5,944, 44.5% men). CRP data and conventional risk factors were collected at
baseline. During an average follow-up of 7.1 years, 308 CVD events (a composite of fatal
and nonfatal events incorporating acute myocardial infarction, coronary artery bypass
surgery, percutaneous coronary angioplasty, stroke, and heart failure) occurred. The log
CRP/SD predicted the risk of CVD events in middle-age (hazard ratio 2.20, 95% confidence
interval 1.34 to 3.61) and older (hazard ratio 1.85, 95% confidence interval 1.23 to 2.78)
participants, after adjustment for the Framingham risk score. Using receiver operating
characteristic (ROC) curves, the area under the curve for the Framingham risk factor
model for predicting CVD events was greater in the younger (ROC 0.78) and middle-age
(ROC 0.72) participants than in the older participants (ROC 0.59), although the discriminative
ability was not substantially improved by adding the CRP data. In conclusion, our
results have demonstrated the steadily decreasing predictive value of conventional risk
factors with advancing age, although CRP has limited additive value for CVD risk
stratification. Our results provide validation of the recently devised Framingham risk factor
algorithm for use in primary care in participants
