Bristol, UK - A new genetic study has cast new doubt on the idea that C-reactive protein (CRP) plays a causative role in metabolic syndrome and cardiovascular disease [1].

In a paper published in the December 3, 2005, issue of the Lancet, the authors, led by Nicholas Timpson (University of Bristol, UK), conclude: "Our findings add to the evidence that CRP is not causal in the pathology of the metabolic-syndrome components and is not a suitable target for direct therapeutic intervention."

They note that raised concentrations of circulating CRP are associated with an increased risk of the metabolic syndrome and cardiovascular disease and it has been suggested that this is a causal relationship, but recent studies have suggested that plasma CRP is not a pathological agent and that the association between raised CRP and heart disease is simply a result of confounding factors.

What is needed to settle the issue is a randomized trial of a drug that lowers CRP, but at present there is no drug that selectively lowers CRP without altering other factors also involved in heart disease. Timpson explained to heartwire that in the absence of drugs that directly influence the concentration of circulating CRP, genetic studies can be used to shed more light on CRP-disease associations. He described this approach, known as a Mendelian randomization, as "Nature's version of a randomized trial."

In this approach, genetic variants that contribute to circulating CRP levels are analyzed. Timpson noted: "As these genotypes are inherited randomly according to Mendelian inheritance, they are not subject to traditional observational confounders. People are just as likely to inherit one of these genotypes if they are from a lower or higher socioeconomic group or if they smoke or do not smoke, so it can be regarded as a parallel to a randomized trial of a CRP-altering effect."

In the current study, Timpson et al used data from 4286 women involved in the British Women's Heart and Health Study, conducted between 1999 and 2001, who had had blood samples taken to measure insulin resistance and levels of triglycerides, HDL cholesterol, and CRP. Blood pressure, height, and weight were also recorded. Timpson et al genotyped the blood samples for four common genetic variants of CRP.

They found that individuals with these genetic variants did produce higher levels of CRP but that this was not associated with increased likelihood of having components of the metabolic syndrome—higher blood pressure, raised triglycerides, low HDL, increased insulin resistance, and increased body-mass index.

Timpson commented to heartwire: "We saw an association between serum CRP and metabolic-syndrome parameters, as expected, and we also saw an association between genetic variants of CRP and serum CRP levels, but we did not see an association between genetic variants of CRP and metabolic-syndrome parameters."

The researchers conclude: "Our findings show that plasma CRP concentrations are associated with all major components of the metabolic syndrome, as previously reported. However, results of analyses that used the genetic determinants of plasma CRP as instruments to enable a nonconfounded and unbiased assessment of the role of CRP in the metabolic-syndrome phenotypes suggest that these associations are not causal."

They add: "If CRP were a causal factor in the metabolic syndrome, we would have noted similar associations when analyzing plasma CRP and metabolic-syndrome outcomes as when incorporating information from the nonconfounded, genetically determined variation in plasma CRP concentration." The researchers suggest therefore that circulating concentrations of CRP are merely marking the presence of other disease risk factors.

They conclude: "If CRP does not determine key components of the metabolic syndrome but is simply predictive of them because of reverse causality or confounding, the suitability of CRP as a potential therapeutic target for the prevention of metabolic disorders and related chronic disease becomes doubtful."

In an accompanying editorial, Drs Aroon Hingorani and Steve Humphries (University College London, UK), who are also working in this field, agree with the conclusions of Timpson et al [2]. "This analysis suggests that the directly observed associations of CRP with the metabolic syndrome are affected by residual confounding and/or reverse causation, leading to the conclusion that CRP is unlikely to be a causal factor in its development," they write.

They point out that Mendelian randomization could be used to study many other circulating biomarkers that may be involved in heart disease but for which specific drugs that lower their concentration do not exist. "If it were possible to distinguish probable causal from noncausal links using Mendelian randomization, this might help prioritize the development of new drugs for cardiovascular disease prevention. With the recent publication of databases of human DNA sequence variation, the availability of genetic tools for Mendelian randomization studies is increasing," they note.

In an interview with heartwire, Hingorani was upbeat about the potential for such studies, but he cautioned that the studies performed so far, including the current one by Timpson et al, were all too small for definitive answers. "These studies are by no means definitive at present, and this is a relatively new approach, but they do offer an exciting prospect of a less biased approach to assessing the relationship between various biomarkers and disease risk," he said.

He explained that as the genetic effect on CRP is actually quite small, large studies are needed for definitive answers. Hingorani's group has conducted a similar study, which was presented earlier this year at the European Atherosclerosis Society meeting and which showed no effect of genetic CRP variants on the risk of MI. And he noted that another study, published in Arteriosclerosis, Thrombosis, and Vascular Biology, has shown no association between one of the genotypes in CRP and raised blood pressure. "So there is a concordance emerging, but all three studies so far reported were all too small to be definitive," he commented. "To exclude a small but potentially important causal effect of CRP on MI risk with reliability would require a study of around 10 000 MI cases and 10 000 controls. This might be aided by pooling genetic studies that are addressing this question," he added.

Commenting on the current study for heartwire, Dr Eric Topol (Cleveland Clinic, OH), said he thought it was a "very interesting" way to perform an analysis using genotypic and phenotypic data as a proxy for a randomized trial. He pointed out a possible flaw—"that the CRP gene was not resequenced and the whole analysis is based on a rather limited look at three [single nucleotide polymorphisms] SNPs in the CRP gene," which he says raises concerns about incomplete ascertainment of the genomics of CRP. But he adds: "The conclusion may well be correct—that high CRP does not drive the metabolic syndrome—but this needs to be further substantiated, and it does not mean that measuring CRP in these patients lacks value."

Also commenting on the study for heartwire, Dr Christopher Cannon (Brigham and Women's Hospital, Boston, MA) said CRP is known to be a marker to identify increased risk: "CRP is great for that—no dispute," and it can also be used as a marker of successful therapy. On whether CRP is actually involved in the disease process itself, Cannon says research is ongoing and not conclusive either way. "But, even if it is not part of the disease itself (and I agree there is not much clinical evidence on this at all), it does not invalidate CRP as a marker in any way either for risk stratification or for monitoring therapies. "