Objective: To determine the value and the optimal cutoff point of waist circumference (WC) in screening diabetes mellitus (DM) and to provide evidence for DM prevention and identifying population at risk in mid-western rural areas of Shandong province.
Methods: A sample consisting 16 341 rural residents was selected and studied. All participants were physically examined on height, weight, WC and fasting plasma glucose (FPG). Oral glucose tolerance test (OGTT) was performed for subjects with FPG valued from 6.1 to 7.0 mmol/L. DM was defined according to the criteria set by WHO in 1999. Area under the curve (AUC), sensitivity, specificity and Youden index were computed based on the receiver operating characteristic (ROC) curve analysis. Optimal cutoff point was determined by the maximum of Youden index.
Results: The prevalence rates of DM for males and females increased along with the rise of WC (trend test chi2 = 72.01, 122.65, P < 0.01). It appeared significantly higher in those with WC > or = 85 cm in females and > or = 80 cm in males, with those WC < 85 cm for females and < 80 cm for males, in particular. AUCs were 0.639 and 0.655 for males and females respectively and both had significant differences (t = 7.22, 11.07, P < 0.01). However, the AUCs did not show significant difference (t = 0.70, P > 0.05) between males and females. The Youden index reached maximum when WC approached 85 cm for females (24.90%) and 80 cm for males (24.39%). The sensitivity and specificity were 58.04% and 66.86% for males, and 67.08% and 57.31% for females.
Conclusion: WC seemed to be an effective indicator for screening the DM. The optimal cutoff point of WC would be 85 cm for females and 80 cm for males in screening DM and defining the population at risk in this area.