Research explained: modeling study with narrative review

No protective effect at low drinking levels: a US modeling study puts the lifetime risk of an alcohol-caused death at 1 in 25 for men at 14 drinks a week

The Alcohol Intake and Health Study combined national survey, census, mortality and morbidity data. Risks rose from relatively low levels, and the authors back a limit of no more than 1 drink a day for men and women.

Drinking alcohol showed no net benefit for health at any low level of intake in a US analysis published in the July 2026 issue of the Journal of Studies on Alcohol and Drugs. For men drinking 14 drinks a week, the upper limit for men in earlier US Dietary Guidelines, the estimated lifetime risk of dying from an alcohol-caused condition was 1 in 25 (4%). The study drew renewed news coverage in recent weeks.

What the researchers did

A team led by researchers at the Centre for Addiction and Mental Health in Toronto, with coauthors from Columbia, Harvard, UCL and other institutions, estimated lifetime risks of alcohol-attributable death and disease in the United States. They combined drinking data from national health surveys with published relative risks, US Census population figures, CDC mortality data and morbidity data from the Institute for Health Metrics and Evaluation, cause by cause. A separate narrative review looked at how much is drunk on a single occasion.

Risk climbed from fairly low amounts. Men drinking more than 6.5 drinks a week and women drinking more than 7.0 had a lifetime risk of an alcohol-caused death above 1 in 1,000. Above 8.5 drinks a week, for both sexes, that risk passed 1 in 100. Drinking more than 1 drink per occasion was linked to steadily higher risks of breast cancer, cardiovascular disease and injury.

How strong is the evidence?

This is a modeling study, not a trial or a new cohort. Its results depend on the survey data and the relative risks it draws on, which come mostly from observational research with self-reported drinking. The confidence intervals around the thresholds are wide. The per-occasion part is a narrative review rather than a systematic one. The authors declare no competing interests, and the abstract does not state funding.

What it means for you

For people who drink, less is better than more, and spreading the same amount over fewer drinks per occasion matters too. The authors suggest no more than 1 drink a day for both men and women. If you enjoy a beer or a glass of red wine or white wine, alcohol-free days help, and a cup of green tea or other options in drinks can take the place of some of them.

Sources

  1. George S, Naimi TS, Keyes K, et al. Alcohol Intake and Health Study: No Protective Effect at Low Levels, With Mortality Increasing to 1 in 25 at 14 Drinks Per Week. J Stud Alcohol Drugs. 2026 Jul;87(4):621-638. doi:10.15288/jsad.25-00435. PMID 42420014.