Hindawi

Behavioural Neurology
Volume 2020, Article ID 6250524, 9 pages
https://doi.org/10.1155/2020/6250524

Clinical Study
Treatment Effects of Upper Limb Action Observation
Therapy and Mirror Therapy on Rehabilitation Outcomes after
Subacute Stroke: A Pilot Study

Yu-Wei Hsieh ,1,2,3 Yu-Hsuan Lin,4 Jun-Ding Zhu ,1 Ching-Yi Wu ,1,2,3 Yun-Ping Lin,5
and Chih-Chi Chen3,6
1
Department of Occupational Therapy and Graduate Institute of Behavioral Sciences, College of Medicine, Chang Gung University,
Taoyuan, Taiwan
2
Healthy Aging Research Center, Chang Gung University, Taoyuan, Taiwan
3
Department of Physical Medicine and Rehabilitation, Chang Gung Memorial Hospital, Linkou, Taiwan
4
Department of Physical Medicine and Rehabilitation, Cathay General Hospital, Taipei, Taiwan
5
Division of Occupational Therapy, Department of Physical Medicine and Rehabilitation, Taipei Hospital, Ministry of Health
and Welfare, New Taipei City, Taiwan
6
School of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan

Correspondence should be addressed to Yu-Wei Hsieh; [email protected]

Received 19 July 2019; Revised 18 November 2019; Accepted 12 December 2019; Published 2 January 2020

Academic Editor: Luigi Trojano

Copyright © 2020 Yu-Wei Hsieh et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. Action observation therapy and mirror therapy, two promising rehabilitation strategies, are aimed at enhancing the
motor learning and functional improvement of stroke patients through different patterns of visual feedback and observation.
Objective. This study investigated and compared the treatment effects of the action observation therapy, mirror therapy, and
active control intervention on motor and functional outcomes of stroke patients. Methods. Twenty-one patients with subacute
stroke were recruited in this study. All patients were randomly assigned to the action observation therapy, mirror therapy, or
active control intervention for 3 weeks. Outcome measures were conducted at baseline, immediately after treatment, and at
3-month follow-up. The primary outcome was the Fugl-Meyer Assessment, and secondary outcomes included the Box and
Block Test, Functional Independence Measure, and Stroke Impact Scale. Descriptive analyses and the number of patients whose
change score achieved minimal clinically important difference were reported. Results. Both the action observation therapy and
active control intervention showed similar improvements on the Fugl-Meyer Assessment, Box and Block Test, and Stroke
Impact Scale. Moreover, the action observation therapy had a greater improvement on the Functional Independence Measure
than the other 2 groups did. However, the mirror therapy group gained the least improvements on the outcomes. Conclusion.
The preliminary results found that the patients in the action observation therapy and active control intervention groups had
comparable benefits, suggesting that the 2 treatments might be used as an alternative to each other. A further large-scale study
with at least 20 patients in each group to validate the study findings is needed. This trial is registered with NCT02871700.

1. Introduction mize arm function and reduce disability in patients [4, 5].
Action observation therapy and mirror therapy are two
Stroke is a common cause of long-term disability among prominent approaches targeting stroke motor and functional
adults worldwide [1]. Most patients with stroke experience recovery, and both are supported by neuroscientific founda-
varying degrees of dysfunction, and one major dysfunction tions [6–9]. However, the relative treatment effects of the
is upper limb motor impairment [2, 3]. There is a need to upper limb action observation therapy versus mirror therapy
develop effective neurorehabilitation interventions to opti- in stroke rehabilitation have not been compared.
2 Behavioural Neurology

The action observation therapy is a developing rehabilita- aspects of clinical outcomes in patients with stroke as
tion approach based on the role of the mirror neuron system compared with a dose-matched active control intervention
in motor learning [6]. It is a motor-based technique with group (i.e., bilateral arm training).
cognitive strategies concerning stroke motor recovery [6].
The mirror neuron system is activated during both the execu- 2. Materials and Methods
tion and observation of an action and is the area responsible
for the action observation therapy [10, 11]. The action obser- 2.1. Research Design and Procedure. This was a three-arm,
vation therapy helps stroke patients improve motor skills single-blind, randomized controlled trial. We followed the
through observing another individual’s normal movements methods of Shih et al. which is Methods/design of this trial
and practicing what they have observed. During the action [31]. However, given the limited funding and study
observation therapy, the participants are commonly asked resources, we only applied and analysed some secondary out-
to carefully observe the actions performed by a healthy comes of the published study protocol [31] in this pilot study.
person in videos (i.e., the observation phase) and then to Each participant received intervention for 15 training ses-
physically practice the same actions (i.e., the execution sions (60 minutes/day, 5 days/week for 3 weeks). Treatment
phase). Neural reorganization and motor relearning of was provided by licensed occupational therapists who were
patients commonly occur in response to different afferent well trained in the treatment protocols. Furthermore, all
inputs and visual feedback [12]. The action observation outcome measures were administered to the patients by
therapy is aimed at promoting these processes to improve the same rater, who was blinded to the group allocation.
motor learning and performance in stroke patients. Recent Measurements were taken at three time points: baseline
studies have found positive effects of the action observation (T0), immediately after treatment (T1), and 3 months after
therapy on improving motor function and daily perfor- treatment (T2).
mance in stroke patients [13–18]. The participants were randomly allocated to 1 of the 3
The mirror therapy was initially developed for alleviating intervention groups in a 1 : 1 : 1 ratio. The two stratification
phantom limb pain after amputation and has been applied to factors were the severity of the upper limb motor deficits of
stroke rehabilitation in the past two decades [19, 20]. The the patients (Fugl-Meyer Assessment (FMA) score: 20–40
mirror therapy has gained much attention as a rehabilitation vs. 40–60 [32]) and the side of lesion (right vs. left) to ensure
strategy to address patients’ arm and hand function following baseline equivalence among the group. The randomization
a stroke [20–26]. During the mirror therapy, the participants was carried out using an online web-based randomization
are instructed to watch the reflection of movements of the tool (freely available at http://www.randomizer.org/). For
unaffected arm in a mirror as if it was the affected one. The concealment of allocation, the randomization procedure
mirror therapy creates the illusory visual image that the and assignment were managed by an independent research
intact arm is the affected arm and is normally moving, so as assistant who was not involved in screening or evaluation
to enhance the movements of the affected arm. It provides of the participants.
visual and proprioceptive feedback of the intact arm,
2.2. Participant Selection. A total of 21 patients with stroke
which may provide a substitute input for absent or
were recruited in this study (Figure 1). The inclusion criteria
reduced proprioceptive feedback from the affected arm
of the patients were the following: (1) diagnosis of cerebral
[27]. In addition, the mirror therapy might be associated
ischemic or hemorrhagic stroke, (2) 1 to 6 months since uni-
with the mirror neuron system and promote reorganiza-
lateral stroke onset, (3) age between 20 and 80 years, (4) base-
tion and functional recovery [25, 28]. A growing number
line score of the FMA between 20 and 60 [32], (5) ability to
of studies have shown that the mirror therapy could be
follow the study instructions (assessed by the Taiwan version
a beneficial approach for enhancing patients’ motor and
of the Montreal Cognitive Assessment) [33], and (6) ability
function after stroke [20, 21, 23–25, 29, 30].
to participate in study therapy and assessment sessions. The
Over the past few years, both the action observation ther-
patients were excluded if they had the following: (1) global
apy and mirror therapy have been considered to hold great
or receptive aphasia, (2) severe neglect, or (3) major medical
promise to promote the motor learning and functional recov-
problems or comorbidities that influenced the usage of the
ery of stroke patients. The two approaches involve different
upper limbs or caused severe pain. All participants provided
patterns of motor observation, imitation, and execution, but
written informed consent forms approved by the institu-
they share a similar neural basis in the mirror neuron system.
tional review boards of participating hospitals.
In the action observation therapy, patients are commonly
asked to observe the actions performed by another person 2.3. Interventions. During treatment, the therapists provided
and to execute the same actions, whereas in the mirror ther- verbal instructions, cues, feedback, and help to patients,
apy, they observe mirror reflections of the unaffected limb’s when needed. The study intervention (Figure 2) was addi-
movement as if it was the affected one. Despite the number tional therapy, and all routine conventional rehabilitation
of recent reports on these two approaches, no studies to date programs (e.g., occupational and/or physical therapy) have
have directly compared the effects of the upper limb action been provided as usual.
observation therapy and mirror therapy in patients with
stroke. This study aimed to investigate the immediate and 2.3.1. Action Observation Therapy. The patients in the action
retained (i.e., at 3-month follow-up) treatment effects of the observation therapy group were required to observe the
action observation therapy and mirror therapy on different upper limb movements or functional actions in video clips
Behavioural Neurology 3

Assessed for eligibility (n = 149)

Excluded (n = 128)
(i) Baseline FMA score did not
meet the criterion (n = 89)
(ii) Poor cognitive function (n = 9)
(iii) Global aphasia (n = 3)
Enrollment (iv) Refused to participate (n = 27)

Randomized (n = 21)

Allocated to action Allocated to mirror Allocated to active

Allocation observation therapy group therapy group control intervention group
(n = 7) (n = 7) (n = 7)

Posttest
outcome Completed posttest (n = 7) Completed posttest (n = 7) Completed posttest (n = 7)
measures

Follow-up Completed assessment Completed assessment Completed assessment

evaluation (n = 6) (n = 6) (n = 5)

Analysis Analyzed (n = 7) Analyzed (n = 7) Analyzed (n = 7)

Figure 1: Flowchart of participants.

Action observation therapy Mirror therapy Active control intervention

Figure 2: Demonstrations of view of action observation therapy, mirror therapy, and active control intervention during active range of
motion exercise.

(i.e., the observation phase) and to execute what they had first-person perspective to make the actions more intuitive
observed to the best of their ability (i.e., the execution phase). and facilitate optimal corticomotor excitability [34]. The
Three common categories of movements and tasks were actors in the videos were healthy young people. Observing
selected in the action observation therapy protocol based the actions from the first-person perspective means that the
on the related literature and clinical expertise: (a) upper limb observers watch the actions as if seeing through the actor’s
active range of motion (AROM) exercises, (b) reaching eyes. It looks like the observers are performing the actions
movement or object manipulation, and (c) upper limb func- themselves as the same directions and space dimensions of
tional tasks. The video movements were displayed from a the actors performed.
4 Behavioural Neurology

During phase 1 (10–15 minutes), the patients watched drawer with bilateral arm and hand movements. During
AROM exercises demonstrated in the video clips on a com- training, the patients were required to move both of their
puter screen and executed the observed exercises with both arms and hands simultaneously as possible. Based on the
arms and hands simultaneously. In phase 2 (15–20 minutes), patient’s level of motor ability and progress, the levels of
the patients were asked to observe one reaching movement or movement and task difficulty could be adjusted accordingly.
one object manipulation task, depending on the patient’s
motor ability, for 2 minutes in a video clip, and afterwards 2.4. Outcome Measures. The patients were assessed at base-
to execute the movements that they had observed for 3 line (T0), immediately after treatment (T1), and at 3 months
minutes. This sequence was repeated 3 times. The reaching after treatment (T2), and the FMA was used as the primary
movements involved reaching for objects of different sizes outcome measure. Secondary outcomes included the Box
and weights at different heights and locations. Object manip- and Block Test (BBT), Functional Independence Measure
ulation included in-hand manipulation, grasp and release, (FIM), and Stroke Impact Scale (SIS) version 3.0.
and transport and turning objects. Phase 3 (30 minutes) con-
2.4.1. Primary Outcome. The upper limb subscale of the
tained one functional task in each session, starting with easy
FMA, which has good psychometric properties, is used to
tasks and continuing with increasingly complex tasks. Each
assess motor impairments in patients [32, 35]. The 33 items
functional task was divided into 3 motor acts. For example,
measure the movements, reflexes, and coordination of the
the action of cleaning the mouth with a tissue paper was
upper limb joints. The maximum total score is 66, indicating
decomposed into the following 3 motor acts: (1) moving
full upper limb motor recovery. The total score of the FMA
hand toward a tissue paper, (2) taking a tissue paper, and
can also be separated into the subscores of the proximal
(3) bringing the tissue paper toward the mouth and wiping.
shoulder/elbow score (0 to 42) and the distal hand/wrist
After observing a motor act in a video clip for 2 minutes,
score (0 to 24). The total, proximal, and distal FMA scores
the patients were asked to execute the action they had
are reported.
observed for 3 minutes. For the next 15 minutes, the patients
observed the functional task as a complete action for 2 2.4.2. Secondary Outcomes. The BBT is a measure of manual
minutes and then executed the entire task for 3 minutes; this dexterity and consists of 150 colored wooden cubes in a box
sequence was repeated 3 times. Examples of the functional with 2 compartments. The subject uses his/her affected hand
tasks are folding a towel, wiping a table, drinking water, to move as many blocks as possible one-by-one from one
opening a small drawer, and using a mobile phone. compartment to the other within 1 minute. The number of
blocks transferred is counted as the BBT score. Its reliability
2.3.2. Mirror Therapy. During the mirror therapy, the and validity are satisfactory in stroke patients [36, 37].
patients were seated in front of a mirror box placed at their The FIM is a commonly used scale for assessing the per-
midsagittal plane. The affected arm of the participants was formance in basic daily activities. It contains 18 items (i.e., 13
placed inside the mirror box, and the unaffected arm was in motor and 5 cognition items), and the total score ranges from
front of the mirror. The patient was instructed to watch the 0 to 126 [38]. A higher score indicates greater independence
mirror reflection of the movement performed by his/her in daily activities. The total score and motor score of the FIM
unaffected hand carefully and to imagine that the movement are reported. The FIM has good reliability, validity, and
was performed by the affected hand. The participant was also responsiveness [39, 40].
encouraged to move the affected arm and hand as much as The SIS 3.0 is a stroke-specific, patient-reported ques-
they could. In the mirror therapy group, treatment activities tionnaire for evaluating the function, participation, and
also contained AROM exercises (10–15 minutes), reaching health-related quality of life of stroke survivors. The SIS 3.0
movement or object manipulation (15–20 minutes), and has sound psychometric properties [41]. It consists of 59
functional task practice (30 minutes). items, and higher scores indicate better function and greater
participation. The total SIS score and the average score of 4
2.3.3. Active Control Intervention—Customary Bilateral Arm physical functional domains of the SIS are reported in this
Training. The patients in the active control intervention study. In addition, an item assessing general perception of
group received dose-matched bilateral arm training provided recovery since stroke onset on a visual analogue scale of 0
by a certified occupational therapist, but no video input or to 100 was also used.
mirror box was provided for this group. In the active control
intervention, the same 3 categories of movements and tasks 2.5. Data Analysis. Baseline characteristics and outcomes of
as provided in the action observation therapy and mirror patients were compared among the 3 groups. This study
therapy groups were used. Treatment programs also included applied an intention-to-treat analysis. For missing data, the
AROM exercises (10–15 minutes), reaching movement or last observation carried forward method was used. Descrip-
object manipulation (15–20 minutes), and functional task tive analyses with change scores from T0 to T1 and from
practice (30 minutes). AROM exercises included bilateral T0 to T2 on each outcome were performed. Further, the
shoulder, elbow, forearm, wrist, and finger movements. number of patients whose change score achieved minimal
Object manipulation tasks were in-hand manipulation, grasp clinically important difference (MCID) on the clinical out-
and release, and transporting and turning objects with both comes in each intervention group was reported. The MCID
hands. Examples of functional tasks were reading a maga- value was adopted for the established value of the outcome
zine, folding clothes, wiping a table, and opening a small or was set as 10% of the maximum score while it has not been
Behavioural Neurology 5

Table 1: Baseline characteristics of the participants.

Action observation therapy Mirror therapy Active control intervention

P
(n = 7) (n = 7) (n = 7)
Gender (male/female) 6/1 6/1 6/1 1.0
Age (years) (mean (SD)) 52.77 (11.25) 46.41 (13.45) 54.30 (13.61) 0.49
Onset (months) (mean (SD)) 2.86 (1.77) 4.86 (1.95) 2.57 (1.72) 0.06
Lesion side (right/left) 2/5 4/3 3/4 0.85
Stroke type (hemorrhagic/ischemic) 4/3 7/0 2/5 0.03
Education (years) (mean (SD)) 12.29 (2.87) 11.86 (2.27) 11.43 (4.12) 0.88
MoCA score (mean (SD)) 24.57 (5.29) 25.29 (3.25) 24.00 (5.51) 0.88
FMA score (mean (SD)) 42.29 (11.03) 43.29 (13.72) 39.57 (6.55) 0.80
Abbreviations: FMA: Fugl-Meyer Assessment; MoCA: Taiwanese version of Montreal Cognitive Assessment.

established. Thus, the MCID values of the FMA total, BBT, T0 to T1 and 8.86, 5, and 8.71 from T0 to T2, respectively
FIM total, and SIS total scores were 6, 5, 22, and 10 points, (Table 2). In addition, there were 4, 2, and 4 patients achiev-
respectively [39, 42, 43]. In addition, two-way repeated ing the MCID of the BBT score after receiving the action
measures ANOVA was used to evaluate the treatment effects observation therapy, mirror therapy, and active control inter-
among the 3 treatment groups at 3 time points of assess- vention, respectively.
ment. The results of inferential statistics were reported in The mean changes in the FIM total scores in the
Supplementary Table 1. Statistical analyses were performed action observation therapy, mirror therapy, and active
in SPSS 19.0 (SPSS Inc., Chicago, Illinois). control intervention groups were 7.43, 2.14, and 2.57 from
T0 to T1 and 12.86, 5.86, and 6.14 from T0 to T2, respec-
3. Results tively (Table 2). In addition, there was 1 patient achieving
the MCID of the FIM total score after the action observa-
The baseline demographic and clinical data are summarized tion therapy, but no patient reached the MCID of the FIM
in Table 1. There were no statistically significant differences total score after the mirror therapy and active control
among the 3 groups except for the stroke type, indicating that intervention.
most of the baseline characteristics among the 3 groups were Furthermore, the mean changes in the SIS total scores in
comparable. In addition, no adverse effect was reported by the the action observation therapy, mirror therapy, and active
participants. control intervention groups were 8.3, 5.67, and 8.41 from
T0 to T1 and 12.83, 13.83, and 9.27 from T0 to T2, respec-
3.1. Primary Outcome. Table 2 shows that the mean changes tively (Table 2). In addition, there were 4, 2, and 4 patients
in the FMA total scores in the action observation therapy, achieving the MCID of the SIS total score after receiving
mirror therapy, and active control intervention groups were the action observation therapy, mirror therapy, and active
5.14, 2.57, and 7.14 from T0 to T1 and 4.43, 4.71, and 9.86 control intervention, respectively.
from T0 to T2, respectively. The mean changes in the FMA According to the descriptive and MCID results on sec-
proximal scores in the action observation therapy, mirror ondary outcomes, both the action observation therapy and
therapy, and active control intervention groups were 1.71, active control intervention had similar improvements on
1.29, and 3 from T0 to T1 and 2.14, 3.14, and 5 from T0 to the BBT and SIS scores. The action observation therapy
T2, respectively. Additionally, the mean changes in FMA dis- group had a greater improvement on the FIM total score than
tal scores in the action observation therapy, mirror therapy, the other 2 groups did. In addition, the patients in the mirror
and active control intervention groups were 3.43, 1.29, and therapy group had the least improvements in most outcomes
4.14 from T0 to T1 and 2.29, 1.57, and 4.86 from T0 to T2, compared with the other 2 groups.
respectively (Table 2).
In addition, there were 4, 1, and 5 patients achieving the 4. Discussion
MCID of the FMA total score after receiving the action
observation therapy, mirror therapy, and active control inter- This is the first study to examine and compare the treatment
vention, respectively. Based on the descriptive and MCID effects of the upper limb action observation therapy and mir-
results, the patients had similar improvements on the FMA ror therapy on different aspects of rehabilitation outcomes in
after receiving the action observation therapy and active con- patients with subacute stroke. Based on the descriptive and
trol intervention. The mirror therapy group gained the least MCID results, we found that both the action observation
improvement on the FMA among the 3 groups. Most of the therapy and active control intervention showed similar
improvements were maintained for 3 months. degrees of improvements on the FMA, BBT, and SIS scores.
In addition, patients in the action observation therapy group
3.2. Secondary Outcomes. The change scores of the BBT in improved most on the FIM score among the 3 groups.
the action observation therapy, mirror therapy, and active However, the mirror therapy group showed the least
control intervention groups were 6.14, 3.43, and 5.14 from improvements on the outcomes compared with the other 2
6 Behavioural Neurology

Table 2: Descriptive statistics for the primary and secondary outcomes.

Assessment time Change score

Variables Group
T0 T1 T2 T1-T0 T2-T0
Action observation therapy 42.29 (11.03) 47.43 (13.38) 46.71 (11.91) 5.14 (5.27) 4.43 (8.06)
FMA-total Mirror therapy 43.29 (13.72) 45.86 (14.66) 48.00 (15.20) 2.57 (3.82) 4.71 (4.50)
Active control intervention 39.57 (6.55) 46.71 (8.44) 49.43 (7.93) 7.14 (3.29) 9.86 (4.45)
Action observation therapy 30.86 (5.15) 32.57 (6.40) 33.00 (5.89) 1.71 (3.15) 2.14 (3.53)
FMA-proximal Mirror therapy 30.57 (5.97) 31.86 (5.87) 33.71 (6.37) 1.29 (3.25) 3.14 (4.26)
Active control intervention 30.86 (3.67) 33.86 (4.56) 35.86 (4.53) 3.00 (2.52) 5.00 (3.46)
Action observation therapy 11.43 (7.66) 14.86 (8.59) 13.71 (6.95) 3.43 (2.51) 2.29 (4.89)
FMA-distal Mirror therapy 12.71 (8.77) 14.00 (9.50) 14.29 (9.69) 1.29 (1.60) 1.57 (1.90)
Active control intervention 8.71 (5.28) 12.86 (5.15) 13.57 (5.29) 4.14 (2.04) 4.86 (2.79)
Action observation therapy 9.86 (11.39) 16.00 (11.14) 18.71 (10.08) 6.14 (5.98) 8.86 (7.29)
BBT Mirror therapy 17.28 (17.50) 20.71 (16.94) 22.29 (18.18) 3.43 (3.55) 5.00 (6.06)
Active control intervention 9.00 (7.66) 14.14 (11.61) 17.71 (14.43) 5.14 (4.22) 8.71 (7.76)
Action observation therapy 103.86 (22.61) 111.29 (11.54) 116.71 (10.87) 7.43 (13.16) 12.86 (17.71)
FIM-total Mirror therapy 110.14 (5.34) 112.29 (5.06) 116.00 (4.69) 2.14 (0.90) 5.86 (3.80)
Active control intervention 113.57 (5.47) 116.14 (4.85) 119.71 (4.50) 2.57 (1.90) 6.14 (4.81)
Action observation therapy 74.00 (16.45) 79.71 (8.73) 84.29 (6.85) 5.71 (8.69) 10.29 (13.43)
FIM-motor Mirror therapy 78.43 (6.05) 80.57 (5.35) 83.86 (3.34) 2.14 (0.90) 5.43 (3.51)
Active control intervention 82.00 (5.26) 84.00 (4.00) 86.00 (3.65) 2.00 (1.53) 4.00 (4.12)
Action observation therapy 64.00 (13.28) 72.31 (11.75) 76.83 (10.43) 8.30 (5.09) 12.83 (10.44)
SIS-total Mirror therapy 62.96 (9.93) 68.63 (7.44) 76.79 (6.25) 5.67 (7.01) 13.83 (10.59)
Active control intervention 68.85 (10.28) 77.26 (8.71) 78.12 (8.97) 8.41 (7.20) 9.27 (9.12)
Action observation therapy 53.37 (13.73) 65.85 (9.77) 72.12 (11.52) 12.49 (9.16) 18.75 (17.84)
SIS-physical function Mirror therapy 56.07 (14.05) 63.96 (14.07) 74.40 (13.83) 7.90 (7.68) 18.33 (11.27)
Active control intervention 62.03 (11.71) 74.29 (11.34) 74.69 (12.09) 12.26 (9.02) 12.66 (8.15)
Action observation therapy 51.43 (17.73) 58.57 (10.69) 58.57 (11.07) 7.14 (12.54) 7.14 (15.77)
SIS-recovery Mirror therapy 49.29 (15.92) 57.14 (14.68) 63.57 (11.07) 7.86 (3.93) 14.29 (8.38)
Active control intervention 61.43 (12.15) 70.71 (15.92) 65.71 (17.18) 9.29 (10.18) 4.29 (5.35)
Abbreviations: FMA: Fugl-Meyer Assessment; BBT: Box and Block Test; FIM: Functional Independence Measure; SIS: Stroke Impact Scale. Note: T0: baseline;
T1: immediately after treatment; T2: 3 months after treatment.

groups. During the period immediately after treatment and at therapy, only static images or pictures were provided during
3-month follow-up, the scores on most outcomes had been the observation phase, which may have given prominence to
maintained in the 3 groups, indicating retention effects. the experimental action observation therapy group. In the
Our preliminary findings showed that the patients in the present study, as compared with the dose-matched active
action observation therapy and active control intervention control intervention (i.e., bilateral arm training), the action
groups exhibited comparable improvements on most out- observation therapy group still gained comparable improve-
comes, but the patients in the mirror therapy group gained ments, suggesting that the 2 treatments might be used as an
the least improvements after treatment. alternative to each other.
The action observation therapy and active control inter- The action observation therapy group (observation plus
vention (i.e., bilateral arm training) had comparable effects execution) spent less time and effort on physical practice
on most rehabilitation outcomes assessed in this study. The and execution than the active control intervention and mir-
patients improved their motor impairment, manual dexter- ror therapy groups because some of the therapy time in the
ity, and quality of life after receiving the action observation action observation therapy group was the period of observa-
therapy and active control intervention. One explanation tion only. This difference indicates that the action observa-
for the similar improvements might be the active control tion therapy might be a suitable alternative for patients
intervention used in this study. Our active control interven- with poor physical endurance or those who have difficulty
tion was a type of bilateral arm training. However, most pre- engaging in mirror imagination to improve their function.
vious studies used sham action observation therapy as the Nevertheless, the action observation therapy requires much
control group [13, 17, 18]. In the sham action observation preparation. Video clips need to be made and prepared
Behavioural Neurology 7

before the action observation therapy is conducted, and video usual care group might be more suitable as a control group
content must be added or modified in line with the patients’ than is an active comparative group in further research with
progress and needs. By contrast, the active control interven- purposes similar to those of this pilot study.
tion is a relatively simple and convenient rehabilitation The limitations of this study need to be addressed. First,
approach with easier access than the action observation this pilot study might not have sufficient total sample size
therapy. Each treatment has its own advantages and limita- to detect significant interaction effects or group main effects.
tions, and therapists can select the most appropriate one for Future studies should endeavor to recruit larger sample sizes
their specific patients in particular practice settings. to compare and verify the treatment effects among the 3
We also found that the patients gained the least groups. About 20 subjects in each group (a total of 60) will
improvements after receiving the mirror therapy among the be required to have enough power to detect statistical signif-
3 intervention groups. There are some possible reasons for icance in a further study (power = 80%, effect size f = 0:42,
explaining the fewest gains from the mirror therapy. Firstly, and α = 0:05). Second, we did not select patients with partic-
the patients were instructed to perform movements and tasks ular stroke types or lesion sites, which may have increased the
within a mirror box during the mirror therapy. There were variability among the patients in this pilot study. Whether
limitations while executing movements in the limited space the stroke types or different lesion sites of patients (e.g., mir-
of a mirror box, such as allowing only a small range of ror neurons affected or not) influence the treatment efficacy
motion and limited types of activities. Secondly, the variabil- of the action observation therapy or mirror therapy will need
ity and range of motor deficits of patients (i.e., FMA baseline to be investigated in the future. In addition, early studies
score) in the mirror therapy group were higher than those in found that the initial motor function of stroke patients was
the other 2 groups which may also affect its efficacy in this a significant predictor or a critical determinant of motor
study. Thirdly, some previous studies also showed that the recovery after receiving rehabilitation therapies [49–51].
mirror therapy did not lead to better outcomes compared Whether the baseline severity of the upper limb motor defi-
with control interventions [44, 45]. Probably, the conflicts cits of the patients affects the treatment efficacy of the action
between vision and proprioception caused by not exactly observation therapy and mirror therapy needs to be further
the same movements between the mirror reflection of nonaf- investigated.
fected hand and the actual affected hand during the mirror
therapy may lead to unpleasant sensations and reduce its 5. Conclusions
treatment effect [44, 46].
We found that greater mean change scores and a higher In conclusion, both the action observation therapy and active
number of patients achieving the MCID values on the motor control intervention had similar degrees of improvements on
and functional outcomes were in the action observation ther- most outcomes. The mirror therapy group gained the least
apy group than in the mirror therapy group. In one previous improvements among the 3 groups. Based on the preliminary
study of lower limb stroke rehabilitation, the action observa- study findings, the action observation therapy is a promising
tion therapy group significantly improved static balance and alternative to the contemporary intervention of bilateral arm
gait function of patients, and the mirror therapy improved training (i.e., active control intervention) in patients with
gait function only [47]. One possible reason may be that subacute stroke. Further large-scale studies to examine
the motor learning process of action observation and imita- the clinical efficacy and neural mechanisms of the action
tion during the action observation therapy might be more observation therapy and mirror therapy in patients with
intuitive and straightforward than that during the mirror stroke are suggested.
therapy. Motor understanding and learning might be easier
for patients with stroke to imitate and learn via video obser- Data Availability
vation than via mirror visual feedback. However, the prelim-
inary results and the mechanisms underlying action Data are available from the corresponding authors on
observation therapy- and mirror therapy-induced clinical reasonable request.
improvements will require further elaboration.
In this pilot study, 85.9% of the patients screened were Conflicts of Interest
excluded. The major reason was that the baseline FMA
scores of the patients were too high or low (i.e., <20 or The authors certify that there is no conflict of interest.
>60). A further study should add different types of partici-
pating sites or hospitals, and it should also recruit partici- Funding
pants from additional sites with higher percentages of
patients with moderate to mild motor impairments (i.e., This study was supported by the Ministry of Science and
FMA scores between 20 and 60). In addition, the therapy Technology (MOST 105-2314-B-182-018 and MOST 106-
contents and programs of the control intervention might 2314-B-182-015-MY3) and partly supported by the Healthy
need to be further revised. This pilot study used bilateral Aging Research Center, Chang Gung University, from the
arm training, a known and effective treatment, as an active Featured Areas Research Center Program within the Frame-
control intervention. However, the design of applying an work of the Higher Education Sprout Project by the Ministry
active control intervention is more suitable for noninferiority of Education (EMRPD1I0451) in Taiwan and the Chang
clinical trials [48]. Thus, a conventional rehabilitation or Gung Memorial Hospital (BMRPD25) in Taiwan.
8 Behavioural Neurology

Supplementary Materials [15] E. Kim and K. Kim, “Effects of purposeful action observation
on kinematic patterns of upper extremity in individuals with
Supplementary Table 1: inferential statistics for the primary hemiplegia,” Journal of Physical Therapy Science, vol. 27,
and secondary outcomes. (Supplementary Materials) no. 6, pp. 1809–1811, 2015.
[16] M. Scott, S. Taylor, P. Chesterton, S. Vogt, and D. L. Eaves,
References “Motor imagery during action observation increases eccen-
tric hamstring force: an acute non-physical intervention,”
[1] G. A. Donnan, M. Fisher, M. Macleod, and S. M. Davis, Disability and Rehabilitation, vol. 40, no. 12, pp. 1443–1451,
“Stroke,” The Lancet, vol. 371, no. 9624, pp. 1612–1623, 2008. 2018.
[2] G. Kwakkel, B. J. Kollen, and R. C. Wagenaar, “Long term [17] P. Sale, M. G. Ceravolo, and M. Franceschini, “Action observa-
effects of intensity of upper and lower limb training after tion therapy in the subacute phase promotes dexterity recovery
stroke: a randomised trial,” Journal of Neurology, Neurosur- in right-hemisphere stroke patients,” BioMed Research Inter-
gery, & Psychiatry, vol. 72, no. 4, pp. 473–479, 2002. national, vol. 2014, Article ID 457538, 7 pages, 2014.
[3] E. S. Lawrence, C. Coshall, R. Dundas et al., “Estimates of the [18] D. Ertelt, S. Small, A. Solodkin et al., “Action observation has a
prevalence of acute stroke impairments and disability in a positive impact on rehabilitation of motor deficits after
multiethnic population,” Stroke, vol. 32, no. 6, pp. 1279– stroke,” NeuroImage, vol. 36, pp. T164–T173, 2007.
1284, 2001.
[19] V. S. Ramachandran, D. Rogers-Ramachandran, and S. Cobb,
[4] A. Pollock, B. St George, M. Fenton, and L. Firkins, “Top ten “Touching the phantom limb,” Nature, vol. 377, no. 6549,
research priorities relating to life after stroke,” The Lancet pp. 489-490, 1995.
Neurology, vol. 11, no. 3, p. 209, 2012.
[20] E. L. Altschuler, S. B. Wisdom, L. Stone et al., “Rehabilitation
[5] L. Brewer, F. Horgan, A. Hickey, and D. Williams, “Stroke of hemiparesis after stroke with a mirror,” The Lancet,
rehabilitation: recent advances and future therapies,” QJM, vol. 353, no. 9169, pp. 2035-2036, 1999.
vol. 106, no. 1, pp. 11–25, 2013.
[21] C. Y. Wu, P. C. Huang, Y. T. Chen, K. C. Lin, and H. W. Yang,
[6] G. Buccino, “Action observation treatment: a novel tool in
“Effects of mirror therapy on motor and sensory recovery in
neurorehabilitation,” Philosophical Transactions of the Royal
chronic stroke: a randomized controlled trial,” Archives of
Society of London: Series B, Biological Sciences, vol. 369,
Physical Medicine and Rehabilitation, vol. 94, no. 6,
no. 1644, 2014.
pp. 1023–1030, 2013.
[7] F. J. Deconinck, A. R. Smorenburg, A. Benham, A. Ledebt,
[22] D. Ezendam, R. M. Bongers, and M. J. Jannink, “Systematic
M. G. Feltham, and G. J. Savelsbergh, “Reflections on mirror
review of the effectiveness of mirror therapy in upper extrem-
therapy: a systematic review of the effect of mirror visual feed-
ity function,” Disability and Rehabilitation, vol. 31, no. 26,
back on the brain,” Neurorehabilitation and Neural Repair,
pp. 2135–2149, 2009.
vol. 29, no. 4, pp. 349–361, 2015.
[8] M. Franceschini, M. Agosti, A. Cantagallo, P. Sale, [23] H. Thieme, M. Bayn, M. Wurg, C. Zange, M. Pohl, and
M. Mancuso, and G. Buccino, “Mirror neurons: action obser- J. Behrens, “Mirror therapy for patients with severe arm pare-
vation treatment as a tool in stroke rehabilitation,” European sis after stroke–a randomized controlled trial,” Clinical Reha-
Journal of Physical and Rehabilitation Medicine, vol. 46, bilitation, vol. 27, no. 4, pp. 314–324, 2013.
no. 4, pp. 517–523, 2010. [24] C. Dohle, J. Pullen, A. Nakaten, J. Kust, C. Rietz, and H. Karbe,
[9] P. Sale and M. Franceschini, “Action observation and mirror “Mirror therapy promotes recovery from severe hemiparesis: a
neuron network: a tool for motor stroke rehabilitation,” Euro- randomized controlled trial,” Neurorehabilitation and Neural
pean Journal of Physical and Rehabilitation Medicine, vol. 48, Repair, vol. 23, no. 3, pp. 209–217, 2009.
no. 2, pp. 313–318, 2012. [25] M. E. Michielsen, R. W. Selles, J. N. van der Geest et al., “Motor
[10] C. Keysers and V. Gazzola, “Social neuroscience: mirror neu- recovery and cortical reorganization after mirror therapy in
rons recorded in humans,” Current Biology, vol. 20, no. 8, chronic stroke patients: a phase II randomized controlled
pp. R353–R354, 2010. trial,” Neurorehabilitation and Neural Repair, vol. 25, no. 3,
[11] G. Buccino, F. Binkofski, G. R. Fink et al., “Action observation pp. 223–233, 2011.
activates premotor and parietal areas in a somatotopic man- [26] C. Colomer, E. NOé, and R. Llorens, “Mirror therapy in
ner: an fMRI study,” The European Journal of Neuroscience, chronic stroke survivors with severely impaired upper limb
vol. 13, no. 2, pp. 400–404, 2001. function: a randomized controlled trial,” European Journal of
[12] A. M. Hughes, J. H. Burridge, S. H. Demain et al., “Translation Physical and Rehabilitation Medicine, vol. 52, no. 3, pp. 271–
of evidence-based assistive technologies into stroke rehabilita- 278, 2016.
tion: users' perceptions of the barriers and opportunities,” [27] K. Sathian, A. I. Greenspan, and S. L. Wolf, “Doing it with mir-
BMC Health Services Research, vol. 14, no. 1, 2014. rors: a case study of a novel approach to neurorehabilitation,”
[13] M. Franceschini, M. G. Ceravolo, M. Agosti et al., “Clinical rel- Neurorehabilitation and Neural Repair, vol. 14, no. 1, pp. 73–
evance of action observation in upper-limb stroke rehabilita- 76, 2000.
tion: a possible role in recovery of functional dexterity. A [28] D. Carvalho, S. Teixeira, M. Lucas et al., “The mirror neuron
randomized clinical trial,” Neurorehabilitation and Neural system in post-stroke rehabilitation,” International Archives
Repair, vol. 26, no. 5, pp. 456–462, 2012. of Medicine, vol. 6, no. 1, p. 41, 2013.
[14] K. Sugg, S. Müller, C. Winstein, D. Hathorn, and A. Dempsey, [29] S. Hoermann, L. Ferreira dos Santos, N. Morkisch et al., “Com-
“Does action observation training with immediate physical puterised mirror therapy with augmented reflection technol-
practice improve hemiparetic upper-limb function in chronic ogy for early stroke rehabilitation: clinical feasibility and
stroke?,” Neurorehabilitation and Neural Repair, vol. 29, integration as an adjunct therapy,” Disability and Rehabilita-
no. 9, pp. 807–817, 2015. tion, vol. 39, no. 15, pp. 1503–1514, 2016.
Behavioural Neurology 9

[30] H. Thieme, N. Morkisch, J. Mehrholz et al., “Mirror therapy with minimal to moderate impairment due to chronic stroke,”
for improving motor function after stroke,” Cochrane Data- Physical Therapy, vol. 92, no. 6, pp. 791–798, 2012.
base of Systematic Reviews, no. 7, article CD008449, 2018. [44] L. C. Rodrigues, N. C. Farias, R. P. Gomes, and S. M. Michael-
[31] T. Y. Shih, C. Y. Wu, K. C. Lin et al., “Effects of action sen, “Feasibility and effectiveness of adding object-related
observation therapy and mirror therapy after stroke on reha- bilateral symmetrical training to mirror therapy in chronic
bilitation outcomes and neural mechanisms by MEG: study stroke: a randomized controlled pilot study,” Physiotherapy
protocol for a randomized controlled trial,” Trials, vol. 18, Theory and Practice, vol. 32, no. 2, pp. 83–91, 2016.
no. 1, p. 459, 2017. [45] W. C. Chan and S. S. Y. Au-Yeung, “Recovery in the severely
[32] A. R. Fugl-Meyer, L. Jaasko, I. Leyman, S. Olsson, and impaired arm post-stroke after mirror therapy: a randomized
S. Steglind, “The post-stroke hemiplegic patient. 1. A method controlled study,” American Journal of Physical Medicine &
for evaluation of physical performance,” Scandinavian Journal Rehabilitation, vol. 97, no. 8, pp. 572–577, 2018.
of Rehabilitation Medicine, vol. 7, no. 1, pp. 13–31, 1975. [46] C. McCabe, “Mirror visual feedback therapy. A practical
[33] C. F. Tsai, W. J. Lee, S. J. Wang, B. C. Shia, Z. Nasreddine, and approach,” Journal of Hand Therapy, vol. 24, no. 2, pp. 170–
J. L. Fuh, “Psychometrics of the Montreal Cognitive Assess- 179, 2011.
ment (MoCA) and its subscales: validation of the Taiwanese
[47] H. J. Lee, Y. M. Kim, and D. K. Lee, “The effects of action
version of the MoCA and an item response theory analysis,”
observation training and mirror therapy on gait and balance
International Psychogeriatrics, vol. 24, no. 4, pp. 651–658,
in stroke patients,” Journal of Physical Therapy Science,
2012.
vol. 29, no. 3, pp. 523–526, 2017.
[34] F. Maeda, G. Kleiner-Fisman, and A. Pascual-Leone, “Motor
facilitation while observing hand actions: specificity of the [48] J. Schumi and J. T. Wittes, “Through the looking glass: under-
effect and role of observer's orientation,” Journal of Neurophys- standing non-inferiority,” Trials, vol. 12, no. 1, 2011.
iology, vol. 87, no. 3, pp. 1329–1335, 2002. [49] M. Brunetti, N. Morkisch, C. Fritzsch et al., “Potential determi-
[35] Y. W. Hsieh, C. Y. Wu, K. C. Lin, Y. F. Chang, C. L. Chen, and nants of efficacy of mirror therapy in stroke patients–a pilot
J. S. Liu, “Responsiveness and validity of three outcome mea- study,” Restorative Neurology and Neuroscience, vol. 33,
sures of motor function after stroke rehabilitation,” Stroke, no. 4, pp. 421–434, 2015.
vol. 40, no. 4, pp. 1386–1391, 2009. [50] Y. Y. Lee, Y. W. Hsieh, C. Y. Wu, K. C. Lin, and C. K. Chen,
[36] K. C. Lin, L. L. Chuang, C. Y. Wu, Y. W. Hsieh, and W. Y. “Proximal Fugl-Meyer assessment scores predict clinically
Chang, “Responsiveness and validity of three dexterous func- important upper limb improvement after 3 stroke rehabilita-
tion measures in stroke rehabilitation,” Journal of Rehabilita- tive interventions,” Archives of Physical Medicine and Rehabil-
tion Research and Development, vol. 47, no. 6, pp. 563–571, itation, vol. 96, no. 12, pp. 2137–2144, 2015.
2010. [51] K. C. Lin, Y. H. Huang, Y. W. Hsieh, and C. Y. Wu, “Potential
[37] V. Mathiowetz, G. Volland, N. Kashman, and K. Weber, predictors of motor and functional outcomes after distributed
“Adult norms for the Box and Block Test of manual dexterity,” constraint-induced therapy for patients with stroke,” Neuror-
The American Journal of Occupational Therapy, vol. 39, no. 6, ehabilitation and Neural Repair, vol. 23, no. 4, pp. 336–342,
pp. 386–391, 1985. 2009.
[38] B. B. Hamilton, C. V. Granger, F. S. Sherwin, M. Zielezny, and
J. S. Tashman, “A uniform national data system for medical
rehabilitation,” in Rehabilitation Outcomes: Analysis and
Measurement, M. J. Fuhrer, Ed., pp. 137–147, Paul H. Brookes
Publishing Company, Baltimore, 1987.
[39] M. Beninato, K. M. Gill-Body, S. Salles, P. C. Stark, R. M.
Black-Schaffer, and J. Stein, “Determination of the minimal
clinically important difference in the FIM instrument in
patients with stroke,” Archives of Physical Medicine and
Rehabilitation, vol. 87, no. 1, pp. 32–39, 2006.
[40] D. Wallace, P. W. Duncan, and S. M. Lai, “Comparison of the
responsiveness of the Barthel Index and the motor component
of the Functional Independence Measure in stroke: the impact
of using different methods for measuring responsiveness,”
Journal of Clinical Epidemiology, vol. 55, no. 9, pp. 922–928,
2002.
[41] P. W. Duncan, R. K. Bode, S. Min Lai, and S. Perera, “Rasch
analysis of a new stroke-specific outcome scale: the stroke
impact scale,” Archives of Physical Medicine and Rehabilita-
tion, vol. 84, no. 7, pp. 950–963, 2003.
[42] H. M. Chen, C. C. Chen, I. P. Hsueh, S. L. Huang, and
C. L. Hsieh, “Test-retest reproducibility and smallest real
difference of 5 hand function tests in patients with stroke,”
Neurorehabilitation and Neural Repair, vol. 23, no. 5,
pp. 435–440, 2009.
[43] S. J. Page, G. D. Fulk, and P. Boyne, “Clinically important dif-
ferences for the upper-extremity Fugl-Meyer scale in people
 MEDIATORS of

INFLAMMATION

The Scientific Gastroenterology Journal of

World Journal
Hindawi Publishing Corporation
Research and Practice
Hindawi
Hindawi
Diabetes Research
Hindawi
Disease Markers
Hindawi
www.hindawi.com Volume 2018
http://www.hindawi.com
www.hindawi.com Volume 2018
2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal of International Journal of

Immunology Research
Hindawi
Endocrinology
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Submit your manuscripts at

www.hindawi.com

BioMed
PPAR Research
Hindawi
Research International
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal of
Obesity

Evidence-Based
Journal of Stem Cells Complementary and Journal of
Ophthalmology
Hindawi
International
Hindawi
Alternative Medicine
Hindawi Hindawi
Oncology
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013

Parkinson’s
Disease

Computational and
Mathematical Methods
in Medicine
Behavioural
Neurology
AIDS
Research and Treatment
Oxidative Medicine and
Cellular Longevity
Hindawi Hindawi Hindawi Hindawi Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018