Hematocrit as a risk factor of cerebral infarction: long-term prospective population survey in a Japanese rural community.

Abstract

To elucidate the relation of hematocrit (Hct) to the incidence of cerebral infarction, a prospective follow-up study of 16 years (1965-81) was performed in a general population sample of 1220 Hisayama residents aged 44 and over, of both sexes. Most of the subjects who died during the follow-up period were autopsied, the rate being 89.0%. Hct decreased with advancing age in men, but not in women. The average value for Hct was significantly lower in women than in men. According to the mean value +/- 1 SD of Hct, the subjects were grouped into 3, in each sex as follows: low (less than 35%), normal (35-45%) and high (greater than or equal to 45%) for men, and, low (less than 30%), normal (30-40%) and high (greater than or equal to 40%) for women. During the follow-up period, cerebral infarction occurred in 117 patients. The cumulative incidence of cerebral infarction in the low Hct group for men was the lowest, even after adjustments for age and blood pressure. Conversely, the incidence in the low Hct group of women was significantly higher than that in the normal Hct group and was consistently increased with time during 2-5 years of the follow-up. After the 6th year or later, however, the incidence was gradually but significantly increased in the high Hct group, compared with the normal Hct group. Since Hct levels were related with other variables such as serum total cholesterol, serum total protein, Quetelet index and prevalence of hypertension in both sexes, heavy alcohol consumption in men, and glucose intolerance in women, such variables were taken into account using Cox's proportional hazards regression model.(ABSTRACT TRUNCATED AT 250 WORDS)