Dual-Task Exercise Improves Walking Ability in Chronic Stroke: A Randomized Controlled Trial

Participants 

Subjects were recruited from community groups. The age, sex, paretic side, and onset time of hemiparesis were obtained from patient interviews and confirmed via medical records review. The inclusion criteria were (1) hemiparetic from a single stroke occurring at least a year earlier, (2) limited (gait velocity between 58 and 80cm/s) or full community ambulatory ability (minimum gait velocity of 80cm/s) by Perry et al’s classification system,6 (3) not presently receiving any rehabilitation services, (4) able to walk 10m independently without an assistive device, (5) functional use of the involved upper extremity, (6) stable medical condition to allow participation in the testing protocol and intervention, and (7) an ability to understand instructions and follow commands. The exclusion criteria were (1) patient with any comorbidity or disability other than stroke that would preclude gait training, (2) any uncontrolled health condition for which exercise is contraindicated, and (3) any neurologic or orthopedic diseases that might interfere with the study.

Study Protocol 

The study protocol was reviewed and approved by the institutional review board. Before data collection, the purposes and procedures were fully explained, and informed consent was obtained from each participant. Thirty-four subjects were identified as potential participants for this study. Seven were excluded because they failed to meet the inclusion criteria (fig 1). Two people did not sign an informed consent form. Twenty-five subjects signed an informed consent form before participating in the study. Participants were randomized to the control group or experimental group by an independent person who picked 1 of the sealed envelopes 30 minutes before the start of the intervention. All subjects were evaluated before the commencement of training (pretraining) and at the end of the 4-week training period (posttraining). Each subject was evaluated individually and tested following a standard protocol. Subjects were tested under the control single-task condition and under the dual-task condition. The 12 subjects in the control group did not receive any rehabilitation training. The remaining 13 subjects in the experimental group received ball exercise training.

View Large Image Download to PowerPoint

Fig 1. Flow diagram of the study.

Measurements 

Subjects were evaluated by a physical therapist who was not involved in the training program and did not know about the subject’s group assignment. For each subject, the gait performance was measured in 2 conditions: (1) preferred walking (single task) and (2) walking carrying a tray with glasses (tray-carrying task). The instructions for each test condition were as follows: (1) “Walk with your comfortable speed right to the end of the walkway” and (2) “Walk with your comfortable speed right to the end of the walkway carrying this tray and glasses in front of you with both hands.” For the tray-carrying task, subjects have to keep the empty glasses (height, 15cm; base diameter, 6cm) on the tray without any dropping otherwise the trial would be considered as failed. The trial order was randomized to compensate for the effects of practice and fatigue. To obtain representative samples, each test condition was repeated 3 times successfully, and the mean of the 3 successful trials was used for further data analysis.

Gait was measured by using GAITRite,a an instrumented walkway. The GAITRite system provided temporal (time) and spatial (distance) gait parameters via an electronic walkway connected to the serial port of a personal computer. The standard GAITRite walkway contained 6 sensor pads encapsulated in a roll-up carpet with an active area of 3.66m long and 0.61m wide. The validity and reliability of the GAITRite system has been well established.16, 17 Subjects were asked to walk at their comfortable speed without an assistive device through a 10-m hallway. The GAITRite walkway was placed in the middle of the 10-m hallway to eliminate the effect of initiating or stopping walking.

The temporospatial parameters recorded were as follows: speed (in cm/s), cadence (in steps/min), stride time (in seconds), stride length (in centimeters), affected single-limb support (percentage of gait cycle), and unaffected single-limb support (percentage of gait cycle). Single-limb support was used only to calculate the temporal symmetry index. The temporal symmetry index was calculated by using the following formula:

unaffected single limb support (% of gait cycle)/affected single limb support (% of gait cycle).18

Interventions 

Subjects in the experimental group participated in 30 minutes of a ball exercise program 3 times a week for 4 weeks. The training program was based on a dual-task concept; subjects walked while manipulating either 1 or 2 balls. The balls used in this study were therapy balls with 45-, 55-, 85-, and 95-cm diameters and a basketball. The training program included (1) walking while holding 1 or 2 balls on both hands, (2) walking to match the rhythm of bouncing 1 ball with 1 hand or both hands, (3) walking while holding 1 ball on 1 hand and concurrently bouncing another ball with the other hand, (4) walking in time while kicking a basketball (the basketball was put into a net, and the net was held by the subject) (fig 2), (5) walking while holding 1 ball and concurrently kicking another basketball within a net, (6) walking while bouncing 1 ball and concurrently kicking another basketball within a net, and (7) walking while reciprocally bouncing 1 ball with both hands. Variable practice for the walking condition involved walking forward, walking backward, walking on a circular route, and walking on an S-shaped route. The subject was challenged with increasingly difficult tasks.

View Large Image Download to PowerPoint

Fig 2. The starting position of kicking a basketball.

Data Analysis 

Information from all subjects was entered into a computerized database and analyzed by using the SPSS statistical package.b Descriptive statistics were calculated for the clinical characteristics of each group. To compare the baseline demographic characteristics and the pretraining variables between groups, independent-samples t tests were used for means and chi-square tests were used for frequencies. To elucidate the effect of training, the differences on all dependent variables between the pre- and posttraining phases within group were analyzed by 1-way multivariate analysis of variance (MANOVA). Difference scores were calculated for each subject by subtracting the pretraining data from the posttraining data. Mean difference scores and the standard deviation (SD) of these changes scores were calculated for each variable. MANOVA was used to determine differences of mean difference scores of each dependent variable between groups. A significance level of .05 was set for all analyses.

Study Limitations 

There are several limitations in this study. First, the major limitation of this study was a lack of follow-up. We do not know whether the subjects in the experimental group were able to maintain these changes. However, ball exercise has the potential to encourage exercise habits on a long-term basis. Second, the sample size used in this trial was small, which implies that caution should be exercised when interpreting the results. Third, precaution must be taken in generalizing the results because our results are based on a selected group of participants suffering from chronic stroke with relatively high motor function. Finally, the study was limited to subjects who volunteered, and, therefore, they were a self-selected group of willing and highly motivated people.