Background: Low health literacy and low cognitive abilities both predict mortality, but no study has jointly examined these relationships.
Methods: We conducted a prospective cohort study of 3,260 community-dwelling adults age 65 and older. Participants were interviewed in 1997 and administered the Short Test of Functional Health Literacy in Adults and the Mini Mental Status Examination. Mortality was determined using the National Death Index through 2003.
Measurements and main results: In multivariate models with only literacy (not cognition), the adjusted hazard ratio was 1.50 (95% confidence of interval [CI] 1.24-1.81) for inadequate versus adequate literacy. In multivariate models without literacy, delayed recall of 3 items and the ability to serial subtract numbers were associated with higher mortality (e.g., adjusted hazard ratios [AHR] 1.74 [95% CI 1.30-2.34] for recall of zero versus 3 items, and 1.32 [95% CI 1.09-1.60] for 0-2 vs 5 correct subtractions). In multivariate analysis with both literacy and cognition, the AHRs for the cognition items were similar, but the AHR for inadequate literacy decreased to 1.27 (95% CI 1.03 - 1.57).
Conclusions: Both health literacy and cognitive abilities independently predict mortality. Interventions to improve patient knowledge and self-management skills should consider both the reading level and cognitive demands of the materials.