For decades, a glass of wine with dinner was seen as potentially healthy. Some studies suggested moderate alcohol use might be fine and might even offer minor health benefits. Increasingly, newer and better science is contradicting that idea.
This new study was carried out by an international team of scientists, led by researchers from the University of Oxford, Yale, and Cambridge. The team found that any amount of alcohol consumption may raise the risk of developing dementia. Their results, published this week in BMJ Evidence-Based Medicine, suggest that the notion of a “safe” level of drinking (at least when it comes to brain health) may be an illusion.
“Our findings challenge the common belief that low levels of alcohol are beneficial for brain health,” said Dr. Anya Topiwala, senior clinical researcher at Oxford Population Health and lead author of the study. “Genetic evidence offers no support for a protective effect—in fact, it suggests the opposite.”
What Past Studies Missed
The idea that light drinking might be protective has been around for years. Older studies appeared to show that people who consumed small amounts of alcohol had a lower risk of dementia than heavy drinkers and even abstainers. This pattern formed a J-shaped curve, seemingly suggesting that moderation was the key and low consumption was alright.
But critics have long warned that such studies were misleading. Many of the people in the “non-drinking” category had, in fact, stopped drinking due to health problems. They were already sick, and that skewed the results.
“A lot of people who don’t currently drink are people who used to drink heavily, or who have health problems that led them to quit,” said Dr. Keith Humphreys, professor of psychiatry at Stanford, in a separate analysis. “That skews the data, making moderate drinkers look healthier by comparison.”
To get around this problem, Dr. Topiwala and her colleagues used a dual approach. First, they analyzed data from 559,559 people in the UK Biobank and the U.S. Million Veteran Program. Then, they layered in genetic data from 2.4 million participants across 45 studies.
Their results were striking.
The Illusion of Safe Drinking
When the team looked only at self-reported drinking patterns, the familiar J-shape emerged. People who drank lightly (less than seven drinks a week) appeared to have the lowest risk of dementia. Both heavy drinkers and abstainers had higher rates.
But things changed when the researchers applied a genetic technique called Mendelian randomization. This method uses genetic markers to estimate a person’s lifetime tendency to drink, offering a less biased lens on long-term alcohol exposure. In these analyses, any increase in genetically predicted alcohol use correlated with a higher risk of dementia. There was no sweet spot, no U-shaped curve—just a steady upward slope. The more you drink, the worse it gets, and no amount is safe.
A genetically predicted threefold increase in weekly drinks was associated with a 15% increase in dementia risk. Similarly, a twofold increase in the likelihood of alcohol use disorder raised the risk by 16%.
This finding dovetails with broader critiques of alcohol science. As Stanford’s Dr. Randall Stafford put it, “We have bought into a storyline about alcohol that, when you really look at the facts, is not there.”
Genetic Clues, Diverse Populations
The research also stands out for its breadth. It included people of European, African, and Latin American ancestry—a rare feat in genetic studies of dementia, which have historically focused on white populations.
In all groups, alcohol use disorder was associated with a higher risk of dementia, although the strength of the signal varied due to smaller sample sizes outside of European ancestry. The authors also ran non-linear genetic models, asking whether there might be a “safe zone” for drinking but found none. Instead, the risk of dementia increased steadily with greater alcohol intake, even at low levels.
However, not all experts are ready to declare alcohol a definitive cause of dementia.
“Neither part of the study can conclusively prove that alcohol use directly causes dementia,” said Professor Tara Spires-Jones of the University of Edinburgh, in commentary for the Science Media Centre. “But this adds to a large amount of similar data showing associations between alcohol intake and increased dementia risk.”
Spires-Jones added: “Fundamental neuroscience work has shown that alcohol is directly toxic to neurons in the brain.”
Yet others urged caution in how the findings are interpreted.
“The authors say that ‘genetic analysis showed a monotonic increasing dementia risk with increased alcohol intake.’ This is untrue,” said Professor Sir David Spiegelhalter, a statistician at Cambridge, as per Science Media Centre. “The relationship is with genetically predicted alcohol intake… and those genetic predictions rely on many unverifiable assumptions.”
Despite these caveats, the study is among the most comprehensive ever conducted on the topic.
No Safe Sip
The new results arrive amid a broader scientific reassessment of alcohol’s role in health. Studies now link even moderate drinking to increased cancer risk, particularly for breast and digestive system cancers. Recent research also shows higher death rates from cardiovascular disease among older adults who drink regularly.
“There is a mythology about alcohol having positive benefits,” said Stanford’s Dr. Stafford. “We don’t have strong evidence of any health benefit from moderate drinking, but we do have strong evidence of harm.”
For people of East Asian descent who carry a common genetic mutation (ALDH2 deficiency), the risks may be even higher. Just one drink can produce toxic levels of acetaldehyde—a compound that damages DNA and raises cancer risk.
Given these concerns, some countries are revising their guidelines. Canada now advises no more than two drinks per week. The World Health Organization has declared that “no level of alcohol consumption is safe.”
This study doesn’t say you must quit drinking. But it does suggest that drinking for brain health is a mistake. If you’re weighing the risks, here’s the key takeaway: the safest level of alcohol for the brain may be none at all.
“Reducing alcohol consumption across the population could play a significant role in dementia prevention,” said Dr. Topiwala.
Public health officials may soon need to rethink how they communicate alcohol’s risks. For decades, the idea of the “healthy drinker” has shaped social norms and medical advice alike.
This new study may finally put that narrative to rest.
