Adhesive capsulitis, commonly referred to as frozen shoulder, represents a profoundly debilitating musculoskeletal condition characterized by progressive glenohumeral pain and a global restriction of both active and passive shoulder kinematics ^[1]. It affects approximately 2% to 5% of the general population, with a noted predilection for women and individuals in their fourth to sixth decades of life ^[2]. The socioeconomic and personal burden is substantial; patients frequently endure severe nocturnal pain, sleep disturbances, and profound occupational and functional impairment that can persist for one to three years ^{[1, 3]}. The underlying pathophysiology is driven by a chronic inflammatory synovial cascade that triggers a robust fibroblast-to-myofibroblast transition, depositing a dense, non-compliant extracellular matrix within the glenohumeral joint capsule ^[4]. Furthermore, the condition is notoriously prevalent and recalcitrant among patients with endocrine comorbidities, particularly diabetes mellitus, which significantly increases the risk of development and symptom severity ^[3].

The management of frozen shoulder encompasses a broad spectrum of therapeutic modalities. Initial treatment typically involves conservative strategies, including non-steroidal anti-inflammatory drugs, structured physiotherapy, and targeted home exercise programs ^[5]. When the condition proves refractory to conservative measures, clinicians frequently escalate to interventional procedures such as intra-articular corticosteroid injections, ultrasound-guided hydrodistension, Manipulation Under Anaesthesia (MUA), or arthroscopic capsular release ^[5].

Among these interventions, Manipulation Under Anaesthesia (MUA) remains a cornerstone technique for refractory adhesive capsulitis ^[5]. The fundamental rationale for MUA is to mechanically rupture the contracted, fibrotic glenohumeral capsule and lyse intra-articular adhesions while the patient's protective reflex muscle guarding is abolished by anaesthesia ^[4]. This procedure offers a cost-effective and relatively rapid restoration of joint mobility ^[5]. Nevertheless, MUA is a blind, forceful intervention associated with a spectrum of recognized iatrogenic risks, including proximal humeral fractures, glenohumeral dislocations, rotator cuff tears, and brachial plexus traction injuries ^[6]. Furthermore, the acute tissue trauma generated by the forced capsular tearing triggers a severe post-procedural inflammatory cascade, characterized by reactive hyperemia, which can lead to reactive scarring and a "refreezing" of the joint if not properly managed ^[4].

To mitigate this reactive inflammatory spike, the adjunctive administration of an intra-articular corticosteroid injection during MUA has been widely advocated ^[4]. Direct intra-articular deposition of a corticosteroid suppresses pro-inflammatory cytokines and inhibits capsular fibroblastic proliferation at the site of the newly torn tissue ^[4]. Clinically, this pharmacological adjunct aims to provide rapid post-procedural pain relief, thereby affording patients a critical window of comfort to participate in the immediate and aggressive physical therapy required to preserve the newly acquired range of motion ^[7].

Despite the theoretical advantages of this combined approach, the current literature presents conflicting evidence regarding its clinical superiority over MUA alone. Certain meta-analyses have demonstrated that adding a corticosteroid injection significantly enhances early pain relief and specifically accelerates the recovery of external rotation, while improvements in forward flexion or internal rotation remain statistically insignificant ^[8]. Conversely, other randomized clinical trials have concluded that the addition of intra-articular steroids yields no statistically significant long-term improvement in overall shoulder function, disability scores, or global range of motion when compared to isolated manipulation ^[9].

A critical appraisal of the existing body of evidence reveals several methodological weaknesses and inconsistencies. Much of the available literature on MUA is compromised by small sample sizes, a lack of robust control groups, retrospective study designs, and variable follow-up periods ^[10]. Moreover, numerous randomized controlled trials evaluating MUA have been assessed as having a high risk of bias, rendering the generalizability of their findings questionable ^[10]. The variability in injection protocols, specifically concerning the dosage and precise timing of the steroid administration, further confounds the ability to draw definitive conclusions from the current data ^[11].

Clarifying this clinical ambiguity is paramount for optimizing patient care and refining treatment algorithms. Immediate post-procedural rehabilitation is the most critical determinant of long-term success following MUA; therefore, identifying whether corticosteroid adjuncts reliably control pain enough to facilitate this rehabilitation is essential ^[7]. Furthermore, considering the known systemic and local side effects of corticosteroids particularly the risk of severe glycemic spikes in diabetic patients, who represent a large subset of the frozen shoulder demographic the decision to inject should be driven by unequivocal evidence of benefit ^[4].

Given the persistent evidentiary gaps, methodological limitations in current literature, and the clinical necessity of maximizing rehabilitation outcomes while minimizing pharmacological risks, further robust investigation is required. Therefore, the present study is designed to critically evaluate and comprehensively compare the functional and symptomatic outcomes of frozen shoulder managed with manipulation under anaesthesia, with and without the addition of an intra-articular corticosteroid injection.

Study Design and Patient Selection: A retrospective comparative cohort analysis was conducted utilizing clinical records from a two orthopaedic referral centres between January 2020 and December 2024 at Bangalore Medical college & research hospital and Nandi Medical College and research institute Chikkaballapura. The study was designed to evaluate the clinical efficacy of isolated manipulation under anaesthesia compared to manipulation augmented with an intra-articular corticosteroid injection. Patients aged 40 to 70 years diagnosed with primary idiopathic adhesive capsulitis were screened for inclusion. Diagnostic criteria required a minimum symptom duration of six months refractory to structured conservative management, which included oral non-steroidal anti-inflammatory drugs and physical therapy. Clinical restriction was defined as a global limitation of both active and passive glenohumeral kinematics, specifically demonstrating forward flexion of less than 100 degrees and a reduction in external rotation of at least 50% compared to the contralateral unaffected limb. Individuals with secondary shoulder stiffness stemming from prior trauma, previous shoulder surgery, established glenohumeral osteoarthritis, or concomitant full-thickness rotator cuff tears verified by magnetic resonance imaging were excluded. Figure 1: Study Flow Diagram Surgical Intervention and Rehabilitation Protocol: All procedures were performed under short-duration general anaesthesia with patients positioned supine. To minimize iatrogenic risks such as proximal humeral fractures or brachial plexus traction injuries, the manipulation was executed utilizing a short lever arm technique while firmly stabilizing the scapula. The controlled mechanical rupture of the fibrotic capsule was achieved through a sequential vector application: forward elevation in the scapular plane, external rotation in both neutral and 90 degrees of abduction, internal rotation, and finally cross-body adduction. Following the capsular release, patients were stratified into two cohorts based on the intraoperative adjunctive treatment documented in their surgical records. The isolated manipulation cohort (Group A) received no further pharmacological intervention. The combined therapy cohort (Group B) received an intra-articular injection via a posterior landmark-guided approach, consisting of 40 mg of triamcinolone acetonide mixed with 5 mL of 1% lidocaine hydrochloride. Postoperatively, both groups were prescribed an identical, aggressive rehabilitation protocol. Passive range of motion (ROM) exercises were initiated in the recovery unit on the day of the procedure to prevent the re-adhesion of the acutely torn capsular fibers. Upon discharge, patients engaged in supervised outpatient physiotherapy focused on active-assisted mobilization and progressive strengthening, supplemented by a mandatory home exercise regimen executed four to five times daily. Outcome Measures and Statistical Analysis: Clinical data were extracted at baseline and at specific postoperative intervals: two weeks, three months, six months, and twelve months. The primary outcome measure was the Oxford Shoulder Score (OSS) to quantify functional disability. Secondary measures included the Visual Analog Scale (VAS) for pain assessment and precise goniometric recordings of forward flexion, abduction, external rotation, and internal rotation. Statistical evaluation was performed using SPSS software. Continuous variables satisfying normality assumptions were compared using independent samples t-tests, while non-parametric data were analyzed via the Mann-Whitney U test. Categorical demographic variables and complication rates were assessed using Chi-square tests. A P-value of less than 0.05 was established as the threshold for statistical significance.

Demographic and Baseline Characteristics: The final analysis included 214 patients who met all inclusion criteria and completed the mandatory 12-month follow-up. Group A (isolated manipulation) consisted of 102 patients, and Group B (manipulation with corticosteroid) comprised 112 patients. The mean age of the collective cohort was 54.6 ± 6.2 years, with a distinct female predominance (65.8%). Type II diabetes mellitus was documented in 18.2% of the total study population. Baseline assessments demonstrated severe symptomatic impairment across both cohorts, with a mean preoperative VAS score of 7.6 ± 1.1 and an average OSS of 14.8 ± 3.2 out of 48. Statistical comparison of baseline demographics, symptom duration, and initial kinematic restrictions revealed no significant differences between the two groups (p>0.05).

Figure 2: Gender Distribution of the Study Population

Figure 3: Distribution of Type II Diabetes Mellitus

Pain and Functional Outcomes: Both procedural pathways yielded profound and enduring improvements in glenohumeral function and pain reduction. In the acute postoperative phase, Group B exhibited a statistically superior trajectory regarding symptomatic relief. At the two-week interval, the mean VAS score in Group B was 2.9 ± 1.0, compared to 4.5 ± 1.2 in Group A (p=0.003). Similarly, early functional capacity measured by the OSS at two weeks favored the corticosteroid cohort (27.4 vs. 21.8, p=0.015).

Figure 4: Trend of Visual Analog Scale (VAS) Scores During Follow-up

As time progressed, the clinical outcomes between the two treatment arms converged. At the three-month evaluation, the difference in VAS scores narrowed considerably (Group A: 2.3 ± 0.8; Group B: 1.9 ± 0.7, p=0.08). By the final 12-month follow-up, parity was achieved in both primary domains. The terminal VAS scores were 1.1 ± 0.6 for isolated manipulation and 0.9 ± 0.5 for the combined approach (p=0.42), while the OSS reached 44.2 ± 3.1 and 45.0 ± 2.8, respectively (p=0.38).

Figure 5: Trend of Oxford Shoulder Score During Follow-up

Kinematic Restoration: Goniometric tracking of shoulder mobility demonstrated robust restoration of forward flexion, abduction, and internal rotation across the entire study population, with no statistically significant variance between the groups at the final follow-up (p>0.05 for all three parameters).

Figure 6: Comparison of Final Shoulder Range of Motion

A highly significant divergence was uniquely isolated to the recovery of external rotation. At three months, Group B achieved a mean external rotation of 69.4° ± 11.2°, which was superior to the 56.1° ± 14.5° observed in Group A (p<0.001). This kinetic advantage in external rotation persisted through the 12-month milestone, with Group B maintaining an average of 76.2° ± 9.8° compared to 68.5° ± 12.3° in the isolated manipulation cohort (p=0.021).

Figure 7: Recovery of External Rotation During Follow-up

Complications and Re-interventions: The overall safety profile was excellent, with no occurrences of iatrogenic humeral fractures, glenohumeral dislocations, or severe systemic adverse events. Minor complications included transient neuropraxia in one patient from Group A and localized post-injection erythema in two patients from Group B, all of which resolved spontaneously within 72 hours. During the study period, clinical failure necessitating a repeat manipulation occurred in five patients (4.9%) from Group A and two patients (1.7%) from Group B, a difference that did not reach statistical significance (p=0.18).

Figure 8: Comparison of Complications and Re-interventions

The principal finding of this retrospective comparative cohort study is that Manipulation Under Anaesthesia (MUA) effectively restores glenohumeral kinematics and alleviates pain in patients with refractory primary adhesive capsulitis, regardless of whether an intra-articular corticosteroid is administered. However, the adjunctive use of a corticosteroid injection yields a distinctly biphasic clinical advantage: it provides statistically superior acute pain relief during the early post-procedural phase and confers a sustained, isolated kinetic superiority in the restoration of external rotation. Long-term overall functional outcomes and pain scores, conversely, demonstrate eventual convergence between the two therapeutic pathways. The acute symptomatic benefit observed in the combined treatment cohort aligns closely with current orthopaedic literature. Previous retrospective analyses have similarly reported that the addition of intra-articular steroids to MUA significantly accelerates the reduction of early Visual Analog Scale (VAS) and disability scores during the initial two to four weeks post-procedure [12, 8]. Similar short-term analgesic superiority has been documented in randomized comparative trials evaluating manipulation augmented with triamcinolone acetonide [1]. The biological rationale for this early divergence is rooted in the suppression of iatrogenic inflammation. MUA achieves mechanical release through the forceful rupture of the contracted, fibrotic joint capsule, which inadvertently triggers a severe, acute inflammatory cascade and reactive hyperemia at the site of tissue failure [4]. Direct intra-articular deposition of a corticosteroid blunts this cytokine-driven pro-inflammatory response [4, 10]. Clinically, this potent acute analgesia is paramount; it affords patients a critical window of comfort necessary to tolerate the immediate, aggressive physical therapy required to preserve the newly acquired joint volume and prevent the formation of bridging scar tissue [4, 10]. Despite this early divergence, the present study found that mid-to-long-term outcomes specifically at 6 and 12 months demonstrated no statistically significant differences in global functional capacity (Oxford Shoulder Score) or terminal pain levels. This convergence suggests that while steroids accelerate early recovery, the ultimate functional ceiling is dictated by the mechanical disruption of the capsule rather than the pharmacological adjunct. This trajectory is supported by numerous randomized controlled trials and systemic reviews, which frequently conclude that MUA augmented with steroids offers no definitive long-term superiority over MUA alone regarding overall disability indices [10, 13]. Furthermore, comparative trials evaluating arthroscopic capsular release against MUA have noted that long-term functional stabilization is highly consistent once the fibrotic contracture is overcome and the initial pathological inflammatory phase naturally burns out [14]. A highly compelling finding of the present analysis is the selective, long-term kinetic advantage in external rotation associated with the corticosteroid adjunct. While forward flexion, abduction, and internal rotation equalized over time, the steroid cohort maintained a significantly wider arc of external rotation at both three and twelve months. This specific kinematic phenomenon has been rigorously validated in a recent comprehensive meta-analysis, which identified external rotation as the sole range-of-motion parameter significantly enhanced by the peri-procedural addition of corticosteroids to MUA [8]. The microanatomy of the anterior capsular structures provides a plausible explanation for this localized outcome. The coracohumeral ligament and the rotator interval act as the primary mechanical restrictors of external rotation with the humerus adducted [4]. These highly vascularized anterior structures are the epicenter of intense synovitis in adhesive capsulitis. Following the mechanical tearing induced by MUA, the unmitigated anterior capsule is highly prone to rapid fibroblastic proliferation and reactive scarring, which can cause the joint to preferentially "refreeze" in internal rotation [4, 9]. By inhibiting capsular myofibroblast activity during the critical early remodeling phase, the corticosteroid prevents robust re-adhesion across the rotator interval, thereby selectively preserving the external rotation gained intraoperatively [8, 4]. Regarding procedural safety, the complication rates in this study were notably low and statistically comparable between groups, with clinical failure necessitating repeat manipulation occurring in a minority of the overall cohort. This reflects the findings of extensive longitudinal registries demonstrating that MUA, when performed with controlled short-lever techniques, carries an exceptionally low long-term recurrence rate and a minimal risk of iatrogenic fractures or instability [15]. However, while the pharmacological adjunct appears to reduce the need for repeat interventions by facilitating early rehabilitation compliance, clinicians must weigh this benefit against the systemic risks of high-dose corticosteroids, particularly regarding glycemic control in the diabetic patient demographic [9, 11]. In clinical practice, these findings support a stratified approach to the surgical management of refractory adhesive capsulitis. While isolated MUA remains a highly effective and economically viable intervention [2], the peri-procedural administration of an intra-articular corticosteroid is strongly justified when acute pain control is a primary concern or when severe preoperative deficits in external rotation threaten the patient's capacity for overhead and functional reach. Strengths: The primary strength of this study lies in its robust sample size of 214 patients, which provides adequate statistical power to detect subtle kinematic differences between the treatment arms. Furthermore, the implementation of a standardized short-lever manipulation technique and an identical, aggressive post-operative rehabilitation protocol across both cohorts minimized performance bias. The structured 12-month follow-up schedule allowed for a nuanced evaluation of both acute temporal trajectories and long-term functional stabilization, clarifying discrepancies often found in studies with shorter observation periods. Limitations: Several limitations must be acknowledged. The retrospective study design introduces inherent selection bias and relies on the accuracy of historical medical documentation. Importantly, while a standardized post-operative physical therapy protocol was prescribed, outpatient compliance with the mandatory home exercise regimen could not be objectively quantified, potentially introducing a confounding variable in the functional recovery rates. Additionally, the study did not stratify outcomes based on the precise duration of pre-operative symptoms or stratify the diabetic subpopulation, both of which are known prognostic variables that can influence capsular compliance and responsiveness to corticosteroids. Finally, the absence of advanced pre-operative imaging, such as magnetic resonance imaging, prevented the correlation of clinical outcomes with the anatomical severity of capsuloligamentous thickening. Future Research Directions: Future investigations should focus on large-scale, prospective randomized controlled trials to eliminate the biases inherent in retrospective data. Specifically, research is needed to determine the optimal dosing of the corticosteroid adjunct, comparing low-dose versus high-dose regimens to maximize capsular remodeling while mitigating systemic endocrine side effects. Additionally, further comparative studies utilizing objective imaging parameters are necessary to evaluate whether specific subpopulations, particularly those with uncontrolled diabetes mellitus or prolonged symptom duration, might benefit more from alternative interventions such as ultrasound-guided hydrodistension or precise arthroscopic capsular release.

In conclusion, both isolated manipulation under anaesthesia and its augmentation with an intra-articular corticosteroid provide excellent, comparable long-term restoration of shoulder function in refractory adhesive capsulitis. However, the pharmacological adjunct offers distinct advantages. Corticosteroids significantly suppress acute iatrogenic inflammation, delivering superior early analgesia that facilitates immediate, aggressive postoperative rehabilitation. Furthermore, this adjunct uniquely preserves long-term gains in external rotation by preventing anterior capsular re-adhesion. Therefore, while isolated manipulation remains a highly effective and safe intervention, the peri-procedural addition of a corticosteroid is strongly recommended to optimize early pain control and maximize specific kinematic recovery, provided no endocrine contraindications exist.

Author Contributions

The first authors conducted the primary study. The Second, third & corresponding author assisted with data collection, interpretation, and manuscript preparation

1. Bora S, Shah V, Sonowal PJ, Islam ZP, Dutta D. A Prospective Randomized Comparison Study in Patients with Frozen Shoulder Treated by Manipulation under Anesthesia with or Without Intra-Articular Steroid Injection. Int J Pharm Clin Res. 2024.
2. Saeed HS, Manzoor B, Qureshi S, Sarwar MU, Rasheed N, Manzoor I. Comparison of efficacy of intra articular steroid injection versus manipulation under anesthesia (MUA) in idiopathic frozen Shoulder. Professional Med J. 2022.
3. Burnham J. Frozen Shoulder (Adhesive Capsulitis): 4 Stages, Symptoms, and Treatment.
4. Clinical and Functional Outcomes of Manipulation Under Anesthesia with or Without Corticosteroid Adjuncts for Adhesive Capsulitis.
5. Mohanty A, Mohanty A. Comparative Efficacy of Manipulation Under Anaesthesia Versus Arthroscopic Capsular Release in Primary Frozen Shoulder: A Prospective Randomized Study. J Orthop Case Rep. 2025.
6. Manipulation Under Anaesthesia (MUA): Shoulder Fix or False Cure? World Frozen Shoulder Clinic. 2021.
7. Joshi SU, Oberai TS, Panat RM, Ingale NA, Gadekar GN, Mapari YM. Comparative Outcomes of Manipulation under General Anesthesia and Ultrasound-Guided Hydrodistension in Adhesive Capsulitis. J Orthop Case Rep. 2025.
8. Boutros M, Awad G, Smadi Z, Tarabay PMDB, Rached RB, Elhassan B. Manipulation under anesthesia with vs without intra-articular corticosteroid injection for frozen shoulder: A systematic review and meta-analysis. Orthop Traumatol Surg Res. 2025.
9. Manipulation Under Anesthesia for Frozen Shoulders. Musculoskeletal Key. 2016.
10. Salomon M, Pastore C, Maselli F, Di Bari M, Pellegrino R, Brindisino F. Manipulation under Anesthesia versus Non-Surgical Treatment for Patients with Frozen Shoulder Contracture Syndrome: A Systematic Review. Int J Environ Res Public Health. 2022.
11. Kim SJ, Park JM, Song J, Yoon SY, Shin JI, Lee SC. High- versus Low-dose Steroid Injection for Adhesive Capsulitis (Frozen Shoulder): A Systematic Review and Meta-analysis. Pain Physician. 2023.
12. Song C, Song C, Li C. Outcome of manipulation under anesthesia with or without intra-articular steroid injection for treating frozen shoulder: A retrospective cohort study. Medicine (Baltimore). 2021;100(13):e23893.
13. Jacobs LG, Smith MG, Khan SA, Smith K, Joshi M. Manipulation or intra-articular steroids in the management of adhesive capsulitis of the shoulder? A prospective randomized trial. J Shoulder Elbow Surg. 2009.
14. Rangan A, Brealey SD, Keding A, et al. Management of adults with primary frozen shoulder in secondary care (UK FROST): A multicentre, pragmatic, three-arm, superiority randomised clinical trial. Lancet. 2020.
15. Woods DA, Loganathan K. Recurrence of frozen shoulder after manipulation under anaesthetic (MUA): the results of repeating the MUA. Bone Joint J. 2017;99-B(6):812-817