Cognition Predicts Mobility Change in Lower Extremity Amputees Between Discharge From Rehabilitation and 4-Month Follow-up: A Prospective Cohort Study.

Abstract

To assess (1) the effect of task (single and dual task), time (discharge and 4mo), and their interaction for mobility; (2) task prioritization during dual-task testing; and (3) the association between cognition on change in mobility between discharge from rehabilitation and 4 months' follow-up. Prospective cohort study. Rehabilitation hospital. People with lower extremity amputations (N=22) were consecutively recruited at discharge from an inpatient prosthetic rehabilitation program. Not applicable. Gait velocity and the L Test of Functional Mobility, single and dual task (serial subtractions by 3), were the primary outcomes. Montreal Cognitive Assessment and Trail Making Test quantified cognition as secondary outcomes. Repeated measures analysis of variance evaluated the effects of task (single task and dual task) and time (at discharge and 4 months' follow-up) and their interaction on each outcome. A performance-resource operating characteristic graph evaluated gait and cognitive task prioritization. Multivariable linear regression evaluated the association between cognition and change in mobility over time. No significant interactions between task and time were found (all P>.121) for L Test and gait velocity. The L Test single task (P=.001) and dual task (P=.004) improved over time. Gait velocity improved over time for both single task and dual task (P<.001). Dual-task performance was slower than single-task performances at each time point. The Trail Making Test B was independently associated with the change in dual-task L Test (P=.012), and single-task (P=.003) and dual-task (P=.006) gait velocity at follow-up. Gait velocity and L Test single and dual task improved over time. No significant interactions indicated that cognitive task did not differentially affect performance over time. Lower executive function scores at discharge were independently associated with lower gains in all gait velocity and dual-task L Test outcomes at follow-up.