Received: 29 September 2018 Revised: 27 February 2019 Accepted: 17 May 2019

DOI: 10.1002/pri.1792

RESEARCH ARTICLE

Unilateral dorsiflexor strengthening with mirror therapy to

improve motor function after stroke: A pilot randomized study

Daniel Simpson1 | Monika Ehrensberger1 | Frances Horgan2 | Catherine Blake3 |

David Roberts1 | Patrick Broderick1 | Kenneth Monaghan1

1
Clinical Health & Nutrition Centre (CHANCE),
School of Science, Institute of Technology, Abstract
Sligo, Ireland
Background: Independently, cross‐education, the performance improvement of
2
School of Physiotherapy, RCSI (Royal College
the untrained limb following unilateral training, and mirror therapy have shown to
of Surgeons in Ireland), Dublin, Ireland
3
School of Public Health, Physiotherapy and improve lower limb functioning poststroke. Mirror therapy has shown to augment
Sport Science, University College Dublin, the cross‐education effect in healthy populations. However, this concept has not yet
Dublin, Ireland
been explored in a clinical setting.
Correspondence
Objectives: This study set out to investigate the feasibility and potential efficacy
Daniel Simpson, Clinical Health & Nutrition
Centre (CHANCE), School of Science, Institute of applying cross‐education combined with mirror therapy compared with cross‐
of technology, Ash Lane, Ballinode, Sligo,
education alone for lower limb recovery poststroke.
Ireland.
Email: daniel.simpson@mail.itsligo.ie Methods: Thirty‐one chronic stroke participants (age 61.7 ± 13.3) completed either

Funding information
Irish Research Council Postgraduate Scholar-
ship, Grant/Award Number: GOIPG/2016/
1662; IT Sligo President's Bursary Fund; IT
Sligo Capacity Building Fund; Institutes of
Technology Ireland Postgraduate Research
Scholarship
a unilateral strength training (ST; n = 15) or unilateral strength training with mirror‐
therapy (MST; n = 16) intervention. Both groups isometrically strength trained the
less‐affected ankle dorsiflexors three times per week for 4 weeks. Only the MST
group observed the mirror reflection of the training limb. Patient eligibility, com-
pliance, treatment reliability, and outcome measures were assessed for feasibility.
Maximal voluntary contraction (MVC; peak torque, rate of torque development, and
average torque), 10‐m walk test, timed up and go (TUG), Modified Ashworth Scale
(MAS), and the London Handicap Scale (LHS) were assessed at pretraining and
posttraining.
Results: Treatment and assessments were well tolerated without adverse effects.
No between group differences were identified for improvement in MVC, MAS,
TUG, or LHS. Only the combined treatment was associated with functional improve-
ments with the MST group showing an increase in walking velocity.
Conclusion: Cross‐education plus mirror therapy may have potential for improving
motor function after stroke. This study demonstrates the feasibility of the combina-
tion treatment and the need for future studies with larger sample sizes to investigate
the effectiveness of the treatment.

K E Y W OR D S

exercise, mobility, physiotherapy, spasticity, stroke

Trial Registration: ClinicalTrials.gov. NCT03497650

Physiother Res Int. 2019;e1792. wileyonlinelibrary.com/journal/pri © 2019 John Wiley & Sons, Ltd. 1 of 9
https://doi.org/10.1002/pri.1792
2 of 9
1 | I N T RO D U CT I O N
Worldwide, 17 million people suffer a stroke annually (Benjamin et al.,
2017), with hemiparesis and spasticity the most common residual
physical impairments (McElwaine, McCormack, & Harbison, 2015;
Urban et al., 2010). Hemiparesis occurs predominantly in distal muscle
groups (Patten, Lexell, & Brown, 2004) and spasticity develops most
frequently in the ankle, resulting in a strength deficit (Barzi & Zehr,
2008) and impaired gait (Wissel et al., 2010). Motor recovery is attrib-
uted to neuro‐plastic adaptations, activation of dormant neurons,
formation of new synapses and pathways, and improved efficiency
of existing networks (Pekna, Pekny, & Nilsson, 2012). Additionally,
correct sensory feedback from the paretic limb and normalized
excitatory‐inhibitory balance between the two brain hemispheres are
important for motor recovery (Howatson, Zult, Farthing, Zijdewind,
& Hortobagyi, 2013; Michielsen et al., 2011; Rossiter, Borrelli,
Borchert, Bradbury, & Ward, 2015).
Cross‐education and mirror therapy are rehabilitation techniques
founded upon the principles of neuroplasticity. Cross‐education, the
performance improvement in the untrained homologous muscle after
unilateral exercise (Farthing, 2009), has beneficial rehabilitative out-
comes in the lower extremity following stroke (Ehrensberger, Simpson,
Broderick, & Monaghan, 2016). To date, only one study has investi-
gated cross‐education in the lower limb poststroke, reporting strength
increases of 34% in the trained limb and 31% in the untrained limb
following maximal isometric dorsiflexion strength training with the
less‐affected limb only (Dragert & Zehr, 2013). The study also found
a positive effect on functional mobility with improvements in timed
up and go (TUG) scores. Although a meta‐analysis of 31 cross‐
education studies reports a more modest average strength increase
of 16% of initial strength for the untrained lower limb (Manca,
Dragone, Dvir, & Deriu, 2017), high intensity unilateral resistance
training is believed to increase excitability in the untrained primary
motor cortex (M1), mediate synaptic connectivity within neural circuits,
and enhance neural drive to the contralateral homologous muscle,
improving force output (Hendy & Lamon, 2017; Hortobagyi, 2005;
Lee & Carroll, 2007). Because contralateral strength gains are facilitated
via neural pathways damaged during stroke, cross‐education seems
appropriately suited to stroke rehabilitation (Zult, Howatson, Kadar,
Farthing, & Hortobagyi, 2014). Mirror therapy has also been shown to
aid motor recovery and improve walking velocity and passive range of
motion for ankle dorsiflexion after stroke (Broderick et al., 2018;
Thieme et al., 2018). Based on visual stimulation, mirror therapy has
demonstrated to activate the mirror neuron system (Carvalho et al.,
2013) and enhance bilateral M1 excitability (Rossiter et al., 2015).
Mirror visual feedback of the training limb overrides proprioception
and increases attention of the resting limb, further enhancing
activation of the untrained hemisphere (Deconinck et al., 2015;
Touzalin‐Chretien, Ehrler, & Dufour, 2010). Additionally, observing
the mirror reflection of the less‐affected limb creates the illusion of
a normal functioning more‐affected limb resulting in appropriate
sensory feedback, vital for achieving poststroke motor recovery
(Rossiter et al., 2015).
SIMPSON ET AL.
Because cross‐education and mirror therapy have individually
shown to induce overlapping neuro‐plastic adaptations resulting in
motor recovery, Howatson et al. (2013) postulated that unilateral
strength training of the less‐affected limb combined with mirror visual
feedback of the training limb may increase neural activation of the
untrained hemisphere, resulting in an enhanced training effect in
stroke patients. Although the optimum dose point of saturation for
each individual treatment and the degree to which the combined
treatment may increase, the threshold of neural activation in the
affected hemisphere leading to recovery of a hemiparetic limb is not
yet established. Zult et al. (2016), however, recently illustrated how
the cross‐education effect in a nonclinical population can indeed be
further augmented by including mirror therapy during unilateral
strength training. Untrained limb strength increased significantly
more in a mirror and strength training group (61%) compared with a
nonmirror group (34%) with a significant reduction (12%) in contralat-
eral silent period in the mirror group, suggesting a mirror‐induced
reduction in intracortical inhibition within the untrained hemisphere.
Thus, cross‐education and mirror‐therapy interventions have the
capacity to influence brain plasticity beneficial to poststroke recovery
but have not yet been applied in combination to the lower limb in a
stroke population. Therefore, the primary aim of this study was to
investigate the feasibility of unilateral strength training plus mirror
therapy in the lower limb poststroke. The primary feasibility objectives
were assessing patient eligibility, intervention compliance, treatment
reliability, and outcome measures. The secondary aim of this study
was to investigate the potential efficacy of cross‐education coupled
with mirror therapy on motor performance compared with cross‐
education alone. The authors hypothesized that unilateral isometric
ankle dorsiflexion strengthening applied to the less‐affected limb,
combined with mirror visual feedback of the training limb, would
result in greater strength transfer and improved motor function in
the more‐affected (untrained) limb compared with unilateral strength
training alone.
2 | METHODS
2.1 | Study design
This study followed a randomized controlled parallel group design with
assessor blinded to treatment allocation. Ethical approval was granted
by Sligo University Hospital Research Ethics Committee and regis-
tered with ClinicalTrials.gov: NCT03497650.
2.2 | Participants
Between January 2015 and March 2016, clinicians at Sligo University
Hospital identified potential participants that met the study inclusion
criteria. Inclusion criteria were adults presenting with lower limb
poststroke hemiparesis, at least 12 months' poststroke (diagnosed by
a physician), discharged from formal rehabilitation services, and not
involved in any other lower limb rehabilitation or strength training
SIMPSON ET AL.
during the study. Exclusion criteria were cardiovascular, neurological,
or musculoskeletal impairments not related to stroke that would pre-
vent strength training, impaired cognition (MMSE <24) and vision
impairments (identified by the referring clinician) that would interfere
with the ability to observe mirror images.
Thirty‐six subjects were contacted for eligibility, one was excluded
for not meeting inclusion criteria and 35 were assessed for baseline
measures. All participants signed informed consent according to the
Declaration of Helsinki. Following baseline assessment, computer‐
generated block‐random numbers were used to randomly assign
participants to either the control group, which performed strength
training only (ST; n = 17), or the experimental mirror and strength
training group (MST; n = 18; Figure 1). An independent assistant
conducted the randomization with notification delivered in numbered
opaque sealed envelopes.
2.3 | Procedure
An assessor, blinded to treatment assignment, performed all pre-
assessment and postassessments at the Institute of Technology Sligo
exercise physiology laboratory. Baseline measurements (T1) were
obtained no more than 7 days prior to the start of the intervention
with postintervention (T2) measurements taken 48 hr after interven-
tion completion.
2.4 | Intervention
The strength training only (ST) intervention consisted of a home‐based
isometric unilateral strength training programme, applied to the less‐
affected limb only, and performed under therapist supervision. The
FIGURE 1 Flow diagram of participant
recruitment and study procedure
3 of 9
warm‐up consisted of 1 min of dynamic dorsiflexion contractions at
a self‐selected speed without resistance, followed by five unilateral
submaximal (<50% maximal voluntary contraction [MVC]) isometric
contractions of the less‐affected limb (Holmback, Porter, Downham,
& Lexell, 1999). To facilitate an isometric training mode, the par-
ticipant's less‐affected lower limb was strapped into an ankle brace
securing the ankle joint at 10° plantarflexion (Dragert & Zehr, 2013).
Participants were seated with back support, with a knee joint at
120°. The strength training programme was designed in line with pre-
vious cross‐education studies in clinical populations (Dragert & Zehr,
2013; Magnus et al., 2013) and maximal strength training guidelines
(ACSM, 2009; Bird, Tarpenning, & Marino, 2005). Training consisted
of four sets of five maximal effort isometric ankle dorsiflexion contrac-
tions, performed with the less‐affected limb only, held for 5 s with 5‐s
rest between repetitions and a 3‐min rest between sets. The same
protocol was followed three times per week for 4 weeks.
The MST group performed the same unilateral strengthening pro-
tocol as the ST group with the addition of the mirror placed in the
participant's midsagittal plane (Figure 2). The MST group observed
the reflection of the training limb in the mirror. Prompts to focus on
the reflection were given to the MST group only; other verbal cues
were identical for all participants of both groups.
2.5 | Evaluation and measurement
2.5.1 | Feasibility outcomes
We evaluated four primary objectives to assess the feasibility of
conducting the cross‐education plus mirror‐therapy protocol. Patient
eligibility was assessed by determining the percentage of referred
patients eligible for the study. Participant compliance was assessed
Colour online, B&W in print
 4 of 9 SIMPSON ET AL.

Medical Systems Inc, 2006). Following familiarization, one set of four

MVC was performed by the less‐affected side followed by the more‐
affected side. Of the four contractions, the contractions with the
highest peak torque (PT), highest rate of torque development (RTD),
and highest average torque (AT) of a single contraction were identified
for analysis.
Secondary, spasticity was assessed according to the valid and
Colour online, B&W in print

reliable Modified Ashworth Scale (MAS; scored 0–4 with 0 indicating

FIGURE 2 The MST participant is set up with the training (less
affected) limb strapped into the isometric strengthening ankle brace
with the reflection of the same limb in the mirror placed between the
participant's legs. The more‐affected limb is hidden behind the mirror
and placed in the same position as the training limb. The ST group
exercised without a mirror entirely; the MST group observed the
reflection of the training (less affected) limb in the mirror while training
normal tone), with a minimal clinically important difference (MCID)
of 1 point (van Wijck, Pandyan, Johnson, & Barnes, 2001; Zipp & Sul-
livan, 2010). The valid and reliable TUG (Faria, Teixeira‐Salmela, Neto,
& Rodrigues‐de‐Paula, 2012), with a minimal detectable change of
2.9 s or 23% (Flansbjer, Holmback, Downham, Patten, & Lexell,
2005), was used to measure basic mobility and balance manoeuvres.
The valid and reliable 10‐m walk test (10MWT; Collen, Wade, &
Bradshaw, 1990; Wolf et al., 1999), with an MCID of 0.06 m s−1
(Perera, Mody, Woodman, & Studenski, 2006), was used to assess
walking velocity. The valid and reliable London Handicap Scale (LHS;
scored 0–1 with 0 indicating most severe disadvantage and 1 indicat-
ing no disadvantage; Park & Choi, 2014) was used to assess self‐
perceived participation.
2.6 | Statistical analysis

by recording training attendance and retention. Preassessment and
postassessment were evaluated by determining the number of assess-
ments participants were able to complete. Treatment reliability was
assessed by recording any adverse events during the intervention.
2.5.2 | Effectiveness outcomes
Effectiveness measures were in accordance to the three levels of
human functioning as outlined in the International Classification of
Functioning, Disability and Health framework (World Health Organiza-
tion, 2001): The primary effectiveness outcome of isometric
dorsiflexion MVC was assessed using the Biodex System 3 Isokinetic
Dynamometer (Biodex Medical Systems Inc., Shirley, NY, USA). Iso-
metric MVC was assessed in stocking feet at 10° plantarflexion with
120° knee flexion (Dragert & Zehr, 2013) and 75° hip flexion (Biodex
Because formal power calculations are not strictly required for a pilot
study (Whitehead, Julious, Cooper, & Campbell, 2016), a minimum of 15
participants per treatment arm was considered a sufficient sample size.
Data were analysed using IBM SPSS for Windows (Version 20,
Chicago, IL, USA). Sample demographics (Table 1) and outcome mea-
sures (Table 2) are described in mean ± SD. Demographic characteris-
tics were tested for between group differences for using the
independent t test, the Mann–Whitney U test, and the chi‐square test.
Data analysis and interpretation were guided by a recent study with
comparable design (Carroll, Volpe, Morris, Saunders, & Clifford,
2017). When tested, certain variables within this pilot study did not
conform to assumptions for normality. Consequently, between group
differences in change over the intervention were tested for using
the independent t test or Mann–Whitney U test as appropriate. Simi-
larly, changes in within group means were analysed using the paired

TABLE 1 Demographic characteristics of participants at baseline mean ± SD (range)

Characteristic ST group baseline (n = 15) MST group baseline (n = 16) ST versus MST differences (p)

Sex, M:F 11:4 9:7 0.54

Age (years) 63.5 ± 12.0 (36–80) 60.0 ± 14.7 (32–90) 0.48
Type of stroke
Ischaemic:haemorrhagic 9:6 11:5 0.61
Time since stroke (months) 90.1 ± 83.3 (16–276) 78.7 ± 75.2 (6–207) 0.48
MA side right:left 8:7 6:10 0.60
Trained side
Dominant:nondominant 7:8 9:7 0.86

Abbreviations: ST, strength training only group; MST, mirror and strength training group; M, male; F, female; MA, more affected; n, number of participants.
 SIMPSON
ET AL.

TABLE 2 Results in mean ± SD, within group and between group differences for all outcome measures

ST group (n = 15) MST group (n = 16)

Within group difference Within group difference Between group difference

T1 mean ± SD T2 mean ± SD P (ES) T1 mean ± SD T2 mean ± SD P (ES) P (ES)

Trained ankle (LA) MVC

PT (Nm) 30.01 ± 7.91 (n = 14)a 30.01 ± 9.31 (n = 14)a 0.993 (0.0)c 24.63 ± 10.57 (n = 16) 23.87 ± 11.59 (n = 16) 0.438 (0.2)c 0.803 (0.05)b
−1 a a c b
RTD (Nm s ) 87.75 ± 41.11 (n = 14) 87.75 ± 45.59 (n = 14) 1.000 (0.0) 84.63 ± 42.59 (n = 16) 86.44 ± 45.53 (n = 16) 0.605 (0.09) 0.852 (0.03)b
AT (Nm) 26.42 ± 7.57 (n = 14)a 26.79 ± 8.55 (n = 14)a 0.633 (0.1)c 20.91 ± 10.48 (n = 16) 20.96 ± 10.74 (n = 16) 0.965 (0.0)c 0.950 (0.01)b
Untrained ankle (MA) MV
PT (Nm) 15.86 ± 15.61 (n = 14)a 15.79 ± 16.15 (n = 14)a 0.756 (0.06)b 14.56 ± 11.93 (n = 15)a 15.37 ± 12.26 (n = 15)a 0.160 (0.4)c 0.965 (0.01)b
RTD (Nm s−1) 39.35 ± 43.05 (n = 14)a 49.89 ± 52.63 (n = 14)a 0.138 (0.3)c 50.27 ± 40.39 (n = 15)a 54.97 ± 42.36 (n = 15)a 0.065 (0.5)c 0.645 (0.08)b
a a c a a b
AT (Nm) 13.10 ± 13.80 (n = 14) 13.05 ± 14.08 (n = 14) 0.981 (0.0) 11.27 ± 10.56 (n = 15) 12.62 ± 10.83 (n = 15) 0.071 (0.33) 0.677 (0.08)b
MAS hip mean 1.64 ± 0.42 (n = 15) 1.14 ± 0.60 (n = 15) 0.009 (0.48)b 1.50 ± 0.40 (n = 16) 1.13 ± 0.57 (n = 16) 0.038 (0.38)b 0.312 (0.18)b
MAS knee mean 1.63 ± 0.51 (n = 15) 0.90 ± 0.57 (n = 15) 0.001 (0.58)b 1.50 ± 0.34 (n = 16) 0.73 ± 0.57 (n = 16) 0.001 (0.61)b 0.763 (0.05)b
b b
MAS ankle mean 1.83 ± 0.65 (n = 15) 1.07 ± 0.64 (n = 15) 0.001 (0.59) 1.80 ± 0.80 (n = 16) 1.22 ± 0.87 (n = 16) 0.002 (0.54) 0.539 (0.11)b
10MWT (m s−1) 0.78 ± 0.45 (n = 14)a 0.81 ± 0.45 (n = 14)a 0.122 (0.4)c 0.82 ± 0.50 (n = 16) 0.91 ± 0.52 (n = 16) 0.000 (1.1)c 0.055 (0.7)c
TUG (s) 18.68 ± 15.02 (n = 14)a 17.95 ± 14.09 (n = 14)a 0.124 (0.29)b 28.05 ± 43.95 (n = 16) 27.22 ± 46.63 (n = 16) 0.079 (0.31)b 0.678 (0.08)b
c c
LHS 0.52 ± 0.23 (n = 15) 0.53 ± 0.27 (n = 15) 0.793 (0.1) 0.44 ± 0.23 (n = 16) 0.53 ± 0.19 (n = 16) 0.030 (0.6) 0.260 (0.4)c

Abbreviations: ST, strength training only; LA, less‐affected limb; MA, more‐affected limb; MST, mirror and strength training; MVC, maximal voluntary contraction; PT, peak torque; RTD, rate of torque develop-
ment; AT, average torque; MAS, Modified Ashworth Scale; TUG, timed up and go; LHS, London Handicap Scale; ES, effect size; n, number of subjects; 10MWT, 10‐m walk test; T1, preintervention baseline
assessment; T2, postintervention assessment.
a
One subject unable to be assessed. bEffect size expressed as r. cEffect size expressed as Cohen's d.
5 of 9
6 of 9
samples t test or Wilcoxon signed‐rank test. A p < 0.05 was con-
sidered statistically significant and effect size was expressed as
Cohen's d or r.
3 | RESULTS
Of the 35 participants randomized, four dropped out during the inter-
vention (Figure 1). Thirty‐one participants (61.7 ± 13.3) completed the
intervention and were included for data analysis. There were no
demographic differences between the ST and MST groups (Table 1).
Preassessment and postassessment results are presented in Table 2.
3.1 | Feasibility outcomes
Thirty‐five subjects were eligible to participate in the study from a
total of 36 referred (97% success rate). The one screen failure was
due to bilateral neuropathy present. Twenty‐two participants
attended 100% of training sessions, eight attended 92% (11 sessions),
and one attended 83% (10 sessions). Non‐attendance was due to ill
health unrelated to the intervention, or unavailable due to personal
reasons. Four participants (two ST and two MST) dropped out during
the intervention, equating to an 11% attrition rate. Thirty‐one partici-
pants attended all assessment sessions. One participant was unwilling
to attempt the strength assessment, the 10MWT, and the TUG. One
additional participant was not assessed for untrained limb strength
due to positioning discomfort. All training and assessment sessions
were completed without any adverse events.
3.2 | Effectiveness outcomes
There were no between group differences for trained or untrained
limb strength for all strength parameters (PT, RTD, and AT). The
MST group showed approaching significant increases in RTD with
medium effect and AT with medium effect in the untrained (more
affected) limb. There were no between group differences in spasticity
reduction but significant reductions in spasticity in the untrained
(more affected) limb for the hip, knee, and ankle in both the ST and
MST groups. There was a significant increase in walking velocity
(0.09 m s−1) for the MST group with large effect with a medium effect
between group difference in favour of the MST group. The MST group
showed a significant improvement in LHS scores (8%) with medium
effect but no between group difference was observed.
4 | DISCUSSION
To our knowledge, this is the first pilot study investigating the feasibil-
ity and potential efficacy of cross‐education combined with mirror
therapy compared with cross‐education only in a stroke population.
In this study, mirror visual feedback of the isometrically training limb
did not augment cross‐education of strength. Small sample sizes often
prevent the detection of statistically significant results, causing Type II
errors (Faber & Fonseca, 2014). Even though the between group
SIMPSON ET AL.
difference for change in walking velocity in favour of the MST group
did not quite reach statistical significance (p = 0.055, d = 0.7), in a pilot
study with a small sample size this may be considered an interesting
finding worthy of further investigation. Although not a high standard
of evidence (Stolberg, Norman, & Trop, 2004), the within group results
of this novel study add valuable information to existing knowledge
regarding cross‐education and mirror therapy in reducing spasticity
and increasing walking velocity. This study also provides critical in-
formation regarding assessing patient eligibility, compliance and
retention, effectiveness measures, and treatment reliability prior to a
definitive randomized controlled trial (RCT).
This pilot study demonstrated a high level of feasibility. A 97% eli-
gibility rate indicates that it is feasible to obtain subjects conforming
to the inclusion/exclusion criteria. The minimum compliance rate was
83% (one participant), with the majority of participants achieving
92–100% treatment compliance. The attrition rate of 11% in this
study is considered acceptable for a high quality RCT (de Morton,
2009). Of all participants, 94% completed all preassessment and
postassessment, indicating good feasibility of effectiveness measures.
There were no adverse events during training or assessment; there-
fore, treatment was considered safe and reliable.
We observed no between group differences for strength gains in
all strength parameters (PT, RTD, and AT); however, only the MST
group showed approaching significant increases with medium effect
in RTD (9.3%) and AT (12%) for the untrained (more affected) limb.
Though cross‐education combined with mirror therapy was not previ-
ously investigated poststroke, Zult et al. (2016) report untrained upper
limb PT gains in both an experimental; unilateral dynamic mirror and
strength training (61%) and control; and unilateral dynamic strength
training only (34%) in healthy subjects training for 3 weeks (15 ses-
sions). The isometric training mode may have contributed to the
substantially lower mirror aided strength improvement in our study.
Isometric strengthening, a safe and easily implemented training mode
(Harbo, Brincks, & Andersen, 2012), has previously shown to elicit
significant strength transfer to the untrained limb following stroke
(Dragert & Zehr, 2013). However, Manca, Cabboi, et al. (2017) report
that dynamic training modes achieve a greater mean strength transfer
(16%) than isometric training (8%). Similarly, Zult et al. (2016) conclude
that a dynamic form of training that allows for a moving mirror image
is more effective in generating a training effect.
This study showed no strength improvement for either the trained
or untrained limb for the ST group, whereas Dragert and Zehr (2013)
reported a 34% and 31% PT strength increase in the trained and
untrained limbs of stroke patients following 6 weeks (18 sessions) uni-
lateral isometric dorsiflexion strengthening. Manca, Cabboi, et al.
(2017) recently demonstrated a linear response to unilateral strength
training, requiring 6 weeks to achieve any significant cross‐education
strength gains in the untrained dorsiflexors in neurologically impaired
(multiple sclerosis) subjects. Our low magnitude of strength transfer
for both groups may in part be due to intervention duration and
training dose (4 weeks, 12 sessions), which compared with previous
cross‐education studies (typically 4–6 weeks, >16 sessions; Dragert
& Zehr, 2013; Manca, Cabboi, et al., 2017; Manca, Pisanu, Ortu, &
SIMPSON ET AL.
Deriu, 2015) is a short intervention with minimal treatment dose.
Furthermore, participant's effort in our study was not objectively mea-
sured during training. It is possible that participants did not consis-
tently reach a high enough intensity (80–100% MVC) needed to
elicit significant PT strength transfer to the untrained limb (Fimland
et al., 2009). However, with favourable RTD and AT, results for the
MST group and considering improvements in RTD or AT may repre-
sent the most important adaptations occurring from training (Aagaard,
Simonsen, Andersen, Magnusson, & Dyhre‐Poulsen, 2002; Wimpenny,
2016) in clinical populations, rather than PT alone, mirror aided con-
tractions warrant further investigation. With no MCID for dorsiflexion
strength, our results cannot be compared with a value representing a
clinically meaningful change.
A surprising finding in our study was the absence of any significant
change in trained limb strength for both the ST and MST groups. This
is a particularly unexpected result considering that the training proto-
col was based on strength specific guidelines (ACSM, 2009). The high
variability in the day to day physical ability of stroke patients may have
impacted training and testing performance (Eng, Kim, & Macintyre,
2002). Future studies should apply a higher treatment dose of mea-
sured dynamic MVC's (80–100%) that ensure mirror visual feedback
of a moving limb.
We observed a significant reduction in spasticity for all lower limb
joints for both the ST and MST groups. However, spasticity reduction
for either group did not reach the MCID of 1‐point decrease in MAS
score (Shaw et al., 2010). Dragert and Zehr (2013) reported no change
in MAS scores, but a significant improvement was perhaps unlikely as
five out of 19 subjects presented with no spasticity whatsoever at
baseline and a further seven with an MAS score of 1 (representing
minimal detection of muscle tone). A beneficial effect of cross‐
education on spasticity/muscle tone might be expected. The under-
lying cause of spasticity is the disrupted balance between excitatory
and inhibitory signals resulting in hyperexcitability of the stretch reflex
(Trompetto et al., 2014). Previous cross‐education studies have
reported a reduction in contralateral H‐reflex excitability during unilat-
eral training (Carson et al., 2004; Hortobagyi, Taylor, Petersen, Russell,
& Gandevia, 2003). Although the most widely reported assessment of
spasticity in the stroke population (Thibaut et al., 2013), the MAS
assesses the resistance to passive movement, not just stretch‐reflex
hyperexcitibility. Spastic cocontraction is another component of
muscle overactivity influencing MAS scores (Vattanasilp, Ada, &
Crosbie, 2000), brought on by abnormal reflex activity, causing simul-
taneous activity in the agonist and antagonist muscles (Gracies, 2005;
Trompetto et al., 2014). The reduced MAS scores for both groups in
this study may be attributed to an improved motor output or firing
pattern to the untrained limb, too minimal to be detected in strength
measures, potentially reducing spastic cocontraction. Perhaps, a fur-
ther reduction in spasticity, facilitated by a higher treatment dose,
would result in more detectable strength improvements.
The associated improvement in walking velocity (10MWT
>0.06 m s−1; Perera et al., 2006) in favour of the MST group warrants
further investigation. In this study, the results of the 10MWT represent
the most statistically significant change over time for the MST group
7 of 9
and the greatest between group difference for any clinical effective-
ness measure. The effect size of the between group difference
(d = 0.7) can be used to determine sample size requirements for a fully
powered trial (significance level 0.05 and power 0.8), indicating 68 par-
ticipants (34 in each arm; AI Therapy Statistics, 2017; Cohen, 1992).
Only the MST group self‐reported a significant improvement (9%
with medium effect) in their self‐perceived impact of stroke (LHS).
This is the first lower limb study to report the beneficial influence of
unilateral strengthening with mirror therapy on self‐perceived partici-
pation poststroke.
4.1 | Limitations
The small sample size may have influenced the ability to detect signif-
icant changes between pre/postassessments and between group
differences; nonetheless, effect sizes for appropriate clinical measures
have been established, which can be used when calculating sample
sizes for future trials. This study did not measure training intensity
or control for placebo effects. Therefore, a fully powered RCT is
needed to substantiate the effectiveness of these therapies.
4.2 | Implications for physiotherapy practice
Cross‐education combined with mirror therapy may improve function
poststroke, with particular interest in patients who are unable to
engage the more‐affected side during physical therapy. In this pilot
study, the combination treatment did not improve strength, spasticity,
or self‐perceived participation to a greater extent than unilateral
strength training alone. However, the combination treatment was asso-
ciated with functional improvements and warrants further investigation
focussing on dynamic strength training modes. Findings also indicate
that the combination treatment has potential to reduce spasticity
poststroke and is easy to implement, without adverse effects. Further
investigations in the form of RCT's are required to test for effectiveness
and the dose–response relationship of cross‐education and mirror
therapy in stroke populations. Similarly, further studies are needed to
measure physiological mechanisms in response to training.
ACKNOWLEDGEMENTS
This study was supported by the Institutes of Technology Ireland
Postgraduate Research Scholarship (D. Simpson and P. Broderick), IT
Sligo Capacity Building Fund (D. Simpson and P. Broderick), IT Sligo
President's Bursary Fund, and Irish Research Council Postgraduate
Scholarship (M. Ehrensberger; Grant GOIPG/2016/1662).
ORCID
Daniel Simpson https://orcid.org/0000-0003-3043-999X
RE FE RE NC ES
Aagaard, P., Simonsen, E. B., Andersen, J. L., Magnusson, P., & Dyhre‐
Poulsen, P. (2002). Increased rate of force development and neural
drive of human skeletal muscle following resistance training. Journal
8 of 9
of Applied Physiology (Bethesda, MD: 1985), 93(4), 1318–1326. https://
doi.org/10.1152/japplphysiol.00283.2002
ACSM (2009). American College of Sports Medicine position stand. Pro-
gression models in resistance training for healthy adults. Medicine and
Science in Sports and Exercise, 41(3), 687–708. https://doi.org/10.
1249/MSS.0b013e3181915670
AI Therapy Statistics (2017). Sample size calculator. Retrieved from https://
www.ai‐therapy.com/psychology‐statistics/sample‐size‐calculator
Barzi, Y., & Zehr, E. (2008). Rhythmic arm cycling suppresses hyperactive
soleus H‐reflex amplitude after stroke. Clinical Neurophysiology,
119(6), 1443–1452. https://doi.org/10.1016/j.clinph.2008.02.016
Benjamin, E. J., Blaha, M. J., Chiuve, S. E., Cushman, M., Das, S. R., Deo, R., …
Jiménez, M. C. (2017). Heart disease and stroke statistics‐2017 update:
A report from the American Heart Association. Circulation, 135(10),
e146–e603. https://doi.org/10.1161/cir.0000000000000485
Biodex Medical Systems Inc. (2006). Biodex system 3 pro application/
operationmanual. Retrieved from http://www.biodex.com/sites/
default/files/835000man_06159.pdf
Bird, S. P., Tarpenning, K. M., & Marino, F. E. (2005). Designing resistance
training programmes to enhance muscular fitness: A review of the acute
programme variables. Sports Medicine, 35(10), 841–851. https://doi.
org/10.2165/00007256‐200535100‐00002
Broderick, P., Horgan, F., Blake, C., Ehrensberger, M., Simpson, D., & Mona-
ghan, K. (2018). Mirror therapy for improving lower limb motor function
and mobility after stroke: A systematic review and meta‐analysis. Gait &
Posture, 63, 208–220. https://doi.org/10.1016/j.gaitpost.2018.05.017
Carroll, L. M., Volpe, D., Morris, M. E., Saunders, J., & Clifford, A. M. (2017).
Aquatic exercise therapy for people with Parkinson disease: A random-
ized controlled trial. Archives of Physical Medicine and Rehabilitation,
98(4), 631–638. https://doi.org/10.1016/j.apmr.2016.12.006
Carson, R., Riek, S., Mackey, D., Meichenbaum, D., Willms, K., Forner, M., &
Byblow, W. (2004). Excitability changes in human forearm corticospinal
projections and spinal reflex pathways during rhythmic voluntary
movement of the opposite limb. The Journal of Physiology, 560(Pt 3,
929–940. https://doi.org/10.1113/jphysiol.2004.069088
Carvalho, D., Teixeira, S., Lucas, M., Yuan, T. F., Chaves, F., Peressutti, C., &
Arias‐Carrion, O. (2013). The mirror neuron system in post‐stroke reha-
bilitation. International Archives of Medicine, 6(1), 41. https://doi.org/
10.1186/1755‐7682‐6‐41
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159.
https://doi.org/10.1037/0033‐2909.112.1.155
Collen, F. M., Wade, D. T., & Bradshaw, C. M. (1990). Mobility after
stroke: Reliability of measures of impairment and disability. Interna-
tional Disability Studies, 12(1), 6–9. https://doi.org/10.3109/0379
0799009166594
de Morton, N. A. (2009). The PEDro scale is a valid measure of the meth-
odological quality of clinical trials: A demographic study. The Australian
Journal of Physiotherapy, 55(2), 129–133. https://doi.org/10.1016/
S0004‐9514(09)70043‐1
Deconinck, F. J., Smorenburg, A. R., Benham, A., Ledebt, A., Feltham, M. G.,
& Savelsbergh, G. J. (2015). Reflections on mirror therapy: A systematic
review of the effect of mirror visual feedback on the brain.
Neurorehabilitation and Neural Repair, 29(4), 349–361. https://doi.org/
10.1177/1545968314546134
Dragert, K., & Zehr, E. P. (2013). High‐intensity unilateral dorsiflexor resis-
tance training results in bilateral neuromuscular plasticity after stroke.
Experimental Brain Research, 225(1), 93–104. https://doi.org/10.1007/
s00221‐012‐3351‐x
Ehrensberger, M., Simpson, D., Broderick, P., & Monaghan, K. (2016). Cross‐
education of strength has a positive impact on post‐stroke rehabilitation:
SIMPSON ET AL.
A systematic literature review. Topics in Stroke Rehabilitation, 23(2),
126–135. https://doi.org/10.1080/10749357.2015.1112062
Eng, J. J., Kim, C. M., & Macintyre, D. L. (2002). Reliability of lower extrem-
ity strength measures in persons with chronic stroke. Archives of
Physical Medicine and Rehabilitation, 83(3), 322–328. https://doi.org/
10.1053/apmr.2002.29622
Faber, J., & Fonseca, L. M. (2014). How sample size influences research
outcomes. Dental Press Journal of Orthodontics, 19(4), 27–29. https://
doi.org/10.1590/2176‐9451.19.4.027‐029.ebo
Faria, C. D., Teixeira‐Salmela, L. F., Neto, M. G., & Rodrigues‐de‐Paula, F.
(2012). Performance‐based tests in subjects with stroke: Outcome
scores, reliability and measurement errors. Clinical Rehabilitation,
26(5), 460–469. https://doi.org/10.1177/0269215511423849
Farthing, J. P. (2009). Cross‐education of strength depends on limb
dominance: Implications for theory and application. Exercise and
Sport Sciences Reviews, 37(4), 179–187. https://doi.org/10.1097/
JES.0b013e3181b7e882
Fimland, M. S., Helgerud, J., Solstad, G. M., Iversen, V. M., Leivseth, G., &
Hoff, J. (2009). Neural adaptations underlying cross‐education after
unilateral strength training. European Journal of Applied Physiology,
107(6), 723–730. https://doi.org/10.1007/s00421‐009‐1190‐7
Flansbjer, U. B., Holmback, A. M., Downham, D., Patten, C., & Lexell, J.
(2005). Reliability of gait performance tests in men and women with
hemiparesis after stroke. Journal of Rehabilitation Medicine, 37(2),
75–82. https://doi.org/10.1080/16501970410017215
Gracies, J. M. (2005). Pathophysiology of spastic paresis. II: Emergence of
muscle overactivity. Muscle & Nerve, 31(5), 552–571. https://doi.org/
10.1002/mus.20285
Harbo, T., Brincks, J., & Andersen, H. (2012). Maximal isokinetic and isomet-
ric muscle strength of major muscle groups related to age, body mass,
height, and sex in 178 healthy subjects. European Journal of Applied Phys-
iology, 112(1), 267–275. https://doi.org/10.1007/s00421‐011‐1975‐3
Hendy, A. M., & Lamon, S. (2017). The cross‐education phenomenon: Brain
and beyond. Frontiers in Physiology, 8, 297. https://doi.org/10.3389/
fphys.2017.00297
Holmback, A. M., Porter, M. M., Downham, D., & Lexell, J. (1999). Reliabil-
ity of isokinetic ankle dorsiflexor strength measurements in healthy
young men and women. Scandinavian Journal of Rehabilitation Medicine,
31(4), 229–239.
Hortobagyi, T. (2005). Cross education and the human central nervous sys-
tem. IEEE Engineering in Medicine and Biology Magazine, 24(1), 22–28.
https://doi.org/10.1109/MEMB.2005.1384096
Hortobagyi, T., Taylor, J. L., Petersen, N. T., Russell, G., & Gandevia, S. C.
(2003). Changes in segmental and motor cortical output with contralat-
eral muscle contractions and altered sensory inputs in humans. Journal
of Neurophysiology, 90(4), 2451–2459. https://doi.org/10.1152/jn.
01001.2002
Howatson, G., Zult, T., Farthing, J. P., Zijdewind, I., & Hortobagyi, T. (2013).
Mirror training to augment cross‐education during resistance training:
A hypothesis. Frontiers in Human Neuroscience, 7, 396. https://doi.
org/10.3389/fnhum.2013.00396
Lee, M., & Carroll, T. J. (2007). Cross education: Possible mechanisms for the
contralateral effects of unilateral resistance training. Sports Medicine,
37(1), 1–14. https://doi.org/10.2165/00007256‐200737010‐00001
Magnus, C. R., Arnold, C. M., Johnston, G., Dal‐Bello Haas, V., Basran, J.,
Krentz, J. R., & Farthing, J. P. (2013). Cross‐education for improving
strength and mobility after distal radius fractures: A randomized con-
trolled trial. Archives of Physical Medicine and Rehabilitation, 94(7),
1247–1255. https://doi.org/10.1016/j.apmr.2013.03.005
SIMPSON ET AL.
Manca, A., Cabboi, M., Dragone, D., Ginatempo, F., Ortu, E., De Natale, E., …
Deriu, F. (2017). Resistance training for muscle weakness in multiple
sclerosis: Direct versus contralateral approach in individuals with ankle
dorsiflexors' disparity in strength. Archives of Physical Medicine and
Rehabilitation, 98(7), 1348–1356. https://doi.org/10.1016/j.apmr.
2017.02.019
Manca, A., Dragone, D., Dvir, Z., & Deriu, F. (2017). Cross‐education of
muscular strength following unilateral resistance training: A meta‐
analysis. European Journal of Applied Physiology, 117(11), 2335–2354.
https://doi.org/10.1007/s00421‐017‐3720‐z
Manca, A., Pisanu, F., Ortu, E., & Deriu, F. (2015). Isokinetic cross‐training
effect in foot drop following common peroneal nerve injury. Isokinetics
and Exercise Science, 23(1), 17–20. https://doi.org/10.3233/IES‐140559
McElwaine, P., McCormack, J., & Harbison, J. (2015). National Stroke Audit
2015. Retrieved from http://www.irishheart.ie/media/pub/
strokestudy2015/ihfhse_national_stroke_audit__mcelwaine.pdf
Michielsen, M. E., Selles, R. W., van der Geest, J. N., Eckhardt, M., Yavuzer,
G., Stam, H. J., … Bussmann, J. B. (2011). Motor recovery and cortical
reorganization after mirror therapy in chronic stroke patients: A phase
II randomized controlled trial. Neurorehabilitation and Neural Repair,
25(3), 223–233. https://doi.org/10.1177/1545968310385127
Park, E., & Choi, Y. (2014). Rasch analysis of the London Handicap Scale in
stroke patients: A cross‐sectional study. Journal of Neuroengineering and
Rehabilitation, 11, 114. https://doi.org/10.1186/1743‐0003‐11‐114
Patten, C., Lexell, J., & Brown, H. E. (2004). Weakness and strength training
in persons with poststroke hemiplegia: Rationale, method, and efficacy.
Journal of Rehabilitation Research and Development, 41(3a), 293–312.
https://doi.org/10.1682/JRRD.2004.03.0293
Pekna, M., Pekny, M., & Nilsson, M. (2012). Modulation of neural plasticity
as a basis for stroke rehabilitation. Stroke, 43(10), 2819–2828. https://
doi.org/10.1161/strokeaha.112.654228
Perera, S., Mody, S. H., Woodman, R. C., & Studenski, S. A. (2006). Mean-
ingful change and responsiveness in common physical performance
measures in older adults. Journal of the American Geriatrics Society,
54(5), 743–749. https://doi.org/10.1111/j.1532‐5415.2006.00701
Rossiter, H. E., Borrelli, M. R., Borchert, R. J., Bradbury, D., & Ward, N. S.
(2015). Cortical mechanisms of mirror therapy after stroke.
Neurorehabilitation and Neural Repair, 29(5), 444–452. https://doi.org/
10.1177/1545968314554622
Shaw, L., Rodgers, H., Price, C., van Wijck, F., Shackley, P., Steen, N., &
Graham, L. (2010). BoTULS: A multicentre randomised controlled trial
to evaluate the clinical effectiveness and cost‐effectiveness of treating
upper limb spasticity due to stroke with botulinum toxin type A.
Health Technology Assessment, 14(26), 1–113. https://doi.org/10.
3310/hta14260
Stolberg, H. O., Norman, G., & Trop, I. (2004). Randomized controlled trials.
AJR. American Journal of Roentgenology, 183(6), 1539–1544. https://
doi.org/10.2214/ajr.183.6.01831539
Thibaut, A., Chatelle, C., Ziegler, E., Bruno, M. A., Laureys, S., & Gosseries, O.
(2013). Spasticity after stroke: Physiology, assessment and treatment.
Brain Injury, 27(10), 1093–1105. https://doi.org/10.3109/02699052.
2013.804202
Thieme, H., Morkisch, N., Mehrholz, J., Pohl, M., Behrens, J., Borgetto, B., &
Dohle, C. (2018). Mirror therapy for improving motor function after
stroke. Cochrane Database of Systematic Reviews, (7), Cd008449.
https://doi.org/10.1002/14651858.CD008449.pub3
9 of 9
Touzalin‐Chretien, P., Ehrler, S., & Dufour, A. (2010). Dominance of vision
over proprioception on motor programming: Evidence from ERP. Cere-
bral Cortex, 20(8), 2007–2016. https://doi.org/10.1093/cercor/bhp271
Trompetto, C., Marinelli, L., Mori, L., Pelosin, E., Currà, A., Molfetta, L., &
Abbruzzese, G. (2014). Pathophysiology of spasticity: Implications for
neurorehabilitation. BioMed Research International, 2014, 1–8. https://
doi.org/10.1155/2014/354906
Urban, P. P., Wolf, T., Uebele, M., Marx, J. J., Vogt, T., Stoeter, P., … Wissel, J.
(2010). Occurence and clinical predictors of spasticity after ischemic
stroke. Stroke, 41(9), 2016–2020. https://doi.org/10.1161/strokeaha.
110.581991
van Wijck, F. M., Pandyan, A. D., Johnson, G. R., & Barnes, M. P. (2001).
Assessing motor deficits in neurological rehabilitation: Patterns of
instrument usage. Neurorehabilitation and Neural Repair, 15(1), 23–30.
https://doi.org/10.1177/154596830101500104
Vattanasilp, W., Ada, L., & Crosbie, J. (2000). Contribution of thixotropy,
spasticity, and contracture to ankle stiffness after stroke. Journal of
Neurology, Neurosurgery, and Psychiatry, 69(1), 34–39. https://doi.org/
10.1136/jnnp.69.1.34
World Health Organization (2001). International classification of function-
ing, disability and health. Retrieved from http://unstats.un.org/unsd/
disability/pdfs/ac.81‐b4.pdf2001
Wimpenny, P. (2016). Theory—Interpretation of results. Retrieved from
http://www.isokinetics.net/index.php/2016‐04‐05‐17‐04‐58/inter-
pretation/general‐interpretation
Wissel, J., Schelosky, L. D., Scott, J., Christe, W., Faiss, J. H., & Mueller, J.
(2010). Early development of spasticity following stroke: A prospective,
observational trial. Journal of Neurology, 257(7), 1067–1072. https://
doi.org/10.1007/s00415‐010‐5463‐1
Wolf, S. L., Catlin, P. A., Gage, K., Gurucharri, K., Robertson, R., & Stephen,
K. (1999). Establishing the reliability and validity of measurements of
walking time using the emory functional ambulation profile. Physical
Therapy, 79(12), 1122–1133.
Zipp, G., & Sullivan, J. (2010). Neurology section StrokEDGE taskforce.
Retrieved from http://www.neuropt.org/docs/stroke‐sig/strokeedge_
taskforce_summary_document.pdf
Zult, T., Goodall, S., Thomas, K., Solnik, S., Hortobagyi, T., & Howatson, G.
(2016). Mirror training augments the cross‐education of strength and
affects inhibitory paths. Medicine and Science in Sports and Exercise,
48, 1001–1013. https://doi.org/10.1249/mss.0000000000000871
Zult, T., Howatson, G., Kadar, E. E., Farthing, J. P., & Hortobagyi, T. (2014).
Role of the mirror‐neuron system in cross‐education. Sports Medicine,
44(2), 159–178. https://doi.org/10.1007/s40279‐013‐0105‐2
Whitehead, A. L., Julious, S. A., Cooper, C. L., & Campbell, M. J. (2016).
Estimating the sample size for a pilot randomised trial to minimise the
overall trial sample size for the external pilot and main trial for a contin-
uous outcome variable. Statistical Methods in Medical Research, 25(3),
1057–1073. https://doi.org/10.1177/0962280215588241
How to cite this article: Simpson D, Ehrensberger M, Horgan
F, et al. Unilateral dorsiflexor strengthening with mirror ther-
apy to improve motor function after stroke: A pilot randomized
study. Physiother Res Int. 2019;e1792. https://doi.org/
10.1002/pri.1792