LAUSR

Objective Body fat percentage (BF%) might be an alternative index of obesity which is the major risk factor for developing type 2 diabetes (T2D). We aim to longitudinally evaluated the relationship between BF% and risk of T2D. Methods A sample of 5,595 adults aged 18–65 who participated in two waves of China Health and Nutrition Survey (CHNS 2015 and 2018) was analyzed. Two level mixed-effects modified Poisson regression with robust estimation of variance stratified by sex was used to evaluate the risk ratios (RRs) for T2D according to quintiles of BF%, and the curves of receiver operating characteristic (ROC) were plotted to identify the optimal total and trunk BF% cut-off points for predicting an increased T2D risk. Results In males, compared with subjects in the first quintile of total BF%, those in the third (RR = 2.03, 95% CI 1.09–3.79), fourth (RR = 2.56, 95%CI 1.46–4.48), and fifth (RR = 2.16, 95%CI 1.22–3.82) quintile had higher risk of T2D after adjusting for all potential confounders (p-trend < 0.001). For females, the RR (95% CI) was 1.92 (1.14, 3.24) in the fifth quintile (p-trend = 0.014). Males and females with a trunk BF% >25.5 and 34.4% (≥ quintile 4), respectively, were at significantly increased risk of T2D (p-trend = 0.001). Besides, the optimal cut-off values of total and trunk BF% were 21.9 and 25.2% for males, and 36.7 and 30.3% for females, respectively. Conclusions The incident risk of T2D significantly increased over specific level of total and trunk BF% in both Chinese males and females, and the optimal BF% cut-off values were valuable for clinical application of BF% based on sex difference, which may be a cost-effective implementation for prevention and treatment of T2D in China.