Study Suggests C-Reactive Protein Doesn't Cause Heart Disease


(Getty Images)

By Anne Harding
WEDNESDAY, June 30, 2009 ( — High levels of C-reactive protein (CRP) in the blood probably don’t cause hardening of the arteries or heart disease, according to the largest study of its kind to focus on the long-suspected culprit.

People with heart disease have high levels of CRP in their blood because it’s released by cells during inflammation, the immune system’s attempt to defend against disease or infection. For this reason, blood tests for CRP levels have become a valuable tool in determining a person’s cardiovascular risk. But there’s a hot debate under way about whether CRP itself hurts the heart or whether it’s just an innocent bystander.

To look at the issue, Paul Elliott, FRCP, of Imperial College London, used a technique called Mendelian randomization, which is based on the arbitrary assignment of genes from parent to child—a bit like assigning someone either a placebo or a real drug.

Some people are genetically predisposed to having high levels of CRP, while others have genes that help keep their levels low. If the same genes associated with lower or higher levels of CRP were found to also influence heart disease risk, it would suggest that the protein does indeed help to cause heart disease.

But in the study, published this week in the Journal of the American Medical Association, Dr. Elliott and his team found the opposite. Their study isn’t the first to look at the CRP–heart disease question in this way, but it is the biggest.

“The power of the new study is its size,” says Mark Pepys, the head of medicine at the Royal Free Campus of the University of London Medical School. “It’s a very compelling null result which tells us that almost certainly CRP does not cause atherosclerosis and does not cause heart attacks.” The findings are consistent with animal studies, he adds, and “should help resolve the controversy.”

Another leading CRP researcher isn’t so sure. “A null Mendelian randomization may not mean very much,” Paul M. Ridker, MD, the director of the Center for Cardiovascular Disease Prevention at Brigham and Women’s Hospital, in Boston, says. “While it does not support causality, I don’t think most genetic statisticians believe it excludes a causal pathway either.”

Either way, the findings don’t mean that blood tests to measure CRP levels should be excluded when taking into account a person’s risk of heart disease or need for treatment, Dr. Ridker says. (The study didn’t consider this issue.)

“We should still measure it because, A, it predicts vascular risk even when cholesterol is low and other risk factors are absent, and, B, because we have clinical trial data demonstrating that if you have an increased level of [reactive protein], you will live longer and have fewer heart attacks and strokes if you take a statin," he says.

In the study, Dr. Elliott and his colleagues first looked at the genomes of 17,967 people to identify chromosome regions that were related to CRP levels. They checked their findings by genotyping an additional 13,165 people. Then they tested an additional 28,112 people with heart disease and 100,823 healthy subjects to see if any of the five gene variants most closely tied to CRP levels were linked to an increased risk of heart disease.

People who had the gene variants had 20% lower CRP levels. Based on observational studies of CRP levels and heart disease risk, this should have translated to a 6% reduction in heart disease risk—but it didn’t.

The current findings, as well as past research that used similar techniques, “strongly challenge” a causal role for CRP in heart disease, Svati H. Shah, MD, of Duke University Medical Center, and James A. de Lemos, MD, of the University of Southwestern Medical Center in Dallas, wrote in an editorial published with the new study.

But this doesn’t mean CRP tests should be tossed, according to Dr. Shah and Dr. de Lemos. “If CRP increases in response to other inflammatory triggers, it may still be a useful tool for personalizing selection of anti-inflammatory therapies, including statins,” they wrote.

Dr. Ridker agrees. “I have always felt that CRP is a good clinical biomarker of inflammation (and high vascular risk) but that it is inflammation that is likely to be causal for atherosclerosis, not CRP itself,” he says.

Dr. Pepys, however, worries that CRP might be too “nonspecific” to be a useful test for heart disease risk. “Anything that’s wrong with you will put up your CRP values, a little or a lot," he says.

He also points out that most variation in CRP levels is accounted for by well-known, modifiable risk factors for heart disease—such as being overweight and having type 2 diabetes. “Those things we know actually cause heart attacks," he explains.