Multimarker Prediction of Coronary Heart Disease Risk: The Women's Health Initiative

Abstract

The aim of this study was to investigate whether multiple biomarkers contribute to improved coronary heart disease (CHD) risk prediction in post-menopausal women compared with assessment using traditional risk factors (TRFs) only. The utility of newer biomarkers remains uncertain when added to predictive models using only TRFs for CHD risk assessment. The Women's Health Initiative Hormone Trials enrolled 27,347 post-menopausal women ages 50 to 79 years. Associations of TRFs and 18 biomarkers were assessed in a nested case-control study including 321 patients with CHD and 743 controls. Four prediction equations for 5-year CHD risk were compared: 2 Framingham risk score covariate models; a TRF model including statin treatment, hormone treatment, and cardiovascular disease history as well as the Framingham risk score covariates; and an additional biomarker model that additionally included the 5 significantly associated markers of the 18 tested (interleukin-6, d-dimer, coagulation factor VIII, von Willebrand factor, and homocysteine). The TRF model showed an improved C-statistic (0.729 vs. 0.699, p = 0.001) and net reclassification improvement (6.42%) compared with the Framingham risk score model. The additional biomarker model showed additional improvement in the C-statistic (0.751 vs. 0.729, p = 0.001) and net reclassification improvement (6.45%) compared with the TRF model. Predicted CHD risks on a continuous scale showed high agreement between the TRF and additional biomarker models (Spearman's coefficient = 0.918). Among the 18 biomarkers measured, C-reactive protein level did not significantly improve CHD prediction either alone or in combination with other biomarkers. Moderate improvement in CHD risk prediction was found when an 18-biomarker panel was added to predictive models using TRFs in post-menopausal women.