Abstract

Background: Cognitive capabilities in childhood and in late life are inversely associated with mortality rates. However, it is unclear if adult cognition, at a time still relatively free from comorbidity, is associated with subsequent mortality, and whether this explains the associations of early life factors with adult mortality. Methods: We used data from the MRC National Survey of Health and Development, a birth cohort study prospectively assessing 5362 participants born in 1946. The present analysis includes participants followed up from age 43 and undergoing cognitive assessment (verbal memory and search speed). Mortality outcomes were notified through linkage with a national register. Cox regression was used to estimate mortality hazards in relation to cognitive performance at age 43, adjusting for early life factors, socioeconomic position and health status. Results: Data were available on 3192 individuals. Univariable analyses indicated that adult verbal memory and search speed, parental factors, childhood cognition and educational attainment were associated with mortality. However, multivariable models showed that the mortality associations with earlier life factors were explained by adult cognitive capability. A standard deviation increase in verbal memory and search speed scores was associated with lower mortality rates [hazard ratio (HR) = 0.86, 95% confidence interval (CI) 0.77-0.97, P = 0.02; HR = 0.88, 95% CI 0.78-1.00, P = 0.05, respectively), after adjustment for adult health. Conclusions: Cognitive capability in early midlife was inversely associated with mortality rates over 25 years and accounted for the associations of family background, childhood cognitive ability and educational attainment with mortality. These findings, in a nationally representative cohort with long-term follow-up, suggest that building cognitive reserve may improve later life health and survival chances.