Accepted Manuscript A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in Patients with Chronic Stroke Jen-Wen Hung , Md Chiung-Xia Chou , OT Yen-Wei Hsieh , OT Wen-Chi Wu , OT Min-Yuan Yu , OT Po-Chih Chen , OT Shiue-Feng Chang , OTA Shan-Er Ding , OTA PII:
S0003-9993(14)00366-9
DOI:
10.1016/j.apmr.2014.04.029
Reference:
YAPMR 55837
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 5 February 2014 Accepted Date: 22 April 2014
Please cite this article as: Hung J-W, Chou C-X, Hsieh Y-W, Wu W-C, Yu M-Y, Chen P-C, Chang S-F, Ding S-E, A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in Patients with Chronic Stroke, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2014), doi: 10.1016/j.apmr.2014.04.029. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Running Head: Wii Fit Training for Balance in Stroke
RI PT
Title: A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in Patients with Chronic Stroke
SC
Authors:
M AN U
Jen-Wen Hung, MD; Chiung-Xia Chou OT; Yen-Wei Hsieh, OT; Wen-Chi Wu, OT; Min-Yuan Yu, OT; Po-Chih Chen, OT; Shiue-Feng Chang, OTA; Shan-Er Ding,
TE D
OTA
Affiliations:
From the Department of Rehabilitation, Chang Gung Memorial Hospital-Kaohsiung
AC C
EP
Medical Center, Chang Gung University College of Medicine, Kaohsiung, Taiwan
Corresponding author Department of Physical Medicine and Rehabilitation, Kaohsiung Chang Gung Memorial Hospital, Taiwan. No.123, Ta-Pei Road, Niao-Sung District, Kaohsiung City 833, Taiwan. E-mail: [email protected]
ACCEPTED MANUSCRIPT Tel: +886-7-7317123 Fax: +886-7-7336988
RI PT
Acknowledgment: This study was supported by a research grant from the Medical Research Center at
SC
Chang Gung Memorial Hospital CMRPG8A0641
M AN U
Conflict of Interest
All authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated AND, if applicable, we certify that all financial and material
TE D
support for this research (eg, NIH or NHS grants) and work are clearly identified in
AC C
EP
the title page of the manuscript.
ACCEPTED MANUSCRIPT
A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in
2
Patients with Chronic Stroke
3
Abstract
4
Objective
5
To compare the effects of exergaming with conventional weight-shifting training on balance function in patients with chronic stroke
M AN U
SC
RI PT
1
Design
8
Single-blind randomized controlled trial
9
EP
7
TE D
6
Setting
11
Rehabilitation department of a medical center
AC C
10
12
1
13
Participants
14
Patients (N=30) with chronic stroke and balance deficits
RI PT
ACCEPTED MANUSCRIPT
SC
15 Intervention
17
Twelve-week Wii Fit training or conventional weight-shifting training
M AN U
16
18 Main outcome measures
20
Static balance was assessed by using posturography. We recorded the stability index and percentage of weight bearing on the affected
21
leg in eight positions. We also used timed-up-and-go and forward reach tests for dynamic balance evaluation, the Falls Efficacy
22
Scale-International questionnaire for fear-of- falling assessment, and physical activity enjoyment scale for estimating the enjoyment of
23
training.
AC C
EP
TE D
19
24
2
ACCEPTED MANUSCRIPT
Results
26
The exergaming group showed more improvement in the stability index than the control group in (1) head straight with eyes open on
27
foam surface, (2) eyes closed on solid surface with head turned at 30° to left or (3) with head up positions (time-group interaction p
28
= .02, .04, and .03, respectively); however, the effects were not maintained. At 3-month follow-up, control group showed more
29
improvement in weight-bearing symmetry in the “head straight with eyes open on solid surface” position than the exergaming group
30
(time-group interaction p = .03). Both groups showed improvement in timed-up-and-go, forward reach tests, and fear of falling. The
31
improvement in fear of falling was not maintained. The exergaming group enjoyed training more than the control group (p = .03).
TE D
M AN U
SC
RI PT
25
32 Conclusions
34
Exergaming is enjoyable and effective for patients with chronic stroke.
36 37
AC C
35
EP
33
Keywords: Stroke, Postural Balance, Wii Fit, Video games, Feedback Sensory
3
ACCEPTED MANUSCRIPT
List of Abbreviation
39
SC
M AN U
TE D
Eyes closed on solid surface with head up Eyes closed on solid surface with head down Stability Index Timed Up and Go Forward Reach Falls Efficacy Scale-International Physical Activity Enjoyment Scale
EP
HB HF ST TUG FR FES-I PACES
Center of Pressure Head straight with eyes open on solid surface Head straight with eyes closed on solid surface Head straight with eyes open on foam surface Head straight with eyes closed on foam surface Eyes closed on solid surface with head turned at 30° to right Eyes closed on solid surface with head turned at 30°o left.
AC C
COP NO NC PO PC HR HL
RI PT
38
4
ACCEPTED MANUSCRIPT
Balance deficits were frequently seen in stroke survivors,1, 2 which were significantly associated with ambulatory dysfunction,
41
dependency in activities of daily living, and increased fall risk.3, 4 Increased postural sway, decreased weight bearing on affected leg,
42
and poor weight shifting are common balance problems post stroke.5-7 Force-plate biofeedback systems have been used for balance
43
training.3, 4, 8 It has been reported using force-plate system can improve stance symmetry in patients with hemiplegia.9 However,
44
force-plate systems are expensive. In addition, patients may quickly lose interest in performing repetitive tasks; lack of interest can
45
decrease the potential effectiveness of the therapy. To enhance the level of biofeedback, a coupling biofeedback exercise linked directly
46
to video games has been applied to rehabilitation.10 The fun, interactive approach of such a system may motivate patients to increase
47
their participation and is likely to result in better therapy results.
SC
M AN U
TE D
EP
48
RI PT
40
Nintendo Wii Fit is a balance-training tool. The Wii Fit system comprises a balance board, which is similar to a force platform but more
50
user friendly.11 Wii Fit games were not originally designed to train disabled persons, but these games have been integrated into
51
rehabilitation programs for patients with neurological disorders.12-18 Only two studies have focused on the use of such games in patients
AC C
49
5
ACCEPTED MANUSCRIPT
with stroke.17, 18 Both studies allowed additional training time for the Wii Fit games, which might have created a bias in favor of Wii Fit
53
intervention because the intensity and frequency of training per se could lead to positive outcomes. In addition, we do not know whether
54
Wii Fit training provides a greater effect than traditional weight shifting training. Moreover, there was no follow-up assessment:
55
information about the maintenance effect was unavailable. Furthermore, neither adverse events nor enjoyment of training were reported;
56
such reports would be important for the successful application of a new intervention. Our hypothesis was that using Wii Fit games might
57
provide a more enjoyable and effective intervention than traditional weight-shifting training, and that the effects could be maintained.
58
Hence, the purpose of this study was to compare the feasibility, enjoyment and effectiveness of Wii Fit balance training with traditional
59
weight-shifting training, with a longitudinal follow-up in patients with chronic stroke.
61
Methods
62 63
AC C
60
EP
TE D
M AN U
SC
RI PT
52
Participants
6
ACCEPTED MANUSCRIPT
Participants were enrolled from the rehabilitation department of a medical center. Inclusion criteria were: patients who had a hemiplegic
65
stroke at least 6 months prior to enrolment; aged over 18 years; a Berg balance score 0.95) in patients with chronic stroke.23
121
The mean of two consecutive trials for each measure was used for statistical analysis. If the subject used a walking aid at the
122
pre-intervention assessment, the same aid was used for all subsequent assessments.
AC C
EP
TE D
M AN U
115
123
11
ACCEPTED MANUSCRIPT
We used the Falls Efficacy Scale-International (FES-I) questionnaire to assess the intervention effect on fear of falling. The FES-I was
125
developed by the Prevention of Falls Network Europe group24 to assess the different levels of concern about falling when carrying out
126
16 specific activities, by using a 4-point scale (1 = not at all concerned, 4 = very concerned). The FES-I questionnaire has excellent
127
psychometric properties, which have been demonstrated both in English and in a cross-cultural context.25
M AN U
SC
RI PT
124
128
At the end of the training, we used the Physical Activity Enjoyment Scale (PACES)26 to evaluate the participants’ enjoyment in training.
130
The scale consists of 18 statements describing bipolar responses associated with physical activity. Respondents were asked to rate “how
131
you feel at the moment about the physical activity you have been doing” by using a 7-point bipolar rating scale. Higher PACES scores
132
reflected greater levels of enjoyment.
AC C
133
EP
TE D
129
134
Statistical analysis
135
We used descriptive statistics to calculate the demographic and clinical characteristics of the participants, and the Kolmogorov–Smirnov
12
ACCEPTED MANUSCRIPT
test to confirm the normality of the data. Inter-group differences among the baseline characteristics were evaluated by using an
137
independent t test or χ2 analysis.
RI PT
136
SC
138
All outcome analyses were conducted on a per-protocol basis, excluding patients with major deviations from the treatment protocol
140
(defined as less than 50% adherence to the treatment). We used analysis of variance with repeated measures to determine the effects of
141
intervention on each dependent variable. Model effects were group (experimental, control) and time (pre-, post training), and their
142
interaction. We calculated partial η2 for each outcome variable to index the magnitude for each effect. A large effect was represented by
143
a partial η2 of at least 0.14, a moderate effect by 0.06, and a small effect by 0.01.27 We used the same analytical procedures to examine
144
the maintenance of intervention effects at 3-month follow-up. Data were analyzed by using SPSS version 12.0, with statistical
145
significance set at p = 0.05.
147
AC C
146
EP
TE D
M AN U
139
Results
13
ACCEPTED MANUSCRIPT
RI PT
148 Baseline characteristics of the participants
150
We screened 150 subjects for eligibility; 30 underwent randomization, 15 for each group (Fig 3). Three participants discontinued the
151
study protocol—all from the Wii Fit group. One participant withdrew after six intervention sections, because she had severe ankle
152
inversion when standing on the balance board. Another subject transferred his rehabilitation training to another hospital and dropped out
153
after eight intervention sections. A subject stopped intervention after 16 training sections because of medical issues unrelated to the
154
intervention; however, he completed the post-intervention and follow-up measures. According to the principle of per-protocol analysis,
155
the first two cases were excluded from data analysis. The two groups were well matched regarding baseline data (Tables 1–4).
156
EP
TE D
M AN U
SC
149
Feasibility, safety, and enjoyment
158
The average attendance throughout the 12-week period was 22 out of 24 sessions for the Wii Fit group (range 16–24) and 23 out of 24
159
for the control group (range 21–24). There was no significant difference in attendance between groups (p = 0.19). Two participants in
AC C
157
14
ACCEPTED MANUSCRIPT
the control group reported increased knee pain when training. Three participants in the Wii Fit group reported increased spasticity when
161
playing the basic stepping games. This symptom subsided when they played other games. Accidental falls or other adverse events did
162
not occur. The Wii Fit group enjoyed the training more than the control group (PACES score 79.15±7.49 vs 72.47±7.63 respectively, p
163
= .03).
M AN U
SC
RI PT
160
164 Balance performance
166
After intervention there were significant time-group interactions in ST in the PO (p = .02), HL (p = .04), and HB (p = .03) positions; all
167
the effect sizes were large (all partial η2 >.14). These findings indicate between-group differences over time, favoring Wii fit training;
168
however, the gain was not maintained at 3-month follow-up. (Table 2) Both groups increased their weight-bearing on the affected leg
169
after intervention; there was no time-group interaction. At 3-month follow-up, the Wii Fit group showed a decline in symmetric weight
170
bearing, in the NO position there was a significant time-group interaction (p = .03); the effect size was large (partial η2 = .16) (Table 3).
AC C
EP
TE D
165
171
15
ACCEPTED MANUSCRIPT
After intervention, both groups showed improvement in TUG and FR tests, and less fear of falling (time effect in TUG p = . 02, partial
173
η2 = .21; FR p
S0003-9993(14)00366-9
DOI:
10.1016/j.apmr.2014.04.029
Reference:
YAPMR 55837
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 5 February 2014 Accepted Date: 22 April 2014
Please cite this article as: Hung J-W, Chou C-X, Hsieh Y-W, Wu W-C, Yu M-Y, Chen P-C, Chang S-F, Ding S-E, A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in Patients with Chronic Stroke, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2014), doi: 10.1016/j.apmr.2014.04.029. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Running Head: Wii Fit Training for Balance in Stroke
RI PT
Title: A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in Patients with Chronic Stroke
SC
Authors:
M AN U
Jen-Wen Hung, MD; Chiung-Xia Chou OT; Yen-Wei Hsieh, OT; Wen-Chi Wu, OT; Min-Yuan Yu, OT; Po-Chih Chen, OT; Shiue-Feng Chang, OTA; Shan-Er Ding,
TE D
OTA
Affiliations:
From the Department of Rehabilitation, Chang Gung Memorial Hospital-Kaohsiung
AC C
EP
Medical Center, Chang Gung University College of Medicine, Kaohsiung, Taiwan
Corresponding author Department of Physical Medicine and Rehabilitation, Kaohsiung Chang Gung Memorial Hospital, Taiwan. No.123, Ta-Pei Road, Niao-Sung District, Kaohsiung City 833, Taiwan. E-mail: [email protected]
ACCEPTED MANUSCRIPT Tel: +886-7-7317123 Fax: +886-7-7336988
RI PT
Acknowledgment: This study was supported by a research grant from the Medical Research Center at
SC
Chang Gung Memorial Hospital CMRPG8A0641
M AN U
Conflict of Interest
All authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated AND, if applicable, we certify that all financial and material
TE D
support for this research (eg, NIH or NHS grants) and work are clearly identified in
AC C
EP
the title page of the manuscript.
ACCEPTED MANUSCRIPT
A Randomized Comparison Trial of Balance Training by Using Exergaming and Conventional Weight-Shifting Therapy in
2
Patients with Chronic Stroke
3
Abstract
4
Objective
5
To compare the effects of exergaming with conventional weight-shifting training on balance function in patients with chronic stroke
M AN U
SC
RI PT
1
Design
8
Single-blind randomized controlled trial
9
EP
7
TE D
6
Setting
11
Rehabilitation department of a medical center
AC C
10
12
1
13
Participants
14
Patients (N=30) with chronic stroke and balance deficits
RI PT
ACCEPTED MANUSCRIPT
SC
15 Intervention
17
Twelve-week Wii Fit training or conventional weight-shifting training
M AN U
16
18 Main outcome measures
20
Static balance was assessed by using posturography. We recorded the stability index and percentage of weight bearing on the affected
21
leg in eight positions. We also used timed-up-and-go and forward reach tests for dynamic balance evaluation, the Falls Efficacy
22
Scale-International questionnaire for fear-of- falling assessment, and physical activity enjoyment scale for estimating the enjoyment of
23
training.
AC C
EP
TE D
19
24
2
ACCEPTED MANUSCRIPT
Results
26
The exergaming group showed more improvement in the stability index than the control group in (1) head straight with eyes open on
27
foam surface, (2) eyes closed on solid surface with head turned at 30° to left or (3) with head up positions (time-group interaction p
28
= .02, .04, and .03, respectively); however, the effects were not maintained. At 3-month follow-up, control group showed more
29
improvement in weight-bearing symmetry in the “head straight with eyes open on solid surface” position than the exergaming group
30
(time-group interaction p = .03). Both groups showed improvement in timed-up-and-go, forward reach tests, and fear of falling. The
31
improvement in fear of falling was not maintained. The exergaming group enjoyed training more than the control group (p = .03).
TE D
M AN U
SC
RI PT
25
32 Conclusions
34
Exergaming is enjoyable and effective for patients with chronic stroke.
36 37
AC C
35
EP
33
Keywords: Stroke, Postural Balance, Wii Fit, Video games, Feedback Sensory
3
ACCEPTED MANUSCRIPT
List of Abbreviation
39
SC
M AN U
TE D
Eyes closed on solid surface with head up Eyes closed on solid surface with head down Stability Index Timed Up and Go Forward Reach Falls Efficacy Scale-International Physical Activity Enjoyment Scale
EP
HB HF ST TUG FR FES-I PACES
Center of Pressure Head straight with eyes open on solid surface Head straight with eyes closed on solid surface Head straight with eyes open on foam surface Head straight with eyes closed on foam surface Eyes closed on solid surface with head turned at 30° to right Eyes closed on solid surface with head turned at 30°o left.
AC C
COP NO NC PO PC HR HL
RI PT
38
4
ACCEPTED MANUSCRIPT
Balance deficits were frequently seen in stroke survivors,1, 2 which were significantly associated with ambulatory dysfunction,
41
dependency in activities of daily living, and increased fall risk.3, 4 Increased postural sway, decreased weight bearing on affected leg,
42
and poor weight shifting are common balance problems post stroke.5-7 Force-plate biofeedback systems have been used for balance
43
training.3, 4, 8 It has been reported using force-plate system can improve stance symmetry in patients with hemiplegia.9 However,
44
force-plate systems are expensive. In addition, patients may quickly lose interest in performing repetitive tasks; lack of interest can
45
decrease the potential effectiveness of the therapy. To enhance the level of biofeedback, a coupling biofeedback exercise linked directly
46
to video games has been applied to rehabilitation.10 The fun, interactive approach of such a system may motivate patients to increase
47
their participation and is likely to result in better therapy results.
SC
M AN U
TE D
EP
48
RI PT
40
Nintendo Wii Fit is a balance-training tool. The Wii Fit system comprises a balance board, which is similar to a force platform but more
50
user friendly.11 Wii Fit games were not originally designed to train disabled persons, but these games have been integrated into
51
rehabilitation programs for patients with neurological disorders.12-18 Only two studies have focused on the use of such games in patients
AC C
49
5
ACCEPTED MANUSCRIPT
with stroke.17, 18 Both studies allowed additional training time for the Wii Fit games, which might have created a bias in favor of Wii Fit
53
intervention because the intensity and frequency of training per se could lead to positive outcomes. In addition, we do not know whether
54
Wii Fit training provides a greater effect than traditional weight shifting training. Moreover, there was no follow-up assessment:
55
information about the maintenance effect was unavailable. Furthermore, neither adverse events nor enjoyment of training were reported;
56
such reports would be important for the successful application of a new intervention. Our hypothesis was that using Wii Fit games might
57
provide a more enjoyable and effective intervention than traditional weight-shifting training, and that the effects could be maintained.
58
Hence, the purpose of this study was to compare the feasibility, enjoyment and effectiveness of Wii Fit balance training with traditional
59
weight-shifting training, with a longitudinal follow-up in patients with chronic stroke.
61
Methods
62 63
AC C
60
EP
TE D
M AN U
SC
RI PT
52
Participants
6
ACCEPTED MANUSCRIPT
Participants were enrolled from the rehabilitation department of a medical center. Inclusion criteria were: patients who had a hemiplegic
65
stroke at least 6 months prior to enrolment; aged over 18 years; a Berg balance score 0.95) in patients with chronic stroke.23
121
The mean of two consecutive trials for each measure was used for statistical analysis. If the subject used a walking aid at the
122
pre-intervention assessment, the same aid was used for all subsequent assessments.
AC C
EP
TE D
M AN U
115
123
11
ACCEPTED MANUSCRIPT
We used the Falls Efficacy Scale-International (FES-I) questionnaire to assess the intervention effect on fear of falling. The FES-I was
125
developed by the Prevention of Falls Network Europe group24 to assess the different levels of concern about falling when carrying out
126
16 specific activities, by using a 4-point scale (1 = not at all concerned, 4 = very concerned). The FES-I questionnaire has excellent
127
psychometric properties, which have been demonstrated both in English and in a cross-cultural context.25
M AN U
SC
RI PT
124
128
At the end of the training, we used the Physical Activity Enjoyment Scale (PACES)26 to evaluate the participants’ enjoyment in training.
130
The scale consists of 18 statements describing bipolar responses associated with physical activity. Respondents were asked to rate “how
131
you feel at the moment about the physical activity you have been doing” by using a 7-point bipolar rating scale. Higher PACES scores
132
reflected greater levels of enjoyment.
AC C
133
EP
TE D
129
134
Statistical analysis
135
We used descriptive statistics to calculate the demographic and clinical characteristics of the participants, and the Kolmogorov–Smirnov
12
ACCEPTED MANUSCRIPT
test to confirm the normality of the data. Inter-group differences among the baseline characteristics were evaluated by using an
137
independent t test or χ2 analysis.
RI PT
136
SC
138
All outcome analyses were conducted on a per-protocol basis, excluding patients with major deviations from the treatment protocol
140
(defined as less than 50% adherence to the treatment). We used analysis of variance with repeated measures to determine the effects of
141
intervention on each dependent variable. Model effects were group (experimental, control) and time (pre-, post training), and their
142
interaction. We calculated partial η2 for each outcome variable to index the magnitude for each effect. A large effect was represented by
143
a partial η2 of at least 0.14, a moderate effect by 0.06, and a small effect by 0.01.27 We used the same analytical procedures to examine
144
the maintenance of intervention effects at 3-month follow-up. Data were analyzed by using SPSS version 12.0, with statistical
145
significance set at p = 0.05.
147
AC C
146
EP
TE D
M AN U
139
Results
13
ACCEPTED MANUSCRIPT
RI PT
148 Baseline characteristics of the participants
150
We screened 150 subjects for eligibility; 30 underwent randomization, 15 for each group (Fig 3). Three participants discontinued the
151
study protocol—all from the Wii Fit group. One participant withdrew after six intervention sections, because she had severe ankle
152
inversion when standing on the balance board. Another subject transferred his rehabilitation training to another hospital and dropped out
153
after eight intervention sections. A subject stopped intervention after 16 training sections because of medical issues unrelated to the
154
intervention; however, he completed the post-intervention and follow-up measures. According to the principle of per-protocol analysis,
155
the first two cases were excluded from data analysis. The two groups were well matched regarding baseline data (Tables 1–4).
156
EP
TE D
M AN U
SC
149
Feasibility, safety, and enjoyment
158
The average attendance throughout the 12-week period was 22 out of 24 sessions for the Wii Fit group (range 16–24) and 23 out of 24
159
for the control group (range 21–24). There was no significant difference in attendance between groups (p = 0.19). Two participants in
AC C
157
14
ACCEPTED MANUSCRIPT
the control group reported increased knee pain when training. Three participants in the Wii Fit group reported increased spasticity when
161
playing the basic stepping games. This symptom subsided when they played other games. Accidental falls or other adverse events did
162
not occur. The Wii Fit group enjoyed the training more than the control group (PACES score 79.15±7.49 vs 72.47±7.63 respectively, p
163
= .03).
M AN U
SC
RI PT
160
164 Balance performance
166
After intervention there were significant time-group interactions in ST in the PO (p = .02), HL (p = .04), and HB (p = .03) positions; all
167
the effect sizes were large (all partial η2 >.14). These findings indicate between-group differences over time, favoring Wii fit training;
168
however, the gain was not maintained at 3-month follow-up. (Table 2) Both groups increased their weight-bearing on the affected leg
169
after intervention; there was no time-group interaction. At 3-month follow-up, the Wii Fit group showed a decline in symmetric weight
170
bearing, in the NO position there was a significant time-group interaction (p = .03); the effect size was large (partial η2 = .16) (Table 3).
AC C
EP
TE D
165
171
15
ACCEPTED MANUSCRIPT
After intervention, both groups showed improvement in TUG and FR tests, and less fear of falling (time effect in TUG p = . 02, partial
173
η2 = .21; FR p
