To determine whether low to moderate alcohol intake is protective against cognitive decline in older people.DESIGN
Prospective cohort study.SETTING
Community-based study in Ireland, the Netherlands, and Scotland.PARTICIPANTS
Five thousand eight hundred four people (3,000 women) aged 70 to 82 and randomized to pravastatin or placebo in the Prospective Study of Pravastatin in the Elderly at Risk.MEASUREMENTS
Alcohol consumption was determined at study baseline. Serial measures of cognitive function over 3.2 years mean follow-up included Mini-Mental-State-Examination (MMSE), speed of information processing (Stroop and Letter-Digit Coding tests), and immediate and delayed memory (Picture-Word Learning test).RESULTS
Forty-two percent of women and 71% of men were alcohol drinkers. Cognitive performance was better for female drinkers than nondrinkers for all cognitive domains over the 3.2-year follow-up; no significant effects were seen for men (linear mixed model, including adjusting for possible confounders). The rate of cognitive decline was similar for drinkers and nondrinkers for all cognitive domains, except for MMSE, which declined significantly less in female drinkers than nondrinkers (linear mixed model attenuated rate of decline=0.05 MMSE units per annum, P=.001).CONCLUSION
Drinking low to moderate amounts of alcohol may delay age-associated cognitive decline in older women (including slowing deterioration in global cognitive function), but these apparent benefits were not clearly seen in older men.