Craft beer in Melbourne, pinot noir in Paris, mojitos in Havana –alcohol is part of nearly every culture around the world. As the evidence of harm has accumulated, the message from health authorities globally is that we need to drink less and even abstain entirely. But the advice on exactly how much less isn’t consistent. Why?

As an addiction medicine specialist and public health physician at the University of Sydney, Prof Kate Conigrave is more aware than most about the damage caused by alcohol. She chaired the group who authored Australia’s latest guidelines on reducing the health risks of alcohol, which in 2020 recommended healthy adults drink no more than 10 standard drinks a week, and no more than four in a day. It’s a reduction from the 2009 guidelines which recommended no more than two drinks a day for men and women.

“The more we know about alcohol, the more we know that the risk of harms can cut in from low levels of drinking,” she says.

The message that alcohol is a social and physical poison is being repeated around the world with increasing confidence, as the evidence of the harms done by its consumption continues to build. But if the evidence is clear, why do guidelines on what is low-risk drinking vary so much?

For example, Belgium labels up to 21 drinks a week for men and 14 for women as low risk, while Ireland goes with up to 17 drinks for men and 11 for women, with two alcohol-free days a week. France recommends no more than 10 standard drinks a week – the same as Australia – but never more than two standard drinks a day and at least one alcohol-free day a week. The UK advises no more than 14 units a week, over at least three days, and “some” alcohol-free days, while the US recommends no more than two drinks a day for men and one for women.

To further confuse things, countries have different definitions of what constitutes a standard drink. In the US, a standard drink contains 14 grams of alcohol, in Australia it’s 10 grams and in the UK it’s about eight grams.

In some countries there is separate advice for men and women, while in others – including Australia – there is not. Conigrave says at the lower levels of drinking advised in the Australian guidelines, the health impact is similar for men and women. At higher levels of consumption, women are more vulnerable to harm.

There’s also the question of alcohol’s harms varying across age groups. “There’s general acceptance that when you reach older age – especially if you start getting frail – you’re probably going to want to drink less,” she says. The Australian guidelines are also clear that those under 18 should not drink alcohol at all.

But the guidelines don’t tailor their advice to subsets of adult age groups, even though there are likely to be different risks based on age. “If we’re going to really individualise it, we’d say, ‘Well, you’re 24, your brain’s still developing … so maybe you should drink a bit less than you should a year later,” Conigrave says. “The trouble is it’s a trade-off between individualising it and having a usable guidelines.”

Drawing the line

The biggest question for those who make these recommendations is where to set the threshold between what is perceived as being low-risk drinking and what is moderate or high risk. If the harms of alcohol – and there are many, ranging from heart disease, liver disease, brain damage, cancer and high blood pressure – increase in a linear or upwardly curved fashion with increasing alcohol intake, how do we choose what constitutes low risk?

“The question of where to draw the line is one of the reasons I think that we have such variation in how guidelines look globally,” says Assoc Prof Michael Livingston, an alcohol policy researcher at Curtin University’s National Drug Research Institute. “There’s no easy answer.”

In Australia the guidelines committee chose a threshold for low risk as being a one-in-100 risk of dying from alcohol-related causes. The reason, according to Prof Robin Room – a sociologist at La Trobe University’s centre for alcohol policy research, who has worked on previous alcohol recommendations – is that this is approximately the same level of risk associated with driving a car. “If you look around at what’s the closest thing to alcohol in terms of being a risky behaviour that’s regularly undertaken by a lot of people… then the closest you can easily come in Australia is what the governments do around driving,” Room says.

Canada has taken an approach based on evidence suggesting a possible health benefit – mainly for heart disease – associated with very low levels of alcohol consumption. This evidence is contested and controversial, because it could be an effect of people who drink low levels of alcohol being more likely to engage in other healthy habits, while those who don’t drink at all doing so because of pre-existing poor health or previous alcohol abuse.

But the Canadian guidelines take this effect into account. “The point where the risk crosses back up above abstaining – so as soon as your risk is higher than not drinking at all – that’s where they’ve drawn the line,” Livingston says.

How low can it go?

Overall, the number of drinks a week considered low risk has been steadily decreasing over the years, as more evidence emerges of the negative health effects, particularly an increased risk of cancer. So how low can it go?

Room says alcohol guidelines are unlikely to ever go to zero, because alcohol is “our drug”. “Wherever you look, there are people who expect to have alcohol as part of their life, essentially, and that’s very different from what would be true for tobacco at this point,” he says.

The Netherlands has pushed the low-risk drinking threshold the furthest. Its guidelines first advise drinking no alcohol at all but, if people choose to drink, they should have not more than one glass a day. “That seems to me a kind of bottoming out,” Livingston says. “It’s hard to imagine a lower guideline than that.”

The elephant in the room is the question of whether alcohol consumption guidelines actually make any difference to drinking behaviour. Livingston says there’s a small amount of evidence that changing guidelines on drinking do have an impact on people’s perceptions of how much alcohol is “safe”. But it’s not strong evidence, he says.

“They might work in a more subtle way, where they gradually can shift the norms around alcohol,” he says. However changing the guidelines on alcohol consumption is not going to solve a country’s alcohol problems.

“It’s not going to lead to dramatic behavioural changes but, if well-promoted and embedded in general health promotion, it’s potentially a lever that could have an influence.”