Objectives To discuss the better predictors of metabolic syndrome in Uygur population of Xinjiang.

Methods Data were gathered from epidemiological survey of Xinjiang Hetian area in 2008 and 2 228 subjects were included. Waist-height ratio (WHtR), body mass index ( BMI), Waist-hip ratio ( WHpR) were calculated. Kappa test was applied to determine the concordance among different methods.

Results The results showed that according to International Diabetes Federation 2005 diagnostic criteria for metabolic syndrome, it was better to apply WHtR in predicting metabolic syndrome by the receiver operating characteristic (ROC) curve than BMI, abdominal circumference, and WHpR. When the cutoff of WHtR was 0. 53, the sensitivity and specificity in diagnosing MS were 91. 2% and 71. 8%, with its area under ROC curve 0.878. Kappa test showed an intensive concordance between WHtR and waist circumference (WC). The results showed that according to International Diabetes Federation 2005 diagnostic criteria for metabolic syndrome, it was better to apply WHtR in predicting metabolic syndrome by the receiver operating characteristic (ROC) curve than BMI, abdominal circumference, and WHpR. When the cutoff of WHtR was 0. 53, the sensitivity and specificity in diagnosing MS were 91. 2% and 71. 8%, with its area under ROC curve 0.878. Kappa test showed an intensive concordance between WHtR and waist circumference (WC).

Conclusions The specificity and sensitivity of diagnosing MS would be significantly raised by using both WHtR and WC.