Abstract
Background
Extracorporeal shockwave therapy (ESWT) is widely applied to manage musculoskeletal disorders, including shoulder pathologies. However, the optimal therapeutic dosage for adhesive capsulitis (AC) remains unclear. This study aimed to evaluate the dose-response relationship of ESWT in patients with AC.
Methods
In this randomized controlled trial, 48 patients with AC refractory to conservative management were assigned to receive a single session of either basic-dose (3,000 pulses) or high-dose (6,000 pulses) ESWT. Shoulder range of motion was assessed at baseline, immediately after treatment, and at 1, 2, 4, 8, and 12 weeks. Functional outcomes were evaluated at 12 weeks using the Constant Murley score. Subgroup analysis was performed based on baseline abduction (cutoff: 90°).
Results
At 12 weeks, the high-dose group demonstrated significantly greater improvements than the basic-dose group in forward flexion (P < .001), abduction (P < .001), external rotation (P = .004), and internal rotation (P = .001). The largest changes occurred immediately after treatment. Subgroup analysis showed that high-dose ESWT resulted in superior gains in forward flexion, abduction, and internal rotation across baseline abduction strata, whereas no significant between-group difference was observed for external rotation.
Conclusion
High-dose ESWT provided significantly enhanced recovery of shoulder range of motion and function compared with a basic-dose protocol in patients with AC. These findings support a dose-response relationship; however, results should be interpreted as short-term outcomes given the 12-week follow-up and absence of a control group.
Keywords: Adhesive capsulitis, Extracorporeal shockwave therapy, Dose-response, Range of motion, Constant Murley score, Shoulder stiffness
The shoulder is the most mobile joint in the human body and is essential for daily activities, work-related tasks, and sports. Patients with shoulder joint injuries often present with pain and reduced range of motion (ROM). Adhesive capsulitis (AC), a debilitating condition characterized by pain and restricted movement, affects 8.2% and 10.1% of working-aged men and women, respectively.25 Typical symptoms of AC include pain, nocturnal discomfort that disrupts sleep, gradual onset of stiffness, and functional impairment.10 The natural course of AC can last for 2-4 years and is traditionally thought to resolve spontaneously. However, many patients experience prolonged and even lifelong functional limitations.9
Management strategies for AC continue to be debated. A stepwise approach is generally recommended, beginning with pharmacologic pain and inflammation control combined with home-based stretching and supervised physical therapy. If symptoms persist, glenohumeral corticosteroid injection or ultrasound-guided saline distension can be used. For patients seeking faster recovery of motion or those with structural capsular thickening, more invasive procedures, such as manipulation under anesthesia or arthroscopic capsulotomy, are considered.
Extracorporeal shockwave therapy (ESWT) is a widely used energy-based treatment for AC. However, the exact mechanism and pathophysiological link of ESWT to AC remain under investigation. ESWT was initially developed in urology for lithotripsy and has since been applied to various musculoskeletal disorders, including lateral epicondylitis and calcific tendinitis of the shoulder.8,32 The proposed mechanisms of ESWT include energy transmission to soft tissues, stimulation of vascularization, enzyme release, inflammatory cytokine reduction, and increased flexibility of collagen fiber.5,33 Given the combined inflammatory and fibrotic pathology of AC, ESWT has emerged as a potential therapeutic option. In a systematic review and meta-analysis, Fabrizio et al reported that ESWT provided better short-term pain reduction, ROM improvement, and functional outcomes compared with other treatment modalities.2 Furthermore, ESWT may provide an alternative treatment option to corticosteroid injections for patients in whom such injections are contraindicated, including those with impaired glucose control.16 A network meta-analysis further revealed the efficacy of ESWT in achieving pain relief and functional improvement.34
Despite the clinical benefits of ESWT, the optimal dosing strategy for AC has not been clearly defined. Evidence from other musculoskeletal conditions suggests that therapeutic effects may be dose-dependent. In poststroke calf muscle spasticity, higher-dose ESWT has been associated with greater functional improvement and more pronounced elastographic changes compared with lower doses.30 Higher-dose protocols have also shown efficacy in conditions such as early osteonecrosis of the femoral head.29 However, this relationship is not universal, as no dose-dependent benefit was observed in knee osteoarthritis.35 These inconsistencies suggest that the optimal ESWT dose may be condition-specific, and its role in AC has not been established.
We conducted a prospective randomized controlled trial to address this unmet need by comparing 2 ESWT dosing—a basic-dose regimen (3,000 pulses) and a high-dose regimen (6,000 pulses)—in patients who were unresponsive to at least 3 months of conservative treatment. The primary objective was to evaluate differences in shoulder ROM and functional outcomes at 12 weeks. A secondary objective was to assess whether baseline shoulder mobility influenced treatment response. We hypothesized that a higher ESWT dose would yield superior clinical outcomes, with potentially greater benefit in patients with more severe motion restriction.
Methods
Study design and ethical approval
This was a prospective, randomized controlled trial designed to evaluate the dose-response relationship of focused ESWT in patients with AC who remained unresponsive to at least 3 months of conservative treatment. The study protocol was approved by the Institutional Review Board of the participating institutions (IRB/REC no. 108-124-E), and written informed consent was obtained from all participants. The trial was registered at ClinicalTrials.gov (Identifier: NCT07325318) and conducted in accordance with the principles of the Declaration of Helsinki.
Participants
Participants were consecutively recruited from an orthopedic outpatient clinic between April and December 2021. Inclusion criteria were as follows: (1) age ≥20 years; (2) first episode of AC; (3) persistent shoulder pain with restricted motion for ≥3 months; (4) prior management limited to analgesic medications (including acetaminophen, non-steroidal anti-inflammatory drugs, and opioid-containing medications); and (5) active abduction in the scapular plane ≤ 135° (measured with a standard long-arm goniometer). All participants underwent shoulder ultrasound or magnetic resonance imaging to confirm characteristic AC findings, defined as thickening of the coracohumeral ligament (> 2.2 mm) and/or the inferior glenohumeral ligament (> 4.0 mm), and to exclude alternative diagnoses such as rotator cuff pathology or calcific tendinopathy. Imaging reports were interpreted by qualified radiologists or physical medicine and rehabilitation physicians.
Exclusion criteria included any prior intra-articular or periarticular corticosteroid injection to the affected shoulder within the preceding 3 months; concomitant shoulder or cervical pathology; history of shoulder fracture, infection, or tumor; previous shoulder surgery; pacemaker presence; pregnancy; coagulopathy; or the inability to complete follow-up assessments.
Randomization and intervention
Focused ESWT (PiezoWave2; Richard Wolf GmbH, Knittlingen, Germany) was introduced to patients during the recruitment period as a treatment alternative for AC. At our institution, the standard initial ESWT treatment dose is 3,000 pulses per session; this was defined as the basic-dose regimen, whereas 6,000 pulses were defined as the high-dose regimen. Patients who elected to undergo ESWT were subsequently randomized in a 1:1 ratio to receive either the basic-dose or high-dose treatment according to a computer-generated randomization sequence.
Participants underwent one treatment session to reduce variability associated with multisession protocols and to minimize dropout risk. All procedures were performed by a single experienced nurse practitioner to ensure consistency. Before the intervention, patients were informed in detail about the treatment procedure, ROM measurement protocols, potential adverse effects, and their management.
Clinical assessments
Baseline demographic and clinical data were collected, including age, sex, body mass index (BMI), affected shoulder side (right/left), presence of night pain (yes/no), and history of prior corticosteroid injection. A long-arm goniometer was used to assess shoulder ROM in terms of forward flexion, abduction in the scapular plane, external rotation with the elbow at the side, and internal rotation. Internal rotation was quantified through the behind-the-back angle method and expressed in degrees, following Sraj et al.21
Functional outcomes were assessed using the Constant Murley score (0-100, with higher scores indicating better function), which evaluates pain, activities of daily living, ROM, and strength, with muscle strength measured using a handheld dynamometer. This instrument is widely used in shoulder research and has demonstrated high reliability.18
ESWT protocol
Focused ESWT was delivered at an energy flux density of 0.182 mJ/mm2 (level 5) with a frequency of 4 Hz. After application of coupling gel, treatment was administered at 3 predefined anatomical sites: (1) one fingerbreadth inferior to the anterior acromion, (2) lateral to the acromion, and (3) posterior to the acromion. A total of either 3,000 or 6,000 pulses were evenly distributed across the 3 sites, resulting in 1,000 or 2,000 pulses per site, respectively. Ultrasound guidance was not used, as treatment was performed using reproducible surface anatomical landmarks. This approach was selected given the diffuse capsular involvement in AC rather than a discrete focal lesion.
Immediately after ESWT completion, shoulder ROM was reassessed according to the study protocol. Participants were then instructed in a home-based stretching program by a nurse practitioner. The program included wall walks to improve forward flexion and abduction, towel stretches to improve internal rotation, and cane-assisted or elastic band‑assisted exercises to improve external rotation.
Follow-up and outcome measures
ROM was reassessed immediately after treatment and at 1, 2, 4, 8, and 12 weeks. At each follow-up visit, the home-based stretching program was reviewed and reinforced to ensure correct performance; formal quantitative adherence was not recorded. The Constant Murley score was reassessed at 12 weeks.
Statistical analysis
Sample size estimation was based on the primary outcome measure of shoulder ROM. Shoulder abduction was selected as the reference variable because of its functional relevance in AC. In our pilot study of 17 patients with AC treated with ESWT, clinically meaningful improvements in shoulder ROM and Constant Murley scores were observed. Previous reliability research in patients with AC reported a smallest detectable difference of approximately 15° for individual shoulder ROM measurements.22 Together with our pilot observations, this threshold was considered a reasonable target difference for study planning. Based on these pilot observations and the anticipated treatment effect, a large effect size (Cohen's d = 0.9) was assumed. Sample size estimation was performed using G∗Power (version 3.1.9.7; Heinrich Heine University Düsseldorf, Düsseldorf, Germany) based on a 2-tailed Wilcoxon–Mann-Whitney test for 2 independent groups, with a significance level (α) of 0.05, statistical power (1−β) of 0.80, and an allocation ratio of 1:1. A minimum of 22 patients per group was required. To account for an anticipated dropout rate of approximately 10%, the target sample size was increased to 24 patients per group to maintain adequate statistical power at study completion. Data analysis was performed with IBM SPSS Statistics version 25.0 (IBM, Armonk, NY, USA). Given the small sample size (<30 patients) and non-normal distribution determined through normality testing, continuous variables are presented as medians with interquartile ranges. Continuous variables were compared between groups according to a Mann-Whitney U test. Within-group comparisons between 2 time points (eg, baseline vs. 12 weeks) were conducted according to a Wilcoxon signed-rank test. A chi-squared test was performed to compare categorical variables. To address potential baseline imbalance in outcome measures, Quade's test, a rank-based non-parametric analog of analysis of covariance, was applied. This method adjusts for baseline values by analyzing ranked residuals, enabling comparison of postintervention outcomes between groups while relaxing assumptions of normality and linearity. This approach is particularly suitable for small sample sizes and non-normally distributed data.
To investigate baseline predictors of treatment response, we stratified patients according to their pretreatment shoulder abduction angle into 2 groups: ≥90° and <90°. Abduction was selected for stratification because of its functional relevance in AC. The 90° was chosen to provide a simple and clearly defined classification criterion. Between-group comparisons (basic-dose vs. high-dose) were performed using the Mann-Whitney U test. Subgroup analyses were considered exploratory.
The effects of ESWT dosage (between-subject factor: 3,000 vs. 6,000 pulses) and time (within-subject factor: baseline; immediately post-treatment; and 1-week, 2-week, 4-week, 8-week, and 12-week post-treatment) on ROM were assessed through 2-way repeated-measures analysis of variance. Mauchly's test was applied to evaluate sphericity; where sphericity was violated, the Greenhouse-Geisser correction was used.
When a significant group × time interaction was observed, simple main effects were analyzed within each group. Bonferroni-adjusted pairwise comparisons were then performed to identify differences between consecutive time points. Changes in ROM between consecutive time points were expressed as ΔT1-ΔT6, defined as follows:
ΔT1: baseline to immediately after treatment.
ΔT2: immediately after treatment to 1 week.
ΔT3: 1 week to 2 weeks.
ΔT4: 2 weeks to 4 weeks.
ΔT5: 4 weeks to 8 weeks.
ΔT6: 8 weeks to 12 weeks.
Blinding and significance threshold
An independent statistician blinded to group allocation and clinical data performed all statistical analyses. A 2-tailed P value of < .05 was considered significant.
Results
A total of 48 patients who completed the 12-week evaluation were included in the final analysis, with 24 patients assigned to the basic-dose group (3,000 pulses) and 24 to the high-dose group (6,000 pulses) (Fig. 1).
Figure 1.
Flow diagram of the study, showing patient enrollment, randomization, allocation to extracorporeal shockwave therapy basic-dose and high-dose groups, follow-up, and inclusion in the final analysis. ESWT, extracorporeal shockwave therapy.
In the basic-dose group, the median age and BMI were 52 years and 25.3 kg/m2, respectively. Median shoulder motion values were 105° for forward flexion, 79° for abduction, 38° for external rotation, and 32° for internal rotation, with a median Constant Murley score of 27. In the high-dose group, the median age was 52 years and BMI was 26.7 kg/m2. Median shoulder motion measurements were 101° for forward flexion, 94° for abduction, 29° for external rotation, and 30° for internal rotation, with a median Constant Murley score of 33. No significant between-group differences were observed in demographic variables or pretreatment ROM values (Table I). Baseline imaging confirmed AC in all participants, with no alternative structural pathology identified.
Table I.
Baseline characteristics of the basic-dose and high-dose groups
| Characteristics | Basic-dose (n = 24) | High-dose (n = 24) | P value |
|---|---|---|---|
| Age (yrs), median (IQR) | 52 (8) | 52 (6) | .649 |
| Female sex, n (%) | 12 (50.0) | 11 (45.8) | .733 |
| BMI (kg/m2), median (IQR) | 25.3 (4.8) | 26.7 (4.4) | .117 |
| Right shoulder affected, n (%) | 13 (54.2) | 14 (58.3) | .771 |
| Night pain, n (%) | 21 (87.5) | 22 (91.7) | .637 |
| Prior corticosteroid injection, n (%) | 0 (0) | 0 (0) | 1.000 |
| Forward flexion (°), median (IQR) | 105 (39) | 101 (25) | .836 |
| Abduction (°), median (IQR) | 79 (29) | 94 (27) | .260 |
| External rotation (°), median (IQR) | 38 (23) | 29 (17) | .671 |
| Internal rotation (°), median (IQR) | 32 (24) | 30 (15) | .649 |
| Constant Murley score, median (IQR) | 27 (5) | 33 (15) | .020∗,† |
IQR, interquartile range; BMI, body mass index.
P < .05. P values were calculated using the Mann-Whitney U test for continuous variables and the chi-squared test for categorical variables.
Baseline Constant Murley scores differed between groups and were adjusted in subsequent analyses using Quade's test.
Following ESWT, ROM and Constant Murley scores improved significantly at 12 weeks in both groups (Table II), with greater ROM in the high-dose group (Table III). Despite higher baseline Constant Murley scores in the high-dose group (Table I), baseline differences were accounted for using the Quade's test, after which the high-dose group still demonstrated significantly higher post-treatment scores (Table III).
Table II.
Within-group changes in shoulder ROM and Constant Murley score from baseline to 12 weeks
| Outcomes | Baseline, median (IQR) | 12 weeks post-treatment, median (IQR) | P value |
|---|---|---|---|
| Basic-dose group | |||
| Forward flexion (°) | 105 (39) | 166 (16) | <.001 |
| Abduction (°) | 79 (29) | 164 (12) | <.001 |
| External rotation (°) | 38 (23) | 78 (16) | <.001 |
| Internal rotation (°) | 32 (24) | 107 (12) | <.001 |
| Constant Murley score | 27 (5) | 78 (4) | <.001 |
| High-dose group | |||
| Forward flexion (°) | 101 (25) | 176 (6) | <.001 |
| Abduction (°) | 94 (27) | 174 (7) | <.001 |
| External rotation (°) | 29 (17) | 82 (10) | <.001 |
| Internal rotation (°) | 30 (15) | 121 (11) | <.001 |
| Constant Murley score | 33 (15) | 86 (6) | <.001 |
ROM, range of motion; IQR, interquartile range.
P values were calculated using the Wilcoxon signed-rank test for within-group comparisons between baseline and 12 weeks post-treatment.
Table III.
Adjusted between-group comparison of shoulder outcomes at 12 weeks
| Outcomes | Basic-dose, median (IQR) | High-dose, median (IQR) | P value |
|---|---|---|---|
| Forward flexion (°) | 166 (16) | 176 (6) | <.001 |
| Abduction (°) | 164 (12) | 174 (7) | <.001 |
| External rotation (°) | 78 (16) | 82 (10) | .004 |
| Internal rotation (°) | 107 (12) | 121 (11) | .001 |
| Constant Murley score | 78 (4) | 86 (6) | <.001∗ |
IQR, interquartile range.
P values were calculated using the Mann-Whitney U test for between-group comparisons at 12 weeks post-treatment.
Constant Murley scores were adjusted for baseline differences using Quade's test.
For patients with baseline abduction ≥90°, high-dose ESWT resulted in significantly greater final forward flexion, abduction, and internal rotation compared with basic-dose group (Table IV). Among patients with baseline abduction <90°, high-dose ESWT similarly yielded significantly greater improvements in final forward flexion, abduction, and internal rotation and higher Constant Murley scores, compared with the basic-dose group (Table V). No significant between-group differences were observed for external rotation in either subgroup.
Table IV.
Exploratory subgroup analysis of outcomes in patients with baseline abduction ≥90°
| Outcomes | Basic-dose (n = 6), median (IQR) | High-dose (n = 15), median (IQR) | P value |
|---|---|---|---|
| Baseline | |||
| Forward flexion (°) | 114 (26) | 108 (36) | .622 |
| Abduction (°) | 101 (6) | 98 (10) | .470 |
| External rotation (°) | 39 (32) | 39 (25) | .677 |
| Internal rotation (°) | 30 (11) | 32 (31) | .622 |
| Constant Murley score | 33 (7) | 37 (6) | .045 |
| 12 weeks post-treatment | |||
| Forward flexion (°) | 168 (9) | 177 (5) | <.001 |
| Abduction (°) | 161 (8) | 175 (7) | <.001 |
| External rotation (°) | 75 (9) | 82 (10) | .080 |
| Internal rotation (°) | 107 (9) | 121 (11) | .002 |
| Constant Murley score | 79 (3) | 86 (8) | .003 |
IQR, interquartile range.
P values were calculated using the Mann-Whitney U test for between-group comparisons (basic-dose vs. high-dose) at each time point. No adjustment for multiple comparisons was performed. Subgroup analyses were exploratory and not powered for formal hypothesis testing.
Table V.
Exploratory subgroup analysis of outcomes in patients with baseline abduction <90°
| Outcomes | Basic-dose (n = 18), median (IQR) | High-dose (n = 9), median (IQR) | P value |
|---|---|---|---|
| Baseline | |||
| Forward flexion (°) | 88 (36) | 87 (19) | .860 |
| Abduction (°) | 78 (20) | 68 (24) | .232 |
| External rotation (°) | 36 (25) | 27 (9) | .212 |
| Internal rotation (°) | 35 (30) | 23 (25) | .176 |
| Constant Murley score | 26 (4) | 24 (8) | .348 |
| 12 weeks post-treatment | |||
| Forward flexion (°) | 163 (20) | 175 (6) | .001 |
| Abduction (°) | 165 (16) | 172 (6) | .001 |
| External rotation (°) | 78 (21) | 82 (10) | .053 |
| Internal rotation (°) | 106 (22) | 121 (11) | .035 |
| Constant Murley score | 78 (6) | 87 (5) | .001 |
IQR, interquartile range.
P values represent between-group comparisons (basic-dose vs. high-dose) at each time point using Mann-Whitney U tests. No adjustment for multiple comparisons was performed. Subgroup analyses were exploratory and not powered for formal hypothesis testing.
Significant group × time interaction effects were observed for forward flexion (P = .047), external rotation (P = .006), and internal rotation (P = .019) (Table VI). The greatest improvements occurred immediately after ESWT (ΔT1), particularly in the high-dose group. Although both groups showed initial increase, sustained progression through week 8 was primarily observed in the high-dose group, most notably for forward flexion. By week 12, the magnitude of improvement appeared to plateau, with some attenuation in rotational, as reflected by smaller ΔT6 values. External and internal rotation demonstrated a more transient pattern, with early gains diminishing over time. Table VII summarizes median joint angle changes for each interval; all P values were Bonferroni-adjusted.
Table VI.
Results of 2-way repeated-measures ANOVA for shoulder range of motion
| Motion | Group effect (P) | Time effect (P) | Group × time interaction (P) |
|---|---|---|---|
| Forward flexion | .066 | <.001 | .047 |
| External rotation | .010 | <.001 | .006 |
| Internal rotation | .021 | <.001 | .019 |
ANOVA, analysis of variance.
Data were analyzed using 2-way repeated-measures analysis of variance. The Greenhouse-Geisser correction was applied when the assumption of sphericity was violated (Mauchly's test, P < .05 for all motions).
Table VII.
Changes in joint angles between follow-up intervals (ΔT1-ΔT6) by treatment groups
| Motion | Group | △T1 | △T2 | △T3 | △T4 | △T5 | △T6 | Post hoc summary |
|---|---|---|---|---|---|---|---|---|
| Forward flexion | Basic-dose Angle changes (°), median (IQR) |
28 (23) | 4 (4) | 6 (11) | 8 (9) | 6 (15) | 4 (5) | △T1 > △T2–△T6 |
| High-dose Angle changes (°), median (IQR) |
37 (11) | 6 (4) | 9 (7) | 9 (7) | 8 (10) | 3 (8) | △T1 > △T2–△T6; △T4, △T5 > △T6 | |
| External rotation | Basic-dose Angle changes (°), median (IQR) |
13 (12) | 3 (4) | 2 (5) | 2 (6) | 4 (7) | 3 (7) | △T1 > △T2–△T6 |
| High-dose Angle changes (°), median (IQR) |
22 (3) | 4 (6) | 5 (5) | 4 (5) | 3 (15) | 2 (3) | △T1 > △T2–△T6; △T3 > △T6 | |
| Internal rotation | Basic-dose Angle changes (°), median (IQR) |
26 (36) | 5 (7) | 6 (5) | 6 (7) | 5 (11) | 7 (15) | △T1 > △T2–△T6 |
| High-dose Angle changes (°), median (IQR) |
41 (12) | 5 (6) | 6 (10) | 11 (7) | 9 (8) | 8 (10) | △T1 > △T2–△T6 |
IQR, interquartile range.
Each △T represents the change in joint angle between 2 consecutive follow-up intervals. Given that the interaction between group and time was significant for these motions (Table VI), post hoc tests were conducted separately within each group. All comparisons are based on Bonferroni-adjusted post hoc tests.
For abduction, repeated-measures analysis of variance revealed a significant main effect of time (P < .001) but not of group (P = .178) or interaction between group and time (P = .183) (Table VIII), indicating that both dosing regimens yielded improvements that followed similar temporal patterns. Median changes in the abduction angle across consecutive intervals are presented in Table IX. The most substantial increase was noted immediately after treatment (ΔT1), with a median increase of 37° in the high-dose group and of 28° in the basic-dose group. Although moderate changes were observed at subsequent intervals (ΔT2-ΔT5), their magnitudes were smaller and did not differ significantly from one another. Bonferroni-adjusted pairwise comparisons confirmed that the increase at ΔT1 was significantly greater than those observed at later intervals for both groups.
Table VIII.
Results of 2-way repeated-measures ANOVA for shoulder abduction
| Motion | Group effect (P) | Time effect (P) | Group × time interaction (P) |
|---|---|---|---|
| Abduction | .178 | <.001 | .183 |
ANOVA, analysis of variance.
Data were analyzed using 2-way repeated-measures analysis of variance. The Greenhouse-Geisser correction was applied when the assumption of sphericity was violated (Mauchly's test, P < .05).
Table IX.
Changes in shoulder abduction angle across follow-up intervals (ΔT1-ΔT6) by treatment groups
| Motion | Group | △T1 | △T2 | △T3 | △T4 | △T5 | △T6 | Post hoc summary |
|---|---|---|---|---|---|---|---|---|
| Abduction | Basic-dose Angle changes (°), median (IQR) |
28 (23) | 4 (4) | 6 (11) | 8 (9) | 6 (15) | 4 (5) | △T1 > △T2–△T6 |
| High-dose Angle changes (°), median (IQR) |
37 (11) | 6 (4) | 9 (7) | 9 (7) | 8 (10) | 3 (8) |
IQR, interquartile range.
Each ΔT represents the change in shoulder abduction angle between 2 consecutive follow-up intervals. Given that the interaction between group and time was non-significant (Table VIII: P = .183), post hoc tests were not conducted separately within each group. All comparisons are based on Bonferroni-adjusted post hoc tests.
No patients experienced immediate pain requiring discontinuation of ESWT during treatment. Transient, mild adverse events—such as localized swelling and bruising—were observed in 5 participants (2 in the basic-dose group and 3 in the high-dose group). These symptoms basically occurred 1-3 days after treatment and resolved spontaneously within a short period with conservative management, including cold therapy. No residual symptoms were noted at the first follow-up (1 week). No permanent iatrogenic injuries were reported.
Two patients in the basic-dose group demonstrated persistent abduction <120° at week 12 and subsequently required additional interventions.
Discussion
In this study, a single session of high-dose ESWT resulted in clinically and statistically significant improvements in shoulder ROM and functional outcomes in patients with AC. Compared with the basic-dose protocol, the high-dose regimen produced greater and more consistent gains across all motion planes and Constant Murley scores, irrespective of baseline abduction status. The most pronounced changes in ROM were observed immediately after treatment, indicating an early treatment response, with forward flexion demonstrating continued improvement during follow-up.
Although ESWT is widely used in the management of musculoskeletal disorders, the optimal dosing strategy for AC remains undefined. To our knowledge, this is the first study directly comparing different ESWT doses in this condition. Previous reports have typically employed 2,000-4,000 pulses per session across various indications.35 In poststroke calf muscle spasticity, higher-dose ESWT has demonstrated superior functional and elastographic outcomes compared with lower-dose protocols.30 Higher-energy ESWT has also been associated with improved outcomes in rotator cuff disease and calcific tendinopathy.7,24 Collectively, these findings support a tissue-level dose responsiveness of ESWT, which our data extend to AC, a condition characterized by combined capsular fibrosis and inflammation.12,16
The dose-dependent effect observed in this study is likely multifactorial, involving both biomechanical and biological mechanisms. Shockwaves generate rapid pressure gradients and cavitation forces that produce micromechanical disruption of fibrotic adhesions within the capsule and rotator interval, key pathological substrates in AC.4,15 Cavitation effects are known to be dose-sensitive,11 suggesting that higher pulse counts may amplify tissue-level remodeling. Supporting evidence from randomized studies in stroke-related spasticity demonstrates that higher-dose ESWT can induce structural changes in connective tissue, accompanied by improvements in ROM and function.31 In parallel, ESWT exerts analgesic effects through modulation of nociceptive signaling, reduction of inflammatory mediators, and activation of central inhibitory pathways.20 The combination of mechanical tissue modulation and rapid analgesia likely explains both the immediate ROM gains and sustained functional improvements observed in the high-dose group.
Given the diffuse and multidirectional involvement of AC, shockwaves were distributed across anterior, lateral, and posterior shoulder regions. The high-dose protocol (6,000 pulses) ensured adequate energy delivery while remaining within established safety limits. This multiregional application strategy may have contributed to the enhanced clinical outcomes observed.
The optimal number of ESWT sessions for AC has not been clearly established. A single-session design was therefore selected to isolate dose effects and avoid cumulative confounding. The immediate ROM improvement is likely driven by combined analgesic and neuromuscular effects, whereas longer-term gains may reflect structural adaptation, including capsular remodeling. Although repeated sessions are often advocated, prior clinical6,13 and experimental27,28 studies suggest that ESWT induces sustained biological effects, including neovascularization and upregulation of regenerative mediators such as endothelial nitric oxide synthase and vascular endothelial growth factor. Consistent with this, improvements persisted throughout the 12-week follow-up in our cohort. However, without a multisession comparator, relative efficacy cannot be determined. Nonetheless, a single-session approach may offer practical advantages, including improved patient adherence and reduced healthcare burden.
A notable finding was the temporal pattern of recovery. The largest ROM gains occurred immediately after treatment, particularly in the high-dose group, followed by a relative plateau between 4 and 8 weeks. This pattern suggests that ESWT may function not only as a biological modulator but also as an acute mechanical facilitator of joint mobility. The delivery of 6,000 pulses may reach a threshold sufficient to induce immediate capsular compliance changes, effectively “unlocking” restricted motion, while concurrent analgesic effects facilitate functional utilization of regained range. Beyond the early phase, improvements tended to stabilize, with some attenuation observed by week 12, particularly in rotational movements. This may reflect a transition from rapid mechanical gains to a more gradual biological recovery phase. Although AC is classically self-resolving, part of the observed improvement may reflect its natural history. However, given the short follow-up, the durability of these effects cannot be confirmed and requires further validation.
ESWT should be considered an adjunctive or second-line intervention for patients with AC refractory to conservative management. While our findings demonstrate a clear dose-response effect, the absence of a control group limits direct comparison with established treatments such as corticosteroid injection or physiotherapy. Head-to-head comparative studies are needed to define its relative position in treatment algorithms.
Importantly, ESWT was not applied in isolation. All patients participated in a structured home-based stretching program, which likely contributed to maintenance and consolidation of treatment gains. We propose that high-dose ESWT functions as a biomechanical and analgesic “primer,” creating a therapeutic window that enhances rehabilitation effectiveness.
Subgroup analysis demonstrated that the superiority of high-dose ESWT was generally consistent across baseline mobility strata. Significant improvements were observed in forward flexion, abduction, and internal rotation in both groups. External rotation showed a weaker between-group difference, suggesting potential plane-specific responsiveness. Although AC is traditionally categorized into freezing, frozen, and thawing phases,17 these stages often overlap and may be difficult to distinguish clinically. A ROM-based classification may therefore be more practical. Baseline abduction was selected as a pragmatic stratification marker due to its clinical accessibility and broader excursion range,1 which may allow more consistent assessment compared with rotational measures that are inherently limited and more susceptible to measurement variability. Although abduction limitation is not specific to AC, diagnostic misclassification was minimized through protocol-based clinical and imaging confirmation of AC. However, the use of abduction as an inclusion criterion may influence the characteristics of the study population and should be considered when interpreting the generalizability of the findings. Despite its use for subgroup classification, the influence of baseline abduction on treatment response appeared limited. These findings suggest that ESWT efficacy is largely independent of baseline restriction severity, although confirmation in larger cohorts is required.
Internal rotation showed sustained improvement in the high-dose group. While external rotation is traditionally emphasized in AC,14,26 internal rotation is functionally important for activities such as personal hygiene and reaching behind the back. Internal rotation deficits in AC are commonly associated with posterior capsular tightness and global capsuloligamentous contracture, which may persist into later stages of recovery.19 The observed improvement may therefore reflect reduced capsular stiffness and improved overall joint extensibility rather than changes in a specific soft-tissue structure. These findings suggest that internal rotation may represent an under-recognized but clinically meaningful domain of recovery, and future studies should place greater emphasis on its assessment and targeted restoration in patients with AC.
Focused ESWT was used in this study due to its higher energy density and greater tissue penetration10 compared with radial devices. This may be advantageous in AC, where pathological changes primarily involve the joint capsule and periarticular soft tissues. Accordingly, focused ESWT may provide a theoretical advantage in targeting fibrotic capsular pathology,3,23 although direct comparative evidence with radial ESWT remains limited.
This study has several strengths. Treatment delivery was fully standardized and performed by a single experienced operator using a uniform protocol, minimizing interoperator variability. Objective ROM measurements and predefined follow-up intervals enabled robust longitudinal assessment. No loss to follow-up occurred, eliminating attrition bias and strengthening internal validity.
Several limitations should be acknowledged. First, although the sample size was adequately powered for the primary outcome, it was relatively small and may limit subgroup analyses; therefore, these findings should be interpreted as exploratory. Second, although patients were randomized, enrollment was limited to individuals who elected to receive ESWT, which may introduce selection bias and limit generalizability, as such patients may have higher treatment expectations or motivation. In addition, adherence to the home-based stretching program was not formally quantified and may represent a potential confounding factor. Third, the absence of a sham or conventional-treatment control group precludes definitive conclusions regarding comparative efficacy and does not fully exclude placebo effects. Given the natural history of AC, some observed improvement may reflect spontaneous recovery, although our cohort comprised patients refractory to initial conservative management. Finally, the follow-up period was limited to 12 weeks; therefore, long-term durability, recurrence, and potential effects on disease progression have yet to be determined.
Conclusion
Focused ESWT provides clinically meaningful early gains in ROM and function in patients with AC. A high-dose protocol (6,000 pulses) demonstrated greater improvements across multiple ROM domains and functional outcomes compared with a basic-dose regimen, with generally consistent benefits across different baseline abduction levels. Treatment effects were rapid in onset and sustained over 12 weeks, likely facilitated by adjunctive stretching. These findings suggest that higher-dose ESWT may represent a more effective strategy for overcoming capsular restriction and enabling early functional recovery. However, given the absence of a control group and the limited follow-up duration, these results should be interpreted as reflecting short-term outcomes. Further studies with larger cohorts, longer follow-up, and comparative designs are required to confirm durability and optimize treatment parameters.
Acknowledgments
The authors thank the staff of the Department of Medical Research, National Taiwan University Hospital Hsin-Chu branch, for their assistance with the statistical analysis.
Disclaimers:
Funding: No funding was disclosed by the authors.
Conflicts of interest: The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
Footnotes
The study protocol was approved by the Institutional Review Board of the National Taiwan University Hospital Hsin-Chu branch (IRB/REC no. 108-124-E). All participants were informed that data from the research would be submitted for publication and gave their consent.
Clinical trial registration: This study was registered at ClinicalTrials.gov (Identifier: NCT07325318).
Study site: National Taiwan University Hospital Hsin-Chu Branch, Hsinchu, Taiwan.
References
- 1.Achilova F., Daher M., Nassar J.E., Daniels A.H., Abboud J.A. Frozen shoulder: diagnosis and treatment of adhesive capsulitis. Am J Med. 2026;139:598–605. doi: 10.1016/j.amjmed.2026.01.021. [DOI] [PubMed] [Google Scholar]
- 2.Brindisino F., Girardi G., Crestani M., Fiore A., Giovannico G., Garzonio F., et al. Effectiveness of electrophysical agents in subjects with frozen shoulder: a systematic review and meta-analysis. Disabil Rehabil. 2024;46:3513–3534. doi: 10.1080/09638288.2023.2251880. [DOI] [PubMed] [Google Scholar]
- 3.Chen C.Y., Hu C.C., Weng P.W., Huang Y.M., Chiang C.J., Chen C.H., et al. Extracorporeal shockwave therapy improves short-term functional outcomes of shoulder adhesive capsulitis. J Shoulder Elbow Surg. 2014;23:1843–1851. doi: 10.1016/j.jse.2014.08.010. [DOI] [PubMed] [Google Scholar]
- 4.Cheng J.H., Wang C.J. Biological mechanism of shockwave in bone. Int J Surg. 2015;24(Pt B):143–146. doi: 10.1016/j.ijsu.2015.06.059. [DOI] [PubMed] [Google Scholar]
- 5.El Naggar T., Maaty A.I.E., Mohamed A.E. Effectiveness of radial extracorporeal shock-wave therapy versus ultrasound-guided low-dose intra-articular steroid injection in improving shoulder pain, function, and range of motion in diabetic patients with shoulder adhesive capsulitis. J Shoulder Elbow Surg. 2020;29:1300–1309. doi: 10.1016/j.jse.2020.03.005. [DOI] [PubMed] [Google Scholar]
- 6.Farr S., Sevelda F., Mader P., Graf A., Petje G., Sabeti-Aschraf M. Extracorporeal shockwave therapy in calcifying tendinitis of the shoulder. Knee Surg Sports Traumatol Arthrosc. 2011;19:2085–2089. doi: 10.1007/s00167-011-1479-z. [DOI] [PubMed] [Google Scholar]
- 7.Fatima A., Ahmad A., Gilani S.A., Darain H., Kazmi S., Hanif K. Effects of high-energy extracorporeal shockwave therapy on pain, functional disability, quality of life, and ultrasonographic changes in patients with calcified Rotator Cuff Tendinopathy. Biomed Res Int. 2022;2022 doi: 10.1155/2022/1230857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gerdesmeyer L., Wagenpfeil S., Haake M., Maier M., Loew M., Wortler K., et al. Extracorporeal shock wave therapy for the treatment of chronic calcifying tendonitis of the rotator cuff: a randomized controlled trial. JAMA. 2003;290:2573–2580. doi: 10.1001/jama.290.19.2573. [DOI] [PubMed] [Google Scholar]
- 9.Griggs S.M., Ahn A., Green A. Idiopathic adhesive capsulitis. A prospective functional outcome study of nonoperative treatment. J Bone Joint Surg Am. 2000;82:1398–1407. [PubMed] [Google Scholar]
- 10.Hameedi S.A., Shadmehr A., Malmir K., Fereydounnia S., Shiravi Z. Comparison of application of the radial Vs. Focused probes of extracorporeal shockwave therapy on pain, range of motion and function in patients with adhesive capsulitis. NeuroQuantology. 2022;20:2726–2736. doi: 10.14704/nq.2022.20.6.NQ22266. [DOI] [Google Scholar]
- 11.Huber P., Debus J., Jochle K., Simiantonakis I., Jenne J., Rastert R., et al. Control of cavitation activity by different shockwave pulsing regimes. Phys Med Biol. 1999;44:1427–1437. doi: 10.1088/0031-9155/44/6/301. [DOI] [PubMed] [Google Scholar]
- 12.Itoi E., Arce G., Bain G.I., Diercks R.L., Guttmann D., Imhoff A.B., et al. Shoulder stiffness: current concepts and concerns. Arthroscopy. 2016;32:1402–1414. doi: 10.1016/j.arthro.2016.03.024. [DOI] [PubMed] [Google Scholar]
- 13.Kudo P., Dainty K., Clarfield M., Coughlin L., Lavoie P., Lebrun C. Randomized, placebo-controlled, double-blind clinical trial evaluating the treatment of plantar fasciitis with an extracoporeal shockwave therapy (ESWT) device: a North American confirmatory study. J Orthop Res. 2006;24:115–123. doi: 10.1002/jor.20008. [DOI] [PubMed] [Google Scholar]
- 14.Le H.V., Lee S.J., Nazarian A., Rodriguez E.K. Adhesive capsulitis of the shoulder: review of pathophysiology and current clinical treatments. Shoulder Elbow. 2017;9:75–84. doi: 10.1177/1758573216676786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ogden J.A., Toth-Kischkat A., Schultheiss R. Principles of shock wave therapy. Clin Orthop Relat Res. 2001;387:8–17. doi: 10.1097/00003086-200106000-00003. [DOI] [PubMed] [Google Scholar]
- 16.Redler L.H., Dennis E.R. Treatment of adhesive capsulitis of the shoulder. J Am Acad Orthop Surg. 2019;27:e544–e554. doi: 10.5435/JAAOS-D-17-00606. [DOI] [PubMed] [Google Scholar]
- 17.Reeves B. The natural history of the frozen shoulder syndrome. Scand J Rheumatol. 1975;4:193–196. doi: 10.3109/03009747509165255. [DOI] [PubMed] [Google Scholar]
- 18.Rocourt M.H., Radlinger L., Kalberer F., Sanavi S., Schmid N.S., Leunig M., et al. Evaluation of intratester and intertester reliability of the Constant-Murley shoulder assessment. J Shoulder Elbow Surg. 2008;17:364–369. doi: 10.1016/j.jse.2007.06.024. [DOI] [PubMed] [Google Scholar]
- 19.Shaffer B., Tibone J.E., Kerlan R.K. Frozen shoulder. A long-term follow-up. J Bone Joint Surg Am. 1992;74:738–746. [PubMed] [Google Scholar]
- 20.Simplicio C.L., Purita J., Murrell W., Santos G.S., Dos Santos R.G., Lana J. Extracorporeal shock wave therapy mechanisms in musculoskeletal regenerative medicine. J Clin Orthop Trauma. 2020;11(Suppl 3):S309–S318. doi: 10.1016/j.jcot.2020.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sraj S.A. Internal rotation behind-the-back angle: a reliable angular measurement for shoulder internal rotation behind the back. Sports Health. 2015;7:299–302. doi: 10.1177/1941738113502159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tveita E.K., Ekeberg O.M., Juel N.G., Bautz-Holter E. Range of shoulder motion in patients with adhesive capsulitis; intra-tester reproducibility is acceptable for group comparisons. BMC Musculoskelet Disord. 2008;9:49. doi: 10.1186/1471-2474-9-49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vahdatpour B., Taheri P., Zade A.Z., Moradian S. Efficacy of extracorporeal shockwave therapy in frozen shoulder. Int J Prev Med. 2014;5:875–881. [PMC free article] [PubMed] [Google Scholar]
- 24.Verstraelen F.U., den Kleef N.J., Jansen L., Morrenhof J.W. High-energy versus low-energy extracorporeal shock wave therapy for calcifying tendinitis of the shoulder: which is superior? A meta-analysis. Clin Orthop Relat Res. 2014;472:2816–2825. doi: 10.1007/s11999-014-3680-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Walker-Bone K., Palmer K.T., Reading I., Coggon D., Cooper C. Prevalence and impact of musculoskeletal disorders of the upper limb in the general population. Arthritis Rheum. 2004;51:642–651. doi: 10.1002/art.20535. [DOI] [PubMed] [Google Scholar]
- 26.Walmsley S., Osmotherly P.G., Rivett D.A. Movement and pain patterns in early stage primary/idiopathic adhesive capsulitis: a factor analysis. Physiotherapy. 2014;100:336–343. doi: 10.1016/j.physio.2014.02.001. [DOI] [PubMed] [Google Scholar]
- 27.Wang C.J., Wang F.S., Yang K.D., Weng L.H., Hsu C.C., Huang C.S., et al. Shock wave therapy induces neovascularization at the tendon-bone junction. A study in rabbits. J Orthop Res. 2003;21:984–989. doi: 10.1016/S0736-0266(03)00104-9. [DOI] [PubMed] [Google Scholar]
- 28.Wang C.J., Wang F.S., Yang K.D. Biological effects of extracorporeal shockwave in bone healing: a study in rabbits. Arch Orthop Trauma Surg. 2008;128:879–884. doi: 10.1007/s00402-008-0663-1. [DOI] [PubMed] [Google Scholar]
- 29.Wang C.J., Huang C.C., Yip H.K., Yang Y.J. Dosage effects of extracorporeal shockwave therapy in early hip necrosis. Int J Surg. 2016;35:179–186. doi: 10.1016/j.ijsu.2016.09.013. [DOI] [PubMed] [Google Scholar]
- 30.Yang S.M., Chen Y.H., Lu Y.L., Wu C.H., Chen W.S., Lin M.T. The dose effectiveness of extracorporeal shockwave on plantar flexor spasticity of ankle in stroke patients: a randomized controlled trial. J Neuroeng Rehabil. 2024;21:176. doi: 10.1186/s12984-024-01473-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Yang S.M., Lin H.H., Chen Y.H., Lu Y.L., Wu C.H., Chen W.S., et al. Dose-response effectiveness of focused shockwave therapy on ultrasonographic muscular properties in patients with stroke exhibiting ankle spasticity. J Neuroeng Rehabil. 2025;22:184. doi: 10.1186/s12984-025-01724-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Yao G., Chen J., Duan Y., Chen X. Efficacy of extracorporeal shock wave therapy for lateral epicondylitis: a systematic review and meta-analysis. Biomed Res Int. 2020;2020 doi: 10.1155/2020/2064781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zhang J., Zhong S., Tan T., Li J., Liu S., Cheng R., et al. Comparative efficacy and patient-specific moderating factors of nonsurgical treatment strategies for frozen shoulder: an updated systematic review and network meta-analysis. Am J Sports Med. 2021;49:1669–1679. doi: 10.1177/0363546520956293. [DOI] [PubMed] [Google Scholar]
- 34.Zhang R., Wang Z., Liu R., Zhang N., Guo J., Huang Y. Extracorporeal shockwave therapy as an adjunctive therapy for frozen shoulder: a systematic review and meta-analysis. Orthop J Sports Med. 2022;10 doi: 10.1177/23259671211062222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zhang Y.F., Liu Y., Chou S.W., Weng H. Dose-related effects of radial extracorporeal shock wave therapy for knee osteoarthritis: a randomized controlled trial. J Rehabil Med. 2021;53 doi: 10.2340/16501977-2782. [DOI] [PMC free article] [PubMed] [Google Scholar]