Use of body mass index to identify obesity-related metabolic disorders in the Chinese population

Abstract

To identify the body mass index (BMI; in kg/m2) cutoff that predicts the risk for obesity-related metabolic disorders for the Chinese population. Community-based cross-sectional survey. Rural regions of Jiangxi and Anhui provinces and an urban community of Jing'an District of Shanghai, China. Five hundred and twenty-nine non-pregnant, non-lactating urban and rural adults, aged 20-64 years without diagnosed diabetes. Subjects were divided into two groups: with or without obesity-related metabolic disorders, which was defined as having at least one of the following: hypertension, insulin resistance, high plasma triacylglycerol, low-density lipoprotein-cholesterol or glucose. Gender-specific multiple logistic regression analysis demonstrated a significant dose-response relationship between BMI and obesity-related metabolic disorders, after adjusting for potential confounders. The lowest BMI interval associated with significant risk for both men and women (odds ratios of 2.67 and 3.46, respectively) was that of 22.5-24.4. Receiver-operating characteristic (ROC) curve analysis indicated that a BMI cutoff of 23 had the best combination of sensitivity and specificity and the shortest distance in the ROC curve, with positive and negative predictive values of 0.6-0.7 in both genders. A BMI cutoff of 23 might be appropriate for use in identification of high risk of obesity-related metabolic disorders and serve as a public health action threshold in the Chinese population. Center of a Livable Future, John Hopkins Bloomberg School of Public Health.