Background: Coronary artery calcium (CAC) has been suggested to improve prediction of coronary
events beyond traditional risk factor assessment. We determined whether CAC scoring
improves prediction of coronary events when added to traditional risk factor analysis
in the general population. Methods: We used data of the Heinz Nixdorf Recall cohort study, including 4129 randomly selected
subjects aged 59.4±7.7 years (53% women) without overt coronary artery disease at
baseline (2000 to 2003). Traditional risk factors and CAC, measured by electron-beam
computed tomography, were assessed. Participants were stratified into low, intermediate
or high risk categories using the Framingham risk score (FRS), and modified ATPIII
risk score categories (modATPIII). Participants were followed for hard coronary events,
i.e. coronary death and non-fatal myocardial infarction for 5.02 years. Results: The cumulative event rate was 2.3% (95% confidence intervals (CI): 1.8–2.8). Relative
risks (RR) increased from the lowest to the intermediate and highest FRS and modATPIII
risk categories. Crude and modATPIII-adjusted RRs in the highest versus lowest CAC
quartile were 6.40 (95% CI: 3.37–12.16) and 4.25 (95% CI: 2.14–8.45), respectively.
Adding CAC scores to the FRS and modATPIII-categories significantly improved the areas
under the ROC curves from 0.681 and 0.667 to 0.749 (p=0.003) and 0.753 (p=0.0001),
respectively. Intermediate modATPIII risk subjects with CAC<100 had an event rate
similar to low risk subjects, i.e. 1.5 (95% CI: 0.7–2.6) versus 1.0 (95% CI: 0.7–1.6)
(p=0.33). Event rates in intermediate risk subjects with CAC ≥400 were similar to
high risk subjects, i.e. 8.9 (95% CI: 5.1–14.3) versus 4.7 (95% CI: 3.3–6.4). Conclusion: CAC scoring improves risk stratification in the general population above and beyond
traditional risk factors. Adding CAC scoring to the FRS results in a reclassification
improvement of coronary risk, especially when focused on intermediate risk individuals.
