Hong Kong Physiotherapy Journal (2016) 34, 27e32

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.hkpj-online.com

RESEARCH REPORT

The effect of mirror therapy on balance

ability of subacute stroke patients
Myoung-Kwon Kim, PT, PhD a, Sang-Gu Ji, PT, PhD b,
Hyun-Gyu Cha, PT, PhD c,*

a
Department of Physical Therapy, Daegu University, 201, Daegudae-ro, Gyeongsan-si,
Gyeongsangbuk-do, Republic of Korea
b
Department of Physical Therapy, Eulji University Hospital, Dunsan-dong 1306, Seo-gu, Daejeon,
302-799, Republic of Korea
c
Department of Physical Therapy, Kyungbuk College, Hyucheon-dong, Yeongju, Gyungbuk 750-050,
Republic of Korea

KEYWORDS Abstract Background: Rehabilitation of subacute stroke patients represents a major chal-
balance; lenge.
mirror therapy; Objective: This study was conducted to examine the effects of mirror therapy on balance abil-
subacute stroke ity among subacute stroke patients.
Methods: The patients were assigned to a mirror therapy group (n Z 17) or a sham therapy
group (n Z 17). Participants in the experimental group received mirror therapy and conven-
tional rehabilitation therapy for a total of 60 minutes (mirror therapy: 30 minutes; conven-
tional rehabilitation therapy: 30 minutes) per day, with a 10 minute rest period halfway
through the session. Participants in the experimental group received training 5 days/week
for 4 weeks. Participants in the control group received sham therapy and conventional rehabil-
itation therapy for a total of 60 minutes (sham therapy: 30 minutes, conventional rehabilita-
tion therapy: 30 minutes) per day on the same day. Balance Index (BI) scores were obtained
using a balance measurement system.
Results: A significant difference in post-training gains for the overall stability index and medial
and lateral stability index was observed between the experimental group and the control group
(p < 0.05).
Conclusion: We conclude that mirror therapy may be beneficial in improving balance ability
among subacute stroke patients.
Copyright ª 2016, Hong Kong Physiotherapy Association. Published by Elsevier (Singapore) Pte
Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.
org/licenses/by-nc-nd/4.0/).

* Corresponding author. Department of Physical Therapy, Kyungbuk College, Hyucheon-dong, Yeongju, Gyungbuk 750-050, Republic of
Korea.
E-mail address: [email protected] (H.-G. Cha).

http://dx.doi.org/10.1016/j.hkpj.2015.12.001
1013-7025/Copyright ª 2016, Hong Kong Physiotherapy Association. Published by Elsevier (Singapore) Pte Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
28 M.-K. Kim et al.

Introduction (Mini-Mental State Examination score  24) [14]; (2) only

mild spasticity in all joints of the affected limb (Modified
Stroke, the number one cause of neurological disability Ashworth Scale score < 3) [15]; and (3) a higher than fair
worldwide, is characterised by both cognitive and motor im- score on the Manual Muscle test [16].
pairments, as well as balance impairment, which may lead to The exclusion criteria were: (1) muscular-skeletal dis-
functional dependence and reduced quality of life [1]. order and operation of the lower extremities; (2) limited
In particular, because of abnormal muscle tone, coor- range of motion of the lower extremity; (3) unilateral
dination disorder, and associated reactions, hemiplegic neglect, hemianopsia, or apraxia; and (4) psychological or
patients experience a considerable decrease of motor emotional problems.
function in affected limbs [1]. The impaired lower ex- Thirty four patients with stroke met the criteria
tremity motor function commonly leads to functional limi- (Figure 1). The Research Ethics Committee of Eulji Uni-
tations and disability [2]. versity Hospital approved the study, and all participants
Damage to balance ability after a stroke decreases sta- provided written informed consent prior to inclusion in the
bility in the standing position, causing difficulty in walking study. After completion of the initial assessment, the par-
and performing functional activities. Hence, balance ability ticipants were randomly assigned to an experimental group
is the basis for independent movement and functional (n Z 17) or a control group (n Z 17). For randomisation,
performance [3]. sealed envelopes were prepared in advance and marked
Treatments to improve function of stroke patients inside with A or B, indicating the experimental group or the
include constraint-induced therapy [4], mental imagery control group respectively. The randomisation was done by
training [5], and robotic-assisted rehabilitation [6]. How- a third party who was completely unaware of the study
ever, most intervention protocols are labour intensive, content. The participants’ characteristics and all outcome
making the provision of intensive treatment for all patients measures before and after the treatment were assessed by
difficult. Physician 1, who was blinded to the treatment allocations.
Mirror therapy, with its low cost and simplicity, may be a The mirror therapy and sham therapy were performed by
suitable alternative [7,8]. Ramachandran and Rogers- Physician 2, who was not involved in the assessment of the
Ramachandran [9] introduced the use of visual illusions participants, in a closed room. Both physicians were
created by a mirror as a treatment for phantom pain. instructed not to communicate with the participants about
A mirror is placed in the patient’s midsagittal plane, the possible goals or the rationale of either treatment.
with the unaffected limb in front of it, so that the affected
limb is blocked and the patient can only see the reflection
of the intact limb [10]. Movement of the intact limb gives Intervention
the patient the illusion in which inputs are perceived
through the affected limb behind the mirror. Participants in the experimental group received mirror
Studies on the effectiveness of mirror therapy in stroke therapy and conventional rehabilitation therapy for a total
patients, including a 2012 Cochrane analysis [11], identified of 60 minutes (mirror therapy: 30 minutes; conventional
methodologically appropriate randomised controlled trials rehabilitation therapy: 30 minutes) per day, with a 10 mi-
[12]. In addition, Altschuler et al [13] implemented mirror nutes rest period halfway through the session. Participants
therapy with stroke patients and proved the beneficial ef- in the experimental group received training 5 days/week
fects of mirror therapy on the range of motion of upper for 4 weeks. Conventional rehabilitation therapy consists of
extremity joints, motion velocity, and accuracy. Sathian neurodevelopmental facilitation techniques. Subjects in
et al [8] found that, after 2 weeks of mirror therapy, a the control group received sham therapy and conventional
chronic stroke patient recovered hand function and grip rehabilitation therapy for a total of 60 minutes (sham
strength in the paretic limb. therapy: 30 minutes, conventional rehabilitation therapy:
In previous research, mirror therapy focused mainly on 30 minutes) per day on the same day.
the upper limb functions. However, since damage to lower Mirror therapy was performed using a modified version of
limb function after stroke affects the independent daily that described in the study by Sütbeyaz et al [17]. A mirror
functions, more studies are required to explore the issue of was mounted on a stand tilted toward the paretic side of
mirror therapy and balance ability which is very much the body to prevent the participant from viewing the
related to the lower limb functions. paretic limb. For the experimental group, the reflective
This study was conducted to examine the effect of mirror surface was kept facing the non-paretic leg. The exercises
therapy on balance ability in subacute patients (onset within performed in a semiseated position were: (1) hip-knee-
6 months after a stroke). ankle flexion; (2) knee extension with ankle dorsiflexion;
and (3) knee flexion beyond 90 degrees. The control group
performed the same exercise for the same duration as the
Methods experimental group, but the reflective side of the mirror
was covered with white fabric. Both the experimental and
Participants the control group performed only non-paretic lower limb
movements during mirror therapy. Participants did not
Patients (n Z 60) with stroke were screened for this study move their paretic limb. Second, for the balance training in
from March 2014 to August 2014. The inclusion criteria front of a mirror, a method proposed by Vaillant et al [18]
were: (1) sufficient cognitive ability to follow instructions was modified. The participants performed the following
 29

Figure 1. Study flowchart. CRT Z conventional rehabilitation therapy; MT Z mirror therapy; ST Z sham therapy.

actions: placing a chair in front of a full-length mirror and refers to the participant’s ability to maintain the vertical
sitting and standing up from the chair, and standing while axis of the body within a suitable range of the balance
keeping feet in parallel back and forth. The control group centre of the platform’s angle of tilt. A low BI score implies
performed the identical training in a space without a excellent balance ability [19].
mirror. In evaluation of balance ability, overall index captures
the change of overall movement, anterior/posterior sta-
Outcome measurement bility index captures the change on the sagittal plane, and
medial/lateral stability index captures the change on the
Balance Index frontal plane. That is, the stability index means the
Balance Index (BI) scores were obtained using a balance movement change of the platform, where a higher value
measurement system (Balance System SD, Biodex, NY, indicates more movement during the test.
USA), which incorporates a specific monitor and a movable The BI has a strong internal consistency, and acceptable
force platform, which provides up to 20 of surface tilt in a intrarater (r Z 0.82) and interrater (r Z 0.70) reliability
360 range of motion, with a visual feedback system. BI [19].
30 M.-K. Kim et al.

experiment. A significant difference in post-training gains

Table 1 General and medical characteristics of partici-
for the overall stability index and medial and lateral sta-
pants (n Z 34).
bility index was observed between the experimental group
EG CG t/X2 p and the control group (p < 0.05) (Table 2).
(n Z 17) (n Z 17)
Age (y) 52.4 54.6 0.548 0.453 Discussion
(7.2) (10.2)
Height (cm) 165.6 165.3 0.315 0.882 This study was conducted to investigate the effect of mirror
(6.4) (7.1) therapy on the balance in patients with subacute stroke.
Weight (kg) 65.6 67.7 0.432 0.478 According to the results, the treatment group showed a
(9.3) (6.9) significant decrease of degree in overall stability index and
Since onset (mo) 4.7 4.5 0.443 0.663 in-and-out stability index after treatment compared to the
(1.3) (1.1) control group, implying that mirror therapy is effective in
Sex 8/9 7/10 0.336 0.739 improving balance ability.
(male/female) (n) Several underlying mechanisms have been suggested for
Affected side 8/9 10/7 0.671 0.507 the effect of mirror therapy on motor recovery after stroke.
(left/right) (n) Stevens and Stoykov [20] suggested that mirror therapy is
Type of 12/5 14/3 0.792 0.434 related to motor imagery and that the mirror provides vi-
stroke (ischaemia/ sual feedback of the successful performance of the imag-
haemorrhage) (n) ined action with the affected limb. Motor imagery itself,
MMSE-K (score) 26.8 26.5 0.427 0.681 the mental performance of a movement without overt
(2.2) (1.9) execution of this movement, has proven to be beneficial in
CG Z control group; EG Z experimental group; MMSE- the rehabilitation of stroke patients [21].
K Z Korean version of the Mini-Mental State Examination. The effect of mirror therapy on brain activity has been
Values are presented as mean (SD) unless indicated otherwise. investigated in a number of studies. Galeazzi et al [22]
reported improvement of balance ability in stroke pa-
tients who performed exercise training in front of a full-
length mirror. Lynch [23] argued that instant self-analysis
Statistical analysis and correction for the patients are possible, as they visu-
ally confirm their own physical alignment through mirrors.
The general characteristics of the participants were tested Sütbeyaz et al [17] randomly assigned 40 stroke patients to
for homogeneity using the independent t test. Paired t tests a mirror therapy group and a control group. The partici-
were used to verify statistical significance in performances pants repeatedly performed dorsiflexion of the ankles with
before and after the experiment. An independent t test was the non-affected lower limb placed in front of a mirror. In
performed for comparison between groups. The statistical measurement of their functional ambulation categories,
significance level was set at a Z 0.05. the mirror therapy group achieved a statistically significant
higher level of improvement in gait ability than the control
group.
Results Functional magnetic resonance imaging (fMRI) images
related to mirror therapy found that the primary motor
As shown in the summary of clinical and demographic fea- cortex, which is involved in the movements of the other
tures of the sample (n Z 34) in Table 1, there were no side extremities, was excited by only observing the moving
significant differences in the baseline characteristics be- extremities in mirrors [24e26]. This finding shows that the
tween the two groups (p > 0.05). One participant in the functional organisation of motor systems is achieved not
control group gave up the training due to extremely poor only by the active movements of the ipsilateral extrem-
health. Therefore, 17 participants (experimental) and 16 ities, but also by passive observation of the movements of
participants (control) in each group completed this the contralateral extremities [25,26].

Table 2 Comparison of balance within and between groups (n Z 33) (unit: degree).
EG (n Z 17) CG (n Z 16) t p
Pre Post CWG Pre Post CWG
a
Overall 5.29 (0.79) 4.32 (1.12) 0.97 (0.59e1.34)* 5.31 (0.87) 5.08 (0.88) 0.23 (0.19e0.45)* 2.155 0.038
A/P 4.42 (0.43) 4.10 (0.51) 0.32 (0.10e0.54)* 4.43 (0.45) 4.27 (0.51) 0.16 (0.01e0.31)* 0.983 0.333
M/La 4.27 (0.42) 3.33 (0.67) 0.94 (0.67e1.21)* 4.06 (0.44) 3.89 (0.45) 0.17 (0.37e0.30)* 2.816 0.008
Values are presented as mean (SD) except for changes within groups, which are expressed as mean differences (95% confidence in-
tervals).
Overall Z overall stability index; A/P Z anterior and posterior stability index; M/L Z medial and lateral stability index; CG Z control
group; CWG Z changes within groups; EG Z experimental group.
a
Significant difference in gains between two groups, p < 0.05.
 31

In another study, Altschuler et al [27] reported that Conflicts of interest

application of mirror therapy to chronic stroke patients
caused an increase in range of motion and moving speed of All contributing authors declare that they have no conflicts
upper limbs, along with minute improvement of task per- of interest.
formance ability that uses fingers. Cacchio et al [28], who
applied mirror therapy to stroke patients, reported signifi-
cant improvement of upper limb functions, as indicated by Funding/support
the Motor Activity Log and Wolf motor function test scores.
Feltham et al [29] showed that visual feedback treatment
No financial or material support of any kind was received
using a mirror has positive effects on the bilateral coordi-
for the work described in this article.
nation ability and neuromuscular activity among spastic
hemiplegic patients.
These findings are generally in line with the results of
the current study. As a cognitive intervention method, References
mirror therapy causes activation of the frontal or parietal
lobe in the corresponding motor region, an area known as [1] Erkinjuntti T, Bowler JV, DeCarli CS, Fazekas F, Inzitari D,
O’Brien JT, et al. Imaging of static brain lesions in vascular
the mirror neurons, through only observing the behaviour of
dementia: implications for clinical trials. Alzheimer Dis Assoc
others [30]. Disord 1999;13(Suppl. 3):S81e90.
Among the assumptions regarding the functional role of [2] Luke C, Dodd KJ, Brock K. Outcomes of the Bobath concept on
mirror neuron, the first one is expressing actions related to upper limb recovery following stroke. Clin Rehabil 2004;18:
those entered into visual and auditory sense in exercise. 888e98.
This means understanding of the observed exercise goal and [3] Berg KO, Maki BE, Williams JI, Holliday PJ, Wood-
projecting it in exercise. The ventral premotor cortex and Dauphinee SL. Clinical and laboratory measures of postural
inferior parietal lobe control these reactions. The second balance in an elderly population. Arch Phys Med Rehabil 1992;
assumption is about is realising the conceived emotional 73:1073e80.
situation into motor reaction. Insula and rostral cingulate [4] Page SJ, Sisto S, Johnston MV, Levine P, Hughes M. Modified
constraint-induced therapy in subacute stroke: a case report.
control these reactions [30].
Arch Phys Med Rehabil 2002;83:286e90.
Therefore, mirror therapy could play an important role [5] Liu H, Song L, Zhang T. Changes in brain activation in stroke
in learning a new skill or understanding the behaviour of patients after mental practice and physical exercise: a func-
others, which can be used to motivate stroke patients [17]. tional MRI study. Neural Regen Res 2014;9:1474e84.
Mirror therapy is performed on the non-paralysed side by [6] Masiero S, Celia A, Rosati G, Armani M. Robotic-assisted
moving the limbs of stroke patients; the movement of rehabilitation of the upper limb after acute stroke. Arch Phys
paralysed limbs through the reflection in the mirror shows a Med Rehabil 2007;88:142e9.
visual illusion of normal movement. This method for [7] Michielsen ME, Selles RW, van der Geest JN, Smits M,
treating a brain injury is based on the principle of synaptic Ribbers GM, Bussmann JB. Motor recovery and cortical reor-
plasticity [17]. ganization after mirror therapy in chronic stroke patients: a
phase Ⅱ randomized controlled trial. Neurorehabil Neural
The current study has some limitations. First, the small
Repair 2011;25:223e33.
sample size may have influenced the results. Second, the [8] Sathian K, Greenspan AI, Wolf SL. Doing it with mirrors: a case
absence of follow-up after the end of the mirror therapy study of a novel approach to neurorehabilitation. Neuro-
does not allow for determination of the durability of the rehabil Neural Repair 2000;14:73e6.
effect of this therapy. Third, individual musculoskeletal [9] Ramachandran, Rogers-Ramachandran D. Synaesthesia in
and neurologic variations, which is important in patient phantom limbs induced with mirrors. Proc R Soc Lond B Bilo Sci
evaluation, was not considered. This can be a significant 1996;263:377e86.
factor affecting the research results regarding the [10] Dohle C, Pullen J, Nakaten A, Küst J, Rietz C, Karbe H. Mirror
improvement of balance ability. Finally, the study does not therapy promotes recovery from severe hemiparesis: a ran-
include diverse evaluation of the walking ability and inde- domized controlled trial. Neurorehabil Neural Repair 2009;23:
209e17.
pendent daily routine which are important in quality of life.
[11] Thieme H, Mehrholz J, Pohl M, Behrens J, Dohle C. Mirror
Therefore, these results can be generalised to subacute therapy for improving motor function after stroke. Cochrane
stroke patients. Further studies, including a long-term Database Syst Rev 2012;3:CD008449.
follow-up assessment, are needed to evaluate the long- [12] Lee MM, Cho HY, Song CH. The mirror therapy program en-
term benefits of mirror therapy. hances upper-limb motor recovery and motor function in
acute stroke patients. Am J Phys Med Rehabil 2012;91:
689e700.
[13] Altschuler EL, Sidney B, Wisdom SL, Foster C, Galasko D,
Conclusion Llewellyn DM, et al. Rehabilitation of hemiparesis after stroke
with a mirror. Lancet 1999;353:2035e6.
[14] Skidmore ER, Whyte EM, Holm MB, Becker JT, Butters MA,
Mirror therapy may be beneficial in increasing the balance
Dew MA, et al. Cognitive and affective predictors of rehabil-
ability of subacute stroke patients. The data should be itation participation after stroke. Arch Phys Med Rehabil 2010;
interpreted with caution due to the small sample size. A 91:203e7.
large-scale randomised controlled trial is required in order [15] Bohannon RW, Smith MB. Interrater reliability of a modified
to confirm our findings and further evaluate the long-term ashworth scale of muscle spasticity. Phys Ther 1987;67:
effects of mirror therapy in patients after stroke. 206e7.
32 M.-K. Kim et al.

[16] Aitkens S, Lord J, Bernauer E, Fowler Jr WM, Lieberman JS, [24] Luft AR, Smith GV, Forrester L, Whitall J, Macko RF,
Berck P. Relationship of manual muscle testing to objective Hauser TK, et al. Comparing brain activation associated
strength measurements. Muscle Nerve 1989;12:173e7. with isolated upper and lower limb movement across cor-
[17] Sütbeyaz S, Yavuzer G, Sezer N, Koseoglu BF. Mirror therapy responding joints. Hum Brain Mapp 2002;17:131e40.
enhances lower-extremity motor recovery and motor func- [25] Lipert J, Dettmers C, Terborg C, Weiller C. Inhibition of ipsi-
tioning after stroke: a randomized controlled trial. Arch Phys lateral motor cortex during phasic generation of low force.
Med Rehabil 2007;88:555e9. Clin Neurophysiol 2001;112:114e21.
[18] Vaillant J, Vuillerme N, Janvy A, Louis F, Juvin R, Nougier V. [26] Muellbacher W, Ziemann U, Boroojerdi B, Cohen L, Hallett M.
Mirror versus stationary cross feedback in controlling the Role of the human motor cortex in rapid motor learning. Exp
center of foot pressure displacement in quiet standing in Brain Res 2001;136:431e8.
elderly subjects. Arch Phys Med Rehabil 2004;85:1962e5. [27] Altschuler EL, Wisdom SB, Stone L, Foster C, Galasko D,
[19] Srivstava A, Taly AB, Gupta A, Kumar S, Murali T. Post-stroke Llewellyn DM, et al. Rehabilitation of hemiparesis after stroke
balance training : role of force platform with visual feedback with a mirror. Lancet 1999;353:2035e6.
technique. J Neurol Sci 2009;287:89e93. [28] Cacchio A, De blasis E, De Blasis V, Santilli V, Spacca G. Mirror
[20] Stevens JA, Stoykov ME. Using motor imagery in the rehabili- therapy in complex regional pain syndrome type 1 of the
tation of hemiparesis. Arch Phys Med Rehabil 2003;84:1090e2. upper limb in stroke patients. Neurorehabil Neural Repair
[21] Sharma N, Pomeroy VM, Baron JC. Motor imagery: a backdoor 2009;23:792e9.
to the motor system after stroke? Stroke 2006;37:1941e52. [29] Feltham MG, Ledebt A, Deconinck FJ, Savelsbergh GJ. Mirror
[22] Galeazzi GM, Monzani D, Gherpelli C, Covezzi R, Guaraldi GP. visual feedback induces lower neuromuscular activity in chil-
Posturographic stabilisation of healthy participants exposed dren with spastic hemiparetic cerebral palsy. Res Dev Disabil
to full-length mirror image is inversely related to body-image 2010;31:1525e35.
preoccupations. Neurosci Lett 2006;410:71e5. [30] Gallese V, Keysers C, Rizzolatti GA. Unifying view of the basis
[23] Lynch JA. The effect of mirror feedback in learning a frontal of social cognition. Trends Cogn Sci 2004;8:396e403.
plane motor skill on students in a pilates mat program.
Western Washington University; 2006.