0
0
It is an ordinary Tuesday. Dinner is cleared, the day is finally quiet, and you pour a second glass. And somewhere behind the pour there is a small question you have probably never said out loud: is this still fine?
If you went looking for a number to answer that, you found this one: two drinks a day is linked to roughly a 1-in-25 lifetime risk of dying from an alcohol-related cause.
That figure is real, it comes from a real study, and this article is not going to soften it. But it also does not mean what most people reading it assume it means. And the most useful parts of that research are the parts almost nobody quoted.
So: here is the number, what it covers, what it cannot see, and one detail about it that quietly changes the answer for a lot of people.
What the study on two drinks a day actually found
The figure comes from the Alcohol Intake and Health Study, published in the Journal of Studies on Alcohol and Drugs in June 2026. More than twenty researchers, led by Kevin Shield of the University of Toronto, reviewed over 7,200 scientific publications on alcohol-related illness and injury, estimated the risk for each condition, then combined those with large national health data sets to model what different drinking levels do across a population.
Three findings carry the weight:
- At an average of 14 drinks a week — two a day — the estimated lifetime risk of dying from an alcohol-related cause was about 1 in 25.
- At up to 7 drinks a week — one a day — risks were only minimally elevated for most of the conditions examined.
- No protective effect appeared at any level. “We did not observe a significant protective effect of alcohol on health at any level of consumption,” Shield said.
That last point is the one that broke with decades of received wisdom about a glass of red being good for the heart. The researchers did see a possible reduced risk of ischemic heart disease and stroke at low levels, and they kept it in the paper rather than quietly dropping it. They concluded it was outweighed by the other risks, with cancers of the oesophagus, mouth and breast leading that list.
One thing to hold onto before anything else: these are associations, estimated with a model. Nobody followed you, or anyone like you, for a lifetime and counted. The team took relative risks from existing research and applied them to national drinking and death data. That is a legitimate way to answer the question. It is not the same thing as a measurement.
What counts as “a drink” — and why your glass may not be one
Here is the detail that changes the answer for a lot of people, and it appeared in almost none of the coverage.
The study counts in US standard drinks. One US standard drink is 14 grams of pure alcohol — roughly 355 ml of ordinary beer, 150 ml of wine, or a 45 ml measure of spirits.
That definition is not universal. A unit in the United Kingdom is about 8 grams. Australia, Japan and the World Health Organization use 10. Austria uses 20. The same word can mean more than twice as much alcohol depending on which country printed the advice you are reading.
Which sets a genuine trap. The study’s threshold is 14 drinks a week — about 196 grams of pure alcohol. The UK’s own long-standing guidance is to stay under 14 units a week — about 112 grams. Both numbers are “14”. One is nearly double the other.
Then there is the pour. A 150 ml glass of 12% wine is one standard drink. A 250 ml glass of 13% wine is closer to 1.8 of them — so two generous glasses poured at home sits nearer three and a half standard drinks than two. That is arithmetic off the bottle, not a finding from the study, and you can check it yourself: multiply the millilitres by the percentage to get the millilitres of pure alcohol, and 14 grams is about 17.7 ml of it.
None of this makes the risk larger or smaller than it is. It means the honest version of the question was never “how many glasses do I have?” It is “how much alcohol is actually in them?” — and plenty of people who believe they are sitting at two a day are not.
Why nobody will simply give you a number
There is a reason this figure took effort to find.
The study was not a free-standing academic exercise. It was commissioned by the US federal government and explicitly invited to help inform the 2025–2030 Dietary Guidelines for Americans. It did the job it was asked to do.
The guidelines, when they arrived, told people to “limit alcoholic beverages” and named no threshold at all. The previous edition had given one: up to two drinks a day for men, one for women. The new one gives nothing.
In an editorial published alongside the study, Robert M. Vincent, a former associate administrator at the US Substance Abuse and Mental Health Services Administration, wrote that “despite the study’s adherence to its mandate, its findings were sidelined.”
So the reason you had to go hunting for a number is not that nobody worked one out. Someone was asked to work one out, did it, and the number never made it to the label. You were left to find it on your own, at night, in a search bar.
What the researchers said about your risk
This is the part someone frightened by the headline needs most, and it comes from the lead author rather than from a critic.
“We can’t assume that means one person’s individual health risk is the same as what is reported here,” Shield said.
That is not a polite hedge. It is the actual boundary of the method. The study estimates what happens across a whole population of drinkers. Your own risk is shaped by things the model knows nothing about: your genetics, whether you smoke, what else is happening in your body, and how the drinking is spread across the week rather than just how much of it there is. Pattern matters as much as volume — the first real study of student binge drinking in gallon jugs found the same thing coming from the opposite extreme.
The authors also left out conditions where the evidence is still unsettled, pancreatic cancer among them, which they said “still needs further work.” And the paper has critics; some have publicly challenged how the risks were modelled.
What a 1-in-25 lifetime risk honestly tells you is this. It is not nothing. It is also not a sentence. It is a reason to take the question seriously — not a verdict already passed on your life.
The question underneath the question
Which leaves the one thing the number cannot do. It cannot tell you whether you have a problem.
Because almost nobody drinks two a day for the taste of the second one. The second glass usually has a job. It is the day formally declaring itself over. It is the noise in your head going quiet. It is a switch you have not found another way to reach.
That is why “how much” is the wrong question to be afraid of. Some people drink well under the threshold and are in real trouble. Others sit close to it and are not. The number measures volume. It cannot see the job. Which may be part of why younger drinkers have been walking away from alcohol faster than anyone predicted — not because they were frightened by a statistic, but because they found other ways to get the same thing.
And it is a strangely old observation. The ancient wisdom people assume is simply anti-drink is not. It treats wine as a genuine gift, something that gladdens the heart — and the people who believed that gift came from God were the same ones who warned about it most sharply, often within a few lines of the same page. What is easy to read past is that the warnings almost never set a quantity. They are about something else entirely — who has the controls. Being filled by a thing. Whether you are holding it or it is quietly holding you. Whoever wrote that down was asking about custody, not volume. Which happens to be the question the modern number cannot reach, and the one that actually tells you where you stand.
A simple way to find out where you stand
If you want the answer the statistics cannot give you, it is not a calculation. It is a pause.
- Work out what you are actually drinking. Not glasses — millilitres and percentage, across one ordinary week. Most people are surprised, in one direction or the other.
- Stop for two weeks. Not forever. Two weeks.
- Notice how hard it is. This is the entire test. If it is easy and a little dull, you have your answer. If you find yourself negotiating, rescheduling it, or deciding that this particular fortnight is a bad one to try — that is information, and it is far more useful to you than 1 in 25 is.
- Find out what the glass was doing. If the evenings turn long and loud without it, the drink was never the point. The thing it was switching off is the point. If what it was switching off is a mind that will not go quiet at bedtime, our free Why Can’t I Sleep? assessment walks through the likeliest reason in a couple of minutes.
And if any of that reads as frightening rather than interesting, that is worth taking seriously rather than arguing with. Most countries run a free and confidential alcohol helpline, and a doctor will have had this exact conversation several times this month without a flicker of judgement. Asking early is not an admission of anything. It is just a shorter road.
Where that leaves the second glass
The number is worth knowing. Two drinks a day is not the safe harbour the older guidance implied, and the research no longer supports treating it as one.
But a lifetime risk is a statement about twenty-five people. It is not a statement about you. The thing that is about you is quieter, and easier to check: pour the second glass tonight and notice whether you chose it, or whether it had already been decided for you.
Nobody can publish that number. It is just worth knowing which one you are.
Questions people ask about two drinks a day
Is two drinks a day bad for you?
The Alcohol Intake and Health Study, published in the Journal of Studies on Alcohol and Drugs in June 2026, estimated that an average of 14 drinks a week — two a day — is associated with roughly a 1-in-25 lifetime risk of dying from an alcohol-related cause. The researchers found no significant protective effect from alcohol at any level of drinking. These are population-level estimates from a statistical model, not predictions for any individual person, and the lead author has said an individual’s risk cannot be assumed to match the figure.
What is a standard drink?
A US standard drink contains 14 grams of pure alcohol, which is about 355 ml of regular beer, 150 ml of wine, or a 45 ml measure of spirits. The definition varies widely by country: a unit in the United Kingdom is about 8 grams, Australia, Japan and the World Health Organization use 10 grams, and Austria uses 20. Because of this, the same stated number of “drinks” can represent very different amounts of alcohol depending on which country’s guidance is being quoted.
How many drinks a week is considered low risk?
In the Alcohol Intake and Health Study, drinking up to 7 US standard drinks a week — an average of one a day — was associated with only minimally elevated risks for most of the conditions examined, and risk rose above one drink a day for both men and women. The 2025–2030 Dietary Guidelines for Americans advise limiting alcoholic beverages without naming a specific threshold, while the United Kingdom’s long-standing guidance is to stay under 14 units a week, which is about 112 grams of pure alcohol.
Does moderate drinking have any health benefits?
The Alcohol Intake and Health Study found no significant protective effect of alcohol on health at any level of consumption. The researchers did observe a possible reduced risk of ischemic heart disease and stroke at low levels of drinking, but concluded that any such benefit was outweighed by increased risks of other conditions, including cancers of the oesophagus, mouth and breast. This contradicts the long-popular belief that a daily glass of wine is good for the heart.
Does a 1-in-25 lifetime risk mean I will die from alcohol?
No. A 1-in-25 lifetime risk is a population-level estimate, meaning that out of 25 people with that drinking pattern, roughly one death would be attributable to alcohol over a lifetime. It is not a prediction about any particular person. Individual risk is shaped by genetics, smoking, other health conditions, and how drinking is spread across the week — factors the statistical model does not account for. The study’s lead author, Kevin Shield, cautioned that one person’s individual health risk cannot be assumed to be the same as the reported figure.
Something to think about
Here is the open question. Should health guidelines hand people a specific number, even one that will be argued over for years — or is “limit alcoholic beverages” the more honest answer while the science is still moving? One side says people deserve a figure they can actually use to make a decision. The other says any number people can clear quietly becomes permission. Which do you think serves people better? Tell us in the comments — we read them.
Share this
- Turns out the study everyone quoted counts in US standard drinks — 14 grams of alcohol. A UK unit is 8. Both guidelines say “14 a week.” One is nearly double the other. https://bgodinspired.com/index.php/health-and-wellness/two-drinks-a-day-risk/
- Two drinks a day is linked to a 1-in-25 lifetime risk of an alcohol-related death. But the detail that actually got me: the study was commissioned by the government to set a guideline number, and the guideline came out with no number in it. https://bgodinspired.com/index.php/health-and-wellness/two-drinks-a-day-risk/
- Best thing I’ve read on moderate drinking in a while, because it doesn’t lecture. The number can’t tell you if you have a problem — but stopping for two weeks and noticing how hard it is, can. https://bgodinspired.com/index.php/health-and-wellness/two-drinks-a-day-risk/
