Prospective study of moderate alcohol consumption and mortality in US male physicians

Arch Intern Med. 1997 Jan 13;157(1):79-85.


Background: Although moderate alcohol consumption decreases the risk of myocardial infarction, its impact on all-cause mortality among apparently healthy men is unclear.

Methods: We performed a prospective cohort study in 22071 men in the Physicians' Health Study. Participants were aged 40 to 84 years and had no history of myocardial infarction, stroke, transient ischemic attack, or cancer.

Results: There were 1206 deaths (394 cardiovascular, 488 cancer, and 324 other) during 10.7 years of follow-up. Compared with participants who consumed less than 1 drink per week, the relative risk (95% confidence interval) of all-cause mortality for men who consumed 2 to 4 drinks per week was 0.72 (0.59-0.87); 5 to 6 drinks per week, 0.79 (0.64-0.99); 1 drink per day, 0.98 (0.84-1.15); and the highest drinking group (> or = 2 drinks per day), 1.51 (1.17-1.95). This association was similar with either nondrinkers or occasional drinkers used as the reference group and was not subject to material confounding or effect modification by any factor examined. The overall relationship was the result of a J-shaped association with cardiovascular mortality, an increase in cancer deaths for the highest drinking group, and a U-shaped association with other causes of mortality.

Conclusions: Risk of all-cause mortality varies by level of alcohol consumption. In this apparently healthy cohort, men who consumed 2 to 6 drinks per week had the most favorable mortality profile and men who had 2 or more drinks per day the most unfavorable mortality profile. The difference between consumption of small and large amounts of alcohol may mean the difference between preventing and causing excess mortality.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Alcohol Drinking / mortality*
  • Cause of Death
  • Humans
  • Male
  • Middle Aged
  • Physicians / statistics & numerical data*
  • Prospective Studies
  • Randomized Controlled Trials as Topic
  • Risk
  • Risk Factors
  • United States / epidemiology