Associations of alcohol use in early and middle adulthood with mid- and late-life cognition - a synthetic cohort approach.

OBJECTIVE

Using two cohorts and synthetic datasets, we estimated effects of prospectively reported alcohol use on memory outcomes across middle age.

METHODS

Data were from National Longitudinal Study of Youth 1979 (NLSY79, n=7540, alcohol reports from ages 18-26), Health and Retirement Study (HRS age 50-56 at enrollment, n=13,090), and a synthetic cohort matching early life exposure information from 3,259 NLSY79 participants to later life memory information from 5,451 HRS participants. Covariate-adjusted linear mixed models regressed memory (word list recall) on alcohol use (none, light/moderate, heavy).

RESULTS

In NLSY, we found no evidence that associations between light/moderate drinking in early adulthood and mid-life memory score significantly differed from associations between drinking abstention (β = -0.09 (95% CI: -0.30, 0.11)) or heavy drinking (β = -0.26 (-0.48, -0.04)) with memory score. In HRS, both abstaining from alcohol (β = -0.14 (-0.25, -0.02)) and heavy drinking (β = -0.25 (-0.42, -0.07)) were negatively associated with cognitive level. Results from the synthetic cohort mirrored NLSY, suggesting no significant association between abstention (β = 0.13 (-0.10,0.36)) nor heavy drinking (β = 0.02 (-0.25,0.28)) with mid-to-late life memory score.

DISCUSSION: Alcohol consumption may not have an effect on memory until later life, though associations may be affected by residual confounding.