NON-INSULIN-DEPENDENT DIABETES AND ITS METABOLIC CONTROL ARE IMPORTANT PREDICTORS OF STROKE IN ELDERLY SUBJECTS

Background and Purpose Non-insulin-dependent diabetes mellitus (NIDDM) is a major risk factor for stroke in the middle-aged population, but few prospective population-based studies are available in the elderly. Moreover, the importance of metabolic control and the duration of dia betes in diabetic subjects has remained controversial. There are no pr evious studies on association of insulin with the risk of stroke. The present study examined whether NIDDM, its metabolic control and durati on, and insulin level predict stroke. Methods We measured cardiovascul ar risk factors including glucose tolerance, plasma insulin, and glyco sylated hemoglobin A(1c) in a Finnish cohort of 1298 subjects aged 65 to 74 years and investigated the impact of these risk factors on the i ncidence of both fatal and nonfatal stroke during 3.5 years of follow- up. Results Of 1298 subjects participating in the baseline study, 1069 did not have diabetes and 229 had NIDDM. During the 3.5-year follow-u p, 3.4% (n=36) of nondiabetic subjects and 6.1% (n=14) of NIDDM subjec ts had a nonfatal or fatal stroke. The incidence of stroke was signifi cantly higher in diabetic women compared with nondiabetic women (odds ratio [OR], 2.25; 95% confidence interval [CI], 1.65 to 3.06). In cont rast, the risk of stroke was not significantly higher in diabetic men than in nondiabetic men (OR, 1.36; 95% CI, 0.44 to 4.18). In multivari ate logistic regression analyses including all study subjects, fasting and 2-hour glucose (P<.01 and P<.05, respectively), glycosylated hemo globin A(1c) (P<.01), atrial fibrillation (P<.05), hypertension (P<.05 ), and previous stroke (P<.01) predicted stroke events. In diabetic su bjects, fasting and 2-hour glucose (P<.01 and P<.05, respectively), gl ycosylated hemoglobin A(1c) (P<.05), the duration of diabetes (P<.05), and atrial fibrillation (P<.05) were the baseline variables predictin g stroke events. Finally, fasting insulin (P<.05), hypertension (P<.05 ), and previous stroke (P<.01) were associated with stroke incidence i n nondiabetic subjects. Conclusions Our 3.5-year follow-up study provi des evidence that NIDDM, its metabolic control, and the duration of di abetes are important predictors of stroke in elderly subjects, particu larly in women. Moreover, fasting insulin level appears to be a risk f actor for stroke in elderly nondiabetic subjects.