Duration of Diabetes and Risk of Ischemic Stroke

Text-only Preview

Duration of Diabetes and Risk of Ischemic Stroke
The Northern Manhattan Study
Chirantan Banerjee, MBBS, MPH; Yeseon P. Moon, MS; Myunghee C. Paik, PhD;
Tatjana Rundek, MD, PhD; Consuelo Mora-McLaughlin, BS; Julio R. Vieira, MD, MS;
Ralph L. Sacco, MD, MS; Mitchell S.V. Elkind, MD, MS
Background and Purpose--Diabetes increases stroke risk, but whether diabetes status immediately before stroke improves
prediction and whether duration is important are less clear. We hypothesized that diabetes duration independently
predicts ischemic stroke.
Methods--Among 3298 stroke-free participants in the Northern Manhattan Study, baseline diabetes and age at diagnosis
were determined. Incident diabetes was assessed annually (median, 9 years). Cox proportional hazard models were used
to estimate hazard ratios (HR) and 95% CI for incident ischemic stroke using baseline diabetes, diabetes as a
time-dependent covariate, and duration of diabetes as a time-varying covariate; models were adjusted for demographic
and cardiovascular risk factors.
Results--Mean age was 69 10 years (52% Hispanic, 21% white, and 24% black); 22% had diabetes at baseline and 10%
had development of diabetes. There were 244 ischemic strokes, and both baseline diabetes (HR, 2.5; 95% CI, 1.9 -3.3)
and diabetes considered as a time-dependent covariate (HR, 2.4; 95% CI, 1.8 -3.2) were similarly associated with stroke
risk. Duration of diabetes was associated with ischemic stroke (adjusted HR, 1.03 per year with diabetes; 95% CI,
1.02-1.04). Compared to nondiabetic participants, those with diabetes for 0 to 5 years (adjusted HR, 1.7; 95% CI,
1.1-2.7), 5 to 10 years (adjusted HR, 1.8; 95% CI, 1.1-3.0), and
10 years (adjusted HR, 3.2; 95% CI, 2.4 - 4.5) were
at increased risk.
Conclusions--Duration of diabetes is independently associated with ischemic stroke risk adjusting for risk factors. The risk
increases 3% each year, and triples with diabetes
10 years. (Stroke. 2012;43:1212-1217.)
Key Words: acute stroke
cerebral infarct
diabetes mellitus
risk factors
Diabetes mellitus is a major public health burden. In the such as coronary heart disease, cardiovascular mortality,
United States,
23.6 million people have diabetes.1
peripheral arterial disease, carotid wall thickness, and thin
Prevalence is increasing as the population ages and certain
cap fibroatheroma,12-18 the association between duration of
populations, such as minority groups, are more vulnerable.
diabetes and stroke risk is less well-studied.
The number of people with diabetes diagnosed is estimated
The Northern Manhattan Study prospective cohort, with its
to increase 165% between 2000 and 2050.2 The indepen-
annual evaluation of diabetes diagnosis, provides an oppor-
dent association of diabetes with stroke is now well-
tunity to investigate the utility of incorporating serial interval
documented.3-10 Notably, almost all prospective cohort
evaluations of diabetes into risk prediction estimates. We
studies have evaluated diabetes as an exposure at baseline,
investigated the effect of updated diabetes status on stroke by
or the time of participant enrollment, and measured its effect
incorporating diabetes as a time-dependent covariate. We
on stroke outcomes during follow-up. In large cohorts with
further examined the effect of duration of diabetes on
long periods of follow-up, a significant proportion of partic-
ischemic stroke risk.
ipants who did not have diabetes at baseline had development
of diabetes over the period of follow-up.11 Taking this change
Materials and Methods
in status into consideration may be expected to provide a
Northern Manhattan Study is a prospective population-based cohort
more precise estimate of stroke risk.
study designed to determine stroke incidence, risk factors, and
Although some studies have assessed the relationship of
prognosis in an urban multiethnic population. Northern Manhattan is
duration of diabetes to risk of cardiovascular outcomes
defined as the area in New York City north of 145th Street, south of
Received October 15, 2011; accepted January 11, 2012.
From the Department of Epidemiology (C.B., J.R.V., M.S.V.E.), Mailman School of Public Health, Columbia University, New York, NY; Department
of Neurology (Y.P.M., C.M.M., M.S.V.E.), College of Physicians and Surgeons, Columbia University, New York, NY; Department of Neurology (T.R.,
R.L.S.), Miller School of Medicine, University of Miami, Miami, FL; Department of Biostatistics (M.C.P.), Mailman School of Public Health, Columbia
University, New York, NY; Departments of Public Health and Epidemiology and Human Genetics (R.L.S.), Miller School of Medicine, University of
Miami, Miami, FL.
Correspondence to Mitchell S.V. Elkind, MD, MS, 710 West 168th Street, Room 642, New York, NY 10032. E-mail [email protected]
(c) 2012 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org
DOI: 10.1161/STROKEAHA.111.641381

Banerjee et al
Diabetes and Risk of Stroke
218th Street, bound on the west by the Hudson River, and separated
the diagnosis of diabetes was consistent with a diagnosis of diabetes
from the Bronx on the east by the Harlem River. The cohort has a
in 92% of cases. We reflected these changes in the analyses.
racial/ethnic mixture consisting of 52.3% Hispanic, 24.3% non-
Similarly, there were 87 participants who self-reported the diag-
Hispanic black, and 20.9% non-Hispanic white residents.
nosis of diabetes once but never reported any subsequent treatment.
All these cases were confirmed to have diabetes on medical record
Selection of Prospective Cohort
The study has been previously described in detail.19-22 Briefly,
To avoid potential bias caused by intensive medical record review
community participants were eligible for enrollment if they: (1) had
selectively among those with possible diabetes conversion during
never had a stroke diagnosed; (2) were 40 years of age or older; and
follow-up, and to validate the nonreport of diabetes diagnosis or
(3) resided for
3 months in a household with a telephone in
treatment among the remainder of the cohort, medical record review
northern Manhattan. Subjects were identified with random digit
of a random computer-generated list of 50 participants who never
dialing using dual-frame sampling to identify both published and
reported diabetes during follow-up was performed. Two participants
unpublished phone numbers. The protocol was approved by the
had the diagnosis in the medical record but it had not been detected
Institutional Review Board at Columbia University Medical Center
in interview that year. On further review, it was determined that these
and the Miller School of Medicine, University of Miami, and
participants did not report the diagnosis because it was "diet-
participants provided informed consent.
controlled diabetes" at that time, and they reported the diagnosis on
subsequent follow-up once they started using medication.
Baseline Evaluation
Measurement of Outcome
Baseline data were collected via interviews by trained research
assistants, medical record review, physical and neurological exami-
Stroke was defined during follow-up by the first symptomatic
nation by study investigators, in-person measurements, and collec-
occurrence of any type of stroke including intracerebral hemorrhage,
tion of fasting blood specimens for glucose and lipid measurements.
subarachnoid hemorrhage, and cerebral infarction, as previously
A standardized questionnaire was adapted from the Behavioral Risk
described.19 Medical records were reviewed to verify details of
Factor Surveillance System23 developed by the Centers for Disease
suspected events. At least 2 neurologists reviewed data indepen-
Control and Prevention regarding the following conditions: diabetes,
dently and classified strokes. Any disagreements were adjudicated
hypertension, hypercholesterolemia, smoking, peripheral vascular
by the principal investigators (R.L.S. and M.E.). We used only
disease, transient ischemic attack, and cardiac disease (including
ischemic strokes for current analyses.
angina, myocardial infarction, coronary artery disease, atrial fibril-
lation, and valvular heart disease).
Statistical Analyses
All analyses were performed using SAS version 9.1 (SAS Institute).
Interval Evaluation
The distributions of diabetes at baseline were calculated, both overall
All participants were prospectively followed-up annually through
and by subject characteristics, including demographics and risk
telephone interviews, and mean duration of follow-up at time of
factors. Cox proportional hazard regression models were fitted to
calculate hazard ratios (HR) and 95% CI for ischemic stroke as the
analysis was 9.0 3.7 years. The yearly contact rate was 99%.
outcome. Main predictors were baseline diabetes, diabetes as time-
Subjects were interviewed to determine changes in vital status, detect
varying covariate (incorporating new-onset diabetes during follow-
cardiac and neurological symptoms and events, and review any
up), and duration of diabetes as time-varying covariate. We used
hospitalizations. The telephone assessment served as a screen for
duration of diabetes as a continuous measure as well as categorized
vascular events. The simple stroke question ("Since your last contact
at 5 and 10 years to examine for threshold effects.
have you been diagnosed with a stroke?") during telephone interview
Models unadjusted and adjusted for demographic factors (age, sex,
was 92% sensitive and 95% specific with in-person assessment,
race- ethnicity, insurance status, and educational level) and behav-
physician interviews, medical records, and neuroimaging data used
ioral and medical risk factors (hypertension, cardiac disease, high-
as the gold standards.21 Participants with affirmative responses to
density lipoprotein, low-density lipoprotein, current smoking, past
neurological symptoms underwent examination and review by a
smoking, alcohol consumption, waist circumference, and physical
study neurologist or had medical records reviewed. Hospital surveil-
activity) were constructed. Assessment for 2-way interactions was
lance of admission and discharge were performed to provide data
conducted. We compared the Akaike Information Criterion among
that may have been missed during the annual telephone follow-up.
the baseline and the time-varying diabetes models.
Measure of Exposure
Diabetes at baseline was defined if a participant reported a history of
medical diagnosis of diabetes mellitus or treatment with oral hypo-
Baseline Characteristics
glycemic agents or insulin. In addition, fasting blood glucose
Baseline characteristics of the cohort are shown in Table 1.
mg/dL (6.5mmol/L) was used among those who did not self-report
The mean age was 69 10 years; 62.8% of the cohort were
diabetes to adjudicate diabetic status and find "unaware" cases.
Fasting blood glucose was measured using a Hitachi 747 automated
women, 52.3% were Hispanic, 20.9% were non-Hispanic
spectrometer (Boehringer). Age at time of diagnosis was also
white, and 24.3% were non-Hispanic black.
recorded for those with self-report of diabetes at baseline, and
At baseline, 574 participants (17.4%) self-reported diabe-
diabetes duration was calculated. During follow-up evaluations
tes and 142 subjects had fasting blood glucose
126 mg/dL,
among nondiabetic participants, the first follow-up contact at which
for a total of 716 (21.8%) with diabetes at baseline. Approx-
there was self-report of new diagnosis of diabetes, treatment with
oral hypoglycemic drugs, or insulin therapy was used to define
imately 93% to 96% of participants visited their primary care
conversion to diabetes during follow-up. Duration of diabetes was
physician at least once during the previous year for each year
calculated from the onset of diabetes up to the date of ischemic
of follow-up. Among those who did not have diabetes at
stroke or censoring.
baseline (n 2582), 338 subjects (13.1%) reported new-onset
diabetes during a mean 9.0 years of follow-up.
Validation of Diabetes Status Determined
During Follow-Up

Diabetes at Baseline and as a
There were 74 instances in which a participant reported treatment
Time-Dependent Covariate
with antidiabetic drugs or insulin without ever reporting the diagno-
sis of diabetes. Medical records were reviewed to validate these
There were 244 incident ischemic strokes. Baseline diabetes
responses. The self-report of medications or insulin without report of
was associated with risk of stroke (unadjusted HR, 2.6; 95%

May 2012
Table 1.
Baseline Sociodemographic and Cardiovascular Risk Factors in the Northern Manhattan Study
Sociodemographic and
Cardiovascular Risk Factors,
Diabetes at
No Diabetes at
New Diabetes
N (%) or Mean SD
Without Diabetes
During Follow-Up
P Value*
716 (21.7%)
2582 (78.2%)
2244 (68.0%)
338 (10.2%)
Age (y)
69 10
69 8
69 11
70 11
66 9
2071 (62.8%)
438 (61.2%)
1633 (63.2%)
1425 (63.5%)
208 (61.5%)
1227 (37.2%)
278 (38.8%)
949 (36.8%)
819 (36.5%)
130 (38.5%)
690 (20.9%)
100 (14.0%)
590 (22.9%)
556 (24.8%)
34 (10.1%)
803 (24.3%)
196 (27.4%)
607 (23.5%)
552 (24.6%)
55 (16.3%)
1726 (52.3%)
408 (57.0%)
1318 (51.0%)
1076 (48.0%)
242 (71.6%)
Completed high school education
1511 (45.8%)
282 (39.4%)
1229 (47.6%)
1107 (49.3%)
122 (36.1%)
Medicaid/no insurance
1435 (43.5%)
375 (52.4%)
1060 (41.1%)
870 (38.8%)
190 (56.2%)
Medicare/private insurance
1841 (55.8%)
335 (46.8%)
1506 (58.3%)
1358 (60.5%)
148 (43.8%)
Smoking status
1548 (46.9%)
317 (44.3%)
1231 (47.7%)
1076 (48.0%)
155 (45.9%)
1179 (35.7%)
273 (38.1%)
906 (35.1%)
787 (35.1%)
119 (35.2%)
569 (17.3%)
126 (17.6%)
443 (17.2%)
379 (16.9%)
64 (18.9%)
Moderate alcohol consumption
1075 (32.6%)
181 (25.3%)
894 (34.6%)
796 (35.5%)
98 (29.0%)
Any physical activity
1909 (57.9%)
380 (53.1%)
1529 (59.2%)
1352 (60.2%)
177 (52.4%)
Cardiac disease
792 (24.0%)
210 (29.3%)
582 (22.5%)
500 (22.3%)
82 (24.3%)
Waist circumference (in)
36.8 5.0
38.4 5.0
36.3 4.9
36 5.0
38.4 4.6
High-density lipoprotein (mg/dL)
46.8 14.6
43.7 14.0
47.6 14.6
48.4 14.7
42.7 13.3
Low-density lipoprotein (mg/dL)
129.1 36.1
126.3 39.1
130.0 35.3
129.8 35.3
130.8 35.0
Systolic blood pressure (mm Hg)
143.7 21.0
146.7 20.0
142.9 21.3
142.7 21.2
144.2 22.0
SD indicates standard deviation.
*Diabetes at baseline vs no diabetes at baseline.
CI, 2.0 -3.3). In the fully adjusted model, adjusted for
group. Without assuming linearity, the trichotomized dura-
demographic and other cardiovascular risk factors including
tion of diabetes variable was fitted adjusting for other risk
smoking, alcohol consumption, low-density lipoprotein, high-
factors. The null hypothesis that all 3 groups had the same
density lipoprotein, blood pressure, waist circumference,
risk of stroke was rejected ( 2 test with 2 degrees of freedom,
history of cardiac disease, and physical activity, the associa-
P 0.01). Compared to those without diabetes, those with
tion was unchanged (adjusted HR, 2.5; 95% CI, 1.9 -3.3;
Table 2.
Risk of Ischemic Stroke Associated With Baseline
Table 2). When new-onset diabetes was taken into account as
Diabetes and Diabetes as Time-Dependent Covariate
a time-varying covariate, diabetes was still associated with
risk of ischemic stroke (adjusted HR, 2.4; 95% CI, 1.8 -3.2).
95% Confidence
The Akaike Information Criteria for the models were similar
(3333.1 for the model using baseline diabetes and 3335.2 in
Baseline diabetes
the model using diabetes as a time-dependent covariate).
Unadjusted (diabetes only)
There were no interactions between diabetes and age, race--
Adjusted for demographic variables*
ethnicity, or sex.
Adjusted for demographic variables
and cardiovascular risk factors
Duration of Diabetes
Diabetes as time-dependent covariate
The mean duration of diabetes among people who self-
Unadjusted (diabetes only)
reported diabetes at baseline was 17.3 11.6 years (median,
Adjusted for demographic variables*
13.7 years). Among the 338 subjects with diabetes diagnosed
Adjusted for demographic variables
during follow-up, mean duration was 4.5 3.2 years (median,
and cardiovascular risk factors
4.2 years). With each year of diabetes, stroke risk increased
*Age, sex, education, race- ethnicity, and insurance.
by 3% (adjusted HR per year, 1.03; 95% CI, 1.02-1.04).
Age, sex, education, race- ethnicity, insurance, waist circumference,
Duration was also categorized as
5 years, 5 to 10 years,
alcohol consumption, smoking status, physical activity, high-density lipoprotein,
10 years, with nondiabetic participants as a reference
low-density lipoprotein, history of cardiac disease, and systolic blood pressure.

Banerjee et al
Diabetes and Risk of Stroke
Table 3.
Risk of Ischemic Stroke Associated With Duration
follow-up does not confer added information. Second, sub-
of Diabetes
jects with newly diagnosed diabetes may be more compliant
with treatment, which has been shown to be beneficial for
95% Confidence
primary stroke prevention in our cohort.25 Third, the median
Diabetes Duration
Hazard Ratio*
P Value
duration of follow-up for those with diabetes at baseline in
Continuous model
our cohort was 13.7 years, but it was only 4.2 years for those
Per year
with development of diabetes after baseline, which may not
Categorical model
be sufficient to manifest cerebrovascular events. Last, we
5 y
used both self-report and laboratory results to identify diabe-
5-10 y
tes at baseline. However, self-report alone was used to define
10 y
diabetes during telephone follow-up, which may have led us
to miss cases, because nearly one-third of diabetes cases may
*Adjusted for age, sex, race- ethnicity, education, insurance, waist circum-
ference, smoking status, alcohol consumption, physical activity, systolic blood
be undiagnosed.26
pressure, history of cardiac disease, low-density lipoprotein cholesterol,
Among those with diabetes
10 years, risk of ischemic
high-density lipoprotein cholesterol.
stroke is 3-times the risk among those without diabetes. Our
Participants without diabetes as reference group.
study provides evidence that the risk of ischemic stroke
increased continuously with duration of diabetes mellitus.
5 years (adjusted HR, 1.7; 95% CI, 1.1-2.7), 5 to
The increase is not as much during the second half of the first
10 years (adjusted HR, 1.8; 95% CI, 1.1-3.0), and
10 years
decade, but it increases steeply as the disease enters its second
(adjusted HR, 3.2; 95% CI, 2.4 - 4.5) had an increased risk of
decade. This must, however, be interpreted keeping in mind
stroke (Table 3 and Figure).
that true onset of diabetes may be 4 to 7 years earlier than
clinical diagnosis.27
This is the first prospective cohort study to address the
In this prospective cohort, diabetes at time of enrollment was
association of diabetes duration and ischemic stroke among
associated with ischemic stroke risk, consistent with esti-
both men and women. The Nurses' Health study reported an
mates of association reported in other studies, varying from
association between diabetes duration and various stroke
1.3 to 4.0.3-10,24 Contrary to our hypothesis, however, the
subtypes among women,3 in which the risk of ischemic stroke
magnitude of the association for diabetes with stroke risk was
increased from 1.5 (0 - 4 years) to 4.1 ( 20 years). The
no different when we included diabetes as a time-dependent
maximum increase in the risk was seen at the 10-year mark,
covariate. In traditional epidemiological analyses, the use of
similar to our findings. Our cohort has men and women 40
only baseline assessments of a risk factor, such as diabetes,
years of age or older and representation by Hispanic, white,
could potentially bias study results toward the null. Our
and black participants, as compared to the Nurses' Health
findings suggest that there is marginal incremental value to
study cohort, comprising predominantly white women ages
including further assessments of diabetes during follow-up in
30 to 55 years at the time of enrollment.
analyses of its effect on stroke risk. One practical implication
Several potential mechanisms could explain the association
for future epidemiological studies would be potential cost-
of diabetes duration and stroke in our study. There is evidence
savings through avoiding additional lengthy interviews and
of association between diabetes duration and atherosclerotic
assessments of risk factors during follow-up. Whether these
lesions, including intimal medial thickness and thin cap
findings are transferable to other risk factors and cardiovas-
fibroatheromas.13,16 Carotid plaque thickness has been shown
cular outcomes is not clear from these analyses.
to predict ischemic stroke in our cohort.28 In addition,
There are several possible explanations for this absence of
hypertension is twice as prevalent among those with diabetes
additional information from follow-up assessments in our
as in people without diabetes,29 and long-term hypertension
analyses. The cardiovascular risk factor burden carried by the
causes accelerated microvascular and macrovascular compli-
participants at enrollment (mean age, 69 years) could already
cations among those with diabetes.29 The risk of microalbu-
be high enough that development of diabetes during
minuria has been shown to increase with increasing duration
of diabetes30,31 and microalbuminuria has been reported as a
strong and independent risk factor of stroke among patients
with diabetes.32 Other potential mediators may be endothelial
dysfunction33 and abnormalities in fibrinogen and clotting
Our study has public health implications. Although stroke
rates have been declining among those with diabetes,36 the
rapid increase in diabetes incidence over the same period is
leading to a higher overall stroke burden.36 In recent decades,
the age of onset of type 2 diabetes has decreased, paralleling
the obesity epidemic in young adults.37 As the population
ages and the elderly live longer, more and more people will
live with longer duration of the disease. It is thus important to
Figure. Risk of ischemic stroke and duration of diabetes.
better-understand the dynamics between diabetes, time, and

May 2012
stroke, and to emphasize the importance of interventions to
prevent early diabetes. Minimizing the number of years a
patient has diabetes would help combat the increase in stroke
risk with each year of the disease.
Our study has several strengths. Northern Manhattan Study
1. American Diabetes Association. Diabetes Basics: Diabetes statistics.
is designed to focus on risk factors for stroke in whites,
American Diabetes Association website. Available at: www.diabetes.org/
diabetes-statistics.jsp. Accessed October 4, 2009.
blacks, and Hispanics living in the same community. The
2. Narayan KM, Boyle JP, Thompson TJ, Sorensen SW, Williamson DF.
study has a large sample size, long duration of follow-up,
Lifetime risk for diabetes mellitus in the United States. JAMA. 2003;290:
minimal loss to follow-up, and detailed information on
1884 -1890.
potential confounding factors. Our study design also allowed
3. Janghorbani M, Hu FB, Willett WC, Li TY, Manson JE, Logroscino G,
us to use diabetes as a time-dependent covariate to model risk
et al. Prospective study of type 1 and type 2 diabetes and risk of stroke
subtypes: the Nurses' Health Study. Diabetes Care. 2007;30:1730 -1735.
of stroke and to study the relationship between diabetes
4. Jorgensen H, Nakayama H, Raaschou HO, Olsen TS. Stroke in patients
duration and ischemic stroke among both men and women.
with diabetes: The Copenhagen Stroke Study. Stroke. 1994;25:
However, this study is not without limitations. First,
because we were limited to using self-report to determine
5. Rodriguez BL, D'Agostino R, Abbott RD, Kagan A, Burchfiel CM, Yano
K, et al. Risk of hospitalized stroke in men enrolled in the Honolulu Heart
diabetes during follow-up, we may have misclassified "un-
Program and the Framingham Study: a comparison of incidence and risk
aware" individuals with diabetes as not having diabetes,
factor effects. Stroke. 2002;33:230 -236.
leading to a bias toward the null. Use of quantifiable
6. Kissela BM, Khoury J, Kleindorfer D, Woo D, Schneider A, Alwell K, et
measures of glycemic status such as fasting blood glucose or
al. Epidemiology of ischemic stroke in patients with diabetes: the greater
hemoglobin A1c as time-dependent covariates may have
Cincinnati/Northern Kentucky Stroke Study. Diabetes Care. 2005;28:
added additional prognostic information. However, our co-
7. Folsom AR, Rasmussen ML, Chambless LE, Howard G, Cooper LS,
hort may be atypical in that there was a high degree of
Schmidt MI, et al. Prospective associations of fasting insulin, body fat
follow-up with primary physicians (93%-96% annually),
distribution, and diabetes with risk of ischemic stroke: the Atherosclerosis
which may have led to a higher likelihood of diagnosis with
Risk in Communities (ARIC) Study investigators. Diabetes Care. 1999;
diabetes. However, although we do have data collected
8. Lehto S, Ronnemaa T, Pyorala K, Laakso M. Predictors of stroke in
during the annual follow-up interview on visits to primary
middle-aged patients with non-insulin-dependent diabetes. Stroke. 1996;
care doctors, we do not have data on whether diabetes
27:63- 68.
screening occurred during those visits. Second, the duration
9. Almdal T, Scharling H, Jensen JS, Vestergaard H. The independent effect
of follow-up may not have been long enough to bring out the
of type 2 diabetes mellitus on ischemic heart disease, stroke, and death:
a population-based study of 13,000 men and women with 20 years of
difference between baseline and time-dependent models.
follow-up. Arch Intern Med. 2004;164:1422-1426.
Third, duration of diabetes at baseline was calculated based
10. Davis TM, Millns H, Stratton IM, Holman RR, Turner RC. Risk factors
on participants' self-reported age of onset, which is vulnera-
for stroke in type 2 diabetes mellitus: United Kingdom Prospective
ble to inaccuracy, because there is a lag time between onset
Diabetes Study (UKPDS) 29. Arch Intern Med. 1999;159:1097-1103.
11. Yeh HC, Duncan BB, Schmidt MI, Wang NY, Brancati FL. Smoking,
and diagnosis.27 The apparent threshold of 10 years identified
smoking cessation, and risk for type 2 diabetes mellitus: a cohort study.
in this study, therefore, may be an underestimate. The
Ann Intern Med. 2010;152:10 -17.
analysis of duration and stroke are complicated by the fact
12. Fox CS, Sullivan L, D'Agostino RB Sr, Wilson PW, Framingham Heart
that longer duration is associated with older age, and residual
Study. The significant effect of diabetes duration on coronary heart
disease mortality: the Framingham Heart Study. Diabetes Care. 2004;27:
confounding cannot be ruled out. We did not have sufficient
704 -708.
numbers to detect interaction by sex, age, and race/ethnicity,
13. Wagenknecht LE, D'Agostino R Jr, Savage PJ, O'Leary DH, Saad MF,
and thus cannot comment on the differential effect of diabetes
Haffner SF. Duration of Diabetes and Carotid Wall Thickness The Insulin
or its duration in segments of the cohort. However, our
Resistance Atherosclerosis Study (IRAS). Stroke. 1997;28:999 -1005.
findings are in agreement with other large population-based
14. Brun E, Nelson RG, Bennett PH, Imperatore G, Zoppini G, Verlato G, et
al. Diabetes duration and cause-specific mortality in the Verona Diabetes
Study. Diabetes Care. 2000;23:1119 -1123.
In conclusion, use of diabetes as a time-dependent covari-
15. Natarajan S, Liao Y, Sinha D, Cao G, McGee DL, Lipsitz SR. Sex
ate adds little incremental value to using diabetes at baseline
differences in the effect of diabetes duration on coronary heart disease
as a risk factor for stroke. Duration of diabetes, however,
mortality. Arch Intern Med. 2005;165:430 - 435.
16. Lindsey JB, House JA, Kennedy KF, Marso SP. Diabetes duration is
increases the risk of ischemic stroke, independent of coexist-
associated with increased thin-cap fibroatheroma detected by intra-
ing risk factors. As more people have development of
vascular ultrasound with virtual histology. Circ Cardiovasc Interv. 2009;
diabetes earlier and live longer, this relationship assumes
public health importance and warrants steps to institute
17. Spijkerman AM, Dekker JM, Nijpels G, Jager A, Kostense PJ, van
long-standing and sustainable lifestyle changes for primary
Hinsbergh VW, et al. Impact of diabetes duration and cardiovascular risk
factors on mortality in type 2 diabetes: the Hoorn Study. Eur J Clin
prevention and appropriate long-term management after
Invest. 2002;32:924 -930.
18. Emanuele NV. Duration of diabetes, glucose control and cardiovascular
risk. Diabetologia. 2010;53:214 -215.
19. Sacco RL, Elkind MS, Boden-Albala B, Lin IF, Kargman DE, Hauser
WA, et al. The protective effect of moderate alcohol consumption on
The authors thank Janet DeRosa, NOMAS Project Coordinator.
ischemic stroke. JAMA. 1999;281:53- 60.
20. Sacco RL, Gan R, Boden-Albala B, Lin IF, Kargman DE, Hauser WA, et
Sources of Funding
al. Leisure-time physical activity and ischemic stroke risk: the Northern
The Northern Manhattan Study is funded by National Institutes of
Manhattan Stroke Study. Stroke. 1998;29:380 -387.
Health/National Institute of Neurological Disorders and Stroke grant
21. Eguchi K, Boden-Albala B, Jin Z, Di Tullio M R, Rundek T, Rodriguez
R37 NS 29993.
C, et al. Usefulness of fasting blood glucose to predict vascular outcomes

Banerjee et al
Diabetes and Risk of Stroke
among individuals without diabetes mellitus (from the Northern Man-
30. Orchard TJ, Dorman JS, Maser RE, Becker DJ, Drash AL, Ellis D, et al.
hattan Study). Am J Cardiol. 2007;100:1404 -1409.
Prevalence of complications in IDDM by sex and duration. Pittsburgh
22. Elkind MS, Sciacca RR, Boden-Albala B, Rundek T, Paik MC, Sacco RL.
Epidemiology of Diabetes Complications Study II. Diabetes. 1990;39:
Moderate alcohol consumption reduces risk of ischemic stroke: The
1116 -1124.
Northern Manhattan Study. Stroke. 2006;37:13-19.
31. Tuomilehto J, Borch-Johnsen K, Molarius A, Forsen T, Rastenyte D, Sarti
C, et al. Incidence of cardiovascular disease in Type 1 (insulin-dependent)
23. Gentry EM, Kalsbeek WD, Hogelin GC, Jones JT, Gaines KL, Forman
diabetic subjects with and without diabetic nephropathy in Finland.
MR, et al. The behavioral risk factor surveys: II. Design, methods, and
Diabetologia. 1998;41:784 -790.
estimates from combined state data. Am J Prev Med. 1985;1:9 -14.
32. Guerrero-Romero F, Rodriguez-Moran M. Proteinuria is an independent
24. Tuomilehto J, Rastenyte D, Jousilahti P, Sarti C, Vartiainen E. Diabetes
risk factor for ischemic stroke in non-insulin-dependent diabetes mellitus.
mellitus as a risk factor for death from stroke. Prospective study of the
Stroke. 1999;30:1787-1791.
middle-aged Finnish population. Stroke. 1996;27:210 -215.
33. Clarkson P, Celermajer DS, Donald AE, Sampson M, Sorensen KE,
25. Boden-Albala B, Cammack S, Chong J, Wang C, Wright C, Rundek T, et
Adams M, et al. Impaired vascular reactivity in insulin-dependent
al. Diabetes, fasting glucose levels, and risk of ischemic stroke and
diabetes mellitus is related to disease duration and low density lipoprotein
vascular events: findings from the Northern Manhattan Study (NOMAS).
cholesterol levels. J Am Coll Cardiol. 1996;28:573-579.
Diabetes Care. 2008;6:1132-1137.
34. Meigs JB, Mittleman MA, Nathan DM, Tofler GH, Singer DE, Murphy-
26. Thorpe LE, Upadhyay UD, Chamany S, Garg R, Mandel-Ricci J,
Sheehy PM, et al. Hyperinsulinemia, hyperglycemia, and impaired hemo-
Kellerman S, et al. Prevalence and control of diabetes and impaired
stasis: the Framingham Offspring Study. JAMA. 2000;283:221-228.
fasting glucose in New York City. Diabetes Care. 2009;32:57- 62.
35. Kannel WB, D'Agostino RB, Wilson PW, Belanger AJ, Gagnon DR.
Diabetes, fibrinogen, and risk of cardiovascular disease: the Framingham
27. Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM
experience. Am Heart J. 1990;120:672- 676.
occurs at least 4 -7 yr before clinical diagnosis. Diabetes Care. 1992;15:
36. Booth GL, Kapral MK, Fung K, Tu JV. Recent trends in cardiovascular
815- 819.
complications among men and women with and without diabetes.
28. Rundek T. Arif H Boden-Albala B Elkind MS Paik MC Sacco RL Carotid
Diabetes Care. 2006;29:32-37.
plaque, a subclinical precursor of vascular events: the Northern Man-
37. Alberti G, Zimmet P, Shaw J, Bloomgarden Z, Kaufman F, Silink M, et
hattan. Study Neurology. 2008;70:1200 -1207.
al. Type 2 diabetes in the young: The evolving epidemic. The interna-
29. Epstein M, Sowers JR. Diabetes mellitus and hypertension. Hypertension.
tional diabetes federation consensus workshop. Diabetes Care 2004;27:
1992;19:403- 418.
1798 -1811.