C-Reactive Protein Makes It to Center Stage

Abstract & Commentary

Synopsis: C-reactive protein levels are stronger predictors of first cardiovascular events than low-density lipoprotein cholesterol levels.

Source: Ridker PM, et al. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002;347:1557-1565.

Using a portion of the database of the Women’s Health Study (WHS), Ridker and colleagues previously demonstrated that among a dozen markers of inflammation, C-reactive protein (CRP) was best at identifying women at risk for cardiovascular events.1 This study finds them tackling the entire database to compare the efficacy of CRP and low-density lipoprotein cholesterol (LDL) as predictors of first cardiovascular events and to define population-based cutoff points for CRP. WHS is an ongoing study of aspirin and vitamin E as primary prevention of cardiovascular events in women 45 years or older (average age 54.7 years at study entry). Twenty-five percent were hypertensive, 12% were current smokers, and 2.5% were diabetic. The average body mass index was 25.9. Blood samples from 27,939 participants were analyzed for CRP and LDL levels. Approximately 400 other samples could not be analyzed. The patients were followed for these end points: nonfatal myocardial infarction, nonfatal ischemic stroke, coronary revascularization, and death from cardiovascular events. Knowing that hormone replacement therapy (HRT) can affect CRP and LDL levels, the total population was divided into those taking (12,139) and those not taking (15,745) HRT at the beginning of the study. The HRT status of 55 women was not known. The CRP and LDL data from both HRT and non-HRT were grouped into quintiles to establish cutoff points and subjected to Kaplan-Meier analysis to construct event-free survival curves.

The relative risks (RR) for first cardiovascular event (after adjusting for age, smoking status, diabetes status, blood pressure, and HRT status) were calculated for each quintile for CRP and LDL, using the first quintile as the reference. The results for CRP were 1.0, 1.4, 1.6, 2.0, and 2.3 by increasing quintile. The results for LDL were 1.0, 0.9, 1.1, 1.3, and 1.5. These values were all statistically significant. The data were also analyzed for each end point separately, and CRP continued to be more predictive than LDL. HRT reduced the RR as estimated by CRP but not by LDL. Seventy-seven percent of events occurred in women with LDL < 160 mg/dL and 44% in women with LDL < 130 mg/dL. LDL and CRP levels did not correlate well with each other. To explore the interactions between CRP and LDL, the women were divided into 4 groups, and the relative risks were calculated with the low CRP/low LDL as the reference, using the mean values of CRP (1.52 mg/L) and LDL (123.7 mg/dL). The RRs were 1.0 (low CRP/low LDL), 1.5 (low CRP/high LDL), 1.5 (high CRP/low LDL), and 2.1 (high CRP/high LDL). The age-adjusted event rates per 1000 person-years were 1.3, 2.0, 2.6, and 3.9.

Further analysis of CRP levels and the Framingham risk score showed that CRP levels are independent predictors of risk.

Comment by Allan J. Wilke, MD

CRP is the character actor of laboratory tests; it performs very well in bit parts and supporting roles but is still waiting for its shot at the big time. The generally recognized normal value is < 8 mg/dL,2 which is considerably higher than the mean value in this study. CRP is useful for monitoring the course of inflammatory disorders, detecting and monitoring infection, and staging chronic lymphoid leukemia. It is elevated with tissue injury or necrosis. Is cardiovascular event prediction its starring role? I’m not sure, but I don’t think so. As the accompanying editorial notes,3 there are hundreds of known risk factors for coronary heart disease. Which ones alone are the most predictive? Which ones in combination?

What have we learned from this study? First, CRP levels are stronger predictors of cardiovascular events than LDL levels. Second, CRP and LDL levels are not well correlated and probably predict risk in different populations. Third, the combination of CRP and LDL is better than either test by itself. It is well to remember, however, that the RRs are not great, at best 2-3 times the lowest quintile.

Are the results of this study plausible? Yes. More and more, we are learning that inflammation is central to the development of cardiovascular disease.4,5 Can we treat an elevated CRP level? Maybe. There are some intriguing hints from a previous study by the same authors that showed that lovastatin might prevent coronary events in patients with a total-to-HDL-cholesterol ratio that was lower than the median and a C-reactive protein level higher than the median.6

A weakness of this study is that it is derived from data from the WHS; it may not be wise to extrapolate the results to men. More importantly, WHS was designed to evaluate aspirin and vitamin E as primary prevention for cardiovascular events in women 45 years of age or older, not to look at CRP and LDL as cardiovascular risk factors. Therefore, the possibility of bias exists. I think that the sheer magnitude of the database makes bias unlikely, however.

The main question is whether we add CRP to our list of screening labs. The same editorialist suggests we should not because the "predictive value is markedly diminished when adjusted for other risk factors." CRP will have to wait for its close-up. In the meantime, we need to work to identify our patients at risk for cardiovascular disease and treat or modify known risk factors that improve those risks (hyperlipidemia, hypertension, tobacco abuse, etc).


1. Ridker PM, et al. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000;342:836-843.

2. Interpretation of Diagnostic Tests. 6th ed. 1996. 73-74.

3. Mosca L. C-reactive protein—To screen or not to screen? N Engl J Med. 2002;347:1615-1617.

4. Kalayoglu MV, et al. Chlamydia pneumoniae as an emerging risk factor in cardiovascular disease. JAMA. 2002;288:2724-2731.

5. Danesh J, et al. Low-grade inflammation and coronary heart disease: Prospective study and updated meta-analyses. BMJ. 2000;321:199-204.

6. Ridker PM, et al. Measurement of C-reactive protein for the targeting of statin therapy in the primary prevention of acute coronary events. N Engl J Med. 2001;344:1959-1965.

Dr. Wilke is Assistant Professor of Family Medicine at the Medical College of Ohio in Toledo.