Risk factors for stroke in middle-aged men in Göteborg, Sweden.

To evaluate stroke risk factors in Göteborg, Sweden, during 1970-1973 a cohort of 7,495 participating men from a general population sample of 9,998 men aged 47-55 years were examined with respect to cardiovascular risk factors. Men with hypertension and hypercholesterolemia and men who were heavy smokers were treated. We assessed stroke end points and cause-specific mortality using a stroke register and death certificates. During a mean follow-up of 11.8 years, 230 strokes occurred in the entire population sample (participants and nonparticipants) (7% subarachnoid hemorrhages, 13% intracerebral hemorrhages, 42% cerebral infarctions, and 38% unspecified strokes). Using univariate analysis, we found measured high blood pressure (systolic and diastolic), smoking, known hypertension, diabetes mellitus, stroke in either parent, severe psychological stress, marital status, atrial fibrillation, previous transient ischemic attacks, previous myocardial infarction, effort-induced chest pain, and intermittent claudication to be significantly related to all stroke. Of the stroke types, subarachnoid hemorrhage was not related to any of these indicators, and intracerebral hemorrhage was related only to measured high blood pressure. Using multivariate analyses, we found measured high blood pressure, smoking, and severe psychological stress as well as atrial fibrillation, previous transient ischemic attacks, and intermittent claudication to be independent risk factors for nonhemorrhagic stroke. Serum cholesterol concentration, occupational and leisure-time physical activity, body mass index, alcohol abuse, and low occupational class were not risk factors for stroke.