Study after study has shown that high blood levels of C-reactive protein (CRP), a marker of inflammation, are associated with increased risk of cardiovascular disease. But what hasn’t been clear is whether CRP actually causes heart disease, or just indicates the presence of artery-blocking atherosclerosis.
Researchers from Imperial College London set out to answer this question using a genetic approach. They reasoned that if high CRP levels really do cause heart disease then genetic variants associated with increased CRP should also be associated with increased heart disease risk. Their results, published today in Journal of the American Medical Association, indicate that CRP might not be a key player after all.
Using a sample of 28,000 heart disease patients and 10,000 controls, Paul Elliott and colleagues examined several SNPs in the CRP gene. They found a significant association between the SNPs and CRP levels, and between CRP levels and heart disease – but not between the genetic variations and heart disease.
For example, each T at rs1205 lowered CRP by 20%. Based on previous studies of CRP, this should have translated into 6% lower odds of heart disease. But there was no association of rs1205 with heart disease at all.
The researchers also identified four SNPs in other genes that were associated with CRP levels. One was associated with heart disease in the expected direction (that is, the version associated with lower CRP was also associated with decreased heart disease risk). But two others had an inverse relationship with heart disease risk. The final SNP affected CRP levels, but had no relationship with heart disease risk either way.
(The SNPs studied, their effects on CRP and heart disease risk, and other notes are included in a table at the end of this post. 23andMe customers will be able to check their data for three of the five SNPS addressed in this study.)
The JUPITER trial, published in the New England Journal of Medicine in November, showed that middle aged men and women with low LDL cholesterol (below 130 mg/dL), but high CRP (greater than 2 mg/L) reduced their CRP levels by 37% and cut their risk of heart attack by 54% by taking rosuvastatin (Crestor ®).
These results suggest that CRP should be taken into account when deciding whether to prescribe statins to a patients, a move that would result in millions more Americans taking the drugs.
But the results of this new genetic study suggest that lower CRP doesn’t necessarily translate into a lower heart disease risk. These findings support skeptics of the JUPITER trial, who argued that the improvements in cardiovascular disease risk seen with rosuvastatin treatment could have been due solely to their reduction (50%) of the study participants’ already low cholesterol levels.
But the new study doesn’t necessarily mark end of the road for CRP testing. In an editorial accompanying the new report in JAMA, Svati Shah and James de Lemos argue that even if CRP does not cause heart disease, it might still be valuable as a marker for identifying people who are at risk but do not exhibit traditional indicators such as high cholesterol.
|SNP||Gene||Less Common Version||CRP effect/ Copy of Less Common Version||Odds of Heart Disease/ Copy of Less Common Version||Notes|
|rs6700896||LEPR||T||-14.8%||1.06||Also associated with increased weight and body mass index|
|rs1183910||HNF1A||T||-13.8%||No effect||Also associated with higher total cholesterol, LDL choesterol and HDL choesterol.|
|rs4420638||APOE-CI-CII cluster||G||-21.8%||1.16||Also associated with higher total cholesterol, LDL choelsterol and triglycerides, and lower HDL cholesterol|