RESUMEN
Objective: Obesity is a known risk factor for diabetes, but the effect of weight changes on the incidence of diabetes is not yet determined. This study aims to evaluate the long-term effects of weight change [based on body mass index (BMI)] on the incidence of diabetes mellitus (DM) in a middle eastern population. Method: In the Isfahan Cohort Study (ICS) 6504 adults equal or greater than 35 years of age were recruited at 2001 and were followed until 2013. Absolute BMI changes (ΔBMI) were calculated by subtracting the baseline BMI from the BMI measured at follow-ups. To compare participants with different baseline BMI easier, relative changes in BMI were quantified as the percentage of changes from baseline. DM was assessed based on standard definitions. Multivariable Cox regression was used to determine the association between ΔBMI and the incidence of diabetes. Results: During follow-ups, 261 new cases of diabetes were recorded, with an IR of 3401.29 per 100,000 P-Y. The highest number of new cases of type 2 DM belongs to participants with overweight and obesity who had minimal BMI changes (less than 5% of their baseline BMI limits; 42 and 38 new cases, respectively). Participants who were obese at baseline and had lost more than 10% or gained 5-10% of baseline BMI were in the groups with the highest IR [360.05-95% CI (239.3-541.8) and 322.39-95% CI (178.5-582.1) respectively]. There was no significant association between BMI changes and the incidence of DM in the participants with normal BMI, overweight, and obesity at baseline in cure and adjusted models. Conclusions: This study showed there was no significant association between diabetes mellitus incidence and BMI changes.