Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282

Available online at

ScienceDirect
www.sciencedirect.com

Review

Effectiveness of static stretching positioning on post-stroke

upper-limb spasticity and mobility: Systematic review
with meta-analysis
Ana Paula Salazar a,b, Camila Pinto a,b, Joao Victor Ruschel Mossi c, Bruno Figueiro c,
Janice Luisa Lukrafka a,c, Aline Souza Pagnussat a,b,c,*
a
Rehabilitation Sciences Graduate Program, Universidade Federal de Ciências da Saúde de Porto Alegre, Porto Alegre, Brazil
b
Movement Analysis and Rehabilitation Laboratory, Universidade Federal de Ciências da Saúde de Porto Alegre, Porto Alegre, Brazil
c
Department of Physiotherapy, Universidade Federal de Ciências da Saúde de Porto Alegre, Porto Alegre, Brazil

A R T I C L E I N F O A B S T R A C T

Article history: Objective: To systematically review the effects of static stretching with positioning orthoses or simple
Received 28 May 2018 positioning combined or not with other therapies on upper-limb spasticity and mobility in adults after
Accepted 29 November 2018 stroke.
Methods: This meta-analysis was conducted according to PRISMA guidelines and registered at
Keywords: PROSPERO. MEDLINE (Pubmed), Embase, Cochrane CENTRAL, Scopus and PEDro databases were
Physical therapy modalities searched from inception to January 2018 for articles. Two independent researchers extracted data,
Physiotherapy
assessed the methodological quality and rated the quality of evidence of studies.
Muscle stretching exercises
Muscle spasticity
Results: Three studies (57 participants) were included in the spasticity meta-analysis and 7
Range of motion (210 participants) in the mobility meta-analysis. Static stretching with positioning orthoses reduced
Rehabilitation wrist-flexor spasticity as compared with no therapy (mean difference [MD] = 1.89, 95% confidence
interval [CI] 2.44 to 1.34; I2 79%, P < 0.001). No data were available concerning the spasticity of other
muscles. Static stretching with simple positioning, combined or not with other therapies, was not better
than conventional physiotherapy in preventing loss of mobility of shoulder external rotation (MD = 3.50,
95% CI 3.45 to 10.45; I2 54.7%, P = 0.32), shoulder flexion (MD = 1.20, 95% CI 8.95 to 6.55; I2 0%,
P = 0.76) or wrist extension (MD = 0.32, 95% CI 6.98 to 5.75; I2 38.5%, P = 0.92). No data were available
concerning the mobility of other joints.
Conclusion: This meta-analysis revealed very low-quality evidence that static stretching with positioning
orthoses reduces wrist flexion spasticity after stroke as compared with no therapy. Furthermore, we found
low-quality evidence that static stretching by simple positioning is not better than conventional
physiotherapy for preventing loss of mobility in the shoulder and wrist. Considering the limited number of
studies devoted to this issue in post-stroke survivors, further randomized clinical trials are still needed.
Clinical Trial Registration: PROSPERO (CRD42017078784).
C 2018 Published by Elsevier Masson SAS.

1. Introduction is defined as a sensorimotor disorder implicated in some level of

involuntary muscle activation [3,4]. Changes in mechanical
Stroke is the third leading cause of disability worldwide muscle-fiber properties such as loss of sarcomeres and enhanced
[1]. Cerebral stroke may result in several motor impairments intrinsic stiffness in muscle fibers may contribute to increasing
including spasticity, weakness (spastic paresis) and contractures, the muscle tone [5]. Spasticity limits muscle lengthening, which
which impose significant challenges for patient care [2]. Spasticity may have 2 consequences:
is one of the consequences of upper motor-neuron syndrome and
 spastic muscles have a tendency to stay in a shortened position
* Corresponding author. Universidade Federal de Ciências da Saúde de Porto
for longer periods;
Alegre, 245, Sarmento Leite Street, 90050-170 Porto Alegre RS, Brazil.  voluntary activities of antagonist muscles are frequently
E-mail address: [email protected] (A.S. Pagnussat). restricted [6].

https://doi.org/10.1016/j.rehab.2018.11.004
1877-0657/ C 2018 Published by Elsevier Masson SAS.
 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282 275

There is an implicit assumption that spasticity results in soft- meta-analysis clinical trials that did not provide information on
tissue changes that lead to contractures, pain, weakness, activity the effects of the intervention, for the experimental or control
limitations and participation restrictions [4]. group. Authors were contacted by e-mail to obtain missing data.
Muscle stretching is one of the physical therapy modalities
most used to manage spasticity and improve the viscoelastic 2.3. Quality assessment
properties of the muscle-tendon units [7,8]. Static stretching is a
widely used type of stretching and may be applied in different Two reviewers (APS and CP) independently rated the quality of
ways, including by physiotherapist’s hands, splints, orthoses, and studies. Disagreements in ratings were discussed until consensus
plaster casts [8–11]. Studies have investigated stretching therapy was achieved or with a third reviewer (JVRM). The methodological
alone [10,12,13] or combined with other therapies [14–16] to treat quality and risk of bias of included studies were assessed by
spasticity [10,11,17] and increase mobility by enhancing the range extracting PEDro scores from the Physiotherapy Evidence Database
of motion (ROM) [18,19] after stroke. Even though widely used, the (www.pedroorg.au). This tool consists of 11 items, and except for
ideal frequency, intensity, velocity, and duration of stretching item one, which pertains to external validity, each item represents
therapy lacks consensus. Moreover, we do not know whether one point in the total score (range = 0–10) [23]. Scores 9 or 10 are
stretching programs alone or combined with other therapies are considered excellent, 6 to 8 good, 4 or 5 fair, and < 4 poor. Scores
effective to reduce spasticity and/or improve joint ROM after 0 to 5 correspond to high risk of bias and 6 to 10 low risk of bias
stroke [8]. [24]. We used the grading of recommendation, assessment,
Hypertonia has both a neural and a biomechanical component development, and evaluation (GRADE) system to assess the quality
that can negatively affect the quality of life of both patients and of the body of evidence [25] as high, moderate, low or very low. The
caregivers [6]. This systematic review aimed to examine the effect system considers susceptibility to bias, directness of evidence,
of static stretching therapy, combined or not with other therapies, heterogeneity or inconsistencies in the results, imprecision, and
on upper-limb spasticity and mobility of adults after stroke. The probability of publication bias.
effect of static stretching with simple positioning and static
stretching with positioning orthoses was investigated. 2.4. Data extraction

Two reviewers (APS and CP) separately and independently

2. Methods evaluated the titles and abstracts of all articles identified via the
search strategy. After deleting duplicates, they screened the full
This systematic review and meta-analysis was performed in text of the remaining studies. The 2 investigators independently
accordance with the Cochrane Collaboration [20] guidelines. It is extracted data on the methodological characteristics of studies,
presented according to the Preferred Reporting Items for interventions, and outcomes by using standardized forms. Inter-
Systematic Review and Meta-Analyses (PRISMA) statement [21] ventions were detailed regarding measurement, intensity, type of
(Appendix A). The protocol was registered at PROSPERO (http:// activity, and frequency. The primary outcome extracted was
www.crd.york.ac.uk/PROSPERO/display_record. spasticity assessed by conventional scales, including the Ashworth,
php?ID=CRD42017078784). modified Ashworth, and Tardieu Scales. The secondary outcome
was upper-limb mobility assessed by passive ROM (PROM) trough
2.1. Search strategy goniometry.

Articles in the following electronic databases were searched 2.5. Statistical analysis
from inception to April 2017: MEDLINE via PubMed, EMBASE,
Cochrane Central Register of Controlled Trials (CENTRAL), Scopus, For quantitative synthesis, pooled-effect estimates were
and Physiotherapy Evidence Database (PEDro). The search terms obtained by comparing the change from baseline to the end of
included ‘‘stroke’’ or ‘‘brain ischemia’’ and ‘‘muscle stretch the study in each group. Regarding continuous outcomes, if the
exercise,’’ and a string of words previously proposed, which unit of measurement was consistent across studies, the results
yielded a high sensitivity in the search for randomized controlled are presented as the weighted mean difference (MD) with 95%
trials [22], was included for the PubMed search (Appendix B). The confidence intervals (95% CIs). If the unit of measurement was
search was limited to papers written in English, Spanish and inconsistent, the results are expressed as the standardized mean
Portuguese and the terms were adjusted to fit the requirements of difference (SMD) with 95% CIs. Calculations involved using the
each electronic database. Words relating to the outcomes of random effects method.
interest were not included so as to enhance the sensitivity of the The statistical heterogeneity of the treatment effects among
search. We also searched the reference lists of the included studies studies was assessed by the Cochran’s Q and inconsistency I2 tests.
to identify other relevant studies. There were no restrictions on These 2 tests consider values > 25% and 50% as indicating mode-
stroke type (ischemic or hemorrhagic) or time since stroke (acute, rate and high heterogeneity, respectively. The effect size was
subacute or chronic). calculated by the SMD for clinical post-treatment scores in
experimental and control trials [26]. P  0.05 was considered
2.2. Selection criteria statistically significant. All analyses involved using R v1.0.153
(metafor package v2.0-0) [27].
The inclusion criteria was randomized controlled trials focused
on determining the effect of static stretching combined or not
with other therapies for upper-limb spasticity or mobility in 3. Results
adults. Studies that included participants with acute/subacute
(< 6 months) or chronic (> 6 months) stroke were selected. The initial electronic search identified 306 studies. After
Comparator groups had to receive no intervention, other therapies screening titles and abstracts, 20 articles were considered poten-
or conventional physiotherapy. Conventional physiotherapy tially relevant (Fig. 1). Eleven studies met the eligibility criteria, but
was defined as usual care without specific static stretching. only 10 met the criteria for the meta-analyses: 3 [10,11,13] reported
The exclusion criterion was no comparator. We excluded from the data on wrist spasticity and 7 [9,12,15,18,19,28,29] reported data on
[(Fig._1)TD$IG]
276 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282

Fig. 1. The flow of studies included in the review.

upper-limb mobility after stroke. Among the 7 studies, 6 assessed Three studies examined wrist and hand static stretching
shoulder external rotation PROM [9,12,15,18,19,28], 3 shoulders with positioning orthoses to treat spasticity [10,11,13]. These
flexion PROM [9,12,15] and 3 wrists extension PROM studies delivered the therapy for 6 [10,11] or 7 [13] days a
[15,19,29]. One study was not included in the meta-analysis of week. The total time of stretching ranged from 20 [13] to 42 min
spasticity because the protocol of the intervention was quite [10] per day. Intervention protocols lasted from 3 [11] to 4 weeks
different (12 hr of daily stretching for 4 weeks) [30]. [10,13].
Seven studies used static stretching with simple positioning
3.1. Participants [9,12,15,18,19,28,29]. These studies combined static stretching
with conventional physiotherapy. This type of stretching was
The mean age of participants ranged from 36 to 75.4 years delivered 5 [9,12,15,28,29] or 7 [18] days a week; one study did not
(Table 1). All studies included in the wrist spasticity meta-analysis report this information [19]. The daily time of stretching ranged
reported data for participants in the chronic phase after from 30 [29] to 120 min [15,19]. Intervention protocols lasted from
stroke [10,11,13]. However, studies assessing upper-limb 4 [9,29] to 8 [15,19] weeks; one study did not provide this
mobility included participants with acute/subacute stroke information [18].
[9,12,15,18,19,28,29]. Three studies included both ischemic and
hemorrhagic stroke patients [10,18,29] and 8 did not report the 3.3. Quantitative data synthesis
type of stroke [9,11–13,15,19,28,30].
3.3.1. Spasticity
3.2. Interventions (Table 1) Three studies [10,11,13] (n = 57 participants) assessed the
effect of static stretching with positioning orthoses on wrist
Regarding the stretching protocols, meta-analyses were divided spasticity and were included in the meta-analysis. Static stretching
in 2 intervention types: static stretching with positioning orthoses alone was effective to reduce wrist flexion spasticity as compared
and static stretching with simple positioning. Details on inter- with no therapy (MD = 1.89, 95% CI 2.44 to 1.34; I2 79%,
ventions are described below. P < 0.001) (Fig. 2).
Table 1
Summary of included studies.

Study Design Participants Protocol Outcome measures

Frequency and duration for the different groups Characteristics

Ada et al., 2005 [9] RCT n = 36 IG IG Passive range of

Age (yr) 60 min  5 days/wk  4 wk total Static stretch positioning + shoulder exercises and motion
IG: 70  7 CG upper-limb care -Goniometer
CG: 64  9 10 min  5 days/wk  4 wk CG Timing: 0, 4 wk
Sex = 13 M, 18 F Conventional physiotherapy with shoulder exercises
Type of stroke = n/s without specific static stretching
Severity: not reported Muscles stretched:
Time since stroke (month) = < 1 Internal rotators and extensors of shoulder
(acute/subacute)
Follow-up: not reported

A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282
de Jong et al., 2006 [12] RCT n = 19 IG IG Spasticity
Age (yr) 60 min (30 min twice a day)  5 days/wk  5 wk total Static stretch positioning + conventional physiotherapy Ashworth Scale
36–63 (min–max) CG CG Passive range of
Sex = 9 M, 8 F Not reported Conventional physiotherapy without specific static motion
Type of stroke = n/s stretching Hidrogoniometer
Severity: severe Muscles stretched: Timing: 0, 5, 10 wk
Time since stroke Adductors, extensors and internal rotators of shoulder;
(month) = >3 < 6 (subacute) elbow flexors and forearm pronators
Follow-up: 5 weeks
de Jong et al., 2013 [15] RCT n = 46 IG IG Spasticity
Age (yr) 120 min (45 min twice a day)  5 days/wk  8 wk total Arm static stretch positioning + NMES + conventional Tardieu Scale
IG: 56.6  14.2 CG physiotherapy + multidisciplinary rehabilitation Passive range of
CG: 58.4  9.6 Not reported (rehabilitation nurses, occupational therapists and motion
Sex = 27 M, 19 F speech therapists) Goniometer
Type of stroke = n/s CG Timing: 0, 4, 8, 20 wk
Severity: severe Conventional physiotherapy and multidisciplinary
Time since stroke (month) = < 2 rehabilitation (rehabilitation nurses, occupational
(acute/subacute) therapists and speech therapists)
Follow-up: 12 weeks + sham arm positioning (ie, without stretching) + sham
conventional TENS
Muscles stretched: Internal rotators, extensors and
adductors of shoulder; elbow flexors; forearm
pronators; wrist flexors
Dean et al., 2000 [28] RCT n = 23 IG IG Passive range of
Age (yr) 60 min (20 min in each of 3 positions)  5 days/ Static prolonged stretch positioning motion
IG: 58.1  12.5 wk  6 wk total (3 positions) + conventional physiotherapy and Goniometer
CG: 58.2  10.5 CG multidisciplinary rehabilitation (psychology, Active range of motion
Sex = 16 M, 7 F Not reported occupational therapy and social worker) Goniometer
Type of stroke = n/s CG Timing: 0, 6 wk
Severity: not reported Conventional physiotherapy without specific static
Time since stroke (month) = < 3 stretching and multidisciplinary rehabilitation
(acute/subacute) (Psychology, occupational therapy and social worker)
Follow-up: not reported Muscles stretched:
Adductors and internal rotators of shoulder
Gustafsson and RCT n = 32 IG IG Passive range of
McKenna, 2006 [18] Age (yr) 40 min (20 min twice a day) of stretching  7 days/wk Shoulder static stretch positioning + conventional daily motion
IG: 65.9  15.6 (not reported weeks in total) upper-limb (affected) physiotherapy Goniometer
GC: 67.1  13.9 CG CG Timing: 0 (admission),
Sex = n/s 30 min  7 days/week (not reported weeks in total) Conventional daily upper-limb (affected) physiotherapy 1 (discharge)
Type of stroke = I/H without specific static stretching
Severity: not reported Muscles stretched:
Time since stroke (month) = < 1 Internal rotators and adductors of shoulder; elbow
(acute/subacute) flexors
Follow-up: not reported

277
 278
Table 1 (Continued )

Study Design Participants Protocol Outcome measures

Frequency and duration for the different groups Characteristics

Horsley et al., 2007 [29] RCT n = 40 IG IG Passive range of

Age (yr) 30 min  5 days/wk  4 wk total Static stretch positioning + conventional upper-limb motion
GI: 61  21 CG physiotherapy Goniometer
GC: 62  17 5 days/wk  4 wk total CG Timing: 0, 4, 5, 9 wk
Sex = 19 M, 21 F Conventional upper-limb physiotherapy without
Type of stroke = I/H specific static stretching
Severity: not reported Muscles stretched:
Time since stroke (month) = < 3 Wrist and fingers flexors
(acute/subacute)
Follow-up: 5 weeks
Jang et al., 2016 [10] RCT n = 21 IG IG Spasticity

A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282
Age (yr) 42 min (14 min in each 3 positions 3 times a Wrist and hand static stretching device MAS
IG: 48.8  14.8 day)  6 days/wk  4 wk total CG Active range of motion
CG: 49.5  14.2 CG There was no training program Goniometer
Sex = 17 M, 4 F There was no training program Muscles stretched: Timing: 0, 2, 4, 6 wk
Type of stroke = I/H Wrist and fingers flexors
Severity: severe spasticity
Time since stroke (month) = > 6
(chronic)
Follow-up: not reported
Jung et al., 2011 [11] RCT n = 21 IG IG Spasticity
Age (yr) 40 min (20 min twice a day)  6 days/wk  3 wk total Hand static stretching device (Splint) MAS
IG: 45.6  8.1 CG CG Timing: both groups:
CG: 47.5  13.3 There was no training program There was no training program 6 times in 1 week; IG:
Sex = 15 M, 6 F Muscles stretched: 2 times before starting
Type of stroke = n/s Finger flexors the stretching program
Severity: severe weakness and
spasticity
Time since stroke (month) = > 6
(chronic)
Follow-up: 1 week
Kim et al., 2013 [13] RCT n = 15 IG IG Spasticity
Age (yr) 20 min (10 min twice a day)  7 days/week  4 wk total Hand static stretching device MAS
IG: 47.7  8.0 CG CG Timing: 0, 4, 8 wk
CG: 55.1  14.0 There was no training program There was no training program
Sex = 10 M, 5 F Muscles stretched: Fingers flexors
Type of stroke = n/s
Severity: severe weakness
Time since stroke (month) = > 6
(chronic)
Follow-up: not reported
Lannin et al., 2007 [30]a RCT n = 62 G1 (Exp) G1(Exp) Spasticity
Age (yr) 12 hr overnight  4 wk total Wrist static stretching device (Splint) in neutral wrist Tardieu scale
G1: 70.3  12.6 G2 (Exp) position + routine therapy Passive range of
G2: 68.7  12.1 12 hr overnight  4 wk total G2 (Exp) motion
G3: 75.4  11.0 G3 (Con) Wrist static stretching device (splint) in extended wrist Goniometer (assessed
Sex = 30 M, 32 F 4 wk total position plus routine therapy by lateral photographs)
Type of stroke = n/s G3 (Con) Timing: 0, 4, 6 wk
Severity: not reported Conventional physiotherapy without specific static
Time since stroke (month) = < 1 stretching
(acute/subacute) Muscles stretched:
Follow-up: 2 weeks Wrist and fingers flexors
 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282 279

3.3.2. Upper-limb mobility

Timing: 0, 4, 8, 12 wk
Six studies [9,12,15,18,19,28] (n = 170 participants) were
Outcome measures

included in the meta-analysis of mobility of shoulder external

Passive range of
rotation. Static stretching with simple positioning (without any
Goniometer device), combined or not with other therapies, was ineffective to
motion

prevent PROM reduction in shoulder external rotation as

compared with conventional physiotherapy (MD = 3.50, 95% CI
3.45 to 10.45; I2 55%, P = 0.32) (Fig. 3A).

CG or Con: control group; Exp: experimental group; H: hemorrhagic; I: ischemic; IG: intervention group; MAS: Modified Ashworth Scale; n/s: no stated; RCT: randomized controlled trial; wk: week.
Three studies [9,12,15] (n = 90 participants) were included in
Shoulder adductors and internal rotators; elbow flexors;
Static stretch positioning + standard arm care without

the meta-analysis of mobility of shoulder flexion. Static stretching

Standard arm care without specific static stretching

with simple positioning, combined or not with other therapies,

was ineffective to prevent PROM reduction in shoulder flexion as
compared with conventional physiotherapy (MD = 1.20, 95% CI
8.95 to 6.55; I2 0%, P = 0.76) (Fig. 3B).
Finally, 3 studies [15,19,29] (n = 105 participants) were
included in the meta-analysis of mobility of wrist extension.
Static stretching with simple positioning, combined or not with
wrist flexors; finger flexors

other therapies, was ineffective to prevent PROM reduction in

wrist extension as compared with conventional physiotherapy
(MD = 0.32, 95% CI 6.39 to 5.75; I2 38%, P = 0.92) (Fig. 3 C).
Muscles stretched:
Characteristics

3.3.3. Quality of studies

stretching

The median PEDro score of included studies was 6 (range 4–9),

which characterizes studies with low risk of bias (Table 2). All
CG
IG

studies were randomized and had similar groups at baseline. Most

studies reported point estimates and variability (91%), had blinded
assessors (82%) and concealed the allocation list (64%). Five (45.5%)
120 min (30 min in each 2 different positions twice a

studies had adequate follow-up, used intention-to-treat analysis

and reported between-group differences. Only 2 (18.2%) studies
Frequency and duration for the different groups

had blinded participants and no trial had blinded therapists. The

day)  8 wk total (not reported days a week)

quality of evidence was low for spasticity of wrist flexors by the

GRADE system and was low for PROM in shoulder external
rotation, shoulder flexion and wrist extension. The evidence profile
is presented in Appendix C.

4. Discussion

This systematic review with meta-analysis aimed to verify the

Not reported

effect of static stretching therapy combined or not with other

Protocol

therapies on upper-limb spasticity and mobility after stroke. Our

results showed that static stretching with positioning orthoses was
CG
IG

better than no therapy for reducing spasticity of wrist flexors. Also,

static stretching with simple positioning, combined with conven-
tional physiotherapy, had similar effects as conventional physio-
Time since stroke (month) = < 1

therapy on mobility of the shoulder (external rotation and flexion

PROM) and wrist extension. To the best of our knowledge, this is
the first systematic review with meta-analysis including only
Severity: not reported

Follow-up: 4 weeks
Type of stroke = n/s

randomized controlled trials focused on assessing the effect of this

(acute/subacute)

approach on upper-limb spasticity and mobility after stroke.

Sex = 17 M, 8 F
Participants

Both spasticity and reduced mobility are common issues after

CG: 70  10
IG: 66  14
Age (yr)

stroke and result in decreased function due to muscle tightness of

n = 25

the affected extremities [13]. Static stretching is widely used as

treatment after stroke and may be delivered with different
approaches such as static stretching orthoses or simple positioning
Not included in the meta-analysis.
Design

[13]. Our meta-analysis found very low-quality evidence for the

RCT

effectiveness of static stretching with orthoses to reduce spasticity

of wrist flexors as compared with no therapy. This result suggests
that protocols of static stretching using hand devices in neutral or
extended wrist position might be better than no therapy to reduce
Table 1 (Continued )

Turton and Britton,

wrist-flexor spasticity in chronic post-stroke individuals.

All studies included in the spasticity meta-analysis examined
2005 [19]

positioning orthoses. In these studies, the duration and frequency

of stretching therapy were similar [10,11,13]. These studies
Study

delivered 3 [10,13] and 4 [11] weeks of treatment. Orthoses were

a

used 6 [10,11] and 7 [13] days per week for at least 20 min per day
[(Fig._2)TD$IG]
280 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282

[(Fig._3)TD$IG]
Fig. 2. Forest plot for meta-analysis of static stretching through positioning devices versus control (no therapy) on the spasticity of wrist flexors.

Fig. 3. Forest plot for meta-analysis of static stretching through simple positioning versus control (conventional physiotherapy without specific static stretching) on upper-
limb mobility after stroke. Mobility of A) shoulder external rotation, B) shoulder flexion and C) wrist extension.

[13]. One study was excluded from the meta-analysis because it position as compared with a control group. Regarding other
used an overnight hand device for 4 weeks [30], so the duration of muscles, no data were available on the effect of static stretching
stretching was very different from that in other studies. In this with orthoses to reduce spasticity.
excluded study, the authors did not find differences in spasticity or Static stretching is also widely used to increase ROM after
PROM after the use of hand orthoses in neutral or extended stroke [31]. Our meta-analysis showed a low quality of evidence
 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282 281

Table 2
PEDro scores of included studies.

Study Random Concealed Groups Participant Therapist Assessor Adequate Intention- Between-group Point estimate Total score
allocation allocation similar at blinding blinding blinding follow-up to-treat difference and variability (0 to 10)
baseline analysis reported reported

Ada et al., 2005 [9] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 4

DeJong et al., 2006 [12] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 7
De Jong et al., 2013 [15] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 9
Dean et al., 2000 [28] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 6
Gustafsson and [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 6
McKenna, 2006 [18]
Horsley et al., 2007 [29] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 7
Jang et al., 2016 [10] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 4
Jung et al., 2011 [11] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 4
Kim et al., 2013 [13] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 4
Lannin et al., 2007 [30] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 7
Turton and Britton, 2005 [19] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] [TD$INLE] 6

Median (range) 6 (4–9)

that static stretching with simple positioning, combined with focused on the reliability of the Ashworth, modified Ashworth and
conventional physiotherapy, was not better than conventional Tardieu Scales for the adult population [31,34], these scales have
physiotherapy to prevent reduced mobility of shoulder external limitations that may affect their results, including anatomic and
rotation, shoulder flexion or wrist extension. No data were biomechanical characteristics of joints and muscle groups, inter
available on the effect of static stretching with simple positioning rater and intra rater changes as well as the environment and
to prevent the loss of mobility in other upper-limb joints. general condition of the patient [35]. Another important limitation
Studies included in the mobility meta-analysis examined static of this study is the significant heterogeneity presented in the
prolonged positioning combined with conventional physiotherapy spasticity meta-analysis. The high heterogeneity is probably
[9,12,15,18,19,28,29] and adopted different protocols. Stretching related to the small number of participants. Previous studies also
was delivered 5 [9,12,15,28,29] or 7 [18] days a week, and the total found large heterogeneity among studies, including methodology,
time of stretching ranged from 30 [29] to 120 min per day population, intervention, and outcome measures [7,8]. Even if the
[15,19]. Intervention protocols varied from 4 [9,29] to 8 [15,19] included studies presented a low risk of bias, the quality of
weeks. Thus, the duration and frequency of intervention were quite evidence was rated ‘‘very low’’ or ‘‘low’’ according to the GRADE
varied among studies. Considering the heterogeneity of studies system. Hence, we suggest further research assessing the effects of
included in our meta-analysis, it was impossible to reach static stretching protocols to elucidate their effectiveness in both
conclusive results. One study [15] showed positive results of spasticity and mobility outcomes. Finally, future research should
stretching in preventing loss of mobility of shoulder external also assess the relationship between the increase in mobility
rotation. However, according to the authors, these results were not (PROM) and functional recovery after stroke.
clinically relevant [15] (i.e., static stretching through simple
positioning did not provide benefits for participants considering 5. Conclusions
the PROM of shoulder external rotation).
Regarding the maintenance of the stretching effects on This meta-analysis revealed very low-quality evidence that
spasticity, only one study found reduced hand spasticity across static stretching with positioning orthoses is effective to diminish
all times evaluated (i.e., during the stretching protocol [2 evalua- wrist flexor spasticity as compared with no therapy in individuals
tions, 1 per week], immediately after the end of the protocol and in the chronic phase after stroke. Furthermore, we found low-
1 week post-intervention [11]). Considering the maintenance of quality evidence that static stretching with simple positioning,
the stretching effects on mobility, 3 studies compared static combined with conventional physiotherapy, is not better than
stretching positioning plus conventional physiotherapy versus conventional physiotherapy to prevent the loss of upper-limb
conventional physiotherapy: the authors did not find any positive mobility (shoulder external rotation and flexion and wrist flexion)
effect over time [15,19,29]. We did not perform a meta-analysis in individuals in the acute/subacute phase of stroke. Considering
regarding follow-up effects because the time of follow-up was the limited number of studies devoted to this issue in post-stroke
varied greatly among studies [15,19,29]. survivors, further randomized clinical trials with low risk of bias,
The 2 most common types of stretching are dynamic and static adequate sample size and follow-up evaluation are still necessary.
[32]. Dynamic stretching involves moving the body and gradually Future studies addressing the inconclusive elements indicated in
increasing the reach and/or speed of movement via functional this review may help endorse or refute the use of static stretching
ROM. However, static stretching means holding the stretch in for improving upper-limb spasticity and mobility after stroke.
maximum ROM for a certain time [32]. In our review, we included
only static stretching because most studies use this type of
stretching in individuals with stroke. Future studies could compare Funding
these 2 types of stretching to reduce spasticity and prevent PROM
loss in people with stroke. This study was financed in part by the Coordenação de
The studies included in this meta-analysis were similar Aperfeiçoamento de Pessoal de Nı́vel Superior – Brasil (CAPES) –
regarding the severity of spasticity, time since stroke and duration Finance Code 001.
of the stretching treatment. However, our results should be
carefully interpreted because of the small number of participants Disclosure of interest
included. Furthermore, the studies used a method to measure
spasticity that has been questioned [33]. Although several studies The authors declare that they have no competing interest.
282 A.P. Salazar et al. / Annals of Physical and Rehabilitation Medicine 62 (2019) 274–282

Appendix A. Supplementary data [16] Lai JM, Francisco GE, Willis FB. Dynamic splinting after treatment with
botulinum toxin type-A: a randomized controlled pilot study. Adv Ther
2009;26:241–8.
Supplementary data associated with this article can be found, in [17] Baricich A, Carda S, Bertoni M, Maderna L, Cisari C. A single-blinded, random-
the online version, at https://doi.org/10.1016/j.rehab.2018.11.004. ized pilot study of botulinum toxin type A combined with non-pharmacologi-
cal treatment for spastic foot. J Rehabil Med 2008;40:870–2.
[18] Gustafsson L, McKenna K. A programme of static positional stretches does not
reduce hemiplegic shoulder pain or maintain shoulder range of motion–a
References randomized controlled trial. Clin Rehabil 2006;20:277–86.
[19] Turton AJ, Britton E. A pilot randomized controlled trial of a daily muscle
[1] Hankey GJ. The global and regional burden of stroke. Lancet Glob Health stretch regime to prevent contractures in the arm after stroke. Clin Rehabil
2013;1:e239–40. 2005;19:600–12.
[2] Li S. Spasticity, motor recovery, and neural plasticity after stroke. Front Neurol [20] Higgins J, Green S. Cochrane handbook for systematic reviews of interventions,
2017;8:120. 5 ed., Chichester: John Wiley & Sons; 2011.
[3] Li S, Francisco GE. New insights into the pathophysiology of post-stroke [21] Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items
spasticity. Front Hum Neurosci 2015;9:192. for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg
[4] Pandyan AD, Gregoric M, Barnes MP, Wood D, Van Wijck F, Burridge J, et al. 2010;8:336–41.
Spasticity: clinical perceptions, neurological realities and meaningful mea- [22] Robinson KA, Dickersin K. Development of a highly sensitive search strategy
surement. Disabil Rehabil 2005;27:2–6. for the retrieval of reports of controlled trials using PubMed. Int J Epidemiol
[5] Dietz V, Sinkjaer T. Spastic movement disorder: impaired reflex function and 2002;31:150–3.
altered muscle mechanics. Lancet Neurol 2007;6:725–33. [23] Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the
[6] Barnes MP. An overview of the clinical management of spasticity. In: Barnes PEDro scale for rating quality of randomized controlled trials. Phys Ther
MPaJ GR, editor. Upper motor neurone syndrome and spasticity: clinical 2003;83:713–21.
management and neurophysiology. second ed., Cambridge: Cambridge Uni- [24] Olivo SA, Macedo LG, Gadotti IC, Fuentes J, Stanton T, Magee DJ. Scales to assess
versity Press; 2008 [p. 1–8]. the quality of randomized controlled trials: a systematic review. Phys Ther
[7] Harvey LA, Katalinic OM, Herbert RD, Moseley AM, Lannin NA, Schurr K. Stretch 2008;88:156–75.
for the treatment and prevention of contracture: an abridged republication of a [25] Balshem H, Helfand M, Schunemann HJ, Oxman AD, Kunz R, Brozek J, et al.
Cochrane Systematic Review. J Physiother 2017;63:67–75. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol
[8] Bovend’Eerdt TJ, Newman M, Barker K, Dawes H, Minelli C, Wade DT. The 2011;64:401–6.
effects of stretching in spasticity: a systematic review. Arch Phys Med Rehabil [26] Cohen J. Statistical power analysis for the behavioral sciences, 2nd ed.,
2008;89:1395–406. Hillsdale: L. Erlbaum Associates; 1988.
[9] Ada L, Goddard E, McCully J, Stavrinos T, Bampton J. Thirty minutes of [27] Viechtbauer W. Conducting meta-analyses in R with the metafor package. J
positioning reduces the development of shoulder external rotation contrac- Stat Soft 2010;36:1–48.
ture after stroke: a randomized controlled trial. Arch Phys Med Rehabil [28] Dean CM, Mackey FH, Katrak P. Examination of shoulder positioning after
2005;86:230–4. stroke: a randomised controlled pilot trial. Aust J Physiother 2000;46:35–40.
[10] Jang WH, Kwon HC, Yoo KJ, Jang SH. The effect of a wrist-hand stretching [29] Horsley SA, Herbert RD, Ada L. Four weeks of daily stretch has little or no effect
device for spasticity in chronic hemiparetic stroke patients. Eur J Phys Rehabil on wrist contracture after stroke: a randomised controlled trial. Aust J Physio-
Med 2016;52:65–71. ther 2007;53:239–45.
[11] Jung YJ, Hong JH, Kwon HG, Song JC, Kim C, Park S, et al. The effect of a [30] Lannin NA, Cusick A, McCluskey A, Herbert RD. Effects of splinting on wrist
stretching device on hand spasticity in chronic hemiparetic stroke patients. contracture after stroke: a randomized controlled trial. Stroke 2007;38:111–6.
NeuroRehabilitation 2011;29:53–9. [31] Ansari NN, Naghdi S, Moammeri H, Jalaie S. Ashworth Scales are unreliable
[12] de Jong LD, Nieuwboer A, Aufdemkampe G. Contracture preventive position- for the assessment of muscle spasticity. Physiother Theory Pract 2006;22:
ing of the hemiplegic arm in subacute stroke patients: a pilot randomized 119–25.
controlled trial. Clin Rehabil 2006;20:656–67. [32] Behm DG, Chaouachi A. A review of the acute effects of static and dynamic
[13] Kim EH, Chang MC, Seo JP, Jang SH, Song JC, Jo HM. The effect of a hand- stretching on performance. Eur J Appl Physiol 2011;111:2633–51.
stretching device during the management of spasticity in chronic hemiparetic [33] Fleuren JF, Voerman GE, Erren-Wolters CV, Snoek GJ, Rietman JS, Hermens HJ,
stroke patients. Ann Rehabil Med 2013;37:235–40. et al. Stop using the Ashworth Scale for the assessment of spasticity. J Neurol
[14] Carda S, Invernizzi M, Baricich A, Cisari C. Casting, taping or stretching after Neurosurg Psychiatry 2010;81:46–52.
botulinum toxin type A for spastic equinus foot: a single-blind randomized [34] Li F, Wu Y, Li X. Test-retest reliability and inter-rater reliability of the Modified
trial on adult stroke patients. Clin Rehabil 2011;25:1119–27. Tardieu Scale and the Modified Ashworth Scale in hemiplegic patients with
[15] de Jong LD, Dijkstra PU, Gerritsen J, Geurts AC, Postema K. Combined arm stroke. Eur J Phys Rehabil Med 2014;50:9–15.
stretch positioning and neuromuscular electrical stimulation during rehabili- [35] Mutlu A, Livanelioglu A, Gunel MK. Reliability of Ashworth and Modified
tation does not improve range of motion, shoulder pain or function in patients Ashworth Scales in children with spastic cerebral palsy. BMC Musculoskelet
after stroke: a randomised trial. J Physiother 2013;59:245–54. Disord 2008;9:44.