Objective: To investigate which anthropometric index is the best predictor of diabetes in relation to age.
Methods: In this longitudinal study 4479 non-diabetic men and women aged > or =20 years were followed for 3.6 years. Diabetes with its risk factors and obesity were defined according to the ADA and the WHO criteria, respectively. Logistic regression analysis was used to estimate the odds ratio (OR) of developing diabetes in model 1 including only the anthropometric measure and in model 2 adjusted for common diabetes risk factors and in model 3 adjusted for other anthropometric indices plus all the variables in model 2.
Results: A total of 166 new cases of type 2 diabetes were diagnosed. In subjects aged <60 years general obesity and high waist-to-hip ratio (WHR) predicted diabetes in all three models with OR of 2.4 and 2.6 in model 3, respectively, while high waist circumference (WC) lost it association with diabetes in the full model. In subjects aged > or =60 years, however, high WC was the only independent predictor of diabetes in model 3 with OR of 3.8 while high WHR and general obesity predicted diabetes in models 1 and 2, respectively.
Conclusion: General obesity and high WHR in Iranian subjects aged <60 years and high WC in older ones are the important predictors of type 2 diabetes.