Is alcohol good for your heart?

A report published in The Lancet this week has made a major contribution to our understanding. News reports initially suggested the study showed that drinking anything above 12.5 units a week will shorten your life, but that’s not really what it said. Like everything around alcohol and health risks, the story is more complicated than that.

The study looked at the risk of heart conditions and stroke associated with drinking at various levels. For many years, research seemed to point to a ‘J-curve’, showing that while very heavy drinkers were at the highest risk of suffering a serious heart condition, non-drinkers also had increased risk.

More recent analysis has tried to account for why non-drinkers might appear to have worse outcomes: in particular, whether the figures were skewed by ex-drinkers who perhaps gave up drinking because of another illness. Adjusting for this, a 2016 review suggested the J-curve simply didn’t exist. However, another study from University College London still found a J-curve after adjusting for ex-drinkers – though with different effects for different conditions.

But where does this new study fit in? The Lancet study is enormous: comparing the health records of just under 600,000 people against their drinking levels. It should therefore be seen as one of the most important recent contributions to the science.

Unlike many other studies the authors didn’t compare everyone against the non-drinkers, but against people drinking very lightly (in the chart below, up to about 3 units a week). When looked at this way, the J-curve seemed to flatten dramatically, but with the lowest risk overall at around 100 grams (or 12.5 units) per week. That’s about five pints of beer, or four large glasses of wine.

^{Source: Wood, A. M. et al. (2018). Risk thresholds for alcohol consumption: combined analysis of individual participant data for 599,912 current drinkers in 83 prospective studies – Supplementary appendix. The Lancet, v391.}

Risk increases slightly from there, up to around 25 units a week - after which it shoots up.

While the cardiovascular disease risk shows a curved line, closer analysis shows that this is mostly linked to incidence of heart attack, especially when non-fatal. The lines for strokes are almost straight, suggesting risk increases directly with the amount consumed.

^{Source: Wood, A. M. et al. (2018). Risk thresholds for alcohol consumption: combined analysis of individual participant data for 599,912 current drinkers in 83 prospective studies – Supplementary appendix. The Lancet, v391.}

When non- and ex-drinkers were included in the analysis, the J-curve reappeared. Non-drinkers had overall mortality risk similar to that for people drinking around 35 units a week.

_{Source: Wood, A. M. et al. (2018). Risk thresholds for alcohol consumption: combined analysis of individual participant data for 599,912 current drinkers in 83 prospective studies – Supplementary appendix. The Lancet, v391.}

We still don’t know why this is the case: some argue that non-drinkers are just a poor comparison because even those who have never consumed alcohol are more likely to have pre-existing conditions. Others argue that, until proved otherwise, this remains evidence for a protective effect.

What is the take-home from all this? Well, firstly that the science is complicated, much as we’d like it to be otherwise. It’s not possible to say, ‘drinking x amount will make your life longer, or shorter.’ Risk just doesn’t work that way (even if headline-writing does), and there are too many other factors to consider.

Secondly, it looks like the lowest risk for overall mortality is somewhere close to the Chief Medical Officers’ low-risk guidelines of 14 units a week. Within that, the risks for specific conditions vary.

Does this mean abstainers should start drinking to improve their heart health? Well, firstly there are many better ways to protect your heart (not smoking, regular exercise, good diet etc.). Also, this idea would only make sense if your sole reason to drink is to prevent particular heart conditions (and you’re not so concerned about cancer risks); only if you’re going to stick resolutely to two units a day (the risks increase when consumption is packed into fewer drinking sessions); and only if you’re convinced the reduction in those particular risks outweighs your other reasons for not drinking in the first place.

For drinkers, the message is that if heart health is the reason you are drinking (is it, really?) then that 14 unit a week level is the best call, spread over a number of days. But if you pour yourself that second, or third, glass, saying, ‘Ah well, it’s good for the heart,’ you’re probably using a bit of wishful thinking.