But what if it’s all a big mistake?

For some scientists, the question will not go away. No study, these critics say, has ever proved a causal relationship between moderate drinking and lower risk of death — only that the two often go together. It may be that moderate drinking is just something healthy people tend to do, not something that makes people healthy.

“The moderate drinkers tend to do everything right — they exercise, don’t smoke, eat right and drink moderately,” said Kaye Middleton Fillmore, a retired sociologist from the University of California, San Francisco, who has criticised the research. “It’s very hard to disentangle all of that and that’s a real problem.”

Some researchers say they are haunted by the mistakes made in studies about hormone replacement therapy, which was widely prescribed for years on basis of observational studies similar to the kind done on alcohol. Questions have also been raised about the financial relationships that have sprung up between the alcoholic beverage industry and many academic centers, which have accepted industry money to pay for research, train students and promote their findings.

“The bottomline is there has not been a single study done on moderate alcohol consumption and mortality outcomes that is a “gold standard” kind of study,” said Dr Tim Naimi, an epidemiologist with the Centers for Disease Control and Prevention.

Even avid supporters of moderate drinking temper their recommendations with warnings about the dangers of alcohol, which has been tied to breast cancer, accidents, liver disease, cancers, heart damage and strokes.

“It is very difficult to form a single-bullet message because one size doesn’t fit all,” said Dr Arthur L Klatsky, a cardiologist in Oakland, California, who wrote a landmark study in the early 1970s finding that members of the Kaiser Permanente healthcare plan who drank in moderation were less likely to be hospitalised for heart attacks than abstainers.

“People who would not be able to stop at one to two drinks a day shouldn’t drink and people with liver disease shouldn’t drink,” Klatsky said. On the other hand, “the man in his 50s or 60s who has a heart attack and gives up his glass of wine at night — that person is better off being a moderate drinker.”

Health organisations have phrased their recommendations gingerly.

The American Heart Association says people should not start drinking to protect themselves from heart disease. The 2005 US dietary guidelines say “alcohol may have beneficial effects when consumed in moderation.”

The association was first made in the early 20th Century. In 1924, a Johns Hopkins biologist, Raymond Pearl, published a graph with a U-shaped curve, its tall strands on either side representing the higher death rates of heavy drinkers and nondrinkers; in the middle were moderate drinkers, with the lowest rates. Dozens of other observational studies have replicated the findings, particularly with respect to heart disease.

“With the exception of smoking and lung cancer, this is probably the most established association in the field of nutrition,” said Eric Rimm, an associate professor of epidemiology and nutrition at the Harvard School of Public Health. “There are probably at least 100 studies by now.”

Alcohol is believed to reduce coronary disease because it has been found to increase the “good” HDL cholesterol and have anti-clotting effects.

Other benefits have been suggested, too. A study in China found that cognitively impaired elderly patients who drank in moderation did not deteriorate as quickly as abstainers.

A report from the Framingham Offspring Study found that moderate drinkers had greater mineral density in their hipbones than non-drinkers.

Researchers have reported that light drinkers are less likely than abstainers to develop diabetes and that those with Type 2 diabetes who drink lightly are less likely to develop coronary heart disease.

But the studies comparing moderate drinkers with abstainers have come under fire in recent years. Critics ask: Who are these abstainers? Why do they avoid alcohol? Is there something that makes them more susceptible to heart disease? Some researchers suspect the abstainer group may include “sick quitters,” people who stopped drinking because they already had heart disease.

In 2006, shortly after Fillmore and her colleagues published a critical analysis saying a vast majority of the alcohol studies they reviewed were flawed, Dr R Curtis Ellison, a Boston University physician who has championed the benefits of alcohol, hosted a conference on the subject. There, scientists had reached a “consensus” that moderate drinking “has been shown to have predominantly beneficial effects on health.”

The meeting, like much of Ellison’s work, was partly financed by industry grants. And the summary was written by him and Marjana Martinic, a senior vice president for the International Center for Alcohol Policies, a group supported by the industry.

Two central questions, meanwhile, remain unresolved: whether abstainers and moderate drinkers are fundamentally different and, if so, whether it is those differences that make them live longer, rather than their alcohol consumption.

Dr Naimi of the CDC, who did a study looking at the characteristics of moderate drinkers and abstainers, says moderate drinkers are healthier, wealthier and more educated, and they get better healthcare, even though they are more likely to smoke. They are even more likely to have all of their teeth, a marker of well-being.

“Moderate drinkers tend to be socially advantaged in ways that have nothing to do with their drinking,” Naimi said. And simply advising the non-drinkers to drink won’t change that, he said.

Some scientists say the time has come to do a large, long-term randomised controlled clinical trial, like the ones for new drugs. But experts who believe in the health benefits of alcohol say this is an implausible idea. Large randomised trials are expensive and they might lack credibility. And there are practical and ethical problems in giving alcohol to abstainers without making them aware of it and without contributing to accidents.

“The last thing we want to do is expose people to something that might harm them,” said Sei Lee of the University of California, San Francisco, who recently proposed a large trial on alcohol and health.