The effect of cognitive dual-tasks on dynamic postural sway during gait using inertial measurement units

Objective: To investigate the influence of cognitive dual-tasks on dynamic postural sway variability, measured with a single inertial measurement unit, in healthy collegiate athletes during dual-task gait. Approach: Thirty-four athletes ambulated down and back a 10-meter course. Root mean square values of postural sway acceleration were calculated for four sway-planes. Single and dual-task trial times were calculated from gait initiation to termination. Single and dual-task root mean square values were compared using Wilcoxon signed rank tests for each sway plane to determine significance. The relative difference between each sway measures and trial times for single and dual-task gait (i.e. dual-task cost was calculated). Spearman's correlation was used to analyze the association between dual-task cost of sway and trial time data. Main results: During dual-task conditions, participants ambulated slower and demonstrated a significant reduction in root mean square sway in all planes of motion. The greatest decreases were in the medial-lateral (Z = -4.83, p < 0.001, ES = -0.83) and superior-inferior planes (Z = -4.93, p < 0.001, ES = -0.85). Increased trial time cost was associated with a decrease in dual-task cost of postural sway (p < 0.05). Significance. Performing a cognitive task while ambulating decreased postural sway variability in task-relevant planes of motion compared to single-task conditions. Reduced dual-task cost of sway variability during gait was associated with maintained trial times between single and dual-tasks. These data support the use of instrumented dual-task gait assessment to evaluate postural sway variability.