Abstract
Previous epidemiologic studies debated the association of body mass index (BMI) trends with cardiovascular disease and mortality. This study aimed to evaluate the association of BMI variability and slope with the incidence of Type 2 diabetes mellitus (T2DM) in a sex-stratified 15.8-year follow-up in the population-based Tehran Lipid and Glucose Study (TLGS). Of 10,911 individuals aged 20-60 years, 4981 subjects were included and followed for 15.8-years. The slope coefficient of BMI in the linear regression model represented individuals' BMI trends up to the incidence of DM. The root mean squared error (RMSE) of the BMI linear trend was selected to reflect BMI variability through six follow-ups. Cox proportional hazards regression was used to investigate the association of the baseline BMI, BMI slope and RMSE with the incidence of T2DM among men and women. Multivariable-adjusted HRs of T2DM for each SD increment in BMI slope was 1.18 (95% CI: 0.94-1.48, p = 0.161) in normal weight men and 1.26 (95% CI: 1.10-1.44, p = 0.001) in overweight and obese men. However, in women, each SD increment in BMI slope increased the risk of T2DM with a HR of 1.19 (95% CI: 1.01-1.40, p = 0.039) in normal weight, and 1.14 (95% CI: 1.08-1.19, p < 0.001) in women with BMI ≥ 25 kg/m(2). In men with a baseline BMI ≥ 25 kg/m(2), BMI-RMSE was associated with a decreased risk of T2DM (HR: 0.71, 95% CI: 0.53-0.93, p = 0.015). Baseline BMI was not associated with the risk of diabetes in men and women. Positive BMI slope is associated with the development of diabetes in both sexes. The association of BMI variability with incident T2DM differs according to sex and baseline BMI. BMI variability is associated with a lower risk of T2DM in overweight and obese men. BMI variability in women and baseline BMI in both gender are not related to the risk of T2DM.