If you think a little red wine is good for the heart, you're not alone - but you're wrong.

"Pretty much everyone believes that red wine may have health benefits," said Professor Tim Stockwell, the former director of the Canadian Institute for Substance Use Research (CISUR).

"It's a lovely idea, and people love to hear it, but the evidence doesn't stack up."

Prof Stockwell is speaking to the Herald ahead of a talk in Edinburgh on April 17, jointly organised by Scottish Health Action on Alcohol Problems (SHAAP), the Institute of Alcohol Studies (IAS), and the Royal College of Physicians Edinburgh (RCPE).


For roughly 20 years he has been at the vanguard of research challenging the widely held belief that alcohol in moderation has cardiac benefits, although he admits that he was once on the other side of the fence.

"I published a piece in the Medical Journal of Australia where I likened people who doubted the protective effects of alcohol to members of the Flat Earth Society, or people who doubted that there were manned landings on the Moon," said Prof Stockwell, who headed up Australia's National Drug Research Institute during the 1990s.

He changed his mind after collaborating with the San Francisco-based medical sociologist, Kaye Fillmore, who was convinced the evidence that moderate drinkers had better health outcomes than abstainers was flawed.

Their landmark 2007 meta-analysis found that a "systemic error" was behind the famous J-curve, a epidemiological conundrum which implied that light to moderate alcohol intake had a more protective effect against coronary heart disease than avoiding alcohol altogether.

Prof Stockwell said: "I started out in a position of scepticism and I think the broad consensus, 25 years ago, was that that's how it was [that moderate alcohol was protective].

"There are still holdouts now, but I think the majority of people in epidemiology - especially alcohol epidemiology - are sceptical about the health benefits."

The Herald: An example of the famous 'J curve' found in alcohol studies which compares consumption to all-cause mortality rates, suggesting that one drink per day may be optimal for healthAn example of the famous 'J curve' found in alcohol studies which compares consumption to all-cause mortality rates, suggesting that one drink per day may be optimal for health (Image: House of Commons Library)

The British-born academic says the notion that alcohol might do us good has a long history - "there's references in the Bible and Greek philosophers talk about how alcohol is good for health" - but that the contemporary concept really took off in 1991, when America's CBS broadcast a famous 60 Minutes programme on the so-called 'French Paradox'.

"We didn't have the internet then but it 'went viral'," said Prof Stockwell, who is currently based at the University of Victoria in Canada.

"The French Paradox was this idea that French people don't take much exercise and they have quite a fatty diet, and yet they have less chance of heart disease than the Irish for example, or the Scots, so why is that?

"The 60 Minutes programme said 'it's red wine'."

Within a year, sales of wine in the United States had soared by 40%.

More than 30 years on, research suggests that the paradox is more probably explained by comparatively high levels of fruit, vegetables, wholegrains, olive oil, and lean protein in the French diet, but scientists have also identified that resveratrol - a chemical found most abundantly in red grapes - does have antioxidant and anti-inflammatory effects.

Human research trials have suggested that a 500mg dose can deliver health benefits; the problem is that you would have to consume around 40 litres of red wine in one sitting to get the same amount, by which point you might be dead from alcohol poisoning.

The Herald: The idea that red wine promotes heart health and longevity has taken hold in popular cultureThe idea that red wine promotes heart health and longevity has taken hold in popular culture (Image: Getty)

Other studies have suggested that wine drinkers have a lower risk of heart disease than beer and spirit drinkers, but - as Prof Stockwell notes - wine drinkers "are not the same as beer or spirit drinkers".

He added: "People who like wine tend to be better off, more educated, have more moderate lifestyles, so there is a lot of bias and confounding."

Extracting these "biases" is one of the biggest obstacles for alcohol epidemiology, which relies on observational studies.

This means you take a large cohort of adults at a given point in time and follow them up over decades to find out what they die of, when, and how this correlates against their alcohol consumption.

Ideally you want to control for every other confounding variable - from smoking and deprivation status to obesity and exercise levels - that might also increase their likelihood of dying prematurely, but this is not easy to do.

Time and again, the data will throw up a perplexing J-shape curve which suggests that drinking a moderate amount of alcohol is healthier than being teetotal, with similar patterns occurring in other areas of health research - for example, suggesting that being slightly overweight is better for you than a 'normal' BMI.

When Prof Stockwell began to unpick the alcohol puzzle he found that the 'abstainer' group was misleading, because it was skewed by "sick quitters" - people who had cut down or stopped drinking due to illness or alcohol dependency.

In addition, abstainers tend to be poorer and more likely to be overweight, have a disability, poor dental health, a worse diet, and lower activity levels than the low-to-moderate drinkers.

The Herald: Prof Stockwell was also the first to propose linking alcohol price to ethanol content, in what became known as minimum unit pricingProf Stockwell was also the first to propose linking alcohol price to ethanol content, in what became known as minimum unit pricing (Image: CISUR)

In this interpretation, the J-curve - with its implication that moderate alcohol consumption reduces premature mortality - actually reflects how difficult it is to control for all these risk factors.

"The simple statistical fixes don't go far enough," said Prof Stockwell.

There is also an issue that the data can be skewed by a "healthy surviving drinker" effect. In other words, that some heavier drinkers are naturally more robust and predisposed to longevity.

Intriguingly, recent research by Prof Stockwell and colleagues - which has attempted to control for all these methodological issues - found that the appearance of an alcohol benefit "either shrinks so that it's no longer significant or completely vanishes".

If they are correct, it has potentially far-reaching implications for how we evaluate alcohol harms.

Prof Stockwell said: "It doesn't just have implications for heart disease, it has implications for our estimates of how risky alcohol is for cancer, or liver cirrhosis, because we're still comparing drinkers to unhealthy abstainers and you haven't fully controlled for all these lifestyle factors.

"The implication is that we may have really underestimated alcohol harm, which is a bit scary. But we're still talking quite small risks at low levels of consumption."

Putting that into context, Prof Stockwell says that the evidence suggests consuming as little as two servings of alcohol per week (where one serving equals a bottle of beer, a 140ml glass of wine, or a 43ml shot of spirits) will increase your risk of dying from an alcohol-related illness before the age of 75 by 0.1% versus abstaining.

Six drinks per week increase that by 1%.


Prof Stockwell's talk in Edinburgh on Wednesday is also expected to coincide with MSPs voting on a motion to increase Scotland's minimum unit price (MUP) levy from 50 to 65 pence.

Back in 2006, when the UK's then-Labour government was exploring options for minimum pricing, it was Prof Stockwell who advised them to link it to ethanol content - a novel approach which formed the basis for the minimum unit price model subsequently adopted by Scotland in a world first.

He said he is confident that MUP has made a difference to heavy drinkers, despite some surveys suggesting they had not cut down. 

He said: "If you look at total sales data: 50% of the alcohol sold is consumed by heavy drinkers.

"You don't get population level reductions - a 3% total reduction in consumption compared to England and Wales - unless heavy drinkers are cutting down. It's mathematically impossible.

"All the surveys done, however good they seem, are observational studies - not control studies.

"So it really has worked, especially in people with heavy alcohol use."