Do conventional risk factors predict subclinical coronary artery disease? Results from the prospective army coronary calcium project

Background Recent guidelines recommend against the routine use of coronary artery calcification (CAC) detection because the additive value over clinical prediction tools is uncertain. We compared CAC, with use of electron-beam computed tomography (EBCT), with clinical and serologic coronary risk factors for the identification of patients with increased coronary heart disease risk. Methods and Results We studied 630 active-duty CIS Army personnel (39-45 years old) without known coronary artery disease (CAD) who were undergoing a routine physical examination as required by regulations. Each participant underwent clinical and serologic risk factor screening and EBCT. The cohort (mean age 42 +/- 2 years, 82% male) had a low predicted risk of coronary events (mean 5-year Framingham risk index [FRI] 1.6% +/- 1.2%). The prevalence of coronary calcification was 17.6% (male 20.6%, female 4.3%). Significant univariate correlates of CAC were total and low-density lipoprotein [LDL] cholesterol, triglycerides, systolic blood pressure, and body mass index. However, only LDL cholesterol was independently associated with CAC. There was a significant but weak relationship between CAC and the Framingham risk index (FRI) (receiver-operator characteristic [ROC] curve area 0.62 +/- 0.03, P<.001), which was not different from the relationship between CAC and LDL cholesterol alone (ROC curve area 0.61 +/- 0.03, P<.001). The prevalence of any CAC in men increased slightly across increasing quartiles of FRI: 17.0%, 20.8%, 33.0%, and 29.2% (P=.033). Other risk factors (family history, homocysteine, insulin, lipoprotein[a], and fibrinogen) were not related to CAC. Conclusions In this age-homogeneous, low-risk screening cohort, conventional coronary risk factors significantly underestimated the presence of premature, subclinical calcified coronary atherosclerosis. These data support the potential of CAC detection as on anatomic, plaque-burden diagnostic test to identify patients who may require more intensive risk-reduction therapies, independent of predicted clinical risk.