Dual-task walking performance : relationship to stroke characteristics

Abstract

Background: Previous work suggested that outdoor walking is one of the top concerns among community-dwelling stroke individuals. And most outdoor mobility activities involve dual-tasking. When a cognitive task was imposed during walking, there may be degradation of performance of the walking or/and the cognitive task, in a phenomenon called dual-task interference. How the extent and pattern of dual-task interference is influenced by component task complexity and stroke characteristics remain understudied. Objective: To examine (1) how complexity of the component tasks influence dual-task cognitive and mobility performance in individuals with chronic stroke; (2) the association between dual-task performance and stroke characteristics (location of lesion, severity); and (3) the association between dual-task performance and satisfaction with community reintegration. Study design: This was a cross-sectional study. Individuals with chronic stroke were tested on various combinations of dual-task conditions during walking. Main outcome measure: Participants were classified to two groups: cortical involved stroke and subcortical stroke based on their CT or MRI reports. The severity of cognitive deficit was measured by Montreal Cognitive Assessment (MoCA) and Wisconsin Card Sorting Test (WCST). Mini-Balance Evaluation System Test (Mini-BESTest) and Fugl-Meyer Assessment (FMA) were used to test the balance and motor control deficits. The Reintegration to Normal Living Index (RNLI) was used to quantify the degree of satisfaction with community reintegration after stroke. The dual task protocol used in this study involved a combination of the mobility task and cognitive task. The former had two different complexity levels [low: walking on level ground (LGW) for 1 minute vs. high: obstacle crossing walking (OBW) for 1 minute]. Four aspects of gait performance were measured: velocity (distance, stride length, stride time), variability (stride length and stride time variability), asymmetry (stride velocity asymmetry) and postural stability (peak frontal trunk velocity). The cognitive component task used in the testing protocol was a serial subtraction task which also had 2 levels of complexity [low: serial subtraction by three (SS3) vs. high: serial subtraction by seven (SS7)]. Results: Eighty participants [44 men; mean (SD) age: 62.2 (6.5)] were included in the final analysis, with 27 cortical involved stroke and 53 pure subcortical stroke individuals. The cognitive performance, and velocity related gait parameters (walking distance, stride time, stride length) under DT conditions deteriorated significantly when comparing with the respective values in the single-task condition (p<0.01). On the other hand, compared with single-task walking, better postural stability (i.e., smaller peak frontal trunk velocity) was observed under DT conditions (p<0.01). Also, the increased difficulty level of the mobility task (level ground walking vs obstacle crossing) among DT conditions did not change the cognitive performance significantly. Likewise, the increased complexity level of the cognitive task (SS3 vs. SS7) also did not impact the gait performance significantly. Overall, there was no significant difference in DT gait and cognitive performance between cortical involved stroke and subcortical stroke group (p>0.05). Negative associations were found between stride length during DT walking and perseverative errors (%) on the WCST (p<0.05). Lower MoCA scores were significantly associated with poorer DT cognitive performance as measured by the correct response rate (NCR). Lower Mini-BESTest and FMA scores were associated with poorer DT gait performance. Finally, poorer DT performance was associated with lower RLNI scores. Conclusion: Significant dual-task interference occurred in individuals with chronic stroke, when a serial subtraction task was imposed during walking, regardless of the difficulty level of the component tasks used. Those who have more severe motor and cognitive deficits tended to have poorer DT performance, which in turn was related to lower level of satisfaction with community reintegration.