The objective of this study is to investigate the effect of dual-task training at impairment, disability, and participation levels for stroke survivors and investigate possible factors affecting cognitive-motor interference (CMI) under dual-task conditions. Specifically, we will compare the immediate and retention effects of dual-task balance training and single-task balance training on CMI, balance confidence, fall rate, functional performance, and quality of life in individuals with stroke (Aim 1). The second aim of this study is to explore how lateralization, stroke chronicity, task type, and task difficulty may influence patterns of CMI (Aim 2). The third aim of this study is to investigate whether dual-task balance assessments are more sensitive than single-balance assessment in distinguishing stroke fallers from stroke non-fallers (Aim 3).

Gait-related dual-task impairments are common problems experienced in people with stroke by walking more slowly or taking shorter steps in dual-task walking. Decrements in dual-task capacity may predispose stroke survivors to risk of falls and affect negatively on functional independence and quality of life. Currently there is a need to explore whether people with stroke practice dual-tasking in standing and walking will lead to enhance dual-task performance and reduce falls rate. However, there is a lack of prospective, controlled studies which quantify balance, gait, cognition, and fall-risk after dual-task training. The objective of this study is to investigate the effect of dual-task training at impairment, disability, and participation levels for stroke survivors and investigate possible factors affecting cognitive-motor interference (CMI) under dual-task conditions. Specifically, we will compare the immediate and retention effects of dual-task balance training and single-task balance training on CMI, balance confidence, fall rate, functional performance, and quality of life in individuals with stroke (Aim 1). The second aim of this study is to explore how lateralization, stroke chronicity, task type, and task difficulty may influence patterns of CMI (Aim 2). The third aim of this study is to investigate whether dual-task balance assessments are more sensitive than single-balance assessment in distinguishing stroke fallers from stroke non-fallers (Aim 3).

A prospective randomized control trial will be conducted at 2 medical centers in this study. Forty-eight people with stroke will be randomly allocated to either a single- or dual-task balance training group. Both groups will receive 12 hours of training over 4 weeks at progressively increasing task difficulty (60 minutes per session, three times a week, for four weeks). The single-task training group will undertake balance training and gait training. The dual-task training group will train the same set of balance activities as the single-task training while simultaneously performing a variety of added cognitive or motor tasks with instruction of a variable priority. During each session, participants in dual-task training practice shifting attention between tasks by spending half of the training attention focus on the balance task, and half focus on the secondary cognitive or motor task. A blinded assessor will conduct three assessments at baseline, post intervention, and 1-month follow-up. At baseline, all participants will be examine balance, gait, and cognitive performance under single-task (standing/walking at preferred speed and fast speed only, cognitive task only) and dual-task conditions (standing/walking at preferred speed and fast speed and concurrent cognitive task). Three cognitive tasks (simple auditory reaction time task, counting backward by 3s task, and Stroop task) will be used to examine patterns of CMI because these tasks represent different domains of cognitive function (information processing speed, working memory, and executive function). The primary outcome measure of cognition will be the composite score of accuracy and reaction time. The primary outcome measure of balance will be the sway index in mCTSIB test on a BioSway balance system under single- and dual-task standing. The primary outcome measure of gait will be gait speed on 10 m walk test under single- and dual-task walking. Secondary outcome measures will include: Berg Balance Scale, the Timed Up and Go test under single- and dual-task conditions, Activity-specific Balance Confidence Scale, Fugl-Meyer Assessment, Functional Independence Measure, Nottingham Extended Activities of Daily Living Scale, and Stroke Impact Scale. Additionally, falls will be prospectively monitored monthly and recorded during the following 6 months after the training. Repeated measure ANOVA will be used to compare measurements baseline, after training, and follow-up between the groups.

Inclusion Criteria:

• first-ever stroke with onset duration more than 3 months
• able to walk with or without walking aids
• self-selected gait speed at least 0.6 m per second determined during a 10m walk test
• no severe vision, hearing, and language problems
• have experienced at least one fall in the previous year

Exclusion Criteria:

• have orthopedic condition affecting their gait or other diseases that might interfere with participation in the training program
• could not stand for at least two minutes without assistance
• score less than 24 on the Mini-Mental State Examination