A 15.5-YEAR FOLLOW-UP-STUDY OF STROKE IN A JAPANESE PROVINCIAL CITY -THE SHIBATA STUDY

Background and Purpose Change toward Western lifestyles, particularly during the high economic growth period (approximately 1960 to 1975), d ynamically altered stroke frequency and the distribution of risk facto rs in the Japanese. We reexamined their association after this environ mental change by a cohort study.Methods The cohort (2302 subjects) com prised residents aged 40 years or older of the Akadani-Ijimino distric t in Shibata City, Niigata Prefecture, Japan, who were followed up fro m 1977 for 15.5 years. Results Crude incidence rates per 1000 person-y ears for all strokes were 5.22 for men and 4.36 for women (3.02 and 2. 18 for cerebral infarction, 0.65 and 1.06 for intracerebral hemorrhage , and 0.41 and 0.34 for subarachnoid hemorrhage, respectively). Multiv ariate analyses performed with the Cox proportional hazard model revea led these risk factors to be independently significant: for cerebral i nfarction in men, age, blood pressure, atrial fibrillation, albuminuri a, funduscopic abnormality, and current smoking; for cerebral infarcti on in women, age, atrial fibrillation, and history of ischemic heart d isease; for intracerebral hemorrhage in men, age and funduscopic abnor mality; for intracerebral hemorrhage in women, age, blood pressure, an d light physical activity; for all strokes in men, age, blood pressure , atrial fibrillation, albuminuria, funduscopic abnormality, current s moking, and heavy physical activity; and for all strokes in women, age , atrial fibrillation, and light physical activity. Conclusions Most t raditional risk factors, including blood pressure and its related orga n diseases, were confirmed, but serum total cholesterol had almost no effect. Physical activity had both negative and positive effects on st roke risk. In these findings, however, some differences related to sex were also observed.