To compare the suitability of metabolic syndrome definitions in predicting cardiovascular disease (CVD) risk.
We analyzed data from a population-based prospective cohort of 3598 participants from Jiangsu, China. Waist circumference was replaced with central obesity [index of central obesity (ICO), a ratio of waist circumference and height] in Cholesterol Education Program Adult Treatment Panel III (ICO-replaced ATPIII) and International Diabetes Federation (ICO-replaced IDF), respectively. Cox proportional-hazards regression model and the receiver operating characteristic curve (ROC curve) was used to evaluate the suitability of ATPIII, IDF, ICO-replaced IDF and ICO-replaced ATPIII in predicting CVD risk.
ICO was a better parameter in predicting CVD risk by ROC curve analysis. The ROC curve analysis also showed that although ICO-replaced IDF and IDF had the higher degree of specificity, lower sensitivity, longer ROC curve distance, less area under the curve to identify CVD than ATPIII and ICO-replaced ATPIII, therefore ICO-replaced IDF and IDF seemed to be undesirable. However, there was no significant difference in area under the curve between ATPIII and ICO-replaced ATPIII in predicting CVD risk. But it seems that odds ratios for abnormal triglyceride and high-density lipoprotein levels increase slightly when using ICO, but decrease for hyperglycemia and hypertension when using ICO.
ICO was a better predictor of abnormal triglyceride and high-density lipoprotein levels than waist circumference, but waist circumference was a better predictor of hyperglycemia and hypertension than ICO. However, we failed to support ICO as a better parameter for metabolic syndrome definition in predicting CVD risk compared with waist circumference.