Coronary disease risk prediction algorithm warranting incorporation of C-reactive protein in Turkish adults, manifesting sex difference

Abstract

Background and aim: An algorithm is needed for predicting coronary heart disease (CHD) risk in Turkish adults who have a high prevalence of metabolic syndrome (MetS). Methods and results: Ten-year risk of CHD was estimated in 2232 middle-aged adults free of CHD at baseline, followed over 7.6-years. Cox proportional hazard regression was used to predict CHD. Discrimination was assessed with area under receiver operating characteristics curve (AROC). CHD developed in 302 subjects. In multivariable analysis, high-density lipoprotein (HDL)-cholesterol levels were borderline predictive in men; smoking status and HDL-and low-density lipoprotein (LDL)-cholesterol levels were not predictive in women. Age, presence of diabetes, systolic blood pressure and C-reactive protein (CRP) were predictors in both sexes, while smoking status and LDL-cholesterol were so inmenonly. AROC of the model was 0.789 in men, and 0.806 in women (p<0.001 each). An algorithm using the stated seven variables was derived separately for each sex. After age adjustment, men and women in the highest quintile of risk score were significantly and 20-27-fold more likely to develop CHD than those in the lowest quintile. Conclusions: In a population with prevalent MetS, low-grade inflammation is independently relevant for CHD, as are serum lipoproteins and smoking status. The derived algorithm is effective in estimating CHD risk among Turkish adults. (C) 2010 Elsevier B. V. All rights reserved.