Evaluation of newer risk markers for coronary heart disease: The Rotterdam study

Design: Prospective cohort study in the general population of Rotterdam, the Netherlands (The Rotterdam Study). Methods: Data on measurements taken between 1997 and 2001 in 5,933 persons free of CHD (40.6% men; mean age 69.1 years) were collected. We studied the predictive ability of 12 newer risk markers (N-terminal pro-brain natriuretic peptide [NT-proBNP] levels, von Willebrand factor antigen levels, fibrinogen levels, chronic kidney disease, leukocyte count, C-reactive protein levels, homocysteine levels, uric acid levels, coronary artery calcification [CAC] scores obtained by means of CT, carotid intima-media thickness, peripheral arterial disease, and aortic pulse wave velocity). The predictive value was determined by adding a newer marker to a prediction model that was based on traditional cardiovascular risk factors. Results: Risk discrimination improved the most with the addition of CAC scores. A net 23.5% of the individuals who developed CHD were reclassified to a higher risk category, but also 4.2% of those who did not develop CHD. This resulted in a net reclassification improvement (NRI) of 0.193. The CAC score was followed by NT-proBNP (NRI 0.076) in terms of the most improvement to risk classification. Improvements in risk predictions with the other newer markers were marginal. Conclusions: Classification of CHD risk predictions improved most with the addition of the CAC scores to the risk model. Further research is needed to assess whether refinements in risk prediction will actually lead to more effective prevention of cardiovascular disease together with justifiable costs and efforts.