Duration and degree of weight change and risk of incident diabetes: Japan Epidemiology Collaboration on Occupational Health Study.

We prospectively examined diabetes risk in association with a summary measure of degree and duration of weight change. The study participants were 51,777 employees from multiple companies in Japan, who were aged 30-59years, free of diabetes at baseline, and followed up for 7years (2008-2015). Exposure was cumulative body mass index (BMI)-years, which was defined as the area of BMI units above or below baseline BMI during follow-up, and was treated as a time-dependent variable in the Cox proportional hazards regression models. During the 263,539 person-years of follow-up, 3465 participants developed diabetes. The adjusted hazard ratio (HR) of diabetes for a 1-unit increase in cumulative BMI-years was 1.11 (95% confidence interval (CI): 1.09, 1.12). The association was more pronounced among overweight (HR=1.11; 95% CI: 1.08, 1.14) and obese (HR=1.12; 95% CI: 1.08, 1.15) adults compared with normal- and under-weight (HR=1.07; 95% CI: 1.03, 1.11) adults (P for interaction of cumulative BMI-years X baseline BMI-group=0.002). The association of higher cumulative BMI-years with incident diabetes did not substantially differ by metabolic phenotype. The present results emphasize the importance of avoiding additional weight gain over an extended period of time for the prevention of type 2 diabetes, especially among overweight and obese adults, irrespective of metabolic health status.