< Back to Menu

C REACTIVE PROTEIN

A simple new test for heart disease and strokes

(March 2001; updated June 2002)

The best predictors of increased risk of heart attack have been an elevated blood cholesterol level, elevated blood pressure, and cigarette smoking. Now, Paul M. Ridker, MD and colleagues from Harvard Medical School present evidence that a simple blood test measuring inflammation called C reactive protein is as good a predictor as blood cholesterol levels. They obtained blood specimens from 28,263 women and then compared 122 women who developed coronary heart disease or stroke with 244 controls with no such disease during a three-year follow-up period.

Subjects were divided into four groups (quartiles) according to their baseline levels of C reactive protein. Those women in the top quartile of C reactive protein levels (highest levels) had a four-fold increased risk of developing coronary heart diseases or stroke. But, even a modest increase in C reactive protein levels above the lowest group doubled the risk of coronary heart disease or stroke. At baseline, those destined to have either coronary heart disease or stroke during the next three years had about a 50 percent higher C reactive protein level than found in those who did not.

The authors then suggest that women (or men) with high levels of C reactive protein in their blood are candidates for cholesterol-lowering drugs that, according to a study from this same group of investigators, lower both cholesterol and C reactive protein blood levels.

Commentary: this is an interesting study that is supported by several other studies, but there are some substantial problems and issues. First, they combined coronary heart disease and stroke, but never gave the breakdown of how many of each. The risk factors for the two diseases are somewhat different; we should know if the potential usefulness of C reactive protein is the same for both diseases.

Second, the issue of false positive C reactive protein is not even discussed. The most critical question is what percentage of normal people who never will develop coronary heart disease or stroke (at least not for the next decade or two) will have at least somewhat elevated C reactive protein levels. It is very possible that there will be a very significant number of people who will be incorrectly labeled as having increased risk of coronary heart disease and stroke, especially because any recent infection can raise C reactive protein levels. And that mislabeling could have very serious consequences for them - unnecessary medications, etc.

Third, the authors do not suggest a cutoff level of C reactive protein that might justify administering a cholesterol-lowering drug (even if the cholesterol level is normal!). The data they present could be interpreted as suggesting that 75 percent of middle-aged and older adults should be given a cholesterol-lowering drug for life! (Because the top 75 percent of C reactive protein levels had a significant increase in risk of coronary heart disease or stroke compared to the lowest 25 percent.)

C reactive protein measures tissue inflamation. There is increasing interest in the role of three infections (chlamydia, herpes simplex, and cytomegalovirus) as having a role in coronary heart disease. In this study, none of these was related to C reactive protein levels. That finding in no way diminishes this study, but it does suggest caution.

This requires a lot more study before any recommendations can be made. We must know the false positive rates and establish definitions of what C reactive protein level constitutes an elevation that should result in some action. It could well be that elevated C reactive protein levels on two successive annual prevention examinations, in the absence of obvious infection, will result in a special focus on doing something about that individual's risk factors for coronary heart disease and stroke. But, that is for the future when more studies are available. Besides, we should all be paying attention to and doing something about risk factors for coronary heart disease and stroke - regardless of C reactive protein levels.

A thoughtful editorial by Dr Irving Kushner, a long time expert in this area, appeared in the Archives of Internal Medicine of April 22, 2002. He and his coauthor point out that we do not know the actual extent of risk of heart attack in those with moderate elevation of CRP, and we do not know how often a positive test is a false positive. CRP values can fluctuate from day to day, and values can increase with aging, alcohol use, and depression, among other causes. They conclude that we do not have the evidence to recommend routine use of CRP to predict heart attack risk.

Kushner, I. And Sehgal, A.R. Is high sensitivity C reactive protein an effective screening test for cardiovascular risk. Archives of Internal Medicine. Vol 162 (April 22) Pgs 867-869. 2002.

Ridker, P.M., et al. C reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. New England Journal of Medicine. Vol 342. (March 23) Pgs 836-843. 2000.

Supported by