Spasticity: mixed neurological conditions · Review · 2024
Does shockwave ease muscle stiffness (spasticity) after stroke or in cerebral palsy?
In this review of 16 studies with children who have cerebral palsy and adults after a stroke, stiffness scores were about 1 point lower (on a 0–4 scale) with shockwave than in control groups for the arms, and about 0.4 lower for the legs. With shockwave, stiffness stayed below its starting level for up to 12 weeks.
Helped more than control groups- studies
- 16
- children with cerebral palsy and adults after stroke
- 704
- weeks after treatment: stiffness still below its starting level with shockwave
- 12
Good to know: Stiffness was rated by how much a muscle resists being stretched; studies mixed children and adults and used different settings.
Otero-Luis I, Cavero-Redondo I, Álvarez-Bueno C, et al. Journal of Clinical Medicine, 2024 · Read the paper ↗
How this study worked
What happened
- Compared with
- Control groups, which differed from study to study
Muscle stiffness compared with control groups
Modified Ashworth scale, 0–4 · below zero means less stiffness with shockwave
Show the numbers
| Difference (points on the 0–4 stiffness scale) | likely range | |
|---|---|---|
| Arm muscles | −1.05 | −1.39 to −0.71 |
| Leg muscles | −0.4 | −0.77 to −0.03 |
What the trials used
- Modality
- Radial or focused ESWT were eligible; the review table does not report each study’s modality.
- Reported protocol range
- Across Table 1, measurement times range from immediately after therapy to week 16; Results §3.1 describes included study lengths of 1–16 weeks. These figures describe assessment/study duration, not ESWT treatment duration. Other study-level ranges remain descriptive: 4–15 Hz, 0.03–0.84 mJ/mm² where numerical, and 800–4,000 impacts per muscle. Results §3.1 gives 10 Hz as the maximum frequency, but Table 1 lists 15 Hz for Tabra et al. (2021).
- Comparator and co-intervention
- Control groups were required; comparator procedures varied. Studies combining ESWT with other treatments were excluded under review eligibility criteria.
- Treatment targeting
- Targets varied by trial and included upper- and lower-limb muscles; Table 1 lists the study-specific targets.
Who took part
Children with cerebral palsy and adults with long-term stroke (704 people), plus 68 people with multiple sclerosis in one study.
Side effects
The review did not count side effects. Some included studies described the treatment as safe, painless or free of complications.
All the numbers and the full appraisal
The results, in numbers.
Review of 16 controlled studies reported as 704 participants: children with cerebral palsy, adults with chronic stroke and a small multiple-sclerosis component. Author subgroup counts: 242 cerebral-palsy and 462 chronic-stroke participants. The Table 1 sample sizes add up to 772, including 68 in the one multiple-sclerosis study; the reported 704 equals the two subgroup counts combined.
Modified Ashworth Scale (MAS)9 reported outcomes
Upper-limb MAS pooled mean difference
Modified Ashworth Scale (MAS) MAS points (mean difference)
ESWT versus control across upper-limb studies
I²=7.7%
Random-effects pooled estimate; adjustment not reported
The pooled estimate favored ESWT; the confidence interval excludes no difference.
Analysis details and source
Review-level pooled comparison; this is not a within-arm change. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
Lower-limb MAS pooled mean difference
Modified Ashworth Scale (MAS) MAS points (mean difference)
ESWT versus control across lower-limb studies
I²=0.0%
Random-effects pooled estimate; adjustment not reported
The pooled estimate favored ESWT; the confidence interval excludes no difference.
Analysis details and source
Review-level pooled comparison; this is not a within-arm change. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
MAS pooled mean difference within ESWT groups: immediately post-intervention
Modified Ashworth Scale (MAS) MAS points (mean difference)
Within-ESWT-group pooled change (time-period meta-analysis performed in the intervention group); not an ESWT versus control comparison
I²=85.8%
Random-effects pooled estimate; adjustment not reported
The authors report a significant decrease in MAS score in the pooled ESWT groups at this time point; the confidence interval excludes no change. This is a within-group change and does not show a benefit over control.
Analysis details and source
Methods §2.5 states that the time-period meta-analysis was performed in the intervention group (IG), so this is a pooled within-group change in ESWT arms, not a between-group contrast; a significant within-group change does not establish superiority over control. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
MAS pooled mean difference within ESWT groups: 1 week
Modified Ashworth Scale (MAS) MAS points (mean difference)
Within-ESWT-group pooled change (time-period meta-analysis performed in the intervention group); not an ESWT versus control comparison
I²=76.6%
Random-effects pooled estimate; adjustment not reported
The authors report a significant decrease in MAS score in the pooled ESWT groups at this time point; the confidence interval excludes no change. This is a within-group change and does not show a benefit over control.
Analysis details and source
Methods §2.5 states that the time-period meta-analysis was performed in the intervention group (IG), so this is a pooled within-group change in ESWT arms, not a between-group contrast; a significant within-group change does not establish superiority over control. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
MAS pooled mean difference within ESWT groups: 2–4 weeks
Modified Ashworth Scale (MAS) MAS points (mean difference)
Within-ESWT-group pooled change (time-period meta-analysis performed in the intervention group); not an ESWT versus control comparison
I²=74.8%
Random-effects pooled estimate; adjustment not reported
The authors report a significant decrease in MAS score in the pooled ESWT groups at this time point; the confidence interval excludes no change. This is a within-group change and does not show a benefit over control.
Analysis details and source
Methods §2.5 states that the time-period meta-analysis was performed in the intervention group (IG), so this is a pooled within-group change in ESWT arms, not a between-group contrast; a significant within-group change does not establish superiority over control. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
MAS pooled mean difference within ESWT groups: 5–12 weeks
Modified Ashworth Scale (MAS) MAS points (mean difference)
Within-ESWT-group pooled change (time-period meta-analysis performed in the intervention group); not an ESWT versus control comparison
I²=86.4%
Random-effects pooled estimate; adjustment not reported
The authors report a significant decrease in MAS score in the pooled ESWT groups at this time point; the confidence interval excludes no change. This is a within-group change and does not show a benefit over control.
Analysis details and source
Methods §2.5 states that the time-period meta-analysis was performed in the intervention group (IG), so this is a pooled within-group change in ESWT arms, not a between-group contrast; a significant within-group change does not establish superiority over control. Heterogeneity is reported by the authors. Do not interpret a nonsignificant estimate as equivalence.
MAS pooled mean difference within ESWT groups: >12 weeks
Modified Ashworth Scale (MAS) MAS points (mean difference)
Within-ESWT-group pooled change (time-period meta-analysis performed in the intervention group); not an ESWT versus control comparison
Not stated for this stratum in narrative
Random-effects pooled estimate; adjustment not reported
The confidence interval includes no difference; the source reports no statistically significant result for this time stratum. This is a within-group estimate in ESWT groups, not a comparison with control.
Analysis details and source
Methods §2.5 states that the time-period meta-analysis was performed in the intervention group (IG), so this is a pooled within-group change in ESWT arms, not a between-group contrast; a significant within-group change does not establish superiority over control. No heterogeneity statistic is given in the text for this stratum. Do not interpret a nonsignificant estimate as equivalence.
MAS in adults with chronic stroke
Modified Ashworth Scale (MAS) Mean difference (95% CI)
ESWT versus control, chronic-stroke subgroup
The authors report a subgroup estimate favoring ESWT; the underlying supplementary table was unavailable for independent cell-level verification. This population-specific finding should not be generalized across diagnoses.
Analysis details and source
Review-reported subgroup summary; supplementary Table S2 was cited but unavailable in the retrieved XML.
MAS in children with cerebral palsy
Modified Ashworth Scale (MAS) Mean difference (95% CI)
ESWT versus control, cerebral-palsy subgroup
The authors report a subgroup estimate favoring ESWT; the underlying supplementary table was unavailable for independent cell-level verification. This population-specific finding should not be generalized across diagnoses.
Analysis details and source
Review-reported subgroup summary; supplementary Table S2 was cited but unavailable in the retrieved XML.
How treatment was delivered
Radial or focused ESWT were eligible; the review table does not report each study’s modality.
Across Table 1, measurement times range from immediately after therapy to week 16; Results §3.1 describes included study lengths of 1–16 weeks. These figures describe assessment/study duration, not ESWT treatment duration. Other study-level ranges remain descriptive: 4–15 Hz, 0.03–0.84 mJ/mm² where numerical, and 800–4,000 impacts per muscle. Results §3.1 gives 10 Hz as the maximum frequency, but Table 1 lists 15 Hz for Tabra et al. (2021).
Control groups were required; comparator procedures varied. Studies combining ESWT with other treatments were excluded under review eligibility criteria.
Targets varied by trial and included upper- and lower-limb muscles; Table 1 lists the study-specific targets.
Harms were not synthesized as a review outcome. The discussion cites included studies describing ESWT as safe, painless or without complications, but no pooled adverse-event counts or surveillance method are reported.
Author-reported: clinical heterogeneity in population, age, aetiology and treatment parameters; small samples; dependence on MAS alone; possible assessor and participant bias; insufficient evidence for a multiple-sclerosis subgroup.
The authors report possible publication bias for upper-limb results by Egger test (p=0.062; threshold 0.1), but not lower-limb results (p=0.237).
Extractor observation: the Discussion gives conflicting directions for treatment duration/energy and MAS reduction compared with the Results meta-regression statement. The direction is unresolved and no dose conclusion is drawn.
Referenced supplements were unavailable in the retrieved XML package, limiting direct review of detailed search strategies, risk-of-bias figures, and supplementary meta-regression tables.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Author-reported: clinical heterogeneity in population, age, aetiology and treatment parameters; small samples; dependence on MAS alone; possible assessor and participant bias; insufficient evidence for a multiple-sclerosis subgroup.
- The authors report possible publication bias for upper-limb results by Egger test (p=0.062; threshold 0.1), but not lower-limb results (p=0.237).
- Extractor observation: the Discussion gives conflicting directions for treatment duration/energy and MAS reduction compared with the Results meta-regression statement. The direction is unresolved and no dose conclusion is drawn.
- Referenced supplements were unavailable in the retrieved XML package, limiting direct review of detailed search strategies, risk-of-bias figures, and supplementary meta-regression tables.
Source-linked critical reading
Systematic review with meta-analysis
Technical source record
Complete article XML, all 26 section records, Table 1 with its footnote, and captions for Figures 1–3 read. The original XML was checked against the parsed extraction. Supplementary Table S1–S3 and Figures S1–S4 are referenced but absent from the source package; they were not visually reviewed.
This review found lower average spasticity scores after ESWT in several analyses, with the clearest reported follow-up through 12 weeks. The studies and treatment methods varied, and the review did not show that these score changes reliably improve daily function or identify the best dose.
In children with cerebral palsy and adults with chronic stroke, what effect does extracorporeal shock-wave therapy have on spasticity measured by the Modified Ashworth Scale, by limb and after treatment?
PRISMA systematic review and random-effects meta-analysis, registered in PROSPERO (CRD42023436889). Searches covered four databases through December 2023. The authors included 16 controlled trials: 14 randomized trials and two crossover trials as described in Results; the paper’s study citations and some table details are inconsistent, so this count follows the authors’ narrative.
The review reports 704 participants across children with cerebral palsy, adults with chronic stroke, and one multiple-sclerosis study. The authors’ subgroup totals are 242 children with cerebral palsy and 462 adults with chronic stroke. The Table 1 sample sizes add up to 772, including 68 in the multiple-sclerosis study, so the reported 704 matches only the two subgroup totals. These pooled review totals are not one trial cohort; the age and disease groups differ.
Both radial and focused ESWT were eligible. Table 1 shows varied target muscles, frequency (4–15 Hz across listed rows; the Results text gives 10 Hz as the maximum), energy (0.03–0.84 mJ/mm² where numeric), impacts (800–4,000 per muscle), and measurement times from immediately after therapy to week 16; the Results describe study lengths of one to 16 weeks. Some entries report NR or ranges; this review does not define one protocol. Co-interventions and sham details vary by trial and are not consistently specified in the extracted table.
Eligible studies had a control group, but the review does not pool one uniform comparator. Combined treatments were excluded by the review’s eligibility criteria. Control procedures vary among the included trials and should not be treated as a single sham or usual-care comparison.
The primary synthesis used the Modified Ashworth Scale (MAS). The article describes the source scale as 0 (normal tone) to 4 (rigid), though it does not clearly establish whether every included trial used an identical scoring convention. Pooled estimates are mean differences with reported 95% confidence intervals.
Pooled limb-specific estimates favored ESWT: upper limb MD −1.05 (95% CI −1.39 to −0.71; I² 7.7%) and lower limb MD −0.40 (95% CI −0.77 to −0.03; I² 0%). By follow-up, the Methods state that the meta-analysis was performed in the intervention group, so these estimates describe MAS change within ESWT groups, not differences from control: immediate −1.35 (−1.89 to −0.81; I² 85.8%), 1 week −0.79 (−1.16 to −0.42; I² 76.6%), 2–4 weeks −0.92 (−1.25 to −0.60; I² 74.8%), 5–12 weeks −0.75 (−1.22 to −0.27; I² 86.4%), and >12 weeks −0.47 (−1.30 to 0.35; not statistically significant). These are review-reported estimates; the source does not provide a common patient-important threshold. The authors also report population subgroup estimates: chronic stroke MD −1.11 (95% CI −1.42 to −0.80) and cerebral palsy MD −0.43 (−0.79 to −0.06); Supplementary Table S2 was cited but unavailable, so these are retained as review-reported values.
The review did not evaluate harms as a pooled outcome. The discussion cites included studies describing treatment as safe or painless, but this is not systematic adverse-event surveillance and absence of pooled reporting is not evidence of zero harm.
Authors identify clinical heterogeneity in age, cause of spasticity, dose, frequency and impact count; small samples; reliance on one tone scale; possible assessor and participant bias; and too few multiple-sclerosis studies for subgroup analysis. Egger’s test suggested possible publication bias for upper-limb studies (p=0.062 using the stated 0.1 threshold); no signal was reported for lower-limb studies (p=0.237). Supplementary Tables S1–S3 and Figures S1–S4 are referenced but their source assets were not present in the retrieved XML package, limiting checks of search strings, risk-of-bias judgments, meta-regression and publication-bias plots.
A lower MAS score indicates less resistance on passive movement, but this impairment measure does not itself establish better walking, hand use, comfort, care burden, or independence. Results combine distinct ages and diagnoses, and high heterogeneity affects several follow-up estimates. The review does not establish an optimal protocol or a durable functional benefit.
Future adequately powered trials should stratify by diagnosis, age and limb; prespecify clinically meaningful functional outcomes alongside tone; report allocation, assessor blinding, adverse events and complete time-specific group data; and compare protocol parameters directly rather than infer dose-response from across-study meta-regression.
Read the original article.
Effectiveness of Extracorporeal Shock Wave Therapy in Treatment of Spasticity of Different Aetiologies: A Systematic Review and Meta-Analysis.
Otero-Luis I, Cavero-Redondo I, Álvarez-Bueno C, Martinez-Rodrigo A, Pascual-Morena C, Moreno-Herráiz N, Saz-Lara A. · Journal of clinical medicine · 2024
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →