Researchers have not conducted long-term, randomized trials of the effect of alcohol on heart health, as such a study would be expensive and hard to control. Instead, they generally rely on what participants say they drink.
So when past studies have linked light alcohol use with a lower risk of heart disease, that could be due to some factor other than alcohol, said Penn's Michael V. Holmes, the co-lead author of the study. People who drink one glass of wine a day might take overall better care of themselves than those who drink four, for example.
Instead, Holmes and his coauthors relied on a proxy for alcohol consumption: a genetic variant that causes alcohol to be metabolized much faster.
People with this gene variant tend to feel nauseated when they drink, and they drink less as a result. The idea was that this variant would be randomly distributed throughout the population, and not linked to other traits that affect heart health, such as smoking and exercise, Holmes said.
The results: over the long term, those with the genetic variant had a 10 percent lower risk of developing coronary heart disease. The findings came from an analysis of more than 50 studies with 260,000 participants.
"It doesn't matter where you are on the drinking spectrum," Holmes said. "If you reduce your alcohol intake, your risk of heart disease is lower."
Not so fast, said William Ghali, director of the Institute for Public Health at the University of Calgary.
Ghali oversaw a 2011 review of long-term observational studies and of short-term experimental studies in which scientists measured cholesterol and other biomarkers. The conclusion: a drink a day appeared to lower the risk of heart disease.
As for the new study, Ghali said it was a stretch for the authors to use the genetic variant as ground for advocating a reduction in drinking. He said it is possible that the variant is connected to yet some other gene or trait that explains the lower risk of heart disease found in the study.
Patrick O'Gara, president of the American College of Cardiology and a Harvard professor, shared that skepticism: "This is is at odds with most of what I have understood over the course of the past 10 to 15 years."
That said, O'Gara cautioned that alcohol can have other ill effects, citing another new study that linked moderate wine drinking to atrial fibrillation.