Abstract
BACKGROUND: The relation between body mass index (BMI) and mortality is not well established. The objective of this study was to examine the association in Japanese adults. METHODS: In 1990, 18,740 men and 20,870 women in Miyagi Prefecture in rural northern Japan (40-64 years of age) completed a self-administered questionnaire including height and weight. Cox regression was used to estimate relative risk (RR) of mortality according to levels of BMI, with adjustment for age, marital status, smoking, drinking, walking, and weight change since 20 years of age. RESULTS: During 11 years of follow-up, 1,121 men and 567 women had died. Compared with the referent BMI category (23.0-24.9), women in the highest BMI category (BMI>30.0) had a RR of death of 1.64 (95% confidence interval (CI), 1.09-2.49) and men and women in the lowest BMI categories (BM1<18.5) had a RR of death of 2.06 (95% CI, 1.49-2.84) and 1.83 (95% CI, 1.17-2.88), respectively, after adjustment for potential confounders and after exclusion of deaths occurring in the first three years of follow-up. We did not observe significant differences in mortality for subjects with wide range of BMI (18.5 or higher in men and 18.5 to 29.9 in women). CONCLUSIONS: The risk of death from all causes increases in lean men and women, and obese women in this cohort.