The effects of attention resources and dual-task strategies on gait performance in aging: A comparison of two walking while talking paradigms

Abstract

Objective: Using non-demented community residing older adults, we aimed to (1) examine dual-task costs in gait and cognitive performance in two walking while talking tasks: walking while reciting alternate letters of the alphabet (WWR) and walking while counting backward by sevens (WWC), and (2) assess dual-task strategies by examining interactions between cognitive performance and dual-task costs in gait. Measurements: Gait and cognitive performances were tested in 217 healthy older adults (mean age 75.7 + 8.78, 56.2% female) under single and dual-task conditions. Quantitative gait measures including velocity, swing time, and stride length variability were obtained using an instrumented walkway. Cognitive performance was assessed by calculating accuracy ratio: [number of correct responses] / [number of total responses]. Results: Linear mixed effects models revealed significant dual-task costs including slower velocity (p < .001), longer swing time (p < .001), greater stride length variability (p < .001), and decreased accuracy ratio (< .001) in WWR and WWC compared to the single task conditions. Greater dual-task costs in velocity (p < .001) and stride length variability (p = .001) were observed in WWC compared to WWR. In the context of the linear mixed effects models significant interactions were observed between cognitive performance (predictor) and dual-task costs in gait (outcome measure) only in WWR (velocity, p = .010; swing time, p = .026; stride length variability, p = .002). Visual depictions of these interactions revealed that decline in cognitive performance were associated with greater dual-task costs in gait performance. Conclusion: Dual-task performance costs in walking while talking paradigms are attributed to limited attentional resources but not to dual-task strategies where older participants show a preference to either the gait or cognitive tasks.