As little as half a drink every other day is enough to reduce the risk, regardless of whether it is beer, red wine, white wine or liquor, the study indicates. Whether you drink it with your meal or at some other time also appears irrelevant.
"It was a surprise that — almost regardless of other factors associated with drinking — frequency of use seemed to be what reduced the subsequent risk of a heart attack," said Dr. Kenneth Mukamal of Harvard University Medical School, who led the study.
Those who drank at least three days a week had about one-third fewer heart attacks than did non-drinkers. And it made almost no difference whether the drinking consisted of half a drink or four. Those who imbibed only once or twice a week had only a 16 percent lower risk of a heart attack.
Some studies have indicated that alcohol raises the level of "good" cholesterol and also thins the blood, warding off the clots that cause heart attacks. But alcohol breaks down fairly rapidly in the body and its effects on red blood cells are short-lived, according to Mukamal.
Mukamal speculated that regular, moderate drinking is beneficial because it helps keep the blood thinned.
"We think it may be much like people who take aspirin every day or every other day. A little bit of alcohol on a regular basis helps keep the platelets from becoming sticky and prevents heart attacks," he said.
He noted that other studies have found that people with a gene that keeps alcohol longer in their system seem to benefit the most from moderate drinking. "That helps reinforce the notion that maintaining a low level of exposure is the way to go," he said.
Mukamal and other doctors emphasized that the study applies only to moderate drinkers. The dangers of heavy drinking are well-established and include alcoholism, drunken driving, and damage to the liver and brain. Studies have also found that women who have two or more drinks a day are 41 percent more likely to develop breast cancer than women who do not drink.
"I don't think any doctor would advise a patient to start drinking to prevent heart disease," said Dr. Gary Francis, director of the coronary intensive care unit at the Cleveland Clinic.
"We all, both the media and physicians who see patients, need to be careful that this isn't construed as a license to drink heavily. Certainly we don't want to exchange one public health problem for another."
Mukamal and every other doctor interviewed for comment emphasized that people should talk to their physicians about drinking.
Doctors have long believed that red wine explains the so-called French Paradox — the fact that the French have fewer heart attacks than Americans even though their food is richer. But the new study adds to the evidence that it is the alcohol itself, and not something found only in red wine, such as red pigment, that is good for the heart.
Mukamal analyzed data from a long-term study of 51,529 male health professionals — doctors, osteopaths, dentists, optometrists, podiatrists and veterinarians. He looked at 38,077 of them, eliminating people who had stopped drinking within the previous 10 years and those with histories of cancer and diseases of the heart or blood vessels.
Dr. Lynn Smaha, a cardiologist in Sayre, Pa., and a past president of the American Heart Association, said the study leaves some important questions unanswered. Among them: whether it would apply to a wider group than health professionals, who may take better care of themselves than the average drinker.
Smaha also noted that the overall rate of heart attacks — 1,418 cases out of 38,000 people — was very low, for reasons that were not clear.
He said he will continue to be very cautious in what he tells his patients about drinking.
And he worried about how some drinkers might interpret the study: "I can see the scene in a bar: `The paper says I should drink more. Where are my keys?'"