In an accompanying editorial, Dr Alan Tall (Columbia University Medical Center, New York, NY) points out that the study by Danesh et al provides by far the largest number of cases that have been examined in such analyses.2 Other strengths of the study include the fact that it was conducted in an Icelandic population with high rates of participation and follow-up, and the consistency in the relationship of traditional risk factors to CHD suggests that, in terms of cardiovascular risk, this is a typical European population, he adds.

He says that although these findings confirm pioneering work showing that the CRP level is an indicator of the risk of CHD, they call into question the magnitude of the effect. And he suggests that the strength of the relationship of CRP levels to the risk of CHD may have been overestimated in earlier reports.

Tall believes these new findings "call into question the clinical value of measuring CRP as a predictor of the risk of CHD and indicate that further research is needed to clarify the place of this approach in clinical medicine." He adds that: "Recent recommendations that CRP measurements be offered as an optional adjunct to a global assessment of risk factors in persons whose calculated 10-year risk of CHD is 10% to 20% may have to be revisited."

However, Ridker takes issue with these statements. He commented to heartwire that the current paper "confirms exactly what we have been saying for years, namely, that CRP is a strong, independent predictor of heart-disease risk even after adjustment for all other known coronary risk factors. In fact, in the new data, the predictive value of CRP is exactly the same as that of high blood pressure and almost as large as that of smoking. So unless one is going to argue that high blood pressure and smoking don't matter, then it is now completely clear that CRP is a major and important risk factor for heart disease. After all, this is the 22nd study of CRP and every single one has been positive."

In fact there is no controversy here at all, only an attempt . . . to create one.

Ridker adds: "In fact there is no controversy here at all, only an attempt by the [journal] authors to create one. A responsible reporting of the exact same data would not have generated anywhere near the media interest level." He notes that in the March 23, 2004 issue of Circulation there is a "lovely" article from Dr Wolfgang Koenig (University of Ulm, Germany) et al that confirms prior work in this field and argues that the time has come for CRP to be added to global risk prediction.3 "However, as that paper was published (appropriately) as a confirmatory study, it has not generated any media interest at all." He adds, "We ourselves have a paper in Circulation this week that shows CRP clearly adds to predictive models at all levels of the Framingham Risk Score. Thus, every study availableincluding the new data from Icelandshows that CRP adds predictive value over and above Framingham Risk covariates."

Ridker maintains that by using a cut-off value of 2.0 mg/L rather than the recommended value of 3.0 mg/L, Danesh et al have simply underestimated the utility of CRP. And he points out that Danesh's study shows an odds ratio for CRP in the first 10 years of 1.8, which is "exactly what we had said." But, he adds, "That would not be controversial, it would simply have confirmed everything we have argued for over a decade."

"We have an ongoing epidemic of heart disease, obesity, and diabetes and we are failing as a society to take needed steps to improve risk detection. When we have a simple, inexpensive, and proven technology like CRP, we should be willing to use it," Ridker concludes.

In response, Danesh says that as the largest such study ever reported, his results speak for themselves and suggest that CRP does not give much additional predictive value on top of established risk factors. "Our findings are clearly set out in detail in the paper, and I will leave it to the readers to decide for themselves if they are reported responsibly or not," he commented to heartwire.