The relative temporal sequence of decline in mobility and cognition among initially unimpaired older adults: Results from the Baltimore longitudinal study of aging

Abstract

Background: most older individuals who experience mobility decline, also show cognitive decline, but whether cognitive decline precedes or follows mobility limitation is not well understood. Objective: examine the temporal sequence of mobility and cognition among initially unimpaired older adults. Methods: mobility and cognition were assessed every 2 years for 6 years in 412 participants aged ≥ 60 with initially unimpaired cognition and gait speed. Using autoregressive models, accounting for the dependent variable from the prior assessment, baseline age, sex, body mass index and education, we examine the temporal sequence of change in mobility (6 m usual gait speed, 400 m fast walk time) and executive function (visuoperceptual speed: Digit Symbol Substitution Test (DSST); cognitive flexibility: Trail Making Test part B (TMT-B)) or memory (California Verbal Learning Test (CVLT) immediate, short-delay, long-delay). Results: there was a bidirectional relationship over time between slower usual gait speed and both poorer DSST and TMTB scores (Bonferroni-corrected P < 0.005). In contrast, slower 400 m fast walk time predicted subsequent poorer DSST, TMT-B, CVLT immediate recall and CVLT short-delay scores (P < 0.005), while these measures did not predict subsequent 400 m fast walk time (P > 0.005). Conclusions: among initially unimpaired older adults, the temporal relationship between usual gait speed and executive function is bidirectional, with each predicting change in the other, while poor fast walking performance predicts future executive function and memory changes but not vice versa. Challenging tasks like the 400 m walk appear superior to usual gait speed for predicting executive function and memory change in unimpaired older adults.