RESUMEN
BACKGROUND: The link between adiposity and type 2 diabetes (T2D) is well known. However, it remains controversial as to which index and cutoff point is the best predictor in different populations. METHODS: A total of 2981 urban and rural Iranian adults over 35 years of age, and free of cardiovascular disease and diabetes were followed for 7 years. Anthropometric indices included body mass index (BMI), body adiposity index [BAI=(hip circumference/height¹·5)-18], waist-to-height ratio (WHtR), waist-to-hip ratio (WHpR), and waist and hip circumferences. T2D was defined as fasting plasma glucose ≥ 126 mg/dL or 2-h post-prandial plasma glucose ≥ 200 mg/dL, or the use of antidiabetic agents. Receiver operating characteristic curve analysis determined the best cutoff point for each adiposity index. RESULTS: After 7 years of follow-up, 389 new cases of diabetes were found. Most indices were linearly associated with increased risk of diabetes but the best continuous predictor was WHtR in men [odds ratio: 1.10 (95% confidence interval: 1.07-1.12) for one unit] and BMI in women [1.08 (1.04-1.11) for 0.1 kg/m²]. BMI cutoffs of 26 kg/m² in men and 30 kg/m² in women were the best binary predictors in adjusted models, and showed increased T2D risks of 2.91 (2.06-4.12) and 1.94 (1.42-2.66) times, respectively. All central-obesity indices in men and WHpR in women were also significantly associated with T2D independent of BMI. BAI was significantly associated with T2D in men but not in women. CONCLUSION: BMI at the appropriate cutoffs in both genders and WHtR in men and BMI in women as continuous factors were the best predictors of incident T2D in this Iranian population.