Strong Association of Waist Circumference (WC), Body Mass Index (BMI), Waist-to-Height Ratio (WHtR), and Waist-to-Hip Ratio (WHR) with Diabetes: A Population-Based Cross-Sectional Study in Jilin Province, China

BACKGROUNDS: The prevalence of diabetes has increased with the increase of obesity, and finding indicators to predict diabetes risk has become an urgent need. The purpose of this study is to compare the correlation between four anthropometric indices and the prevalence of diabetes. METHODS: A total of 4052 participants aged 40 years and above were selected in Dehui City, Jilin Province, using a multistage stratified whole group sampling method. Face-to-face interviews and physical examinations were conducted. Multivariate logistic analysis was used. The values of BMI, waist circumference (WC), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR) were divided into quartiles (Q1: <25%; Q2: ~25%; Q3: ~50%; and Q4: ~75%). The median of each quartile was used for a linear trend test. RESULTS: For all four body fat-measuring indices of body mass index (adjusted OR: 3.300, 95% CI: 2.370, 4.595), WC (adjusted OR: 5.131, 95% CI: 3.433, 7.669), WHR (adjusted OR: 3.327, 95% CI: 2.386, 4.638), and WHtR (adjusted OR: 5.959, 95% CI: 3.922, 9.054), patients in the highest quartile were more likely to have diabetes than those in the lowest quartile. The areas under the curve of WHtR, WC, WHR, and BMI for diabetes were 0.683, 0.669, 0.654, and 0.629, respectively. In female participants, the areas under the curve of the waist-height ratio and WC were 0.710 (95% CI: 0.679-0.741) and 0.701 (95% CI: 0.670-0.732), respectively. CONCLUSIONS: The WC and WHtR were more closely related to diabetes than BMI and WHR among study participants ≥ 40 years of age, especially in females.