ORIGINAL INVESTIGATION

Risk Factors for Stroke in Type 2 Diabetes Mellitus

United Kingdom Prospective Diabetes Study (UKPDS) 29
Timothy M. E. Davis, FRACP; Helen Millns, PhD; Irene M. Stratton, MSc;
Rury R. Holman, FRCP; Robert C. Turner, MD, FRCP; for the UK Prospective Diabetes Study Group

Objective: To investigate modifiable and nonmodifi-
able risk factors for stroke in type 2 diabetes mellitus.
Patients and Methods: A total of 3776 patients aged
25 to 65 years newly diagnosed as having type 2 diabe-
tes mellitus without known cardiovascular or other se-
rious disease were studied for a median of 7.9 years. An
initial stepwise evaluation of risk factors was done in 2704
patients with all risk factors measured, with the final Cox
model analysis being of 3776 patients who had com-
plete data on the selected variables.
Results: Of 3776 patients, 99 (2.6%) had a stroke. Sig-
nificant risk factors for stroke in a multivariate model were
age (estimated hazard ratio [95% confidence interval],
4.78 [2.56-8.92] for $60 vs ,50 years), male sex (1.63
[1.08-2.47)] vs female), hypertension (2.47 [1.64-
3.74)] vs normotension), and in 3728 patients who had
electrocardiography at study entry, atrial fibrillation (8.05
[3.52-18.44] vs sinus rhythm). Obesity, lack of exer-
cise, smoking, poor glycemic control, hyperinsuline-
mia, dyslipidemia, and microalbuminuria were not sig-
nificantly associated with stroke in the model.
Conclusion: In patients with type 2 diabetes, aggres-
sive antihypertensive therapy and routine anticoagula-
tion therapy for atrial fibrillation may reduce the risk of
stroke.
Arch Intern Med. 1999;159:1097-1103

From the Department of
Medicine, University of
Western Australia, Fremantle
(Dr Davis); and Diabetes
Research Laboratories,
University of Oxford, Oxford,
England (Drs Millns, Holman,
and Turner and Mss Stratton).
A complete list of study
participants was published
previously (BMJ.
1998;317:703-713).
T
HE INCIDENCE of cardiovas-
cular disease is increased 2-
to 3-fold in patients with
type 2 diabetes mellitus, and
this increase cannot be ex-
plained by the presence of classic risk fac-
tors for atherosclerosis, such as smoking,
hypertension, and dyslipidemia.1 In patients
For editorial comment
see page 1033
with cerebrovascular disease, the presence
of type 2 diabetes increases the risk of is-
chemic cerebral infarction, which accounts
for more than three quarters of all strokes,
but is not associated with an increased risk
of cerebral hemorrhage.2,3 The crude inci-
dence of stroke among patients with type
2 diabetes can be more than 3 times that in
the general population,4-14 with particularly
high rates reported in Sweden11 and the
southeastern United States.8,15 The relative
risk of stroke in patients with type 2 dia-
betes reaches a maximum in the 40- to 60-
year-old group, and women comprise a
greater proportion of patients with stroke
than in the nondiabetic population.16
Nonmodifiable risk factors such as age,
sex, race, and heredity are associated with
stroke in subjects with16 and without17 dia-
betes. Hypertension, cardiac disease (in-
cluding atrial fibrillation), and cigarette and
alcohol use are modifiable risk factors in
patients without diabetes,17 but an associa-
tion with dyslipidemia is less clear.18 Stud-
ies19 of diabetic patients have consistently
identified hypertension as the major risk
factor. Associations have been found be-
tween stroke and other manifestations of
atherosclerosis, cardiac failure, and non-
rheumatic atrial fibrillation,20-23 but in dia-
betic patients, inadequate glycemic con-
trol, dyslipidemia, obesity, smoking, and
microvascular disease have not been iden-
tified as independent risk factors.16 Many
studies, however, have used relatively small
numbers of selected patients from cross-
sectional studies and a limited number of
risk factors in the assessment of stroke in
patients with type 2 diabetes.
We have investigated risk factors for
stroke in patients with newly diagnosed type
2 diabetes mellitus recruited to the United
Kingdom Prospective Diabetes Study
(UKPDS).24 White patients without known
atheromatous disease were observed for a

ARCH INTERN MED/ VOL 159, MAY 24, 1999

1097

©1999 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023

 PATIENTS AND METHODS
PATIENTS
The UKPDS recruited 5102 patients aged 25 to 65 years
with newly diagnosed type 2 diabetes (fasting plasma glu-
cose levels, .6 mmol/L [.108 mg/dL] on 2 occasions) be-
tween December 1, 1977, and March 31, 1991. An addi-
tional 2006 patients of similar age, sex, and fasting plasma
glucose levels were excluded because they had severe vas-
cular disease (myocardial infarction in the past year, cur-
rent angina, or heart failure), accelerated hypertension, pro-
liferative or preproliferative retinopathy, renal failure with
plasma creatinine levels of greater than 175 µmol/L (.2.0
mg/dL), other life-threatening disease such as cancer, an
illness requiring parenteral steroid therapy, an occupa-
tion precluding insulin treatment, language difficulties, or
the presence of ketonuria (urine ketone bodies, .3 mmol/
L), suggestive of type 1 diabetes mellitus.
There were 4178 white patients, of whom 381 had
known cardiovascular disease (previous myocardial infarc-
tion or electrocardiographic [ECG] Q-wave abnormality [202
patients], angina [7 patients], heart failure [1 patient], in-
termittent claudication [120 patients], and a previous stroke
or transient ischemic attack [51 patients]) and were there-
fore excluded. Biochemical measurements were not per-
formed until 1981, and some patients had no valid data for
1 or more of the other variables, leaving 2704 patients with
a complete set of risk factor information at baseline. The fi-
nal analysis was done in 3776 patients who had data relat-
ing to age, sex, and hypertension or normotension catego-
rization. The study protocol was approved by the institutional
Ethics Committee in each of the 23 centers. All recruited pa-
tients gave informed consent to participation.
PATIENTS AND METHODS
During an initial 3-month period in which patients re-
ceived dietary therapy alone, contraceptive or hormone re-
placement therapy was stopped and a loop diuretic (furo-
semide [frusemide]) was substituted for benzothiadiazide
treatment unless these changes were considered inappro-
priate on clinical grounds. After the initial dietary therapy,
patients were randomly allocated to different hypoglyce-
mic treatments according to the UKPDS protocol.24 Those
allocated to diet formed a conventional-therapy group, and
those allocated to receive sulfonylurea, insulin, or metfor-
min hydrochloride comprised an intensive-therapy group.
Patients were seen every 3 months in UKPDS clinics,
and any possibly diabetes-related clinical events were re-
corded. The administrator requested full information
from the center, general practitioner, or other health care
professionals. A file without details of randomized or ac-
tual therapies was evaluated by 2 independent clinical as-
sessors to ascertain whether predetermined criteria for
such end points were met. If the 2 assessments did not
agree, the information was presented to a panel of 3 other
median of 7.9 years. The results indicate that appropri-
ate management of hypertension and atrial fibrillation may
prove the most effective primary prevention strategies in
patients with type 2 diabetes.
ARCH INTERN MED/ VOL 159, MAY 24, 1999
1098
©1999 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023
independent senior physicians for a final decision. Each
end point was classified according to the International
Classification of Diseases, Ninth Revision,25 the codes for
stroke being 430 to 434.9 and 436.
RISK FACTOR ASSESSMENT
The height and waist and hip circumferences were mea-
sured at the time of the diagnosis of type 2 diabetes melli-
tus, and tobacco smoking and amount of exercise taken were
assessed by questionnaire. Blood pressure was recorded as
the mean of measurements taken 2 and 9 months after di-
agnosis using an electronic sphygmomanometer. Hyperten-
sion was diagnosed if the patient had a systolic blood pres-
sure of 160 mm Hg or greater, a diastolic blood pressure of
90 mm Hg or greater, or both, or if antihypertensive therapy
had already been prescribed. At the visit 3 months after ini-
tial dietary therapy, patients attended a UKPDS clinic in the
morning after a 10-hour overnight fast to have a blood speci-
men drawn for standard biochemical tests and to provide a
spot urine specimen for the assessment of albuminuria. Reti-
nopathy was assessed by 4 horizontal 30° color photo-
graphs per eye with modified Wisconsin grading.24 Electro-
cardiographic tracings were evaluated by 2 trained coders,
and a Minnesota code was assigned.24
BIOCHEMICAL MEASUREMENTS
Fasting plasma glucose levels were measured in each cen-
ter, with a monthly quality assurance scheme showing a
coefficient of variation of less than 4%. Glycosylated he-
moglobin (A 1c ) levels were measured by a high-per-
formance liquid chromatography analyzer (Bio-Rad Dia-
mat; Bio-Rad Laboratories, Hemel Hempstead, England)
with a reference range of 4.5%-6.2%. Plasma triglyceride
levels were measured using a commercial kit (GPO-PAP
[glycerol phosphate oxidase-p-aminophenazone]; Boeh-
ringer Mannheim, Lewes, East Sussex, England), without
correction for free glycerol, on a centrifugal analyzer (Co-
bas FARA; Roche Diagnostica, Welwyn Garden City, Herts,
England). Plasma cholesterol levels were measured using
a high-performance cholesterol oxidase-p-aminophe-
nazone (CHOD-PAP) method with Preciset cholesterol
standard (Kit C system; Boehringer Mannheim) on the cen-
trifugal analyzer, with low-density lipoprotein and
high-density lipoprotein cholesterol levels after precipita-
tion.26 Plasma insulin levels were measured with a radio-
immunoassay that cross-reacted 100% with proinsulin. The
urine albumin concentration was expressed relative to the
mean urine creatinine concentration of 11 mmol/L in men
and 8 mmol/L in women to allow for urine dilution.26 The
urine albumin concentration was measured by radioim-
munoassay or immunoturbidimetry.
STATISTICAL ANALYSES
Age was categorized as younger than 50 years, 50 to 54
years, 55 to 59 years, or 60 years or older. Other continuous
RESULTS
Of 3776 patients in the study, 99 (2.6%) had a stroke dur-
ing follow-up. The median duration of follow-up was 7.9
 Table 1. Baseline Characteristics of 2704 Patients
variables were categorized by tertiles calculated from Who Had Each Variable Measured*
3797 persons with no prior indication of atherosclero-
sis. Cox proportional hazards models,27 with censor- Men Women
ing at 10 years’ follow-up, were used to assess the ef- Variable (n = 1569) (n = 1135)
fect of potential risk factors on fatal or nonfatal stroke. Age, y 52 ± 9 53 ± 9
The dependent variable was the time to the first end point Body mass index, kg/m2 27.1 ± 4.7 29.4 ± 6.4
or to censoring. Relationships of single risk factors with Waist-hip ratio 0.95 ± 0.06 0.87 ± 0.08
events, after adjusting for age and sex, were assessed in Systolic BP, mm Hg 133 ± 18 139 ± 20
2704 patients with data on all risk factors. A multivar- Diastolic BP, mm Hg 82 ± 10 83 ± 10
iate selection of risk factors was made by a stepwise pro- Hypertensive, No. (%) 518 (33.0) 511 (45.0)
cedure, after adjusting for age and sex. The final mul- Fasting plasma glucose, 8.3 ± 2.8 9.2 ± 3.0
tivariate analysis was done in the 3776 patients who had mmol/L (mg/dL) (148.2 ± 50.0) (164.2 ± 53.6)
data for all the selected variables. Estimated hazard ra- Hemoglobin A1c, % 6.9 ± 1.7 7.4 ± 1.8
tios are represented graphically with 95% confidence in- Serum total cholesterol, 5.2 ± 1.0 5.7 ± 1.2
tervals estimated for each group by treating the rela- mmol/L (mg/dL) (198.4 ± 38.5) (219.2 ± 46.2)
tive risks as floating absolute risks.28 Serum LDL cholesterol, 3.3 ± 0.9 3.8 ± 1.1
The association of urine albumin concentration (cat- mmol/L (mg/dL) (126.9 ± 34.6) (146.1 ± 42.3)
egorized as ,50 or $50 mg/L for microalbuminuria and Serum HDL cholesterol, 1.04 ± 0.23 1.11 ± 0.24
,300 or $300 mg/L for proteinuria) with strokes was mmol/L (mg/dL) (40.0 ± 8.8) (42.7 ± 9.2)
determined by adding urine albumin concentrations to Serum triglyceride, 1.5 (0.9-2.5) 1.6 (1.0-2.6)
mmol/L (mg/dL) (132.2 [79.3-220.3]) (141.0 [88.1-229.1])
the multivariate model obtained by the stepwise selec-
Fasting plasma insulin, 11.2 (6.4-19.7) 13.1 (7.4-23.3)
tion procedure in 2973 patients who had a valid mea- mU/L (µU/mL)
surement. The same method was used for the presence Exercise, sedentary/ 17/32/44/7 22/39/38/1
of atrial fibrillation in 3728 patients who had a base- moderate/active/fit, %
line ECG. In 2161 patients who had retinal photo- Smoking, never/ 22/46/32 44/26/29
graphs assessed by a modified Wisconsin 191 grading ex-smoker/ current, %
of 4 color photographs per eye,24 3 categories—no reti- Atrial fibrillation, No. (%)† 19 (1.2) 9 (0.9)
nopathy, the presence of microaneurysms in 1 or both
eyes, and more severe retinopathy (eg, hard exudates, *Data are given as mean ± SD or geometric mean (SD interval) unless
hemorrhages, or intraretinal microvascular abnormali- otherwise indicated. BP indicates blood pressure; LDL, low-density lipoprotein;
ties)—were used in analysis. and HDL, high-density lipoprotein.
To assess the effect of regression to the mean on †Patients also analyzed in Table 3.
systolic blood pressures, a second measurement was
taken in a subset of 477 patients who were randomly
allocated to dietary therapy and continued this treat-
justing for age and sex, hypertension and systolic blood
ment alone for a further 12 months after randomiza- pressure were significantly and independently associ-
tion. Patients were categorized according to their base- ated with stroke (Figure), whereas the body mass in-
line value into the tertiles calculated from the 3797 dex (BMI or Quetelet index; calculated as weight in ki-
persons with no prior indication of coronary artery lograms divided by the square of the height in meters),
disease. In the highest and lowest blood pressure waist-hip ratio, smoking, exercise level, fasting plasma
groups, the difference in the means of the 2 samples glucose levels, hemoglobin A1c level, serum insulin level,
taken 12 months apart gave an estimate of regression serum lipid variables, and diastolic blood pressure were
to the mean. These differences were then applied to not (Table 2).
the respective baseline mean values of the upper and In a multivariate analysis of 3776 patients, age, sex,
lower tertiles in the 3776 patients, to correct for a
regression to the mean. To estimate the effect of a
and the presence of hypertension were the major risk fac-
10-mm Hg increment of systolic blood pressure on the tors. When hypertension was replaced by the systolic blood
risk of stroke, the systolic blood pressure was fitted in pressure, this was also a significant risk factor, with esti-
the stepwise selected Cox model as a continuous vari- mated hazard ratios of 1.96 (95% confidence interval, 1.01-
able, leaving other risk factors as categorical variables. 3.81) for the middle tertile and 2.99 (1.57-5.69) for the
This estimate was also adjusted for a regression to the top tertile. When fitted as a continuous variable, after ad-
mean. justing for a regression to the mean, for a 10-mm Hg rise
Statistical analyses were performed using commer- in blood pressure, the hazard ratio increased by 0.54.
cial software (SAS Institute, Inc, Cary, NC). They did Of the 2973 patients with a valid urine albumin mea-
not include any reference to allocated or actual therapy surement at baseline, 80 patients (2.7%) subsequently had
during the 10-year follow-up period. Data are reported
as mean ± 1 SD, geometric mean (1 SD interval), or
a nonfatal or fatal stroke. After adjusting for age and sex,
percentages. a urine microalbumin excretion of greater than 50 mg/L
was associated with stroke, with an estimated hazard ra-
tio of 2.3 (1.4-4.0) compared with a urine albumin ex-
cretion of 50 mg/L or less. When urine albumin excre-
tion was added to the model with age, sex, and the
years with an interquartile range of 6.0 to 10.0 years (the presence of hypertension, neither it nor retinopathy was
data of 27.1% of patients were censored at 10 years). Base- a significant predictor of stroke (P..10).
line data aggregated by sex for the 2704 patients with com- Of 3728 patients with a baseline ECG, 98 (2.6%) had
plete risk factor data are shown in Table 1. After ad- a stroke during follow-up. Atrial fibrillation was present

ARCH INTERN MED/ VOL 159, MAY 24, 1999

1099

©1999 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023

 7.0 7.0
6.0 6.0
5.0 5.0
Estimated Hazard Ratios 4.0 4.0

3.0 3.0

2.0 2.0

1.0 1.0

0.4 0.4
40 45 50 55 60 65 Men Women
Age, y

7.0 7.0
6.0 6.0
5.0 5.0
4.0 4.0
Estimated Hazard Ratios

3.0 3.0

2.0 2.0

1.0 1.0

0.4 0.4
No Yes 110 120 130 140 150 160
Hypertension Systolic BP, mm Hg

The estimated hazard ratios for the significant risk factors for stroke that occurred in 99 of 3776 patients with type 2 diabetes mellitus, expressed as floating
absolute risks. These assign the appropriate variances to each tertile or category to allow visual comparison of the associated risks. BP indicates blood pressure.

in 28 patients initially, and 6 (21.4%) of these had a stroke
within 10 years. Thus, of all patients with stroke (n = 99),
atrial fibrillation had been documented in 6 (6.1%). Atrial
fibrillation was a significant risk factor, with an esti-
mated hazard ratio of 8.65 (3.5-18.4), when added to the
model with age, sex, and the presence of hypertension
(P,.001) (Table 3).
COMMENT
Our data confirm the importance of nonmodifiable and
modifiable risk factors for stroke in diabetes mellitus. In-
creasing age and male sex were the nonmodifiable risk
factors present in our cohort, and hypertension and atrial
fibrillation were the only modifiable factors present among
a broad range of clinical and biochemical variables en-
tered in the model.
The risk of stroke increased progressively for the 4
age categories in the present study; a patient older than
60 years at diagnosis had almost 5 times the risk of a pa-
tient younger than 50 years at baseline. Male sex was also
associated with stroke in our diabetic patients, with men
having a relative risk of more than 1.5 times that of women.
Nevertheless, a woman with diabetes probably has less ce-
rebrovascular “protection” associated with sex because the
relative risk of stroke in women with diabetes compared
with those without diabetes (between 2.6 and 13) is gen-
erally higher than that of diabetic vs nondiabetic men.12,14
Consistent with our findings, hypertension, espe-
cially systolic hypertension, has been recognized previ-
ously16 as a major risk factor for stroke in patients with
diabetes. An evaluation of the effect of changes in blood
pressure and antihypertensive treatment from the time of
diagnosis on stroke incidence was beyond the scope of the
present study and has been reported recently else-
where.29 Nevertheless, given the difficulty in achieving ac-
ceptable blood pressure control in diabetic patients with
hypertension,30 it is not surprising that the blood pres-
sure at the time of diagnosis was a strong predictor of the
occurrence of stroke. Other features of the “metabolic syn-
drome,” including obesity, hyperinsulinemia, and dyslip-
idemia, were not associated with stroke in our patients,
consistent with previous studies of type 2 diabetes.16 A re-
cent meta-analysis18 has indicated that hypertension, but
not elevated blood cholesterol levels, is related to the in-
cidence of stroke in the general population.
Recent intervention studies have provided appar-
ently conflicting data concerning the relationship be-
tween blood pressure control and the incidence of stroke
in patients with type 2 diabetes. An analysis of UKPDS
data on an intention-to-treat basis has revealed that hy-
pertensive patients allocated to tight blood pressure con-
trol (target, ,150/85 mm Hg; mean achieved blood pres-
sure during a median of 8.4 years of follow-up, 144/82
mm Hg) had a 44% reduction in the incidence of fatal
and nonfatal stroke compared with patients in a group
with less tight control (target, ,180/105 mm Hg; achieved,
154/87 mm Hg).29 Preliminary data from the Systolic Hy-
pertension in Europe Treated With Nitrendipine-based
Antihypertensive Therapy Trial show that the excess risk

ARCH INTERN MED/ VOL 159, MAY 24, 1999

1100

©1999 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023

 Table 2. Relationship of Potential Risk Factors to Stroke, Table 3. Estimated Hazard Ratios for Stroke in 3776 Patients
Adjusting for Age and Sex in 2704 Patients With Type 2 With Age, Sex, and Hypertension/Normotension Categorization
Diabetes Who Had Each Variable Measured*
Hazard Ratio (95% CI*)
Breakpoint Between Tertile Dependent Variable for Nonfatal or Fatal Stroke

Variable Lower/Middle Middle/Upper P Age, y

,50 1.00
Body mass index† 24.8 29.0 .10 50-54 1.37 (0.62-3.00)
Waist-hip ratio 0.87 0.94 .43 55-59 3.22 (1.68-6.19)
Systolic blood pressure,‡ 125 142 .01 $60 4.78 (2.56-8.92)
mm Hg Sex
Diastolic blood pressure, 79 87 .10 Female 1.00
mm Hg Male 1.63 (1.08-2.47)
Hypertension,§ no or yes ... ... .001 Hypertension
Fasting plasma glucose, 7.3 (103.3) 9.7 (173.1) .46 No 1.00
mmol/L (mg/dL)
Yes 2.47 (1.64-3.74)
Hemoglobin A1c, % 6.2 7.5 .88
Atrial fibrillation†
Serum total cholesterol, 4.88 (187.7) 5.77 (221.9) .22
No 1.00
mmol/L (mg/dL)
Yes 8.05 (3.52-18.44)
Serum LDL cholesterol, 3.02 (116.1) 3.89 (149.6) .20
mmol/L (mg/dL)
Serum HDL cholesterol, 0.95 (36.5) 1.15 (44.2) .82 *CI indicates confidence interval.
†From 3728 patients who had electrocardiography and in whom a
mmol/L (mg/dL)
multivariate hazard ratio model was fitted.
Serum triglyceride, mmol/L 1.22 (107.4) 1.87 (164.7) .87
(mg/dL)
Fasting plasma insulin, 9.7 15.6 .61
mU/L (µU/mL) The association between macroproteinuria and mac-
Exercise, sedentary vs ... ... .56 rovascular disease that includes stroke is well recog-
moderate vs active vs fit nized.34,35 Nevertheless, microalbuminuria was not a risk
Smoking, never vs ... ... .45
factor for stroke in the present study when included in a
ex-smoker vs current
multivariate model with hypertension. This has also been
*Data are given as tertiles to categorize data recorded as continuous reported for coronary artery disease in patients with type
variables. Sixty-four of the 2704 patients had a stroke during follow-up. Ellipses 2 diabetes36 and suggests that microalbuminuria is, in part,
indicate not categorized into tertiles. a marker for hypertension rather than an independent
†Calculated as described in the second footnote in Table 1.
‡Significant estimated hazard ratios are 1, 2.30, and 3.25 for the 3 tertiles of
risk factor for atherosclerosis. Although physical activ-
blood pressure, with the lowest standardized to 1. ity appears to have a beneficial effect on the incidence of
§Significant estimated hazard ratios are 1 and 2.27 (the hazard ratios for no stroke in middle-aged British men,37 we found no such
and yes, with no hypertension standardized to 1). association in the present study.
Patients in our cohort who had atrial fibrillation on
of stroke associated with diabetes was abolished by an- ECG at the time of diagnosis were more than 8 times as
tihypertensive treatment of older patients with type 2 dia- likely to suffer a stroke during the first 8 years of diabe-
betes mellitus and isolated systolic hypertension.31 In the tes than those who were in sinus rhythm. This finding
Hypertension Optimal Treatment (HOT) trial,32 how- suggests that atrial fibrillation is the most important risk
ever, blood pressure reduction in a subgroup with dia- factor, either modifiable or nonmodifiable, in patients hav-
betes was not associated with a reduced risk of stroke. ing newly diagnosed type 2 diabetes mellitus. The strength
There are several possible explanations for the ap- of association between atrial fibrillation and subsequent
parently discordant findings in the HOT study. First, fewer stroke has not been identified by previous studies and
diabetic patients were entered in the HOT study than in has significant implications for clinical management. Be-
the UKPDS, and the number of cardiovascular events was cause atrial fibrillation can be intermittent and because
less than the investigators anticipated,32 thus reducing it increases in prevalence with increasing age, more pa-
the power of the study to detect a beneficial effect of blood tients in the UKPDS than identified at the time of diag-
pressure reduction on stroke. Second, the average fol- nosis likely had atrial fibrillation during the follow-up.
low-up in the HOT study was only 3.8 years, and this Identification of these patients is clearly difficult with-
may have been inadequate to show a treatment-related out specific intensive ECG monitoring beyond that done
effect on the risk of stroke. The UKPDS data29 show that every 3 years under the UKPDS protocol. Nevertheless,
the Kaplan-Meier plots for cumulative stroke in the 2 treat- it is possible that, if this information were available, the
ment groups were coincident for the first 3 years, with association between atrial fibrillation and stroke would
increasing separation thereafter. Third, the UKPDS used be even stronger than that evident from cross-sectional
atenolol or captopril as primary therapy in the tight- data alone.
control group, whereas the HOT study used a calcium In some previous reports of risk factors for stroke in
channel blocker. The possible adverse effects of calcium diabetes, atrial fibrillation did not appear in logistic re-
channel blockers on cardiovascular outcomes33 may gression analyses. This includes longitudinal studies35,38
also be a factor underlying the differences in the risk of of approximately 1000 Finnish patients aged 45 to 64 years
stroke between hypertension intervention studies in with type 2 diabetes mellitus and a smaller-scale prospec-
type 2 diabetes. tive study39 of a cohort of 133 patients of similar age with

ARCH INTERN MED/ VOL 159, MAY 24, 1999

1101

©1999 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023

 newly diagnosed diabetes. Proteinuria, glycemic control,
dyslipidemia, and autonomic neuropathy were among the
independent risk factors for stroke identified in at least 1
of these 3 studies.35,38,39 The apparent differences be-
tween risk factor profiles in Finnish and UKPDS patients
might reflect relative sample sizes and differences in the
true underlying risk of stroke in the study populations11
because, even allowing for differences in age and dura-
tion of diabetes, the incidence of stroke appeared higher
in the Finnish patients (eg, 3.8% after 5 years and 14.7%
after 10 years in those who were newly diagnosed39) than
in our cohort (2.6% after a median of 8 years). The selec-
tion of variables used in regression analysis, including the
presence or absence of atrial fibrillation, may have been a
key contributing factor.
Where data concerning atrial fibrillation have been
available, only limited conclusions about their impor-
tance can be drawn. In a study40 of 428 unselected hos-
pital inpatients with cerebrovascular disease, 77 (18.0%)
patients had diabetes mellitus and were also signifi-
cantly more likely to have had a history of atrial fibril-
lation than the patients without diabetes. In an 8-year
population-based study21 of subjects aged 35 to 74 years,
15.1% of diabetic patients with stroke had previously
documented atrial fibrillation compared with 10.7% of
those without diabetes. Although these studies confirm
that atrial fibrillation increases the risk of stroke in pa-
tients with diabetes, our data suggest that it is the most
important risk factor in younger patients with newly di-
agnosed type 2 diabetes mellitus who are relatively free
of vascular disease.
Warfarin sodium therapy is associated with a two-
thirds reduction in the risk of stroke when given for atrial
fibrillation in the general population.41 Although there
are no equivalent data for patients with diabetes, it is pos-
sible that the benefits may be even greater. Neverthe-
less, only 8 (30%) of our patients with atrial fibrillation
on ECG at study entry were taking warfarin therapy at
some stage during the first year of follow-up. The un-
deruse of anticoagulant therapy for chronic atrial fibril-
lation is well recognized, with a similar proportion of pa-
tients with atrial fibrillation in the general population
(about a third) receiving warfarin.42,43 These patients have
an increased risk of hemorrhagic stroke of 3 such events
per 1000 patient-years of treatment.44 Because the risk
of cerebral hemorrhage in diabetic patients may be less
than that in persons without diabetes,2,3 warfarin therapy
is not likely to markedly increase the incidence of this
complication of anticoagulation therapy in type 2 dia-
betes mellitus.
CONCLUSIONS
Both aggressive treatment of hypertension and routine
anticoagulation therapy for atrial fibrillation may re-
duce the risk of stroke in patients with newly diagnosed
type 2 diabetes mellitus who are younger than 65 years.
Although data from published intervention trials29,45 in
diabetic patients with hypertension have inconsisten-
cies, the beneficial effects of antihypertensive treatment
on the incidence of cerebrovascular disease that have been
documented in subjects without diabetes appear even
ARCH INTERN MED/ VOL 159, MAY 24, 1999
1102
©1999 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023
greater for diabetic patients,31 including those who are
older than the patients in the present study, in whom the
risk of stroke is increased. Other strategies may include
aspirin therapy in any patient with atheromatous car-
diovascular disease and prophylactic anticoagulation in
those with cardiac failure or who have had a large ante-
rior myocardial infarction.21,46
Accepted for publication November 2, 1998.
This work has been supported in part by grants from
the United Kingdom Medical Research Council, London; Brit-
ish Diabetic Association, London; the United Kingdom De-
partment of Health, London; The National Eye Institute and
The National Institute of Digestive, Diabetes, and Kidney
Disease of the National Institutes of Health, Bethesda, Md;
The British Heart Foundation, London; The Health Promo-
tion Research Trust, London; Charles Wolfson Charitable
Trust, London; The Alan and Babette Sainsbury Trust,
London; The Clothworkers’ Foundation, London, and The
Oxford University Medical Research Fund Committee, Ox-
ford. It has also been supported by grants from pharmaceu-
tical companies, including Novo Nordisk A/S, Bagsværd, Den-
mark; Bayer plc, Newbury, England; Bristol-Myers Squibb
Ltd, Hounslow, England; Hoechst Marion Roussel, Bridge-
water, NJ; Eli Lilly and Co, Indianapolis, Ind; Lipha, Lyon,
France; and Farmitalia Carlo Erba, St Albans, England.
Additional assistance was provided by Boehringer Mann-
heim, Livingston, Scotland; Becton Dickinson, Oxford; Owen
Mumford, Woodstock, England; Securicor, Sutton, En-
gland; Kodak, Hemel Hempstead, England; and Cortecs Di-
agnostics, Deeside, England.
The cooperation of the patients and many National
Health Service staff and nonstaff at the centers is much
appreciated.
Reprints: Philip Bassett, MA, University of Oxford, Dia-
betes Research Laboratories, Radcliffe Infirmary, Wood-
stock Road, Oxford OX2 6HE, England.
REFERENCES
1. Kannel WB, McGee DL. Diabetes and cardiovascular disease: the Framingham
Study. JAMA. 1979;241:2035-2038.
2. Aronson SM. Intracranial vascular lesions in patients with diabetes mellitus.
J Neuropathol Exp Neurol. 1973;23:183-196.
3. Biller J, Love BB. Diabetes and stroke. Med Clin North Am. 1993;77:95-110.
4. Hadden WC, Harris MI. Prevalence of diagnosed diabetes, undiagnosed diabe-
tes, and impaired glucose tolerance in adults 20-74 years of age. Vital Health
Stat 11. 1987;No. 237:1-55.
5. Barker WH, Feldt KS, Felbel JH. Community surveillance of strokes in persons
under 70 years old. Am J Public Health. 1983;73:260-265.
6. Loug S, Melamed E, Cahane E, Carmen A. Hypertension and diabetes as risk fac-
tors in stroke patients. Stroke. 1973;4:751-759.
7. Kagan A, Popper JS, Rhoads GG. Factors related to stroke incidence in Hawaii in
Japanese men. Stroke. 1980;11:14-21.
8. Stallones R, Dyken M, Fang HCH. Epidemiology for stroke facilities planning. Stroke.
1972;3:360-371.
9. Roehmodt ME, Palumbo PJ, Whisnant JP, Elusback LR. Transient ischemic at-
tack and stroke in a community-based diabetic cohort. Mayo Clin Proc. 1983;
58:56-58.
10. Ostfeld AM, Shekelle RB, Kiawans H, Tufo HM. Epidemiology of stroke in an el-
derly welfare population. Am J Public Health. 1974;64:450-458.
11. Lindegard B, Hillbom M. Associations between brain infarction, diabetes and al-
coholism: observations from the Gothenburg population cohort study. Acta Neu-
rol Scand. 1987;75:195-200.
12. Kuller L, Anderson H, Peterson D. Nationwide cerebrovascular disease morbid-
ity survey. Stroke. 1970;1:86-99.
 13. Wolf PA, Kannel WB, Verter J. Current status of risk factors for stroke. Neurol
Clin. 1983;1:317-343.
14. Herman B, Leyten ACM, van Luijk JH, Frenken CW, Op de Coul AA, Schulte BP.
An evaluation of risk factors for stroke in a Dutch community. Stroke. 1982;13:
334-339.
15. Lackland DT, Moore MA. Hypertension-related mortality and morbidity in the South-
east. South Med J. 1997;90:191-198.
16. Bell DSH. Stroke in the diabetic patient. Diabetologia. 1994;17:213-219.
17. Sacco RL. Risk factors and outcomes for ischemic stroke. Neurology. 1995;45
(suppl 1):S10-S14.
18. Cholesterol, diastolic blood pressure, and stroke: 13 000 strokes in 450 000 people
in 45 prospective cohorts: prospective studies collaboration. Lancet. 1995;346:
1647-1653.
19. Abbott RD, Donahue RP, MacMahon SW, Reed DM, Yano K. Diabetes and the
risk factors of stroke: the Honolulu Heart Program. JAMA. 1987;257:949-952.
20. Palumbo PJ, Elveback LR, Whisnant JP. Neurological complications of diabetes
mellitus: transient ischemic attack, stroke, and peripheral neuropathy. Adv Neu-
rol. 1978;19:593-601.
21. Stegmayr B, Asplund K. Diabetes as a risk factor for stroke: a population per-
spective. Diabetologia. 1995;38:1061-1068.
22. Tunbridge WMG. Factors contributing to deaths of diabetics under fifty years of
age: on behalf of the Medical Services Study Group and British Diabetic Asso-
ciation. Lancet. 1981;2:569-572.
23. Webster P. The natural history of stroke in diabetic patients. Acta Med Scand.
1980;207:417-424.
24. UKPDS Group. UK Prospective Diabetes Study, VIII: study design, progress, and
performance. Diabetologia. 1991;34:877-890.
25. World Health Organization. International Classification of Diseases, Ninth Revi-
sion. Geneva, Switzerland: World Health Organization; 1977.
26. UKPDS Group. UK Prospective Diabetes Study, XI: biochemical risk factors in
type 2 diabetic patients at diagnosis compared with age-matched normal sub-
jects. Diabet Med. 1994;11:534-544.
27. Cox DR. Regression models and life-tables. J R Stat Soc Ser B. 1972;34:187-201.
28. Easton DF, Peto J, Babiker AG. Floating absolute risk: an alternative to relative
risk in survival and case-control analysis avoiding an arbitrary reference group.
Stat Med. 1991;10:1025-1035.
29. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of
macrovascular and microvascular complications in type 2 diabetes: UKPDS 38.
BMJ. 1998;317:703-713.
30. Hypertension in Diabetes Study Group. HDS 4: therapeutic requirements to main-
tain tight blood pressure control. Diabetologia. 1996;39:1554-1561.
31. Tuomilehto J, Rastenyte D, Lutgarde T, Staessen J. Reduction in mortality and
ARCH INTERN MED/ VOL 159, MAY 24, 1999
1103
©1999 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a University of Bradford User on 09/19/2023
cardiovascular events in older diabetic patients with isolated systolic hyperten-
sion in Europe treated with nitrendipine-based antihypertensive therapy (Syst-
Eur Trial) [abstract]. Diabetes. 1998;47:54A.
32. Hansson L, Zanchetti A, Carruthers SG, et al, for the HOT Study Group. Effects
of blood-pressure lowering and low-dose aspirin in patients with hypertension:
principal results of the Hypertension Optimal Treatment (HOT) randomised trial.
Lancet. 1998;351:1755-1762.
33. Cutler JA. Calcium-channel blockers for hypertension: uncertainty continues. N Engl
J Med. 1998;338:679-681.
34. Morrish NJ, Stevens LK, Fuller JH, Jarrett RJ, Keen H. Risk factors for macro-
vascular disease in diabetes mellitus: the London follow-up to the WHO Multi-
national Study of Vascular Disease in Diabetics. Diabetologia. 1991;34:590-
594.
35. Miettinen H, Haffner SM, Lehto S, Ronnemaa T, Pyörälä K, Laakso M. Protein-
uria predicts stroke and other atherosclerotic vascular disease events in non-
diabetic and non–insulin-dependent diabetic subjects. Stroke. 1996;27:2033-
2039.
36. Turner RC, Millns H, Neil HA, et al, for the UKPDS Group. Risk factors for coro-
nary artery disease in non–insulin dependent diabetes mellitus: United Kingdom
Prospective Diabetes Study (UKPDS: 23). BMJ. 1998;316:823-828.
37. Wannamethee G, Shaper AG. Physical activity and stroke in British middle aged
men. BMJ. 1992;304:597-601.
38. Lehto S, Ronnemaa T, Pyörälä K, Laakso M. Predictors of stroke in middle-aged
patients with non–insulin-dependent diabetes. Stroke. 1996;27:63-68.
39. Toyry JP, Niskanen LK, Lansimies EA, Partanen KPL, Uusitupa MIJ. Autonomic
neuropathy predicts the development of stroke in patients with non–insulin-
dependent diabetes mellitus. Stroke. 1996;27:1316-1318.
40. Lithner F, Asplund K, Eriksson S, Hagg E, Strand T, Wester PO. Clinical charac-
teristics in diabetic stroke patients. Diabetes Metab. 1988;14:15-19.
41. Morley J, Marinchak R, Rials SJ, Kowey P. Atrial fibrillation, anticoagulation, and
stroke. Am J Cardiol. 1996;77:38A-44A.
42. Munschauer FE, Priore RL, Hens M, Castilone A. Thromboembolism prophy-
laxis in chronic atrial fibrillation: practice patterns in community and tertiary-
care hospitals. Stroke. 1997;28:72-76.
43. Antani MR, Beyth RJ, Covinsky KE, et al. Failure to prescribe warfarin to patients
with nonrheumatic atrial fibrillation. J Gen Intern Med. 1996;11:713-720.
44. Carey PA, More R. Atrial fibrillation and its management. Br J Cardiol. 1997;4:
56-63.
45. Yudkin JS. Factors influencing threshold and choice of treatment for hyperten-
sion in NIDDM: cardiovascular factors. Diabetes Care. 1991;14(suppl 4):27-32.
46. Loh E, Sutton MS, Wun CC, et al. Ventricular dysfunction and the risk of stroke
after myocardial infarction. N Engl J Med. 1997;336:251-257.