Tighter alcohol guidelines? No net benefit found

Clinical takeaway: The lowest-risk alcohol intake is no more than one drink a day with food, rather than fewer, heavier occasions. Measurable mortality risk emerges around seven drinks a week and climbs steeply above that.

When a US federal panel set out to update the 2025–2030 Dietary Guidelines, it commissioned a study to settle a simple question the current guidelines dodge: how much alcohol is too much? The findings undercut prior decades of reassurance: no level of drinking carries a net health benefit.

The current 2020–2025 guidelines tell Americans to "limit" alcohol but set no weekly ceiling, advising men to cap intake at two drinks a day and women at one on days they drink. The new study argues that framing misses cumulative weekly risk, and that the two-drink allowance for men sits well above where harm becomes measurable.

For men drinking that upper limit of 14 standard drinks a week, the modeled lifetime risk of an alcohol-attributable death was about one in 25. Risk crossed one in 1,000 at roughly seven drinks a week for both sexes and rose to one in 100 above 8.5 drinks a week.

Small protective associations appeared below three drinks a week but none reached statistical significance. For years, observational data suggested light drinking lowered cardiovascular risk, producing the familiar J-shaped curve. Here, low-level intake still tracked with reduced ischemic heart disease and stroke, and with lower diabetes risk in women. But once cancer, liver disease, and injury were added across the full range of outcomes, the net effect was harmful.

Drinking concentrated into heavy occasions offset or reversed the protective associations seen with steady low-level intake even at the same weekly volume. Women faced sharply higher liver-related mortality at higher intake, and adults under 40 saw no net protective effect at any level with most of their alcohol-attributable deaths coming from injuries.

Researchers selected 16 systematic reviews and meta-analyses through an expert ranking process and applied those relative risks to US mortality, census, and survey data, using lifetime abstainers as the reference group. Drinking estimates were adjusted upward to match alcohol sales, since surveys capture only 40% to 60% of what is sold. The analysis also left beverage type, frequency, and speed of consumption unexamined, while social context was outside its scope entirely.

For clinicians, the message is less a bright line than a slope. Risk rises with volume and with per-occasion intensity, and the old two-drinks-a-day ceiling for men sits well into the elevated range. The authors translate that into a single benchmark for counseling: one drink a day or less.

"It turns out that two drinks per day, which might be considered 'moderate' from a social standpoint, is associated with a substantially elevated risk of a premature death caused by alcohol," said study co-author Timothy Naimi, MD, MPH, director of the University of Victoria's Canadian Institute for Substance Use Research and an adjunct professor at Boston University.

Source: George S. J Stud Alcohol Drugs. 2026 Jul. Alcohol Intake and Health Study: No protective effect at low levels, with mortality increasing to 1 in 25 at 14 drinks per week