Clinical Question: Is C-reactive protein a good predictor of coronary artery disease (CAD)?

Setting: Population-based

Study Design: Case-control

Synopsis: C-reactive protein is recommended increasingly as a screening test for CAD. This large study included healthy men (n = 8,888) and women (n = 9,681) born between 1907 and 1935 in Reykjavik, Iceland. Participants were recruited between 1967 and 1991 and were followed until 1995, at which point 2,459 persons had experienced a major coronary event. A group of 3,969 control subjects were selected from those who had not experienced an event, and were matched by sex, year of recruitment, and age.

Logistic regression was used to estimate the odds ratios (ORs) for the relationship between elevated levels of C-reactive protein and CAD, and adjusted for age, sex, CAD risk factors, and socioeconomic status. Although the participants did not specifically report it, the authors thought that the routine use of aspirin or statins was uncommon in this population during the study period.

According to the study results, C-reactive protein was not a very strong predictor of CAD, with an OR of 1.45, compared with an OR of 1.30 for erythrocyte sedimentation rate, 2.35 for total cholesterol levels, 1.87 for smoking, and 1.50 for elevated systolic blood pressure. An update of a meta-analysis of 22 previous prospective studies found similar and homogeneous results.

Bottom Line: C-reactive protein is a moderately accurate independent predictor of CAD. Its role as a screening tool has not been firmly established. (Level of Evidence: 3b)