 |
Printable
PDF
Version
Summary
N. Tasnime Akbaraly,1
Henri Faure,2 Veronique
Gourlet,3 Alain Favier,2
and Claudine Berr
1INSERM U888 and
Universite´ Montpellier
1, France.
2De´partement de
Biologie Inte´gre´e, CHU
de Grenoble, France.
3INSERM U708 and
Universite´ Paris 6,
France.
Background.
The hypothesis of
carotenoids having a
preventive role in
cognitive impairment is
suggested by their
antioxidant properties.
Methods. We examined, in
a cross-sectional
analysis, the
relationship between
cognitive performance
(assessed by the
Mini-Mental State
Examination, Trail
Making Test Part B,
Digit Symbol
Substitution, Finger
Tapping Test, and Word
Fluency Test) and
different plasma
carotenoids (lutein,
zeaxanthin, b-cryptoxanthin,
lycopene, a-carotene,
and trans-bcarotene and
cis-b-carotene) in a
healthy elderly
population (the
EVA,‘‘Etude du
Vieillissement Arte´riel,’’
study; n¼589, age ¼ 73.5
6 3 years). Results.
Logistic regression
showed that participants
with the lowest
cognitive functioning
(,25th percentile) had a
higher probability of
having low levels of
specific plasma
carotenoids (,1st
quartile): lycopene and
zeaxanthin. For
zeaxanthin, odds ratios
(ORs) were as follows:
ORDSS¼1.97 (95%
confidence interval
[CI]¼1.21–3.20),
ORFTT¼1.70
(CI¼1.05–2.74), and
ORWFT¼1.82
(CI¼1.08–3.07); for
lycopene, ORDSS¼1.93
(CI¼1.20–3.12) and
ORTMTB¼1.64 (CI ¼
1.04–2.59).
Conclusion. Even if it
is not possible to
affirm if these low
levels of carotenoids
precede or are the
consequence of cognitive
impairment, our results
suggest that low
carotenoid levels could
play a role in cognitive
impairment. The
biological significance
of our findings needs
further research.
AMONG the leading causes
of cognitive impairment,
an increase in brain
oxidative stress is well
documented
(1). In fact, the brain
is particularly prone to
free radical attacks
owing to its relatively
low antioxidant content,
a considerable amount of
polyunsaturated fatty
acid chains in the
neuronal membrane
lipids, and its high
oxygenconsumption rate
(2). The hypothesis of
carotenoids having a
preventive role in
cognitive impairment is
suggested by their
ability to trap peroxyl
radicals and their
singlet oxygenquenching
properties, which
enables them to prevent
lipid peroxidation
(3,4). Epidemiological
studies (5–9) and
clinical trials (10,11)
on cognitive impairment
and plasma carotenoids
mainly concern
b-carotene, which is a
major
carotenoid. Some
biological studies,
however, showed that
antioxidant activity of
other carotenoids could
be more effective than
b-carotene activity
(12,13). This study’s
aim is to examine the
relationship between
cognitive performance
and a large variety of
carotenoids including
xanthophylls (lutein,
zeaxanthin, b-cryptoxanthin)
and carotenes (lycopene,
a-carotene,
trans-b-carotene and
cis-b-carotene)in a
healthy elderly
population.
MATERIALS AND METHODS
Study Population
The EVA (‘‘Etude du
Vieillissement Arte´riel’’)
study isa 9-year
longitudinal study with
six waves of follow-up
(14). During the first 2
years (1991–1993; EVA0),
1389 volunteers (574 men
and 815 women) born
between 1922 and 1932
(mean age ¼ 65 years)
residing in the town of
Nantes (Western France)
were recruited from
electoral rolls, and to
a lesser extent, via
information campaigns.
The sixth and last
follow-up of the EVA
study (EVA6) was
conducted between June
2000 and December 2001.
During the 9-year
follow-up, 101 deaths
occurred. The first
leading cause of death
was cancer (n ¼ 45,
44.5%); the second was
cardiovascular diseases
(n ¼ 22, 21.8%). The
main factors related to
mortality were found to
be, as reported in the
literature: male gender,
smoking (current and
prior), alcohol intake,
medication use, obesity,
diabetes, hypertension,
and cardiovascular
diseases (15). At EVA6,
blood samples after a
12-hour fast were
obtained from 773
participants. Those
participants who did not
complete the whole study
(n¼616, 44.3%) were
significantly more
frequently men (those
who did not complete:
44.3% vs those who
completed: 38.9%,
p¼.04), obese or
overweight (57.9% vs
47.3%, p¼.0004), or
hypertensive (52.6% vs
47.2%, p ¼ .05). They
were statistically more
frequently participants
in the lowest cognitive
performance class (,25th
percentile of the
distribution) for the
Mini-Mental State
Examination (MMSE)
(25.2% vs 18.2%, p ¼
.002), Digit Symbol
Substitution (DSS)
(28.3% vs 20.3%,
p¼.0006), Trail Making
Test Part B (TMTB)
(30.5% vs 21.4%,
p¼.0002), and Word
Fluency Test (WFT)
(25.1% vs 19.3%, p ¼
.009), but we did not
observe that for the
Finger Tapping Test (FTT)
(p ¼ .37). The present
analysis was restricted
to the 589 participants
who underwent at EVA6 a
cognitive evaluation and
blood sampling. The
study protocol was
approved by the Ethical
Committee of the
University Center
Hospital of Kremlin-Biceˆtre,
Paris. Signed informed
consent was obtained
from all participants at
enrollment.
Data Collection
Cognitive evaluation and
depressive
symptoms.—Trained
neuropsychologists
evaluated cognition with
a neuropsychological
battery of tests
including a global test,
the MMSE (16), and an
assessment of a range of
cognitive domains.Visual
conceptual and
visuomotor tracking were
assessed by TMTA and
TMTB (17). Involving
motor speed and
attention functions, the
TMT is highly vulnerable
to the effects of brain
injury (18). Part A is
considered as exploring
motor speed, control,
and working memory,
whereas Part B assesses
executive functioning
such as set shifting.
The variables of
interest are the time in
seconds (19). These
tests (Parts A and B)
were performed with a
maximum allotted time to
perform the test of 180
and 240 seconds,
respectively. When
participants exceeded
the time allotted for
each part, the maximum
allotted time was
imputed. The DSS from
the Wechsler Adult
Intelligence
Scale-Revised (WAIS-R)
measured sustained
attention and logical
reasoning (20). Manual
dexterity and
psychomotor speed were
evaluated
with the FTT. Verbal
fluency was evaluated
with the WFT. Depression
symptoms were assessed
by the Center for
Epidemiological
Studies-Depression (CESD)
scale using score (17
for men and 23 for
women) for high risk of
depression to define
depressive
symptomatology (21).
Carotenoid measurement.—Retetinol
and carotenoids were
measured with a Biotek-Kontron
high-performance liquid
chromatography (HPLC)
system (UVK Lab, Trappes,
France), which consists
of a 525 dual pump, a
465 autosampler, and a
540 diode array
detector. Retinol and
bcarotene were purchased
from Fluka
(Sigma-France, L’Isle
d’Abeau), and other
carotenoids were
provided by Hoffman-La
Roche (Hoffman-La Roche,
Baˆle, Switzerland). The
LC separation was run
with an Alltech
Adsorbosphere C18 column
(15034.5 mm ID, 3 lm
particle size;
Templemars, France),
which was thermostated
at 288C with a 402
column oven. Carotenoids
and retinol were
measured by HPLC after
two extractions with a
hexane–tetrahydrofurane
mixture. For
quantification, we used
the method of Steghens
and colleagues (22) with
minor modifications.
Indeed, we used a single
150
mm-long column instead
of two, and we added 10
ppm water in mobile
phase A to improve the
separation of retinol,
lutein, and zeaxanthin.
The laboratory
participates in the NIST
(National Institute of
Standards and
Technology, New York,
NY) and in the French
Society for Vitamins and
Biofactors (SFVB)
external quality
assurance programs, and
ChromSystems (Munchen,
Germany) internal
controls were analyzed
in every series of
measurements (one
control every 10 unknown
serums). The limits of
detection were
calculated as 5-fold the
maximum baseline noise
in the region of the
peaks. Hence we found
limits of detection of
0.05 lM for
retinol, and 0.02 lM for
the carotenoids. All
concentrations of
retinol, lycopene, and
b-carotene were above
these respective limits,
and only 5% of lutein,
8% of zeaxanthin, and 2%
of b-cryptoxanthin were
under. Total plasma
carotenoid level was
obtained by summing
levels of lutein,
zeaxanthin, b-cryptoxanthin,
lycopene, and and
b-carotenes.
Questionnaire and
medical examination.—The
general questionnaire
allowed us to obtain
information on
sociodemographic factors
such as sex, age, and
educational achievement
plus consumption habits
such as smoking status,
alcohol consumption
(which was determined
from the participant’s
estimated average amount
of alcoholic beverages
ingested weekly), and
medication use. In
addition, height and
weight were measured.
Two independent measures
of systolic and
diastolic blood pressure
were taken with a
digital electronic
tensiometer after a
10-minute rest. Total
plasma cholesterol and
plasma glucose levels
were also measured using
routine methods. The
apolipoprotein Egenotype
of participants was
determined on DNA
samples.
Statistical Methods
The characteristics of
the 589 participants
included in the analysis
were described compared
to the 184 participants
who had blood sampling
but not the cognitive
evaluation at EVA6;
results were expressed
as percentages and means
with their standard
deviation (SD). To test
the differences between
these two groups, the
Chi-square test and the
Student t test were
used. These
characteristics
comprised sex, age,
educational achievement
( primary school vs high
school), smoking status
(current or ex-smokers
vs nonsmokers), alcohol
consumption (, 20 mL/d
vs 20 mL/d), and
medicine use (, 3/d vs
3/d). Health
characteristics were
body mass index (BMI)
classes [underweight:
BMI , 21 kg/m2 (23);
‘‘normal weight’’: 21
BMI , 25; overweight: 25
BMI , 30 (24); obesity
30 kg/m2 (24)], diabetes
(plasma glucose level
7.80 mmol/L or use of
antidiabetic drugs or
diabetes medical
history), dyslipidemia
(total cholesterol 6.20
mmol/L or use of
lipid-lowering drugs or
dyslipidemia medical
history), hypertension
(systolic or diastolic
blood pressure 140 or 90
mmHg, respectively, or
use of hypertensive
drugs or hypertension
medical history),
history of vascular
diseases (self-reported
history of myocardial
infarction, angina
pectoris, stroke, or use
of vascular drugs),
depressive
symptomatology (yes/no),
and apolipoprotein E
genotype (e4 allele
þ/–). To calculate the
correlation between
carotenoids, we
calculated Pearson
correlation coefficients
on log-transformed
carotenoids. The graphic
representation of the
percentage of
participants with low
cognitive functioning
for the different
carotenoid levels showed
that the relationship
between cognition and
these biological
variables was not
linear, so we considered
them to be categorical
variables. We compared
the characteristics of
participants in the
different levels of
carotenoids (, 25th
percentile vs 25th
percentile)
by using the Student t
test or the Chi-square
test for both continuous
and categorical
variables. To define
participants with the
lowest cognitive
performances in this
well-educated cohort, we
chose two cutoffs:
participants who had
cognitive test scores
below the 25th
percentile (and above
the 75th percentile for
the TMTB), and
participants who had
cognitive test scores
below the 10th
percentile (and above
the 90th percentile for
the TMTB). Values for
these 25th and 10th
percentile cutoffs for
each test are presented
in Table 2. Classical
multivariate logistic
regressions were
performed to test
associations between
probability of
participants to have the
lowest cognitive
functioning and levels
of plasma carotenoids (,
25th percentile vs .
25th percentile),
adjusting for all
potential confounding
variables. Results were
expressed as odds ratios
(OR) with their 95%
confidence intervals
(CI). All interactions
between a carotenoid and
each consumption habit
and health variable
werecalculated and were
not statistically
significant. Statistical
analyses were performed
using SAS software
(version 9.1; SAS
Institute, Inc., Cary,
NC).
RESULTS
Characteristics of the
589 EVA Participants The
589 participants in the
EVA study (361 women and
228 men, age ¼ 73.5 6
2.9 years), were well
educated (52.1% had a
high school or superior
degree). Among them,
39.0% were smokers or
ex-smokers, 24% consumed
alcohol regularly ( 2
glasses/day). In
dividing participants
among BMI classes, we
observed 13.8%
underweight, 34.0%
overweight, and 8.5%
obese. Concerning health
status, 8.3% showed
depressive
symptomatology, 8.1%
were diabetics, 65.0 %
were dyslipidemic, 78.8%
had hypertension, 19.7%
had a history of
cardiovascular disease,
and 20.5% carried at
least one allele e4 of
the apolipoprotein E.
For the different
carotenoids and scores
on neuropsychological
tests, concentrations
and ranges are described
in Table 1, and
percentile distributions
in Table 2. We also
compared the
characteristics of these
589 participants
included in the analysis
to the 184 participants
for whom we obtained a
blood sample but not a
cognitive evaluation.
Results showed that
these 184 participants
were significantly older
(74.2 6 3.1 years vs
73.5 6 2.9 years), and
proportions of
participants with
hypertension (p ¼ .05)
or with a history of
cardiovascular disease
(p ¼ .03) were higher.
In contrast, proportions
of participants with low
levels of b-cryptoxanthin
and trans- and cis-b-carotene
were significantly lower
in this group (results
not shown).
Description of
Carotenoids: Correlation
andAssociated Factors
Plasma carotenoids were
highly and significantly
intercorrelated. The
highest correlations
were found among the
following carotenes:
a-carotene/trans-b-carotene
(r¼0.78),trans-b-carotene/cis-b-carotene
(r ¼ 0.61), a-carotene/lycopene
(r¼0.60) and among
lutein/zeaxanthin
(r¼0.58). The other
correlation coefficients
ranged from 0.16 for
lycopene/
b-cryptoxanthin to 0.50
for trans-b-carotene/lycopene.
In Table 3 and Table 4,
we describe the
characteristics of
participants according
to their total plasma
carotenoid, the
different xanthophylls,
and carotenes levels (,
25th percentile vs 25th
percentile). We observe
that the profiles of
factors associated with
a-carotene,
trans-b-carotene, and
cis-b-carotene were
identical. Between the
other carotenoids,
however, the associated
factors can differ.
Gender was associated
with all carotenoids,
which (with the
exception of zeaxanthin)
reach higher levels in
women. Smoking status
and alcohol consumption
were significantly
associated with lower
levels of total plasma
carotenoids, b-cryptoxanthin,
acarotene, and trans-
and cis-b-carotene.
Diabetes was associated
with low levels of all
carotenoids, and
hypertension was
significantly associated
with low levels of
lutein, acarotene, and
trans- and cis-b-carotene.
Being obesity or
overweight was
associated with low
levels of all
carotenoids except for
zeaxanthin and b-cryptoxanthin.
Age, education,
depressive
symptomatology,
dyslipidemia, history of
cardiovascular disease,
and apolipoprotein E
genotype were not
associated with any
plasma carotenoids.
Association Between
Cognition and
Carotenoids Table 5
shows the results of
crude associations
between
cognitive performance
and carotenoid levels,
obtained by univariate
logistic regression
analyses. Participants
with the lowest
cognitive performance
(neuropsychological test
scores , 25th
percentile) had a higher
probability of having
low levels of some
carotenoids (level , 1st
quartile). Significant
associations were
observed between
zeaxanthin and all
cognitive tests except
the MMSE (for the TMTA,
OR ¼ 1.66 [CI ¼
1.08–2.55]; for the TMTB,
OR ¼ 1.60[CI¼1.04–2.44];
for the DSS, OR¼1.87
[CI¼1.21–2.89]; for the
FTT, OR¼1.70
[CI¼1.10–2.62], and for
the WFT, OR ¼ 1.87 [CI ¼
1.16–3.00]). Low levels
of lycopene were
associated with low
performance on the TMTB
(OR ¼ 1.76 [CI ¼
1.16–2.67]) and on the
DSS (OR ¼ 2.02 [CI ¼
1.32– 3.11]). A
significant association
was found between low
performance on the TMTB
and low levels of total
plasma carotenoids and
trans-b-carotene (OR ¼
1.57 [CI ¼ 1.03– 2.40]
and OR ¼ 1.58 [CI ¼
1.04–2.41]). After
taking into account
sociodemographic factors
(sex, age, education),
consumption habits
(tobacco, alcohol),
diabetes, hypertension,
and BMI class (Table 6),
associations between
zeaxanthin and cognitive
tests remained
statistically
significant for the TMTA
(OR¼1.67
[CI¼1.06–2.66]), the DSS
(OR ¼ 1.92 [CI ¼
1.18–3.14]), the FTT (OR
¼ 1.69 [CI ¼ 1.05–2.72])
and the WFT (OR¼1.80
[CI¼1.07–3.05]), but not
for the TMTB (OR ¼ 1.51
[CI ¼ 0.95–2.39]; p ¼
.08).
Lycopene remained
associated with the TMTB
(OR ¼ 1.54 [CI ¼
0.97–2.43]; p ¼ .06) and
the DSS (OR ¼ 1.85 [CI ¼
1.14–2.98]). The other
associations between
carotenoids and
cognitive performance
observed in the crude
analyses did not remain
statistically
significant after
adjustment. Total plasma
carotenoids, a-carotene,
and b-carotene (trans or
cis) levels were not
statistically associated
to low cognitive
performance, nor were
lutein and b-cryptoxanthin.
Sensitivity Analysis
The same analyses were
first performed by
removing participants
with depressive
symptomatology (n ¼ 49),
then by removing
participants who were
underweight (BMI , 21
kg/m2) (n ¼ 81), and
then by adjusting for
levels of plasma
retinol. In the three
situations, we obtained
similar results. We also
performed analyses on
participants who had an
MMSE score 25 (n ¼ 570)
to exclude participants
with potential
clinically significant
cognitive impairment (n
¼ 19). Results were
identical (data not
shown). We also repeated
analyses between
cognitive performances
and carotenoids levels
by defining participants
with the lowest
cognitive performances
as having cognitive test
scores below the 10th
percentile. Only
associations between
zeaxanthin and FTT (p ¼
.003) and TMTB (p ¼
.06), and between
lycopene and DSS (p ¼
.07) remained marginally
statistically
significant (data not
shown). In these
analyses, the number of
participants who had the
lowest cognitive
performance scores was
2.2 to 2.8 smaller than
for the preceding
analyses (n ¼ 54 for the
MMSE, 49 for the DSS, 55
for the TMTA, 53 for the
TMTB, 52 for the FTT,
and 38 for the TEL).
Lack of statistical
power could explain the
drop in statistically
significant results. All
of our analyses were
conducted with
dichotomized variables
for both the plasma
carotenoids and the
cognitive outcomes; in
supplemental analyses,
we tested whether
similar findings were
observed when a
continuous measure for
cognition was used. We
calculated Spearman
correlation coefficients
between cognitive
variables and
log-transformed
carotenoids. Results
showed a significant
association for
zeaxanthin and lycopene
and DSS (r¼0.08, p¼.01
and r¼ 0.11, p ¼ .006
for zeaxanthin and
lycopene, respectively),
TMTA (r ¼ 0.11, p ¼ .01
and r ¼ 0.08, p ¼ .05),
TMTB (r¼ 0.08, p¼.05 and
r¼ 0.12, p¼.004), and
FTT (r¼0.09, p¼.04 and
r¼0.09, p¼.04). These
correlations confirm
that our findings were
not driven by the chosen
dichotomous
classifications for
cognition and carotenoid
levels.
DISCUSSION
To our knowledge, this
study is the first that
investigated, in a
healthy elderly
population, the
relationship between
cognitive performance
measured by five
neuropsychological tests
and the different plasma
carotenoids:
xanthophylls (lutein,
zeaxanthin, b-cryptoxanthin)
and carotenes (lycopene,
a-carotene,
trans-b-carotene, and
cis-b-carotene). The EVA
study included
volunteers with higher
educational status,
higher incomes, and
greater cognitive
function than the
average elderly French
population. Despite this
selection, plasma
carotenoid
concentrations in the
EVA study population
were in the same ranges
as those in different
European or American
populations (4). In this
present study, low
levels of specific
plasma carotenoids—lycopene
and zeaxanthin—were
associated to poor
cognitive functioning in
a highly educated,
community-dwelling
elderly population.
Carotenoids are found in
fruits and vegetables.
Some studies have shown
that plasma carotenoid
levels could be related
to dietary fruit and
vegetable intake
(25,26). More
specifically, it seems
that mangos, papayas,
peaches, prunes, squash,
oranges, and green
fruits and vegetables
are source of zeaxanthin,
whereas tomatoes, pink
grapefruit, and
watermelon are sources
of lycopene (3). Two
cross-sectional studies
(27,28) showed an
association between
greater intake of fruits
and vegetables and
better
cognitive performance.
In a large prospective
study of older women
(n¼13,388) (29), the
authors reported a
relationship between low
vegetable intake and
cognitive decline but no
relationship with fruits
(the strongest
association being with
greater intake of green
leafy vegetables and
cruciferous vegetables).
Even if some studies
show that intake of
foods with high
carotenoid contents are
correlated with their
corresponding plasma
concentrations (26),
fruits and vegetables
are sources of many
other nutrients [vitamin
E (30), folates (31,32),
flavonoids (33)] that
have been associated
with cognitive function.
It is now impossible to
know if the association
between carotenoids and
cognitive function is
the result of a specific
effect of carotenoids or
if it is the result of
combined effects of the
different fruit and
vegetable compounds.
For each cognitive test,
we tested an
eight-association
hypothesis (Ho) between
cognitive function and
carotenoids. To ensure
that significant
observed associations
were not hazard related
(alpha risk ¼ 5%),
specific multiple test
(Bonferroni, Sidak)
corrections were applied
to the data.
Considering, however,
that the tested
hypotheses are not
independent, submitting
them to these
corrections entails an
overcorrection of
threshold significance
thus not necessarily
leading to rejection of
the Ho hypothesis, which
we know to
be false. In this
exploratory analysis,
the fact that one
specific carotenoid was
associated with more
than one cognitive
function, and that these
associations remained
statistically
significant after
controlling for
potential confounding
factors or after
removing some
participants (those with
depressive
symptomatology, those
with an MMSE score , 25,
or undernourished
participants) seems to
us more likely to ensure
that our significant
observed associations
were not hazard related.
Although there were
important correlations
among all carotenoids,
we observed significant
associations between low
cognitive performances
and some (but not all)
carotenoids. More
specifically, we found
no associations between
b-carotenes (trans or
cis) or a-carotene,
which are the most
studied carotenoids in
epidemiological
literature on cognitive
impairment (5–11). These
studies give conflicting
results. Previously, in
the EVA study, we showed
that a low level of
baseline total plasma
carotenoid (, 25th
percentile) was not
significantly associated
with a 4-year cognitive
decline (5). In a
cross-sectional study,
Perrig and colleagues
(9) showed that a higher
b-carotene plasma level
was associated with
better memory
performances (free
recall, recognition, and
vocabulary) in 442
healthy persons aged
65–94 years. Studying
dietary intake of
b-carotene, Morris and
colleagues (8) found no
association between
carotene intake and
cognitive decline,
whereas the Rotterdam
Study showed that a
lower intake of
b-carotene was
associated with impaired
cognitive function
measured by the MMSE
(7). Three studies
investigated the link
between cognitive
performance and
supplementation of
antioxidants including
b-carotene (6,10,11).
All, except the work of
the Age- Related Eye
Diseases Study Research
Group (11), found that
use of these supplements
reduced the risk of
cognitive decline. In
these studies with
multiantioxidant
supplementation it is,
however, impossible to
isolate the specific
effect of b-carotene on
cognitive impairment.
Only one study, with
1769 participants,
focused on the
association between a
large spectrum of
carotenoids and
cognitive performance in
elderly participants
without neuropsychiatric
disease, but it found no
association (34).
Discrepancies can be
explained by
methodological
differences (e.g.,
neuropsychological tests
to assess cognitive
performances, number of
tests used, choice of
modelization of
carotenoids). Finally,
we found two clinical
epidemiological studies
(35,36) that focused on
the comparison of plasma
carotenoids and retinol
levels in elderly
patients with or without
Alzheimer’s disease
(AD). One study showed
that levels of vitamin
A, lutein, zeaxanthin,
b-cryptoxanthin, and
a-carotene were lower in
AD patients (n¼63) than
in controls (n¼56).
However, the researchers
found no difference for
lycopene and b-carotene
(36). In the second
study, levels of
carotenoids were lower
in AD patients (n ¼ 40)
than in controls (n ¼
39) for zeaxanthin, b-cryptoxanthin,
lycopene, and b-carotene
but not for lutein and
a-carotene (35). These
conflicting results
could be explained by
the limited sample size
of these studies. The
major problem in
interpreting these two
studies is that they are
studying AD cases for
which nutritional
habits, and
consequently, biological
status, can be modified,
as a consequence of the
disease progression. The
different measurement
methods and the
bioavailability of
carotenoids, which is
influenced by several
factors [such as
characteristics of the
food sources,
interactions with other
dietary factors, and
various participant
characteristics (3)]
could also explain the
differences between
studies. Many
epidemiological studies
have associated high
carotenoid status with a
decrease in the
incidence of chronic
diseases (heart diseases
and cancer); the
biological mechanism for
such protection is,
however, currently
unclear (3). Multiple
possibilities exist.
Among them, certain
carotenoids can be
converted to retinoids
and have a provitamin A
activity. To assess if
the relationship we
showed could be
explained partly by this
hypothesis, we adjusted
our models for retinol
concentration, but
results remained
unchanged. Our results
are supported by those
of a biological study,
led by Woodall and
colleagues (37), on
oxidation of carotenoids
by free radicals. The
authors found that
lycopene, lutein, and
zeaxanthin all reacted
rapidly with oxidizing
agents, and must also be
considered as potential
dietary antioxidants. A
possible explanation for
low levels
of plasma carotenoids in
AD or cognitive
impairment is that they
might be consumed
because of a higher rate
of free radical
production in the brain.
Our results on the
relationship between
carotenoids and
cognitive performance
were not modified when
we excluded the
participants with BMI ,
21 kg/m2 for which the
hypothesis of
undernutrition is
probable. This finding
allows us to say that
this relationship could
not be limited by the
influence of
undernourished
participants. However,
in this cross-sectional
framework, it is not
possible to affirm
whether these low levels
of carotenoids preceded
or were the consequence
of cognitive impairment
in a context in which
poor cognitive status
may be a risk factor for
poor nutrition. In this
study, we used various
tests to explore
cognitive performance;
this approach is more
powerful than using only
MMSE, which is not a
very sensitive
measurement of cognitive
impairment. With the
other tests, the
psychometric scores
ranges are large and
more powerful to study
cognitive impairment. In
addition, because the
scores were not normally
distributed, a percent
cut-off is appropriate.
The observed association
will probably have no
functional significance
yet, because
participants had only a
subtle impairment.
Moreover, we have no
basis to expect specific
association between
carotenoids and
psychometric evaluation.
However, it is well
known that low plasma
lutein and zeaxanthin
concentrations were
implicated in agerelated
macular degeneration
(38). Although the
retina is a puzzle (the
ultimate solution of
which lies on the other
side of the optic nerve
in its connection with
the brain), a highly
specific accumulation of
lutein and zeaxanthin in
the retina and in the
macula is described
(39). Could other areas
of the brain have the
same affinity for some
specific carotenoids?
The biological
significance of our
findings needs further
research by biological
studies, longitudinal
epidemiological studies,
and by specific clinical
trials with carotenoid
supplementation.
Printable
PDF
Version
|
