A glass of red wine a day helps you live longer
Short answerNot the way you think
On this page
Not the way you think. For decades light drinkers seemed to outlive non-drinkers, but the non-drinkers included people who’d quit because they were already ill. In 107 studies of 4.8 million people, once lifelong non-drinkers were kept apart from former drinkers, light drinking didn’t go with a longer life. If you don’t drink, don’t start.
What we’d do
If you don’t drink, don’t start. If you do, keep it to a few drinks a week with days off, and treat it as a pleasure with a cost rather than a health habit. If you have more than two a day, cutting back is one of the changes with real years attached. If you drink heavily every day, talk to a doctor before you stop, since stopping suddenly can be dangerous.
If you remember three things
- Sick quitters made light drinking look good. In 7,735 British men, the benefit only showed up in men who already had heart-related illness when the study began.
- Once former drinkers were counted apart, light drinking added nothing. People who had up to about 24 g a day, under two US drinks, didn’t have a lower death rate than lifelong non-drinkers.
- Heavy drinking costs years. 45 to 64 g a day went with a 19% higher death rate and 65 g or more with 35% higher.
Why everyone believes it
Because the studies said so, for a long time. Plot death rates against drinks in almost any older study and you get a U shape. Non-drinkers at one end, heavy drinkers at the other, and light drinkers in the dip with the lowest risk.
Even the big Harvard study this site’s calculator uses found it. People who didn’t drink at all had a 27% higher death rate than people who had 5 to 15 g a day, about one drink. Add a story about red wine and a compound called resveratrol, and a glass with dinner started to sound like medicine.
A picture from daily life
Imagine judging whether gyms are good for you by comparing members with people who cancelled. The cancelled group includes everyone who quit after an injury or an illness. Of course the members look healthier. That says nothing about what the gym did.
Who were the non-drinkers?
In 1988 a team of British researchers looked closely at 7,735 middle-aged British men followed for 7.5 years, of whom 504 died. They found the familiar U shape. Then they split the men by their health when the study began. The U shape was there only in men who already had heart disease or a related illness diagnosed by a doctor. Many of them had moved into the non-drinking and occasional drinking groups because they were sick. The authors wrote that the idea of a protective effect of drinking “is likely to be ill founded”.
That was the problem with the “non-drinkers” in many studies. They mixed people who’d never drunk with people who’d stopped, often because of illness, heavy drinking or a doctor’s advice.
What happened when they were kept apart?
In 2023 Zhao and colleagues pooled 107 studies with 4,838,825 people and 425,564 deaths, and adjusted for former drinkers and other known biases.
| Daily alcohol | About how much | Death rate against lifelong non-drinkers |
|---|---|---|
| Under 1.3 g | An occasional drink | No real difference (4% lower, not significant) |
| 1.3 to 24 g | Up to nearly 2 US drinks | No real difference (7% lower, not significant) |
| 25 to 44 g | About 2 to 3 drinks | 5% higher, not significant |
| 45 to 64 g | About 3 to 4.5 drinks | 19% higher |
| 65 g or more | About 4.5 drinks or more | 35% higher |
A US standard drink has about 14 g of alcohol, which is a 12 ounce beer, a 5 ounce glass of wine or a 1.5 ounce shot. For women, the extra risk started at lower amounts than for men.
What about resveratrol?
Red wine contains resveratrol, the compound behind much of the anti-aging hype. When researchers at three labs fed it to genetically mixed mice, at two doses, it did nothing for how long they lived. Do longevity pills like resveratrol add years?
Why does our calculator count drinking at all?
Because the Harvard study counted moderate drinking (5 to 15 g a day for women, 5 to 30 g for men) as one of its five habits, so people who didn’t drink scored as missing it. Our calculator counts no more than 7 drinks a week for women and 14 for men, none included, as meeting the habit. Drinking only ever shows up as a cost, at 15 drinks a week or more, where the Harvard study found 25% more risk at 30 g a day and up. It never suggests starting.
Aren’t there still studies showing a benefit?
The sceptic says
Light drinkers still look a bit better. Even in the 2023 pooled study, low-volume drinkers had a 7% lower death rate.
The reply
That 7% was within chance, after the researchers adjusted for former drinkers and study quality. Every one of these studies follows people rather than assigning drinks, so a small leftover difference could come from anything else light drinkers have in common. Nobody has shown that starting to drink adds years.
5 sources
- Zhao J and others, Association Between Daily Alcohol Intake and Risk of All-Cause Mortality: A Systematic Review and Meta-analyses, JAMA Network Open, 2023pubmed.ncbi.nlm.nih.gov
- Shaper AG and others, Alcohol and mortality in British men: explaining the U-shaped curve, Lancet, 1988pubmed.ncbi.nlm.nih.gov
- Li Y and others, Impact of Healthy Lifestyle Factors on Life Expectancies in the US Population, Circulation, 2018pmc.ncbi.nlm.nih.gov
- Miller RA and others, Rapamycin, but not resveratrol or simvastatin, extends life span of genetically heterogeneous mice, Journals of Gerontology Series A, 2011pmc.ncbi.nlm.nih.gov
- National Institute on Alcohol Abuse and Alcoholism, What is a standard drink?niaaa.nih.gov
