RESEARCH ARTICLES

A Cognitive-Balance Control Training Paradigm Using Wii

Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

Fit to Reduce Fall Risk in Chronic Stroke Survivors

nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

Savitha Subramaniam, PT, MS, Christina Wan-Ying Hui-Chan, PT, PhD, and Tanvi Bhatt, PT, PhD

Video Abstract available. See Video (Supplemental Digital Con-

Background and Purpose: The impaired ability to maintain balance
while performing higher-level cognitive tasks (cognitive-motor inter- tent 1, http://links.lww.com/JNPT/A80) for more insights from the
ference) significantly predisposes stroke survivors to risk of falls. We authors.
investigated adherence and intervention-related effects of gaming to Key words: balance, balance and stroke, cognition, dual task
improve balance control and decrease cognitive-motor interference
in stroke survivors. (JNPT 2014;38: 216–225)
Methods: Community-dwelling individuals with hemiparetic stroke
(N = 8) received balance control training using Wii Fit in conjunction INTRODUCTION
with cognitive training for approximately 110 min/d for 5 consecutive
days. Changes in balance and cognitive performance were evaluated
by the limits of stability test performed under single-task (ST) and
S troke is one of the major causes of adult disability, leading
to dependence in activities of daily living, with more than
800 000 incidences per year.1,2 A stroke event causes a number
dual-task (DT) conditions. The outcome measures from the limits of of impairments that contribute toward poor balance,3 leading
stability test included reaction time and movement velocity of the to a decreased functional status and increased disability.4 Forty
center of pressure. The cognitive performance was quantified by the percent to 70% of community-dwelling stroke survivors expe-
number of errors. The DT cost was computed for the balance and rience detrimental falls each year5 and tend to have 1.5 to 4
cognitive outcome measures using [(ST − DT)/ST × 100]. Adher- times higher risk of hip fracture than their healthy counter-
ence was assessed by change on the Intrinsic Motivation Inventory parts; with only less than 40% of those individuals regaining
scores postintervention. No commercial party having a direct finan- independent mobility. Falls thus not only affect activities of
cial interest in the research findings reported here has conferred orwill daily living but also hamper community reintegration.6
confer. Recent studies7 demonstrate that virtual reality (VR)
Results: Posttraining, reaction time cost in the forward direction im- rehabilitation, in comparison with conventional methods, pro-
proved from 31 ± 8.02 to ±8.7 ± 6.6. Similarly, movement velocity vides enhanced sensory feedback about movement character-
cost improved from 33.7 ± 12.3 to 11 ± 1. Cognitive cost also istics and improves both motor task learning and execution.8
decreased from 47.9 ± 13.9 to 20 ± 18.8. There were similar im- Virtual reality rehabilitation methods offer highly customiz-
provements in the backward direction for all the outcome measures. able, controllable, multimodal simulations that give the subject
Scores on the Intrinsic Motivation Inventory improved from 16.6 ± high levels of motivation and adherence and a strong sense of
1.3 to 23.5 ± 1.5. presence in the virtual environment.9,10 These methods could,
Discussion and Conclusions: The results demonstrate good adher- potentially, allow more engaging forms of interventions to
ence and evidence of clinical value of this high-intensity, short- be accessed under reduced supervision, along with increased
duration protocol for reducing cognitive-motor interference and im- functional recovery.11 Despite the advantages, because of high
proving balance control in stroke survivors. Future studies should costs, VR systems are still unavailable in many rehabilita-
examine the dose-response effects and long-term changes of such DT tion settings.12 Off-the-shelf, low-cost video gaming systems
training paradigm applied to improve fall efficacy. such as Wii Fit (Nintendo Co, Ltd. Kyoto, Japan) and Kinect
(Microsoft, Inc., Redmond, WA, U.S.A) provide the stroke
Department of Physical Therapy, University of Illinois at Chicago. survivor a similar environment and effectiveness in balance
No commercial party having a direct financial interest in the research findings rehabilitation compared with VR systems13 across various re-
reported here has conferred or will confer a benefit on the authors or on habilitation settings.9,14 For example, balance training using
any organization with which the authors are associated. Wii Fit has been shown to result in significant improvements
Supplemental digital content is available for this article. Direct URL citation in static and dynamic balance control, which corresponds with
appears in the printed text and is provided in the HTML and PDF versions
of this article on the journal’s Web site (www.jnpt.org). functional improvements.13
The authors declare no conflicts of interest. Although balance control rehabilitation is fundamental,
Correspondence: Tanvi Bhatt, PT, PhD, Department of Physical Therapy, recent findings indicate that impairments in cognitive func-
University of Illinois at Chicago, 1919, W Taylor St (M/C 898), Chicago, tion poststroke may interfere with community mobility and
IL 60612 (tbhatt6@uic.edu).
Copyright C 2014 Neurology Section, APTA.
reintegration. In persons with stroke performing a motor and
ISSN: 1557-0576/14/3804-0216 cognitive task concurrently, if both the tasks share the same
DOI: 10.1097/NPT.0000000000000056 attentional resources (due to an overlap in structural cortical

216 JNPT r Volume 38, October 2014

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 JNPT r Volume 38, October 2014
networks15 ) or if they are limited (due to age-related deteriora-
tion or pathologically compromised central nervous system16 ),
performance on either one or both of the tasks is reduced
(cognitive-motor interference). Because of the stroke-induced
cortical lesion, there is a decrease in capacity of process-
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
ing resources and the cognitive-motor interference is signifi-
cantly greater,17 especially while performing the physical task
concurrently with a cognitive task that challenges working
memory or executive abilities.18 These findings suggest that
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024
integrating cognitive-motor rehabilitation might be crucial in
addressing balance recovery among chronic stroke survivors.
Although dual-task (DT) training (training in perfor-
mance of motor task with a concurrent cognitive task) in vari-
ous populations such as older adults,19 persons with Parkinson
disease,18 and persons with traumatic brain injury21 has been
very promising in improving balance control, there is little re-
search on effects of such rehabilitation in stroke survivors. Re-
cent research demonstrates that rehabilitation strategies with
DT might produce an efficient integration of the 2 tasks22
and “automate” the performance of the primary motor task,23
providing resources to focus on other tasks.24,25
In this study, we examined a novel and cost-effective
cognitive-motor training paradigm, comprising Wii Fit gaming
performed in conjunction with various higher-order cognitive
tasks, to lay the foundation for a multidimensional treatment
paradigm. This could subsequently be translated into a home
therapy program, contributing to economical health care man-
agement. The purpose of this study was to investigate the
adherence and intervention-induced effects of such training in
improving intentional balance control under DT conditions.
We hypothesized that there would be a significant decrease, in
both balance and cognitive costs, when comparing post- with
pretraining values, along with significantly greater scores on
Intrinsic Motivation Inventory (IMI) postintervention.
METHODS
Subjects
Eight ambulatory adults with self-reported chronic
hemiparetic stroke were recruited after obtaining informed
consent, approved by institutional review board of the Univer-
sity of Illinois. Demographics of the stroke subjects are shown
in Table 1.
Subject Eligibility
Subjects with hemiparetic stroke (>6 months, without
any presence of aphasia) as confirmed by participants’ physi-
Table 1. Demographics and Stroke Characteristics of Study Participants
Affected
Age, y Sex (M/F) Side (L/R)
54 M L
38 F L
64 M R
54 M R
55 M R
28 F R
56 F R
65 F L
Abbreviations: F, female; H, hemorrhagic; I, ischemic; M, male; L, left; R, right; SOMC, Short Orientation-Memory-Concentration.

C 2014 Neurology Section, APTA
Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
Dual-Task Training Paradigm
cian were included. They were required to have the ability to
stand independently for at least 5 minutes without the use of
an assistive device and no incidence of falls in last 6 months.
Participants with cognitive deficits, as measured by the Short
Orientation-Memory-Concentration test of cognitive impair-
ment (ie, >8), were excluded.26 Short Orientation-Memory-
Concentration is also positively correlated with screening tests
for aphasia, suggesting that individuals with higher score on
the Short Orientation-Memory-Concentration test would show
worse language functioning.27 Subjects’ mean ± SD disabil-
ity status, quantified using the Modified Rankin Scale, ranged
from mild to moderate disability (2.62 ± 0.51) (Table 1). Three
of the subjects reported having slipped during their activities
of daily living but were able to regain balance with external
support. Participants with other neurological (eg, Parkinson
disease, vestibular deficits, peripheral neuropathy, or unstable
epilepsy) and musculoskeletal disorders were excluded. Indi-
viduals with cardiovascular disorders as assessed by resting
heart rate (>85% of age-predicted maximal) and resting oxy-
gen saturation (<95) were also excluded.
Protocol
Intervention
The cognitive-motor training consisted of VR balance
training (using the Nintendo Wii Fit console) performed in
conjunction with cognitive training for 5 consecutive days of
approximately 110 minutes per session. Training consisted of
4 balance board games performed in randomized order: Table
tilt, Tightrope, Soccer, and Balance bubble. These games were
to be played while performing cognitive tasks, which included
memory tasks such as word list generation, letter-number se-
quencing, and question-answer and memory recall games. In
the word list generation task, subjects were asked to recite as
many words as they can, from a given category, as quickly
as possible for a given duration (eg, participants were asked
to say as many names of “animals,” “fruits,” or words from
a given alphabet letter, such as “S”). In the letter-number se-
quencing task, subjects were given a paired letter and number
(eg, B2) and asked to recite subsequent letter-number pair-
ings in ascending or descending order (eg, C3, D4, E5). In
the question-answer task, subjects were asked to answer a set
of questions (eg, what is the date today, what is your favorite
season, what is the fastest mode of transportation for you to go
home, etc). In the memory recall game, subjects were period-
ically given a set of 3 simple words (eg, watch, summer, and
Stroke Ischemic/ Modified Rankin SOMC Test
Onset, y Hemorrhagic Scale (/6) (/28)
2 H 3 6
7 I 2 5
10 I 3 5
3 I 3 4
7 I 2 5
2 H 3 4
5 H 2 7
13 I 3 6
217
 Subramaniam et al JNPT r Volume 38, October 2014

book) to recite and which they were asked to recall after about
1 minute. During the initial session, participants underwent
a brief orientation on how to use the Wii Fit gaming system
and the objective of each game (Table tilt: getting the marble
in the hole; Tightrope: walk across a rope; Soccer: heading
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
function. This is a relatively complex task that requires main-
tenance, retrieval, and reorganization of information.36 In this
cognitive task, subjects were given predetermined numbers
and asked to perform serial subtraction (eg, subtract 7 from
40 as many times as you can until stopped). After 2 trials of

soccer balls; and Balance bubble: navigating through the maze familiarization, another 2 trials were conducted using 2 new
without popping the bubble), which a research assistant then number sets. These trials served as the ST cognitive condition.
demonstrated. Subjects learned the mechanisms of the gaming The number of accurate responses recited within the given
system with practice, verbal feedback from the therapist lead- time period of each trial was recorded.
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

ing the session, and visual feedback from the gaming system,
as seen on the television. Training began once they were able DT Condition
to demonstrate ability to play the game by achieving a score Under DT conditions, subjects performed the limits of
of more than zero on each game of the 4 games, without any stability test in conjunction with the counting backward task.
verbal feedback or physical assistance from the investigator. Upon cue, subjects were required move their avatar into the
All subjects were able to achieve this goal within 10 minutes. target by shifting their weight in the intended direction while
Subjects participated in all 4 games (23 minutes each) with 5 simultaneously reciting as many numbers as possible through
minutes rest in between games, totaling up to approximately the given serial subtraction. To allow for accurate comparison
110 minutes of balance training time per session. Subjects of the cognitive performance between ST and DT conditions,
wore a gait belt and a researcher provided external assistance the same 2 number sets (start number and subtracting digit)
(contact guard support) and supervised the subjects so that no from the seated cognitive task were utilized for the DT task.
falls occurred during the intervention period. During testing, subjects were not specifically instructed to pri-
oritize either the cognitive task or the balance task and the
Primary Outcome Measurements numbers for each trial remained same for all the subjects.
All the subjects completed assessments pre- and postin-
tervention, which were performed by a research assistant, who Trial Order. Subjects first performed the cognitive task
was blinded to prevent balance evaluation bias. The details of (ST condition). They were then given a 30-minute break as
each test are described in the following text. a washout period. The washout period was followed by the
limits of stability test. Each trial for the limits of stabil-
ST Condition ity test was coded on the basis of the task condition (ST
Balance Control Task. Balance control was assessed vs DT) and direction (forward vs backward). The order for
using the limits of stability test,28 a protocol of the Equi- each trial was randomized using a random number generator
Test (Computerized Dynamic Posturography),29,30 with evi- computer algorithm and followed for each subject. To avoid
dence of its reliability31 and validity.32,33 Limitations in the fatigue, a 1- to 2-minute rest period was provided between
ability to perform the test strongly correlated with other trials.
performance-based functional balance measures.32,35 Sub-
jects, secured with a safety harness system, stood on the Eq- Secondary Outcome Measurements
uiTest balance platform and were asked to lean their body in In addition to the primary outcome measures, adherence
either forward or backward directions to move their “avatar” to therapeutic activities37 was measured with the IMI scale.
from the center (center of pressure vector projected on the It has been used in several experiments related to intrinsic
screen in front of them) to the respective target box. Subjects motivation and self-regulation38,39 and has good evidence of
were asked to do this without losing balance, stepping, or being reliable.40 We also recorded the number of missed train-
reaching for assistance. All subjects performed 2 familiariza- ing days. Standardized clinical outcome measures such as the
tion trials, after which data for one trial, in both the directions, Berg Balance Scale and Timed Up and Go were also examined
were collected. pre- and postintervention. The gaming scores were recorded
The outcome measures to quantify balance performance from the Wii Fit software after each game (during training).
recorded by the software included reaction time, movement Schematic of the study design, demonstrating the chronologi-
velocity, maximum excursion, and directional control. Reac- cal sequence of events, consisting of the pretest, intervention,
tion time is the time, in seconds, between the command to and posttest, is represented in Figure 1.
move and the onset of the user’s movement. Movement ve-
locity is the average speed of center-of-gravity movement in Statistical Analysis
degrees per second. Maximum excursion is the maximum ex- To determine any change in performance between DT
cursion amplitude, in percentage, of center of mass achieved and ST (cognitive-motor interference) conditions, we calcu-
during the trial. Directional control is the comparison, in per- lated DT costs for each of the balance outcome measures (re-
centage, of the amount of movement of the center of mass action time, movement velocity, maximum excursion, and di-
in the intended direction (toward the target) to the amount of rectional control) and the cognitive measure (number of words
extraneous movement (away from the target). recited) pre- and postintervention. Dual-task cost equals [(ST
− DT)/DT] × 100.41 Positive cost (higher cost) indicated
Cognitive Task. While seated, subjects performed a lower performance for all outcome measures except reaction
counting backward task, which is a test of working memory time. Reaction time was expected to increase more under DT

218 
C 2014 Neurology Section, APTA

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 JNPT r Volume 38, October 2014 Dual-Task Training Paradigm
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

Figure 1. Schematic of the study design, demonstrating the chronological sequence of events, consisting of the pretest,
intervention, and posttest. The preintervention tests were performed 2 days before beginning the intervention; similarly, the
postintervention tests were performed 2 days after completion of the 5th day of intervention. For training, the question-answer
game and memory games were interspersed and the order of the word list generation, letter-number sequencing, and
question-answer memory recall games were randomized between the 5 training session but were kept same between subjects.
LOS, limits of stability; Fwd, Forward; Bwd, Backward; BBS, Berg Balance Scale; TUG, Timed Up and Go; IMI, Intrinsic Motivation
Inventory.

conditions than under ST conditions; a higher cost (lower per-
formance) would be indicated by increased negative values.
It was postulated that after training DT cost would be lower,
indicating significantly greater performance under DT condi-
tions for balance and cognitive measures. Paired t tests were
performed on these computed costs, along with changes in the
gaming scores, Berg Balance Scale, Timed Up and Go, and
IMI scores, to analyze the changes in postintervention com-
pared with preintervention values. A significance level (α) of
0.05 was chosen for statistical comparisons performed using
SPSS software version 17.0 for analysis. Also the magnitude
of effect between the pre and post intervention were calculated
with Cohen’s d formula where: d = Pre-intervention
√ mean -
Post-intervention mean/SD pooled and = d/ (d2 + 4). To
adjust for changes in intervention Cohen’s benchmarks were
used to indicate small (≤0.20), medium (≤0.50), and large
(≥0.80) effect sizes.
RESULTS
All subjects were present for all 5 days of training. Each
subject’s raw scores on the balance outcome measures (from
the limits of stability test) are presented in Figure 2. All sub-
jects were present for all 5 days of training. Each subject’s raw
scores on the balance outcome measures (from the limits of
stability test) are presented in Figure 2A-H.
Between pre- and postintervention scores, the reaction
time cost decreased significantly for forward (P < 0.01; η2 =
−0.83; 95% confidence interval [CI], −30.96 to −13.76) and
backward (P < 0.01; η2 = −0.92; 95% CI, −68.13 to −45.23)
directions (Figure 3A and B). There was a significant decrease
postintervention in movement velocity cost for forward (P <
0.01; η2 = 0.71; 95% CI, 9.66-35.62) and backward (P < 0.01;
η2 = 0.98; 95% CI, 36.34-46.96) directions (Figure 3C and
D). There was a similar decrease in maximum excursion cost
for forward (P < 0.05; η2 = 0.93; 95% CI, 24.39-39.91) and
backward (P < 0.05; η2 = 0.93; 95% CI, 28.27-42.41) direc-
tions (Figure 3E and F) and directional control cost for forward
(P < 0.05; η2 = 0.92; 95% CI, 19.66-29.75) and backward (P
< 0.05; η2 = 0.81; 95% CI, 8.85-25.30) directions (Figure 3G
and H). The cognitive cost was significantly lower for forward
(P < 0.05; η2 = 0.71; 95% CI, 10.41-44.1) and backward (P
< 0.05; η2 = 0.98; 95% CI, 2.30-40.61) directions (Figure 4).
For the gaming scores between pre- and postinterven-
tion (first day and last day), there was a significant increase in
Soccer (P < 0.01; η2 = −0.62; 95% CI, (−5.87 to −1.87),
Table tilt (P < 0.01; η2 = −0.70; 95% CI, −18.39 to −5.1),
Balance bubble (P < 0.01; η2 = −0.78; 95% CI, −90.40 to
−34.01), and Tightrope (P < 0.01; η2 = −0.87; 95% CI,
−20.24 to −12.75) scores (Table 2). There was a signifi-
cant improvement in both the Berg Balance Scale (P < 0.05;
η2 = −0.36; 95% CI, −3.71 to −1.53) and Timed Up and Go
scores postintervention (P < 0.05; η2 = 60; 95% CI, 1.63-
4.07). Postintervention scores on the IMI were significantly
greater than preintervention scores (P < 0.01; η2 = −0.92;
95% CI, −7.91 to −5.58) (Table 3). For all the outcome vari-
ables, the values of resulted in >50% of the variance due to
intervention, indicating large practical significance.
DISCUSSION
The results supported our hypothesis that there would be
a significant decrease in both balance and cognitive costs post-
training, suggesting an improvement in these functions and
corresponding decrease in cognitive-motor interference. Fur-
thermore, as hypothesized, subjects’ adherence to intervention


C 2014 Neurology Section, APTA 219

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 Subramaniam et al JNPT r Volume 38, October 2014
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

Figure 2. Change in pre- and posttraining scores of individual subject performance under dual-task condition on the limits of
stability test for reaction time (RT) in seconds (A and B), movement velocity (MV) in degrees per second (C and D), maximum
excursion (MXE) in percentage, and directional control (DC) in percentage. The left panel represents the forward (Fwd) direction
(A, C, E, and G), and the right panel represents the backward (Bwd) direction (B, D, F, and H).

220 
C 2014 Neurology Section, APTA

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 JNPT r Volume 38, October 2014 Dual-Task Training Paradigm
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

Figure 3. Means and standard deviations for the pre- and posttraining costs (in percentage) for reaction time (RT) cost (A
and B), movement velocity (MV) cost (C and D), maximum excursion (MXE) cost (E and F), directional control (DC) (G and H).
The cost for each variable represents the difference in performance on the limits of stability test between single-task and
dual-task conditions. The left panel represents the forward (Fwd) direction (A, C, E, and G) and the right panel represents the
backward (Bwd) direction (B, D, F, and H).

was maintained and postintervention motivation was signifi-
cantly greater.
There was a significant improvement in training-induced
performance on the Wii Fit game scores, suggesting that sub-
jects acquired the ability to successfully execute the balance-
related requirements of the gaming tasks: weight shifting, sym-
metric foot stepping, controlled movements near the limits
of stability, and adequate attention, memory, and decision-
making skills. In addition, the game scores (results) on each
trial provided performance feedback to the subjects, which
may have facilitated their decision making for improving
performance on subsequent trials.42 This reinforcement could


C 2014 Neurology Section, APTA 221

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 Subramaniam et al JNPT r Volume 38, October 2014
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024

Figure 4. Pre- and posttraining scores (correct responses) for individual subjects for the counting backward (CB) task under
dual-task condition for the (A) forward (Fwd) and (B) backward (Bwd) directions. The means and standard deviations of the
cognitive (Cog) cost in percentage are also presented for the (C) forward and (D) backward directions. The cognitive cost
represents the difference between the single-task (seated only) and dual-task conditions.

be particularly important for adherence to therapeutic activities Table 2. The Means (and Standard Deviations) for the
in community-dwelling stroke survivors with reduced motiva- Wii Fit Balance Gaming Scores for Day 1, Day 3, and Day 5
tion levels for rehabilitation programs.9,14 Significantly im-
proved scores on the IMI postintervention further lend support Day 1 Day 3 Day 5
to the aforementioned findings. Table tilt 19.41 (3.97) 31.95 (8.21) 35.66 (5.16)a
Better performance on the games indeed translated to an Tightrope 21.66 (6.44) 23.37 (10) 33.37 (5.42)a
Soccer 9.58 (2.84) 11.95 (3.14) 13.20 (1.79)a
improvement in balance control under DT conditions. Postin- Balance bubble 72.91 (16.65) 139 (68.10) 135.12 (30.19)b
tervention, subjects decreased DT cost in temporal (reaction a
P < 0.05, between day 1 and day 5.
time), spatial (movement velocity, maximum excursion, and b
P < 0.01, between day 1 and day 5.
directional control), and cognitive abilities, indicating im-
proved performance in these functions. These improvements
could have been a result of better allocation of cognitive re- Until recently, cognitive training and its influence on
sources, which might have further facilitated attentional con- motor behavior received little attention. A decline in cognitive
trol when performing cognitive tasks concurrently with bal- function was related to irreversible structural changes asso-
ance activities. ciated with aging or pathology, rather than a clinical symp-
Recent research recommends VR training to pro- tom that could benefit from rehabilitation.48,49 A recent study
mote improved integration of motor and cognitive skills for by Chapman et al50 has demonstrated significant plasticity-
improving physical function.43 However, other evidence in- induced improvements in cortical activations after a 12-week,
dicates that although VR interfaces allow for improved body 1 h/wk, cognitive training intervention in a group of older
awareness, movement processing, and overall motor skill ac- adults. The study demonstrated significant gains in resting state
quisition, they do not provide sufficient cognitive stimulation activation and increased connectivity and structural integrity
for addressing higher-order functions such as working mem- in the default mode network (a brain system that corresponds
ory and executive abilities.44,45 In fact, some studies have dis- to task introspection and active in the absence of focused task
cussed the difficulties in transferring cognitive skills trained performance) and the central executive network (regions of the
by video games to real tasks.46,47 This study thus explored a brain active while performing executive functions). We believe
combined cognitive-motor training paradigm, consisting of be- that DT training utilized in this study challenged higher-order
havioral training focused on improving higher-order cognitive functions including working memory and semantic mem-
skills along with balance skills. ory, which could have led to greater provision of attentional

222 
C 2014 Neurology Section, APTA

Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
 JNPT r Volume 38, October 2014
Table 3. Means (and Standard Deviations) of Pre- and
Postintervention Scores for the Secondary Outcome
Measures
Mean (SD)
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
Pretest Posttest
BBS (/56) 46.5 (3.46) 49.12 (3.13)
TUG, s 20.28 (2.4) 17.42 (1.39)
IMI (/25) 16.875 (1.35) 23.62 (1.30)
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024
Abbreviations: BBS, Berg Balance Scale; IMI, Intrinsic Motivation Inventory; TUG,
Timed Up and Go.
resources toward improving both balance control and cogni-
tive performance while performing these tasks concurrently.
Given the gradual and progressive decline in working memory,
especially in the chronic stages poststroke, such training could
have significant impacts to slow this decline.
Recovering from stroke involves relearning complex bal-
ance tasks requiring stability and mobility in the presence of
motor and cognitive deficits. Individuals with stroke therefore
use greater attentional resources to perform activities previ-
ously performed skillfully. In other words, these activities will
be performed at an associative or cognitive stage as opposed to
autonomous stage.14 When the benefits of movement automa-
tion are lost, balance control can be expected to be more vulner-
able to cognitive distractions (cognitive-motor interference),
subsequently increasing the risk of falls.51,52 Consequently,
the central element of successful cognitive-motor rehabilita-
tion for stroke survivors should be designed to compensate for
damaged cortical regions through the activation of compen-
satory reserves. Also, as balance control centers in the brain-
stem are postulated to be influenced by descending cortical
inputs, addressing practice-induced plasticity changes in these
networks may decrease cognitive-motor interference and im-
prove DT capacity.18,24,53 Interventions should therefore focus
on higher-order cognitive functions such as working memory
and executive function,18 and provide physical activities with
decision-making opportunities to facilitate the complex cog-
nitive processing required for community living.24 Improved
performance posttraining, as indicated by a decrease in both
balance and cognitive costs, suggests that DT rehabilitation
challenging higher-order functions could promote automatic-
ity of the DT performance.
Although the entire intervention was conducted over a
week, the overall dosage of approximately 9 hours of training is
similar to many other protocols that are spread over longer du-
rations. Also, recent rehabilitation research has shown the im-
portance of high-intensity, short-duration training in improv-
ing functional recovery in people who have had a stroke.14,53,54
The post-intervention improvement of 2.6 for the Berg Balance
Scale and 2.9s for the Timed up and Go test, were greater or
equal to the minimal detectable change of 2.5 for Berg Bal-
ance Scale and 2.9s for Timed Up and Go test respectively
established for people with chronic stroke. This suggests that
the post-intervention change in outcome measures could in-
deed be a meaningful clinical change that can lower fall-risk in
this population. These results further support to shorter-term

C 2014 Neurology Section, APTA
Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
Dual-Task Training Paradigm
effects of such high-intensity training protocols, lend further
support to shorter-term effects of such high-intensity training
protocols.55 While most current literature on adherence to par-
ticipation includes intervention durations ranging from a few
months to years, our study duration was only for a week. To
P have a wider application of this approach in stroke survivors,
0.001 future studies should explore dosage requirements for indi-
0.001 viduals at different stages of recovery, specifically addressing
0.000 long-term adherence to the protocol and retention of training-
induced changes.
Overall, the effect sizes in our study resulted in >50%
of the variance due to intervention, indicating medium to large
improvement in both balance and cognitive costs. Similar
effects were also seen in the gaming scores, Berg Balance
Scale, Timed Up and Go, and Intrinsic Motivation Inven-
tory scores post-training. This indicates relevant clinical sig-
nificance and promotes the importance of integrating this
paradigm into a clinical treatment program and assessing it’s
efficacy for translation into a home therapy program.
Limitations
The present study design did not provide any estima-
tion of learning or maturation trends during pretest due to
the lack of interrupted time-series design. However, previous
literature on reliability studies for the limits of stability test
demonstrates no learning effects when the test is used for as-
sessment once the patient is sufficiently familiarized with the
procedure prior to data collection.29,56,57 Thus, as part of our
study design, we provided 2 familiarization trials before the
actual testing, ensuring that the test was highly reliable and
did not have any maturation trends during the balance testing.
Furthermore, baseline tests were done on day 1, followed by the
intervention phase from day 4 to day 8, and postintervention as-
sessment performed on day 11 to ensure a sufficient time inter-
val between preassessment, intervention, and postassessment.
The lack of control group in the study might pose a
threat to its internal validity. However, the study protocol
was designed to address this limitation. It was ensured that
the training games selected for intervention used a wide
range of balance control mechanisms, such as lateral weight
shifting (Soccer and Tightrope game), and controlling the
displacement and variability of their center of pressure within
their existing base of support (Table tilt and balance bubble).
Further to minimize the threat to internal validity, the pre- and
postassessments were, however, conducted on the limits of
stability test specifically in the forward and backward direc-
tions. Similarly, the cognitive task used for the DT assessment
was different than the tasks used for training. Furthermore,
subjects did not participate in any other physical activity or
therapeutic interventions other than carrying on their typical
activities of daily living. The pre- and postassessment sessions
were performed at the same time of the day. Nonetheless,
future studies should also compare balance improvements
between equal doses of Wii Fit training with and without dual
tasking to determine whether the improvement in balance
control is purely due to the DT gaming paradigm or due to
the high-intensity Wii Fit training session.
223
 Subramaniam et al
CONCLUSIONS
The dual-task training paradigm proposed in this study
could effectively promote the ability to maintain optimal func-
tion on balance and cognitive tasks under challenging ‘real
life’ circumstances without compromising compliance to the
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
practice schedule. Further studies with larger sample sizes
are needed to assess efficacy of this intervention for potential
translation into a clinical treatment program.
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024
ACKNOWLEDGMENTS
The authors thank the American Heart Association,
National Affiliate, for the Scientific Development Grant (PI,
Dr Bhatt). The authors also thank Kaitlyn Reinwald for edit-
ing the manuscript, Tejal Kajrolkar for assisting with the data
collection, and Prakruti Patel for her thoughtful comments on
the manuscript.
REFERENCES
1. Grimby G, Andren E, Daving Y, Wright B. Dependence and perceived
difficulty in daily activities in community-living stroke survivors 2 years
after stroke: a study of instrumental structures. Stroke. 1998;29(9):1843-
1849.
2. Mathers C, Boerma T. The Global Burden of Disease. Geneva,
Switzerland: World Health Organization; 2004.
3. Gresham GE, Fitzpatrick TE, Wolf PA, McNamara PM, Kannel WB,
Dawber TR. Residual disability in survivors of stroke—the Framingham
study. New Engl J Med. 1975;293(19):954-956.
4. Mol VJ, Baker CA. Activity intolerance in the geriatric stroke patient.
Rehabil Nurs. 1991;16(6):337-343.
5. Belgen B, Beninato M, Sullivan PE, Narielwalla K. The association
of balance capacity and falls self-efficacy with history of falling in
community-dwelling people with chronic stroke. Arch Phys Med Rehabil.
2006;87(4):554-561.
6. Baseman S, Fisher K, Ward L, Bhattacharya A. The relationship of
physical function to social integration after stroke. J Neurosci Nurs.
2010;42(5):237-244.
7. Deutsch JE MM, Kafri M, Ranky R, Sivak M, Mavroidis C, Lewis JA.
Feasibility of virtual reality augmented cycling for health promotion of
people poststroke. J Neurol Phys Ther. 2013;37(3):118-124.
8. Todorov E, Shadmehr R, Bizzi E. Augmented feedback presented in a
virtual environment accelerates learning of a difficult motor task. J Mot
Behav. 1997;29(2):147-158.
9. Celinder D, Peoples H. Stroke patients’ experiences with Wii Sports(R)
during inpatient rehabilitation. Scand J Occup Ther. 2012;19(5):457-463.
10. Moreira MC, de Amorim Lima AM, Ferraz KM, Benedetti Rodrigues
MA. Use of virtual reality in gait recovery among post stroke patients—a
systematic literature review. Disabil Rehabil Assist Technol. 2013;8(5):
357-362.
11. Rand D KR, Weiss PT. The Sony PlayStation II EyeToy: low-cost virtual
reality for use in rehabilitation. J Neurol Phys Ther. 2008;32(4):155-163.
12. Cho KH, Lee KJ, Song CH. Virtual-reality balance training with a video-
game system improves dynamic balance in chronic stroke patients. Tohoku
J Exp Med. 2012;228(1):69-74.
13. Lange B, Flynn S, Proffitt R, Chang C-Y, Rizzo AS. Development of an
interactive game-based rehabilitation tool for dynamic balance training.
Top Stroke Rehabil. 2010;17(5):345-352.
14. Saposnik G, Mamdani M, Bayley M, et al. Effectiveness of Virtual Reality
Exercises in STroke Rehabilitation (EVREST): rationale, design, and pro-
tocol of a pilot randomized clinical trial assessing the Wii gaming system.
Int J Stroke. 2010;5(1):47-51.
15. Lundin-Olsson L, Nyberg L, Gustafson Y. Attention, frailty, and falls:
the effect of a manual task on basic mobility. J Am Geriatr Soc.
1998;46(6):758-761.
16. Montero-Odasso M, Verghese J, Beauchet O, Hausdorff JM. Gait and
cognition: a complementary approach to understanding brain function
and the risk of falling. J Am Geriatr Soc. 2012;60(11):2127-2136.
224
Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
JNPT r Volume 38, October 2014
17. Plummer P, Eskes G, Wallace S, et al. Cognitive-motor interference during
functional mobility after stroke: state of the science and implications for
future research. Arch Phys Med Rehabil. 2013;94(12):2565-2574.
18. Cicerone KD, Dahlberg C, Kalmar K, et al. Evidence-based cognitive
rehabilitation: recommendations for clinical practice. Arch Phys Med Re-
habil. 2000;81(12):1596-1615.
19. Theill N, Schumacher V, Adelsberger R, Martin M, Jancke L. Effects of
simultaneously performed cognitive and physical training in older adults.
Arch Phys Med Rehabil. 2000;81(12):1596-1615.
20. Fok P, Farrell M, McMeeken J. The effect of dividing attention between
walking and auxiliary tasks in people with Parkinson’s disease. Hum Mov
Sci. 2012;31(1):236-246.
21. Fritz NE, Basso DM. Dual-task training for balance and mobility in a
person with severe traumatic brain injury: a case study. J Neurol Phys
Ther. 2013;37:37-43.
22. Glasauer S, Stein A, Günther AL, Flanagin VL, Jahn K, Brandt T. The
effect of dual tasks in locomotor path integration. Ann N Y Acad Sci.
2009;1164:201-205.
23. Ruthruff E, Van Selst M, Johnston JC, Remington R. How does practice
reduce dual-task interference: integration, automatization, or just stage-
shortening? Psychol Res. 2006;70(2):125-142.
24. Robertson IH, Murre JM. Rehabilitation of brain damage: brain plas-
ticity and principles of guided recovery. Psychol Bull. 1999;125(5):
544-575.
25. Collette F, Van der Linden M. Brain imaging of the central executive
component of working memory. Neurosci Biobehav Rev. 2002;26(2):
105-125.
26. Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Val-
idation of a short Orientation-Memory-Concentration Test of cognitive
impairment. Am J Psychiatry. 1983;140(6):734-739.
27. AI-Khawaja I, Wade D, Collin CF. Bedside screening for aphasia: a com-
parison of two methods. J Neurol. 1996(243):201-204.
28. NeuroCom. The NeuroCom Smart EquiTest R System Limits of Stabil-
ity Test—Clinical Interpretation Guide. Clackamas, OR: NeuroCom;
2009.
29. Clark S, Rose DJ. Evaluation of dynamic balance among community-
dwelling older adult fallers: a generalizability study of the limits of stability
test. Arch Phys Med Rehabil. 2001;82(4):468-474.
30. Chien CW, Hu M-H, Tang P-F, Sheu C-F, Hsieh CL. A comparison
of psychometric properties of the smart balance master system and the
postural assessment scale for stroke in people who have had mild stroke.
Arch Phys Med Rehabil. 2007;88(3):374-380.
31. Liston RA, Brouwer BJ. Reliability and validity of measures obtained
from stroke patients using the Balance Master. Arch Phys Med Rehabil.
1996;77(5):425-430.
32. Wallmann HW. Comparison of elderly nonfallers and fallers on perfor-
mance measures of functional reach, sensory organization, and limits of
stability. J Gerontol A Biol Sci Med Sci. 2001;56(9):M580-M583.
33. Au-Yeung SS, Ng JT, Lo SK. Does balance or motor impairment of limbs
discriminate the ambulatory status of stroke survivors? Am J Phys Med
Rehabil. 2003;82(4):279-283.
34. Alonte A GK, Brenneman SK. Relationship of scores on the Berg Balance
Scale to results of the limits of stability portion of the Smart Balance
Master R suite of tests. Neurol Rep. 1995:19:22-23.
35. Allison LK, Rose DJ. The relationship between postural control system
impairments and disabilities in older adults. Phys Ther. 1998;78:5.
36. Unsworth N, Engle RW. The nature of individual differences in working
memory capacity: active maintenance in primary memory and controlled
search from secondary memory. Psychol Rev. 2007;114(1):104-132.
37. Medalia A, Saperstein A. The role of motivation for treatment success.
Schizophr Bull. 2011;37(2):122-128.
38. Ryan RM, Connell JP, Plant RW. Emotions in non-directed text learning.
Learn Individ Diff. 1990(2):1-17.
39. Plant RW, Ryan RM. Intrinsic motivation and the effects of self-
consciousness, self-awareness, and ego-involvement: an investigation of
internally-controlling styles. J Pers Soc Psychol. 1985(53):435-449.
40. McAuley E, Duncan T, Tammen VV. Psychometric properties of the In-
trinsic Motivation Inventory in a competitive sport setting: a confirmatory
factor analysis. Res Q Exerc Sport. 1989;60(1):48-58.
41. Bock O. Dual-task costs while walking increase in old age for some, but
not for other tasks: an experimental study of healthy young and elderly
persons. J Neuroeng Rehabil. 2008;5:27.

C 2014 Neurology Section, APTA
 JNPT r Volume 38, October 2014
42. Brown P, Chen CC, Wang S, et al. Involvement of human basal gan-
glia in offline feedback control of voluntary movement. Curr Biol.
2006;16(21):2129-2134.
43. Kim BR, Chun MH, Kim LS, Park JY. Effect of virtual reality on cognition
in stroke patients. Ann Rehabil Med. 2011;35(4):450-459.
44. Pompeu JE, Mendes FA, Silva KG, et al. Effect of Nintendo Wii-
Downloaded from http://journals.lww.com/jnpt by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW
based motor and cognitive training on activities of daily living in pa-
tients with Parkinson’s disease: a randomised clinical trial. Physiotherapy.
2012;98(3):196-204.
45. dos Santos Mendes FA, Pompeu JE, Modenesi Lobo A, et al. Mo-
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/04/2024
tor learning, retention and transfer after virtual-reality-based training in
Parkinson’s disease—effect of motor and cognitive demands of games:
a longitudinal, controlled clinical study. Physiotherapy. 2012;98(3):
217-223.
46. Basak C, Boot WR, Voss MW, Kramer AF. Can training in a real-time
strategy video game attenuate cognitive decline in older adults? Psychol
Aging. 2008;23(4):765-777.
47. Boot WR, Basak C, Erickson KI, et al. Transfer of skill engendered by
complex task training under conditions of variable priority. Acta Psychol
(Amst). 2010;135(3):349-357.
48. Reuter-Lorenz PA, Lustig C. Brain aging: reorganizing discov-
eries about the aging mind. Curr Opin Neurobiol. 2005;15(2):
245-251.
49. Mahncke HW, Bronstone A, Merzenich MM. Brain plasticity and func-
tional losses in the aged: scientific bases for a novel intervention. Prog
Brain Res. 2006;157:81-109.

C 2014 Neurology Section, APTA
Copyright © 2014 Neurology Section, APTA. Unauthorized reproduction of this article is prohibited.
Dual-Task Training Paradigm
50. Chapman SB, Aslan S, Spence JS, et al. Neural mechanisms of
brain plasticity with complex cognitive training in healthy seniors
[published online ahead of print August 28, 2013]. Cereb Cortex.
doi:10.1093/cercor/bht234.
51. Barra J, Bray A, Sahni V, Golding JF, Gresty MA. Increasing cognitive
load with increasing balance challenge: recipe for catastrophe. Exp Brain
Res. 2006;174(4):734-745.
52. Broglio SP, Tomporowski PD, Ferrara MS. Balance performance with
a cognitive task: a dual-task testing paradigm. Med Sci Sports Exerc.
2005;37(4):689-695.
53. Liepert J, Bauder H, Wolfgang HR, Miltner WH, Taub E, Weiller C.
Treatment-induced cortical reorganization after stroke in humans. Stroke.
2000;31(6):1210-1216.
54. Combs SA, Miller EW. Effects of a short burst of gait training with
body weight-supported treadmill training for a person with chronic
stroke: a single-subject study. Physiother Theory Pract. 2011;27(3):
223-230.
55. Maeda N, Kato J, Shimada T. Predicting the probability for fall inci-
dence in stroke patients using the Berg Balance Scale. J Int Med Res.
2009;37(3):697-704.
56. Clark S, Rose DJ, Fujimoto K. Generalizability of the limits of stability
test in the evaluation of dynamic balance among older adults. Arch Phys
Med Rehabil. 1997;78(10):1078-1084.
57. Rose DJ, McKillop J. Assessment of balance and mobility functions: a ref-
erence study based on the Balance Master. 1998. Located at: NeuroCom,
Inc, Clackamas, OR.
225