Limitations in the design and analysis of C-reactive protein (CRP) data presented by Wilson et al1 within the Framingham Heart Study need careful consideration before any conclusions regarding the clinical utility of this inflammatory biomarker are drawn.
The authors used an old assay for CRP that was previously deemed inaccurate for the evaluation of high-sensitivity CRP2 and as a consequence was abandoned by other researchers and was not validated by the Centers for Disease Control and Prevention.3 The Framingham analysis also includes the soft end points of angina pectoris, “coronary insufficiency,” intermittent claudication, and reported congestive heart failure, a surprising choice given prior work by the same authors emphasizing the importance of hard end points for risk prediction. The analysis also combines sexes despite evidence of statistical interaction, a choice presumably made owing to a lack of power to address the primary issues being raised. Last, the data contradict prior work from the same authors where CRP levels were found to be a highly effective method for stratifying risk within the Framingham population, at least for the hard end point of stroke.4
Ridker P, Rifai N, Koenig W, Blumenthal RS. C-Reactive Protein and Cardiovascular Risk in the Framingham Study. Arch Intern Med. 2006;166(12):1327-1328. doi:10.1001/archinte.166.12.1327-b