Bidirectional associations between word memory and one-legged balance performance in mid and later life.

Abstract Background Age-related changes in cognitive and balance capabilities are well-established, as is their correlation with one another. Given limited evidence regarding the directionality of associations, we aimed to explore the direction and potential explanations of associations between word memory and one-legged balance performance in mid-later life. Methods A total of 3062 participants in the Medical Research Council National Survey of Health and Development, a British birth cohort study, were included. One-legged balance times (eyes closed) were measured at ages 53, 60–64 and 69 years. Word memory was assessed at ages 43, 53, 60–64 and 69 with three 15-item word-recall trials. Autoregressive cross-lagged and dual change score models assessed bidirectional associations between word memory and balance. Random-effects models quantified the extent to which these associations were explained by adjustment for anthropometric, socioeconomic, behavioural and health status indicators. Results Autoregressive cross-lagged and dual change score models suggested a unidirectional association between word memory and subsequent balance performance. In a sex-adjusted random-effects model, 1 standard deviation increase in word memory was associated with 9% (7,12%) higher balance performance at age 53. This association decreased with age (−0.4% /year (−0.6,-0.1%). Education partially attenuated the association, although it remained in the fully-adjusted model (3% (0.1,6%)). Conclusions There was consistent evidence that word memory is associated with subsequent balance performance but no evidence of the reverse association. Cognitive processing plays an important role in the balance process, with educational attainment providing some contribution. These findings have important implications for understanding cognitive-motor associations and for interventions aimed at improving cognitive and physical capability in the ageing population.