Humans have been drawn to the idea that alcohol may have health benefits for almost as long as they have been drinking it. In ancient China, rice wine was widely used for medicinal purposes, while Hippocrates, the ancient Greek “Father of Medicine”, advocated moderate amounts of alcohol for the mind, body and soul.

Later, proponents of the temperance movement, who urged 19th century workers to quit booze, were met with resistance by those who thought beer was necessary for good health.

Surprisingly, the theory has since received plenty of support from modern science. When looking at how the amount you drink relates to your risk of cardiovascular disease and death, studies have come up with a puzzling but consistent “J-shaped curve”, suggesting that drinking a small amount of alcohol is healthier than abstaining completely.

However, the idea has been contentious from the start and now, as research techniques become more sophisticated, a different picture is emerging; one that regular drinkers may not be happy to hear.

The first evidence for alcohol’s salutary effect came from a small 1974 study of 474 people, which found that people who drank small amounts had a lower risk of heart attack than both heavy drinkers and teetotallers. Over the next few decades, as scientists were able to interview larger numbers of people, crunch more data and account for more factors, studies repeatedly came to the same conclusion.

As recently as 2011, a meta-analysis published in the BMJ concluded that the lowest risk of coronary heart disease could be achieved with a rather jolly 1-2 drinks a day. And, two years ago, the extensive global burden of disease study suggested alcohol may benefit people older than 40, contradicting their own 2018 finding that any amount of drink is bad.

Accompanying these studies, though, has been a steady chorus of scientists pointing to major methodological flaws. One core issue is that the risk of death among the non-drinking group is often inflated by significant numbers of “sick quitters” – people who have given up drinking for health reasons. Compared with these, it seems obvious that sensible types who drink in moderation are more likely to live longer.

Another problem is that many studies ask participants to self-report their alcohol intake, and their response is likely to be inaccurate and liable to change over time. When researchers account for these factors, the J-shaped curve tends to become a simple straight line. A paper published earlier this week re-analysed data from previous studies and found that the lowest mortality risk was in those who had never drunk at all.

However, according to Tim Stockwell, the lead author of the latest study, this should not be taken as a conclusive answer; rather, an indication of how much more work there is to do. “At the end of the day, we don’t take our findings literally, we don’t think this is necessarily an accurate picture,” he said. “It’s more like putting up a mirror to how bad the research is.”

A key reason why untangling the factors involved has proved so difficult is that alcohol affects everyone differently and everyone drinks for different reasons. “It’s tricky with alcohol because it has many different biological effects,” Iona Millwood, a researcher at the University of Oxford, said. “People’s drinking patterns are also determined by many other characteristics which themselves are going to have effects on health.”

To avoid these issues, Millwood’s study used a novel approach which separated people who are genetically predisposed to drink more or less alcohol, rather than relying on their self-reported drinking habits. They found that for 61 different outcomes – including many cancers, liver disease, stroke and overall mortality – the conclusion was simple: the more you drink, the higher your risk. Other genetic analyses have found similar results.

It is important to note that alcohol’s purported benefits only apply to cardiovascular disease and overall mortality (although Stockwell suspects the influence of the former drives the pattern in the latter). With many other diseases, the evidence is far more clear-cut. The overwhelming consensus on cancer, for example, is that any amount of alcohol increases your risk, as supported by a large analysis that compared light drinkers with lifetime abstainers and found significantly lower rates of mouth, intestinal and breast cancer in non-drinkers.

So if the tide of evidence is turning against alcohol’s beneficial effects, why does the idea persist in the public’s conscience? Mark Petticrew, a researcher at the London School of Hygiene & Tropical Medicine, said he believed the efforts of the drinks industry explained a lot. “One reason why there’s a public belief in these protective effects is because the industry has funded and promoted research, like the tobacco industry did.”

As evidence, Petticrew points to a 2021 analysis that looked at 60 different reviews of alcohol’s impact on cardiovascular disease risk. It found that 14 of them were either directly funded by the alcohol industry or involved researchers with links to the alcohol industry. All 14 concluded that small amounts of drink could protect against cardiovascular disease.

In the end then, much as people, and the drinks companies, might want to believe that a daily drink is the key to a long life, scientific support for the idea looks to be gradually fading.