Relation of smoking and low-to-moderate alcohol consumption to change in cognitive function: a longitudinal study in a defined community of older persons.

To determine whether smoking habits and alcohol consumption are related to changes in cognitive function, the authors conducted a prospective, community-based study of persons aged 65 years and over in East Boston, Massachusetts. In 1982 and again in 1985, the subjects were given three brief tests of cognitive function: immediate memory, digit span, and a mental status questionnaire, which primarily assessed orientation. The 1,201 individuals who performed well in 1982 were included in linear regression analyses of 3-year change in performance, adjusted for age, sex, education, and income. Relative to nonsmoking, current smoking, past smoking, and pack-years were not significantly related to change in immediate memory. None was significantly related to change in orientation. Only pack-years was significantly related to normal change score in digit span (normal change score change per unit of predictor = 0.001, 95% confidence interval 0.0003-0.002). Low-to-moderate alcohol consumption during the month preceding baseline testing was not significantly related to a subsequent 3-year change in performance in two of the three tests. However, people who consumed a very small amount of alcohol had a normal change score that was 0.088 (95% confidence interval 0.015-0.160) better for digit span than did nondrinkers. This study provides evidence that the reported levels of smoking and alcohol use among older persons are not consistent or substantial predictors of the longitudinal change in cognitive function observed in a community.