Moderate Drinking and a Longer Life

To examine the relationship between alcohol consumption and mortality among older adults, researchers conducted a 23-year prospective study of approximately 12,000 male British doctors aged 48–78 years. Mean daily alcohol intake was about 1.5–2 standard drinks. During follow-up, 7364 deaths occurred. Analyses were adjusted for age, smoking, and follow-up duration.

• Compared with nondrinkers (including ex-drinkers), current drinkers had significantly lower risks of death from ischemic heart disease (relative risk [RR] 0.7), other vascular diseases (RR 0.8), and respiratory disease (RR 0.8). Because these causes accounted for most deaths, current drinkers also had a lower risk of total mortality (RR 0.8).
• The relationship between alcohol consumption and mortality was U-shaped (i.e., nondrinkers and heavier drinkers had the highest rates) for all causes except alcohol-augmentable causes (i.e., those increased by heaving drinking, such as liver disease). Risk of death from alcohol-augmentable causes increased as consumption increased from <5 to >20 drinks per week.
• When recent ex-drinkers were grouped with current drinkers, the association between alcohol consumption and total mortality was attenuated (RR 0.9).

Comments:

This study from a leading epidemiologist confirms the relationship between moderate drinking and lowered risk of total and heart disease mortality. However, its finding of decreased risk of respiratory disease mortality is surprising. An editorialist notes that the main results may be due partially to a greater risk of death among ex-drinkers (e.g., those who stopped drinking because of illness) and/or insufficient adjustment for confounding factors.

R. Curtis Ellison, MD

Reference:

Doll R, Peto R, Boreham J, et al. Mortality in relation to alcohol consumption: a prospective study among male British doctors. Int J Epidemiol. 2005;34(1):199–204.

Wannamethee SG. Commentary. Alcohol and mortality: diminishing returns for benefits of alcohol. Int J Epidemiol. 2005;34(1):205–206.