Estimating the risk of type-2 diabetes using obese-years in a contemporary population of the Framingham Study

Background: We have recently demonstrated that an obese-years construct is a better predictor of the risk of diabetes than the severity of body weight alone. However, these risk estimates were derived from a population cohort study initiated in 1948 that might not apply to the current population. Objective: To validate an obese-years construct in estimating the risk of type-2 diabetes in a more contemporary cohort study. Design: A total of 5,132 participants of the Framingham Offspring Study, initiated in 1972, were followed up for 45 years. Body mass index (BMI) above 29 kg/m(2) was multiplied by the number of years lived with obesity at that BMI to define the number of obese-years. Time- dependent Cox regression was used to explore the association. Results: The risk of type-2 diabetes increased significantly with increase in obese-years. Adjusted hazard ratios increased by 6% (95% CI: 5-7%) per additional 10 points of obese-years. This ratio was observed to be similar in both men and women, but was 4% higher in current smokers than in never/ex-smokers. The Akaike Information Criterion confirmed that the Cox regression model with the obese-years construct was a stronger predictor of the risk of diabetes than a model including either BMI or the duration of obesity alone. Conclusions: In a contemporary cohort population, it was confirmed that the obese-years construct is strongly associated with an increased risk of type-2 diabetes. This suggests that both severity and the duration of obesity should be considered in future estimations of the burden of disease associated with obesity.