Shiro Tanaka1, Sachiko Tanaka, Satoshi Iimuro, Yasuo Akanuma, Yasuo Ohashi, Nobuhiro Yamada, Atsushi Araki, Hideki Ito, Hirohito Sone. 1. Department of Pharmacoepidemiology (Sh.T., Sa.T.), Graduate School of Medicine and Public Health, Kyoto University, Yoshida-Konoe-cho, Sakyo-ku, Kyoto 606-8507, Japan; Department of Biostatistics (S.I., Y.O.), School of Public Health, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan; The Institute for Adult Diseases (Y.A.), Asahi Life Foundation, Tokyo 100-0005, Japan; Department of Internal Medicine (N.Y.), University of Tsukuba Institute of Clinical Medicine, Tsukuba, Ibaraki 305-8577, Japan; Tokyo Metropolitan Geriatric Hospital (A.A., H.I.), Tokyo 173-0015, Japan; and Department of Hematology, Endocrinology, and Metabolism (H.S.), Niigata University Faculty of Medicine, Chuo-ku, Niigata 951-8510, Japan.
Abstract
CONTEXT: Previous studies on the association between body mass index (BMI) and mortality in diabetes do not necessarily provide a comprehensive view in terms of the global population because of the exclusion of individuals with a BMI less than 18.5 kg/m(2). OBJECTIVE: The objective of the study was to examine the association between BMI and mortality. DESIGN, SETTING, AND PARTICIPANTS: We analyzed pooled data from 2 cohorts of 2620 Japanese patients with type 2 diabetes followed up for 6.3 years. Patients with a history of cardiovascular disease or cancer were excluded. MAIN OUTCOME MEASURE: The end point was all-cause mortality. Hazard ratios were estimated by Cox regression adjusted for age, smoking, leisure-time physical activity, and other confounders. RESULTS: An analysis using BMI categories of 14.4-18.5 (5.2%), 18.5-22.4 (37.3%), 22.5-24.9 (31.0%), and 25.0-37.5 kg/m(2) (26.6%) revealed no significant trend in mortality among patients with a BMI of 18.5 kg/m(2) or greater (trend P = .69). In contrast, the hazard ratio of patients with a BMI less than 18.5 kg/m(2) vs 22.5-24.9 kg/m(2) was 2.58 (95% confidence interval 1.38-4.84; P < .01). Exclusion of deaths in the first 4 years of follow-up decreased the hazard ratio only slightly. CONCLUSIONS: The lowest mortality rate was observed among patients with a BMI of 18.5-24.9 kg/m(2), and obesity had no benefits regarding mortality. Further research is needed in lean patients, especially among aging populations in East Asia.
CONTEXT: Previous studies on the association between body mass index (BMI) and mortality in diabetes do not necessarily provide a comprehensive view in terms of the global population because of the exclusion of individuals with a BMI less than 18.5 kg/m(2). OBJECTIVE: The objective of the study was to examine the association between BMI and mortality. DESIGN, SETTING, AND PARTICIPANTS: We analyzed pooled data from 2 cohorts of 2620 Japanese patients with type 2 diabetes followed up for 6.3 years. Patients with a history of cardiovascular disease or cancer were excluded. MAIN OUTCOME MEASURE: The end point was all-cause mortality. Hazard ratios were estimated by Cox regression adjusted for age, smoking, leisure-time physical activity, and other confounders. RESULTS: An analysis using BMI categories of 14.4-18.5 (5.2%), 18.5-22.4 (37.3%), 22.5-24.9 (31.0%), and 25.0-37.5 kg/m(2) (26.6%) revealed no significant trend in mortality among patients with a BMI of 18.5 kg/m(2) or greater (trend P = .69). In contrast, the hazard ratio of patients with a BMI less than 18.5 kg/m(2) vs 22.5-24.9 kg/m(2) was 2.58 (95% confidence interval 1.38-4.84; P < .01). Exclusion of deaths in the first 4 years of follow-up decreased the hazard ratio only slightly. CONCLUSIONS: The lowest mortality rate was observed among patients with a BMI of 18.5-24.9 kg/m(2), and obesity had no benefits regarding mortality. Further research is needed in lean patients, especially among aging populations in East Asia.