplements for athletes
Examples of claims that are made for products whose
target is to improve performance-limiting factors are more
power, improved performance, more muscle mass, less
body fat, faster recovery, quick rehydration, reduces
muscle cramps and improved hormone release. Clearly,
the food and supplement industry tries to make such
claims on the product packaging as well as in advertising by
promoting the product as a solution with a proven benefit.
The ultimate goal is to give athletes the feeling that consump-
tion of the product will improve performance, or result in a
benefit that is of importance to health and well-being.
The basic question in this respect is: What is the evi-
dence? There are valid examples of nutritional interven-
tions in exercise and sport that indeed resulted in
significant effects on the study endpoint of interest. How-
ever, there are also a very large number of products avail-
able on the market promoted with benefit statements that
have never been validated by sound scientific studies.
Below, a short discussion is given on some of the
above-mentioned examples. Test methods and biomarkers
thought to be suitable for the measurement of interventions
in which any of such nutritional factors are tested against a
suitable placebo are discussed briefly. References are given
for the reader who is interested in detailed information.
Muscle/fat mass
The importance of a relatively high muscle mass and
low fat mass for most power and strength events is based
on the observation of a significant relationship between
muscle cross-section and maximal muscle strength
measured either as maximal strength or dynamic strength.
The relative strength or power of an athlete declines with
increasing fat mass. As such, all athletes who move/replace
their own body weight will benefit from having a high lean
body mass/low fat mass. Therefore, athletes involved
in strength and power events will benefit from training
regimens and nutritional factors that boost muscle protein
synthesis and reduce fat mass (Lemon, 1993). Athletes
involved in sports in which low body fat is a key factor,
such as bodybuilding, will do almost anything to reduce
fat mass and increase visible muscle mass (Kleiner et al.
1990; Walberg-Rankin et al. 1993).
Methods used to measure muscle mass/fat mass and
muscle protein synthesis concern the following biomarkers.
1. Selected muscle mass by muscle circumference
measurements (surrogate marker, not reliable due to
a number of confounding factors, e.g. level of sub-
cutaneous fat, muscle glycogen and water content).
2. Site-specific fat mass by circumference measurements
(as above, not reliable).
3. Selected muscle mass or site-specific fat mass by scan-
ning techniques such as computed tomography (CT)
and magnetic resonance imaging (MRI; reliable but
complex and expensive, CT is invasive, MRI is non-
invasive; Gadian, 1995).
4. Total muscle mass by body composition techniques.
The reliability depends strongly on the method of
choice. So-called mechanistic methods, such as under-
water weighing, isotope dilution or both methods
S179
combined, are the most reliable (non-invasive, although
underwater weighing may be problematic in some
groups, is time-consuming, requires special laboratory
facilities and is relatively expensive). The descriptive
methods, such as dual X-ray absorptiometry (DXA),
bio-impedance analyses (BIA) and skin folds (SF),
which are derived from the mechanistic ones, are
less reliable with respect to the absolute values (non-
invasive although DXA uses a small doses of X-rays,
fast, DXA relatively expensive, BIA and SF are
cheap). Body composition changes resulting from the
intervention are also detected most accurately with
mechanistic methods. Finally, imaging techniques,
such as CT and MRI, are used increasingly (accurate
but interpretation of images may be complicated, CT
invasive, MRI non-invasive, special laboratory facilities
needed, often time-consuming, expensive). For more
comprehensive overviews of methods and systematics,
see Wang & Heymsfield (1995) and van Marken
Lichtenbelt & Fogelmholm (1999).
5. Protein synthesis, muscle protein degradation and turn-
over rates by measuring the incorporation rate and
quantity of selected markers such as stable isotope-
labelled amino acids: leucine, glycine, phenylalanine
(reliable but complex and expensive; Wagenmakers,
1999b; Rennie & Tipton, 2000).
Dehydration/hydration
The importance of appropriate hydration and rehydration
stems from a number of observations. With continuous exer-
cise the water content of all body compartments will
decrease as a result of fluid loss by sweating and insensible
water loss from the lungs. Depending on the exercise inten-
sity, training status, climatic circumstances and body size,
sweat losses may range from a few hundred millilitres to
more than two litres per hour. Such large sweat losses will
reduce circulating body fluids. Because a normal plasma
volume is of prime importance in maintaining an appropri-
ate blood flow through exercising tissues, it may be
deduced that a significant decrease in plasma volume will
impair blood flow. This will in turn lead to a reduced trans-
port of substrates and oxygen, which are needed for energy
production, to the muscles. Also, the transport of metabolic
waste products and heat, from the muscle to eliminating
organs such as the liver and skin, will be impaired. The
decreased heat transfer from the muscles to the skin results
in an increased core temperature in all cases where heat pro-
duction exceeds heat dissipation from the body. This may
lead to a decreased energy-production capacity, fatigue,
impaired performance and health threats (Brouns, 1997).
Clearly, not only the total amount of body water (TBW) is
important, but also the water distribution. Therefore deter-
mination of intracellular water (ICW) and extracellular water
(ECW), or the ratio between these two, is often practised.
Methods used to measure the hydration status of the
whole body concern the following biomarkers.
1. TBW content by dilution of stable isotopes, deuterium
or the more expensive oxygen-18 (accurate, non-
invasive, time-consuming, special laboratory facilities
S180 F. Brouns et al.
needed, moderately expensive; Westerterp et al. 1995;
Westerterp, 1999).
2. ECW content by bromide dilution (accurate but
bromide dilution space is not identical to ECW and
a correction factor is needed, non-invasive, time-
consuming, moderately expensive; van Marken
Lichtenbelt et al. 1996).
3. TBW and ECW by bio-impedance spectroscopy
(accuracy is still a matter of controversy but non-
invasive, fast and cheap; van Marken Lichtenbelt,
2001).
Methods used to measure shifts in body fluids concern
the following biomarkers.
1. Changes in plasma volume by determination of
haemocrit and haemoglobin values (reliable and easy
to measure; Dill & Costill, 1974).
2. ECW:ICW with the dilution techniques mentioned
above (unsuitable for short-term interventions).
3. Bioelectrical impedance (see above; Boileau &
Horswill, 2000; van Marken Lichtenbelt, 2001). The
method may pose problems during the study of
short-term (sports) intervention, because of its sensi-
bility to changes in body temperature and electrolyte
concentrations in body fluids.
4. Muscle biopsy analysis (not reliable for rehydration
effects).
Methods used to determine fluid uptake rate by the body
concern techniques to measure biomarkers of gastric
emptying and intestinal absorption. It should be noted
that gastric emptying alone does not give a valid answer
with respect to rehydration efficacy since it is only the
first factor in a chain of determinants of bioavailability.
For example, water is emptied rapidly from the stomach
but is absorbed slowly in the gut. The best data are
obtained from a battery of tests in which gastric emptying,
gut segmental absorption and fluid retention measurements
are combined.
Gastric emptying may be measured by:
1. Intubation techniques applying the double sampling
technique of George (invasive, needs routine, reliable
to quantify during exercise, cheap; Beckers et al.
1988).
2. Use of stable isotopes (non-invasive, complex, can be
used during exercise but quantification not possible,
only qualitative differences can be shown, expensive;
van Nieuwenhoven et al. 1999).
3. Gamma scintillation (non-invasive, complex, isotopes
required, reliable but not suitable during exercise,
expensive; Leiper & Maughan, 1988; Beckers et al.
1992).
Intestinal absorption may be measured by:
1. Intubation/perfusion techniques using non-absorbable
markers (invasive, routine required, difficult during
exercise, reliable to quantify for the test segment but
not for the whole intestine, relatively cheap; Leiper
& Maughan, 1988).
2. Determining the appearance of labels in blood using
stable isotopes (non-invasive, but complex and diffi-
cult to quantify due to label shifts; Leiper & Maughan,
1988).
Glycogen depletion/muscle glycogen content
Since the classical Scandinavian studies using muscle
biopsy techniques, a lot of work has focused on the import-
ance of muscle glycogen for energy homeostasis and per-
formance. Several lines of evidence show that intense
and lasting muscle work cannot be performed without
appropriate availability of carbohydrate. As soon as
specific muscles or muscle fibres become glycogen-
depleted, they will be impaired in their ability to perform
repeated high-intensity contractions. Glycogen depletion,
caused either by exercise or a combination of exercise
and low carbohydrate intake, leads to a reduction in
work capacity to a level of about 50 % of the normal maxi-
mal working capacity. Alternatively, when the carbo-
hydrate stores in muscle and liver are increased by diet
manipulation, athletes are able to perform longer at high
exercise intensity. Thus, the availability of carbohydrate
and the size of the glycogen stores are important and limit-
ing factors for endurance performance.
Methods used to measure muscle and liver glycogen
concern the following biomarkers.
1. Muscle biopsy analysis for muscle glycogen (reliable
and easy but invasive, cheap; Hultman, 1967).
2. NMR for muscle glycogen (reliable for qualitative
changes, difficult for quantification, expensive, com-
plex but non-invasive; Price et al. 1999).
3. NMR for liver glycogen (as above; Price et al. 1999).
Hormone-releasing agents
There are observations that intensive training may induce
lower circulating levels of stress hormones, insulin and
glucagon. This is most probably caused by enhanced sensi-
tivity of tissues/cells for the hormones. It has also been
observed that intense training/over-training may lower
secretion of androgens and growth hormone, most probably
because of endocrine disruption effects. However, the
physiology of many hormones as well as the effects of
exercise on hormonal adaptations are still incompletely
understood (McMurray & Hackney, 2000) and make it
difficult to link secretion modifications of many of them,
if any, to health performance benefits. Nevertheless, a sub-
stantial number of products are promoted to enhance
hormone release for athletic benefits (Brouns, 2002). For
example, it has been hypothesized that the ingestion of
arginine and ornithine may stimulate the release of
human growth hormone, which is thought to stimulate
muscle growth.
Reliable measurement of blood hormone levels after
supplementation is complex. Because of diurnal effects,
24 h profiles will be required to make any statement on
the physiological relevance of observations.
Gastrointestinal tolerance
A substantial number of endurance athletes (40 60 %)
 Food supplements for athletes S181

may be prone to developing gastrointestinal disturbances Fat utilization and oxidation

during exercise, especially when exercising in the heat
An enhanced oxidation of fatty acids during exercise will
and developing dehydration (Brouns, 1991a,b). To avoid
induce a sparing of endogenous carbohydrate stores. The
such problems recommendations are given to reduce the
latter may reduce the development of glycogen depletion
intake of poorly fermentable dietary fibres and fat-rich
and hypoglycaemia and improve endurance capacity.
foods prior to competition and to ingest easily digestible
Many attempts have been made to modify fat metabolism
and absorbable energy sources during exercise. Because
(Brouns & van der Vusse, 1998; Hawley et al. 1998;
the drink/food composition can have a significant influence
Jeukendrup et al. 1998; Jeukendrup, 1999).
on the occurrence of gastrointestinal symptoms (Brouns,
Biomarkers to measure fat metabolism are mostly
1991a,b, 1997; Brouns & Beckers, 1993; Brouns &
related to (1) the rate of lipolysis, (2) the rate of fat
Kovacs, 1997a,b; van Nieuwenhoven et al. 2000), the
uptake in muscle and mitochondria and (3) ultimately the
industry tries to make claims related to this aspect.
fatty acid oxidation rate.
The tolerance to such products during exercise can be
assessed as follows.
Fat lipolytic rate
1. By questionnaire. Registration of the occurrence of
gastrointestinal symptoms during exercise after con- Methods used to determine lipolytic rate concern the
sumption of the test product in comparison to a control following biomarkers.
product (for fluids, usually artificially sweetened and
1. Determination of release of fatty acids and glycerol
coloured water; for solids, a normal food item such
from adipose tissue by microdialysis (invasive, needs
as an equicaloric amount of bread). Symptoms
routine, reliable; Arner, 1999; Frayn, 1999; Henriksson,
scored usually are related to degree of fullness,
1999).
stomach pains, regurgitation, intestinal cramps, bor-
2. Determination of the rate of appearance of glycerol
borigmy, flatulence and diarrhoea/loose stools. It has
(and fatty acids) in the circulation (surrogate marker,
to be noticed that controlled laboratory studies do not
reliable, requires stable isotopes, expensive).
mimic the situation in the field. Laboratory studies
3. Determination of changes in plasma fatty acid and gly-
have the advantage of controlling variables such as
cerol concentrations (surrogate marker, easy, reliable
work intensity, environmental temperature and humid-
only in combination with stable isotopes; Coggan,
ity, and quantitative timed intake of the test product.
1999; Landau, 1999; van Hall, 1999).
Field conditions make these controls very difficult but
are more reliable with respect to competition stress
and fluctuating exercise intensities during a compe-
tition. In addition, the food intake two or three days Fatty acid uptake rate by muscle
prior to the test may be of influence on the occurrence
Methods used to determine fatty acid uptake concern the
of lower intestinal symptoms. This means that food
following biomarkers.
intake should be standardized along with exercise pro-
grammes and the consumption of caffeine- and alcohol- 1. Determination of rate of free fatty acid disappearance
containing beverages during the days prior to the test from blood by stable isotopes (surrogate marker,
(easy but very labour-intensive testing, reliability of reliable, requires stable isotopes, expensive; Coggan,
laboratory data for field circumstances is speculative). 1999; Landau, 1999; Rennie, 1999; van Hall, 1999).
2. By motility measurements. These can be done reliably 2. Determination of net uptake by muscle using arterio-
in a laboratory but the outcome of the data in terms of venous techniques (invasive, reliable; McDonald,
relevance for competition circumstances is speculat- 1999) or microdialysis (Arner, 1999; Henriksson,
ive. Moreover, data are specific for the tested bio- 1999).
markers/endpoints only and do not allow one to
make statements on gastrointestinal tolerance in
general.
Fatty acid oxidation rate
3. Oesophageal sphincter pressure/reflux episodes (inva-
sive, needs routine, reliable, can be measured ambulat- 1. Determination of respiratory quotient (surrogate
ory; Schoeman et al. 1995). marker, easy, reliable for group means, less reliable
4. Contractile activity of the stomach and duodenum by for quantification of small changes within a subject).
antroduodenal manometry (invasive, needs routine, 2. Quantification by appearance of label in expired gas,
reliable, can be measured ambulatory but not suitable using stable isotopes (reliable, non-invasive, expens-
during exercise; Penning et al. 2001). ive; Coggan, 1999; Rennie, 1999; van Hall, 1999;
5. Intestinal transit rate by breath hydrogen from small van Hall et al. 1999).
intestine (non-invasive, easy, allows ambulatory 3. Quantification of intramuscular fat use by biopsy
measurement, cheap; Levitt, 1969) or whole-gut transit analysis (invasive, difficult to quantify, not reliable
measurements using scintigraphic techniques (non- for intervention effects; Hoppeler et al. 1999) or by
invasive, reliable, but expensive and exposure to NMR (non-invasive, complex technique, not reliable
radioactivity and measurement takes a long time; for quantification of treatment effects; Boesch et al.
Charles et al. 1995). 1999).
S182 F. Brouns et al.
Carbohydrate oxidation
An enhanced oxidation of ingested carbohydrate during
exercise will spare the endogenous carbohydrate stores and
this is generally thought to enhance endurance capacity.
It has been shown that when carbohydrate oxidation rates
drop below critical levels during exercise, this will
coincide with fatigue (Coyle et al. 1986). In an attempt
to optimize the carbohydrate oxidation from ingested
carbohydrate sources, numerous studies have been per-
formed in which the timing, amount and type of carbo-
hydrate were varied. These studies were reviewed
recently by Hawley et al. (1992) and Jeukendrup &
Jentjens (2000). Carbohydrate consumed during exercise
is oxidized in small amounts during the first hour of exer-
cise (, 20 g) and thereafter reaches a peak rate of , 1 g/
min (Jeukendrup & Jentjens, 2000). Even ingestion of
very large amounts of carbohydrate will not result in
higher oxidation rates. Generally, the timing of ingestion
has little or no influence as long as the amount of carbo-
hydrate ingested is sufficient (60 g/h). Glucose, sucrose,
maltose, maltodextrins and amylopectin starch are oxidized
at high rates (up to a maximum of 1 g/min) whereas fruc-
tose, galactose and amylose starch are oxidized at 25
50 % lower rates. Combinations of carbohydrates, how-
ever, may give higher exogenous carbohydrate oxidation
rates than single carbohydrates (Adopo et al. 1994).
Methods to determine carbohydrate oxidation concern
the following biomarkers.
1. Oxidation of ingested carbohydrate is usually
measured by using a 13C (or 14C) isotope of the carbo-
hydrate of interest.
2. A 13C or 14C carbohydrate tracer (i.e. glucose, fruc-
tose) is ingested and excretion of 13C (or 14C) in
breath is measured. In combination with a measure-
ment of total CO2 production, an accurate measure
of exogenous carbohydrate oxidation can be obtained.
The drawback of this method is usually the cost of the
tracer, although in some cases the natural 13C enrich-
ment of the carbohydrate (e.g. from cornstarch or cane
sugar) can be of use for testing (reliable, complex
method, expensive).
Nutritional status
There are a number of high-risk sports for marginal nutri-
tional intakes that may impact on performance capacity
and health (Brouns, 2002). The endpoints used for nutri-
tional status in healthy athletes mostly concern circulating
levels of vitamins, minerals and trace elements. Develop-
ment of analytical techniques has made it possible to
study biochemical vitamin deficits that occur at an early
stage of poor intakes. These measurements include the
determination of plasma vitamin levels by HPLC and
enzymatic stimulation tests (easy but invasive and also
expensive). Although minerals are fixed components of tis-
sues such as bone or muscle, this does not necessarily
mean that they are freely available for metabolic purposes.
The major fraction of the metabolic mineral pool is pre-
sent in blood plasma and interstitial fluid. The amount of
minerals circulating in body fluids is a resultant of different
ongoing processes. Absorption from food, on the one hand,
and uptake or release by tissues as well as losses/excretions
(by sweat, urine, faeces), on the other, determine the actual
mineral content in the blood. This mineral level remains
within a narrow range. Any mineral shortage will, in the
first instance, be compensated by reduced excretion or/
and increased release from tissues. With a continuing
shortage, plasma mineral levels will start to fall. The
latter will influence the uptake or release of minerals by
cells and thus the cell mineral status. During prolonged
periods of mineral deficits, cell growth and cell function
will become impaired, as will performance capacity. How-
ever, a short period with a relative shortage of one or more
minerals, e.g. as may occur during an ultra-endurance
event or during a multi-day cycling event, will not necess-
arily mean that health and performance will become
affected measurably.
With respect to trace minerals, it is possible to obtain a
certain figure or a certain level from samples such as
serum, tissue, hair and nails. The question one should ask
is about how representative these samples are for whole-
body status and for muscle in particular. Moreover, hair
and nail analyses give a figure related to intake in the
past and do not necessarily correlate with current status,
while in the case of hair analysis environmental factors
such as water trace minerals content (swimmers, frequent
showering) and use of hair products (shampoos, gels,
colour agents) may be confounding factors.
Growing knowledge over the years has indicated differ-
ent sample sites as being most representative for specific
trace elements. This makes it practically impossible to
perform a simple body fluid analysis with the goal of
making a statement on trace element status, as is the case
for vitamins and minerals as well.
Adenine nucleotides depletion
A number of important metabolic functions are related to
the availability of appropriate levels of phosphocreatine
and total adenine nucleotides (TAN; Balsom et al. 1994;
Williams et al. 1999; Brouns, 2002). The maintenance of
an appropriate level of ATP, through a rapid rephosphoryl-
ation of ADP from phosphocreatine, supports the creatine
phosphate shuttle to enhance the exchange of high-
energy phosphate from the site of the mitochondria to the
site of the cytosol. It also helps reduce acidosis in
muscle cells by buffering hydrogen ions, and indirectly
helps regulate the activation of carbohydrate breakdown
(glycolytic) processes in muscle by activation through the
products resulting from the hydrolysis of phosphocreatine,
which are inorganic phosphate and free creatine.
In the case of depletion of the phosphocreatine store
during all-out exercise, a number of metabolic changes
may occur in muscle that will lead to a change in TAN:
(1) depletion of the creatine phosphate store; (2) break-
down of ATP ! ADP ! AMP ! end products; (3) increase
in muscle and blood lactate; (4) increase in muscle and blood
ammonia; and (5) increase in blood xanthine, hypo-
xanthine, adenine and uric acid. After exercise the break-
down products mentioned might be lost from muscle.
 Food supplements for athletes
This results in a decreased TAN content. The resynthesis of
adenine nucleotides is a slow process, causing recovery to
a normal level to take up to three or four days. It has been
hypothesized that oral supply of ATP or ribose can lead to
a more rapid recovery of the TAN pool after intensive
training sessions or competitions.
Markers of TAN are:
1. Changes in plasma levels of ammonia, uric acid,
xanthine and hypoxanthine (surrogate markers for
qualitative changes in TAN, easy to perform, not
reliable to make any statements about muscle TAN
contents; Op tEinde et al. 2001).
2. Muscle TAN levels by biopsy analysis (invasive, rela-
tively easy, reliable) or NMR (non-invasive, complex
methodology, only for qualitative changes, not reliable
for quantification; Heerschap et al. 1999).
Muscle/tissue damage
There are a number of excellent reviews that highlight the
effect of exercise-induced free radical formation on various
tissues, as well as their measurement in biological fluids
(Halliwell & Gutteridge, 1985; Jenkins, 1988; Kanter,
1994; Duthie, 1999; Jackson, 1999; Poulsen et al. 1999;
Li, 2000). Muscle soreness after an intensive bout of exer-
cise is caused by an inflammation process. The micro
trauma (disruption at the Z band level of the sarcomeres)
that results from acute overload cannot be avoided by anti-
oxidant systems, because it is mechanical in nature. How-
ever, the repair process of the mechanically damaged
muscle fibres involves an inflammatory process, which
causes muscle pain, stiffness and loss of muscle strength,
especially two to five days after the sport event (Smith &
Miles 2000). It is suggested that free radicals play an
important role during this inflammatory process and that
supply with adequate amounts of antioxidants may lessen
both the severity and the duration of this delayed muscle
soreness.
Endurance exercise in polluted air, such as running a
major city marathon on a hot summers day in the smog,
has been suggested to lead to damage to the lung tissue
induced by ozone (Folinsbee, 2000). Free radical formation
and reactions are also suspected here to be the mediating
mechanism. Accordingly, nutritional substances such as
vitamin E supplementation are suggested to reduce such
damage (Evans, 1991) and lung function impairment
(Folinsbee, 2000).
Biomarkers used in tissue damage assessments are listed
below.
1. Free radical formation (complex assays, not reliable
for tissue damage quantification; Duthie, 1999;
Poulsen et al. 1999; Powers & Lennon, 1999).
2. Histological damage by biopsy analysis (reliable, inva-
sive, but difficult to quantify for whole muscle).
3. Damage markers in the circulation: myoglobin, muscle
enzymes (creatinphosphokinase; surrogate markers,
reliable for qualitative changes, difficult as quantitat-
ive measure, long-lasting post-damage adaptation
effects (up to six months) make cross-over studies
S183
very difficult, easy to measure; Clarckson, 1992;
Volfinger et al. 1994).
4. Strength performance and subjective pain (surrogate
marker, easy to measure, no quantification of
damage possible; Clarckson, 1992).
Muscle cramps
Low resting and exercise plasma Mg levels have repeat-
edly been reported in athletes who are involved in regular
endurance exercise. This has been thought to lead to
impaired energy metabolism, greater fatigue and to the
occurrence of muscle cramps. Cramps are also thought to
be caused by acute energy deficits as in the case of phos-
phocreatine depletion and reductions of ATP during very
intensive, all-out metabolism. Accordingly, attempts have
been made to study the effects Mg, creatine and ribose sup-
plementation. The clinical endpoint here, muscle cramp,
seems to occur infrequently and is very difficult to quan-
tify. Many factors may be involved in its aetiology.
1. Mineral levels in body fluids and muscle and TAN
levels do not seem to correlate with cramp occurrence
(invasive, easy to measure but unreliable; Maughan,
1986).
Performance
For many products, the claim will be enhanced perform-
ance. Of course, there are various forms of performance
and no single laboratory or field test could be used to gen-
eralize across all sports performances. Often the distinction
is made between endurance performance and high-intensity
exercise or sprint performance. In addition, in the litera-
ture, the terms endurance performance and endurance
capacity are often used as synonyms. However, endurance
capacity refers to the exercise time to volitional fatigue
whereas endurance performance relates to completing a
certain task (running a certain distance, cycling a certain
distance) as fast as possible. The latter is obviously a
more realistic approach since there are very few events
where athletes are asked to exercise for as long as they
can. Studies have investigated the reliability of both of
these test protocols and found coefficients of variation of
between 1 and 3 % for performance trials and up to 26 %
for time to exhaustion measurements (Jeukendrup et al.
1995). Performance trials have been developed and vali-
dated for the treadmill (self-paced runs for a fixed dis-
tance), intermittent running (Loughborough Intermittent
Shuttle Run Test; Nicholas et al. 2000), the cycle (self-
paced time trial; Jeukendrup et al. 1997), the rowing
ergometer (Schabort et al. 1999) and various other inter-
mittent sports such as soccer, squash (Romer et al. 2001)
and tennis (Vergauwen et al. 1998). Studies that try to
simulate a real event, especially in the field rather than in
the laboratory, are harder to conduct. The most compli-
cated types of performance belong to unpredictable team
games or sports involving complex decision-making and
motor skills. It is hard to find a way to measure adequately
all components of performance, and it is complicated to
design a protocol in which the same event is conducted
twice, before and after an intervention, or with a treatment
S184 F. Brouns et al.
and a placebo. Despite the difficulties in conducting
studies, strategies that enhance carbohydrate availability
have been shown to enhance cycling and running endur-
ance, cycling and running performance, and the perform-
ance in complex games such as tennis, soccer and ice
hockey.
When studying the effects of a nutrition supplement on
performance, an ergometer, treadmill or similar tool should
introduce negligible random error (variation) in its measure-
ments. Random error in the performance measurement
should also be minimized by choice of an appropriate
type of test. Tests based on physiological measures (e.g.
maximum oxygen uptake, anaerobic threshold) and tests
requiring self-selection of pace (e.g. constant-duration
and constant-distance tests) usually produce a random
error of at least 2 to 3 % in the measure of power output
(Paton & Hopkins, 2001). Random error for measures of
power in all-out sprints, incremental tests and very
short (1 5 min) constant-power tests to exhaustion may
be as low as 1 %. Measures with such low error might be
suitable for tracking the small changes in competitive per-
formance that matter to elite cyclists.
In all performance tests it is extremely important that
all confounding factors are removed or standardized
where possible (music, encouragement, feedback). In
addition it must be kept in mind that some of these per-
formance tests may not always pick up the small
improvements in performance that are relevant to an
elite athlete.
1. Sprint performance can be assessed by a Wingate test
(easy to perform, reliable).
2. Endurance capacity can be assessed by a time to
exhaustion test (poor reproducibility, limited practical
value).
3. Endurance performance can be assessed by a constant-
duration or constant-distance test (good reproducibil-
ity, needs extremely strict standardization).
4. Sport-specific performance and field tests could
include measurements of performance in real life
such as during a soccer match or simulations of a par-
ticular sport or discipline in field conditions (very few
established tests, needs extremely strict standardiz-
ation; a review describing details of sport-specific
tests is given by Kearney et al. 2000).
Claims and legal aspects
Generally, no food regulation exists (yet) to control per-
formance or health benefit statements with respect to the
supportive scientific evidence. Also, depending on the
dosage, many of the FC used are in the grey zone between
foods and drugs. The question whether an effective dosage
of a certain FC that never will be consumed in that part-
icular quantity with the normal diet is nutritional or pharma-
cological is relevant in this respect. Many of these
products are sold as dietary supplements, mostly by direct
mail order systems and the Internet, which are difficult to
control.
Development of the sports foods and drinks market is
highly attractive to the food industry. However, one
should be aware of the fact that current developments in
the field of consumer protection, regulatory environment
and product liability aspects do call for a more careful
and well defined product development and marketing
programme than is usually the case. Obtaining scientific
support for a product claim is an essential issue in this
respect.
The establishment of a code of practice among sports
food/supplement companies is another must, as long as
an appropriate legislation is not in place.
References
Adopo E, Peronnet F, Massicotte D, Brisson GR & Hillaire-
Marcel C (1994) Respective oxidation of exogenous glucose
and fructose given in the same drink during exercise. Journal
of Applied Physiology 76, 1014 1019.
Arner P (1999) Microdialysis use in human exercise studies. Pro-
ceedings of the Nutrition Society 58, 913 917.
Balsom PD, Soderlund B & Ekblom B (1994) Creatine in
humans, with special reference to creatine supplementation.
Sports Medicine 18, 268280.
Beckers E, Leiper JB & Davidson J (1992) Comparison of aspira-
tion and scintigraphic techniques for the measurement of gast-
ric emptying rates of liquids in humans. Gut 31, 115 117.
Beckers E, Rehrer NJ & Brouns F (1988) Determination of total
gastric volume, gastric secretion and residual meal using the
double sampling technique of George. Gut 29, 1725 1729.
Boesch C, Decombaz J, Slotboom J & Kreis R (1999) Obser-
vation of intramyocellular lipids by means of H magnetic
resonance spectroscopy. Proceedings of the Nutrition Society
58, 841 850.
Boileau RA & Horswill CA (2000) Body composition in sports:
measurement and applications for weight gain. In Exercise
and Sport Science, pp. 319 338 [WE Garret and DT Kirkendall,
editors]. Philadelphia, PA: Lippincott, Williams & Wilkins.
Brouns F (1991a) Etiology of gastrointestinal disturbances during
endurance events. Scandinavian Journal of Medicine and
Science in Sports 1, 66 77.
Brouns F (1991b) Heatsweatdehydrationrehydration: a praxis
oriented approach. Journal of Sports Science 9, 143152.
Brouns F (1997) Aspects of dehydration and rehydration in sport.
World Reviews in Nutrition and Diet 82, 63 80.
Brouns F (2002) Essentials of Sports Nutrition. Chichester: John
Wiley & Sons (in press).
Brouns F & Beckers E (1993) Is the gut an athletic organ: diges-
tion, absorption and exercise. Sports Medicine 15, 242 257.
Brouns F & Kovacs E (1997a) Functional drinks for athletes.
Trends in Food Science and Technology 8, 414 421.
Brouns F & Kovacs E (1997b) Functional foods for athletes.
Trends in Food Science and Technology 8, 358 363.
Brouns F & van der Vusse G (1998) Utilization of lipids during
exercise in human subjects: metabolic and dietary constraints.
British Journal of Nutrition 79, 117 128.
Charles F, Camilleri M, Phillips SF, Thomforde GM & Forstrom
LA (1995) Scintigraphy of the whole gut: clinical evaluation of
transit disorders. Mayo Clinical Proceedings 70, 113.
Clarckson PM (1992) Exercise induced muscle damage: animal
and human models. Medicine Science Sports Exercise 24,
510 511.
Coggan AR (1999) Use of stable isotopes to study carbohydrate
and fat metabolism at the whole-body level. Proceedings of
the Nutrition Society 58, 953 961.
Coyle EF, Coggan AR, Hemmert MK & Ivy JL (1986) Muscle
glycogen utilization during prolonged strenuous exercise
 Food supplements for athletes
when fed carbohydrate. Journal of Applied Physiology 61,
165172.
Dill DB & Costill DL (1974) Calculation of percentage changes
in volumes of blood, plasma and red cells in dehydration. Jour-
nal of Applied Physiology 37, 247 248.
Duthie GG (1999) Determination of activity of antioxidants in
human subjects. Proceedings of the Nutrition Society 58,
10151024.
Evans WJ (1991) Muscle damage: nutritional considerations.
International Journal of Sports Nutrition 1, 214 224.
Folinsbee LJ (2000) Effects of air pollutants on exercise. In
Exercise and Sport Science, pp. 285 298 [WE Garret and
DT Kirkendall, editors]. Philadelphia, PA: Lippincott, Williams
& Wilkins.
Frayn KN (1999) Macronutrient metabolism of adipose tissue at
rest and during exercise: a methodological viewpoint. Proceed-
ings of the Nutrition Society 58, 877 886.
Gadian DG (1995) NMR and its Applications to Living Systems.
Oxford: Oxford University Press.
Halliwell B & Gutteridge JMC (1985) Free Radicals in Biology
and Medicine. Oxford: Clarenden Press.
Hawley JA, Brouns F & Jeukendrup A (1998) Strategies to
enhance fat utilization during exercise. Sports Medicine 25,
241257.
Hawley JA, Dennis SC & Noakes TD (1992) Oxidation of carbo-
hydrate ingested during prolonged endurance exercise. Sports
Medicine 14, 27 42.
Heerschap A, Houtman C, int Zandt HJA, Van den Bergh AJ &
Wieringa B (1999) Introduction to in vivo 31P magnetic reson-
ance spectroscopy of (human) skeletal muscle. Proceedings of
the Nutrition Society 58, 861 870.
Henriksson J (1999) Microdialysis of skeletal muscle at rest. Pro-
ceedings of the Nutrition Society 58, 919 923.
Hoppeler H, Billeter R, Horvath PJ, Leddy JJ & Pendergast DR
(1999) Muscle structure with low and high fat diets in well
trained male runners. International Journal of Sports Medicine
20, 522526.
Hultman E (1967) Studies on muscle metabolism of glycogen and
active phosphate in man with special reference to exercise and
diet. Scandinavian Journal of Clinical and Laboratory Investi-
gation 19, Suppl. 94, 1 64.
Jackson MJ (1999) An overview of methods for assessment of
free radical activity in biology. Proceedings of the Nutrition
Society 58, 1001 1006.
Jenkins RR (1988) Free radical chemistry: relationship to exer-
cise. Sports Medicine 5, 156 170.
Jeukendrup AE (1999) Dietary fat and physical performance.
Current Opinions in Clinical Nutrition and Metabolism Care
2, 521 526.
Jeukendrup AE, Brouns F, Wagenmakers AJM & Saris WHM
(1997) Carbohydrate feedings improve 1 h time trial cycling per-
formance. International Journal of Sports Medicine 18, 125129.
Jeukendrup AE & Jentjens R (2000) Oxidation of carbohydrate
feedings during prolonged exercise: current thoughts, guide-
lines and directions for future research. Sports Medicine 29,
407424.
Jeukendrup AE, Saris WHM, Brouns F & Kester ADM (1995) A
new validated endurance performance test. Medicine Science
Sport Exercise 27, 266 270.
Jeukendrup AE, Saris WHM & Wagenmakers AJM (1998) Fat
metabolism during exercise: a review. Part III: The effects of
nutritional interventions. International Journal of Sports Medi-
cine 19, 371 379.
Kanter M (1994) Free radicals, exercise and antioxidant supple-
mentation. International Journal of Sport Nutrition 4, 205220.
Kearney JT, Rundell KW & Wilber RL (2000) Measurement
of work and power in sport. In Exercise and Sport Science,
S185
pp. 31 52 [WE Garret and DT Kirkendall, editors]. Philadel-
phia, PA: Lippincott, Williams & Wilkins.
Kleiner SM, Bazzarre TL & Litchford MD (1990) Metabolic pro-
files, diet, and health practices of championship male and
female bodybuilders. Journal of the American Dietetic Associ-
ation 90, 962 967.
Landau BR (1999) Glycerol production and utilization measured
using stable isotopes. Proceedings of the Nutrition Society 58,
973 978.
Leiper JB & Maughan RJ (1988) Experimental models for the
investigation of water and solute transport in man: implications
for oral rehydration solutions. Drugs 36, Suppl. 4, 65 79.
Lemon PWR (2000) Protein metabolism during exercise. In
Exercise and Sport Science, pp. 19 30 [WE Garret and DT
Kirkendall, editors]. Philadelphia, PA: Lippincott, Williams
& Wilkins.
Levitt MD (1969) Production and excretion of hydrogen gas in
man. New England Journal of Medicine 281, 122 127.
Li LJ (2000) Free radicals and antioxidants in exercise and sports.
In Exercise and Sport Science, pp. 299 317 [WE Garret and
DT Kirkendall, editors]. Philadelphia, PA: Lippincott, Williams
& Wilkins.
McDonald IA (1999) Arterio-venous differences to study macro-
nutrient metabolism: introduction and overview. Proceedings
of the Nutrition Society 58, 871 875.
McMurray RG & Hackney AC (2000) Endocrine responses to
exercise and training. In Exercise and Sport Science, pp.
135162 [WE Garret and DT Kirkendall, editors]. Philadelphia,
PA: Lippincott, Williams & Wilkins.
Maughan R (1999) Nutritional ergogenic aids and exercise per-
formance. Nutrition Research Reviews 12, 255 276.
Maughan RJ (1986) Exercise induced muscle cramp: a prospec-
tive biochemical study in marathon runners. Journal of Sports
Science 4, 31 34.
Nicholas CW, Nuttall FE & Williams C (2000) The Loughbor-
ough Intermittent Shuttle Test: a field test that simulates the
activity pattern of soccer. Journal of Sports Science 18,
97 104.
Op tEindeB, van Leempurtte B, Brouns F & van der Vusse G
(2001) No effects of oral ribose supplementation on repeated
maximal exercise and de novo ATP resynthesis. Journal of
Applied Physiology 91, 2278 2281.
Paton CD & Hopkins WG (2001) Tests of cycling performance.
Sports Medicine 31, 489 496.
Penning C, Gielkens HA, Hemelaar M, Lamers CB & Masclee
AA (2001) Reproducibility of antroduodenal motility during
prolonged ambulatory recording. Neurogastroenterology and
Motility 2, 133 141.
Poulsen HE, Weimann A & Loft S (1999) Methods to detect
DNA damage by free radicals: relation to exercise. Proceedings
of the Nutrition Society 58, 1007 1014.
Powers SK & Lennon SL (1999) Analysis of cellular responses to
free radicals: focus on exercise and skeletal muscle. Proceed-
ings of the Nutrition Society 58, 1025 1033.
Price TB, Rothman DL & Shulman RG (1999) NMR of glycogen
in exercise. Proceedings of the Nutrition Society 58, 851859.
Rennie MJ (1999) An introduction to the use of tracers in nutri-
tion and metabolism. Proceedings of the Nutrition Society 58,
935 944.
Rennie MJ & Tipton KD (2000) Protein and amino acid metab-
olism during and after exercise and the effects of nutrition.
Annual Review of Nutrition 20, 457 483.
Romer LM, Barrington JP & Jeukendrup AE (2001) Effects of
oral creatine supplementation on high intensity, intermittent
exercise performance in competitive squash players. Inter-
national Journal of Sports Medicine 22, 546 552.
Schabort EJ, Hawley JA, Hopkins WG & Blum H (1999) High
S186 F. Brouns et al.
reliability of performance of well-trained rowers on a rowing
ergometer. Journal of Sports Science 17, 627 632.
Schoeman MN, Tippett MD, Akkermans LMA, Dent J & Holloway
RH (1995) Mechanisms of gastroesophageal reflux in ambulant
healthy human subjects. Gastroenterology 108, 8391.
Smith LL & Miles MP (2000) Exercise induced muscle injury and
inflammation. In Exercise and Sport Science, pp. 401 412
[WE Garret and DT Kirkendall, editors]. Philadelphia, PA:
Lippincott, Williams & Wilkins.
Spriet L (1999) Ergogenic aids: recent advances and threats. In
Optimizing Sports Performance, pp. 185 238 [DR Lamb and
D Murray, editors]. Carmel, CA: Cooper Publishing.
Van Hall G (1999) Correction factors for 13C-labelled substrate
oxidation at whole-body and muscle level. Proceedings of the
Nutrition Society 58, 979 986.
Van Hall G, Gonzalez-Alonso J, Sacchetti M & Saltin B (1999)
Skeletal muscle substrate metabolism during exercise: method-
ological considerations. Proceedings of the Nutrition Society
58, 899 912.
Van Marken Lichtenbelt WD (2001) The use of bioelectrical
impedance analysis (BIA) for estimation of body composition.
In Body Composition Analysis of Animals. A Handbook of Non-
destructive Methods, pp. 161 187 [J Speakman, editor]. Edin-
burgh: Cambridge University Press.
Van Marken Lichtenbelt WD & Fogelmholm M (1999) Body
composition. In Regulation of Food Intake and Energy Expen-
diture, pp. 383 404 [MS Westerterp-Plantenga, AB Steffens
and A Tremblay, editors]. Milan: Medical Publishing & New
Media (EDRA).
Van Marken Lichtenbelt WD, Kester A, Baarends EM &
Westerterp KR (1996) Bromide dilution in adults: optimal
equilibration time after oral administration. Journal of Applied
Physiology 81, 653 656.
Van Nieuwenhoven MA, Brouns F & Brummer RJ (2000)
Exercise and gastrointestinal function. In Exercise and Sport
Science, pp. 191 216 [WE Garret and DT Kirkendall, editors].
Philadelphia, PA: Lippincott, Williams & Wilkins.
Van Nieuwenhoven MA, Wagenmakers AJM, Senden JMG, Brouns
F & Brummer R-JM (1999) Performance of the [13C]-acetate
gastric emptying breath test during physical exercise. European
Journal of Clinical Investigation 29, 922928.
Vergauwen L, Brouns F & Hespel P (1998) Carbohydrate
supplementation improves stroke performance in tennis. Medi-
cine Science Sports Exercise 30, 1289 1295.
Volek JS (2000) Enhancing exercise performance: nutritional
implications. In Exercise and Sport Science, pp. 471 486
[WE Garret and DT Kirkendall, editors]. Philadelphia, PA:
Lippincott, Williams & Wilkins.
Volfinger I, Lassourd V, Michaux JM, Braun JP & Toutain PL
(1994) Kinetic evaluation of muscle damage during exercise
by calculation of amounts of creatine kinase released. American
Journal of Physiology 266, R434 R441.
Wagenmakers AJM (1999a) Nutritional supplements: effects
on exercise performance and metabolism. In Perspectives in
Exercise Science and Performance, Vol. 12, pp. 207 259
[DR Lamb, editor]. Carmel, CA: Cooper Publishing.
Wagenmakers AJM (1999b) Tracers to investigate protein and
amino acid metabolism in human subjects. Proceedings of
the Nutrition Society 58, 987 1000.
Walberg-Rankin J, Ecksteien Edmonds C & Gwazdauskas FC
(1993) Diet and weight changes of female bodybuilders
before and after competition. International Journal of Sport
Nutrition 3, 87 102.
Wang Z & Heymsfield SB, et al. (1995) Systematic organisation
of body-composition methodology: an overview with emphasis
on component-based methods. American Journal of Clinical
Nutrition 61, 457 465.
Westerterp KR (1999) Body composition, water turnover and
energy turnover assessment with labelled water. Proceedings
of the Nutrition Society 58, 945 951.
Westerterp KR, van Marken Lichtenbelt WD & Wouters L (1995)
The Maastricht protocol for the measurement of body compo-
sition and energy expenditure with labeled water. Obesity
Research 3, 49 57.
Williams MH, Kreider RB & Branch JD (1999) Creatine: The
Power Supplement. Champaign, IL: Human Kinetics.
S166 P. C. Calder and S. Kew
molecule CD4 on their surface) and cytotoxic T cells (these
are distinguished by the presence of CD8 on their surface).
All cells of the immune system originate in bone marrow.
They are found circulating in the bloodstream, organised
into lymphoid organs such as the thymus, spleen, lymph
nodes and gut-associated lymphoid tissue, or dispersed in
other locations around the body.
Innate and acquired immunity
Innate immunity is the first line of defence against infec-
tious agents. It is present prior to exposure to pathogens
and its activity is not enhanced by such exposures. Innate
immunity is concerned with preventing entry of infectious
agents into the body and, if they do enter, with their rapid
elimination. Elimination can occur by:
1. direct destruction of pathogens by complement, by
toxic chemicals (e.g. superoxide radicals and hydrogen
peroxide) released by phagocytes or by toxic proteins
released by natural killer cells;
2. engulfing pathogens by the process of phagocytosis,
which is made more efficient by coating the invading
pathogen with host proteins like complement or
antibodies, and their subsequent destruction.
Acquired immunity involves the specific recognition of
molecules (antigens) on an invading pathogen, which
distinguish it as being foreign to the host. The recognition
of antigens is by antibodies (immunoglobulins (Ig) pro-
duced by B lymphocytes) and by T lymphocytes. T
lymphocytes are only able to recognise antigens displayed
on cell surfaces. Therefore, infection of a cell by an intra-
cellular pathogen is signalled to T lymphocytes by cell
surface expression of peptide fragments derived from the
pathogen. These fragments are transported to the surface
of the infected cell and expressed there in conjunction
with proteins termed major histocompatibility complex
(MHC); in man MHC is termed human leucocyte antigen.
It is the combination of the pathogen-derived peptide
fragment bound to MHC that is recognised by T lympho-
cytes. There are two classes of MHC, MHC I and MHC
II, and the source of the peptide bound to each differs.
MHC I binds peptides that originate from pathogen pro-
teins synthesised within the host cell cytosol; typically
these are from viruses or certain bacteria. The peptides
bound to MHC II are derived from pathogens that have
been phagocytosed by macrophages or endocytosed by
antigen-presenting cells (macrophages, dendritic cells, B
lymphocytes). The MHC peptide complex is recognised
by the T cell receptor on T lymphocytes. T lymphocytes
expressing CD8 recognise MHC I, while T lymphocytes
expressing CD4 recognise MHC II. Thus, intracellular
pathogens stimulate cytotoxic T lymphocytes to destroy
the infected cell, while extracellular pathogens stimulate
a helper T cell-mediated response.
The acquired immune system includes a component of
memory, such that if the antigen is encountered again
(i.e. there is re-infection) the response is faster and stronger
than the initial response. Although the immune system as a
whole can recognise tens of thousands of antigens, each
lymphocyte can recognise only one antigen and so the
number of lymphocytes specific for a particular antigen
must be very low. However, when an antigen is encoun-
tered it binds to the small number of lymphocytes that
recognise it and causes them to divide so as to increase
the number of cells capable of mounting a response to
the antigen; this is the process termed lymphocyte
expansion or proliferation. B lymphocytes proliferate and
mature into antibody-producing cells (plasma cells) and
T lymphocytes proliferate and are able directly to destroy
virally infected cells (cytotoxic T lymphocytes) or control
the activity of other cells involved in the response (helper
T cells). The B lymphocyte response to antigen is termed
humoral immunity and the T cell response is termed
cell-mediated immunity.
Integration of the immune response
Communication within the acquired immune system and
between the innate and acquired systems is brought about
by direct cell-to-cell contact involving adhesion molecules
and by the production of chemical messengers. Chief
among these chemical messengers are proteins called
cytokines, which can act to regulate the activity of the
cell that produced the cytokine and/or of other cells.
Each cytokine can have multiple activities on different
cell types. Cytokines act by binding to specific receptors
on the cell surface and thereby induce changes in
growth, development or activity of the target cell.
When an immunological stimulus is encountered, the
innate response, including its inflammatory component,
responds initially, acting directly to eliminate it by the
activities of complement, phagocytosis, etc. Cytokines
(e.g. tumour necrosis factor-a (TNF-a), interleukin (IL)-1
and IL-6) produced by the cells involved in the innate
response, especially monocytes and macrophages, will
regulate this response and also act systemically on the
liver to promote acute phase protein synthesis, on skeletal
muscle and adipose tissue to promote proteolysis and lipo-
lysis, respectively (this is believed to be the bodys way of
providing fuels to the immune system) and on the brain to
reduce appetite and induce fever. These cytokines will also
interact with T lymphocytes. Antigen-presenting cells,
which include activated monocytes and macrophages,
will present antigen to T lymphocytes and so the acquired
immune response will be triggered. Now there will be a
cell-mediated response to the antigen. T lymphocytes will
produce cytokines which will regulate the activity of the
cells involved in the innate response (monocytes, macro-
phages, natural killer cells), promote the proliferation of
B and T lymphocytes and promote antibody production
by B lymphocytes. By virtue of the integrated innate and
acquired responses the source of the antigen should be
eliminated and a component of immunological memory
will remain (Fig. 1).
Biomarkers of immune function
There is a wide range of methodologies by which to assess
the immune response and the impact of nutrient supply on
immune function (for a discussion see Cunningham-
Rundles, 1998). Assessments can be made of cell functions
 Functional foods and immunity S167

ex vivo (i.e. of the isolated cells outside the body and Ex vivo measures
studied in short- or long-term culture), of indicators of
Animal studies often investigate the functions of immune
immune function in vivo (e.g. by measuring the concen-
cells isolated from the blood, thymus, spleen, lymph
trations of proteins relevant to immune function in the
nodes, peritoneal cavity and, in some cases, from the
bloodstream), of responses to an immunological challenge
bone marrow, lungs, liver and gastrointestinal tract.
(e.g. inoculation with an antigen, a vaccine or live bacteria)
Human studies are more limited and routinely only the
or, in human studies, of the incidence and severity of infec-
blood pool is sampled. In some cases other sources of
tious diseases. Clearly, animals offer greater access to the
human immune cells have been studied: for example,
immune system, but it is important that observations
studies of asthma and respiratory illness often use cells col-
made in animal studies be confirmed in man. Although
lected by bronchoalveolar lavage. Measures of immune
ex vivo measures of immune function are made frequently,
function that can be made on cells cultured ex vivo include:
changes in these may not necessarily result in an altered
immune response in vivo. Thus, in order to ascertain the 1. Phagocytosis of bacteria, sheep red blood cells or
effect of an intervention on the immune response, measures yeast particles by neutrophils, monocytes and macro-
reflecting in vivo activity of the immune system are phages; this can be coupled with measures of bacterial
superior to ex vivo measures. However, the two approaches killing.
should be used in combination to understand better the 2. Respiratory burst (superoxide generation) by neutro-
mechanism of impact of an intervention. phils, monocytes and macrophages in response to

Fig. 1. Integration of the immune response.

S168 P. C. Calder and S. Kew
bacteria or bacterial peptides; this can be coupled with
measures of bacterial killing.
3. Natural killer cell activity, measured as killing of
tumour cells known to be specific targets for natural
killer cells.
4. Cytotoxic T lymphocyte activity, measured as killing
of virally infected cells known to be specific targets
for cytotoxic T cells.
5. Lymphocyte proliferation. This is the increase in
number of lymphocytes in response to a stimulus.
Most often this is measured as the incorporation of
radioactively labelled thymidine into the DNA of the
dividing lymphocytes, although a number of other
measures are available. Agents used to stimulate lym-
phocyte proliferation include concanavalin A and
phytohaemagglutinin, which stimulate T lymphocytes.
These agents are all known as mitogens and the pro-
cess as mitogen-stimulated lymphocyte proliferation.
If the individual has been sensitised to an antigen (or
allergen), then the antigen can be used to stimulate
lymphocyte proliferation.
6. Production of cytokines by lymphocytes, monocytes
and macrophages. This usually requires the cells to
be stimulated. For lymphocytes, mitogens (or antigens,
if the individual has been sensitised) are used, while
for monocytes and macrophages bacterial lipopolysac-
charide is most often used.
7. Production of total or specific Ig by lymphocytes.
8. Cell surface expression of molecules involved in anti-
gen presentation (e.g. MHC) and in cellular activation
(e.g. cytokine receptors).
In vivo measures
Measures of immune function that reflect in vivo activity
include:
1. Size of lymphoid organs.
2. Cellularity of lymphoid organs.
3. Number and types of immune cells circulating in the
bloodstream.
4. Cell surface expression of molecules involved in anti-
gen presentation (e.g. MHC) and in cellular activation
(e.g. cytokine receptors).
5. Circulating concentration of thymulin and its activity.
6. Circulating concentrations of total Ig and of the Ig
subclasses.
7. Circulating concentrations of Ig specific for antigens
after an antigen challenge.
8. Concentration of secretory IgA in saliva, tears and
intestinal washings.
9. Circulating concentrations of cytokines.
10. Delayed-type hypersensitivity (DTH) response to
intradermal application of an antigen to which the
individual has already been exposed; this measures
the cell-mediated immune response. The response is
measured as the size of the swelling (termed indura-
tion) around the area of application at a period (usually
forty-eight hours) after the application.
11. Resistance to challenge with live pathogens; the out-
come is usually survival, although this can be coupled
with some of the above measures and with measures of
the numbers of pathogens found in various organs (e.g.
spleen, lymph nodes, liver). This has been used in
animal experiments.
12. Incidence and severity of infectious diseases. This has
been used in some human studies to suggest inter-
actions between nutrient status and immune function.
In vitro studies
In addition to the above approaches, which can be used to
assess the impact of a nutrient supplied in the diet on
immune function, in vitro studies adding the nutrient in
pure form directly to immune cells in culture can be
used. Each of the cell functions listed above under ex
vivo measures can be studied in this way. In vitro studies
use conditions that are highly controlled, although they are
often rather unphysiological in nature. For example, the
cells are cultured in isolation from the other types of cell
that they would come into contact with in the body (this
is also a problem with many ex vivo measures) and the con-
centrations of the pure nutrient added to the cultures are
often greatly in excess of those that can be attained in
vivo. Also, the exact form of the nutrient added directly
to cell cultures might be different from the form that is
available to the cells in vivo. Nevertheless, in vitro studies
are useful to identify the potential effects of dietary com-
ponents and to study their mechanisms of action. However,
it is necessary that effects identified in in vitro studies be
confirmed in controlled dietary studies.
There is significant inter-individual variation in
immune biomarkers
Variation in cellular immune responses among individuals
is not a great problem in animal studies since these most
often use inbred strains. However, it is important to note
that some responses do differ among animal species and
even among different strains within a species. It is possible
that immune cells from different species and strains will
exhibit different sensitivities to the amount of a nutrient
in the diet. Thus, extrapolations from animal studies to
man should be made cautiously. There is wide variation
in immune cell responses among healthy human subjects
(Table 1; see also Yaqoob et al. 1999), and immune
responses can be affected by the presence of disease.
Individuals with deficient immune responses are more sus-
ceptible to infectious agents and suffer greater morbidity
and mortality as a result of infections. Such individuals
have one or more key immune responses that fall below
the threshold that represents normality. However, it is
not at all clear whether variation in immune responses
within the normal range results in variable susceptibility
to infection. Understanding the relationship between the
variations in immune responses and in susceptibility to
infection is complicated by a number of other factors.
For example, vaccination is widely used to prime the
immune system to efficiently eliminate certain pathogens
that may or may not exist in the environment at some
later stage, antibiotics are used widely to help the host
 Functional foods and immunity S169
Table 1. Variation in immune responses among healthy human subjects. Blood was collected from healthy human volunteers (aged 40 to 60 years) in the fasting state. Phagocytosis and oxi-

ture of lymphocytes and monocytes) were prepared using standard procedures and were cultured under standard conditions in the presence of either bacterial lipopolysaccharide (LPS;

40
40
37
37
57

57
52
56

56
53
60
measured by specific ELISA (see Yaqoob et al. 2000; Thies et al. 2001a,b). LPS was used to stimulate production of tumour necrosis factor-a (TNF-a), interleukin (IL)-1b and IL-6, while Con
dative burst in response to Escherichia coli were determined in whole blood by flow cytometry (see Thies et al. 2001a); data are expressed as the % of active cells. Mononuclear cells (a mix-

15 mg/ml) or concanavalin A (Con A; 25 mg/ml; see Yaqoob et al. 2000; Thies et al. 2001b). After 24 h of culture, the concentrations of cytokines in the cell culture supernatants were

A was used to stimulate the production of IL-2, interferon-g (IFN-g) and IL-4. Lymphocyte proliferation was determined as the incorporation of [3H]thymidine over the last 18 h of a 66 h culture

n
eliminate pathogens once infection takes hold, and
populations (at least in developed countries) exist in
increasingly hygienic (i.e. pathogen-free) environments.
Nevertheless, studies in the developing world demonstrate

64

216
37

60

643
21

10
7

16 665

120
that strategies which improve immune function in popu-
SD

lations with impaired immunity do lower the incidence

and severity of some infections in at-risk groups (Calder
& Jackson, 2000). Despite this most important observation,
Women
period (see Yaqoob et al. 2000; Thies et al. 2001b); data are expressed as cpm of thymidine incorporated/2 105 cells in the initial culture

it does not necessarily follow that increasing immune

105

337

1177
55

83

499
Mean

function per se will improve host resistance, and individ-

73
90
20
91

31 747 uals with functions in the normal range may not benefit
from increased immune function. In other words,
increasing the activity of one or more components of the
immune system (as tested by the above methodologies)
4024 73 058

may not necessarily be of any benefit to the individual,

24 5789
30 3898
10 436
03 174

13 281
15 737

just as decreasing the activity of one or more components

Range

12 95
67 98
4 46
45 99

of the immune system (as tested by the above

methodologies) may not be detrimental to the individual.
The wide variation in any given immune response
among individuals is in part related to genetic polymorph-
isms, which regulate the expression of cytokines, cytokine
76
50
56
51
50

80

82
78
72

82
83

receptors, human leucocyte antigen, adhesion molecules

n

and so on. However, other factors such as age, gender,

smoking habits, habitual levels of exercise, alcohol con-
sumption, diet, stage in the female menstrual cycle,
812
33
213

61
79

stress, and history of infections and vaccinations are

16

11

131
14 612
7

4
SD

likely to be important contributors to the observed vari-

ation (Fig. 2). Some studies suggest that the prenatal and
early life environments can affect later immune compe-
tence (Beach et al. 1982; Calder & Yaqoob, 2000), resist-
Men

1273
111
50

85

696
365

ance to infections (Moore et al. 1997; Rayon et al. 1997)

Mean

32 241
73
90
21
92

and response to vaccination (McDade et al. 2001). Con-

sideration of the variation in immune response among indi-
viduals is important when designing and interpreting
human studies. It is not clear whether there is variation
in responsiveness of the human immune system to the
475370 583

424931
354857

effects of nutrients, although some studies suggest that

05145

22437
12488

09838
Range

4193
6898
3 47
7698

this varies with age (e.g. Meydani et al. 1991).

Nutrition and immune function

General
Oxidative burst to E. coli (% of cells)

Oxidative burst to E. coli (% of cells)

Phagocytosis of E. coli (% of cells)

Phagocytosis of E. coli (% of cells)

Undernutrition impairs the immune system, suppressing

Proliferation (cpm/2 105 cells)

immune functions that are fundamental to host protection

against pathogenic organisms. Undernutrition leading to
TNF-a production (ng/ml)

impairment of immune function can be due to insufficient

IFN-g production (IU/ml)
IL-1b production (ng/ml)

IL-4 production (pg/ml)

IL-6 production (ng/ml)
Function

IL-2 production (IU/ml)

intakes of energy and macronutrients and/or due to

deficiencies in specific micronutrients (vitamins and
minerals). Often these occur in combination. Clearly the
impact of undernutrition is greatest in developing
countries, but it is also important in developed countries
especially amongst the elderly, individuals with eating dis-
orders, alcoholics, patients with certain diseases, and
premature and small-for-gestational-age babies. Many
standard deviation.

studies of the interaction between nutrient availability

and immune function have been performed in animals
Lymphocyte

and these have often compared the effects of diets

Neutrophil

Monocyte
Cell type

containing insufficient, sufficient and, in some cases,

excess amounts of an individual nutrient under study.
SD ,

Such studies are valuable because they allow the effect

S170 P. C. Calder and S. Kew
Fig. 2. Sources of variation in immune function.
of the nutrient to be studied in a highly controlled setting,
and because the immune system of experimental animals is
more accessible than that of man. In man, studies of the
immunological impact of nutrient deficiencies have been
imposed by habitual consumption of diets deficient in
one or more nutrients, or, more rarely, through individuals
with diseases that result in inability to absorb nutrients.
These studies have made it clear that there are a number
of nutrients whose availability at an appropriate level is
essential if the immune response is to operate efficiently.
Nutrients that have been demonstrated (in either animal
or human studies) to be required for the immune system
to function efficiently include essential amino acids, the
essential fatty acid linoleic acid, vitamin A, folic acid, vita-
min B6, vitamin B12, vitamin C, vitamin E, Zn, Cu, Fe and
Se (for reviews see Gross & Newberne, 1980; Chandra,
1991; Kuvibidila et al. 1993; Scrimshaw & SanGiovanni,
1997; Calder & Jackson, 2000). Practically all forms of
immunity may be affected by deficiencies in one or more
of these nutrients and the deficient animal or individual
becomes more susceptible to infections. Animal and
human studies have demonstrated that adding the deficient
nutrient back to the diet can restore immune function and
resistance to infection (for reviews see Chandra, 1991;
Scrimshaw & SanGiovanni, 1997; Calder & Jackson,
2000).
Vitamin E and zinc: well researched examples of the
impact of the dietary supply of micronutrients on immune
function
Vitamin E. Vitamin E is the major lipid-soluble anti-
oxidant in the body and is required for protection of
membrane lipids from peroxidation. Since free radicals
and lipid peroxidation are immunosuppressive, it is
considered that vitamin E should act to optimise and
even enhance the immune response (for reviews see
Bendich, 1993; Meydani & Beharka, 1998). In laboratory
animals, vitamin E deficiency decreased spleen lymphocyte
proliferation, natural killer cell activity, specific antibody
production following vaccination, and phagocytosis by
neutrophils (for reviews see Bendich, 1993; Meydani &
Beharka, 1998). The effects of vitamin E depletion are
more marked if animals are fed a diet containing a high
level of polyunsaturated fatty acids (Corwin & Schloss,
1980). Vitamin E deficiency increases susceptibility of
animals to infectious pathogens (for references see
Meydani & Beharka, 1998). Vitamin E supplementation
of the diet of laboratory animals enhances antibody pro-
duction, lymphocyte proliferation, natural killer cell
activity, and macrophage phagocytosis (for references see
Meydani & Beharka, 1998). Adding vitamin E to the diet
of aged mice increased lymphocyte proliferation, IL-2
production and the DTH response (Meydani et al. 1986).
A high level of vitamin E in the diet (500 mg/kg food)
also increased natural killer cell activity of spleen cells
from old, but not young, mice (Meydani et al. 1988).
Dietary vitamin E promotes resistance to pathogens in
chickens, turkeys, mice, pigs, sheep and cattle (for refer-
ences see Meydani & Beharka, 1998; Han & Meydani,
1999); some of these studies report improved immune
cell functions in the animals receiving additional vitamin
E (Han & Meydani, 1999). Vitamin E prevented the
retrovirus-induced decrease in production of IL-2 and
interferon-g (IFN-g) by spleen lymphocytes and in natural
killer cell activity in mice (Wang et al. 1994). In another
study, young and old mice were fed diets containing
adequate (30 mg/kg diet) or high (500 mg/kg diet) levels
of vitamin E for 6 weeks and infected with influenza A
virus. Young or old mice fed the high level of vitamin E
had lower lung titres of virus than old mice fed the ade-
quate vitamin E diet (Hayek et al. 1997). The high level
of vitamin E caused increased production of IL-2 and
IFN-g by spleen lymphocytes from influenza-infected old
mice (Han et al. 1998). These observations suggest that
increasing vitamin E intake above habitual levels might
enhance immune function and improve resistance and
that vitamin E supplementation might be particularly
beneficial in the elderly.
Studies in man also suggest a relationship between vita-
min E supply and immune function. Canadian 3-year-olds
with the lowest serum vitamin E levels had the lowest lym-
phocyte proliferative responses and serum IgM concen-
trations (Vobecky et al. 1984). Chavance et al. (1989)
 Functional foods and immunity S171

found a positive association between plasma vitamin E

levels and DTH responses, and a negative association
between plasma vitamin E levels and incidence of infec-
tions, in healthy adults aged over sixty. Administration of
vitamin E to premature infants enhanced neutrophil
phagocytosis (Baehner et al. 1977; Chirico et al. 1983)
but decreased the ability of neutrophils to kill bacteria
(Baehner et al. 1977); this latter effect is most likely due
to a vitamin E-induced decrease in the production of free
radicals and related reactive species. Supplementation of
the diet of elderly subjects with 800 mg vitamin E/d for 4
weeks increased lymphocyte proliferation, IL-2 production
and the DTH response, but did not affect IL-1 production,
the number of CD4 cells or circulating Ig concentrations
(Meydani et al. 1990). In a more recent study, 60, 200 and
800 mg vitamin E/d increased the DTH response in elderly
subjects, with 200 mg/d having the maximal effect (Fig. 3;
Meydani et al. 1997). The 200 mg/d dose increased the
antibody responses to hepatitis B, tetanus toxoid and pneu-
mococci vaccinations (Fig. 3; Meydani et al. 1997). In
some cases the 800 mg vitamin E/d supplement decreased
the antibody response to below that of the placebo group
(Fig. 3). The authors concluded that 200 mg of vitamin E
daily represents the optimal level for the immune response.
Some studies report that high levels of vitamin E in the
human diet (. 300 mg/d) decrease the ability of neutrophils
to undergo phagocytosis (Boxer, 1986) and to kill bacteria
(Baehner et al. 1977; Prasad, 1980) and decrease monocyte
respiratory burst and IL-1b production (Devaraj et al.
1996).
Zinc. Zn deficiency in animals is associated with a
wide range of immune impairments (for reviews see
Fraker et al. 1993; Wellinghausen et al. 1997; Shankar & Fig. 3. Vitamin E and human immune function. Healthy elderly sub-
Prasad, 1998). Zn deficiency has a marked impact on jects supplemented their habitual diet with a placebo or with vitamin
bone marrow, decreasing the number of nucleated cells E at levels of 60, 200 or 800 mg/d. The delayed-type hypersensitiv-
and the number and proportion of cells which are lymphoid ity (DTH) response to seven recall antigens was assessed at base-
precursors (for reviews see Fraker et al. 1993; Fraker & line (p) and after 4 months (B). Subjects were then vaccinated
against hepatitis B (with subsequent boosters; A), diphtheria (B),
King, 1998). In patients with Zn deficiency related to tetanus toxoid (W) and three different pneumococci (PN6B (X),
sickle cell disease, natural killer cell activity is decreased, PN14 (K) and PN19F (O)); antibodies to the vaccines were
but can be returned to normal by Zn supplementation measured prior to vaccination and 3 months later. Plasma vitamin E
(Tapazoglou et al. 1985). In acrodermatitis enteropathica, concentrations (mM ) after 4 months supplementation were 23 (pla-
cebo), 38 (60 mg/d), 51 (200 mg/d) and 71 (800 mg/d). Vitamin E
which is characterised by reduced intestinal Zn absorption, intakes for adults in the UK vary between 35 and 195 mg/d (men)
thymic atrophy, impaired lymphocyte development, and 25 and 152 mg/d (women). (From Meydani et al. 1997.)
decreased numbers of CD4 cells and reduced lymphocyte
responsiveness and DTH are observed (Chandra & Dayton,
1982; Fraker et al. 1986). Moderate or mild Zn deficiency is considered a symptom of Zn deficiency. Malnourished,
or experimental Zn deficiency (induced by Zn consumption Zn-deficient children given Zn (2 mg/kg body weight
of , 35 mg/d; habitual intakes among adults in the UK daily for 10 days) had increased thymus size as judged
are 9 to 12 mg/d) in man results in decreased thymulin by radiography (Golden et al. 1977). Topical application
activity, decreased natural killer cell activity, a lowered of Zn to malnourished children improved the DTH
CD4:CD8, and decreased lymphocyte proliferation, response in the area of skin on which the application was
IL-2 production and DTH response; all can be corrected made (Golden et al. 1978). Zn administration (2 mg/kg
by Zn repletion (Shankar & Prasad, 1998). Experimental body weight daily) to malnourished children decreased
Zn deficiency in man decreased IL-2, IFN-g and TNF-a the incidence of diarrhoea by more than 50 %, decreased
production by mitogen-stimulated lymphocytes but did the incidence of respiratory and skin infections, and
not affect IL-4, IL-6 or IL-10 production by these cells resulted in threefold increased growth compared with chil-
or IL-1b production by lipopolysaccharide-stimulated dren given low-dose Zn (35 mg daily; Castillo-Duran et al.
cells (Beck et al. 1997). 1987). There are now a number of studies showing that Zn
Low plasma Zn levels predicted the subsequent develop- supplementation decreases the incidence of childhood diar-
ment of lower respiratory tract infections and diarrhoea rhoea and respiratory illness (for references see Scrimshaw
among Indian infants (Bahl et al. 1988). Indeed, diarrhoea & SanGiovanni, 1997; Shankar & Prasad, 1998; Calder &
S172
Jackson, 2000), although some studies fail to show benefit
of Zn supplementation in respiratory disease (for refer-
ences see Calder & Jackson, 2000). As well as decreasing
the risk of young infants developing diarrhoea (Rosado
et al. 1997), Zn supplementation (20 mg/d) to mal-
nourished children reduced diarrhoea-induced growth
faltering (Roy et al. 1999). Zn administration to preterm
low-birth-weight infants (1 mg/kg daily for 30 days)
increased the number of circulating T lymphocytes and
lymphocyte proliferation (Chandra, 1991). Providing
5 mg Zn/d to low-birth-weight, small-for-gestational-age
infants for 6 months increased measures of cell-mediated
immune function and decreased the incidence of gastro-
intestinal and upper respiratory tract infections (Lira et al.
1998); a Zn dose of 1 mg/d was without effect.
As observed for vitamin E, excessive Zn intakes impair
immune responses. For example, giving 300 mg Zn/d for 6
weeks to young adult human subjects decreased lympho-
cyte and phagocyte function (Chandra, 1984). High Zn
intakes can result in Cu depletion, and Cu deficiency
impairs immune function (for reviews see Prohaska &
Failla, 1993; Failla & Hopkins, 1998).
Probiotics: functional food components that
impact on immune function
Indigenous bacteria are believed to contribute to the
immunological protection of the host by creating a barrier
against colonisation by pathogenic bacteria. This barrier
can be disrupted by disease and by use of antibiotics, so
allowing easier access of the host gut by pathogens. It is
now believed that this barrier can be maintained by provid-
ing supplements containing live desirable bacteria: such
Fig. 4. Potential roles of probiotic bacteria in the human intestinal tract. Probiotic bacteria may act in a variety of ways to prevent the growth
and colonisation of pathogenic bacteria.
P. C. Calder and S. Kew
supplements are called probiotics (for extensive reviews
see Goldin, 1998; Naidu et al. 1999). Probiotic organisms
are found in fermented foods, including traditionally cul-
tured dairy products and newer kinds of fermented milks.
The organisms included in commercial probiotics include
lactic acid bacteria (Lactobacillus acidophilus, Lacto-
bacillus casei, Enterococcus faecium ) and Bifidobacterium
spp. These organisms colonise the gut only temporarily and
so their regular consumption is necessary. In addition to
creating a barrier effect, some of the metabolic products
of probiotic bacteria (e.g. lactic acid and a class of anti-
biotic proteins termed bacteriocins, produced by some bac-
teria) may inhibit growth of pathogenic organisms (Fig. 4).
Also, the desirable bacteria may compete for nutrients with
the pathogens (Fig. 4). Finally, there is some evidence that
probiotic bacteria may enhance the gut immune response
against pathogenic bacteria (Fig. 4).
Studies in rats and mice reveal that lactic acid bacteria
administered orally increase the numbers of T lympho-
cytes, CD4 cells and antibody-secreting cells, including
those in the intestinal mucosa, and enhance lymphocyte
proliferation, natural killer cell activity, IL-1, TNF and
IFN-g production, antibody production (including
secretory IgA), phagocytic activity and the respiratory
burst of macrophages and the DTH response (for a
review see Naidu et al. 1999). However, not all strains of
lactic acid bacteria are equally effective (Naidu et al.
1999). Animal studies also show that orally administered
lactic acid bacteria protect against challenges with patho-
genic bacteria such as Salmonella typhimurium, reverse
some of the immunosuppressive effects of malnutrition
and cause the symptoms of enterocolitis to be less
severe (Naidu et al. 1999). Co-colonisation of rats with
 Functional foods and immunity
Fig. 5. Potential patterns of relationship between nutrient supply
and immune function. All patterns assume that a deficiency of the
nutrient impairs the immune response. In (a) the recommended
intake of the nutrient is an intake at which the maximal immune
response occurs and intakes somewhat above the recommended
intake do not impair immune function. In (b) the recommended
intake is below the intake at which the maximal immune response
occurs and intakes somewhat above the recommended intake do
not impair immune function. In (c) the recommended intake of the
nutrient is an intake at which the maximal immune response occurs
and intakes somewhat above the recommended intake impair
immune function. In (d) the recommended intake is below the intake
at which the maximal immune response occurs and intakes
somewhat above the recommended intake impair immune function.
Lactobacillus plantarum and Escherichia coli resulted in
higher circulating concentrations of total IgA and of E.
coli-specific IgA and IgM than if the rats were colonised
with E. coli alone; there was also increased expression of
the IL-2 receptor in the lamina propria (Herias et al. 1999).
Despite the extensive animal studies, the effects of
probiotic bacteria on human immune function are still
controversial (Naidu et al. 1999). However, several studies
in man have reported an enhancement in markers of
immune function following probiotic administration.
Fig. 6. Different components of the immune system may respond to intakes of a
particular nutrient in different ways.
S173
Healthy Japanese children consuming a probiotic formula
containing bifidobacteria had increased faecal levels of
total and anti-poliovirus IgA than prior to taking the
formula (Fukushima et al. 1998). Healthy adults consum-
ing probiotic bacteria showed enhanced phagocytosis by
neutrophils and monocytes (Schiffrin et al. 1997; Yoon
et al. 1999). Adults fed fermented milk containing lacto-
bacilli and bifidobacteria and exposed to S. typhimurium
showed increased total and S. typhimuruim-specific IgA
concentrations in their serum (Link-Amster et al. 1994).
Administration of probiotic bacteria reduced the incidence
and severity of diarrhoea in children attending daycare
centres and the duration of diarrhoea in infants hospitalised
with gastroenteritis; in some studies this was associated
with an increase in levels of IgG, IgA and IgM and in
anti-rotavirus IgA secreting cells in the bloodstream (for
references see Naidu et al. 1999). Probiotic bacteria have
also been shown to reduce the incidence of antibiotic-
induced diarrhoea in children (Arvola et al. 1999;
Vanderhoof et al. 1999). Although some studies have
shown that consumption of probiotics can protect against
travellers diarrhoea, other studies do not demonstrate
such protection (for references see Naidu et al. 1999).
The immune system as a target for functional foods?
As described earlier, deficiency of total energy or of one or
more essential nutrients impairs immune function and
increases susceptibility to infectious pathogens. This is
most likely because these nutrients are involved in the
molecular and cellular responses to challenge of the
immune system. Providing these nutrients to deficient
individuals restores immune function and improves resist-
ance to infection. For some nutrients (e.g. vitamin E) the
dietary intakes that result in greatest enhancement of
immune function are greater than recommended intakes.
However, excess intake of some nutrients (e.g. vitamin
E, Zn) also impairs immune responses. Thus, four potential
general relationships appear to exist between the intake of
a nutrient and immune function (Fig. 5). These different
types of relationship might in part reflect interactions
between nutrients such that an excess of one nutrient
negatively affects the status of a second nutrient (e.g.
S174 P. C. Calder and S. Kew
Zn and Cu). There are likely to be interactions between
similar classes of nutrients (e.g. n 2 6 and n 2 3
polyunsaturated fatty acids) that have yet to be unravelled
fully, and there are most likely interactions between nutri-
ents which contribute to oxidative stress (e.g. polyunsatur-
ated fatty acids) and those which protect against it (e.g.
vitamin E).
It is often assumed when defining the relationship
between nutrient intake and immune function that all
components of the immune system will respond in the
same dose-dependent fashion to a given nutrient. This is
not correct, at least as far as some nutrients are concerned,
and it appears likely that different components of the
immune system show an individual dose response
relationship to the availability of a given nutrient (Fig. 6).
Indeed, some immune functions might be relatively
insensitive to nutrient supply.
Some consider that the immune system does not
respond optimally to challenge in apparently healthy,
free-living human individuals even if they are not
deficient in any single nutrient. Individuals may be mar-
ginally deficient in one or more nutrients or they may be
consuming some nutrients (e.g. fat) in excess. There is
now much interest in optimising the immune response
in such individuals not simply by correcting marginal
deficiencies but by increasing the intake of certain nutri-
ents and probiotics. At this stage the immuno-enhancing
effect of such supplementation is generally unproven.
Indeed, as far as micronutrients are concerned, such
supplementation might even be dangerous. Before the
immune system can be considered a genuine target for
functional foods, more needs to be known about the
role of variation in the immune response among appar-
ently healthy individuals in determining their suscepti-
bility to infection; about the influence of altered supply
of specific nutrients and of nutrient combinations on
aspects of immune function in different populations;
about the impact of genotype, gender, age and early
life experiences on immune function and on determining
the sensitivity of the immune system to nutrients; and,
most importantly, about whether enhanced immune
responses really translate into increased resistance to
infection.
Acknowledgements
The experimental results shown in Table 1 are taken from a
research project supported by the Ministry of Agriculture
Fisheries and Food, the Biotechnology and Biological
Sciences Research Council, Hoffmann-La Roche and Uni-
lever Research Vlaardingen under the Agri-Food LINK
Program (Grant Number AFQ111). The authors are
indebted to Miss T. Banerjee and to collaborators at the
Hugh Sinclair Unit of Human Nutrition, School of Food
Biosciences, University of Reading for their role in this
research.
References
Arvola T, Laiho K, Torkkeli S, Mykkanen H, Salminen S,
Maunula L & Isolauri E (1999) Prophylactic Lactobacillus
GG reduces antibiotic-associated diarrhea in children with
respiratory infections: a randomised study. Pediatrics 104,
L1 L4.
Baehner RL, Boxer LA, Allen JM & Davis J (1977) Auto-
oxidation as a basis for altered function by polymorphonuclear
leukocytes. Blood 50, 327 335.
Bahl R, Bhandari N, Hambridge KM & Bhan MK (1988) Plasma
zinc as a predictor of diarrhoeal and respiratory morbidity in
children in an urban slum setting. American Journal of Clinical
Nutrition 68, 414S 417S.
Beach RS, Gershwin ME & Hurley LS (1982) Gestational zinc
deprivation in mice: persistence of immunodeficiency for
three generations. Science 218, 469 471.
Beck FWJ, Prasad AS, Kaplan J, Fitzgeral JT & Brewer GJ
(1997) Changes in cytokine production and T cell sub-
populations in experimentally induced zinc deficient humans.
American Journal of Physiology 272, E1002 E1007.
Bendich A (1993) Vitamin E and human immune functions. In
Nutrition and Immunology, pp. 217 228 [DM Klurfeld,
editor]. New York/London: Plenum Press.
Boxer LA (1986) Regulation of phagocyte function by a-toco-
pherol. Proceedings of the Nutrition Society 45, 333 344.
Calder PC & Jackson AA (2000) Undernutrition, infection and
immune function. Nutrition Research Reviews 13, 3 29.
Calder PC & Yaqoob P (2000) The level of protein and type of fat
in the diet of pregnant rats both affect lymphocyte function in
the offspring. Nutrition Research 20, 995 1005.
Castillo-Duran C, Heresi G, Fisberg M & Uauy R (1987)
Controlled trial of zinc supplementation during recovery from
malnutrition: effects on growth and immune function.
American Journal of Clinical Nutrition 45, 602 608.
Chandra RK (1984) Excessive intake of zinc impairs immune
responses. Journal of the American Medical Association 252,
1443 1446.
Chandra RK (1991) 1990 McCollum Award Lecture. Nutrition
and immunity: lessons from the past and new insights into
the future. American Journal of Clinical Nutrition 53,
1087 1101.
Chandra RK & Dayton DH (1982) Trace element regulation of
immunity and infection. Nutrition Research 2, 721 733.
Chavance M, Herbeth B, Fournier C, Janot C & Vernhes G (1989)
Vitamin status, immunity and infections in an elderly popu-
lation. European Journal of Clinical Nutrition 43, 827 835.
Chirico G, Marconi M, Colombo A, Chiara A, Rondini G &
Ugazio A (1983) Deficiency of neutrophil phagocytosis in
premature infants: effect of vitamin E supplementation. Acta
Pediatrica Scandanavica 72, 521524.
Corwin LM & Schloss J (1980) Role of antioxidants on the
stimulation of the mitogen response. Journal of Nutrition
110, 2497 2505.
Cunningham-Rundles S (1998) Analytical methods for evaluation
of immune response in nutrient intervention. Nutrition Reviews
56, S27 S37.
Devaraj S, Li D & Jialal I (1996) The effects of a-tocopherol
supplementation on monocyte function. Journal of Clinical
Investigation 98, 756 763.
Failla ML & Hopkins RG (1998) Is low copper status
immunosuppressive? Nutrition Reviews 56, S59 S64.
Fraker PJ, Gershwin ME, Good RA & Prasad A (1986) Inter-
relationships between zinc and immune function. Federation
Proceedings 45, 1474 1479.
Fraker P & King L (1998) Changes in regulation of lymphpoiesis
and myelopoiesis in the zinc-deficient mouse. Nutrition
Reviews 56, S65 S69.
Fraker PJ, King LE, Garvy BA & Medina CA (1993) The
immunopathology of zinc deficiency in humans and rodents:
 Functional foods and immunity
a possible role for programmed cell death. In Nutrition and
Immunology, pp. 267 283 [DM Klurfeld, editor]. New York/
London: Plenum Press.
Fukushima Y, Kawata Y, Hara H, Terada A & Mitsuoka T (1998)
Effect of probiotic formula on intestinal immunoglobulin A
production in healthy children. International Journal of Food
Microbiology 42, 39 44.
Golden MHN, Golden BE, Harland PSEG & Jackson AA (1978)
Zinc and immunocompetence in protein energy malnutrition.
Lancet i, 1226 1228.
Golden MHN, Jackson AA & Golden BE (1977) Effect of zinc on
thymus of recently malnourished children. Lancet ii,
10571059.
Goldin BR (1998) Health benefits of probiotics. British Journal of
Nutrition 80, S203 S207.
Gross RL & Newberne PM (1980) Role of nutrition in
immunologic function. Physiological Reviews 60, 188 302.
Han SN & Meydani SN (1999) Vitamin E and infectious disease
in the aged. Proceedings of the Nutrition Society 58, 697 705.
Han SN, Wu D, Smith DE, Beharka A, Wang H, Bender BS &
Meydani SN (1998) Vitamin E supplementation increases
splenocyte IL-2 and IFN-g production in old mice infected
with influenza virus. FASEB Journal 12, A819.
Hayek MG, Taylor SF, Bender BS, Han SN, Meydani M, Smith
DE, Eghtesada S & Meydani SN (1997) Vitamin E supple-
mentation decreases lung virus titers in mice infected with
influenza. Journal of Infectious Diseases 176, 273 276.
Herias MV, Hessle C, Telemo E, Midtvedt T, Hanson LA &
Wold AE (1999) Immunomodulatory effects of Lactobacillus
plantarum colonizing the intestine of gnotobiotic rats. Clinical
and Experimental Immunology 116, 283 290.
Kuvibidila S, Yu L, Ode D & Warrier RP (1993) The immune
response in protein energy malnutrition and single nutrient
deficiencies. In Nutrition and Immunology, pp. 121 155 [DM
Klurfeld, editor]. New York/London: Plenum Press.
Link-Amster H, Rochat F, Saudan KY, Mignot O & Aeschlimann
JM (1994) Modulation of a specific humoral immune response
and changes in intestinal flora mediated through fermented
milk intake. FEMS Immunology and Medical Microbiology
10, 55 64.
Lira PI, Ashworth A & Morris SS (1998) Effect of zinc
supplementation on the morbidity, immune function and
growth of low birth weight full-term infants in northeast
Brazil. American Journal of Clinical Nutrition 69, 418S 424S.
McDade TW, Beck MA, Kuzawa C & Adair SS (2001) Prenatal
undernutrition, postnatal environments, and antibody response
to vaccination in adolescence. American Journal of Clinical
Nutrition 74, 543 548.
Meydani SN, Barklund MP, Liu S, Meydani M, Miller RA,
Cannon JG, Morrow FD, Rocklin R & Blumberg JB (1990)
Vitamin E supplementation enhances cell-mediated immunity
in healthy elderly subjects. American Journal of Clinical
Nutrition 52, 557 563.
Meydani SN & Beharka AA (1998) Recent developments in
vitamin E and immune response. Nutrition Reviews 56,
S49 S58.
Meydani SN, Endres S, Woods MM, Goldin BR, Soo C, Morrill-
Labrode A, Dinarello C & Gorbach SL (1991) Oral (n 2 3)
fatty acid supplementation suppresses cytokine production
and lymphocyte proliferation: comparison between young and
older women. Journal of Nutrition 121, 547 555.
Meydani SN, Meydani M, Blumberg JB, Leka LS, Siber G,
Loszewski R, Thompson C, Pedrosa MC, Diamond RD &
Stollar BD (1997) Vitamin E supplementation and in vivo
immune response in healthy subjects. Journal of the American
Medical Association 277, 1380 1386.
Meydani SN, Meydani M, Verdon CP, Shapiro AC, Blumberg JB
S175
& Hayes KC (1986) Vitamin E supplementation suppresses
prostaglandin E2 synthesis and enhances the immune response
in aged mice. Mechanisms of Ageing and Development 34,
191 201.
Meydani SN, Yogeeswaran G, Liu S, Baskar S & Meydani M
(1988) Fish oil and tocopherol-induced changes in natural
killer cell-mediated cytotoxicity and PGE2 synthesis in old
and young mice. Journal of Nutrition 118, 1245 1252.
Moore SE, Cole TJ, Poskitt EME, Sonko BJ, Whitehead RG,
McGregor IA & Prentice AM (1997) Season of birth predicts
mortality in rural Gambia. Nature 338, 434.
Naidu AS, Bidlack WR & Clemens RA (1999) Probiotic spectra
of lactic acid bacteria (LAB). Critical Reviews in Food Science
and Nutrition 38, 13 126.
Prasad JS (1980) Effect of vitamin E supplementation on leukocyte
function. American Journal of Clinical Nutrition 33, 606608.
Prohaska JR & Failla ML (1993) Copper and immunity. In
Nutrition and Immunology, pp. 309 332 [DM Klurfeld,
editor]. New York/London: Plenum Press.
Rayon JI, Carver JD, Wyble LE, Wiener D, Dickey SS, Benford
VJ, Chen LT & Lim DV (1997) The fatty acid composition of
maternal diet affects lung prostaglandin E2 levels and survival
from Group B Streptococcal sepsis in neonatal rat pups.
Journal of Nutrition 127, 1989 1992.
Rosado JL, Lopez P, Munoz E, Martinez H & Allen LH (1997)
Zinc supplementation reduced morbidity, but neither zinc nor
iron supplementation affected growth or body composition of
Mexican pre-schoolers. American Journal of Clinical Nutrition
65, 13 19.
Roy SK, Tomkins AM, Haider R, Behra RH, Akramuzzaman SM,
Mahalanabis D & Fuchs GJ (1999) Impact of zinc supple-
mentation on subsequent growth and morbidity in Bangladeshi
children with acute diarrhoea. European Journal of Clinical
Nutrition 53, 529534.
Schiffrin EJ, Brassart D, Servin AL, Rochat F & Donnet-Hughes
A (1997) Immune modulation of blood leukocytes in humans
by lactic acid bacteria: criteria for strain selection. American
Journal of Clinical Nutrition 66, S515 S520.
Scrimshaw NS & SanGiovanni JP (1997) Synergism of nutrition,
infection and immunity: an overview. American Journal of
Clinical Nutrition 66, 464S 477S.
Shankar AH & Prasad AS (1998) Zinc and immune function: the
biological basis of altered resistance to infection. American
Journal of Clinical Nutrition 68, 447S 463S.
Tapazoglou E, Prasad AS, Hill G, Brewer GJ & Kaplan J (1985)
Decreased natural killer cell activity in patients with zinc
deficiency and sickle cell disease. Journal of Laboratory and
Clinical Medicine 15, 19 22.
Thies F, Miles EA, Nebe-von-Caron G, Powell JR, Hurst TL,
Newsholme EA & Calder PC (2001a) Influence of dietary sup-
plementation with long chain n 2 3 or n 2 6 polyunsaturated
fatty acids on blood inflammatory cell populations and func-
tions and on plasma soluble adhesion molecules in healthy
adults. Lipids 36, 1183 1193.
Thies F, Nebe-von-Caron G, Powell JR, Yaqoob P,
Newsholme EA & Calder PC (2001b) Dietary supplemen-
tation with g-linolenic acid or fish oil decreases T lymphocyte
proliferation in healthy older humans. Journal of Nutrition 131,
1918 1927.
Vanderhoof JA, Whitney DB, Antonson DL, Hanner TL, Lupo JV
& Young RJ (1999) Lactobacillus GG in the prevention of anti-
biotic-associated diarrhea in children. Journal of Pediatrics
135, 564 568.
Vobecky JS, Vobecky J, Shapcott D & Rola-Pleszczynski M
(1984) Nutritional influences on humoral and cell-mediated
immunity in healthy infants. Journal of the American College
of Nutrition 3, 265.
S176 P. C. Calder and S. Kew
Wang Y, Huang DS, Eskelson CD & Watson RR (1994)
Long-term dietary vitamin E retards development of retro-
virus-induced dysregulation in cytokine production. Clinical
Immunology and Immunopathology 72, 70 75.
Wellinghausen N, Kirchner H & Rink L (1997) The immuno-
biology of zinc. Immunology Today 18, 519 521.
Yaqoob P, Newsholme EA & Calder PC (1999) Comparison of
cytokine production in cultures of whole human blood and
purified mononuclear cells. Cytokine 11, 600 605.
Yaqoob P, Pala HS, Cortina-Borja M, Newsholme EA &
Calder PC (2000) Encapsulated fish oil enriched in a-
tocopherol alters plasma phospholipid and mononuclear
cell fatty acid compositions but not mononuclear cell
functions. European Journal of Clinical Investigation 30,
260 274.
Yoon H, Dubarry M, Bouley C, Meredith C, Portier A, Tome D,
Renevot O, Blachon JL, Dugas B, Drewitt P & Postaire E
(1999) New insights in the validation of systemic biomarkers
for the evaluation of the immunoregulatory properties of milk
fermented with yoghurt culture and Lactobacillus casei
(Actimel (R)): a prospective trial. International Journal of
Immunotherapy 15, 79 89.
S160 B. O. Schneeman
The sensory, macronutrient composition and physical
properties of the diet modulate the response to a meal
during the cephalic, gastric and intestinal phases of
digestion. In the cephalic phase the sight, smell and taste
of foods stimulate the secretion of digestive juices into
the mouth, stomach and intestine, essentially preparing
these organs to digest the foods to be consumed. Experi-
ments, in which animals are sham-fed so that food con-
sumed does not actually enter the stomach or intestine,
demonstrate that the cephalic phase accounts for about
20 % or more of the secretion into the gut. The gastric
and intestinal phases occur when food and its components
are in direct contact with the stomach or intestine, respect-
ively. During these phases the distension of the organs with
food, as well as the specific composition of the food, can
stimulate a GIT response. The GIT is the richest endocrine
organ in the body, containing a large array of peptides;
however, the exact physiological function of each of
these compounds has not yet been established. Five
peptides, gastrin, cholecystokinin (CCK), secretin, gastric
inhibitory peptide and motilin, are established as regulatory
hormones in the GIT and several factors affect their release
and subsequent actions. Among the established gastro-
intestinal peptides, secretin stimulates secretion of fluid
and bicarbonate from the pancreas, gastrin stimulates
secretion in the stomach, gastric inhibitory peptide inhibits
gastric acid secretion and stimulates insulin release, and
motilin stimulates the motility of the upper GIT (Bloom,
1983). In addition to the various factors causing release
of these hormones and the response to them, physiologists
are also interested in the interactions among hormones, as
well as with the nervous system, since the response to a
meal involves release of many factors. Of these established
hormones, CCK provides a useful example to examine
some implications of meal-induced responses.
CCK is released from cells of the upper small intestine
into the blood. Dietary protein and fat stimulate its release
from the intestine, while digestible carbohydrates have no
effect on release of the hormone. Gastric acid inhibits
CCK secretion. Once released CCK can inhibit gastric
emptying, stimulate secretion of acid in the stomach and
pancreatic juice into the small intestine, and stimulate
contraction of the gallbladder, which releases bile into
the intestine. In addition, it stimulates motility and
growth in the GIT. Newer identified functions of CCK
include inhibition of food intake and stimulation of insulin
release. Thus CCK is a hormone that prepares the GIT to
digest food by stimulating secretion, regulates the rate of
digestion and absorption by controlling gastric emptying,
helps signal the termination of eating once food is in the
intestine, and helps facilitate the storage of energy con-
sumed in the meal. Only recently have reliable immuno-
assays become available for CCK that allow investigation
of typical meal-stimulated CCK responses. Test meals
that provide about 25 % energy from fat stimulate a peak
increase in CCK concentration of about 3 6 pmol/l
(Bourdon et al. 1999, 2001). Increasing the amount of fat
in an infusion into the intestine of rats increases peak
CCK response as well as the length of time that plasma
CCK concentrations are elevated above baseline. Including
a source of b-glucan, a viscous polysaccharide, in a test
meal does not change the peak concentration of CCK
released after a meal but prolongs the time that CCK is
significantly elevated above baseline (Bourdon et al.
1999). When cooked dry beans are included in the test
meals, the CCK response is doubled compared with a
test meal that is low in fibre-rich foods (Bourdon et al.
2001). This effect of dry beans to enhance CCK release
may be related to their fibre content and to the presence
of trypsin inhibitor (Schwartz et al. 1994). In a subsequent
study we demonstrated in women that addition of fat or
fibre to a test meal that is relatively low in fat or fibre
will significantly (P , 0.05) increase the CCK response
to a test meal (Burton-Freeman et al. 2002). Thus while
protein and fat are the primary macronutrients that stimu-
late the release of CCK, the presence of non-digestible
polysaccharides appears to modulate the pattern of CCK
release. This meal-induced response of CCK has impli-
cations for understanding the potential effects of fibre on
insulin release and/or satiety (Holt et al. 1992; Rushakoff
et al. 1993; Burton-Freeman, 2000). The subjective
measures of satiety recorded by subjects after consuming
test meals were correlated with the CCK response. Thus
meals that elicit the highest CCK response are perceived
by human subjects as the most filling.
This example illustrates that the meal-induced response
causes secretion of digestive juices essential for the
digestion of food components; however, the response
also influences other metabolic responses to the meal.
The GIT is one of the richest endocrine organs in the
body, and CCK is just one example of a variety of
hormones released during the response to a meal (Bloom,
1983; Johnson, 1997). These hormones have metabolic
effects beyond the GIT. The system is simultaneously
preparing the GIT to digest and absorb the meal that has
been consumed in an efficient manner and also is signalling
short-term satiety so that feeding is terminated at an
appropriate point.
Digestive and absorptive functions of the
gastrointestinal tract
The GIT breaks down food into absorbable compounds
through mechanical and biochemical processes. Chewing
breaks food into smaller particles that can mix more readily
with the GIT secretions. In the mouth saliva lubricates the
food bolus so that it passes readily through the oesophagus
to the stomach. The sensory aspects of food stimulate the
flow of saliva, which not only lubricates the bolus of
food but also is protective and contains digestive enzymes.
Swallowing is regulated by sphincter actions to move the
bolus of food into the stomach. The motility of the stomach
continues the process of mixing food with the digestive
secretions, now including gastric juice, which contains
acid and some digestive enzymes. The action of the
stomach continues to break down food into smaller par-
ticles prior to passage to the intestine. The stomach,
which after a meal may contain over a litre of material,
regulates the rate of digestion by metering chyme into
the small intestine over several hours. Several factors can
slow the rate of gastric emptying; for example, solids
take longer to empty than liquids, mixtures relatively
 Gastrointestinal physiology
high in lipid take longer to empty, and viscous or thick
mixtures take longer to empty than watery, liquid contents.
Once digesta is in the small intestine, peristaltic motor
activity propels it along the length of the intestine and
segmentation allows mixing with digestive juices in the
intestine, which include pancreatic enzymes, bile acids
and sloughed intestinal cells. Digestion of macronutrients,
which began in the mouth, continues in the small intestine
primarily through the action of enzymes. Each of the
macronutrients has a unique set of enzymes that break
the macromolecules into sub-units that can be taken up
by the absorptive systems in the intestinal cells. The
phytochemicals that have been investigated for lowering
risk of chronic disease are part of the cell matrix in plant
foods and depend on the mechanical and biochemical
disruption of the food matrix to become available for
absorption from the GIT. Thus factors that alter the rate
and site of digestion and absorption will alter their bio-
availability as well. To illustrate how the GIT functions
of digestion and absorption are important for bio-
availability and subsequent metabolism, a useful example
is to consider some of the factors that affect carbohydrate
digestion and utilization.
Carbohydrates are categorized as digestible or non-
digestible. Digestible carbohydrates are the various sugar-
containing molecules that can be digested by amylase or
the saccharidases of the small intestine to sugars that can
be absorbed from the intestine. The predominant digestible
carbohydrates in foods are starch, sucrose, lactose (milk
sugar) and maltose. Foods may also contain simple
sugars such as glucose or fructose that do not need to be
digested before absorption from the gut. a-Amylase,
which hydrolyses the a(1 4) linkages in starch, is secreted
in the mouth from salivary glands and from the pancreas
into the small intestine. The action of amylase produces
smaller carbohydrate segments that can be hydrolysed
further to sugars by enzymes at the brush border of the
intestinal cells. This hydrolysis step is closely linked with
absorption of sugars into the intestinal cells.
The glucose absorbed from digestible carbohydrate
stimulates the release of insulin. In healthy, normal-
weight individuals, the postprandial increase in blood
insulin concentration is correlated with the intake of carbo-
hydrate (Lee & Wolever, 1998; Lu et al. 2000). Insulin is
important for regulation of energy metabolism in the body
and its release after a meal promotes energy storage. Thus,
in healthy individuals, the postprandial release of insulin
results in glucose and amino acid uptake by cells and
stimulates glycolysis and synthesis of glycogen, protein
and fatty acids. The rate and pattern of carbohydrate
digestion and absorption will influence the appearance of
glucose in blood and stimulation of insulin secretion.
Consequently there is considerable interest in how factors
that influence carbohydrate digestion and absorption
might affect energy metabolism during the postprandial
period. Slowing gastric emptying, which slows entry of
digestible carbohydrate into the small intestine, blunts
the glycaemic and insulinaemic response to a meal. The
hormone CCK, discussed above, also modulates the
release of insulin. In healthy individuals dietary factors
that enhance CCK reduce insulin and glucose response
S161
(Liddle et al. 1988; Rushakoff et al. 1993; Liddle, 2000).
CCK may function by stimulating an early release of
insulin before plasma glucose begins to increase. This
priming of the system results in an overall lower glycaemic
and insulinaemic response to a meal (Bloom, 1983). In
certain types of obesity, cells become insulin-resistant,
which results in elevated plasma levels of insulin, leading
to non-insulin dependent diabetes mellitus. To help
manage this condition, dietary factors such as viscous
polysaccharides or a-amylase inhibitors that can blunt
the postprandial rise in glucose have been investigated.
Non-digestible carbohydrates cannot be digested by the
enzymes in the small intestine and are the primary
component of dietary fibre. Most of these non-starch poly-
saccharides are part of the plant cell wall. The most
abundant polysaccharide in plant tissue is cellulose,
which is a glucose polymer with b(1 4) links between
the sugars. Amylase, the starch-digesting enzyme of the
small intestine, can hydrolyse only a linkages. The non-
digestible carbohydrates also include hemicelluloses,
pectins, gums, oligofructose and inulin. While non-
digestible, they do affect the digestive process because
they provide bulk in the intestinal contents, hold water,
can become viscous or thick in the intestinal contents,
and delay gastric emptying. Viscosity is correlated with
the ability of polysaccharides to lower plasma cholesterol
levels (Gallaher & Hassel, 1995; Carr et al. 1996; Gallaher
et al. 2000). Slowing gastric emptying is associated with
blunting of the glucose and insulin response to a meal as
well as enhancing feelings of satiety. In addition, non-
starch polysaccharides are the primary substrate for
growth of the micro-organisms in the large intestine and
contribute to stool formation and laxation. Products of
microbial action include ammonia, gas and short-chain
fatty acids (SCFA). SCFA are used by cells in the large
intestine for energy and some appear in the circulation
and can be used by other cells in the body for energy
as well. Thus, while dietary fibre is classified as non-
digestible carbohydrate, the eventual digestion of these
polysaccharides by microbes does provide energy to the
body. Current research is focused on the potential effect
of SCFA on the health of the intestine and their possible
role in the prevention of gastrointestinal diseases.
In summary, the rate at which macromolecules in food
are digested and absorbed influences metabolism in the
body. Additionally, physical properties of foods such as
viscosity will influence the bioavailability and utilization
of nutrients and other compounds in foods by the body.
The potential impact of the gastrointestinal
tract on health
The examples above illustrate that the GIT modulates
metabolic responses through meal-induced responses to
diet properties and composition, and through the rate of
digestion and absorption. These examples indicate that
the physiology and function of the GIT can influence risk
factors for non-communicable diseases and thus the GIT
has an important role in maintaining health. Research on
dietary fibre has illustrated that actions within the GIT
modify risk factors for non-communicable diseases since
S162 B. O. Schneeman
fibre is non-digestible and exerts its effects on metabolism
during its transit through the GIT. For example, the ability
of certain fibres to reduce plasma cholesterol is associated
with an increase in viscosity of digesta (Carr et al. 1996)
and with increased excretion of bile acids due to binding
or adsorption by fibre fractions (Buhman et al. 2000).
Non-starch polysaccharides are the primary substrates for
microflora in the large intestine.
The fact that fibre is utilized by the GIT microflora has
stimulated research on the potential impact of the
metabolism of this microflora on reducing the risk of dis-
ease and promoting health. Fermentable carbohydrates
increase the microbial mass in the large intestine, which
contributes to stool bulk and aids laxation (Cummings
et al. 1992; Chen et al. 1998). Fermentation also produces
SCFA, which provide energy from highly fermented poly-
saccharides. SCFA are utilized by the intestinal epithelial
cells, liver and muscle and have a trophic effect on the
intestinal epithelium. Of considerable research interest is
the potential role of butyrate in protecting the colon from
developing cancer because of its role in cell differentiation
and growth (Salminen et al. 1998). Cummings et al. (2001)
have summarized the effects of various substrates, includ-
ing inulin, oligofructose, resistant starch, polydextrose,
pectin and arabinogalactan, on SCFA production. While
each of these substrates can be fermented by the microflora
from the large intestine, it is more difficult to change
markedly the molar ratio of SCFA produced. The fer-
mentability of carbohydrates may also be associated with
metabolic enzymes, including cell signal transduction
pathways in colonic mucosal cells (Pajari et al. 2000)
and the enzymes associated with microbial metabolism of
mutagenic compounds (Cummings et al. 2001).
Approximately 1012 micro-organisms reside in the GIT,
most of which are in the large intestine. Some of these
microbes are considered harmful and have been associated
with intestinal diseases, while others are considered
beneficial and have been associated with decreasing risk
for chronic disease as well as the synthesis of vitamins,
facilitating mineral absorption and immune stimulation.
The microflora of the intestine is complex and diverse
and attempts to modify the composition by dietary means
have not been very effective. However, feeding cultured
products that contain certain strains of Bifidobacterium or
Lactobacillus have been reported to have metabolic effects
that impact health. Typically, the response is seen only
while the product containing a live culture is fed,
suggesting that the bacteria do not colonize the GIT
permanently but remain metabolically active during their
transit thorough the GIT. At the simplest level these
cultures are reported to improve lactose digestion in
lactose-intolerant individuals and, at a more complex
level, stimulate several facets of immune response
(Salminen et al. 1998; Gill et al. 2000). The functional
significance and sustainability of the immuno-stimulation
have not been reported. However, certain live cultures
have been reported to help ameliorate symptoms of
diarrhoea. The emerging research in this area suggests
that we still have a poor understanding of the significance
of the microbial population in the GIT and its role in
promoting health and reducing risk for disease.
Conclusions and need for future research
While the health benefit of a functional food may be a
metabolic response that lowers risk for disease, the
actual target for the food or food component may be
on the functioning of the GIT. This situation raises inter-
esting challenges for assessing the response to functional
foods and the validation of health claims. For example,
there may be a health benefit from slowing absorption
from the intestine; slower absorption might be measured
by examining the time course for appearance of the
nutrient or food component in the blood, measuring
the hormone response associated with absorption of the
compound, or measuring parameters related to excretion
of the compound. However, the food component may
slow absorption by delaying gastric emptying, altering
the mixing within the intestinal contents or decreasing
the availability of digestive enzymes in the intestine.
These measures of GIT function provide validation of
the mechanisms by which the functional food or food
component affects metabolism. Such validation is essen-
tial to demonstrate that the food or food component has
a specific effect on health or risk for disease. For
example, health claims for sources of dietary fibre and
reduced risk of cardiovascular disease are strengthened
by demonstrating the mechanisms by which fibre
lowers plasma cholesterol levels. Research is needed to
validate the links between various aspects of gastrointes-
tinal function and subsequent health effects of functional
foods.
Understanding the GIT response to foods and meals is
critical to understanding the metabolic effects of functional
foods. Three important research needs are evident from a
review of the function and physiology of the GIT in
relation to foods. A primary need is to determine the
bioavailability of physiologically active compounds
from foods. Bioavailability will be determined by the
digestibility of foods that contain these compounds, their
subsequent absorption and their utilization by tissues. For
certain compounds their mode of action will be in the
GIT and the focus will be on interaction with organs of
the GIT; however, in other cases, absorptive mechanisms
will be important in determining the metabolic effects of
functional foods. The physical structure of foods has
often been overlooked in understanding the metabolic
response to diet. However, this structure contributes to
the functional effects of foods as well as to the availability
of compounds from foods. For example, recent studies
have demonstrated that changing the viscosity of the gut
contents alters absorption and GIT response. The impli-
cations of these studies need to be investigated further.
Additionally, food structures such as the plant cell wall
change the availability of absorbable compounds along
the gastrointestinal contents and raise interesting research
questions about the importance of plant cell walls beyond
their content of dietary fibre. A third area of research
with emerging importance relative to the GIT concerns
the function of microflora along the intestines. Studies on
probiotics and prebiotics suggest that this microflora is
responsive to dietary influences. Understanding the
influence of the GIT microflora on health promotion and
 Gastrointestinal physiology
disease prevention is a complex but important research
agenda.
References
Bloom SR (1983) Endocrinology of nutrient entry. In Delaying
Absorption as a Therapeutic Principle in Metabolic Diseases,
pp. 19 27 [W Creutzfeldt and UR Folsch, editors]. Stuttgart/
New York: Georg Thieme Verlag.
Bourdon I, Olson B, Backus R, Richter D, Davis PA &
Schneeman BO (2001) Beans, as a source of dietary fiber,
increases cholecystokinin and apo B48 response to test meals
in men. Journal of Nutrition 131, 1485 1490.
Bourdon I, Yokoyama W, Davis PA, Hudson C, Backus R,
Richter BD, Knuckles B & Schneeman BO (1999) Postprandial
lipid, glucose, insulin, and cholecystokinin responses in men
fed barley pasta enriched with b-glucan. American Journal of
Clinical Nutrition 69, 55 63.
Buhman K, Furumoto EJ, Donkin SS & Story JA (2000) Dietary
psyllium increases expression of ileal apical sodium-dependent
bile acid transporter mRNA coordiantely with dose-responsive
changes in bile acid metabolism in rats. Journal of Nutrition
130, 2137 2142.
Burton-Freeman B (2000) Dietary fiber and energy regulation.
Journal of Nutrition 130, 272S 275S.
Burton-Freeman B, Davis P & Schneeman BO (2002) Plasma chole-
cystokinin is associated with subjective measures of satiety in
women. American Journal of Clinical Nutrition in press.
Carr T, Gallaher D, Yang CH & Hassel CA (1996) Increased
intestinal contents viscosity reduces cholesterol absorption
efficiency in hamsters fed hydroxypropyl methylcellulose.
Journal of Nutrition 126, 1463 1469.
Chen HL, Haack VS, Janecky CW, Vollendorf NW & Marlett JA
(1998) Mechanisms by which wheat bran and oat bran increase
stool weight in humans. American Journal of Clinical Nutrition
68, 711719.
Cummings JH, Bingham SA, Heaton KW & Eastwood MA
(1992) Fecal weight, colon cancer risk, and dietary intake of
non-starch polysaccharides (dietary fiber). Gastroenterology
103, 1783 1789.
Cummings JH, Macfarlane GT & Englyst HN (2001) Prebiotic
digestion and fermentation. American Journal of Clinical
Nutrition 73, Suppl. 2, 415S 420S.
Gallaher CM, Munion J, Hesslink R, Wise J & Gallaher DD
(2000) Cholesterol reduction by glucomannan and chitosan
is mediated by changes in cholesterol absorption and bile
acid and fat excretion in rats. Journal of Nutrition 130,
27532759.
S163
Gallaher DD & Hassel CA (1995) The role of viscosity in the
cholesterol-lowering effect of dietary fiber. In Dietary Fiber
in Health and Disease, pp. 106 114 [D Kritchevsky and C
Bonfield, editors]. New York: Plenum Press.
Gill HS, Rutherford KJ, Prasad J & Gopal PK (2000) Enhance-
ment of natural and acquired immunity by Lactobacillus
rhamnosus (HN001), Lactobacillus acidophilus (HN017) and
Bifidobacterium lactis (HN019). British Journal of Nutrition
83, 167 176.
Holt S, Brand J, Soveny C & Hansky J (1992) Relationship of
satiety to postprandial glycaemic, insulin and cholecystokinin
responses. Appetite 18, 129 141.
Johnson LR (1997) Gastrointestinal Physiology. St Louis, MO:
Mosby-Year Book, Inc.
Lee BM & Wolever TMS (1998) Effect of glucose, sucrose,
fructose on plasma glucose and insulin responses in normal
humans: comparison with white bread. European Journal of
Clinical Nutrition 52, 924 928.
Liddle RA (2000) Regulation of cholecystokinin secretion in
humans. Journal of Gastroenterology 35, 181 187.
Liddle RA, Rushakoff RJ, Morita ET, Beccaria L, Carter JC &
Goldfine ID (1988) Physiological role for cholecystokinin in
reducing postprandial hyperglycemia in humans. Journal of
Clinical Investigation 81, 1675 1681.
Lu ZX, Walker KZ, Muir JG, Mascara T & ODea K (2000)
Arabinoxylan fiber, a byproduct of wheat flour processing,
reduces the postprandial glucose response in normoglycemic
subjects. American Journal of Clinical Nutrition 71,
1123 1128.
Pajari AM, Oikarinen S, Grasten S & Mutanen M (2000) Diets
enriched with cereal brans or inulin modulate protein kinase
C activity and isozyme expression in rat colonic mucosa.
British Journal of Nutrition 84, 635 643.
Rushakoff RA, Goldfine ID, Beccaria LJ, Mathur A, Brand RJ &
Liddle RA (1993) Reduced postprandial cholecystokinin
(CCK) secretion in patients with noninsulin-dependent dia-
betes mellitus: evidence for a role for CCK in regulating post-
prandial hyperglycemia. Journal of Clinical Endocrinology
and Metabolism 76, 489 493.
Salminen S, Bouley C, Boutron-Ruault MC, Cummings JH,
Franck A, Gibson GR, Isolauri E, Moreau MC, Roberfroid
M & Rowland I (1998) Functional food science and gastro-
intestinal physiology and function. British Journal of Nutrition
80, Suppl. 1, S147 S171.
Schwartz JG, Guan D, Green GM & Phillips WT (1994)
Treatment with an oral proteinase inhibitor slows gastric
emptying and acutely reduces glucose and insulin levels after
a liquid meal in type II diabetic patients. Diabetes Care 17,
255 262.
S152
composition of the food supply (Leclercq et al. 2001).
Likewise, increasing evidence points to genetic vari-
ability as a significant determinant of the response to
foods and food components. While delineating the most
effective diet for an individual is a daunting task,
recent scientific discoveries as reviewed briefly below
reinforce the belief that a personalized dietary interven-
tion approach to health promotion and disease prevention
is feasible.
Functional foods: a health perspective
The intriguing term functional foods has arisen from a
general belief in the health benefits of foods. While this
term has no legal meaning, it nevertheless signifies a
proactive appreciation that some foods may confer health
benefits (Milner, 2000). While functional foods have
been defined in various ways, the International Life
Sciences Institute (ILSI) North America defines them as:
Foods, that by virtue of physiologically active food
components, provide health benefits beyond basic nutri-
tion. Classifying which foods fall within the health
benefits category and which do not has been an interesting
challenge for some, especially since many factors can
influence the overall response to a food including the
quantity consumed, accompanying foods ingested, duration
of exposure, physiological state, etc. At this point, it is best
to indicate that all foods are probably functional in some
capacity and that, under specific circumstances, some may
provide immediate and/or long-term benefits.
Widespread interest in functional foods continues to
abound in North America among scientists, legislators
and consumers. About 60 % of adults residing in the
United States are believed to select foods for health pur-
poses, regardless of their age or gender (http://www.IFIC.
org). While younger individuals may be selecting foods
for mental and physical performance, older individuals
appear to be selecting foods for their potential merits in
reducing disease risk or improving the quality of life.
Regardless of why foods are chosen, it is refreshing to
have a positive proactive belief, since during the past
few decades campaigns to alert people about diet and
health issues have largely been negative.
Market value of functional foods in the United States
Increasing consumer demand for healthful foods and
beverages, coupled with scientific discoveries about the
physiological consequences of selected foods and ingredi-
ents, has fostered the development of an active functional
foods and beverage market. Some of these products cer-
tainly extend beyond conventional fortified items because
they contain specific ingredients designed to have health
or structure function benefits for the consumer.
Statistics about the size of the US natural foods market
vary considerably depending on the source of information
and which products are included or excluded, as evident
by a recent report entitled The Natural/Organic Food
Market in the United States (http://atnriae.agr.ca/info/us/
e3164.htm#MARKET OVERVIEW). Many estimate that
natural and organic foods account for annual sales of
J. A. Milner
between $163 billion and $297 billion, which undeniably
represents a large variance. The market value of functional
foods is also largely ill defined. Frost & Sullivan (http://
www.food.frost.com) suggest the total market for func-
tional ingredients, functional foods, functional beverages,
dietary supplements and foods for special dietary use is
approximately $50 billion. Trade publications for the
food industry in 2000 indicated that sales of natural pro-
ducts including foods, beverages and supplements grew
by about 7 % over the previous year.
Health and structure function claims in North
America
The United States Federal Food, Drug and Cosmetic Act
does not provide a statutory definition of functional
foods. Thus, the Food and Drug Administration (FDA)
has no authority to establish a formal regulatory category
for such foods. The primary determinant of the regulatory
status of these foods is thus their intended use. The distinc-
tion between a food and a drug has been blurred because of
recent scientific findings suggesting that the former may
have some medicinal properties. Nevertheless, according
to the FDA, a drug is a substance that is used to prevent,
treat, cure or mitigate disease. Foods, on the other hand,
and according to Websters Dictionary, are defined as pro-
tein, carbohydrates, fats, minerals, vitamins and other sup-
plementary substances that nourish and sustain life.
According to the FDA, a food is defined as: Articles
used for food or drink for man or other animals, chewing
gum, and articles used for components of any such article.
In the United States, food claims can take a number of
approaches, i.e. those related to content, health or to struc-
ture function. In marketing these foods, manufacturers
may come under one of several existing regulatory options.
The first decision manufacturers will make that will help
determine their products regulatory status is whether the
product is a food or a drug. Thus, manufacturers and retail-
ers have a range of legal and regulatory categories into
which their products may be classified. Fifteen health
claims have been approved for manufacturers to describe
the relationship between a food substance and a disease
or health-related condition. Some of the broad areas
Table 1. Areas associated with health claims in the United States
Calcium and osteoporosis
Dietary lipids and cancer
Sodium and hypertension
Dietary saturated fat and cholesterol and the risk of CHD
Fibre-containing grain products, fruits and vegetables and cancer
Fruit, vegetables and grain products that contain fibre, particularly
soluble fibre, and the risk of CHD
Fruits and vegetables and cancer
Folate and neural tube defects (1996)
Dietary sugar alcohols and dental caries (1996)
Soluble fibre from certain foods and the risk of CHD (1997)
Soya protein and the risk of CHD (1999)
Stanols and heart disease
CHD, coronary heart disease.
 Diet and health
addressed by these health claims are presented in Table 1.
Health claims may arise in three ways. First, the 1990
Nutrition Labeling and Education Act allows the FDA to
authorize a health claim for a food or dietary supplement
based on the FDAs careful review of the scientific litera-
ture. Second, the 1997 Food and Drug Administration
Modernization Act allows a health claim to be made for
a food based on an authoritative statement of a scientific
body of the US government or the National Academy of
Sciences. And third, the 1999 court decision in the case
of Pearson vs. Shalala allows certain qualified health
claims to be used. The relatively small number of health
claims for foods that have been approved serves to
emphasize the incomplete or inconsistent information that
precludes firm conclusions.
Structure function claims have been allowed for foods
under the US Federal Food, Drug and Cosmetic Act, but
because they are allowed differently for dietary sup-
plements and conventional foods their use is stirring con-
siderable controversy in the USA. Some current uses and
controversies about structure function claims focus on
how best to identify an action of the food or bioactive
food component without simultaneously imply a health
or drug outcome. Thus, structure function statements
are often vague and imply the food or components helps
maintain normalcy, whatever that might signify.
Historically, the Food Regulations under the Canadian
Food and Drugs Act were developed with the objectives
to ensure that food that was sold was safe and nutritious
and for the prevention of fraud. The recognition of the
health effects of various food constituents has sparked
legislative interest in functional foods. In Canada foods
containing the beneficial ingredients, whether naturally
occurring or as a result of addition of an isolated
component, are termed functional foods. The proposed
Health Canada definition of a functional food (http://
www.hc-sc.gc.ca/main/drugs/zfiles/english/ffn/ffdscdoc_e.
html) is one that is: similar in appearance to, or may be, a
conventional food, is consumed as part of a usual diet, and
is demonstrated to have physiological benefits and/or
reduce the risk of chronic disease beyond basic nutritional
functions. Currently a proposal is pending that would
allow health claims in Canada.
Regardless of the claims that are allowed, consumers are
eager to learn more truthful information about food, food
components and health. To assist in clarifying the issue,
ILSI North Americas Technical Committee on Food
Components for Health Promotion developed in 1997 a
Road Map as a strategy to Improve the Health of the
Public through Consumer Acceptance of Safe Food
Products that Provide Significant Health Benefits. A
number of approaches were identified to accomplish this
goal including: the creation of a comprehensive science
base, the promotion of public trust, the development of
consumer preferred functional foods, the optimization of
a regulatory framework and the creation of marketplace
incentives to develop functional foods. Because of the
scientific and legal mandates involving organizations
such as ILSI, it was apparent that other groups were
needed if progress was truly going to be made. Delight-
fully, progress has occurred during the past four years in
S153
not only showcasing this area but also with building a
number of important and effective linkages. Undeniably
the database regarding foods and their components as
factors in health has continued to blossom during this
period. Likewise, new and exciting linkages have surfaced
with various groups with a commitment to research and
outreach about foods and health. The turn of the twenty-
first century was a point to not only reflect on what had
been accomplished by the Road Map, but to rekindle inter-
est in foods and bioactive components as a continuing
focus. While the exact plan remains to be resolved, key
components will probably be to stimulate fundamental
science to evaluate the health effects of functional foods,
develop a framework to utilize science for decision
making and optimize the scientific impact of these
undertakings through key collaborations.
Numerous bioactive food components may be involved
Numerous foods are already associated with health pro-
motion and disease prevention. The diversity of these
foods suggests a variety of components may be involved.
It has been estimated that about 25 000 different chemical
compounds occur in fruits, vegetables and other plants
eaten by man. To date, more than 500 compounds have
been identified as potential modifiers of the cancer process.
Both essential and non-essential allelochemicals occurring
in plants, as well as zoochemicals found in animal pro-
ducts, may be physiologically important modifiers of a
host of biological processes. Compounds encompassing
such diverse categories as carotenoids, dithiolthiones,
flavonoids, glucosinolates, isothiocyanates, allyl sulf-
hydryls and fermentable fibres have been found to influ-
ence a variety of cellular processes that would be
expected to influence health (Hasler et al. 2000; Milner,
2000; Prosky, 2000; Go et al. 2001). Many have been
reported to modify the redox status of cells and therefore
may have far-reaching implications as determinants of
health and well-being (Diplock et al. 1998; Clarkson &
Thompson, 2000; Roberfroid, 2000). Numerous reviews
extolling the merits and/or possible risks of numerous bio-
active food components have been presented in recent
years (Potter, 1997; Abdulla & Gruber, 2000; Gill &
Cross, 2000; Milner, 2000).
Moving beyond observational studies
While observations about dietary bioactive food com-
ponents and health are exceedingly tantalizing, the future
of nutrition research probably resides in the ability to
move beyond these studies to a more probing and molecu-
lar approach that will allow for tailored recommendations
to individuals (Greenwald & Milner, 2001). DellaPenna
(1999) coined the term nutritional genomics to describe
work at the interface of plant biochemistry, genomics and
human nutrition aimed at understanding and manipulating
nutrient reactions and interactions at the molecular or
genomic level. For the purposes of this review, nutritional
genomics refers to the study of any genetic or epigenetic
interaction with a nutrient that leads to phenotypic changes.
Unquestionably, the study of nutritional genomics offers
S154 J. A. Milner
the potential to identify definitively which components in
foods influence growth, development and health, and to
clarify their specific mechanisms of action. About 26 000
to 38 000 genes were proposed in the first draft of the
human genome. While this is only about double the
number found in the fruit fly and worm, it points to
the importance of the expression of these genes and to
their regulation (Paabo, 2001). A web site (http://www.
cgap.nci.nih.gov/), developed jointly by the National
Cancer Institute and the National Library of Medicine,
was part of the recently launched Cancer Genome Anat-
omy Project (CGAP). The database offers scientists a
powerful new tool to study how various factors including
dietary components might alter a host of cellular events.
Queries to the CGAP database are examined quickly
through a gene index system that a few years ago might
have taken years or even lifetimes to compile. Genomic
data for man and mouse, including expressed sequence
tags, gene expression patterns, single nuclear polymorph-
isms, cluster assemblies and cytogenetic information, are
included. In addition to genetic information, this web site
contains informatics tools to query and analyse the data
and information on methods and resources for reagents
developed by the project. All CGAP resources including
cDNA libraries, clones and sequence data are publicly
accessible to scientists.
Rather compelling evidence already exists revealing that
a variety of nutrients can influence genetic and epigenetic
processes that determine cellular metabolism, differen-
tiation and apoptosis (Bradlow et al. 1999; Knowles &
Milner, 2000; Pan et al. 2000; Blanchard et al. 2001;
Merrill et al. 2001). It is known that cell homeostasis is
regulated by a finely controlled balance among pro-
liferation, growth arrest, differentiation and apoptosis
(programmed cell death). Disruption in this balance can
lead to profound phenotypic changes ranging from
growth suppression to the transformation of normal into
neoplastic cells. Dysregulation of apoptosis is frequently
accompanied with the pathogenesis arising from a wide
array of conditions including neurodegeneration, auto-
immunity, heart disease, cancer and others. Several nutri-
ents are already identified as factors influencing cellular
homeostasis. For example, while vitamin A has repeatedly
been shown to be involved with differentiation, other
nutrients such as vitamin D and lignan may also play a
role (Ward et al. 2000; Gray et al. 2001; Lee & Pelletier,
2001). Since various nutrients can influence the same
process, it becomes of increasing importance to understand
the ideal balance of these nutrients that brings about a
desired effect and whether a shift in this balance leads to
changes in physiological processes and/or phenotypic
characteristics. These dynamic interactions are not limited
to differentiation but are also evidenced by the ability of
diverse nutrients such as plant sterols, Se and butyrate
arising from fermentable fibres to promote apoptosis in
intestinal cells (Awad & Fink, 2000; Chapkin et al.
2000; Schrauzer, 2000). One of the major issues to be
clarified in the future is the minimum quantity of these
dietary components required to bring about phenotypic
effects and if genetic differences within tissues determines
their physiological consequences. Unfortunately, far too
often many pre-clinical studies utilize concentrations of
bioactive food components that would be virtually
impossible to achieve with typical daily eating behaviours.
Historically, evidence has surfaced that cells are effective
in acclimatizing to insults resulting from exposures to
excessive quantities of nutrients (Fafournoux et al. 2000;
Jackson, 2000). It remains to be determined if intermittent
intake of nutrients offers advantages or if sustained intake
is a method to promote health and well-being.
Excesses are not the only way that nutrients may change
cellular events. Caloric restriction is associated with a
reduction in a number of age-specific chronic diseases
(Turturro et al. 1994). Evidence already exists that the tran-
scriptional silencing of selected genes by DNA methylation
plays a crucial role in ageing and a number of diseased
states (Issa, 2000). Evidence with yeast indicates that
caloric restriction can lead to the silencing of a variety of
genes (Guarente & Kenyon, 2000).
While the study of nutritional genomics is still in its
infancy, it is starting to reveal that nutrient excesses and
deficiencies can bring about a host of genomic and proteo-
nomic changes. Regardless of whether the molecular target
is at the transcription, translation or post-translational level,
the net result is an up- or down-regulation of specific gene
products. Unravelling the multitude of interactions among
nutrients with these key events makes the challenge of
nutritional genomics extremely daunting. Inter-individual
differences probably reflecting genetic polymorphisms
may mask the response to an individual nutrient and
thereby complicate this undertaking to an even greater
degree (Hegelea et al. 1997; Hegele, 1998; Rapuri et al.
2001). In some cases nutrients that are generally thought
to be protective may actually increase risk (Hilakivi-
Clarke et al. 1999; Ross, 2000). Deciphering the dynamic
relationship between dietary components and genes is
fundamental to optimizing health. With the information
gained, it should be possible to determine why inconsis-
tencies occur in the nutrition and health literature and to
develop meaningful and tailored strategies to assist
individuals.
Complementary and overlapping mechanisms appear to
account for the response to bioactive food components.
These biological responses encompass such diverse func-
tions as serving as an antioxidant, promoting the activity
of detoxification enzymes, blocking carcinogen formation
and metabolism, shifting hormonal homeostasis, retarding
cell division, and inducting apoptosis. Since more than
one of these processes may be influenced simultaneously,
it is difficult to determine which is most important in
explaining any phenotypic changes.
Some of the most compelling evidence that diet can
influence the cancer process comes from the intervention
study by Clark et al. (1996) with Se yeast as a supplement.
Based on these findings and a host of pre-clinical studies,
the National Cancer Institute and a network of researchers
known as the Southwest Oncology Group initiated the
largest-ever prostate cancer prevention study, The
Selenium and Vitamin E Cancer Prevention Trial, to
determine if these two dietary supplements can offer
protection (http://newscenter.cancer.gov/pressreleases/
SELECTQandA.html). Although much of the attention
 Diet and health
on Se during the past decade has focused on its anti-
oxidant activity, this trace element is known to bring
about a diverse set of biological effects including
suppressing cell proliferation, enhancing immuno-
competence, blocking carcinogen metabolism, and
induction of apoptosis. By identifying which one of
the events is most important in altering the phenotypic
characteristics under specific circumstances, it should
be possible to identify who might and might not benefit
from exaggerated intakes of this trace element and
during what periods of life that efficacy might be
greatest. The characterization of the specific molecular
target(s) for this and other nutrients represents a major
hurdle for the science of nutrition. Nevertheless, it is
fundamental to truly understanding the involvement of
nutrients in cancer prevention as well as in other
health anomalies.
Another nutrient with apparent significance in the cancer
process is folate (folic acid). Its essential role in the de
novo biosynthesis of purines and pyrimidines, and thus in
DNA replication and cell division, and for the synthesis
of S-adenosylmethionine, a methyl donor for more than
100 biochemical reactions including methylation of
DNA, places it in a unique position relative to DNA
stability (Kim et al. 1997; Molloy & Scott, 2001). These
biosynthetic pathways, each of which is important to
DNA metabolism, appear to compete when the dietary
methyl supply is inadequate, as in folate deficiency, poss-
ibly resulting in altered DNA methylation (an epigenetic
event), disruption of DNA integrity, disruption of DNA
repair and, consequently, increased risk for several dis-
orders (Molloy & Scott, 2001). Hypomethylation and
DNA strand breaks arising from folate inadequacy may
actually promote the incorporation of viruses such as
human papilloma virus into human DNA (Choi &
Mason, 2000). Additional research is need to determine
if the response to folate inadequacy is also observed
when methyl donor supply is suppressed by depression in
choline, vitamin B12, pyridoxine and a variety of other
nutrients.
A number of non-essential phytonutrients have also been
found to impact health (Clydesdale, 1998; Milner, 2000).
Recent studies have revealed that the use of transgenic
and knockout animals offers exciting opportunities to
define the mechanisms by which nutrients including phyto-
chemicals function. Diethiolethione represents one class of
nutrients that has been reported to influence a variety of
molecular targets associated with cancer (Kensler et al.
2000). One of the major mechanisms of protection against
carcinogenesis, mutagenesis and other forms of toxicity
mediated by carcinogens is the induction of enzymes
involved in their metabolism, particularly phase 2 enzymes
such as glutathione S-transferases, UDP-glucuronosyl
transferases and quinone reductases. The use of a knockout
animal model has revealed that 1,2-dithiole-3-thione trig-
gers nuclear accumulation of the transcription factor Nrf2
and its enhanced binding to the antioxidant response
element, leading to transcriptional activation of a score
of genes involved in carcinogen detoxification and attenu-
ation of oxidative stress (Ramos-Gomez et al. 2001). More
recently, Gupta et al. (2001) found that polyphenols in
S155
green tea were effective in reducing the incidence and
metastasis in the transgenic adenocarcinoma mouse pros-
tate model and increased the overall longevity of these
animals. Additional studies are needed that employ a
range of gene expressions to determine what impact
genetic backgrounds have on the response to individual
nutrients.
Foods and food components may also influence the
microenvironment within the gastrointestinal tract. Inulin
and oligofructose are intriguing dietary fermentable fibres
that may have an impact on a number of processes directly
and indirectly, and thereby influence health (Roberfroid,
1993). Inulin and oligofructose are fructans with a degree
of polymerization of 2 to 60 and 2 to 20, respectively.
Owing to their structural conformation, they are resistant
to hydrolysis by human alimentary enzymes and therefore
are fermented almost exclusively by colonic bifidobacteria
and bacteroides. This fermentation increases faecal bac-
terial biomass, decreases caeco-colonic pH and produces
a large amount of fermentation products, among which
are short-chain fatty acids. While the long-term impli-
cations of changing the intestinal microflora remain
unknown, the changes are consistent with an induction of
genes in these micro-organisms and with a reduction in
gastrointestinal distress (Roberfroid, 1993; Kleessen et al.
2001).
The ability of several nutrients to influence the same
biological processes, as mentioned above, raises issues
about possible synergy, as well as antagonistic interactions,
among dietary components. Future studies must character-
ize nutrients in terms of their relative effectiveness, dose
dependency, temporality, consistency and specificity. The
defining of diet-specific molecular targets in terms of gen-
etic and epigenetic events that lead to phenotypic changes
should assist in the development of new and creative diet-
ary intervention strategies for not only reducing diseases
but also improving the overall quality of life.
Dynamics between biomarkers and long-term intervention
Unquestionably, scientifically sound and probing inter-
vention studies must be viewed as the cornerstone for
developing nutrition guidance for individuals. Regrettably,
the sheer number of long-term intervention studies that will
be needed to define nutrient interactions will surely be
impractical in terms of speed of discovery and overall
cost to society. Alternative procedures will thus be required
to predict benefit and risk of selected interventions. These
approaches will necessitate the use of sensitive and reliable
biomarkers. Factors similar to those evaluated by environ-
mental toxicologists (Suk & Collman, 1988; Sakai, 2000)
will be needed to evaluate the benefits/risks of functional
foods and/or their components. Fundamental to this process
will be biomarkers that evaluate: (1) the bioactive food
component capable of modifying a molecular target
(intake/exposure biomarker), (2) biological responses that
evaluate directly or indirectly disease risk or health main-
tenance (effect biomarker), and (3) factors modifying the
response such as genetic and the environment (suscepti-
bility biomarker; Milner, 1999). To assess adequately
whether a food or its component has a physiological
S156
effect it is imperative that stringent experimental design
characteristics are followed. Several factors, including
appropriateness of controls, randomization of subjects,
blinding, statistical power of study, presence of bias, attri-
tion rates, recognition and control of confounding factors
(e.g. weight change or nutrition status) and appropriateness
of statistical tests and comparisons, are addressed in a
guidance document published in the Federal Register
(1999). Each of these factors must become the mainstay
for all investigations. Pre-clinical investigations will
require that many if not most of these same factors be con-
sidered in their experimental design. Whatever biomarkers
are established, it is clear that they must be readily acces-
sible, easily and reliably assayed, differentially expressed
in normal and diseased conditions, directly associated
with disease progression, modifiable and, most important,
predictive (Fig. 1). Similar to the US Department of Agri-
cultures pyramid that is used for dietary guidance, it is
likely that the early predicative biomarkers will not be at
the top because of the lateness of the observation but be
focused at the base where they will be more specific and
timelier. Thus, the future of nutritional biomarkers is
likely to reside in the enhanced use of molecular tech-
nologies to help decode who will and will not benefit
from intervention strategies and, just as important, who
might be placed at harm.
Genetic polymorphisms and dietary variability
Increasing genetic polymorphisms are thought to have a
role in the ability of individuals to withstand exposure to
exogenous carcinogens or to inhibit initiation, promotion
or proliferation in carcinogenesis. It is certainly plausible
that polymorphic differences have been a contributing
Fig. 1. Priority of potential nutritional biomarkers for evaluating the
health benefits associated with bioactive food components. The top
of the pyramid should be used sparingly while the molecular target
base should be used extensively. Regardless of the biomarker
used, it should ideally be readily accessible, easily and reliably
assayed, differentially expressed in normal and abnormal
conditions, directly associated with progression of the abnormal
condition, modifiable and, most importantly, predictive.
J. A. Milner
factor in the inconsistencies surrounding dietary com-
ponents and health (Cotton et al. 2000). For example, in
the Alpha-Tocopherol, Beta-Carotene Cancer Prevention
Study (ATBC Study), there was a low prevalence of poly-
morphisms in genes coding for activation (phase I)
enzymes CYP1A1 (007) and CYP2E1 (002) and a high
prevalence in genes coding for detoxification (phase II)
enzymes GSTM1 (040) and NQO1 (020) (Woodson
et al. 1999). Seven of ten members of this sample carried
the VDR Taq1 polymorphism (t) associated with lower
risk for prostate cancer, which may account in part for
lower cancer rates in Finland compared with the United
States (Woodson et al. 1999). Further, in a nested case
control study within the ATBC Study, glutathione peroxi-
dase 1 (hGPX1), an Se-dependent enzyme involved in
detoxification of hydrogen peroxide, was found to have a
polymorphism exhibiting a proline to leucine replacement
at codon 198. This polymorphism conferred a relative
risk for lung cancer risk of 18 for heterozygotes and 23
for homozygous variants, compared with homozygote
wild types (Ratnasinghe et al. 2000).
The future
Research in nutrition and health in this new millennium
must give top priority to studies that will aid in understand-
ing the basic molecular events by which nutrients influence
biological process. A well-coordinated, multidisciplinary
effort among scientists including nutritional scientists,
molecular biologists, geneticists, statisticians and clinical
cancer researchers may be needed to advance this mol-
ecular approach to nutrition-related cancer research. Many
research questions and issues will need to be addressed for
this approach to become a reality, including the minimum
intake of bioactive components to bring about phenotypic
change and how genetically controlled processes, including
acclimatization, may change the overall response. While
the challenges to researchers will be enormous, the poten-
tial rewards in terms of morbidity and mortality could also
be enormous.
References
Abdulla M & Gruber P (2000) Role of diet modification in cancer
prevention. Biofactors 12, 45 51.
Aldana SG (2001) Financial impact of health promotion
programs: a comprehensive review of the literature. American
Journal for Health Promotion 15, 296 320.
Awad AB & Fink CS (2000) Phytosterols as anticancer dietary
components: evidence and mechanism of action. Journal of
Nutrition 130, 2127 2130.
Blanchard RK, Moore JB, Green CL & Cousins RJ (2001) Modu-
lation of intestinal gene expression by dietary zinc status:
effectiveness of cDNA arrays for expression profiling of a
single nutrient deficiency. Proceedings of the National
Academy of Sciences USA 98, 13507 13513.
Bradlow HL, Telang NT, Sepkovic DW & Osborne MP (1999)
Phytochemicals as modulators of cancer risk. Advances in
Experimental Medicine and Biology 472, 207 221.
Chapkin RS, Fan Y & Lupton JR (2000) Effect of diet on colonic-
programmed cell death: molecular mechanism of action.
Toxicology Letters 112 113, 411 414.
 Diet and health
Choi SW & Mason JB (2000) Folate and carcinogenesis: an
integrated scheme. Journal of Nutrition 130, 129 132.
Clark LC, Combs GF Jr, Turnbull BW, Slate EH, Chalker DK,
Chow J, Davis LS, Glover RA, Graham GF, Gross EG,
Krongrad A, Lesher JL Jr, Park HK, Sanders BB Jr, Smith
CL & Taylor JR (1996) Effects of selenium supplementation
for cancer prevention in patients with carcinoma of the skin.
A randomized controlled trial. Nutritional Prevention of
Cancer Study Group. Journal of the American Medical
Association 276, 19571963.
Clarkson PM & Thompson HS (2000) Antioxidants: what role do
they play in physical activity and health? American Journal of
Clinical Nutrition 72, Suppl. 2, 637S 646S.
Clydesdale FM (1998) Science, education, and technology: new
frontiers for health. Critical Reviews in Food Science and
Nutrition 38, 397 419.
Cotton SC, Sharp L, Little J & Brockton N (2000) Glutathione
S-transferase polymorphisms and colorectal cancer: a HuGE
review. American Journal of Epidemiology 151, 7 32.
DellaPenna D (1999) Nutritional genomics: manipulating plant
micronutrients to improve human health. Science 285,
375379.
Diplock AT, Charleux JL, Crozier-Willi G, Kok FJ, Rice-Evans
C, Roberfroid M, Stahl W & Vina-Ribes J (1998) Functional
food science and defence against reactive oxidative species.
British Journal of Nutrition 80, Suppl. 1, S77 S112.
Eisenberg DM, Davis RB, Ettner SL, Appel S, Wilkey S, Van
Rompay M & Kessler RC (1998) Trends in alternative
medicine use in the United States, 1990 1997: results of a
follow-up national survey. Journal of the American Medical
Association 80, 1569 1575.
Fafournoux P, Bruhat A & Jousse C (2000) Amino acid regulation
of gene expression. Biochemistry Journal 351, 1 12.
Federal Register (1999) Guidance for industry: significant
scientific agreement in the review of health claims for conven-
tional foods and dietary supplements. Federal Register 64,
71794.
Gilbert L (1999) 1999 Health Focus Trend Report. DesMoines,
IA: HealthFocus.
Gill HS & Cross ML (2000) Anticancer properties of bovine milk.
British Journal of Nutrition 84, Suppl. 1, S161 S166.
Go VL, Wong DA & Butrum R (2001) Diet, nutrition and cancer
prevention: where are we going from here? Journal of Nutrition
131, Suppl. 11, 3121S 3126S.
Gray T, Koo J & Nettesheim P (2001) Regulation of mucous
differentiation and mucin gene expression in the tracheo-
bronchial epithelium. Toxicology 160, 35 46.
Greenwald P & Milner JA (2001) Nutrition: a major player in
chronic diseases. The Chronicle of Higher Education, B15
Guarente L & Kenyon C (2000) Genetic pathways that regulate
ageing in model organisms. Nature 408, 255 262.
Gupta S, Hastak K, Ahmad N, Lewin JS & Mukhtar H (2001)
Inhibition of prostate carcinogenesis in TRAMP mice by oral
infusion of green tea polyphenols. Proceedings of the National
Academy of Sciences USA 98, 10350 10355.
Hasler CM, Kundrat S & Wool D (2000) Functional foods and
cardiovascular disease. Current Atherosclerosis Reports 2,
467475.
Hegele RA (1998) A review of intestinal fatty acid binding
protein gene variation and the plasma lipoprotein response to
dietary components. Clinical Biochemistry 31, 609 612.
Hegelea RA, Jugenberg M, Connelly PW & Jenkins DJA
(1997) Evidence for gene diet interaction in the response
of blood pressure to dietary fibre. Nutrition Research 17,
12291238.
Hilakivi-Clarke L, Cho E, Onojafe I, Raygada M & Clarke R
(1999) Maternal exposure to genistein during pregnancy
S157
increases carcinogen-induced mammary tumorigenesis in
female rat offspring. Oncology Report 6, 1089 1095.
Issa JP (2000) CpG-island methylation in aging and cancer.
Current Topics in Microbiology and Immunology 249,
101 118.
Jackson AA (2000) Nutrients, growth, and the development of
programmed metabolic function. Advances in Experimental
Medicine and Biology 478, 41 55.
Kensler TW, Curphey TJ, Maxiutenko Y & Roebuck BD (2000)
Chemoprotection by organosulfur inducers of phase 2 enzymes:
dithiolethiones and dithiins. Drug Metabolism and Drug
Interactions 17, 3 22.
Kim YI, Pogribny IP, Basnakian AG, Miller JW, Selhub J, James
SJ & Mason JB (1997) Folate deficiency in rats induces DNA
strand breaks and hypomethylation within the p53 tumor
suppressor gene. American Journal of Clinical Nutrition 65,
46 52.
Kleessen B, Hartmann L & Blaut M (2001) Oligofructose and
long-chain inulin: influence on the gut microbial ecology of
rats associated with a human faecal flora. British Journal of
Nutrition 86, 291300.
Knowles LM & Milner JA (2000) Allyl sulfides modify cell
growth. Drug Metabolism and Drug Interactions 17,
81 107.
Leclercq C, Valsta LM & Turrini A (2001) Food composition
issues implications for the development of food-based
dietary guidelines. Public Health Nutrition 4, 677 682.
Lee TH & Pelletier J (2001) Functional characterization of WT1
binding sites within the human vitamin D receptor gene
promoter. Physiological Genomics 7, 187 200.
Merrill AH Jr, Sullards MC, Wang E, Voss KA & Riley RT
(2001) Sphingolipid metabolism: roles in signal transduction
and disruption by fumonisins. Environmental Health
Perspectives 109, Suppl. 2, 283 289.
Milner JA (1999) Functional foods and health promotion. Journal
of Nutrition 129, Suppl. 7, 1395S 1397S.
Milner JA (2000) Functional foods: the US perspective. American
Journal of Clinical Nutrition 71, Suppl. 6, 1654S 1659S.
Molloy AM & Scott JM (2001) Folates and prevention of disease.
Public Health Nutrition 4, 601 609.
National Center for Health Statistics (1997) Report of final
mortality statistics, 1995. Monthly Vital Statistics Report 45,
Suppl. 2(11), 12 June.
Paabo S (2001) Genomics and society. The human genome and
our view of ourselves. Science 291, 1219 1220.
Pan DA, Mater MK, Thelen AP, Peters JM, Gonzalez FJ & Jump
DB (2000) Evidence against the peroxisome proliferator-
activated receptor alpha (PPARalpha) as the mediator for
polyunsaturated fatty acid suppression of hepatic L -pyruvate
kinase gene transcription. Journal of Lipid Research 41,
742 751.
Potter JD (1997) Cancer prevention: epidemiology and
experiment. Cancer Letters 114, 7 9.
Prosky L (2000) When is dietary fiber considered a functional
food? Biofactors 12, 289 297.
Ramos-Gomez M, Kwak MK, Dolan PM, Itoh K, Yamamoto M,
Talalay P & Kensler TW (2001) From the cover: sensitivity to
carcinogenesis is increased and chemoprotective efficacy of
enzyme inducers is lost in nrf2 transcription factor-deficient
mice. Proceedings of the National Academy of Sciences USA
98, 3410 3415.
Rapuri PB, Gallagher JC, Kinyamu HK & Ryschon KL (2001)
Caffeine intake increases the rate of bone loss in elderly
women and interacts with vitamin D receptor genotypes.
American Journal of Clinical Nutrition 74, 569 570.
Ratnasinghe D, Tangrea JA, Andersen MR, Barrett MJ, Virtamo
J, Taylor PR & Albanes D (2000) Glutathione peroxidase
S158 J. A. Milner
codon 198 polymorphism variant increases lung cancer risk.
Cancer Research 60, 6381 6383.
Roberfroid M (1993) Dietary fiber, inulin, and oligofructose: a
review comparing their physiological effects. Critical Reviews
in Food Science and Nutrition 33, 103.
Roberfroid MB (2000) Concepts and strategy of functional food
science: the European perspective. American Journal of
Clinical Nutrition 71, Suppl. 6, 1660S 1664S.
Ross JA (2000) Dietary flavonoids and the MLL gene: a pathway
to infant leukemia? Proceedings of the National Academy of
Sciences USA 97, 4411 4413.
Sakai T (2000) Biomarkers of lead exposure. Industrial Health
38, 127 142.
Schrauzer GN (2000) Anticarcinogenic effects of selenium.
Cellular and Molecular Life Sciences 57, 1864 1867.
Suk WA & Collman GW (1988) Genes and the environment:
their impact on childrens health. Environmental Health
Perspectives 106, 817 820.
Turturro A, Blank K, Murasko D & Hart R (1994) Mechanisms of
caloric restriction affecting aging and disease. Annals of the
New York Academy of Science 719, 159 170.
Ward WE, Jiang FO & Thompson LU (2000) Exposure to
flaxseed or purified lignan during lactation influences rat
mammary gland structures. Nutrition and Cancer 37,
187 192.
Woodson K, Ratnasinghe D, Bhat NK, Stewart C, Tangrea JA,
Hartman TJ, Stolzenberg-Solomon R, Virtamo J, Taylor PR
& Albanes D (1999) Prevalence of disease-related DNA poly-
morphisms among participants in a large cancer prevention
trial. European Journal of Cancer Prevention 8, 441 447.
S146 F. M. Lajolo
metropolitan areas, but at the same time has caused
increased adoption of inadequate dietary patterns and
decreased physical activity. Diseases of both deficiency
and excess are simultaneously important in the public
health scenario. This has been called the nutrition
transition.
Increased consumption of diets high in fat (particularly
saturated fat), low in complex carbohydrates and less
dense in micronutrients, combined with a more sedentary
life-style, is responsible for spread of diet-related disorders
including obesity, diabetes, cardiovascular diseases, hyper-
tension, osteoporosis and cancer. This situation increases
disabilities and premature death, placing an additional
burden on public health budgets. Over the next ten years
it is estimated that 11 million people in Latin America
and the Caribbean will die from cardiovascular diseases,
57 million from cancer and 34 million from external
causes. Nearly 23 % of cardiovascular deaths and 39 % of
cancer deaths will occur among people under 60 years of
age, representing premature mortality that can be prevented
or postponed (PAHO, 1999b).
Data show that many known risk factors for non-
communicable diseases are becoming highly prevalent in
Latin American and Caribbean countries. Infant mal-
nutrition may coexist with a high prevalence rate of
overweight in some adult populations, as high as
30 55 % in some groups. Obesity among women is 3 %
in Haiti, 8 10 % in eight countries and 29 % in Mexican-
Americans. The prevalence of overweight among 1 5-
year-old children ranges from 6 % in Haiti to 24 % in
Peru, among thirteen countries. Prevalences of overweight
in children and obesity in women are greater in urban areas
and in households of higher economic status (Martorell
et al. 1998). Studies in Brazil have shown overweight
rates as high as 30 % in school children and in urban
areas. The prevalence of hypertension has been estimated
to range between 8 and 30 % and even higher (40 %) in
existing studies. High prevalence of hypercholesterolaemia
(30 40 %) has been documented in some regions. The few
existing studies on physical activity reported rates of
sedentary life as high as 50 90 %. Even diabetes, pre-
viously uncommon in many developing countries, seems
to be growing in importance with prevalence estimates of
between 5 and 10 %. Much higher figures, up to 18 %,
have been found in some countries of the Caribbean. In
the year 2000 the number of people who suffered from dia-
betes in the Americas was estimated as 35 million, of
which 19 million (54 %) live in Latin America and the
Caribbean.
Micronutrient deficiencies
The nutrition transition changes in dietary habits are also
causing nutritional deficit, resulting in marginal micro-
nutrient deficiencies, well-known classical clinical
deficiencies or both. But diseases of both deficiency and
excess are simultaneously important.
Estimates of 1995 indicated over 168 million people at
risk of iodine deficiency disorders (63 million affected by
goitre), 15 million children under 5 years old suffering
from various levels of vitamin A deficiency (20 %
prevalence) and 94 million suffering from deficiency
anaemia (40 % prevalence in women; PAHO, 2000).
Classical deficiency affects millions of people and several
countries have fortification programmes, some of which are
even mandatory, like the addition of Fe to flour in Brazil.
In urban areas, research has indicated that priorities are
deficiencies of Fe, Zn and vitamin A in infants and
pre-school children and of Fe, folate, Ca and vitamin A
in adolescent, pregnant and lactating women (Trugo et al.
1997).
Costa Rica has better nutritional status than most other
Latin American countries, with low levels of anaemia;
yet half of Costa Rican adolescent girls showed anaemia
and low folate intake. This shows the importance of study-
ing specific population groups not only concerning
anaemia, but also for neural tube defects and cardio-
vascular diseases (Monge et al. 2001). In Brazil, too,
evidence based on consumption and clinical and biochemi-
cal surveys has shown the prevalence of Fe deficiency,
followed by low consumption (or low availability) of
folate, Zn and Ca, a trend of reduced fibre intake
and excessive Na consumption (Cozzolino, 1997).
Regarding vulnerable groups and critical nutrients,
the requirement may be above the potential available in
foods.
As for the relationship between micronutrient status and
chronic degenerative diseases, one should focus on folates,
antioxidant nutrients and cardiovascular diseases, and on
Ca and osteoporosis.
Opportunities and challenges
Foods for deficiency and for excess
Nutrition is recognized as one of the major health deter-
minants. We have already seen that the epidemiological
and nutritional situation in Latin America represents a
considerable challenge for everyone involved in the dif-
ferent aspects of food and nutrition: university, industry
and government staff, responsible for research, production
and regulation, respectively.
There are needs and opportunities to produce foods for
the classical clinical deficiencies, and also for the already
known marginal or emerging micronutrient deficiencies
of particular population sub-groups. Enriched, restored
and fortified foods, as well as foods with a high bioavail-
ability of nutrients, must be available either for government
nutrition programmes or on the market. Considering that in
Latin American countries the diet is frequently based on a
staple like cassava, corn, potato or beans with low micro-
nutrient density, there is borderline ingestion of several
nutrients, such as folic acid, Zn, Se and Ca. Thus optimum
consumption of these micronutrients in some groups may
only be achieved by some kind of diet supplementation.
In this respect genetic engineering is an important tool to
increase nutrient contents and bioavailability, and in some
cases may be the best approach. Some examples are the
introduction of vitamin A and also the increase in Fe con-
tent of rice. The production of foods that may improve
immunological protection of infants, to prevent intestinal
infections, is another way to introduce functionality in
 Latin American perspectives
foods and address public health concerns in less developed
countries. A group from the Biomedical Science Institute
of the authors university has engineered a Lactobacillus
to express Escherichia coli intimine that stimulates serum
immunoglobulin A production and thus may protect infants
from diarrhoea.
At the same time there is a need for foods to address
specific health conditions that have become critical in
Latin American countries nowadays, like the ones asso-
ciated with cardiovascular problems, diseases like cancer,
diabetics, obesity, osteoporosis and others. Latin America
also needs foods that may help specific sub-groups of the
population gain control of the risk of disease and of their
well-being.
The concepts of food functionality and functional food
science, as defined in Europe, seem to be the right
approach to generate the knowledge necessary to help
industry offer new foods in order to answer these
challenges and help governments to define nutrition and
adequate regulatory policies.
Bioactive phytochemicals in foods and plants
Another aspect to be considered is associated with
phytochemical compounds of foods and other plants. A
growing number of studies have linked diets rich in fruit
and vegetables with good health, and correlations have
also been found between the ingestion of some compounds,
like flavonoids, carotenoids and glucosinolates, and lower
risks of cancer, cardiovascular and some other chronic
diseases. In Latin America there are very few data concern-
ing the nutrient composition of foods and virtually none on
non-nutrient components. Considering the low fruit con-
sumption and the identified marginal deficiencies of
micronutrients, the consumption of protective compounds
may be low, calling for more research.
One can say that the discoveries of the fifteenth century
introduced Latin America into the global economy and that
globalization has proceeded from then on. Since then food
products such as corn, beans, amaranth, quinoa, potatoes,
tomatoes, cocoa, guava, papaya, avocado and many
others have become an important part of the diet in
many regions of the world, including Europe. The great
potential of unknown or under-exploited Latin American
plants is still waiting adequate research. It is estimated
that in the Amazon region there are 50 000 plants repre-
senting 25 % of total biodiversity, but less than 2 %
have been studied. Many of them may be important
sources of phytochemicals, flavours and colours.
Latin American plants are usually rich in flavonoids,
anthocyanins and carotenoids, and some have interesting
amounts and combinations of specific carotenoids like
zeaxanthin, lutein and others. Some fruits like guava and
palm fruits like aca (Euterpe oleracea ) have three times
more lycopene than tomatoes and are also rich sources
of anthocyanins (Rodriguez-Amaya, 1999). Camu-camu
(Myrciaria dubia ) fruit may contain 2 % ascorbic acid.
Brazil nuts are the richest source of Se. Roots from some
regions produce functional oligosaccharides, polysac-
charides and important oils, sterols, and saponins. Maca
(Lepidium meyenii ), guarana (Paulinia cupana ) and mate
S147
tea (Ieia paraguayensis ) are known for their energy-
giving properties. Many of them have important
physiological and pharmacological actions. Several of
these fruits are no longer just harvested but are cultivated
industrially, and soon they will be economically
significant.
Claims, efficacy validation and human variability
A great variety of products are available in open markets in
Latin America, even in highly urbanized areas like Sao
Paulo. They are whole plants, plant organs like leaves,
tubers, seeds, fruits presented in several forms from
powder to extracts, and juices, alone or in combination.
They are very popular because they are part of the cultural
tradition, they are perceived as natural and therefore
healthy, and because they are cheap. These products are
sold with claims for important diseases, to help nutrition
or simply for bad humour. Not much is known about the
composition, safety and activity of most of them. Only
recently have research projects been undertaken in a
more systematic manner and these products submitted to
review by health authorities.
Latin America has a very diverse cultural and genetic
background. The indigenous population, for instance, can
represent 40 % or more in some regions and less than
1 % in others. Populations with African ancestors in the
Americas may vary from less than 3 % in Mexico to
44 % in Brazil and 84 % in the Dominican Republic. The
first genetic profile of the Brazilian population showed
that the black people are whiter and the white people are
blacker than first thought (white are partly genetically
black and vice versa). Interesting data are emerging in
Brazil to suggest that genes vary from region to region,
where the regions also have different cultural and dietary
habits (Pena, 2000). Here, differences in Apo E4 genes
and LDL A+ A+ genes are linked to coronary diseases
(Cavalli et al. 2000; Salazar et al. 2000) and Apo E4
genes to high cholesterol and diabetes mostly in black
people (Otta et al. 1996). This is important, for instance,
in studies on cholesterol reduction by diet. All this clearly
shows the importance of genetic background and human
ecology considerations when evaluating the efficacy of
functional foods. Even if urbanization has homogenized
some habits, there is still, by and large, a significant vari-
ation in dietary habits and organism physiology in our
populations, pointing to the importance of and need for
local research.
Functional foods should be viewed in Latin America
beyond short-term commercial prospects and consider
long-term research and development, regional needs, and
the cultural habits and economical level of consumers.
Legislation on functional foods in Latin America:
regulatory status
Concerning functional foods, adequate legislation is
essential to protect public health and stimulate techno-
logical development. In Latin America and the Caribbean,
regulation concerning functional foods, functional sup-
plements and functional and health claims differs greatly
S148 F. M. Lajolo

from country to country. There is no official or legal
definition of functional foods or a specific regulation for
functional foods/ingredients. In general, basic nutrient
content/function claims are allowed and subject to some
norms, but only a few countries allow or have norms for
health claims. Even so, health authorities have allowed pro-
duct claims by on a case-by-case basis in several countries.
Brazil is the only country to have a well-defined regu-
lation for functional and health claims for either nutrient
or non-nutrient components and also for the demonstration
of safety and efficacy (Agencia Nacional de Vigilancia
Sanitaria, 1999). In almost all countries functional or
health claims associated with non-nutrient compounds or
non-essential nutrients are neither prohibited nor regulated.
As a consequence, there is no legislation on efficacy and
safety and also a lack of criteria for scientific substantiation
of claims, except for Brazil. Even so, all over Latin
America, in popular markets and even in supermarkets,
hundreds of foods and drugs are sold together with curative
allegation.
Considering the regional economic level, cultural
identity and public health situation, the development of
common functional foods concepts and norms seems to
be very important. Brazilian legislation can be used as a
starting point in this direction since it has been based on
internationally acceptable principles and is considered to
be up-to-date. In Brazil functional foods have not been
defined as such but the norms were based on the idea of
a food that is food and not a drug, that is part of a
normal diet and that can produce benefits beyond basic
nutrition. Legislation essentially rules on demonstration
of the safety and efficacy of novel foods (foods not used
before in Brazil) and foods/ingredients that have a claim
in the label. All of these products should be registered
and approved by health authorities. Safety demonstration
is a priority and should be based on risk analysis, including
risk assessment, management and information. Efficacy
concerning the claims should be based on scientific
evidence obtained from the literature or by new research
and must represent scientific consensus, as has been

Table 1. Examples of food claims approved and not approved in different Latin American countries

Spreads with phytosterols

Brazil Helps to maintain healthy level of cholesterol
When associated with a healthy diet and life-style
(research in Brazil suggested and concluded)
Milk containing phytosterol (different manufacturers)
Argentina Helps in reducing cholesterol
As part of a healthy diet
Milk containing n 2 3 fatty acids (EPA and DHA)
Brazil Helps to control triglyceride levels, blood fluidity, inflammatory and immune response
When associated with a healthy diet and life-style:
Enhances immune defence
Argentina Reduces inflammatory processes, control the level of triglycerides and cholesterol.
Increases blood fluidity
In doing so they:
Contribute to prevention or retardation of atherosclerosis and associated diseases
of the circulatory and nervous systems
Milk with n 2 3 and n 2 6 fatty acids from vegetable oil
Argentina Has less saturated fatty acids that increase cholesterol and more n 2 3 and n 2 6 that help
reducing blood cholesterol
Good for healthy persons who have balanced nutrition and life-style
Brazil Low in saturated fat and enriched with n 2 3 and n 2 6 fatty acids
Needed to keep cholesterol low and healthy heart
Milk with prebiotics
Chile and Argentina Is a functional ingredient that naturally helps your son to have a better digestive
system performance
Allows better absorption of vitamins
Mexico Helps for a better absorption of nutrients to strengthen your defences
Brazil Contributes to a healthy intestinal flora
Helps maintaining the intestinal flora balanced
Probiotics (lactobacilli and bifidobacteria)
Brazil Helps in maintaining the intestinal equilibrium flora balanced
Helps in reducing harmful bacteria
Helps in increasing the beneficial flora
Allegations not approved:
Increases antibodies
Strengthens natural defences against daily aggression and stress
Other claims for specific bacteria are under discussion
Breakfast biscuits
Argentina Slow releasing energy. Energy is released slowly during all the morning
(shows a glycaemic curve on the label)
 Latin American perspectives
discussed by the US Food and Drug Administration and by
the International Life Science Institutes in several forums
and in the Codex Alimentarius.
In Brazil, both functional (enhanced function) and health
claims are defined and allowed. As for health claims, only
health maintenance and risk reduction claims are allowed,
with both prevention and cure allegations being prohibited.
In this case, however, the product can be considered as a
drug and assessed as such. The label on the product
should be clearly understandable by consumers, not
misleading, it should inform the limitation of efficacy,
and should be in accordance with public health policies.
In some specific cases products should clearly inform
restriction of their use to specific groups or to individuals
with specific physiological conditions, as well as any
adverse side-effects that may occur. These rules also
apply for marketing and advertising materials.
Regarding products in pharmaceutical form that contain
bioactive compounds normally present in foods, the label
should inform that the product does not prevent or cure
any disease. Mixtures of foods and food components
with herbals that have pharmacological action are con-
sidered drugs or phytotherapeutic agents, and submitted
to pharmaceutical regulations.
In Brazil, like in many other countries in Latin America,
there is a great use of phytotherapeutics. These include
plants, plant organs, extracts, pills, etc. that by tradition
or by research are known to cure or alleviate some
diseases, and are classified and regulated as drugs by
health authorities. Two years ago, before the existence of
norms in Brazil, there were a large number of these
natural products being sold as foods or food supplements
with unproved and undemonstrated curative claims for all
known diseases. Some of them were complex mixtures of
vegetable extracts containing highly active pharmaco-
logical compounds such as amphetamines. The legislation
issued allowed organization of the market through the
analysis for safety and efficacy of the existing 2000
products, among which only 430 are still being produced
by 115 manufacturers.
Foods with functional or health claims in Latin
America
During the last two years, a number of new foods that may
be considered as functional foods, since they have func-
tional or health properties acknowledged by health
authorities, have appeared on the market in some countries.
The most important are spreads and milks with added
phytosterols, milk containing long-chain n 2 3 fatty acids
and milks with added n 2 3 and n 2 6 fatty acids from
vegetable oils. Milk and products with oligofructose,
margarine and yoghurts containing fibre, cookies engine-
ered to have low glycaemic index, and milks fermented
with selected Lactobacillus and Bifidobacterium strains
have also appeared. Other products are those containing
soyabean proteins and isoflavones, low-cholesterol eggs,
and energy and isotonic drinks containing caffeine and
other herbal extracts.
A number of products in pharmaceutical form should
also be mentioned. These are mostly fibre-containing
S149
products, antioxidants and oils containing freeze-dried
vegetable/fruit extracts.
Some of the claims approved and not approved in dif-
ferent countries can be used as examples to illustrate the
on-going legal situation in Latin America, and are shown
in Table 1. As shown by these examples, almost all of
the claims can be classified more as functional claims
than as claims related to a disease, reflecting the present
position of the health authorities. They are mostly based
on internationally accepted scientific knowledge but some
of them may be too sophisticated even for a well-informed
public.
Latin American consumers, in general, do not know
what functional foods are, although in the more urbanized
areas there is an increasing number of health-conscious
consumers aware of the importance of diet for health and
well-being. This awareness comes either from the media
or cultural tradition. Concepts associated with vitamins,
obesity, cholesterol, fibre, fermented milk and physical
activity are disseminated, even if the general nutrition
knowledge is poor. However, long and complex claims
are difficult to understand and may confuse consumers. It
is necessary to disseminate the best available information
and advice about diet, as part of a sound alimentary
education.
In several cases, marketing information carried by the
media to promote a product goes far beyond what was
approved, misleading the consumer. To solve this problem
some countries like Brazil are also regulating propaganda
on foods. This may help give credibility to the assessment
process and to the functional foods area.
While middle- and upper-class people can get their
health foods at the supermarket, lower-class people can
get theirs in other places where they find their functional
garrafadas, known as natural and therefore healthy pro-
ducts for nutrition and healing. They both seem clearly
to understand how to make use of the concept of improving
their health and well-being through food.
Conclusions
The occurrence of diet-related diseases of deficiency and
excess points to the importance of the development of
functional food (science). In urban areas, due to the
media and cultural tradition, there is increasing consumer
awareness of the importance of foods for health. The
perspectives of Latin America as a potential producer
and consumer of functional foods will depend on the
level of information and income of the population, on the
creditability and pricing of products, on the nutrition and
regulatory policies and on research investments. Functional
food science must be viewed in Latin America beyond the
short-term commercial prospects and should consider
long-term research and development.
References
Agencia Nacional de Vigilancia Sanitaria (1999) Decreto 16, 17,
18 e 19 de abril de 1999. http://www.anvisa.gov.br.
Cavalli SA, Hirata MH, Salazar LA, Diament J, Forti N, Giannini
S150 F. M. Lajolo
SD, Nakandakare ER, Bertolami MC & Hirata RDC (2000)
Apolipoprotein B gene polymorphism: prevalence and impact
on serum lipid concentrations in hypercholesterolemic
individuals from Brazil. Clinica Chimica Acta 302, 189 203.
Cozzolino S (1997) Biodisponibilidade de minerais. Revista Puc-
camp 10, 877 898.
Martorell R, Kan LF, Hujlos ML & Grummer-Strawn LM (1998)
Obesity in Latin American women and children. Journal of
Nutrition 128, 14641473.
Monge R, Raiges F & Rivero A (2001) Iron and folates status in
urban and rural Costa Rican teenagers. Food and Nutrition
Bulletin 22, 45 52.
Otta MI, Cavalli SA, Hirata RDC, Lottenberg S, Quintao E &
Hirata MH (1996) Apolipoprotein E genotype frequencies in
hypercholesterolemic and diabetic individuals. Clinical
Chemistry 42, 839.
Pan American Health Organization (1999a) Assessment of the
1980 1998 health situation and trends in the Americas, by
subregion. Epidemiological Bulletin 20, 2 10.
Pan American Health Organization (1999b) Networking for the
Surveillance of Risk Factors for Non-communicable Diseases
in Latin America and the Caribbean. Paho/HCP/99.08.
Washington, DC: Pan American Health Organization.
Pan American Health Organization (2000) Health situation anal-
ysis in the Americas, 1999 2000. Epidemiological Bulletin
21, 1 3.
Pena SD (2000) Research report on genetic diversity. Minas
Gerais, Brazil. University Federal Minas Gerais.
Rodriguez-Amaya DB (1999) Latin American food sources of
carotenoids. Archivos Latinoamericanos de Nutricion 49,
745 845.
Salazar IA, Cavalli SA, Hirata MH, Diament J, Forti N, Giannini
SD, Nakandakare ER, Bertolami MC & Hirata RDC (2000)
Polymorphisms of the low-density lipoprotein receptor gene
in Brazilian individuals with heterozygous familial hyper-
cholesterolemia. Brazilian Journal of Medical and Biological
Research 33, 1301 1304.
Trugo NMF, Donangels M & Trugo LC (1997) Micronutrient
status and urban lifestyle in Brazil. Archivos Latinoamericanos
de Nutricion 47, Suppl. 1, 1 50.
S140 S. Arai

In this project, several interesting results were obtained. Glance at Asian functional foods
In particular, research into novel phytochemicals eluci-
Asian food processing is characterized by fermentation.
dated the occurrence of new anticarcinogens, e.g. fuco-
This is reflected in the production of functional foods as
sterols in seaweed and auraptenes in citrus fruits. The
well. Actually, FOSHU items (Table 1) include fermented
discovery and characterization of potent antioxidants of
milk, which originates in Caucasus and is produced indust-
plant origin, e.g. curcuminoids in ginger, were also
rially for particular purposes. A good example is offered by
reported. Another unique example was offered by oryza-
a FOSHU product in which reasonable amounts of the
cystatin, a cysteine proteinase inhibitor found in the rice,
lacto-tripeptides Val Pro Pro and Ile Pro Pro, angio-
Oryza sativa L., which was molecularly cloned (Abe
tensin-converting enzyme inhibitors, have been derived
et al. 1987) and demonstrated to inhibit the proliferation
from b-casein by lactic acid bacteria. Another interesting
of human herpes virus in infected animal cells (Aoki
example is provided by fermented soyabean, natto by
et al. 1995). These studies were aimed at finding factors
Japanese name, which has long been consumed in some
with body-defending rather than body-modulating
Asian districts as well as in Japan. Very recently, a
functions.
mutant of Bacillus natto was developed that can produce
For body modulation, food allergies and counter-
a high quantity of menaquinone-7 (vitamin K2) as a
measures were studied extensively. The background was
cofactor of g-glutamic acid carboxylase for biosynthesis
that rice-associated allergy in the form of atopic dermatitis
of the Ca-carrying protein, osteocalcin. The natto pro-
was increasing continually in Japan, a major rice-
duced using this mutant is thus expected to reduce the
consuming country, and countermeasures were needed
risk of osteoporosis and was approved as a FOSHU pro-
urgently. In response to such a social need, an experiment
duct. Also, rice whose wild species per se can be found
was undertaken to design and produce hypoallergenic rice
in a district near Tibet is often used as a material for
grains by enzymatic hydrolysis of main allergenic proteins
fermentation. In Japan, it was found that a rice bran
occurring in the globulin fraction. The development of
extract treated with a mixed Lactobacillus Saccharo-
well-controlled conditions for the hydrolysis gave an
myces culture produces a large amount of g-aminobutyric
immunologically and clinically satisfactory product
acid as a functional food factor for reducing the risk of
(Watanabe et al. 1990) and the production process was
neural disorders as well as hypertension, although the
industrialized to manufacture it. Fine Rice was approved
product remains to be approved.
as the first Food for Specified Health Use (FOSHU) pro-
In China as well, the functional food policy has made a
duct (described below) after careful intervention tests in
good progress. Chinese functional foods are legally
many medical centres. These basic and applied studies
approved, with the logo, in sky blue colour, issued by the
were followed by the discovery and characterization of
Ministry of Public Health. The functions to be considered
soyabean and wheat allergens to develop hypoallergenic
include twenty-four items as follows (Yi et al. 1999):
soyabean and flour products (Tanabe et al. 1996).
The orthodox way of producing functional foods is to
1. immune regulation,
maximize beneficially functional factors that are respon-
2. postponement of senility,
sible for modulation of the immune, endocrine, nerve,
3. memory improvement,
circulatory or digestive system in the body. Sometimes,
4. promotion of growth and development,
however, a unique approach can be adopted. Uniqueness
5. anti-fatigue,
is found in minimizing non-beneficial functional factors
6. body weight reduction,
by enzyme technology or even genetic engineering. The
7. oxygen deficit tolerance,
importance of using this concept for the design and pro-
8. radiation protection,
duction of functional foods has been emphasized since
9. anti-mutation,
1984 when the first project began.
10. anti-tumour,
The Japan Ministry of Health and Welfare in 1991
11. blood lipid regulation,
initiated the worlds first policy of legally permitting the
12. improvement of sexual potency,
commercialization of selected functional foods in terms
13. blood glucose regulation,
of FOSHU, as mentioned above. The new policy is defined
14. gastrointestinal function improvement,
by new legislation and also characterized by approving the
15. sleep improvement,
presentation of a health claim for each FOSHU product.
16. improvement of nutritional anaemia,
Such legal framework is also expected to hinder the pres-
17. protection of liver from chemical damage,
entation of ill-defined and misleading advertisements in
18. lactation improvement,
commercial products. Thus, since 1993, some selected
19. improvement for beauty,
food products have been approved to claim a certain
20. vision improvement,
degree of medical representation never before permitted
21. promotion of lead removal,
for any food. The first was hypoallergenic rice grains
22. removal of intense heat from the throat and
(Fine Rice by commercial name) developed after exten-
moistening of the throat,
sive immunological studies in the MESC project, produced
23. blood pressure regulation and
by enzyme technology and industrialized after clinical
24. enhancement of bone calcification.
intervention tests (Watanabe et al. 1990). A great many
FOSHU products have been approved up to now For some categories among these, a variety of foods are
(Table 1). considered to be approved (Table 2). These foods are also
 Asian perspectives S141

Table 1. Data on 252 foods for specified health uses

Number of products Forms of products on

Health claim* Functional factor approved the market

I Prebiotics 61
Lactosucrose 22 Soft drink, yoghurt, biscuit, cookie,
table sugar
Fructo-oligosaccharides 15 Table sugar, tablet candy, pudding,
soyabean curd
Soyabean oligosaccharides 9 Soft drink, table sugar
Xylo-oligosaccharides 5 Soft drink, vinegar, chocolate, tablet candy
Galacto-oligosaccharides 4 Table sugar, vinegar
Isomalto-oligosaccharides 3 Table sugar
Raffinose 1 Powdered soup
Lactulose 1 Soft drink
Arabinose 1 Table sugar
Probiotics 49
Lactobacillus casei Shirota 26 Lactic acid bacterium drink
Lactobacillus delbrueckii and Steptococcus salivarius 6 Fermented milk
Bifidobacterium lactis 4 Yoghurt
Bifidobacterium breve Yakult 4 Fermented milk
Lactobacillus rhamnosus GG 2 Fermented milk
Bifidobacterium longum 4 Fermented milk
Lactobacillus acidophilus and B. longum 3 Fermented milk
Dietary fibres 58
Indigestible (resistant) dextrin 28 Sausage, cookie, drink, miso soup, curd
Psyllium husks 21 Powdered drink, noodle, cereal
Wheat bran 2 Cereal
Galactomannan 1 Jelly
Partially hydrolysed guar gum 3 Powdered drink, boiled rice
Agar 2 Noodle, jelly
Beer yeast 1 Fermented milk
II Soya protein and peptide 17 Soft drink, humbug, meat ball, sausage,
soya milk
Low-molecular-weight sodium alginate 7 Soft drink, soup
Chitosan 3 Biscuit
Sitosterol ester 1 Margarine
III Peptides 20
Sardine peptides (containing VY) 9 Soft drink, soup
Lacto-tripeptides (VPP and IPP) 5 Lactic acid bacterium drink
Dried bonito oligopeptides (containing IKP) 2 Fermented soyabean soup
Casein dodecapeptides (FFVAPEPEVFGK) 4 Soft drink
Eucommia leaf glycoside (geniposidic acid) 2 Soft drink
IV Diacylglycerol 5 Cooking oil
Diacylglycerol and sitosterol 4 Cooking oil
V Casein phosphopeptide 7 Soft drink, chewing gum, soyabean curd
Calcium citrate malate 3 Soft drink
Heme 4 Soft drink, jelly
Menaquinone-7 producing Bacillus subtilis 2 Fermented soyabean (natto )
Soya isoflavone 1 Tea drink
VI Manitol, palatinose and tea polyphenols 3 Chocolate, chewing gum
Manitol 2 Candy
Xylitol 1 Chewing gum
VII Wheat albumin 2 Powdered soup
Globin digest 1 Soft drink
Guava polyphenols 1 Soft drink

* Claim I, foods that improve gastrointestinal conditions; Claim II, foods for those with high serum cholesterol; Claim III, foods for those with high blood pressure;
Claim IV, foods for those with high serum triacylglycerol; Claim V, foods related to mineral absorption and transport; Claim VI, non-cariogenic foods; Claim VII,
foods for those who begin to feel concerned about their blood sugar level.
As of 29 May 2001.

used as medicines. Thus, distinguishing between foods and
traditional medicines is not an easy task. The difficulty
apparently reflects the Chinese tradition that medicine
and food share a common origin.
Some food factors may be studies for purely medical
rather than food use. In an extensive screening study
covering common vegetables and fruits in south-east
Asian countries, anti-tumour promoting potential, detected
by the inhibitory rate of tumour promoter-induced Epstein-
Barr virus activation, was generally higher in the plants
from tropical zones than in those from Japan. Moreover,
plants in the families Zingiberaceae, Rutaceae, Labiatae
and Cruciferae, which are used as traditional medicines,
were shown to contain potent anti-tumour promoters at
high rates (Arai et al. 2001). For functional food scientists,
Asia may thus be like a treasure house.
S142
Table 2. Representative Chinese functional foods for body
modulation in specified categories (Yi et al. 1999)
Immune regulation
American ginseng (Panax quinquefolius )
Ginseng (Panax ginseng )
Acanthopanax (Acanthopanax senticosus )
Barbary wolfberry fruit (Lycium barbarum )
Astragalus (Astragalus membranaceus, A. membranaceus var.
mongholicus)
Chinese caterpillar fungus (Cordyceps sinensis )
Gingko leaf (Ginkgo biloba )
Walnut (Juglans regia )
Chinese date (Ziziphus jujuba )
Royal jelly
Chinese angelica (Angelica sinensis )
Glossy ganoderma (Ganoderma applanatum )
Postponement of senility
Roxburgh rose (Rosa roxburghii )
Green tea (Camellia sinensis )
Fleeceflower root (Polygonum multiflorum )
Black sesame (Sesamum indicum )
Honey
Mulberry fruit (Morus alba )
Barbary wolfberry fruit (L. barbarum )
Blood lipid regulation
Hawthorn fruit (Crataegus pinnatifida )
Soyabean (Glycine max )
Peach seed (Prunus persica )
Chrysanthemum flower (Chrysanthemum morifolium )
Super vinegar
Spine date seed (Z. jujuba var. spinosa)
Fish oil
Corn oil (Zea mays )
Extract of flax seed (Linum usitatissimum )
Safflower seed (Carthamus tinctorius )
Blood and urine glucose reduction
Fiveleaf gynosttemma (Gynostemma pentaphyllum )
Hawthorn fruit (C. pinnatifida )
Chinese yam (Dioscorea opposita )
Buckweat poria (Fagopyrum esculentum )
Cocos poria (Poria cocos )
Extract of pumpkin (Cucurbita moschata )
Extract of pig pancreas
Powder of spirulina (Spirulina princeps )
International view on functional
food science and policy
Meanwhile, the implementation of the Ministry of Health
and Welfares FOSHU policy as well as the initiation of
functional food science in Japan had a strong impact on
many countries of the world, particularly in Europe.
Early in 1995, the UK Ministry of Agriculture, Fisheries
and Food defined, though temporarily, functional foods
as those that had components incorporated to give specific
medical or physiological benefits, other than nutritional
effects (Richardson, 1996).
Deeper interest was paid by the European body of the
International Life Sciences Institute (ILSI Europe). ILSI
Europe addressed the present status by claiming that we
stand today at the threshold of a new frontier in nutritional
science, and also stating that the concept of food is chang-
ing from a past emphasis on eating to get rid of hunger into
a current emphasis on promising uses of foods to reduce
the risk of chronic illness (Bellisle et al. 1998). This
S. Arai
new concept is of extreme importance in view of the
demands of the elderly population for improved quality
of their late life, increases in life expectancy and costs
for health care, and technical advances in the food
industry.
As recognized in the Japan MESC research project,
up-to-date knowledge in the biosciences, including bio-
chemistry and molecular biology together with sophisti-
cated biotechnologies based on these, supports the
hypothesis that foods can modulate various functions in
the body and thus participate in the maintenance of a
state of health that reduces the risk of life-style related
diseases. Such a hypothesis exists at the origin of the
concept of functional food for the development of a
new discipline, functional food science, in Europe as
well as in Japan.
What may be most important at present is to evaluate the
function. Since any functional food must be based on
science, its evaluation should essentially depend on data
from biochemistry, physiology, molecular and cell biology,
and other most modern biosciences. A key approach to the
development of a functional food is the identification and
validation of relevant markers, including biomarkers that
can predict potential benefits relating to a target function
in the body. If the markers represent an event directly
involved in the process, these should be considered as
functional factors. On the other hand, if the markers
represent correlated events, they should be considered as
indicators. In detail, possible markers have been tenta-
tively listed in relation to target functions (Diplock et al.
1999).
There are many options for selecting appropriate
markers. A promising option may be the use of accumu-
lated knowledge about genomics. As the human genome
project has been completed (Venter et al. 2001), the so-
called post-genomics, involving the use of genomics
data, will follow, which may not be unrelated to functional
food science. A number of sophisticated biotechnologies
will become available, one of the most useful being
DNA tip technology (Marton et al. 1998). It is helpful
because of its high throughput potency to assess the poss-
ible effects and safety of functional food factors by total
gene expression profiles.
Current genome programmes are proceeding on food
organisms as well. These include rice, wheat, corn, soya-
bean, and other major plant seeds for food use. A variety
of post-genomics sciences, to be born shortly before and
after the completion of the genomics, will lead to the
development of new biosciences such as nutrigenomics.
The development will support advances in proteomics,
providing key information about the molecular design of
functional food proteins. It will thus be possible to easily
define their specific functional regions involved in the
prevention of such life-style related diseases as diabetes,
hypertension, hypercholesterolaemia, allergies, cancer and
even infectious diseases. Among a number of trials,
proteomics for the design of antiviral proteins may serve
as an example. One such study involves a three-
dimensional NMR analysis of solution structures of oryza-
cystatin, an anti-herpes protein of rice origin described
above (Abe et al. 1987; Aoki et al. 1995). The analysis
 Asian perspectives
provides information about the minimal effective region
(Nagata et al. 2000), the use of which will contribute
to molecular breeding of new rice cultivated with an
antiviral function.
It is also noted that an ad hoc team headed by the
present author (S.A.) has begun to make a database of
functional foods under the sponsorship of the Japan
MESC.
Finally, it is stressed that Asian people have always
recognized the importance of sensory properties of
foods. They still think that every food must be sensorily
acceptable. Even functional foods cannot be ruled out.
The science should include studies on the senses of
taste and smell. The significance of this is emphasized
since such sensations have been found to be related
more or less to functions of the endocrine and exocrine
systems, as well as to those of the digestive system. An
interesting example can be found in red pepper, whose
unique irritative component, capsaicin, has a hot fla-
vour in seasoning. This substance, shortly after entering
the body per os, interacts with vanilloid receptors such
as Ca channels in sensory neurons inducing a specific
dynamic action, possibly with an anti-obesity result.
Capsaicin can thus be regarded as a functional factor
with such a bilateral effect. Sensory abnormalities and
disorders are other targets of functional food science,
because insights into syndromes of these kinds will
supply patients with organoleptically acceptable tailor-
made foods having body-modulating functions at the
same time.
The design and development of functional foods should
not be carried out independently of purely nutritional and
sensory or cognitive properties of foods. Future advances
in relevant sciences for giving rise to foods with integral
attributes would distinguish functional foods from medi-
cines. Research along this line will add a newer dimension
to functional food science and will also be academically
and socially understood as having great potential for the
development of new food science, technology and industry
in the early twenty-first century.
The reader is advised to consult recently published
papers to learn more about the state-of-the-art concerning
functional foods (Arai et al. 2001; Kwak & Jukes, 2001).
The former paper is the proceedings of a symposium, A
Mainstay of Functional Food Science in Japan
History, Present Status, and Future Outlook, which was
organized by the Japan Branch of the International
Union of Food Science and Technology in collaboration
with ILSI Japan.
S143
References
Abe K, Emori Y, Kondo H, Suzuki K & Arai S (1987) Molecular
cloning of a cysteine proteinase inhibitor of rice (oryza-
cystatin). Journal of Biological Chemistry 262, 16793 16797.
Aoki H, Akaike T, Abe K, Kuroda M, Arai S, Okamura R, Negi A
& Maeda M (1995) Antiviral effect of oryzacystatin, a
proteinase inhibitor in rice, against herpes simplex virus type
1 in vitro and in vivo. Antimicrobial Agents in Chemotherapy
39, 846 849.
Arai S (1996) Studies on functional foods in Japan states of the
art. Bioscience Biotechnology Biochemistry 60, 9 15.
Arai S, Osawa T, Ohigashi H, Yoshikawa M, Kaminogawa S,
Watanabe M, Ogawa T, Okubo K, Watanabe S, Nishino H,
Shinohara K, Esashi T & Hirahara T (2001) A mainstay of
functional food science in Japan history, present status,
and future outlook. Bioscience Biotechnology Biochemistry
65, 1 13.
Bellisle F, Diplock AT, Hornstra G, Koletzko B, Roberfroid M,
Salminen S & Saris WHM (editors) (1998) Functional food
science in Europe. British Journal of Nutrition 80, Suppl. 1,
S1 S193.
Diplock AT, Aggett PJ, Ashwell M, Bornet F, Fern EB &
Roberfroid MB (1999) Scientific concepts of functional foods
in Europe: consensus document. British Journal of Nutrition
81, Suppl. 1, S1 S27.
Kwak S & Jukes DJ (2001) Functional foods, Part 1: the develop-
ment of a regulatory concept. Food Control 12, 99 107.
Marton MJ, DeRisi JL, Bennett HA, Iyer VR, Meyer MR, Roberts
CJ, Stoughton R, Burchard J, Slade D, Dai H, Bassett DE,
Hartwell LH, Brown PO & Friend SH (1998) Drug target
validation and identification of secondary drug target effects
using DNA microarrays. Nature Medicine 4, 1293 1301.
Nagata K, Kudo N, Abe K, Arai S & Tanokura M (2000) Three
dimensional solution structure of oryzacystatin-I, a cysteine
proteinase inhibitor of the rice, Oryza sativa L. japonica.
Biochemistry 39, 14753 14760.
Richardson DP (1996) Functional foods shades of gray. An
industry perspective. Nutrition Reviews 54, S174 S185.
Swinbanks D & OBrien J (1993) Japan explores the boundary
between food and medicine. Nature 364, 180.
Tanabe S, Tesaki S, Yanagihara Y, Mita H, Takahashi K, Arai S
& Watanabe M (1996) Inhibition of basophil histamine release
by a haptenic peptide mixture prepared by chymotryptic
hydrolysis of wheat flour. Biochemistry and Biophysics
Research Communications 223, 492495.
Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG
et al. (2001) The human genome. Science 291, 1304 1351.
Watanabe M, Yoshizawa T, Miyakawa J, Ikezawa Z, Abe K,
Yanagisawa T & Arai S (1990) Production of hypoallergenic
rice by enzymatic decomposition of constituent proteins.
Journal of Food Science 55, 781783.
Yi D, Yong P & Wenkui L (1999) Chinese Functional Food,
pp. 19 20. Beijing: New World Press.
S134
Table 1. List of some challenges to Western societies at the
turn of the new century
Costs of health care
Increase in life expectancy costs
Consumer awareness of the importance of nutrition in health
and well-being
Technical advances in food processing
Improved knowledge of diet-related diseases
Application of new scientific knowledge in nutrition
thus to be understood as a concept (Roberfroid, 2000).
Moreover, if it is function-driven rather than product-
driven, the concept is likely to be more universal and not
too much influenced by local characteristics or cultural trad-
itions (Diplock et al. 1999).
In scientific as well as in more marketing-oriented litera-
ture, functional food has as many definitions as the number
of authors referring to it. These definitions go from simple
statements, such as:
1. foods that may provide health benefits beyond basic
nutrition (IFIC Foundation, 1995) and
2. foods or food products marketed with the message of
the benefit to health (Riemersma, 1996),
to very elaborate definitions such as:
1. food and drink products derived from naturally occur-
ring substances consumed as part of the daily diet and
possessing particular physiological benefits when
ingested (Hillian, 1995);
2. food derived from naturally occurring substances,
which can and should be consumed as part of the
daily diet and which serves to regulate or otherwise
affect a particular body process when ingested
(Smith et al. 1996);
3. food similar in appearance to conventional food,
which is consumed as part of the usual diet and has
demonstrated physiological benefit and/or reduces
the risk of chronic disease beyond basic nutritional
functions (Health Canada, 1997); and
4. food that encompasses potentially helpful products,
including any modified food or food ingredient that
may provide a health benefit beyond that of the traditional
nutrient it contains (Food and Nutrition Board, 1994).
But whatever definition is chosen, functional food
appears as a quite unique concept that deserves a category
of its own, a category different from nutraceutical,
f(ph)armafood, medifood, designer food or vitafood, and
a category that does not include dietary supplements. It
is also a concept that belongs to nutrition and not to
pharmacology. Functional foods are and must be foods,
not drugs. Moreover, their role regarding disease will, in
most cases, be in reducing the risk rather than prevent-
ing it (Roberfroid, 2000).
Functional food: a European consensus
The unique features of a functional food are (Bellisle
et al. 1998; Knorr, 1998):
1. a conventional or everyday food,
M. B. Roberfroid
2. consumed as part of the normal/usual diet,
3. composed of naturally occurring (as opposed to syn-
thetic) components, perhaps in unnatural concen-
trations or present in foods that would not normally
supply them, and
4. having a positive effect on target function(s) beyond
nutritive value/basic nutrition,
5. that may enhance well-being and health and/or reduce
the risk of disease or provide health benefit so as to
improve the quality of life including physical, psycho-
logical and behavioural performances, and
6. have authorized and scientifically based claims.
It is in this general context that the European Commis-
sions Concerted Action on Functional Food Science in
Europe (FUFOSE), which actively involved a large
number of the most prominent European experts in nutri-
tion and related sciences, has been coordinated by the
International Life Sciences Institute (ILSI) Europe. It
developed in early 1996 to reach a European Consensus
on Scientific Concepts of Functional Foods, which was
published in 1999 (Diplock et al. 1999).
As already indicated above, and because functional food
is a concept rather than a well-defined group of food
products, that consensus document proposes a working
definition:
A food can be regarded as functional if it is satisfactorily
demonstrated to affect beneficially one or more target
functions in the body, beyond adequate nutritional
effects, in a way that is relevant to either improved stage
of health and well-being and/or reduction of risk of
disease. A functional food must remain food and it must
demonstrate its effects in amounts that can normally be
expected to be consumed in the diet: it is not a pill or a
capsule, but part of the normal food pattern.
Diplock et al. (1999)
The main aspects of that working definition are:
1. the food nature of functional food not a pill, a cap-
sule or any form of dietary supplement,
2. the demonstration of effects to the satisfaction of the
scientific community,
3. the beneficial effects on body functions, beyond ade-
quate nutritional effects, that are relevant to an
improved state of health and well-being and/or
reduction of risk (not prevention) of disease, and
4. the consumption as part of a normal food pattern.
This definition encompasses all main features of functional
foods identified above, and it is aimed at stimulating
research and development in the field of nutrition so as
to contribute adequately to the scientific knowledge that
will be required to define optimum (optimized) nutrition
by elaborating new dietary guidelines. But it should be
emphasized that a functional food will not necessarily be
functional for all members of the population, and that
matching individual biochemical needs with selected
food component intakes may become a key task as we pro-
gress in our understanding of the interactions between
genes and diet (Kok, 1999).
 European perspectives
From a practical point of view, a functional food can be:
1. a natural food,
2. a food to which a component has been added,
3. a food from which a component has been removed,
4. a food where the nature of one or more components
has been modified,
5. a food in which the bioavailability of one or more
components has been modified or
6. any combination of these possibilities.
Strategy for development of functional food:
functional food science
Being foods, functional foods need to be safe according to
all criteria defined in current food regulations. But in many
cases, new concepts and new procedures will need to be
developed and validated to assess functional food risks.
In Europe, some, but certainly not all, functional foods
will be classified as novel food and, consequently, will
require the decision tree assessment regarding safety that
is described in the EU Novel Food Regulation (European
Commission, 1997).
But that regulation does not concern nutritional proper-
ties or physiological effects of these novel foods. It is
strictly a safety regulation. The requirement for safety is
a prerequisite to any functional food development. Indeed
the risk versus benefit concept, that is familiar to pharma-
cologists developing new drugs, does not apply to func-
tional foods except, maybe, in very specific conditions
for disease risk reduction when the scientific evidence is
particularly strong.
As described in the European consensus document:
The design and development of functional foods is a key
issue, as well as a scientific challenge, which should rely
on basic scientific knowledge relevant to target functions
and their possible modulation by food components.
Functional foods themselves are not universal and a
food-based approach would have to be influenced by
local considerations. In contrast, a science-based
approach to functional food is universal. . . The function-
driven approach has the science base as its foundation
in order to gain a broader understanding of the
interactions between diet and health. Emphasis is then
put on the importance of the effects of food components
on well-identified and well-characterized target functions
in the body that are relevant to well-being and health
issue, rather than, solely, on reduction of disease risk.
Diplock et al. (1999)
By reference to the new concepts in nutrition outlined
above, it is the role of functional food science to stimulate
research and development of functional foods (Fig. 1).
Referring to basic knowledge in nutrition and related
biological sciences, such a development requires the
identification and, at least partly, the understanding of
the mechanism(s) by which a potential functional food
or functional food component can modulate the target
function(s) that is/are recognized or proven to be relevant
to the state of well-being and health, and/or the reduction
of a disease risk. Epidemiological data demonstrating a
S135
Fig. 1. Schematic representation of the strategy for functional food
discovery and development.
statistically validated and biologically relevant relation-
ship between the intake of specific food components
and a particular health benefit will, if available, be
very useful. The conclusion of the first step will be the
demonstration of a functional effect that should serve
to formulate hypotheses to be tested in a new generation
of human nutrition studies aimed to show that relevant
(in terms of dose, frequency, duration, etc.) intake of
the specified food will be associated with improvement
in one or more target functions, either directly or indirectly
in terms of a valid marker of an improved state of well-
being and health and/or reduced disease risk. If well
supported by strong scientific evidence, the conclusion
could be a recommendation for improved or new dietary
guidelines.
The new-generation human nutrition studies should be
hypothesis-driven. But, in many cases, and even though
they will have to follow the same basic rules in terms of
quality of protocol design, quality of data management
and value of statistical analysis, they will differ quite sub-
stantially from what is classically referred to as clinical
studies. The main differences are that nutrition studies
aim at testing the effect of a food as part of the ordinary
diet, that they may concern the general population or
generally large, at-risk target groups, that they are not
diagnostic- or symptoms-based, and that they are not
planned to evaluate a risk versus benefit approach. Most
of these studies will rely on change(s) in validated/relevant
markers to demonstrate a positive modulation of target
functions after (long-term) consumption of the potential
functional foods. A (double) blind-type of design based
on parallel groups, rather than crossing-over, will generally
be appropriate. Finally, the long-term consequences of the
interaction(s) between functional foods and body func-
tion(s) will have to be monitored carefully.
Markers: a key to the development of functional foods
The development of functional foods will, in most cases,
rely on measurements of markers. These markers need
to be identified and validated for their predictive value of
potential benefits to a target function or the risk of a par-
ticular disease. Markers of correlated events are indi-
cators, whereas markers representing an event directly
involved in the process are factors (Diplock et al.
1999). When related to the risk of a disease, indicators
S136 M. B. Roberfroid
and even factors might, in some instances, be equivalent to
surrogate markers, defined as a biological observation,
result or index that predicts the development of a chronic
disease (Keystone, 1997). The more is known about the
mechanisms leading to health outcomes, the more refined
will be the identification of the markers and their appreci-
ation. The markers should be feasible, valid, reproducible,
sensitive and specific. They can be biochemical, physio-
logical, behavioural or psychological in nature. But
dynamic responses might be as useful as, or more useful
than, static or single point measurements. In many cases,
a battery of markers might be needed in order to create a
decision tree from multiple tests.
These markers, most of which still need to be identified
and validated (Fig. 2), should relate to:
1. exposure to the food component under study by
measuring the serum, faecal, urine or tissue level of
the food component itself or its metabolite(s), or the
concentration of an endogenous molecule that is influ-
enced directly by consumption of the food component;
2. target function(s) or biological response, such as the
change in serum or other body fluids of the concen-
tration of a metabolite, specific protein, enzyme or
hormone, etc. (these first two markers are either indi-
cators or factors);
3. an appropriate endpoint of an improved state of well-
being and health and/or reduction of a disease risk
(such a marker is likely to be a factor rather than an
indicator); and
4. individual susceptibility or genetic polymorphism con-
trolling the metabolism and/or the effect of the food
component under study (Kok, 1999).
To develop these markers further, a state-of-the-art
literature review will be necessary to identify, define and
characterize potential markers. Furthermore, the basic
scientific knowledge underpinning these markers should
be evaluated. The next step will include the assessment
of their relevance to physiological functions, well-being
and health, and eventually disease risk. A validation will
then be necessary both for the methodology and the bio-
logical relevance. Finally, the classification as indicator
or factor will be made and potential dietary modulations
will be demonstrated. New techniques such as those used
by molecular and cellular biologists will be useful in
Fig. 2. Schematic representation of the markers that can be used to discover and
develop new functional foods.
identifying target groups who could benefit from the con-
sumption of specific functional foods.
Functional foods and claims: a communication and
scientific challenge
As stated in the European consensus on scientific concepts
of functional foods:
As the relationship between nutrition and health gains
public acceptance and as the market for functional foods
grows, the question of how to communicate the specific
advantages of such foods becomes increasingly
important. Communication of health benefits to the
public, through intermediates such as health
professionals, educators, the media and the food industry,
is an essential element in improving public health and in
the development of functional foods. Its importance also
lies in avoiding problems associated with consumer
confusion about health messages. Of all the different
forms of communication, those concerning claims
made either directly as a statement on the label or
package of food products, or indirectly through
secondary supporting information remain an area of
extensive discussion.
Diplock et al. (1999)
It is also the opinion of Hudson (1994) that: the links
between nutrition science and food product development
will flow through to consumers only if the required com-
munication vehicles are put in place. But the communi-
cation of health benefits and other physiological effects
of functional foods remains a major challenge because:
1. science should remain the only driving force,
2. messages claims must be based on sound,
objective and appropriate evidence, and
3. evidence must be consistent, able to meet established
scientific standards and plausible.
Moreover, communication in nutrition generally comes from
multiple sources that are sometimes contradictory, creating
an impression of chaos. And chaotic information often
generates ignorance and easily becomes misinformation.
Regarding functional foods, claims associated with
specific food products are the preferable means of commu-
nicating to consumers. In application of the fundamental
 European perspectives
principles, any claim must be true and not misleading, it
must be scientifically valid, unambiguous and clear to the
consumer. These basic principles should be safeguarded
without, however, becoming a disincentive to the produc-
tion of functional foods or to the acceptance of these
foods by consumers. Even though a general definition of
claim as widely accepted in the field of nutrition is:
any representation, which states, suggests or implies that
a food has certain characteristics relating to its origin,
nutritional properties . . . or any other quality (Codex
Alimentarius, 1991), one of the difficulties in communicat-
ing the benefits of functional foods is that distinct types
of claim exist, and that especially the term health
claim, traditionally used to communicate the benefits of
foods, is defined differently in different part of the world.
Seeking for clarity, Codex Alimentarius (1991) has classi-
fied and defined four different categories of claim, but
excluding the term health claim, as:
1. relating to dietary guidelines,
2. relating to nutrient content,
3. being comparative (reduced, less, more) and
4. describing nutrient function (contains . . . , that contri-
butes to the development of . . .).
These claims refer to known nutrients and their role in
growth, development and normal function as well as to
the concept of adequate nutrition. They are based on estab-
lished, widely accepted knowledge but they do not refer to
a particular effect over and above that expected from con-
suming a balanced diet. These claims are thus not really
helpful to communicate the specific benefits of functional
foods. Indeed the claims for functional foods should be
based on the scientific classification of markers (indicators
and/or factors) for target functions and on the effects on
these markers. If such an effect, which goes beyond what
could be expected from the established role of diet in
growth, development and other normal function in the
body, concerns a target function or a biological activity
without direct reference to a particular disease or patho-
logical process, claims will be made for enhanced func-
tion. But, if the benefit is clearly a reduction of the risk
of a disease or pathological process, claims will be made
for a disease risk reduction. These two types of claim,
which are specific for functional foods, are the type A
and type B claims, respectively, as they are described
in the European consensus on scientific concepts of func-
tional foods (Diplock et al. 1999). The type A claim is
similar to the structure function claim, whereas the
type B claim can be regarded as equivalent to the health
claim in the USA. Type B claim also corresponds to
health claim in Sweden (Swedish Nutrition Foundation,
1996). In its last proposed draft recommendations for the
use of health claims, Codex Alimentarius (1999) has
included type A and type B claims and defined them as
follows.
1. Type A or claims that concern specific beneficial
effects of the consumption of foods and their con-
stituents on physiological or psychological functions
or biological activities but do not include nutrient
function claims. Such claims relate to a positive
S137
contribution to health or to a condition linked to
health, to the improvement of a function or to modify-
ing or preserving health.
2. Type B or risk of disease reduction claims that con-
cern the reduction of a disease risk related to the con-
sumption of a food or a food constituent in the context
of the daily diet that might help reduce the risk of a
specific disease or condition.
One of the major issues, still to be resolved especially
with these two types of claim, concerns the biological
level at which evidence can be accepted as satisfactorily
demonstrating an enhanced function or a reduction of dis-
ease risk. This evidence should rely on all data available
that can be grouped in three categories:
1. biological observations,
2. epidemiological data and
3. intervention studies, mostly based on markers.
But for any given specific food product, supporting evi-
dence for enhanced function or reduction of disease risk
might not be available or even not necessary from all
three areas (Diplock et al. 1999). All supporting evidence
should, however, be:
1. consistent in itself;
2. meet accepted scientific standards of statistical as well
as biological significance, especially dose effect
relationship, if relevant;
3. plausible in terms of the relationship between inter-
vention and results, especially in terms of mechan-
ism(s) of action; and
4. provided by a number of sources (including obliga-
torily human studies) that give consistent findings
able to generate scientific consensus.
Food technology and its impact on functional food
development
From the point of view of food processing (Knorr, 1998;
Diplock et al. 1999), the development of functional foods
will often require an increased level of complexity and
monitoring of food processing because
1. new raw materials including those produced by bio-
technologies,
2. emerging thermal and non-thermal technologies,
3. new safety issues and
4. integration throughout the entire food chain, especially
to ensure preservation and/or enhancement of
functionality,
will have to be considered carefully.
The following main areas for technological challenge
have been identified.
1. The creation of new food components in traditional
and novel raw materials that add or increase function-
ality. Examples of such challenges are genetic modifi-
cation, the use of under-utilized or unconventional
natural sources (e.g. algae, seaweed) and the develop-
ment of bioreactors based on immobilized enzymes or
live micro-organisms.
S138 M. B. Roberfroid
2. The optimization of functional components in raw
material and in foods, to ensure maximal preservation
of the component(s), modify their function, increase
their bioavailability, etc. Examples of such challenges
are the development of membrane-processing tech-
niques, the use of controlled and modified atmospheres,
the use of high hydrostatic pressure, high-intensity elec-
tric field pulse technology and ultrasound treatments.
3. The effective monitoring, throughout the entire food
chain, of the amount and functionality of the compo-
nent(s) in raw materials and foods. Examples of such
challenges are the monitoring of microbial viability
and productivity for probiotic functions, the develop-
ment of sensitive markers to record changes in specia-
tion and the interactions with food components during
processing, especially fermentation.
Future trends and conclusions
By reference to the conclusions of the FUFOSE concerted
action (Diplock et al. 1999), future trends are as follows.
1. Components in foods have the potential to modulate
target functions in the body so as to enhance these
functions and/or contribute to reduce the risk of dis-
eases, and functional food science will contribute
to human health in the future provided evidence is
supported by sound scientific (mostly human) data.
2. Nutritionists and food scientists have the possibility,
through the development of functional foods, to offer
beneficial opportunities related to well-being and
health and reduction of the risk of diseases. But the
success of this new approach to nutrition will require
the identification, characterization and development
of methodologies to measure, and the validation of,
relevant markers being indicators or factors to be
used in human nutrition studies. The design of such
studies still needs to be analysed carefully and devel-
oped specifically by reference to, but differently from,
classical clinical studies that have been elaborated to
help in the development of drugs, not food products.
3. Major target functions in the body that are or can be
modulated by specific food products will have to be
identified or characterized. The basic science to under-
stand these functions, and how they relate to well-
being and health or a particular pathological process,
needs to be developed so as to give the necessary
scientific base to develop new functional food
products.
4. Progress in food regulation, which is the means to
guarantee the validity of the claims as well as the
safety of the food, will have to be made.
On the road to optimized nutrition that will be one of the
major challenges of nutrition in the twenty-first century,
functional foods have their own role to play. But the devel-
opment of claims for already existing food products, as
well as the development of new products and their own
claims, should remain first a scientific challenge and not
only a marketing challenge. This is the condition for suc-
cess to the benefit of both human health and the food
industry.
References
Bellisle F, Diplock AT, Hornstra G, Koletzko B, Roberfroid M,
Salminen S & Saris WHM (editors) (1998) Functional food
science in Europe. British Journal of Nutrition 80, Suppl. 1,
S1 S193.
Codex Alimentarius (1991) Codex General Guidelines on Claims,
CAC/GL 1-1979 Revision 1. Geneva: WHO.
Codex Alimentarius (1999) Proposed Draft Recommendation for
the Use of Health Claims. Geneva: WHO.
Diplock AT, Aggett PJ, Ashwell M, Bornet F, Fern EB &
Roberfroid MB (1999) Scientific concepts of functional foods
in Europe: consensus document. British Journal of Nutrition
81, Suppl. 1, S1 S27.
European Commission (1997) Novel Food Directive 97/258/CEE.
Food and Nutrition Board, Institute of Medicine, National Acad-
emy of Sciences (1994) Opportunities in the Nutrition and
Food Sciences [PR Thomas and R Earl, editors]. Washington,
DC: National Academy Press.
Health Canada (1997) Policy Options Analysis: Nutraceuticals/
Functional Foods. Ottawa, Ont., Canada: Health Canada,
Health Protection Branch, Therapeutic Products Programme
and Food Directorate.
Hillian M (1995) Functional foods: current and future market
developments. Food Technology International Europe 25 31.
Hudson CB (1994) The food industrys expectation. In Health
Claims: Substantiation and Research Needs, pp. 9 11. Mel-
bourne, Australia: ILSI Australasia.
IFIC Foundation (1995) Functional foods: opening the door to
better health. Food Insight November/December.
Keystone (1997) The Keystone National Policy Dialogue on Food
Nutrition and Health: Executive Summary. Journal of Nutra-
ceuticals, Functional and Medical Foods 1, 11 32.
Knorr D (1998) Functional food science in Europe. Trends in
Food Science and Technology 9, 295 340.
Kok FJ (1999) Functional foods: relevance of genetic susceptibility.
In Proceedings of Forum on Functional Food, pp. 217 229.
Strasbourg: Council of Europe Publishers.
Milner J (2000) Functional foods: the US perspective. American
Journal of Clinical Nutrition 71, 1654S 1659S.
Riemersma RA (1996) A fat little earner. Lancet 347, 775 776.
Roberfroid MB (2000) Defining functional foods. In Functional
Foods, Concept to Product, pp. 9 28. [GR Gibson and CM
Williams, editors]. Cambridge, UK: Woodhead Publishing
Ltd.
Smith BL, Marcotte M & Harman G (1996) A Comparative
Analysis of the Regulatory Framework Affecting Functional
Food Development and Commercialization in Canada, Japan,
the European Union and the United States of America.
Ottawa, Ont., Canada: Intersector Alliance Inc.
Swedish Nutrition Foundation (1996) Health Claims in the Label-
ling and Marketing of Food Products: The Food Industrys
Rules (Self-Regulatory Programme). Lund: Swedish Nutrition
Foundation.
S132 J. Lucas

Table 1. Projects on functional foods

QLK1-1999-42 Nutritional enhancement of probiotics and prebiotics: technology aspects on microbial viability, stability, functionality
and prebiotic function
QLK1-2001-67 Functional food, gut microflora and healthy ageing
QLK1-1999-76 Conjugated linoleic acid (CLA) in functional food: a potential benefit for overweight middle-aged Europeans
QLK1-2000-86 A process for the assessment of scientific support for claims on foods
QLK1-2000-108 Development and application of high-throughput molecular methods for studying the human gut microbiota in relation
to diet and health
QLK1-1999-124 Functional properties, bioactivities and bioavailability of phytochemicals, especially anthocyanins, from processed
foods
QLK1-2001-135 Functional assessment of interactions between the human gut microbiota and the host
QLK1-2000-146 Probiotic strains with designed health properties
QLK1-2001-173 Local Mediterranean food plants: potential new nutraceuticals and current role in the Mediterranean diet
QLK1-1999-179 European research on functional effects of dietary antioxidants
QLK1-2001-221 Isoflavones for reducing risk of coronary heart disease among postmenopausal women
QLK1-2000-266 The role of dietary phyto-oestrogens in the prevention of breast and prostate cancer
QLK1-2000-324 Barley b-D -glucan and wheat arabinoxylan soluble fibre technologies for health-promoting bread products
QLK1-1999-346 Synbiotics and cancer prevention in humans
QLK1-2000-431 The prevention of osteoporosis by nutritional phyto-oestrogens
QLK1-1999-505 Health implications of natural non-nutrient antioxidants (polyphenols): bioavailability and colon carcinogenesis
QLK1-2000-535 Design of foods with improved functionality and superior health effects using cereal b-glucans
QLK1-2000-563 Probiotics and gastrointestinal disorders: controlled trials of European Union patients
QLK1-1999-576 Folate: from food to functionality and optimal health
QLK1-2000-623 Towards a strategy for optimal vitamin D fortification
QLK1-1999-706 Functional foods against colon cancer: development of a genomics- and proteomics-based screening assay
QLK1-1999-888 Nutraceuticals for a healthier life: n 2 3 polyunsaturated fatty acids and 5-methyltetrahydrofolate
QLK1-2001-1179 Molecular analysis and mechanistic elucidation of the functionality of probiotics and prebiotics in the inhibition of
pathogenic micro-organisms to combat gastrointestinal disorders and to improve human health
QLK1-2001-1273 Biosafety evaluation of probiotic lactic acid bacteria used for human consumption
QLK1-1999-1376 Increase in nutritional value of food raw materials by addition, activity or in situ production of microbial nutraceuticals
QLK1-2000-1423 Enhancing the content of beneficial fatty acids in beef and improving meat quality for the consumer
QLK1-2002-2362 Production of CLA-enriched dairy products by natural means
QLK1-2002-2433 Seaweed antioxidants as novel ingredients for better health and food quality
QLK1-2002-2453 Improving health through dietary phyto-oestrogens: a pan-European network on consumers issues and opportunities
for producers
QLK1-2001-70508 Development of new innovative functional foods containing microcrystalline chitosan
QLK1-2002-71361 Development of new food additives extracted from the solid residue of the tomato processing industry for application in
functional foods
QLK1-2002-71714 Developing lignan-enriched functional food from linseed
QLK1-2002-72376 Microencapsulation of probiotic products for human and animal consumption

CLA, conjugated linoleic acid.

on the consumer acceptance of healthy foods and on
functional food science.
Food and nutrition science is dealt with in Key Action 1,
Food, Nutrition and Health, of the specific programme
Quality of Life and Management of Living Resources of
the Fifth Framework Programme (1998 2002). The calls
under FP5 have resulted in thirty-three projects with a con-
tribution of e51 million. The names and contract numbers
of the projects can be read from Table 1.
Under the Sixth Framework Programme (2002 2006),
research projects on functional foods can be supported in
the Thematic Priority Food Quality and Safety, for
which e685 million have been earmarked in total.