Frozen Shoulder (Adhesive Capsulitis)

Frozen shoulder, or adhesive capsulitis, is an idiopathic or post-traumatic inflammatory and fibrotic contracture of the glenohumeral joint capsule that produces progressive pain followed by global loss of active and passive range of motion. On imaging, the hallmark is thickening and fibrosis of the joint capsule and coracohumeral ligament (CHL) at the rotator interval and axillary recess, best appreciated on MR arthrography or high-resolution ultrasound, with obliteration of the subcoracoid fat triangle a classic ancillary sign. Imaging is not required for a straightforward clinical diagnosis but is frequently used to exclude mimics (rotator cuff tear, calcific tendinosis, osteoarthritis) and, increasingly, to guide capsular hydrodilatation.

Quick Reference

• What it is: Fibrotic thickening/contracture of the glenohumeral capsule, rotator interval, and coracohumeral ligament, with progressive painful stiffness.
• Clinical stages (Neviaser): I – painful/pre-adhesive (synovitis, preserved motion); II – freezing/painful (synovitis with early adhesions, motion loss begins); III – frozen/adhesive (dense fibrosis, stiff but less painful); IV – thawing (gradual return of motion).
• Key MRI/MRA thresholds: CHL thickness >4 mm and axillary recess capsule thickness >4 mm are the most cited cutoffs for diagnosis; a rotator interval capsule thickness ≥3 mm on coronal oblique T2 gives ~79–93% sensitivity and 86–100% specificity.
• Key ultrasound thresholds: CHL thickness >1.6–2.3 mm (sensitivity/specificity ~90%); axillary pouch thickness >4 mm (sensitivity up to 100%, specificity ~98% in large series); hypervascularity on power Doppler at the rotator interval is supportive but not required.
• Classic ancillary sign: Obliteration of the subcoracoid fat triangle by thickened rotator interval tissue.
• Main differential considerations: Rotator cuff tear/tendinopathy, glenohumeral osteoarthritis, calcific tendinosis, occult fracture, and — in refractory or atypical cases — an underlying neoplasm or infection.
• Risk factors: Diabetes mellitus (roughly 3–5x increased risk; up to ~13% prevalence in diabetics), thyroid disease, prior immobilization, and shoulder trauma/surgery.
• Why it matters for the report: Confirms the clinical diagnosis when equivocal, excludes a structural mimic requiring different management, and can guide image-guided capsular distension (hydrodilatation) or rotator interval injection.

Background

Adhesive capsulitis affects an estimated 2–5% of the general population, typically presenting between ages 40 and 60, more often in women, and with a strong association with diabetes mellitus and thyroid dysfunction (Le et al., 2017; Zreik et al., 2016). Histologically, the process involves chronic loss of the normal synovial lining, inflammatory infiltrate, and progressive subsynovial fibrosis of the capsule and rotator interval, with resultant capsular contracture (Le et al., 2017). The condition may be primary/idiopathic or secondary to trauma, surgery, or prolonged immobilization.

Imaging Anatomy

The structures a radiologist must interrogate are concentrated in two regions of the glenohumeral joint capsule:

• Rotator interval: The triangular space between the anterior supraspinatus and superior subscapularis tendons, bounded by the coracoid process medially. It contains the coracohumeral ligament, superior glenohumeral ligament, and the long head of the biceps tendon sheath, all enclosed within a capsule that becomes markedly thickened in adhesive capsulitis.
• Coracohumeral ligament (CHL): Extends from the base of the coracoid process to the greater and lesser tuberosities, coursing through the rotator interval. Normal CHL thickness is roughly 2–3 mm or less; thickening here is one of the most reproducible imaging findings in this disease.
• Axillary recess (inferior capsular pouch): The dependent fold of capsule inferior to the humeral head, which normally distends with the axillary pouch of the joint. Loss of normal redundancy and thickening of this recess correlates with restricted abduction and external rotation.
• Subcoracoid fat triangle: A normally visible wedge of fat deep to the coracoid process, between the coracoid, the conjoint tendon, and the joint capsule. This fat is effaced when the rotator interval capsule and CHL thicken and scar down against the coracoid — the “obliterated subcoracoid triangle” sign.

Imaging Findings

MRI and MR arthrography

MR arthrography remains the reference-standard imaging test and outperforms non-contrast MRI for detecting the subtle capsular and ligamentous changes of this disease (Mengiardi et al., 2004). Characteristic findings include:

• Thickening of the coracohumeral ligament, typically reported as >4 mm in the rotator interval on axial or sagittal oblique sequences (Li et al., 2011; Lee et al., 2012).
• Thickening of the joint capsule at the rotator interval, with a coronal oblique T2-weighted cutoff around 3 mm yielding sensitivity of 79–93% and specificity of 86–100% for adhesive capsulitis (Mengiardi et al., 2004).
• Obliteration of the subcoracoid fat triangle by scarred, thickened interval tissue — one of the most specific ancillary signs.
• Thickening of the capsule in the axillary recess, which correlates inversely with the degree of restriction in abduction and external rotation (Li et al., 2011; Lee et al., 2012).
• On MR arthrography specifically: poor capsular distension with contrast, reduced joint volume, and irregular synovial/fibrous thickening at the rotator interval and axillary pouch.
• T2 hyperintensity of the capsule and edema-like signal along the rotator interval are more conspicuous in earlier (painful/freezing) clinical stages, while dense low-signal fibrosis dominates in the frozen stage, although overlap between stages limits reliable MRI staging in practice (Tamai et al., 2024).

Ultrasound

High-resolution ultrasound is a fast, radiation-free, and increasingly validated first-line tool, particularly useful when a point-of-care correlation with symptoms and range of motion is wanted, and when guiding rotator interval or axillary recess injections. Key sonographic findings include:

• Thickening of the coracohumeral ligament, most series citing thresholds between 1.6 mm and 2.3 mm depending on technique, with sensitivity and specificity both in the 79–90% range (Do et al., 2021).
• Hypoechoic, thickened soft tissue filling the rotator interval, often with associated hypervascularity on power Doppler.
• Thickening of the axillary pouch capsule, with a >4 mm cutoff reported to have sensitivity approaching 100% and specificity around 98% in a large cross-sectional series (Stella et al., 2022).
• Effusion within the long head of the biceps tendon sheath and, on dynamic scanning, abnormal folding or impingement of the infraspinatus tendon against the posterior glenoid during passive external rotation.
• Loss of the normal glide of the subacromial-subdeltoid bursa and, occasionally, bursal hypervascularity.

Radiographs and CT

Plain radiographs are typically normal and are obtained mainly to exclude glenohumeral osteoarthritis, calcific tendinosis, or an occult fracture as an alternative cause of pain and stiffness. CT arthrography, when MRI is contraindicated, can demonstrate analogous findings of capsular and rotator interval thickening and reduced joint capacity, though it is used far less often than MR or ultrasound.

Differential Diagnosis and Pitfalls

• Rotator cuff tendinopathy or tear: Can produce overlapping pain and impingement-type restriction; look for cuff signal/discontinuity rather than isolated capsular/CHL thickening.
• Glenohumeral osteoarthritis: Radiographic joint space narrowing, osteophytes, and subchondral change are absent in isolated adhesive capsulitis.
• Calcific tendinosis: Intratendinous calcification on radiographs/CT/ultrasound distinguishes this from primary capsular disease, though the two can coexist.
• Post-surgical or post-traumatic capsular fibrosis: Can mimic idiopathic adhesive capsulitis on imaging; clinical history of prior surgery or immobilization is key.
• Normal CHL variability: CHL thickness measurements vary with technique and slice plane; use the axillary recess and rotator interval capsule findings as corroborating evidence rather than relying on CHL thickness alone.
• Refractory or atypical presentations: Should prompt consideration of an underlying neoplastic or infectious process, since these can rarely masquerade as capsulitis clinically.

Clinical Impact

Correlating capsular and CHL thickening with the degree of motion restriction helps confirm a clinical diagnosis when the presentation is atypical or when a mimic needs to be excluded before treatment. In the report, note the presence and degree of rotator interval and axillary recess capsular thickening, CHL thickening, and any obliteration of the subcoracoid fat triangle; also flag any competing findings (cuff tear, calcific deposit, arthritic change) since these change management. Imaging is also used to plan and guide ultrasound- or fluoroscopically-guided capsular hydrodilatation, a common second-line treatment for patients who fail conservative therapy.

Frequently Asked Questions

What is the most reliable MRI finding for diagnosing frozen shoulder?

Thickening of the coracohumeral ligament and the rotator interval capsule, typically using a threshold around 3–4 mm, is the most consistently reported and validated MRI finding, especially when combined with obliteration of the subcoracoid fat triangle.

Can ultrasound replace MRI for diagnosing adhesive capsulitis?

Ultrasound performs comparably well for detecting coracohumeral ligament and axillary pouch thickening and is faster and less costly, but MR arthrography remains the reference standard, particularly when a mimic such as a rotator cuff tear needs to be excluded in the same study.

Does frozen shoulder need imaging to be diagnosed?

No — the diagnosis is primarily clinical, based on progressive pain followed by global active and passive motion loss. Imaging is used to exclude alternative or coexisting pathology and, in some centers, to confirm equivocal cases or plan hydrodilatation.

Can MRI reliably stage frozen shoulder (painful vs. frozen vs. thawing)?

Not reliably. While capsular T2 hyperintensity and thickness trend differently across the Neviaser clinical stages, substantial overlap exists between stages, so MRI findings should support but not replace clinical staging.

What does a thickened coracohumeral ligament mean if the patient has full range of motion?

Isolated CHL thickening without corresponding clinical restriction should be interpreted cautiously, as measurement technique and normal variation can affect apparent thickness; correlate with axillary recess findings and the clinical exam before diagnosing adhesive capsulitis.

Why is diabetes relevant to a frozen shoulder report?

Diabetic patients have a substantially higher prevalence of adhesive capsulitis (reported around 13%, roughly 3–5 times the general population risk) and may have more severe or bilateral disease, which is useful context for the referring clinician.

References

1. Mengiardi B, Pfirrmann CW, Gerber C, Hodler J, Zanetti M. Frozen shoulder: MR arthrographic findings. Radiology. 2004;233(2):486-492. PMID: 15358849
2. Li JQ, Tang KL, Wang J, et al. MRI findings for frozen shoulder evaluation: is the thickness of the coracohumeral ligament a valuable diagnostic tool? PLoS One. 2011;6(12):e28704. PMID: 22163326
3. Lee SY, Park J, Song SW. Correlation of MR arthrographic findings and range of shoulder motions in patients with frozen shoulder. AJR Am J Roentgenol. 2012;198(1):173-179. PMID: 22194494
4. Le HV, Lee SJ, Nazarian A, Rodriguez EK. Adhesive capsulitis of the shoulder: review of pathophysiology and current clinical treatments. Shoulder Elbow. 2017;9(2):75-84. PMID: 28405218
5. Zreik NH, Malik RA, Charalambous CP. Adhesive capsulitis of the shoulder and diabetes: a meta-analysis of prevalence. Muscles Ligaments Tendons J. 2016;6(1):26-34. PMID: 27331029
6. Do JG, Hwang JT, Yoon KJ, Lee YT. Correlation of ultrasound findings with clinical stages and impairment in adhesive capsulitis of the shoulder. Orthop J Sports Med. 2021;9(5). PMC: PMC8113659
7. Stella SM, et al. Ultrasound features of adhesive capsulitis. Rheumatol Ther. 2022;9(1):151-162. PMID: 34940958
8. Tamai K, Hamada J, Nagase Y, Morishige M, Naito M, Asai H, Tanaka S. Can magnetic resonance imaging distinguish clinical stages of frozen shoulder? A state-of-the-art review. JSES Rev Rep Tech. 2024;4(3):365-370. PMID: 39157226