Vous êtes sur la page 1sur 6

ORIGINAL CONTRIBUTION

Relation of Diabetes to Mild Cognitive Impairment


José A. Luchsinger, MD; Christiane Reitz, MD; Bindu Patel, MPH; Ming-Xin Tang, PhD;
Jennifer J. Manly, PhD; Richard Mayeux, MD

Background: Type 2 diabetes mellitus is an important ate analyses with proportional hazards regression adjust-
risk factor for Alzheimer disease and is more prevalent ing for age, sex, years of education, ethnic group, apoli-
in elderly minority persons compared with non- poprotein E (APOE) ε4 allele, hypertension, low-
Hispanic white persons. density lipoprotein level, current smoking, heart disease,
and stroke.
Objective: To determine whether diabetes is related to
a higher risk of mild cognitive impairment (MCI), a tran- Results: A total of 334 persons had incident MCI, 160
sitional stage between normal cognition and Alzheimer (47.9%) had amnestic MCI, and 174 (52.1%) had non-
disease, in a multiethnic cohort with a high prevalence amnestic MCI. Diabetes was related to a significantly
of diabetes. higher risk of all-cause MCI and amnestic MCI after ad-
justment for all covariates. Diabetes was also related to
Design: Longitudinal cohort study.
a higher risk of nonamnestic MCI, but this association
was appreciably attenuated after adjustment for socio-
Setting: Northern Manhattan in New York, NY.
economic variables and vascular risk factors. The risk of
Participants: We studied persons without prevalent MCI
MCI attributable to diabetes was 8.8% for the whole
or dementia at baseline and with at least 1 follow-up in- sample and was higher for African American persons
terval. Of 1772 participants with a complete neuropsy- (8.4%) and Hispanic persons (11.0%) compared with non-
chological evaluation, 339 (19.1%) were excluded be- Hispanic white persons (4.6%), reflecting the higher
cause of prevalent dementia, 304 were excluded because prevalence of diabetes in minority populations in the
of prevalent MCI (17.2%), and 211 were excluded be- United States.
cause of loss to follow-up (11.9%), resulting in a final
sample of 918 participants for longitudinal analyses. Conclusion: Diabetes is related to a higher risk of am-
nestic MCI in a population with a high prevalence of this
Main Outcome Measures: We related diabetes de- disorder.
fined by self-report to incident all-cause MCI, amnestic
MCI, and nonamnestic MCI. We conducted multivari- Arch Neurol. 2007;64:570-575

A
MONG CARDIOVASCULAR Interest is growing in preclinical tran-
risk factors, type 2 diabe- sitional states of AD as targets for treat-
Author Affiliations: Taub tes mellitus has been con- ment and prevention. Mild cognitive im-
Institute for Research of pairment (MCI), and particularly amnestic
sistently related to a higher
Alzheimer’s Disease and the MCI, has been described as a transitional
Aging Brain (Drs Luchsinger, risk of Alzheimer disease
Tang, Manly, and Mayeux and (AD).1-6 This association is presumed to state between normal cognition and AD14-16
Ms Patel), Gertrude H. be the result of the effects of peripheral hy- that is increasingly used in clinical and re-
Sergievsky Center perinsulinemia on the clearance of brain search settings. We previously demon-
(Drs Luchsinger, Reitz, Manly, amyloid ␤,7,8 the putative main culprit in strated that diabetes is associated with an
and Mayeux), and Departments increased risk of AD and dementia.2,17
the pathogenesis of AD,9 or of other mecha-
of Biostatistics (Dr Tang) and Thus, we hypothesized that diabetes is also
nisms such as advanced products of gly-
Epidemiology (Drs Luchsinger related to a higher risk of MCI, particu-
and Mayeux), Joseph P. cosylation.10 The prevalence of diabetes is larly amnestic MCI.
Mailman School of Public more than 10% in the elderly population
Health; and Departments of in the United States,11 is twice as high in
Medicine (Dr Luchsinger), METHODS
African American and Hispanic popula-
Neurology (Drs Manly and tions compared with non-Hispanic white
Mayeux), and Psychiatry PARTICIPANTS AND SETTING
(Dr Mayeux), College of
populations,12 and is more than 20% in per-
Physicians and Surgeons, sons 65 years or older in the predomi- Participants were from a longitudinal cohort
Columbia University, nantly minority population of northern study of Medicare recipients 65 years or older
New York, NY. Manhattan in New York, NY.13 residing in northern Manhattan (Washington

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


570

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015
Heights, Hamilton Heights, and Inwood).18 Each participant un- subtypes of MCI, those considered for MCI were required to
derwent an in-person interview of general health and function have the following: (1) memory complaints; (2) objective im-
at the time of study entry followed by a standard assessment, in- pairment in at least 1 cognitive domain based on the average
cluding medical history, physical and neurological examination, of the scores on the neuropsychological measures within that
and a neuropsychological battery.19 These assessments were per- domain and a 1.5-SD cutoff using normative corrections for age,
formed at each follow-up interval. Baseline data were collected years of education, ethnicity, and sex; (3) essentially pre-
from 1992 through 1994 and follow-up data every 18 months un- served activities of daily living; and (4) no diagnosis of demen-
til 2003. The institutional review board of Columbia- tia at the consensus conference.
Presbyterian Medical Center (New York, NY) approved this study. To cast the widest net to determine prevalence of MCI and
Our study included persons without prevalent MCI or de- to determine which individuals were more likely to progress
mentia at baseline and with at least 1 follow-up interval. Of 1772 to dementia, we expanded the original Petersen criteria,34
participants with a complete neuropsychological evaluation, 339 which focus on memory impairment, to include mutually
(19.1%) were excluded because of prevalent dementia, 304 be- exclusive subtypes based on cognitive features. Our first sub-
cause of prevalent MCI (17.2%), and 211 because of loss to fol- type, amnestic MCI, corresponds most closely to the original
low-up (11.9%), resulting in a final sample of 918 participants definition used by Petersen and colleagues. Memory impair-
for longitudinal analyses. Compared with the original 1772 par- ment was defined as a score of less than 1.5 SDs below the
ticipants, the final sample without prevalent MCI and dementia demographically corrected mean on an average composite
and with prospective data was younger (mean±SD age, 75.9±6.0 measure that comprised the following learning and memory
vs 77.3±6.8 years; P⬍.001), had a similar distribution of women measures: (1) total recall from the Selective Reminding Test,
(69.4% vs 69.4%) and African American participants (33.6% vs (2) delayed free recall from the Selective Reminding Test, and
32.3%), had a lower proportion of Hispanic participants (43.9% (3) recognition from the Benton Visual Retention Test. Perfor-
vs 47.0%; P⬍.001), had a higher proportion of non-Hispanic white mance on composite scores from all other cognitive domains
participants (22.6% vs 20.4%; P=.008), and had a lower preva- (ie, executive, language, and visuospatial) was required to be
lence of diabetes (23.9% vs 26.4%; P=.004). within normal limits (score must be ⱖ1.5 SDs below the
demographically corrected mean). Other MCI subtypes were
NEUROPSYCHOLOGICAL BATTERY classified that allowed for impairment in a single nonmemory
domain if performance on composite scores from all other
The neuropsychological measures used in this study have been cognitive domains was within normal limits. Executive MCI
previously described.19 The evaluation included measures of was assigned if impairment was demonstrated on an average
learning and memory, orientation, abstract reasoning, lan- composite measure that comprised the following measures:
guage, and visuospatial ability. Specific ability areas and tests (1) letter fluency, (2) category fluency, and (3) the Wechsler
administered included verbal list learning and memory (Selec- Adult Intelligence Scale–Revised similarities subtest. Lan-
tive Reminding Test20), nonverbal memory (multiple-choice ver- guage MCI was defined as isolated impairment on an average
sion of the Benton Visual Retention Test21), orientation (items composite measure that comprised (1) the Boston Naming
from the Mini-Mental State Examination22), verbal reasoning Test, (2) the BDAE repetition test, and (3) the BDAE compre-
(similarities subtest of the Wechsler Adult Intelligence Scale– hension test. Visuospatial MCI was assigned if impairment
Revised23), nonverbal reasoning (identities and oddities sub- was demonstrated on an average composite measure that
test of the Mattis Dementia Rating Scale24), naming (15-item comprised (1) Rosen Drawing and (2) Benton Visual Reten-
version of the Boston Naming Test25), letter fluency (Con- tion Test matching. Finally, we allowed for impairment in
trolled Word Association26), category fluency (animals, food, multiple cognitive domains in the absence of dementia. Mul-
and clothing, using procedures from the Boston Diagnostic Apha- tiple cognitive domains with memory impairment MCI (MCI-
sia Examination [BDAE]27), repetition (high-frequency phrases MCDM) was diagnosed if there was objective impairment on
of the BDAE27), auditory comprehension (first 6 items of the the memory domain composite score and if there was impair-
complex ideational material subtest of the BDAE28), visuocon- ment on at least 1 other cognitive domain. Multiple cognitive
struction (Rosen Drawing Test29), and visuoperceptual skills domains without memory impairment MCI was assigned if
(multiple-choice matching of figures from the Benton Visual there was impairment in 2 or more of the 3 nonmemory
Retention Test21). Norms for these tests to diagnose MCI have domains and if the memory domain composite score was
been previously described.30 within normal limits. Classification into the 6 subtypes was
mutually exclusive. We used 3 outcomes for these analyses:
(1) all-cause MCI; (2) amnestic MCI, which included amnes-
DIAGNOSIS OF DEMENTIA tic MCI and MCI-MCDM; and (3) nonamnestic MCI. The
rationale for grouping amnestic MCI and MCI-MCDM is that
Diagnosis of dementia and assignment of specific cause were amnestic MCI and MCI-MCDM equally predict the develop-
made by consensus of 2 neurologists, 1 psychiatrist, and 2 neu- ment of AD and MCI-MCDM is thought to be a more
ropsychologists based on baseline and follow-up information. advanced form of amnestic MCI involving other cognitive
The diagnosis of dementia was based on Diagnostic and Statis- domains.
tical Manual of Mental Disorders (Fourth Edition) criteria31 and
required evidence of cognitive deficits on the neuropsycho-
logical test battery as well as evidence of impairment in social DEFINITION OF DIABETES
or occupational function (Clinical Dementia Rating of 1 or AND OTHER COVARIATES
more).32 Diagnosis of AD was based on the National Institute
of Neurological Disorders and Stroke–Alzheimer’s Disease and History of type 2 diabetes mellitus was ascertained by self-
Related Disorder Association criteria.33 report or the use of diabetes medications at baseline and each
follow-up visit. Hypertension, heart disease, and smoking
MCI DIAGNOSTIC CRITERIA were defined by self-report. Heart disease included a history
of atrial fibrillation and other arrhythmias, congestive heart
The MCI criteria were retrospectively applied among persons failure, myocardial infarction, and angina pectoris. Smoking
without dementia. Consistent with standard criteria,16 for all was classified into never, current, and past smoking. Fasting

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


571

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015
of diabetes in this sample.41 All analyses were conducted using
Table 1. Demographic and Other Relevant Characteristics SAS statistical software, version 9.1 for Windows (SAS Institute
of Persons With and Without Diabetes: The Washington Inc, Cary, NC).
Heights–Inwood Columbia Aging Project, 1992-2003*

Without Diabetes With Diabetes RESULTS


Characteristic (n = 699) (n = 219)
Age, y 76.2 ± 6.1 75.3 ± 5.6†
Female sex 480 (68.7) 159 (72.6)
There were 334 incident MCI cases, 160 amnestic MCI
African American 237 (33.9) 71 (32.4) cases and 174 cases of nonamnestic MCI in 5556 person-
Hispanic 279 (39.9) 124 (56.6)‡ years of follow-up (mean ± SD, 6.1 ± 3.2 years per per-
Non-Hispanic white 183 (26.2) 24 (10.9)‡ son). The mean±SD age of the sample was 75.9±6.0 years,
Education, y 9.2 ± 4.6 7.8 ± 4.3‡ and the mean ± SD duration of education was 8.9 ± 4.6
APOE ε4 196 (29.8) 48 (23.7) years. Women represented 69.9% of the sample, 43.9%
Hypertension 442 (63.2) 184 (84.0)‡
Heart disease 201 (28.8) 110 (50.2)‡
were Hispanic, 33.6% were African American, and 22.5%
LDL-C, mg/dL 119.6 ± 36.1 122.0 ± 38.0 were non-Hispanic white. Diabetes was reported by 23.9%
Current smoking 78 (11.2) 17 (7.8) of persons, hypertension by 68.2%, heart disease by 33.9%,
Stroke 85 (12.2) 53 (24.2)‡ and stroke by 15.0%; 28.3% of the sample carried the
Follow-up 6.0 ± 3.2 6.2 ± 3.1 APOE ε4 allele. The mean±SD low-density lipoprotein
MCI 241 (34.5) 93 (42.5)† cholesterol level was 120.4 ± 36.7 mg/dL (3.12 ± 0.95
mmol/L).
Abbreviations: LDL-C, low-density lipoprotein cholesterol;
MCI, mild cognitive impairment.
Compared with persons without diabetes, persons with
SI conversion factor: To convert LDL-C to millimoles per liter, multiply by diabetes were younger, were more likely to be Hispanic,
0.0259. were less likely to be non-Hispanic white, had fewer years
*Data are presented as number (percentage) of persons or mean ± SD. of education, and were more likely to have hyperten-
†P⬍.05.
‡P⬍.001. sion, heart disease, and stroke (Table 1). Persons with
incident MCI were more likely to report diabetes and hy-
pertension at baseline and had a longer study follow-up
plasma total cholesterol and triglyceride levels were deter- time (Table 2).
mined at the first follow-up using standard enzymatic tech- In multivariate analyses (Table 3), diabetes was re-
niques. High-density lipoprotein cholesterol levels were deter- lated to a higher risk of all-cause MCI even after adjust-
mined after precipitation of apolipoprotein B–containing ing for age, sex, ethnic group, years of education, APOE
lipoproteins with phosphotungstic acid.35 The low-density ε4, hypertension, low-density lipoprotein level, heart dis-
lipoprotein cholesterol level was recalculated using the for- ease, stroke, and current smoking (HR, 1.4; 95% confi-
mula of Friedewald et al.36 The apolipoprotein E (APOE)
dence interval [CI], 1.1-1.8).
genotypes were determined as described by Hixson and Ver-
nier37 with slight modification.38 We classified persons as When only amnestic MCI was considered the out-
homozygous or heterozygous for the APOE ε4 allele or as not come of interest, the HR was unchanged (HR, 1.5; 95%
having any ε4 allele. CI, 1.0-2.2). When nonamnestic MCI was considered,
diabetes was related to a higher risk in the model ad-
STATISTICAL ANALYSIS justed for age and sex (HR, 1.4; 95% CI, 1.0-1.9), but it
became nonsignificant after adjusting for ethnic group,
Bivariate analyses compared variables between persons with and years of education, and APOE ε4 (HR, 1.3; 95% CI, 0.9-
without diabetes. Continuous variables were compared using 1.8). The HR attenuated further after adjusting for other
analysis of variance, and categorical variables were compared using vascular risk factors, heart disease, and stroke (HR, 1.2;
␹2 tests.39 Cox proportional hazards regression models40 were used 95% CI, 0.9-1.8).
in multivariate analyses exploring the association of diabetes with The association between diabetes and MCI was not
incident all-cause MCI, amnestic MCI, and nonamnestic MCI. The modified by the presence of APOE ε4. Compared with
time-to-event variable was age at diagnosis of MCI. Among in-
persons without diabetes and without APOE ε4, per-
dividuals who did not develop MCI, those who developed de-
mentia were censored at the time of dementia diagnosis, and those sons with diabetes and the APOE ε4 allele had an HR of
who did not develop dementia were censored at the time of last 1.3 (95% CI, 0.8-2.1), whereas persons with diabetes with-
follow-up. We show the results of multivariate analyses for 3 mod- out the APOE ε4 allele had an HR of 1.5 (95% CI, 1.1-
els: 1 adjusted for age and sex; 1 adjusted also for education, eth- 1.9) in the full model. No effect modification by age (cat-
nic group, and APOE ε4; and 1 additionally adjusted for hyper- egorized by the median) or sex was found.
tension, low-density lipoprotein cholesterol level, heart disease, We calculated the risk of MCI attributable to diabe-
stroke, and current smoking, with the caveat that some of these tes for the whole sample and for each ethnic group. The
variables may be in the causal pathway between diabetes and cog- prevalence of diabetes was 23.1% in African American
nitive disorders,17 and attenuation of the hazard ratios (HRs) in persons, 30.8% in Hispanic persons, and 11.6% in non-
this model should be considered evidence of mediation and not
Hispanic white persons, comparable to data from na-
of confounding. The risk of MCI attributable to diabetes was cal-
culated using the following formula: tional surveys.12 The risk of MCI attributable to diabetes
was 8.8% for the whole sample, 8.4% for African Ameri-
PAR=p(RR−1)/[1⫹p(RR−1)], can persons, 11.0% for Hispanic persons, and 4.6% for
where PAR is population attributable risk, RR is the adjusted HR non-Hispanic white persons, reflecting the differences in
obtained from the multivariate models, and p is the prevalence diabetes prevalence by ethnic group.

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


572

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015
Table 2. Relevant Characteristics Among Persons Without and With MCI, Amnestic MCI, and Nonamnestic MCI: The Washington
Heights–Inwood Columbia Aging Project, 1992-2003*

Characteristic No MCI MCI Amnestic MCI Nonamnestic MCI


No. of participants 584 334 160 174
Age, y 75.8 ± 6.2 76.3 ± 5.8 76.9 ± 5.8† 75.7 ± 5.7
Sex 406 (69.5) 233 (69.8) 112 (70.0) 121 (69.5)
African American 196 (33.6) 112 (33.5) 55 (34.4) 57 (32.8)
Hispanic 251 (42.9) 152 (45.5) 65 (40.6) 87 (50.0)
Non-Hispanic white 137 (23.5) 70 (20.9) 40 (25.0) 30 (17.2)
Education, y 9.0 ± 4.5 8.6 ± 4.6 9.0 ± 4.5 8.2 ± 4.6†
APOE ε4 147 (27.4) 97 (29.9) 50 (32.5) 47 (27.5)
Diabetes 126 (21.6) 93 (27.8)† 43 (26.9) 50 (28.7)†
Hypertension 368 (63.0) 258 (77.3)‡ 118 (73.8)† 140 (80.5)‡
Heart disease 197 (33.7) 114 (34.1) 54 (33.8) 60 (34.5)
LDL-C, mg/dL 122.2 ± 36.6 117.0 ± 36.7 117.2 ± 34.1 116.8 ± 39.1
Current smoking 63 (10.8) 32 (9.6) 16 (10.0) 16 (9.2)
Stroke 87 (14.9) 51 (15.3) 27 (16.9) 24 (13.8)
Total follow-up 5.5 ± 3.2 6.9 ± 2.8‡ 6.7 ± 2.9‡ 7.2 ± 2.8‡

Abbreviations: LDL-C, low-density lipoprotein cholesterol; MCI, mild cognitive impairment.


SI conversion factor: To convert LDL-C to millimoles per liter, multiply by 0.0259.
*Data are presented as number (percentage) of persons or mean ± SD. Persons without MCI are the reference for statistical testing.
†P⬍.05.
‡P⬍.001.

Table 3. HRs and 95% CIs Relating Diabetes to MCI, Amnestic MCI, and Nonamnestic MCI*

Model 1 Model 2 Model 3


MCI Cases
(Rate) HR (95% CI) P Value HR (95% CI) P Value HR (95% CI) P Value
All-cause MCI
No diabetes 241 (7.2) 1.0 1.0 1.0
Diabetes 93 (9.4) 1.4 (1.1-1.8) .007 1.3 (1.0-1.7) .03 1.4 (1.0-1.8) .04
Amnestic MCI
No diabetes 117 (3.5) 1.0 1.0 1.0
Diabetes 43 (4.4) 1.4 (1.0-1.9) .05 1.5 (1.0-2.1) .04 1.5 (1.0-2.2) .02
Nonamnestic MCI
No diabetes 124 (3.7) 1.0 1.0 1.0
Diabetes 50 (5.1) 1.4 (1.0-1.9) .04 1.3 (0.9-1.8) .21 1.2 (0.9-1.8) .22

Abbreviations: CI, confidence interval; HR, hazard ratio; MCI, mild cognitive impairment.
*Model 1 is adjusted for age and sex; model 2 is also adjusted for ethnic group, years of education, and APOE ε4; and model 3 is also adjusted for
hypertension, low-density lipoprotein cholesterol level, current smoking, heart disease, and stroke. Rates are per 100 person-years.

COMMENT more likely to be related to vascular cognitive syn-


dromes.42 Diabetes could be related to a higher risk of
Diabetes was associated with a higher risk of incident all- AD and amnestic MCI through direct mechanisms, af-
cause MCI in a population with a high prevalence of this fecting the amyloid accumulation that is the putative cul-
disorder. It was also related to a higher risk of amnestic prit of AD,9 or indirect mechanisms, namely cerebrovas-
MCI after adjustment for vascular risk factors, heart dis- cular disease,17 which is also related to a higher risk of
ease, and stroke, but the relation to nonamnestic MCI AD.45-47 Hyperinsulinemia, which can precede and ac-
was attenuated after adjustment for these covariates. company diabetes,48 may disrupt brain amyloid ␤ clear-
Amnestic MCI has been described as a transitional stage ance by means of the insulin degrading enzyme.7 An-
between normal cognition and AD.14-16 Nonamnestic MCI other potential mechanism is the generation of advanced
may be related to cerebrovascular disease or other cog- products of glycosylation.10,49,50 In our analyses, diabe-
nitive disorders and does not predict the onset of AD.42 tes was related to a higher risk of amnestic MCI even af-
Diabetes is related to a higher risk of AD.1-3,5 Thus, we ter adjusting for stroke and vascular risk factors, which
expected that it would be related to all-cause and am- suggests that the association between diabetes and
nestic MCI, which is a predictor of AD. Because diabe- amnestic MCI is independent of cerebrovascular dis-
tes is related to a higher risk of cerebrovascular disease ease (assuming that potential residual confounding was
and vascular dementia,43,44 we also expected that it would not significant enough to account for the association).
be related to a higher risk of nonamnestic MCI, which is Conversely, the relation of diabetes to nonamnestic MCI

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


573

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015
was appreciably attenuated and became nonsignificant Author Contributions: Study concept and design:
after adjustment for stroke and vascular risk factors, sug- Luchsinger. Acquisition of data: Luchsinger, Manly, and
gesting that cerebrovascular disease may mediate the re- Mayeux. Analysis and interpretation of data: Luchsinger,
lation between diabetes and nonamnestic MCI. Reitz, Patel, and Tang. Drafting of the manuscript:
Our findings have alternative explanations. Diabetes Luchsinger, Reitz, Patel, and Tang. Critical revision of the
is more prevalent in African American, Hispanic, and manuscript for important intellectual content: Luchsinger,
lower socioeconomic groups.12 The risk of AD is also Manly, and Mayeux. Statistical analysis: Luchsinger and
higher in African American and Hispanic populations51 Tang. Obtained funding: Luchsinger, Manly, and Mayeux.
and in persons with lower educational level52; thus, con- Administrative, technical, and material support: Mayeux.
founding by ethnic group and socioeconomic status could Financial Disclosure: None reported.
contribute to our results. We adjusted for ethnic group Funding/Support: Support for this work was provided
and years of education without an appreciable change in by National Institutes of Health grants AG07232,
our results for amnestic MCI. Bias related to the selec- AG07702, 1K08AG20856-01, and RR00645, the Charles
tion of an older cohort of community-dwelling survi- S. Robertson Memorial Gift for Research on Alzheimer’s
vors is also possible. Finally, it is possible that our re- Disease, the Blanchette Hooker Rockefeller Founda-
sults are explained by chance. However, our findings are tion, and the New York City Council Speaker’s Fund for
consistent with published work relating diabetes to AD Public Health Research.
in the context of biologically plausible mechanisms.
The prevalence of diabetes was appreciably higher in
our study compared with others that have examined the REFERENCES
association between diabetes and cognitive impairment,
which may raise the issue of sampling bias. However, the 1. Peila R, Rodriguez BL, Launer LJ. Type 2 diabetes, APOE gene, and the risk for
high prevalence of diabetes in our study is explained by dementia and related pathologies: the Honolulu-Asia Aging Study. Diabetes. 2002;
51:1256-1262.
the fact that it is twice as high in African American and 2. Luchsinger JA, Tang MX, Stern Y, Shea S, Mayeux R. Diabetes mellitus and risk
Hispanic elderly persons,12,53 who constituted 77.5% of of Alzheimer’s disease and dementia with stroke in a multiethnic cohort. Am J
our sample. The prevalence of self-reported diabetes has Epidemiol. 2001;154:635-641.
increased at a 3-fold faster rate in minority compared with 3. Ott A, Stolk RP, van Harskamp F, Pols HA, Hofman A, Breteler MM. Diabetes
mellitus and the risk of dementia: the Rotterdam Study. Neurology. 1999;53:
white populations.54 Thus, the high prevalence of dia- 1937-1942.
betes in our sample would be expected in the elderly popu- 4. Ott A, Stolk RP, Hofman A, van Harskamp F, Grobbee DE, Breteler MM. Asso-
lation of New York City and is in line with national es- ciation of diabetes mellitus and dementia: the Rotterdam Study. Diabetologia.
timates from the national surveys. However, most data 1996;39:1392-1397.
on Hispanic populations from national surveys pertain 5. Leibson CL, Rocca WA, Hanson VA, et al. Risk of dementia among persons with
diabetes mellitus: a population- based cohort study. Am J Epidemiol. 1997;
to Mexican Hispanic populations. Virtually all of our His- 145:301-308.
panic population is Caribbean Hispanic, mostly from the 6. Luchsinger JA, Tang MX, Shea S, Mayeux R. Hyperinsulinemia and risk of Alz-
Dominican Republic, and our study is one of the few heimer disease. Neurology. 2004;63:1187-1192.
sources of information on diabetes prevalence in this 7. Farris W, Mansourian S, Chang Y, et al. Insulin-degrading enzyme regulates the
levels of insulin, amyloid beta-protein, and the beta-amyloid precursor protein
population. One of the consequences of the higher preva- intracellular domain in vivo. Proc Natl Acad Sci U S A. 2003;100:4162-4167.
lence of diabetes in African American and Hispanic people 8. Qiu WQ, Walsh DM, Ye Z, et al. Insulin-degrading enzyme regulates extracellu-
is that the risk of MCI attributable to diabetes in these lar levels of amyloid beta- protein by degradation. J Biol Chem. 1998;273:32730-
groups is approximately twice that of non-Hispanic white 32738.
populations, underscoring the importance of diabetes as 9. Selkoe DJ. The origins of Alzheimer disease: a is for amyloid. JAMA. 2000;283:
1615-1617.
a risk factor for MCI particularly in African American and 10. Smith MA, Sayre LM, Perry G. Diabetes mellitus and Alzheimer’s disease: gly-
Hispanic elderly populations. cation as a biochemical link. Diabetologia. 1996;39:247.
The main limitation of our study is the ascertain- 11. Harris MI. Diabetes in America: epidemiology and scope of the problem. Diabe-
ment of diabetes by self-report. We did not have mea- tes Care. 1998;21(suppl 3):C11-C14.
12. Luchsinger JA Diabetes. In: Aguirre-Molina M, Molina CW, Zambrana RE, eds.
sures of glycemia and could not ascertain undiagnosed Health Issues in the Latino Community. San Francisco, Calif: Jossey-Bass; 2001:
diabetes or glucose intolerance. Had we defined diabe- 277-300.
tes by self-report and fasting glycemia, our estimates of 13. Luchsinger JA, Reitz C, Honig LS, Tang MX, Shea S, Mayeux R. Aggregation of
the prevalence of diabetes would have been higher than vascular risk factors and risk of incident Alzheimer disease. Neurology. 2005;
we report. Thus, it is likely that our results are biased 65:545-551.
14. Petersen RC. Normal aging, mild cognitive impairment, and early Alzheimer’s
toward the null hypothesis and that our findings under- disease. Neurologia. 1995;1:326-344.
estimate the true association between diabetes and MCI. 15. Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med. 2004;
Our results provide further support to the poten- 256:183-194.
tially important independent role of diabetes in the patho- 16. Petersen RC, Doody R, Kurz A, et al. Current concepts in mild cognitive impairment.
Arch Neurol. 2001;58:1985-1992.
genesis of AD. Diabetes may also be a risk factor for non- 17. Luchsinger JA, Mayeux R. Cardiovascular risk factors and Alzheimer’s disease.
amnestic forms of MCI and cognitive impairment, but Curr Atheroscler Rep. 2004;6:261-266.
our analyses need to be repeated in a larger sample. 18. Tang MX, Stern Y, Marder K, et al. The APOE-epsilon4 allele and the risk of Alz-
heimer disease among African Americans, whites, and Hispanics. JAMA. 1998;
Accepted for Publication: April 24, 2006. 279:751-755.
19. Stern Y, Andrews H, Pittman J, et al. Diagnosis of dementia in a heterogeneous
Correspondence: José A. Luchsinger, MD, Department population: development of a neuropsychological paradigm-based diagnosis of
of Medicine, Columbia University, 630 W 168th St, dementia and quantified correction for the effects of education. Arch Neurol. 1992;
PH9E-105, New York, NY 10032 (jal94@columbia.edu). 49:453-460.

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


574

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015
20. Buschke H, Fuld PA. Evaluating storage, retention, and retrieval in disordered 38. Mayeux R, Ottman R, Maestre G, et al. Synergistic effects of traumatic head in-
memory and learning. Neurology. 1974;24:1019-1025. jury and apolipoprotein-epsilon 4 in patients with Alzheimer’s disease. Neurology.
21. Benton AL. The Benton Visual Retention Test. New York, NY: Psychological Corp; 1995;45:555-557.
1955. 39. Fleiss JL. Statistical Methods for Rates and Proportions. 2nd ed. New York: Jo-
22. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical method seph Wiley & Sons; 1981.
for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975; 40. Cox DR, Oakes D. Analysis of Survival Data. London, England: Chapman & Hall;
12:189-198. 1984.
23. Weschler D. Wechsler Memory Scale-Revised Manual. New York, NY: Psycho- 41. Kahn HA, Sempos CT. Statistical Methods in Epidemiology. Oxford, England: Ox-
logical Corp; 1987. ford University press; 1989.
24. Mattis S. Mental Status Examination for Organic Mental Syndrome in the Elderly 42. Luis CA, Loewenstein DA, Acevedo A, Barker WW, Duara R. Mild cognitive im-
Patient. New York, NY: Grune & Stratton; 1976. pairment: directions for future research. Neurology. 2003;61:438-444.
25. Kaplan E, Goodglass H, Weintraub S. Boston Naming Test. Philadelphia, Pa: Lea 43. Curb JD, Rodriguez BL, Abbott RD, et al. Longitudinal association of vascular
& Febiger; 1983. and Alzheimer’s dementias, diabetes, and glucose tolerance. Neurology. 1999;
26. Benton AL, Hamsher KD. Multilingual Aphasia Examination. Iowa City: Univer- 52:971-975.
sity of Iowa; 1976. 44. Tatemichi TK, Desmond DW, Paik M, et al. Clinical determinants of dementia re-
27. Goodglass H, Kaplan E. Assessment of Aphasia and Related Disorders. Phila-
lated to stroke. Ann Neurol. 1993;33:568-575.
delphia, Pa: Lea & Febiger; 1983.
45. Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ, Breteler MM.
28. Goodglass H, Kaplan E. The Assessment of Aphasia and Related Disorders. Phila-
Silent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med.
delphia, Pa: Lea & Febiger; 1972.
2003;348:1215-1222.
29. Rosen W. The Rosen Drawing Test. Bronx, NY: Veterans Administration Medi-
46. Honig LS, Kukull W, Mayeux R. Atherosclerosis and AD: analysis of data from
cal Center; 1981.
the US National Alzheimer’s Coordinating Center. Neurology. 2005;64:
30. Manly JJ, Bell-McGinty S, Tang M-X, Schupf N, Stern Y, Mayeux R. Implement-
494-500.
ing diagnostic criteria and estimating frequency of mild cognitive impairment in
47. Honig LS, Tang MX, Albert S, et al. Stroke and the risk of Alzheimer disease.
an urban community. Arch Neurol. 2005;62:1739-1746.
Arch Neurol. 2003;60:1707-1712.
31. American Psychiatric Association. Diagnostic and Statistical Manual of Mental
48. Reaven GM, Laws A. Insulin Resistance: The Metabolic Syndrome X. Totowa,
Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.
32. Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for NJ: Humana Press; 1999.
the staging of dementia. Br J Psychiatry. 1982;140:566-572. 49. Sasaki N, Fukatsu R, Tsuzuki K, et al. Advanced glycation end products in Alz-
33. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical heimer’s disease and other neurodegenerative diseases. Am J Pathol. 1998;
diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group un- 153:1149-1155.
der the auspices of Department of Health and Human Services Task Force on 50. Markesbery WR. Oxidative stress hypothesis in Alzheimer’s disease. Free Radic
Alzheimer’s Disease. Neurology. 1984;34:939-944. Biol Med. 1997;23:134-147.
34. Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL, DeKosky ST. 51. Tang MX, Cross P, Andrews H, et al. Incidence of AD in African-Americans, Ca-
Early detection of dementia: mild cognitive impairment (an evidence-based re- ribbean Hispanics, and Caucasians in northern Manhattan. Neurology. 2001;
view): report of the Quality Standards Subcommittee of the American Academy 56:49-56.
of Neurology. Neurology. 2001;56:1133-1142. 52. Scarmeas N, Stern Y. Cognitive reserve and lifestyle. J Clin Exp Neuropsychol.
35. Lopes-Virella MF, Stone P, Ellis S, Colwell JA. Cholesterol determination in high- 2003;25:625-633.
density lipoproteins separated by three different methods. Clin Chem. 1977; 53. Harris MI, Flegal KM, Cowie CC, et al. Prevalence of diabetes, impaired fasting
23:882-884. glucose, and impaired glucose tolerance in U.S. adults: The Third National Health
36. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low- and Nutrition Examination Survey, 1988-1994. Diabetes Care. 1998;21:
density lipoprotein cholesterol in plasma, without use of the preparative 518-524.
ultracentrifuge. Clin Chem. 1972;18:499-502. 54. Centers for Disease Control and Prevention. Trends in the prevalence and inci-
37. Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene dence of self-reported diabetes mellitus–United States, 1980-1994. MMWR Morb
amplification and cleavage with HhaI. J Lipid Res. 1990;31:545-548. Mortal Wkly Rep. 1997;46:1014-1018.

Announcement

Visit www.archneurol.com. As an individual sub-


scriber you can send an e-mail to a friend. You may send
an e-mail to a friend that includes a link to an article and
a note if you wish. Links will go to free abstracts when-
ever possible.

(REPRINTED) ARCH NEUROL / VOL 64, APR 2007 WWW.ARCHNEUROL.COM


575

©2007 American Medical Association. All rights reserved.


Downloaded From: http://archneur.jamanetwork.com/ by nopriyan pujokusuma on 04/14/2015