The objective of this study is to investigate psychometric properties of dual-task walking assessments and compare effects of multi-task walking interventions on walking for patients with chronic stroke. Specifically, we will investigate psychometric properties (i.e. reliability, validity, and responsiveness) of dual-task walking assessments with the Stroop task for chronic stroke individuals (Aim 1). The second aim of this study is to compare the effects of multi-task walking training mode to traditional rehabilitation in patients with chronic stroke (Aim 2). The third aim of this study is to compare the immediate, retained, and transfer effect of multi-task overground walking training to multi-task treadmill walking training on walking function, cognitive performance, task coordination, and community ambulation in patients with chronic stroke (Aim 3).

Only 7% patients can walk safely in the community. Functional community ambulation requires the ability of walking while performing other tasks. It is emergent to establish dual-task walking assessment with good psychometric properties and to identify the most effective approach that enhances dual-task walking performance for stroke patients. Our ongoing study found that motor tasks combined with cognitive dual-task training (multi-tasking ) have more improvements on cognitive performance of the Stroop task while walking than cognitive dual-task training alone. Therefore, combined motor and cognitive elements to walking training might be considered as a novel therapeutic strategy to promote functional ambulation and recovery of stroke patients. This combined strategy is based on attentional capacity sharing theory and the research advances on dual-task to enhance active participation and cognitive involvement, strengthen dual-task walking ability, then transfer to promote community ambulation and participation. However, lacking prospective, controlled trials quantify gait and cognition under dual-task conditions after multi-task walking training. Specifically, we will investigate psychometric properties of dual-task walking assessments for chronic stroke individuals. The second aim is to compare the immediate, retained, and transfer effects of of multi-task overground walking training to multi-task treadmill walking and traditional rehabilitation on walking function, cognitive performance, task coordination, and community ambulation in patients with chronic stroke. A metric analysis and comparative efficacy research will be conducted. Sixty chronic stroke patients will receive dual-task walking assessments twice at pretreatment with a 1-week interval for test-retest assessment and investigation of the reliability and validity of outcome measures. The primary outcome measure of the dual-task walking assessments will include walking at preferred speed and fast speed and simultaneously perform the Stroop task. Concurrent validity will be studied to validate the dual-task walking measures with each other and with the item 14 of the mini-Balance Evaluation Systems test (Mini-BESTest), dual-task Timed-up-and-Go test (dual-TUG), and 6-min walk test obtain concurrently for assessing dual-task ability. In addition, we will compare dual-task walking performance between fallers and non-fallers to examine discriminant validity of dual-task walking assessments. A comparative efficacy research is a single-blind, randomized controlled trial, which will be conducted at medical centers. Sixty ambulatory stroke patients will be randomized to multi-task overground walking training or multi-task treadmill walking training or traditional rehabilitation. All three groups will receive interventions 3 times a week for 4 weeks. The multi-task overground walking training group will undertake overground walking training while concurrently perform motor and cognitive tasks. The multi-task treadmill walking training group will train the same set of motor and cognitive tasks while walking on the treadmill. Traditional rehabilitation will train strength, balance, and gait. A blinded assessor will administer three assessments All participants will be examined gait and cognitive performance under single-task (walking only, cognitive tasks only) and dual-task conditions (walking while performing the Stroop task) at baseline, post intervention, and 1-month follow-up. The primary outcome measure of gait and cognition is gait speed and composite score of accuracy and reaction time of the cognitive tasks under single- and dual-task conditions. The secondary outcome measures will be the Mini-BESTest, single- and dual-TUG, Functional Gait Assessment, 6-minute Walk Test, physical activity monitor, and Stroke Impact Scale. Repeated measure ANOVA will be used to compare measurements at baseline, after training, and follow-up among the groups.

Inclusion Criteria:

1. first-ever stroke with onset duration more than 3 months
2. able to walk 10 m
3. no severe vision, hearing, and language problems.

Exclusion Criteria:

1. orthopedic and other neurological disorders that affect walking
2. other treatments that could influence the effects of the interventions (e.g., recent Botulin toxin treatment of the lower extremity)
3. moderate or severe cognitive impairments (score < 24 on Mini-Mental State Examination)
4. severe uncorrected visual deficits.