Ischemic stroke subtypes: a population-based study of incidence and risk factors.

BACKGROUND AND PURPOSE There is scant population-based information on incidence and risk factors for ischemic stroke subtypes. METHODS We identified all 454 residents of Rochester, Minn, with a first ischemic stroke between 1985 and 1989 from the Rochester Epidemiology Project medical records linkage system. We used Stroke Data Bank criteria to assign infarct subtypes after reviewing medical records and brain imaging. We adjusted average annual incidence rates by age and sex to the US 1990 population and compared the age-adjusted frequency of stroke risk factors across ischemic stroke subtypes. RESULTS Age- and sex-adjusted incidence rates (per 100 000 population) were as follows: large-vessel cervical or intracranial atherosclerosis with >50% stenosis, 27; cardioembolic, 40; lacuna, 25; uncertain cause, 52; other or uncommon cause, 4. Sex differences in incidence rates were detected only for atherosclerosis with stenosis (47 [95% CI, 34 to 61] for men; 12 [95% CI, 7 to 17] for women). There was no difference in prior transient ischemic attack and hypertension among subtypes, and diabetes was not more common among patients with lacunar infarction than other common subtypes. CONCLUSIONS The age-adjusted incidence rate of stroke due to stenosis of the large cervicocephalic vessels is nearly 4 times higher for men than for women. There is no association between preceding transient ischemic attack and stroke mechanism. Diabetes and hypertension are not more common among patients with lacunae. Age- and sex-adjusted incidence rates for ischemic stroke subtypes in this population can be compared with similarly determined rates from other populations.