Predicting heart attack risk

By Tom Fortner • Published: March 12th, 2007
Category: Health in a Heartbeat

In the world of medicine, just as in life, there’s always a desire to build a better mousetrap. But sometimes the old one works just as well.

Take the case of C-reactive protein, or C-R-P. A few years ago, doctors hoped the test for this biomarker for arterial inflammation would help them more accurately predict which individuals were likely to suffer a heart attack or stroke. Because half of all heart attacks occur in people whose total cholesterol readings are normal, they were searching for indicators that would give a clearer picture of risk.

But recent research by scientists at Harvard University has dampened hopes that C-R-P and other promising biomarkers add much to the picture. Indeed, these tests are little better than cholesterol at predicting a patient’s risk.

The research grew out of the famous Framingham Heart Study, in which physicians tracked the cardiovascular health of residents of a Boston suburb over several decades. Researchers tested levels of C-R-P and other biomarkers in frozen blood samples drawn from more than three-thousand subjects in the mid-1990s. They compared these readings with the patients’ subsequent health experiences, including major cardiovascular complications.

Turns out the predictive power of seven biomarkers tested was only a couple of percentage points better than a combination of traditional risk factors such as family history, smoking, obesity, lack of exercise, hypertension, diabetes or cholesterol.

Since the biomarker tests are relatively expensive, medical experts have begun to question their value. And the search for a better mousetrap goes on.