Lipid and inflammatory markers for the prediction of coronary artery disease: A multi-marker approach

Background: Many studies have investigated the clinical accuracy of single lipid and inflammatory markers. In contrast, few have evaluated their potential for the detection of CAD using a multi-marker approach. Methods: The concentrations of lipid, lipoproteins, apolipoproteins, high sensitivity C-reactive protein (hs-CRP) and fibrinogen were measured by standard laboratory methods. Apolipoprotein (a) [apo(a)] phenotyping was performed by sodium dodecylsulphate-gel electrophoresis and immunoblotting. The lipid tetrad index (LTI) and the lipid pentad index (LPI) were calculated. Clinical accuracy of the examined parameters, indexes and a logistic regression model was assessed using receiving operative characteristic (ROC) curve analysis. Results: Logistic regression analysis indicated that non-HDL-c, hs-CRP, HDL-c and Lp(a) were significant independent predictors for CAD. The AUC for this model (0.802) was higher than AUCs for any single marker or index tested. Conclusions: We conclude that the performance of a logistic regression model for CAD prediction warrants its use in clinical practice.