Alcohol and cardiovascular risk: what the science says today

For decades, the idea that moderate alcohol consumption could protect the heart was widely held, including in medical discourse. The most recent research invites us to revisit this assumption.

The stakes are far from theoretical. In France, Santé publique France estimates that about 9,900 cardiovascular deaths a year are attributable to alcohol, or a quarter of alcohol-related deaths. In the United States, the CDC has shown that, among women, heart disease and stroke are the leading causes of death linked to excessive drinking. Globally, the World Health Organization (WHO) attributes about 474,000 cardiovascular deaths to alcohol in 2019.

Where the idea of a protective effect came from

Since the 1970s-1980s, several large observational studies had shown a “J” or “U”-shaped curve. People who drank moderately appeared to have a lower cardiovascular risk than non-drinkers, with the risk then rising again for heavier consumption. This observation, repeated across several cohorts, long steered public health messaging toward the idea of a beneficial “reasonable” level of drinking.

The “French paradox” played a major role in spreading this idea. In the early 1990s, researchers noticed that France had relatively few heart attack deaths despite a diet rich in saturated fat. Red wine was put forward as a possible explanation, picked up by a widely watched American TV program in 1991, which caused red wine sales to surge in the United States. Resveratrol, a substance found in grape skin, was then presented as the protective active ingredient. We now know that the amounts of resveratrol in a glass of wine are far too small to have a measurable effect. The “paradox” is also partly explained by differences in how deaths are recorded between countries, and by other lifestyle factors.

A biological explanation also seemed plausible: alcohol modestly raises “good” cholesterol (HDL) and slightly thins the blood. But genetic studies have since shown that raising HDL doesn't, by itself, provide protection against heart attacks. This reasoning has therefore lost much of its force.

What more recent studies call into question

The problem with observational studies is that they can't establish cause and effect. Other factors, such as an overall healthier lifestyle among moderate drinkers (diet, physical activity, socioeconomic status), could explain part of the apparently protective effect, without alcohol itself being the cause.

One particular bias played a central role: that of “sick quitters.” In many older studies, the “non-drinkers” group included people who had stopped drinking precisely because they were ill or had had problems with alcohol. This group therefore had a higher risk for reasons that had nothing to do with abstinence, which made moderate drinkers look healthier by comparison. A meta-analysis published in 2023 in JAMA Network Open (Zhao et al.) reanalyzed 107 studies covering nearly 5 million people while correcting for these biases. It found no reduction in mortality among light or moderate drinkers compared with people who had never drunk.

Another approach made it possible to go further: Mendelian randomization. It uses genetic variants that influence alcohol consumption, randomly assigned at birth, as a kind of natural clinical trial. In 2022, a study published in JAMA Network Open (Biddinger et al.) applied this method to more than 370,000 participants from the UK Biobank cohort. It found that light drinkers did indeed have healthier lifestyles than non-drinkers, which explained their apparent advantage. Once these factors were accounted for, cardiovascular risk rose with consumption, modestly at low doses and then increasingly sharply. A large Chinese study published in The Lancet in 2019 (Millwood et al.) reached a similar conclusion for stroke. It took advantage of a genetic variant common in East Asia that makes alcohol very unpleasant and therefore reduces consumption: stroke risk rose steadily with the amount of alcohol, with no protective effect at low doses.

Overall, this body of work concludes that there is no level of consumption that is genuinely beneficial for the heart. The more consumption increases, the more the risk of hypertension, heart attack, stroke, heart failure, and atrial fibrillation rises, progressively.

It's worth being honest about the nuances. A large analysis of 83 studies published in The Lancet in 2018 (Wood et al.) observed a slightly lower heart attack risk among drinkers, while the risks of stroke, heart failure, and death from hypertension increased with consumption. The overall balance remained unfavorable beyond about 100 grams of alcohol a week, or 10 standard drinks. The scientific debates therefore concern small effects at low doses, not the general direction of the risks.

A risk-factor approach rather than an overall-benefit one

Alcohol acts on the cardiovascular system through several well-identified mechanisms, independent of any supposedly protective effect. It promotes high blood pressure, it can cause heart rhythm disorders (including atrial fibrillation), and at higher doses it can directly damage the heart muscle (alcoholic cardiomyopathy). These harmful effects are well established and are not debated in the scientific literature.

High blood pressure is the most common mechanism. Alcohol activates the sympathetic nervous system and hormones that constrict blood vessels and retain salt. The effect is reversible: a meta-analysis published in The Lancet Public Health in 2017 (Roerecke et al.) showed that, among people drinking more than two drinks a day, cutting consumption roughly in half lowered blood pressure by several millimeters of mercury, an effect comparable to that of certain medications. If you track your blood pressure, the hub's home blood pressure monitoring tool can help you follow it over time.

Atrial fibrillation, the most common rhythm disorder, has a particularly clear link with alcohol. Doctors even speak of “holiday heart syndrome” to describe episodes that occur after heavy drinking during parties or weekends. In people who already have atrial fibrillation, a randomized trial published in the New England Journal of Medicine in 2020 (Voskoboinik et al.) showed that stopping alcohol markedly reduced recurrences. This is one of the rare situations where direct experimental evidence of the benefit of cutting back is available.

Alcoholic cardiomyopathy generally occurs after several years of heavy drinking. The heart muscle, damaged by alcohol and acetaldehyde, dilates and pumps less effectively, which can lead to heart failure. It can improve dramatically after stopping alcohol when caught early enough.

Alcohol also provides “empty” calories, at 7 kilocalories per gram, or about 70 per standard drink. It thus promotes weight gain and raised triglycerides, two other cardiovascular risk factors.

What this means in practice

These findings are not a reason to be alarmed for people with occasional, moderate consumption. At low doses, the increase in risk exists, but it remains small in absolute terms. They are, however, a reason to stop viewing alcohol as an ally of cardiovascular health, even at low doses. The question is no longer “how much can I drink to protect my heart,” but rather “what level of consumption best limits the risks,” knowing that this level trends toward zero rather than toward some “beneficial” threshold to reach.

Medical societies have updated their guidelines accordingly. The European Society of Cardiology, in its 2021 prevention guidelines, advises not exceeding 100 grams of alcohol a week, which matches the French guideline of 10 drinks. The World Heart Federation stated in 2022 that no amount of alcohol is good for the heart. No health authority recommends starting to drink to protect your heart, and this has never officially been the case.

To protect your heart, other levers have a far larger demonstrated effect: quitting smoking, controlling blood pressure, moving regularly, and keeping an eye on weight and cholesterol. The hub's overall cardiovascular risk calculator can estimate your overall risk while accounting for all of these factors together.

To estimate your own risk level, take the “Is my alcohol consumption risky?” test.