REVIEW

URRENT
C
OPINION

Mediterranean diet and cognitive health: an update

of available knowledge
Catherine Feart, Cecilia Samieri, and Pascale Barberger-Gateau

Purpose of review
Nutrition constitutes an interesting approach for the prevention of age-related brain disorders. The objective
of this review was to examine the most recent evidence on the association between adherence to a
Mediterranean diet (MeDi) and cognitive health among elderly individuals.
Recent findings
Based on available epidemiological studies, two meta-analyses published in 2013 have underlined a
protective effect of a greater MeDi adherence on cognitive health, including a reduced risk of Alzheimers
disease and cognitive impairment. Since then, six additional studies, from longitudinal cohorts or post-hoc
analyses of randomized controlled trials conducted in the USA and Europe, have been published and
provided mixed results. Potential reasons for such discrepancies include methodological limitations inherent
to observational studies, and interactions between diet, environmental factors, such as those enhancing
cognitive reserve, chronic diseases, and genetic factors.
Summary
Overall, available evidence suggests that the MeDi might exert a long-term beneficial effect on brain
functioning. However, more high-powered observational studies with long-term follow-up for cognition and
randomized controlled trials assessing the impact of shifting to a MeDi on cognitive functions are still
needed in various populations.
Keywords
aging, cognitive functions, Mediterranean diet

INTRODUCTION
Progressive cognitive decline and neurodegenerative diseases associated with aging are responsible
for a considerable public health challenge because
of the rapid growth of the proportion of older
people all over the world and the limited efficacy
of pharmacological therapies to date [1]. The prevalence of dementia, whose most frequent cause
is Alzheimers disease, increases with age from
approximately 1.0% in the age group 6569 years
to 30% in the older than 90 years age group [2]. The
management of older people with dementia generates a huge financial burden for society, estimated at
$41 000$56 000 per dementia case per year in 2010
in the USA [3]. Therefore, strategies to prevent or
delay the entry into the neurodegenerative process,
to slow down its progression, and avoid conversion
to dementia must be developed. In this context,
prevention by modifiable lifestyle factors, such as
nutrition, has been suggested as potential effective
strategy [47]. More specifically, there has been a
recent interest in healthy dietary patterns, which

consider the synergistic effect of foods and nutrients

consumed together [8]. Initially described as the
typical dietary pattern of populations living around
the Mediterranean basin [9], the so-called Mediterranean diet (MeDi) has gained increasing interest
thanks to its many benefits on health [10]. Since
the first published benefit of higher MeDi adherence n the risk of Alzheimers disease [11], a rapid
accumulation of knowledge has promoted this
dietary pattern as one of the healthiest for cognitive
health, as confirmed by recent critical reviews
[10,12,13 ,14,15].
&

INSERM, ISPED and University Bordeaux, ISPED, Centre INSERM

U897-Epidemiologie-Biostatistique, Bordeaux, France
Correspondence to Catherine Feart, Equipe Epidemiologie de la nutrition
et des comportements alimentaires, INSERM, U897, Universite de
Bordeaux, ISPED, 146 rue Leo-Saignat, CS61292, F-33076 Bordeaux,
Cedex, France. Tel: +33 5 47 30 42 04; fax: +33 5 57 57 14 86; e-mail:
Catherine.Feart@isped.fr
Curr Opin Clin Nutr Metab Care 2015, 18:5162
DOI:10.1097/MCO.0000000000000131

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

www.co-clinicalnutrition.com

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Ageing: biology and nutrition

KEY POINTS

There is a strong biological rationale for a protective
role of the MeDi in brain aging.

Results from previous meta-analyses mostly converge
toward a beneficial effect of the MeDi on cognitive
health, whereas conclusions from last published
longitudinal observational studies are inconsistent.

To date, there is a lack of scientific evidence to ensure
that the MeDi is an optimal dietary strategy of the
prevention of the age-related neurodegeneration.

There is a need for large-scale studies in various
populations with common methodology before
providing convincing conclusions.

The promotion of the healthy MeDi should be
extended, at least for maintaining overall health,
but keeping in mind special features of elderly
individuals.

The purpose of this article was to update

previous reviews with the most recent available
evidence (from 2013 to mid-2014) from longitudinal epidemiological studies and randomized controlled trials (RCTs) addressing the relationship
between MeDi adherence and cognitive health
among elderly people.
First introduced by Ancel Keys in the SevenCountry study, the concept of MeDi was described
as a collection of eating habits traditionally followed by the populations of the Mediterranean
basin [9,16]. This dietary pattern has remained
relatively constant over time and space, and consist
mainly of high consumption of plant foods (i.e.,
fresh or dried fruits and vegetables, legumes, and
cereals), high intake of olive oil as the main source
of added fat, moderate amount of fish, low-tomoderate intake of dairy products, low consumption of meat and poultry and many condiments and
spices, all accompanied by infusions, or wine in
low-to-moderate amounts during meals, always
respecting beliefs of each community [9]. There is
no single MeDi, but a scientific consensus has
now been reached on what constitutes a typical
MeDi today [17]. Beyond dietary habits, the
MeDi is also part of a lifestyle, constitutes a set
of traditional practices, skills, knowledge, and
traditions, from generation to generation, ranging
from the landscape to the table, and providing a
sense of belonging and continuity to the concerned
communities [17]. These features have led to the
inscription of the MeDi on the representative list of
the Intangible Cultural Heritage of Humanity by
UNESCO in 2010.
52

www.co-clinicalnutrition.com

OVERALL HEALTH BENEFITS OF THE

MEDITERRANEAN DIET: ACCRUING
RECENT EVIDENCE
A protective effect of greater MeDi adherence on
general health, and in particular chronic diseases,
has been suggested for a long time. A meta-analysis,
which has been updated with more recent published
studies in 2013, reported that a 2-point increase in
MeDi adherence, assessed by a 10-point scale, was
significantly associated with lower mortality rates
[relative risk (RR) 0.91, 95% confidence interval
(CI) 0.890.93], and reduced risks for fatal and nonfatal cardiovascular diseases (RR 0.90, 95% CI
0.870.92) and cancers (RR 0.95, 95% CI 0.93
0.97) [18 ]. Regarding vascular outcomes, a recent
primary prevention multicenter RCT conducted in
Spain (PREvencion con DIeta MEDiterranea, PREDIMED) reported that individuals at high cardiovascular risk allocated to a MeDi supplemented with
extra virgin olive oil (EVOO) (1 l/week) or mixed
nuts (30 g/day of walnuts, almonds, and hazelnuts)
for 4.8 years had a lower risk of stroke compared
with a low-fat diet [19 ]. Finally, the MeDi has also
been suggested as a protective dietary approach
against the risk of type 2 diabetes mellitus [20].
&

&&

RATIONALE FOR A PROTECTIVE ROLE OF

THE MEDI IN BRAIN AGING
The potential benefit of a greater MeDi adherence
on cognitive health has only recently emerged,
although this dietary pattern provides foods or
nutrients (i.e., virgin olive oil, provider of monounsaturated fatty acids and polyphenols, fish, rich
in long-chain v-3 polyunsaturated fatty acids and
vitamin D, fruits and vegetables, rich in antioxidants, vitamins C and E, carotenoids, folate, and
polyphenols) that alone may contribute to delay
age-related cognitive decline [2123]. Altogether,
these nutrients could have a beneficial impact on
cognition through various mechanisms including
vascular, antioxidant and anti-inflammatory pathways [24]. For instance, the MeDi provides longchain v-3 polyunsaturated fatty acids, which have
well documented anti-inflammatory effects via their
derivatives (i.e., eicosanoids such as prostaglandins
and leukotrienes and docosanoids such as neuroprotectin D1) [25,26]. By providing foods rich in
powerful antioxidants, the MeDi may contribute to
decrease oxidative stress and lipid peroxidation, as
indicated by reported associations between higher
MeDi adherence and lower plasma F2-isoprostane
levels and higher ascorbic acid concentrations
[27,28]. A favorable effect of the MeDi on cardiovascular risk factors (reduced total cholesterol and
low-density lipoproteins) has also been reported in
Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Mediterranean diet, cognitive decline and dementia Feart et al.

seven RCTs conducted in healthy adults and adults

at high risk of cardiovascular diseases [29]. In
addition, a meta-analysis of prospective studies
and RCTs including more than 530 000 individuals
has shown that MeDi adherence was associated with
reduced risk of metabolic syndrome [30], itself
associated with dementia risk, in particular vascular
dementia [31]. Altogether, these data suggest that
the MeDi could be a particularly promising dietary
pattern for lowering dementia risk, by direct or
indirect ways.

LAST PUBLISHED LONGITUDINAL STUDIES

IN THE FIELD
Since 2006 and the first large-scale evidence of a
protective association between MeDi adherence and
the risk of Alzheimers disease among participants
from the Washington Heights-Inwood Columbia
Aging Project [11], many longitudinal studies have
been published, including six studies on the association between MeDi adherence and cognitive
decline [3237], five studies on the incidence of
mild cognitive impairment (MCI) [3842] and three
studies on the risk of Alzheimers disease or dementia [32,40,43]. To help in summarizing the abundant
existing literature, three reviews [13 ,14,44] and
two meta-analyses have already been conducted
[45 ,46 ].
The first meta-analysis published in 2013
focused on cognitive impairment as outcome,
defined as mild or severe cognitive performances,
risk of MCI or Alzheimers disease, and including
eight casecontrol, cross-sectional, and longitudinal studies [45 ]. Higher adherence to a MeDi was
significantly inversely associated with lower risk of
cognitive impairment (pooled effect, RR 0.60, 95%
CI 0.430.83). The magnitude of this association
was slightly decreased when moderate MeDi adherence was considered (pooled effect, RR 0.79, 95%
CI 0.620.95). Similar findings were reported with
Alzheimers disease, despite some heterogeneity in
the methods of diagnosis across eligible studies
[45 ]. The second meta-analysis examined the
association between MeDi adherence and cognitive
health focusing on incident cognitive impairment,
which included any incident outcomes (either MCI
or Alzheimers disease in cognitively normal subjects, or Alzheimers disease in MCI subjects), risk of
MCI from cognitively normal subjects at baseline,
and risk of Alzheimers disease from cognitively
normal subjects at baseline [46 ]. This metaanalysis reported that the highest MeDi tertile, compared with the lowest tertile, was associated with
significantly reduced risks of MCI, Alzheimers disease, and cognitive impairment (hazard ratio 0.73,
&

&&

&&

&&

95% CI 0.560.96; hazard ratio 0.64, 95% CI 0.46

0.89 and hazard ratio 0.67, 95% CI 0.550.81
respectively), with only a dose-response effect
regarding the risk of cognitive impairment [46 ].
These findings add to the growing body of evidence
of a protective effect of the MeDi on cognitive
health, despite the small number of studies included
in both meta-analyses (no more than five longitudinal cohort studies).
Six additional longitudinal studies, not included
in previous meta-analyses, have been published
during the last 18 months [33,34,36,37,39,40].
These studies, including four reports from the
USA and two from Europe, are described in detail
here, and summarized in Table 1.
&&

Reports from the USA

Four studies have recently reported the association
between MeDi adherence and cognitive performances over time among US people, as described
below.
Womens Health Study
The relationship between adherence to the MeDi
and cognitive change was examined in an ancillary
cognitive substudy of the Womens Health Study
(an RCT for the primary prevention of cardiovascular diseases and cancer with low-dose aspirin and
vitamin E supplements in women; N 6174, ages
70 years and older from the cognitive substudy were
analyzed) [33]. An average 5.6 years after the dietary
survey, which served to ascertain MeDi adherence, a
cognitive examination was performed by telephone.
Two repeated visits were conducted at approximately 2-year intervals. There was no significant
association between higher adherence to the MeDi
and the trajectories of cognitive change or averaged
cognition in this sample of older women. The
specific characteristics of the population studied,
with a high socioeconomic level and a narrow range
of score of MeDi adherence in this US-based population, might in part explain these null findings.

&&

&&

Nurses Health Study

The association between long-term MeDi adherence
since midlife and cognitive decline in late life was
also investigated in a very large sample of US women
enrolled in the cognitive substudy of the Nurses
Health Study (N 16 058, ages 70 years and older
and free of stroke at the time of first cognitive
examination) [34]. The long-term MeDi adherence
was assessed from repeated dietary surveys conducted every 4 years for 13 years, on average.
Cognition was assessed by telephone four times at
2-year intervals, using a validated cognitive battery.

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

www.co-clinicalnutrition.com

53

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

54

www.co-clinicalnutrition.com
13 years on average

16 058 women, mean

(SD) age 74.3 (2.3)
years at baseline

Nurses Health Study,

USA, Samieri et al.
[34]

Follow-up (mean (SD)

and/or range)
4 years on average

Participants

Womens Health Study, 6174 women, mean

USA, Samieri et al.
(SD) age 72.0 (4.1)
[33]
years at baseline of
the RCT (dietary
assessment)

Study, location,
authors (reference)

Semiquantitative FFQ
assessed at midlife in
1984, 1986, 1990,
1994, and 1998

Semiquantitative FFQ to
compute the alternate
MeDi adherence score
Five quintiles were
considered

MeDi adherence:
method of assessment
Adjustment variables

Main findings

Composite scores of global Age, education, long-term No association between

alternate MeDi adherphysical activity and
and verbal memory
ence and change in cogenergy intake, BMI,
assessed by a battery of
nition over time in the
six tests (TICS, immediate smoking, multivitamin
TICS (P for trend 0.37),
use, history of depresand delayed recalls of
global cognitive score (P
sion, diabetes, hypertenthe East Boston Memory
for trend 0.33) and
sion, hypercholesteroleTest, delayed recall of
mia, and myocardial
the TICS 10-word list,
verbal memory (P for
infarction
category fluency and
trend 0.36)
digit span backward)
(four repeated measures
between 1995 and
2001)

The overall cognitive status

(averaged measure of
global cognitive function
and of verbal memory)
was not significantly
associated with alternate
MeDi score (P for
trend 0.63 and 0.44,
respectively)

No evidence of an associComposite scores of global Age, race, education,

ation between greater
income, RCT treatment
cognition and verbal
alternate MeDi score
arm, regular vigorous
memory assessed by a
and better trajectories of
battery of five tests (TICS, exercise, BMI, smoking,
global cognition and verdiabetes, history of
immediate and delayed
bal memory (P for quindepression, hypertenrecalls of the East Boston
tiles medsion, hypercholesteroleMemory Test, delayed
ian  time 0.26 for 4mia, and
recall of the TICS 10year trajectories of mean
postmenopausal horword list, and category
repeated global cognimone use
fluency) (three repeated
tive scores and P for
measures at approxiquintiles medmately 2-year intervals)
ian  time 0.40 for 4year trajectories of mean
repeated verbal memory
scores)

Outcomes

Table 1. Adherence to a Mediterranean-like diet and cognitive health: summary of main longitudinal studies published from 2013 to mid-2014

Ageing: biology and nutrition

Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Cache County Study on 3831 community-based 10.6 years for the whole
Memory, Health and
nondemented individcohort, no data for the
Aging, USA, Wenuals ages 65 years
subsample studied
green et al. [36]
and older at baseline

Self-administered FFQ
assessed at baseline in
1995 and caloricderived residuals
Five quintiles were
considered

Long-term alternate MeDi

score was computed as
the mean of scores
across all dietary assessments. Five quintiles
were considered

3MS assessed four times

during follow-up

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

(Continued )

These differences were

maintained over time
and the rates of 3MS
change over time were
not significantly different
across the quintile
groups

Age, sex, BMI, frequency At baseline, higher MeDi

scores were associated
of moderate physical
with higher mean 3MS
activity, multivitamin and
scores. Participants
mineral supplement use,
belonging to the quintile
history of smoking and
five of MeDi score have
drinking, history of diamean performances of
betes, heart attack, and
0.94 (SE 0.29) comstroke
pared with those of
first quintile (P for trend
0.002)

The overall cognitive status

was significantly higher
with higher long-term
alternate MeDi adherence. Participants
belonging to the quintile
five of alternate MeDi
adherence score have
mean performances of
0.06 (95% CI 0.01
0.11) in the TICS (P for
trend 0.004), 0.05
(95% CI 0.010.08) in
the global score of cognition (P for
trend 0.002) and
0.06 (95% CI 0.03
0.10) in the verbal memory score (P for trend
<0.001) compared with
those of first quintile

Mediterranean diet, cognitive decline and dementia Feart et al.

www.co-clinicalnutrition.com

55

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

56

REGARDS, USA, Tsivgoulis et al. [39]

Study, location,
authors (reference)

Table 1 (Continued)

17 478 communitybased nondemented

individuals ages 45
years and older at
baseline

Participants

MeDi adherence:
method of assessment
Outcomes

Adjustment variables

Main findings

www.co-clinicalnutrition.com

High MeDi adherence

was associated with
lower risk of incident
cognitive impairment in
the nondiabetic population (OR 0.81, 95%
CI 0.700.94,
P 0.007) and not in
the diabetic population
(OR 1.27, 95%
CI 0.951.71,
P 0.106)

Interaction with diabetes:

Age, race, sex, region of Higher MeDi adherence

4.0 years (1.5) on average Self-administered FFQ and 1248 incident cognitive
impairment cases,
residence, BMI, waist cir- was associated with
caloric-derived residuals
lower likelihood of inciassessed by the Six-Item
cumference, household
to compute the MeDi
dent cognitive impairScreener
income, education,
score (09 point scale)
ment in partly adjusted
smoking status, alcohol
Two categories:
model (OR 0.87, 95%
use, physical activity
Low adherents
CI 0.761.00,
level, history of heart dis(scores 04)
P 0.046).
ease, diabetes, atrial
High adherents
This relationship was attefibrillation, SBP, DBP,
(scores 59)
nuated after including
high cholesterol, antihydata from cognitive
pertensive regimen, perassessments following
ceived general health,
incident stroke, becomand depressive symping no significant in fully
toms
adjusted models
(OR 0.89, 95% CI
0.771.01, P 0.07)

Follow-up (mean (SD)

and/or range)

Ageing: biology and nutrition

Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

1038 nondemented
12 years on average
men, ages 70 years
and older at baseline

3083 participants of
13 y on average
the SU.VI.MAX RCT,
mean age (SD) 52.0
(4.6) at baseline
(dietary assessment)

Uppsala Longitudinal
Study, Sweden,
Olsson et al. [40]

SU.VI.MAX, France,
Kesse-Guyot
et al. [37]

24-h dietary record, six

records per year

7-day food record to

compute the MeDi
score (08 points)
Three categories were
considered:
Low adherents
(scores 02)
Medium adherents
(scores 35)
High adherents
(score 6 and over)

No association between
MeDi adherence (1
SD) and risk of AD
(HR 1.00, 95% CI
0.751.33) or all-type
dementia (HR 0.94,
95% CI 0.751.18), nor
all-type cognitive impairment (OR 0.82, 95%
CI 0.651.05) in fully
adjusted models
No association between
MeDi adherence considered as categorical
variable and risk of AD,
all-type dementia or
all-type cognitive
impairment

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

(Continued )

No better cognitive perAge, sex, education, folCognitive performances

formances (composite
low-up time between
assessed once by episocognitive score) with
baseline and cognitive
dic memory (48 cued
higher MeDi adherence
evaluation, number of
recall test), forward and
in fully adjusted models
24-h dietary records,
backward digit span,
(P for trend 0.27 and
RCT treatment arm,
trail making test, seman0.12, respectively)
energy intake, BMI,
tic fluency task, phoneoccupational status,
mic fluency task, and
smoking, physical
used to compute a comactivity, memory difficulposite cognitive score
ties at baseline, depressive symptoms, incidence
of diabetes, hypertension, and cardiovascular
disease during follow-up

Incident AD (n 84), inci- Energy, education, ApoE,

dent all-type dementia
living alone, smoking,
(n 143) and incident
and physical activity
all-type cognitive impairment (n 198)

Mediterranean diet, cognitive decline and dementia Feart et al.

www.co-clinicalnutrition.com

57

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

58

www.co-clinicalnutrition.com

Participants

Follow-up (mean (SD)

and/or range)
MeDi score in three
categories:
Low adherents
(score <4)
Medium adherents
(scores 45)
High adherents
(score >5)
MSDPS in three categories
(tertiles):
Low adherents
(score <20.09)
Medium adherents
(scores 20.0925.21)
High adherents
(score >25.21)

MeDi adherence:
method of assessment
Outcomes

Adjustment variables

A lower MeDi adherence

was associated with
poorer backward digit
span performance (P for
trend 0.03) and lower
MSDPS was associated
with poorer phonemic
fluency performance
(P for trend 0.048)
Significant interaction with
occupational status: manual laborers with poorer
MeDi score had lower
composite cognitive
score (mean difference
5.41, 95% CI 9.18
to 1.11, P for
trend 0.01). No similar
significant interaction
was detected with
MSDPS

Main findings

AD, Alzheimers disease; ApoE, apolipoprotein E; CI, confidence interval; FFQ, Food Frequency Questionnaire; HR, hazard ratio; MeDi, Mediterranean diet; MSDPS, Mediterranean-style dietary pattern score; OR, odds
ratio; RCT, randomized controlled trial; REGARDS, Reasons for Geographic and Racial Differences in Stroke; SD, standard deviation; SE, standard error; SU.VI.MAX, Supplementation with Vitamins and Mineral
Antioxidants; TICS, telephone interview for cognitive status; 3MS, Modified Mini Mental State Examination.

Study, location,
authors (reference)

Table 1 (Continued)

Ageing: biology and nutrition

Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Mediterranean diet, cognitive decline and dementia Feart et al.

Main outcomes of this substudy were composite

scores of global cognition and verbal memory. After
multivariate adjustment, there was no significant
association between long-term MeDi adherence and
cognitive change for 6 years. However, a modest but
significant benefit of high MeDi adherence was
observed on overall cognitive status, defined as
the average of the four repeated measures of cognitive function, at older ages. Despite some limitations, including a sample limited to women with a
relatively high socioeconomic status and a low average cognitive decline, which may have limited the
ability to detect associations with cognitive change,
these results suggest a protective impact, although
relatively modest, of the long-term MeDi adherence
on cognitive status.
Cache County Study on Memory, Health,
and Aging
The Cache County Study on Memory is a large
population-based prospective study of the prevalence and incidence of dementia in elderly
residents of Cache County, Utah, USA [36]. Among
5092 participants, ages 65 years and older at baseline, 3580 participants without dementia completed
the baseline interview and a self-administered Food
Frequency Questionnaire. These data enabled us to
assess the adherence to a MeDi and test its association with changes in Modified Mini-Mental State
Examination (3MS) performances, assessed four
times during 10.6 years on average. At baseline, after
adjustment for sociodemographic and clinical
characteristics, higher MeDi scores (quintiles 25),
compared with quintile 1, were significantly associated with higher 3MS performances. Regarding
changes in 3MS performances over time, they were
not statistically different across quintiles of MeDi
adherence groups. These findings suggest that the
benefit of higher MeDi adherence observed at baseline would be maintained over time, but were not
cumulative over a long-term period.
Reasons for Geographic and Racial
Differences in Stroke Study
Among 17 478 participants from the Reasons for
Geographic and Racial Differences in Stroke study
(REGARDS, conducted in the USA), MeDi adherence
was related to cognitive status [39]. Participants
(ages 64.4 years on average at baseline) were followed up for 4 years on average and incident cognitive impairment was identified among 1248
individuals. In a model adjusted for demographic,
environmental, and vascular risk factors, no association was observed between baseline MeDi adherence and risk of incident cognitive impairment
over time. Interestingly, a statistically significant

interaction with diabetes was reported in the

relationship between MeDi adherence and cognitive
impairment. A higher adherence to the MeDi was
associated with a 19% lower risk of incident cognitive impairment in the nondiabetic population
(95% CI 30% to 6%, P 0.007); in contrast, there
was no association among diabetic participants.
Findings from this very large population-based
cohort with relatively short follow-up were intriguing and suggested that the prevention of cognitive
impairment in population at higher risk of dementia, such as diabetic patients, appeared more challenging than in healthy individuals.

Reports from Europe

Only two European studies have focused on the
association between MeDi adherence and cognitive
decline or dementia in the last 18 months, as
described below.
Uppsala Longitudinal Study
In Sweden, a Mediterranean-like dietary pattern has
been computed from a 7-day food record, among
1038 elderly men, participants of the Uppsala Longitudinal cohort, ages 71 years on average at baseline
[40]. Among them, 84 incident cases of Alzheimers
disease, 143 incident cases of all-type dementia and
198 incident cases of all-type cognitive impairment
were identified during 12 years of follow-up on
average (median follow-up 11.6 years). Overall, no
association was reported between MeDi adherence,
considered as a continuous or categorical variable,
and the development of Alzheimers disease or alltype dementia. A modest protective effect of a greater
MeDi adherence was observed against the development of all-type cognitive impairment, in a model
adjusted for energy intake, whereas this association
became no-significant in the fully adjusted model,
including energy intake, education, genotype of
apolipoprotein E (APOE), living alone, smoking,
and physical activity. In sensitivity analyses excluding inadequate reporting of energy intake, a significant inverse association was observed between
MeDi adherence and all-type cognitive impairment
(N 564 men, among whom 105 incident cases of alltype cognitive impairment). Therefore, a major limit
of this study is the highly selected group enrolled in
the sensitivity analysis that precludes the generalization of these positive findings.
Supplementation with Vitamins and
Mineral Antioxidants Study
In France, the Supplementation with Vitamins and
Mineral Antioxidants (SU.VI.MAX) RCT was implemented among healthy men and women between

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

www.co-clinicalnutrition.com

59

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Ageing: biology and nutrition

1994 and 2002 to evaluate the effect of daily supplementation with antioxidants and minerals on the
incidence of cancer and ischemic heart disease [37].
The baseline dietary survey was used to ascertain
MeDi adherence. At the end of the RCT, some
participants (N 3083, ages 65.4 years on average)
completed a single neuropsychological evaluation
(on average 13 years after the initial enrollment in
the trial). Overall, no association between midlife
MeDi adherence and cognitive performance at
the end of follow-up was reported, except for some
isolated tests. The authors also tested the hypothesis
that cognitive reserve markers, that is, education
and occupation, might modify the relationship
between the MeDi and cognition. Participants with
manual occupation and low MeDi adherence have
lower composite cognitive performance, whereas no
significant interaction between education and MeDi
adherence with regard to cognitive function was
observed. In this substudy, cognition was assessed
at a single time point, so the association between
MeDi adherence and cognitive decline could not be
investigated. However, it suggested that some lifestyle characteristics may interfere on the relationship between MeDi adherence and cognitive health.

RANDOMIZED CONTROLLED TRIAL: MEDI

AS NUTRITIONAL INTERVENTION
To answer the issue of residual confounding in
observational studies, RCTs are the gold standard,
but specially difficult to implement in nutritional
epidemiology and even more in the field of dementia, which requires a long period. Nevertheless,
among participants of the PREDIMED RCT, 522
from the Navarra center (74.6 years on average) were
enrolled in a substudy about cognition at the end of
the nutritional intervention [47 ]. Two cognitive
tests [Mini Mental State Examination (MMSE) and
Clock Drawing Test (CDT)] were administered once
after a mean follow-up of 6.5 years, which unfortunately prevented us from studying the cognitive
decline according to the arm of the intervention.
In multivariate analyses, the mean cognitive
scores were higher for participants allocated to the
MeDi EVOO or MeDi mixed nuts groups than for
participants allocated to the arm low-fat diet (i.e.,
control diet). In a second report addressing the issue
of dietgene interactions, both MeDi allocations
have been combined as a single group of intervention, and genotypes of several susceptibility
genes (i.e., APOE, CLU, CR1, PICALM) associated
with the risk of Alzheimers disease have been determined [48 ,49,50]. Overall, significant interactions
between MeDi and either APOE or CLU on MMSE
performance have been highlighted, whereas no
&

&

60

www.co-clinicalnutrition.com

significant interaction between MeDi and each

genetic polymorphism was observed on CDT performance. A beneficial effect of the MeDi intervention, compared with low-fat diet, on MMSE
performances was observed among subjects carrying
the T minor allele on CLU; which has been previously associated with protective effect on risk of
Alzheimers disease. Regarding APOE polymorphism, the MeDi intervention had a beneficial effect
on MMSE for both non-APOE4 and APOE4 carriers,
whereas mixed results were observed on CDT performances (i.e., a protective effect was only observed
among non-APOE4 carriers) [48 ]. Altogether, these
results showed that a nutritional intervention was
feasible at the dietary pattern level, and not only at
the single nutrient level. Moreover, it underlined
that the effect on cognition of an intervention with
MeDi would be greater in participants with a favorable genetic profile.
&

DISCUSSION
Despite promising results from meta-analyses that
highlighted a protective effect of a greater MeDi
adherence on cognitive health, recent findings from
longitudinal studies provided mixed results as with
those from secondary analyses of previous RCTs.
Cumulative exposures to healthy diets over the
long-term might exert slow, cumulative beneficial
effects on the brain, as suggested by associations
between the MeDi and overall cognitive status in
older age found in large recent studies. Yet, several
studies from various populations in Europe and the
USA have also failed to evidence associations with
cognitive change, which may suggest biological
effects at best modest. Nevertheless, these results
should be considered with caution as observational
studies are subjected to bias. Indeed, some residual
confounding factors cannot be dismissed, especially
as the MeDi is part of a healthier lifestyle in general,
beyond dietary intake. Indeed, greater adherence to
a MeDi is generally associated with higher education
and socioeconomic status. It is possible that the
MeDi is more beneficial for brain health in low
socioeconomic populations with lower educational
level and less cognitive reserve (as indicated by
French SU.VI.MAX results) [37]; this may partly
explain stronger effects found in New York cohorts
(low socioeconomic populations) [11] and modest
effects found in Harvard cohorts (with generally
higher socioeconomic status) [33,34]. Moreover,
the interaction of MeDi with endogenous (i.e.,
genetic predisposition factor for dementia) or
exogenous (i.e., chronic disease such as diabetes)
factors leads the picture more complex than
expected and underlines that the benefits of the
Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Mediterranean diet, cognitive decline and dementia Feart et al.

MeDi on cognitive health are only evident in

certain subpopulations.
Several limits could also be attributed to the
tools used to assess MeDi adherence, which limited
the generalization of the results and prevented definite conclusions, as previously reviewed [13 ]. A
major limitation of the original MeDi score [9] is
the use of thresholds based on medians of intake of
each MeDi component, which are, per se, population-specific. Therefore, a MeDi score is, by definition, population-specific and cannot be compared
with a MeDi score computed in a different sample.
This may have led to misclassifications, as low consumers from one cohort could be considered as high
consumers in another cohort for a particular food
group and vice versa [13 ]. An alternate method of
MeDi adherence assessment has been used in the
French SU.VI.MAX study [37]. The major advantage
of the Mediterranean-Style Dietary Pattern Score is
the weighting of the score by the selected food
groups part of the typical MeDi, or not; although
the recent promising proposition of a new method
of computation of a MeDi score, derived from all
available data from the most relevant observational
studies, should still be tested and validated [18 ].
Finally, in a slowly evolving dementia syndrome, imposing a reasonable delay (i.e., at least
several years) between dietary assessment and cognitive evaluation is of utmost importance to avoid
reverse causation. Therefore, midlife is likely to be
the most relevant period to study risk factors for
cognitive decline, although ensuring the stability of
dietary habits over several years, if not decades,
remains an issue. In that sense, the success of the
PREDIMED RCT is to ensure the diet intervention
compliance in a large sample of elderly subjects is
promising and so should encourage the implementation of such an RCT with cognitive health
as a primary outcome.
&

&

&

CONCLUSION
Overall, there is a strong biological rationale for a
protective role of a healthy diet, in particular the
MeDi, rich in fruits, nuts, vegetables, legumes, fish
and seafood, and olive oil, for cognitive health. Yet,
epidemiological evidence of an association between
adherence to the MeDi and better cognitive health
has been mixed, in particular when taking into
account results from the most recent studies. Thus,
more research is clearly needed to better understand
the epidemiological relationship between the MeDi
and brain health, in particular to identify the best
target population for future trials, and the underlying mechanisms. The promotion of a healthy
Mediterranean-type dietary pattern to maintain

general health may be extended to cognitive health

in aging as an even modest positive effect could have
a major public health impact.
Acknowledgements
None.
Financial support and sponsorship
None.
Conflicts of interest
C.F. received fees for conferences from Danone Research
and Nutricia. P.B.-G. reports grants and nonfinancial
support from Danone Research and Vifor Pharma,
personal fees and nonfinancial support from Nutricia,
grants and nonfinancial support from Groupe Lipides et
Nutrition, nonfinancial support from ILSI Europe. C.S.
reported no conflict of interest.

REFERENCES AND RECOMMENDED

READING
Papers of particular interest, published within the annual period of review, have
been highlighted as:
&
of special interest
&& of outstanding interest
1. Hebert LE, Weuve J, Scherr PA, Evans DA. Alzheimer disease in the United
States estimated using the 2010 census. Neurology 2013; 80:17781783.
2. Plassman BL, Langa KM, Fisher GG, et al. Prevalence of dementia in the
United States: the aging, demographics, and memory study. Neuroepidemiology 2007; 29:125132.
3. Hurd MD, Martorell P, Delavande A, et al. Monetary costs of dementia in the
United States. N Engl J Med 2013; 368:13261334.
4. Daviglus ML, Plassman BL, Pirzada A, et al. Risk factors and preventive
interventions for Alzheimer disease: state of the science. Arch Neurol 2011;
68:11851190.
5. Plassman BL, Williams JW Jr, Burke JR, et al. Systematic review: factors
associated with risk for and possible prevention of cognitive decline in later
life. Ann Intern Med 2010; 153:182193.
6. Barnard ND, Bush AI, Ceccarelli A, et al. Dietary and lifestyle guidelines for the
prevention of Alzheimers disease. Neurobiol Aging 2014; 35 (Suppl 2):S74
S78.
7. Morley JE. Cognition and nutrition. Curr Opin Clin Nutr Metab Care 2014;
17:14.
8. Alle`s B, Samieri C, Feart C, et al. Dietary patterns: a novel approach to
examine the link between nutrition and cognitive function in older individuals.
Nutr Res Rev 2012; 25:207222.
9. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a
Mediterranean diet and survival in a Greek population. N Engl J Med
2003; 348:25992608.
10. Sofi F, Abbate R, Gensini GF, Casini A. Accruing evidence on benefits of
adherence to the Mediterranean diet on health: an updated systematic review
and meta-analysis. Am J Clin Nutr 2010; 92:11891196.
11. Scarmeas N, Stern Y, Tang MX, et al. Mediterranean diet and risk for
Alzheimers disease. Ann Neurol 2006; 59:912921.
12. Feart C, Samieri C, Barberger-Gateau P. Mediterranean diet and cognitive
function in older adults. Curr Opin Clin Nutr Metab Care 2010; 13:1418.
13. Feart C, Samieri C, Alle`s B, Barberger-Gateau P. Potential benefits of
&
adherence to the Mediterranean diet on cognitive health. Proc Nutr Soc
2013; 72:140152.
One of first reviews about the relationship between adherence to a Mediterranean
diet and cognitive health, with a comprehensive description of longitudinal studies
published before 2013 and not described in details in the present article.
14. Lourida I, Soni M, Thompson-Coon J, et al. Mediterranean diet, cognitive
function, and dementia: a systematic review. Epidemiology 2013; 24:479489.
15. Shah R. The role of nutrition and diet in Alzheimer disease: a systematic
review. J Am Med Directors Assoc 2013; 14:398402.
16. Keys A, Menotti A, Karvonen MJ, et al. The diet and 15-year death rate in the
seven countries study. Am J Epidemiol 1986; 124:903915.
17. Bach-Faig A, Berry EM, Lairon D, et al. Mediterranean diet pyramid today.
Science and cultural updates. Public Health Nutr 2011; 14 (12A):2274
2284.

1363-1950 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

www.co-clinicalnutrition.com

61

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Ageing: biology and nutrition

18. Sofi F, Macchi C, Abbate R, et al. Mediterranean diet and health status: an
updated meta-analysis and a proposal for a literature-based adherence score.
Public Health Nutr 2014; 17:27692782.
A meta-analysis of the potential benefits of a Mediterranean diet on survival,
cardiovascular and cancer incidence and/or mortality.
19. Estruch R, Ros E, Salas-Salvado J, et al. Primary prevention of cardiovascular
&&
disease with a Mediterranean diet. N Engl J Med 2013; 368:12791290.
This randomized controlled trial (PREDIMED) showed that a nutritional intervention
was feasible at the dietary pattern level, and not just at the single nutrient level.
20. Schwingshackl L, Missbach B, Konig J, Hoffmann G. Adherence to a
Mediterranean diet and risk of diabetes: a systematic review and metaanalysis. Public health nutrition 2014; 18. [Epub ahead of print]
21. Solfrizzi V, Panza F. Mediterranean diet and cognitive decline. A lesson from
the whole-diet approach: what challenges lie ahead? J Alzheimers Dis 2014;
39:283286.
22. Gillette-Guyonnet S, Secher M, Vellas B. Nutrition and neurodegeneration:
epidemiological evidence and challenges for future research. Br J Clin
Pharmacol 2013; 75:738755.
23. Boeing H, Bechthold A, Bub A, et al. Critical review: vegetables and fruit in the
prevention of chronic diseases. Eur J Nutr 2012; 51:637663.
24. Frisardi V, Panza F, Seripa D, et al. Nutraceutical properties of Mediterranean
diet and cognitive decline: possible underlying mechanisms. J Alzheimers Dis
2010; 22:715740.
25. Feart C, Torres MJ, Samieri C, et al. Adherence to a Mediterranean diet and
plasma fatty acids: data from the Bordeaux sample of the Three-City study. Br
J Nutr 2011; 106:149158.
26. Gu Y, Luchsinger JA, Stern Y, Scarmeas N. Mediterranean diet, inflammatory
and metabolic biomarkers, and risk of Alzheimers disease. J Alzheimers Dis
2010; 22:483492.
27. Gaskins AJ, Rovner AJ, Mumford SL, et al. Adherence to a Mediterranean diet
and plasma concentrations of lipid peroxidation in premenopausal women.
Am J Clin Nutr 2010; 92:14611467.
28. Valls-Pedret C, Lamuela-Raventos RM, Medina-Remon A, et al. Polyphenolrich foods in the Mediterranean diet are associated with better cognitive
function in elderly subjects at high cardiovascular risk. J Alzheimers Dis 2012;
29:773782.
29. Rees K, Hartley L, Flowers N, et al. Mediterranean dietary pattern for the
primary prevention of cardiovascular disease. Cochrane Database Syst Rev
2013; 8:CD009825.
30. Kastorini CM, Milionis HJ, Esposito K, et al. The effect of Mediterranean diet
on metabolic syndrome and its components: a meta-analysis of 50 studies
and 534,906 individuals. J Am Coll Cardiol 2011; 57:12991313.
31. Raffaitin C, Gin H, Empana JP, et al. Metabolic syndrome and risk for incident
Alzheimers disease or vascular dementia: the Three-City Study. Diabetes
care 2009; 32:169174.
32. Feart C, Samieri C, Rondeau V, et al. Adherence to a mediterranean diet,
cognitive decline, and risk of dementia. JAMA 2009; 302:638648.
33. Samieri C, Grodstein F, Rosner BA, et al. Mediterranean diet and cognitive
function in older age. Epidemiology 2013; 24:490499.
34. Samieri C, Okereke OI, Devore E, et al. Long-term adherence to the Mediterranean diet is associated with overall cognitive status, but not cognitive
decline, in women. J Nutr 2013; 143:493499.
35. Tangney CC, Kwasny MJ, Li H, et al. Adherence to a Mediterranean-type
dietary pattern and cognitive decline in a community population. Am J Clin
Nutr 2011; 93:601607.
&

62

www.co-clinicalnutrition.com

36. Wengreen H, Munger RG, Cutler A, et al. Prospective study of dietary

approaches to stop hypertension- and mediterranean-style dietary patterns
and age-related cognitive change: the Cache County Study on Memory,
Health and Aging. Am J Clin Nutr 2013; 98:12631271.
37. Kesse-Guyot E, Andreeva VA, Lassale C, et al. Mediterranean diet
and cognitive function: a French study. Am J Clin Nutr 2013; 97:369
376.
38. Cherbuin N, Anstey KJ. The Mediterranean diet is not related to cognitive
change in a large prospective investigation: the PATH Through Life study. Am
J Geriatr Psychiatry 2012; 20:635639.
39. Tsivgoulis G, Judd S, Letter AJ, et al. Adherence to a Mediterranean diet
and risk of incident cognitive impairment. Neurology 2013; 80:1684
1692.
40. Olsson E, Karlstrom B, Kilander L, et al. Dietary patterns and cognitive
dysfunction in a 12-year follow-up study of 70 year old men. J Alzheimers
Dis 2014. [Epub ahead of print]
41. Roberts RO, Geda YE, Cerhan JR, et al. Vegetables, unsaturated fats,
moderate alcohol intake, and mild cognitive impairment. Dement Geriatr
Cogn Dis 2010; 29:413423.
42. Scarmeas N, Stern Y, Mayeux R, et al. Mediterranean diet and mild cognitive
impairment. Arch Neurol 2009; 66:216225.
43. Scarmeas N, Luchsinger JA, Schupf N, et al. Physical activity, diet, and risk of
Alzheimer disease. JAMA 2009; 302:627637.
44. Di Marco LY, Marzo A, Munoz-Ruiz M, et al. Modifiable lifestyle factors in
dementia: a systematic review of longitudinal observational cohort studies.
J Alzheimers Dis 2014; 42:119135.
45. Psaltopoulou T, Sergentanis TN, Panagiotakos DB, et al. Mediterranean diet
&&
and stroke, cognitive impairment, depression: a meta-analysis. Ann Neurol
2013; 74:580591.
A meta-analysis of the potential benefits of a MeDi on cognitive impairment based
on casecontrol, cross-sectional, and longitudinal studies.
46. Singh B, Parsaik AK, Mielke MM, et al. Association of Mediterranean diet with
&&
mild cognitive impairment and Alzheimers disease: a systematic review and
meta-analysis. J Alzheimers Dis 2014; 39:271282.
A meta-analysis of the potential benefits of a MeDi on cognitive health, including
incident cognitive impairment, risk of MCI, and risk of Alzheimer disease, based on
longitudinal studies.
47. Martinez-Lapiscina EH, Clavero P, Toledo E, et al. Mediterranean diet im&
proves cognition: the PREDIMED-NAVARRA randomised trial. J Neurol
Neurosurg Psychiatry 2013; 84:13181325.
This article described a substudy about cognition implemented in the PREDIMED
randomized controlled trial conducted in Spain. Results suggested that a longterm nutritional intervention with MeDi along with EVOO (1 l/week) for 6.5 years
might have improved global cognition.
48. Martinez-Lapiscina EH, Galbete C, Corella D, et al. Genotype patterns at
&
CLU, CR1, PICALM and APOE, cognition and Mediterranean diet: the
PREDIMED-NAVARRA trial. Genes Nutr 2014; 9:393.
This substudy of the PREDIMED trial suggested that an intervention with a MeDi
modulated the association of genetic risk factors on cognition.
49. Harold D, Abraham R, Hollingworth P, et al. Genome-wide association study
identifies variants at CLU and PICALM associated with Alzheimers disease.
Nat Genet 2009; 41:10881093.
50. Lambert JC, Heath S, Even G, et al. Genome-wide association study identifies
variants at CLU and CR1 associated with Alzheimers disease. Nat Genet
2009; 41:10941099.

Volume 18
Number 1
January 2015

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.