Alcohol Intake and Risk of Dementia
Jose A. Luchsinger, MD, z Ming-Xin Tang, PhD, z Maliha Siddiqui, MPH, Steven Shea, MD,z # and
Richard Mayeux, MD w§k#
OBJECTIVES: To examine the association between intake
of alcoholic beverages and risk of Alzheimer’s disease (AD)
and dementia associated with stroke (DAS) in a cohort of
elderly persons from New York City.
DESIGN: Cohort study.
SETTING: The Washington Heights Inwood–Columbia
Aging Project.
PARTICIPANTS: Nine hundred eighty community-dwell-
ing individuals aged 65 and older without dementia at
baseline and with data on alcohol intake recruited between
1991 and 1996 and followed annually.
MEASUREMENTS: Intake of alcohol was measured using
a semiquantitative food frequency questionnaire at base-
line. Subjects were followed annually, and incident demen-
tia was diagnosed using Diagnostic and Statistical Manual
of Mental Disorders, Fourth Edition, criteria and classified
as AD or DAS.
RESULTS: After 4 years of follow-up, 260 individuals
developed dementia (199 AD, 61 DAS). After adjusting for
age, sex, apolipoprotein E (APOE)-e4 status, education,
and other alcoholic beverages, only intake of up to three
daily servings of wine was associated with a lower risk of
AD (hazard ratio 5 0.55, 95% confidence interval 5 0.34–
0.89). Intake of liquor, beer, and total alcohol was not
associated with a lower risk of AD. Stratified analyses by
the APOE-e4 allele revealed that the association between
wine consumption and lower risk of AD was confined to
individuals without the APOE-e4 allele.
CONCLUSION: Consumption of up to three servings of
wine daily is associated with a lower risk of AD in elderly
From the Taub Institute for Research of Alzheimer’s Disease and the
Aging Brain; wGertrude H. Sergievsky Center; zDivision of General
Medicine, Department of Medicine, §Department of Neurology, and
k
Department of Psychiatry, College of Physicians and Surgeons; Departments of
z
Biostatistics and #Epidemiology, Joseph P. Mailman School of Public Health,
Columbia University, New York.
Support for this work was provided by grants from the National Institute of
Aging (AG07232 and AG07702), the Charles S. Robertson Memorial Gift for
research on Alzheimer’s disease, the Blanchette Hooker Rockefeller
Foundation, and the New York City Council Speaker’s fund for Public Health
Research.
Address correspondence to Richard Mayeux, MD, Gertrude H. Sergievsky
Center, PH-19, 630 West 168th St. New York, NY 10032.
E-mail: rpm2@columbia.edu
JAGS 52:540–546, 2004
r 2004 by the American Geriatrics Society
individuals without the APOEe-4 allele. J Am Geriatr Soc
52:540–546, 2004.
Key words: dementia; Alzheimer’s; alcohol; wine; elderly
W ith the aging of the population, the prevalence of
dementia is expected to increase significantly. There
is no established treatment or preventive measure for
dementia, but delaying its onset could significantly decrease
its prevalence, with important public health implications.1
Moderate alcohol intake has been associated with a
decreased risk of dementia, mostly in European studies,2–5
and there are conflicting data on whether moderate alcohol
intake or intake of particular alcoholic beverages is related
to a decreased risk of dementia. There is a paucity of data
on the relationship between alcohol intake and the inci-
dence of dementia in the elderly in the United States, who
may have drinking patterns different from Europeans. This
study was designed to examine the association between the
intake of different types of alcoholic beverages and
dementia in a cohort of persons aged 65 and older from
northern Manhattan.
METHODS
Study Population
Participants in the Washington Heights–Inwood Columbia
Aging Project cohort were drawn by random sampling of
2,126 healthy Medicare beneficiaries aged 65 and older
residing within a geographically defined area of northern
Manhattan.6 At entry, each subject underwent a structured
in-person interview including an assessment of health and
function, a standard medical history, a physical and
neurological examination, and a neuropsychological bat-
tery.7 Subjects were recruited between 1991 and 1996 and
followed annually, repeating the baseline examination at
each follow-up. A food frequency questionnaire was
administered to 1,422 individuals between baseline and
the first follow-up examination. Of these 1,422 individuals,
230 were excluded because of prevalent dementia, 210
because of loss to follow-up, and two because of missing
dietary data. Thus, the analytic sample comprised 980
0002-8614/04/$15.00
JAGS APRIL 2004–VOL. 52, NO. 4
subjects. The 916 individuals excluded due to loss to
follow-up or no dietary information were older than the
final sample (mean age
standard deviation 5 88.2
8.6
vs 73.3
5.8) and had a similar proportion of women
(70% vs 67%).
Diagnosis of Dementia and Cognitive Impairment
A group of neurologists, psychiatrists, and neuropsychol-
ogists made a diagnosis of dementia and assignment of
specific cause by consensus based on the information
gathered at the initial and yearly follow-up visits. The
diagnosis of dementia was based on Diagnostic and
Statistical Manual of Mental Disorders, Fourth Edition
criteria8 and required evidence of cognitive deficit on the
neuropsychological test battery and evidence of impairment
in social or occupational function; persons with a global
summary score on the Clinical Dementia Rating (CDR) of
0.5 or more were considered to have dementia.9 Diagnosis
of Alzheimer’s disease (AD) was based on the National
Institute of Neurological and Cognitive Disorders and
StrokeFAlzheimer’s Disease and Related Disorders Asso-
ciation criteria.10 Diagnosis of dementia associated with
stroke (DAS) was made in all subjects with a history of
stroke. Because moderate alcohol intake is related to a
lower risk of stroke,11 a less-conservative definition of DAS
was used to avoid finding an association between alcohol
and AD because of misclassification of DAS as AD. Brain
imaging was available in 85% of cases of stroke; in the
remainder, World Health Organization criteria were used to
define stroke.12 Subjects without dementia but with a
history of stroke at the baseline examination were included
in the analyses. These criteria and diagnostic methods have
been used extensively in analysis of data in this cohort.
Dietary Data
Dietary data were obtained using a 61-item version of
Willett’s semiquantitative food frequency questionnaire
(SFFQ) (Channing Laboratory, Cambridge, MA).13 Trained
interviewers administered in English or Spanish the SFFQ
by telephone between the baseline and first follow-up
examinations. The questionnaire inquired about servings of
beer, liquor, and wine by serving frequency; possible
answers were one to three servings/month, one serving a
week, two to four servings a week, five to six servings a
week, one serving a day, two to three servings a day, four to
five servings a day, and more than six servings a day. One
serving of beer was equivalent to 12 oz (12.8 g of alcohol),
one serving of liquor was equivalent to 1.5 oz (14 g of
alcohol), and one wine serving was equivalent to 4 oz (11 g
of alcohol). In light of the possibility of nonspecific alcohol
effects and specific nonalcoholic effects of the different
beverages, alcohol intake was examined in two ways: by
examining the association between beer, liquor, and wine
servings separately with incident dementia and by examin-
ing the association between total alcohol servings and
dementia. Individuals were classified as nondrinkers (0
servings reported), light drinkers (1 serving a month to 6
servings a week), moderate drinkers (1 to 3 servings a day),
and heavy drinkers (43 servings a day); this classification
was made trying to resemble previous publications in this
field2 for the sake of comparability and following the
ALCOHOL INTAKE AND RISK OF DEMENTIA 541
format of the questionnaire. Because of a low number of
moderate drinkers, the light and moderate drinkers were
aggregated in one category (1 serving a month to 3 servings
a day). Heavy alcohol intake was included despite the low
number of individuals in this category because heavy
alcohol intake may increase the risk of dementia.14 SFFQs
have been used and validated for the determination of
nutrient intake in the elderly.15–19 The validity of the food
frequency questionnaire used in the Washington Heights–
Inwood Columbia Aging Project cohort was assessed in a
subsample of 78 individuals using two 7-day food records
as the criterion. The correlation for energy-adjusted alcohol
intake was 0.44 (Po.01) (M. Siddiqui, personal commu-
nication, December 7, 2000). The reliability of the alcohol
intake measure was compared between two SFFQs admi-
nistered 2 months apart, and the measurements were not
significantly different.
Definition of Covariates
Ethnic group was based on self-report using the format of
the 1990 census.20 Individuals were also asked whether
they were of Hispanic origin. Participants were then
assigned to one of three groups: black (non-Hispanic),
Hispanic, or white (non-Hispanic). Ethnic group was not
found to significantly alter the models and was not included
in the analyses. Data on years of education and heart
disease were also obtained by self-report. Education was
examined as a continuous variable (education in years) and
as a categorical variable (6 years of education, 7–12 years
of education, 13–16 years of education, and 416 years of
education). Apolipoprotein (APO) genotype was deter-
mined using the method of Hixson and Vernier.21 Partici-
pants were classified as positive for the APOE-e4 allele
genotype if they had one or two e4 alleles.
Data Analysis
Individuals were compared according to their beverage
intake. Chi-square tests were used to compare sex, APOE-
e4 status, and the presence of heart disease. Analyses of
variance and t tests were used to compare mean age and
number of years of education. Cox proportional hazards
regression was used for multivariate analyses, with the
time-to-event variable in the models specified as time from
baseline examination to onset of dementia. Individuals with
dementia not caused by the subtype of interest were
censored at the time of onset of dementia. The final model
was stratified by education category using the STRATA
statement in the SAS procedure PROC PHREG (SAS
Institute, Inc., Cary, NC). Three types of analyses were
conducted: one in which servings of beer, liquor, and wine
were related separately to dementia subtypes; another in
which all beverage types were included in one model; and
another relating servings of total alcohol to dementia and
dementia subtypes. Alcohol intake and other covariates
were treated as baseline time-constant covariates. The
multivariate model examining AD as an outcome included
age, sex, APOE-e4, and education as covariates, and the
models examining dementia and DAS as outcomes also
included the presence of heart disease as a covariate.
Inclusion of diabetes mellitus, hypertension, lipid levels,
and smoking did not add predictive ability and were not
542 LUCHSINGER ET AL. APRIL 2004–VOL. 52, NO. 4 JAGS

included in the final models. SAS version 7 for Windows

75.37
5.78 75.11
5.62 82.41
0 75.48
5.87 75.15
5.49 71.63
3.54
436 g/d
(n 5 11)
was used for all analyses (SAS Institute, Inc.).

9.09
18.18

27.27
12
RESULTS

Total Alcohol

0.1–36 g/d

21.43w
51.79z
The 980 individuals contributed 4,023 years of observation

(280)

26.39
12z
(mean observation time 5 4.1
1.5 years). The mean age
of the sample was 73.3
5.8, and 67% were women. One
hundred forty-four individuals reported beer intake (15%),
138 reported liquor intake (14%), 162 reported wine intake

(n 5 690)

74.20

29.19

32.46
(17%), and 690 reported no intake of alcoholic beverages

0 gm

8
(70%). The incidence of AD was 4.9 cases per 100 person-
years (199 cases), and the incidence of DAS was 1.5 per 100
person years (61 cases).
Table 1 shows the comparison of clinical characteristics

(n 5 1)
Heavy

100
7
0

0
among different levels of intake for beer, liquor, and wine.
Individuals with light to moderate intake of beer were less
likely to be female (70.9% vs 46.0%; Po.001) and have

Moderate
(n 5 138)
Wine Intake
heart disease (31.0% vs 18.7%; P 5.003) than subjects

Light to

19.57w
42.03z

30.53
12z
who did not report drinking beer. Individuals with heavy
intake were less likely to be female (70.9% vs 33.3%;

Note: Light to moderate intake represents of one serving a month to three servings a day; heavy intake represents more than three servings a day.
Table 1. Comparison of Clinical Characteristics Among Drinking Categories Stratified by Beverage Types
P 5.044) than nondrinkers. Individuals who reported light
to moderate intake were less likely to be female (70.5% vs

(n 5 842)
50.3%; Po.001), had more years of education (9 vs 12

71.26

28.04

30.64
None
years; Po.001), and were less likely to have heart disease

9
(30.6% vs 21.5%; P 5.008) than individuals who reported
no intake of liquor. Those who reported light to moderate

(n 5 0)
Heavy
intake of wine were less likely to be female (71.3 vs 42.0;

F
F
F

F
Po.001), had more years of education (9 vs 12; Po.001),
and were less likely to have heart disease (30.6 vs 19.6;

75.52
5.85 74.27
5.07 74.29
5.93 75.31
5.82 75.48
5.46
P 5.019) than those who reported no wine intake. Similar Moderate
(n 5 163)
Light to

21.47
Liquor Intake

50.31z
differences were observed for total alcohol intake.

28.39
12z
Analyses using multivariate proportional hazards
models examining the association between each alcoholic
beverage type individually and the outcomes of interest
revealed that only light to moderate intake of wine was
(n 5 818)

70.54

28.35

30.56
None

significantly associated with a lower risk of incident

9

dementia (hazard ratio (HR) 5 0.64, 95% confidence

interval (CI) 5 0.43–0.96; P 5.031) compared with indivi-
duals who reported no wine intake. In similar analyses
examining AD as an outcome (Table 2), a similar result was
33.33
(n 5 6)
Heavy

found in a model adjusting for age and sex (HR 5 0.59,

14
0

95% CI 5 0.38–0.91; P 5.018), but this association be-

came nonsignificant after including education and APOE-
e4 in the model (HR 5 0.69, 95% CI 5 0.45–1.09;
Beer Intake

Moderate
(n 5 139)

P 5.118). Analyses examining dementia associated with

Light to

18.71w
46.04z

25.37
10

stroke as an outcome (Table 3) showed that, in the full

model, the HR for light to moderate wine intake was
similar to that in the previous analyses but not statisti-
cally significant (HR 5 0.47, 95% CI 5 0.18,1.20;
(n 5 836)

P 5.116). The analyses relating beer and liquor intake

70.93

29.08

30.98
None

did not reveal a significant association with any of the

outcomes.
Po.05; w Po.01; z Po.001.

Analyses were conducted relating total alcohol intake

standard deviation

in grams to dementia. Daily intake of alcohol in grams was

Characteristic

Heart disease, %

categorized to resemble the classification by servings used

Education, years
Apolipoprotein

for the analyses by beverage type assuming that each serving

Age, mean

was approximately 12 g of alcohol: 0 g (no intake), 0.1–36 g

Female, %

E-4, %

(light to moderate intake), and more than 36 g (heavy

intake). None of the HRs and CIs was suggestive of a
significant association between any level of alcohol intake

JAGS APRIL 2004–VOL. 52, NO. 4 ALCOHOL INTAKE AND RISK OF DEMENTIA 543

Table 2. Relationship Between Alcoholic Beverage Intake and Incident Alzheimer’s Disease
Beer Liquor Wine

Adjusted for Adjusted for Adjusted for

Age and Sex Full Model Age and Sex Full Model Age and Sex Full Model

Intake Hazard Ratio (95% Confidence Interval)

None 1.0 1.0 1.0 1.0 1.0 1.0

Light to moderate 1.33 (0.91–1.96) 1.39 (0.95–2.06) 1.15 (0.77–1.71) 1.34 (0.89–2.00) 0.59 (0.38–0.91) 0.69 (0.45–1.09)
P 5 .141 P 5 .094 P 5 .489 P 5 .152 P 5 .018 P 5 .110
Heavy F F F F F F

Note: Full model includes age, gender, education, and apolipoprotein Ee-4 allele presence. Light to moderate intake represents one serving a month to three servings
a day; heavy intake represents more than three servings a day.
Missing data for heavy intake reflects insufficient subjects in that category to calculate a hazard ratio. The light and moderate intake categories were combined because
of insufficient subjects in the moderate intake category.

and dementia or its subtypes. The adjusted HRs of AD and
DAS related to intake of 0.1–36 g of alcohol were 0.97
(95% CI 5 0.69–1.35; P 5.85) and 0.93 (95% CI 5 0.50–
1.72; P 5.82), respectively.
Analyses were conducted using proportional hazards
models including dummy variables for beer, liquor, and
wine in the same model; this is akin to adjusting the
association of each beverage type with the outcome by
intake of other beverage types and their associated traits. It
was found that only light to moderate intake of wine was
associated with a significant lower risk of dementia
(HR 5 0.52, 95% CI 5 0.34–0.80; P 5.003); light to
moderate intake of wine was also associated with a lower
risk of AD (HR 5 0.55, 95% CI 5 0.34–0.89; P 5.015)
(Table 4). The HR relating light to moderate intake of wine
and DAS (Table 4) also suggested a lower risk of dementia
and was close to statistical significance (HR 5 0.42, 95%
CI 5 0.15–1.15; P 5.091).
Stratified analyses by APOE-e4 allele revealed that, in
667 individuals without the allele, light to moderate intake
of wine was associated with a lower risk of dementia
(HR 5 0.44, 95% CI 5 0.25–0.77; P 5.004), and AD
(HR 5 0.45, 95% CI 5 0.24–0.85; P 5.013) compared
with nondrinkers, whereas this association disappeared
for individuals heterozygous or homozygous for the APOE-
e4 allele (HR for dementia 5 1.10, 95% CI 5 0.59–2.04;
P 5.093; HR for AD 5 1.32, 95% CI 5 0.67–2.60;
P 5.427). The interaction terms for alcohol category and
the APOE-e4 allele in the models with dementia and AD as
outcomes were both statistically significant (P 5.056 and
P 5.032, respectively). No effect modification by sex,
diabetes mellitus, hypertension, heart disease, or smoking
was found.
All analyses were repeated with and without the 210
individuals with diet data who were lost to follow-up, using
age at onset of dementia or age at censoring as the time-to-
event variable; the results were virtually unchanged from
those reported above.
DISCUSSION
This study showed that, in a cohort of individuals aged 65
years and older, light to moderate alcohol intake was
associated with a lower risk of dementia and AD, whereas
intake of beer and liquor was not associated with incident
dementia. Analyses examining total alcohol intake also

Table 3. Relationship Between Alcoholic Beverage Intake and Dementia Associated with Stroke
Beer Liquor Wine

Adjusted for Adjusted for Adjusted for

Age and Sex Full Model Age and Sex Full Model Age and Sex Full Model

Intake Hazard Ratio (95% Confidence Interval)

None 1.0 1.0 1.0 1.0 1.0 1.0

Light to moderate 0.68 (0.29–1.60) 0.59 (0.23–1.53) 1.24 (0.62–2.48) 1.14 (0.53–2.47) 0.58 (0.26–1.29) 0.47 (0.18–1.20)
P 5 .377 P 5 .285 P 5 .553 P 5 .737 P 5 .182 P 5 .116
Heavy 3.54 (0.49–25.81) 5.02 (0.67–37.72) F F F F
P 5 .377 P 5 .117

Note: Full model includes age, gender, education, heart disease, and apolipoprotein E-e4 allele presence. Light to moderate intake represents one serving a month to
three servings a day; heavy intake represents more than three servings a day.
Missing data for heavy intake reflects insufficient subjects in that category to calculate a hazard ratio. The light and moderate intake categories were combined because
of insufficient subjects in the moderate intake category.
544 LUCHSINGER ET AL. APRIL 2004–VOL. 52, NO. 4 JAGS

Table 4. Relationship Alcoholic Beverage and Alzheimer’s Disease and Dementia Associated with Stroke
Alzheimer’s Disease Dementia Associated with Stroke

Beer Liquor Wine Beer Liquor Wine

Intake Hazard Ratio (95% Confidence Interval)

None 1.0 1.0 1.0 1.0 1.0 1.0

Light to moderate 1.47 (0.98–2.22) 1.51 (0.98–2.34) 0.55 (0.34–0.89) 0.69 (0.26–1.83) 1.48 (0.64–3.40) 0.42 (0.15–1.15)
P 5 .065 P 5 .062 P 5 .015 P 5 .450 P 5 .360 P 5 .091
Heavy F F F 6.70 (0.79–56.83) F F
P 5 .081

Note: From a model including dummy variables for all alcoholic beverages, age, gender, education, and apolipoprotein E-e4 allele presence. Light to moderate intake
represents one serving a month to three servings a day; heavy intake represents more than three servings a day.
Missing data for heavy intake reflects insufficient subjects in that category to calculate a hazard ratio. The light and moderate intake categories were combined because
of insufficient subjects in the moderate intake category.

indicated that light or moderate intake was not associated
with a lower risk of dementia or AD. Analyses stratified by
the presence of APOE-e4 also showed that the association
between light wine intake and lower risk of AD was
restricted to individuals without the APOE-e4 allele,
whereas in individuals who were heterozygous or homo-
zygous for the allele, there was no association between wine
intake and lower risk of AD.
Moderate alcohol intake has been found to be
associated with a decreased risk of coronary disease22–24
and ischemic stroke11 in observational studies. On this
basis, it would be expected that moderate alcohol intake
would be associated with a lower risk of vascular dementia,
defined in the current study as DAS. Furthermore, the role
of cerebrovascular disease in the development of AD is
unclear, but if cerebrovascular disease is a precipitant of
AD, moderate alcohol intake could also be expected to
decrease the risk of AD. Another mechanism through which
alcohol would be expected to decrease the risk of AD would
be through antioxidant mechanisms. Studies have shown
that oxidation has an important role in AD,25 and there is
ongoing research on the primary prevention of AD with
antioxidant vitamins. Alcohol has antioxidant effects,26,27
and wine in particular has been found to have important
antioxidant effects. One recent report from the Rotterdam
study found, in individuals aged 55 and older, that light to
moderate drinking of alcohol (1–3 drinks per day) was
associated with a lower risk of dementia and that this effect
was not particularly limited to any beverage type.2 A study
from the Bordeaux region of France in individuals aged 65
and older reported that moderate wine consumption (3–4
glasses a day) was associated with a lower risk of AD and
overall dementia, whereas mild consumption (1–2 glasses a
day) was associated with a lower risk of AD but not
dementia.3 Individuals in this study drank wine almost
exclusively. The Canadian Study of Health and Aging also
reported a lower risk of AD with consumption of wine but
not other alcoholic beverages.28 The Copenhagen City
Heart Study recently reported in a nested case control study
that monthly and weekly intake of wine was associated
with a decreased risk of dementia in individuals aged 65 and
older, whereas intake of alcohol was not associated with a
decreased risk of dementia.5 Other smaller studies have
found no association between alcohol intake and a lower
risk of cognitive impairment or dementia.29–31 The findings
of the current study are consistent with the latter two studies,
but as opposed to the Rotterdam study, no association was
found between intake of beer and liquor and dementia. The
statistical power to detect an association between intake of
more than three servings of daily alcohol was limited, but the
results seem to indicate an association with a higher risk of
dementia, consistent with previous findings.14
There were few moderate drinkers in the study, and the
light and moderate intake categories were aggregated; thus,
light intake of alcoholic beverages (1 serving a month to 6
servings a week) mostly affected the results reported for
light to moderate intake. For example, 152 individuals
reported light intake of wine, and 11 individuals reported
moderate intake of wine. The adjusted HR of AD was 0.69
(95% CI 5 0.44–1.09; P 5.065) for light intake of wine and
0.77 (95% CI 5 0.12–5.62; P 5.540) for moderate intake
of wine. Thus, the results for light to moderate intake of
wine mostly reflect the association between light intake and
AD; the HR for moderate intake was in the same direction
as for light intake, but this needs to be explored in a larger
sample.
There are several potential explanations for the finding
of a differential association between beverage type and
incident dementia. Wine has two potential beneficial
effects; one is through alcohol itself, and the other through
substances with antioxidant effects in plasma, such as
polyphenols.26,27 Alcohol has a beneficial effect on vascular
biology through increased levels of high-density lipopro-
tein, decreased platelet adhesiveness, and improved en-
dothelial function that probably explains the association
between moderate alcohol intake and cardiovascular out-
comes,32 irrespective of type of alcoholic beverage; the
finding of a nonsignificant association between moderate
alcohol intake and lower incidence of DAS is consistent
with this explanation. Processes that originate, modulate,
or precipitate the deposition of amyloid beta in the brain,33
such as oxidative stress, rather than vascular processes, may
better explain the development of AD, and the vascular
effects of the alcohol component of alcoholic beverages may
not be enough to cause an association between moderate
intake of any alcoholic beverage and AD. The presence in
wine of nonalcoholic components such as particular
antioxidants could explain a differential effect of wine on
JAGS APRIL 2004–VOL. 52, NO. 4
AD. For example, liquor has been shown to have less
antioxidant activity than wine.34 Another potential ex-
planation for this finding is that of residual confounding.
Drinkers of wine had more years of education than liquor or
beer drinkers. It is possible that the study did not adjust
properly for education and that residual confounding by
educational attainment resulted in the association between
light wine consumption and lower risk of AD. A third
possibility is that of uncontrolled confounding. It is possible
that other behavioral characteristics that are associated
with a lower risk of AD, that are unknown, and that were
not accounted for in the statistical models accompany wine
drinking. Also, drinkers of wine may have patterns of
alcohol intake that are different than those of beer
and liquor drinkers. For example, drinkers of wine may
consume mostly light to moderate amounts of alcohol,
whereas liquor and beer drinkers consume amounts of
alcohol deemed to be unhealthy in shorter periods of time;
because average intake was measured, these patterns
cannot be taken into account. These data have differences
from those from the previously reported studies that may
limit comparisons. The report from the Rotterdam Study
examined a cohort that was younger than this one (55
compared with  65),2 and therefore the two studies may
be measuring different cumulative intakes that are reflected
in the results. Furthermore, the latency phase of dementia
may be at significantly different stages in the two studies
given the age difference, and the effect of any exposure on
dementia may be different in the cohorts. The Bordeaux
study had a cohort similar to the one in the current study in
terms of age, but wine was the only beverage studied. The
Copenhagen City Heart Study is similar in age structure to
the current one and found similar results, but alcohol intake
was measured 15 years before the selection of cases and
controls. Lastly, there may be significant cultural differences
between Europe and the United States in terms of alcohol
intake that may also make this study not comparable with
European studies. In the Rotterdam study, only 21% of
individuals reported no alcohol consumption, whereas
70% of individuals in the current study reported no alcohol
consumption.
The stratified analyses revealed that wine was not
associated with a decreased risk of AD in individuals
homozygous or heterozygous for APOE-e4, whereas there
was a decreased risk for individuals without the allele.
APOE-e4 is a predictor of the development of AD.35
Findings from another study relating moderate alcohol
intake to a decreased risk of cognitive decline in individuals
without the APOE-e4 allele support these results.36 The
mechanism for the effect of APOE-e4 on the risk of AD is not
clear, but it probably involves an increase in the deposition
of amyloid b in the brain.35 One possible explanation is that
the presence of APOE-e4 increases the risk of AD in such a
way that it overwhelms the potential protective mechanism
of wine. For example, if wine were to have an antioxidant
effect in relation to AD pathogenesis, then wine would not
be protective in individuals who have the highest disease
burden. One study reported a decreased risk of cognitive
function associated with alcohol intake in individuals with
diabetes mellitus and cardiovascular disease,37 but this was
not replicated in the current study in stratified analyses by
diabetes mellitus or cardiovascular disease.
ALCOHOL INTAKE AND RISK OF DEMENTIA 545
This study has several strengths. Individuals with very
mild forms of dementia that might be considered cognitive
impairment without dementia (CDR. 5 0.5) were excluded,
thus minimizing the risk of including individuals with
preclinical dementia whose alcohol intake was a result of
behavior modification related to the outcome. Standard
measures of the exposure and the outcome were also used.
The main limitation of this study is common to most
studies of diet and disease, namely misclassification of the
exposure. The coefficient relating report of alcohol to food
records showed a moderate correlation, indicating some
misclassification in the report of alcohol intake. Also, it was
assumed that alcohol intake was stable over time, but it is
probable that there was significant intraperson variation in
alcohol intake that was not taken into account. Assuming
nondifferential misclassification of alcohol intake, this
would result in underestimation of the association between
alcohol intake and incident dementia. Another limitation of
the study is that its primary focus was the detection of
memory-based cognitive impairment and distinguishing
between AD and DAS; thus, it is possible that cases of
dementia not caused by AD or DAS were missed or that
other causes of dementia were misclassified primarily as
AD.38 It is important to point out that this study and others
showing similar results are observational and that, given the
potential risks associated with alcohol intake, no recom-
mendations on alcohol intake should be based on these
results. Furthermore, if the association in this study
represented a true protective effect, it is possible that the
risk of dementia is decreased at the expense of increasing
other comorbidities, including forms of cognitive impair-
ment, and this cannot be addressed with these data.
In conclusion, it was found that light to moderate intake
of wine was associated with a decreased risk of dementia
and Alzheimer’s disease in individuals without the APOE-e4
allele. Similar intakes of beer, liquor, or total alcohol were
not related to a decreased risk of Alzheimer’s disease.
REFERENCES
1. Hake AM, Scherer P. On the brink of the pandemic: Epidemiology and risk
factors for Alzheimer’s. Summaries from the World Alzheimer’s Congress July
9–18, 2000 [on-line]. Available at www.medscape.com/Medscape/CNO/
2000/WAC/Story.cfm?story_id 5 1523 Accessed December 2000.
2. Ruitenberg A, Van Swieten JC, Wittleman JCM et al. Alcohol consumption
and risk of dementia. The Rotterdam Study. Lancet 2002;359:281–286.
3. Orgogozo JM, Dartigues JF, Lafont S et al. Wine consumption and dementia in
the elderly: A prospective community study in the Bordeaux area. Rev Neurol
1997;153:185–192.
4. Huang W, Qiu C, Winblad B et al. Alcohol consumption and incidence of
dementia in a community sample aged 75 years and older. J Clin Epidemiol
2002;55:959–964.
5. Truelsen T, Thudium D, Gronbaek M. Amount and type of alcohol and risk of
dementia. The Copenhagen City Heart Study. Neurology 2002;59:1300–1301.
6. Tang M, Stern Y, Marder K et al. The ApoE-e4 allele and the risk of
Alzheimer’s disease among African Americans, whites, and Hispanics. JAMA
1998;279:751–755.
7. Stern Y, Andrews H, Pittman J et al. Diagnosis of dementia in a heterogeneous
population. I. Development of a neuropsychological paradigm and quantified
correction for education. Arch Neurol 1992;49:453–460.
8. American Psychiatric Association. Diagnostic and Statistical Manual of
Mental Disorders, (DSM-IV). Washington, DC: American Psychiatric
Association, 1994.
9. Hughes CP, Berg L, Danzinger W et al. A new clinical scale for the staging of
dementia. Br J Psychiatry 1982;140:566–572.
10. McKhann G, Drachman D, Folstein M et al. Clinical diagnosis of Alzheimer’s
disease: Report of the NINCDS-ADRDA work group under the auspices of the
546 LUCHSINGER ET AL.
Department of Health and Human Services Task Force on Alzheimer’s Disease.
Neurology 1984;34:939–944.
11. Sacco RL, Elkind M, Boden-Albala B et al. The protective effect of alcohol
consumption on ischemic stroke. JAMA 1999;281:53–60.
12. The ICD-10 Classification of Mental and Behavioral Disorders: Diagnostic
Criteria for Research. Geneva, Switzerland: World Health Organization,
1993.
13. Willett WC, Sampson L, Stampfer JM et al. Reproducibility and validity of
a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:
51–65.
14. Tyas SL. Alcohol use and the risk of developing Alzheimer’s disease. Alcohol
Res Health 2001;25:299–306.
15. Munger RG, Folsom AR, Kushi LH et al. Dietary assessment of older Iowa
women with a food frequency questionnaire: Nutrient intake, reproducibility,
and comparison with 24-hour recall interviews. Am J Epidemiol 1992;135:
192–200.
16. Nes M, Anderson LF, Solvoll K et al. Accuracy of a quantitative food frequency
questionnaire applied in elderly Norwegian women. Eur J Clin Nutr 1992;
46:809–821.
17. Longnecker MP, Lissner L, Holden JM et al. The reproducibility and validity of
a self-administered questionnaire in subjects from South Dakota and
Wyoming. Epidemiology 1993;4:356–365.
18. Lazarus R, Wilson A, Gliksman M et al. Repeatability of nutrient intakes
estimated by a semi-quantitative food frequency questionnaire in elderly
subjects. Ann Epidemiol 1995;5:65–68.
19. Smith W, Mitchell P, Reay EM et al. Validity and reproducibility of a self-
administered food frequency questionnaire in older people. Aust N Z J Public
Health 1998;22:456–463.
20. 1990 Census of Population and Housing. Summary Tape File 1, Technical
Documentation (STF 1A database). Washington, DC: Bureau of the Census,
1991.
21. Hixson J, Vernier D. Restriction isotyping of human lipoprotein E by gene
amplification and cleavage of HhAI. J Lipid Res 1991;31:545–548.
22. Maclure M. Demonstration of deductive meta-analysis: Ethanol intake and
risk of myocardial infarction. Epidemiol Rev 1993;15:328–351.
23. Gaziano JM, Buring JE, Breslow JL et al. Moderate alcohol intake, increased
levels of high density lipoprotein and its subfractions, and decreased risk of
myocardial infarction. N Engl J Med 1993;329:1829–1834.
APRIL 2004–VOL. 52, NO. 4 JAGS
24. Rimm EB. Alcohol consumption and coronary heart disease: Good habits
may be more important than just good wine. Am J Epidemiol 1996;143:
1094–1098.
25. Pitchumoni SS, Doraiwamy PM. Current status of antioxidant therapy for
Alzheimer’s disease. J Am Geriatr Soc 1998;46:1566–1572.
26. Howard A, Chopra M, Thurnman D et al. Red wine consumption an
inhibition of LDL oxidation: What are the important components? Med
Hypotheses 2002;59:101.
27. Bertelli AA, Migliori M, Panichi V et al. Oxidative stress and inflammatory
reaction modulation by white wine. Ann N Y Acad Sci 2002;957:295–301.
28. Linday J, Laurin D, Verreault R et al. Risk Factors for Alzheimer’s disease.
A prospective analysis from the Canadian Study of Health and Aging. Am
J Epidemiol 2002;156:445–453.
29. Broe GA, Creasey H, Jorm AF et al. Health habits and risk of cognitive
impairment and dementia in old age: A prospective study on the effects of
exercise, smoking, and alcohol consumption. Aust N Z J Public Health 1998;
22:621–623.
30. Leibovici D, Ritchie K, Ledesert B et al. The effects of wine and tobacco
consumption on cognitive performance in the elderly: A longitudinal study of
relative risk. Int J Epidemiol 1999;28:77–81.
31. Hebert LE, Scherr PA, Beckett LA et al. Relation of smoking and alcohol
consumption to incident Alzheimer’s disease. Am J Epidemiol 1992;135:347–
355.
32. Belleville J. The French Paradox. Possible involvement of ethanol in the
protective effect against cardiovascular disease. Nutrition 2002;18:173–177.
33. Selkoe DJ. The origins of Alzheimer’s disease: A is for amyloid. JAMA 2000;
283:1615–1616.
34. Heinonen IM, Lehtonen PJ, Hopia AI. Antioxidant activity of berry and fruit
wines and liquors. J Agric Food Chem 1998;46:25–31.
35. Selkoe DJ. Alzheimer’s disease. Genotypes, phenotype, and treatments. Science
1997;275:630–631.
36. Dufouil C, Tzourio C, Brayne C et al. Influence of apolipoprotein E genotype
on the risk of cognitive deterioration in moderate drinkers and smokers.
Epidemiology 2000;11:280–284.
37. Launer LJ, Feskens EJ, Kalmihn S et al. Smoking, drinking, and thinking. The
Zutphen Elderly Study. Am J Epidemiol 1996;143:219–227.
38. Royall DR. The ‘Alzheimerization’ of dementia research. J Am Geriatr Sci
2003;51:277–278.