The massive global hunt for genes for common diseases, such as cardiovascular diseases (CVD) and type 2 diabetes, has raised huge expectations. The idea that knowledge of disease genes can remarkably improve prediction of risk, and lead to better ways of prevention and targeted treatment has excited the world, and resulted in a pouring of resources into this field.
A major new study on prediction of CVD with genetic markers was reported earlier this week by Paynter and colleagues in the Journal of the American Medical Association (February 17, 2010).
A genetic risk score – comprising of 101 nucleotide polymorphisms reported to be associated with CVD – was evaluated in a prospective study of over 19,000 healthy white women followed for an average of 12 years in the Women’s Genome Health Study. The findings were striking. The 101 genetic markers together did not at all improve the prediction of CVD over what was possible with simple traditional risk factors (age, blood pressure, smoking history, diabetes, and cholesterol). In contrast, self-reported family history of CVD independently improved prediction of CVD when added to traditional risk factors. Paynter and colleagues tortured the data in every way they could, but the results remained convincing – 101 currently known genetic markers together simply do not help us predict CVD any better than a medical history, blood pressure examination, and a simple inexpensive blood test. A major international case control study called INTERHEART, done across 52 countries, showed that that nine known risk factors explain approximately 90% of the risk of acute myocardial infarction.
These results from Paynter’s investigation are reminiscent of another major study that evaluated genetic markers for type 2 diabetes. Meigs and colleagues studied 2400 participants of the Framingham Offspring Study, and asked whether a genetic score, composed of 18 nucleotide polymorphisms, would improve prediction of diabetes. Using traditional risk factors (age, family history, body mass index, fasting glucose, systolic blood pressure, high density lipoprotein cholesterol, and triglyceride levels), they were able to predict diabetes, over 28 years of follow up, with an accuracy of 90%. The addition of information on the 18 genetic markers improved the accuracy of prediction to 90.1%. Furthermore, several randomized controlled trials have indicated that the risk of type 2 diabetes can be reduced by 30-60% through lifestyle modification, and a number of studies have suggested that even when genetic risk is present, lifestyle modification can eliminate the excess risk conferred by genes.
But the gene hunters remain unconvinced. In an interview with heartwire, the lead author of the recent CVD genetic score study, Dr. Paynter sounded disappointed that 101 genetic markers together could not add to the prediction of CVD. But she was still hopeful, “However there are options for it to work in the future,” she told heartwire. “The first would be to find even more genes, maybe still with weak individual effects, and keep adding them together until they are strong enough in aggregate.”
One wonders whether the myth of Sisyphus may have some lessons for us here. Sisyphus, a figure in Greek mythology who was condemned to repeat forever the same task of pushing a boulder up a mountain, only to see it roll down again.
The pursuit of genes may, no doubt, improve our understanding of major diseases like CVD and diabetes, and may also lead to better treatments. Should we not, however, aim for a far better balance between spending an extraordinary amount of resources chasing the promise of genes and of implementing the huge body of knowledge that we already have about predicting and preventing common diseases, such as CVD and diabetes?
K.M. Venkat Narayan is Ruth and O.C. Hubert Professor of Global Health and Professor of Epidemiology and Medicine at Emory University Atlanta. He is a product of three continents, having lived and worked in India, United Arab Emirates, United Kingdom, and United States of America.
Read a Research Methods and Reporting article on a similar topic on bmj.com