Narrowing of the Coracohumeral Interval
- Subcoracoid impingement is ANTERIOR shoulder pain arising from narrowing of the CORACOHUMERAL INTERVAL - the space between the tip of the coracoid process and the lesser tuberosity of the humerus - in which the SUBSCAPULARIS tendon is compressed (and abraded) between the coracoid and the humeral head, often described as a 'roller-wringer' effect on the tendon.
- The interval is quantified by the CORACOHUMERAL DISTANCE (CHD) on axial (and sagittal-oblique) MRI; a narrowed CHD and increased coracoid overlap define a tight subcoracoid space, and arthroscopic decompression (coracoplasty) increases the CHD and reduces coracoid overlap.
- Causes of a tight interval include a coracoid that is congenitally/developmentally too LONG or too LATERALLY placed, and IATROGENIC/post-surgical change - particularly after coracoid-transfer stabilisation (Latarjet/Bristow) or glenoid osteotomy that brings the coracoid closer to the humerus - as well as dynamic narrowing in provocative positions.
- The provocative position reproduces pain when the shoulder is brought into FORWARD FLEXION, ADDUCTION and INTERNAL ROTATION (the coracoid impingement test), which drives the lesser tuberosity towards the coracoid; pain is anterior, over the coracoid, and may be associated with subscapularis signs (a positive belly-press or lift-off if the tendon is involved).
- There is a genuine CONTROVERSY about causation: subcoracoid impingement is classically implicated in upper-border SUBSCAPULARIS tears, but comparative MRI studies have found NO significant difference in coracohumeral distance or coracoid overlap between shoulders with isolated subscapularis tears and the opposite healthy shoulders, with coracoacromial-ligament degeneration (a subacromial process) present in most - so a narrowed interval is not proven to cause the tear, and subscapularis pathology should not be attributed to it uncritically.
- THE CONTROVERSY LARGELY RESOLVES ON TWO DISTINCTIONS. First, TEAR MECHANISM: measured on axial MRI, DEGENERATIVE subscapularis tears had a mean CHD of 8.6 mm against 10.2 mm for TRAUMATIC tears and 10.4 mm for controls (p = 0.0003), with traumatic tears INDISTINGUISHABLE from controls - and a CHD below 6 mm occurred ONLY in degenerative tears. A study pooling both mechanisms will therefore find nothing. Second, WHAT IS MEASURED: an absolute CHD is weak because the normal range is wide (roughly 7-14 mm), whereas the SIDE-TO-SIDE DISCREPANCY measured bilaterally in 213 patients predicted subscapularis tears with an AUC of 0.882, a 0.5 mm cut-off giving 76.4% sensitivity and 99.3% specificity.
- CORACOID OVERLAP, though widely quoted, did NOT differ between sides in either the tear or the control group of that 213-patient bilateral study - so it does not discriminate and should not be leaned on, even though coracoplasty does reduce it. Use the patient's own contralateral shoulder as the reference rather than any published absolute value, and remember that an asymmetric interval identifies a mechanical association, not the source of the pain.
- MANAGEMENT is CONSERVATIVE first - activity/biomechanical modification, physiotherapy and corticosteroid injection - reserving ARTHROSCOPIC SUBCORACOID DECOMPRESSION (CORACOPLASTY, resecting the posterolateral coracoid to widen the interval) for recalcitrant, genuinely subcoracoid pain, often combined with addressing any subscapularis tear; coracoplasty reliably increases the CHD and improves pain and subscapularis strength in selected patients.
- “Subcoracoid impingement = anterior shoulder pain from a narrowed CORACOHUMERAL interval; subscapularis pinched between coracoid tip and lesser tuberosity ('roller-wringer').
- “Quantify with coracohumeral distance (CHD) on MRI; provocative position = forward flexion + adduction + internal rotation; causes incl. long/lateral coracoid and post-Latarjet.
- “Controversial as a cause of subscapularis tears. Conservative first; arthroscopic coracoplasty for recalcitrant cases.
- “The conflict resolves on mechanism: CHD narrowed in DEGENERATIVE tears (8.6 mm) but NORMAL in traumatic ones (10.2 vs 10.4 mm controls); under 6 mm occurred only in degenerative tears.
- “Absolute CHD is weak (normal ~7-14 mm) - the SIDE-TO-SIDE discrepancy discriminates (AUC 0.882; 0.5 mm = 76% sens, 99% spec). Coracoid overlap did NOT differ between sides.
Anatomy and Mechanism
The interval. Subcoracoid impingement is anterior shoulder pain from narrowing of the coracohumeral interval, the space between the coracoid tip and the lesser tuberosity of the humerus. The subscapularis tendon lies in that space and is compressed, and abraded, between the coracoid and the humeral head: the 'roller-wringer' effect.
Causes. A tight interval may be:
- Congenital or developmental: a coracoid that is too long or too laterally placed
- Post-traumatic or post-surgical: particularly after coracoid-transfer stabilisation (Latarjet/Bristow) or a glenoid osteotomy that brings the coracoid closer to the humerus
- Dynamic: narrowing in the provocative positions
Clinical Assessment
The provocative position. The pain is anterior, over the coracoid. The coracoid impingement test brings the shoulder into forward flexion, adduction and internal rotation, which drives the lesser tuberosity towards the coracoid and reproduces it. If the tendon is involved there may be subscapularis signs, a positive belly-press or lift-off.
The differential. Subcoracoid impingement is easily confused with subacromial impingement, and the broader skill is a structured differential of anterior shoulder pain and how each cause is told apart.
- Discriminating features
- Pain over the coracoid; provocative position forward flexion + ADDuction + internal rotation; asymmetric coracohumeral distance versus the other side
- The test that settles it
- A diagnostic subcoracoid injection that abolishes the pain - the most useful single manoeuvre
- Discriminating features
- Anterolateral pain, painful arc; a coracoacromial-arch process
- The test that settles it
- Positive Neer and Hawkins with relief after a subacromial injection - a different injection site from the above
- Discriminating features
- Pain over the bicipital groove; deep clicking with a SLAP lesion
- The test that settles it
- Speed's and O'Brien's tests; MR arthrography or arthroscopy for the labrum
- Discriminating features
- Increased passive external rotation, a positive bear-hug
- The test that settles it
- Weak belly-press and lift-off, confirmed on MRI - and remember subscapularis tearing and subcoracoid narrowing often coexist
- Discriminating features
- Pain localised to the AC joint
- The test that settles it
- Cross-body adduction (scarf) test, with an AC joint injection if still uncertain
- Discriminating features
- Acute, severe, often non-positional pain out of proportion to the examination
- The test that settles it
- A plain radiograph showing the deposit - do this before attributing severe acute pain to impingement
- Discriminating features
- Younger patient, history of dislocation or subluxation
- The test that settles it
- Apprehension and relocation tests; the history usually gives it away
Each of these is developed separately in subacromial impingement, biceps tendon pathology, SLAP tears, subscapularis tears, rotator cuff tears, AC joint injuries, calcific tendinitis of the shoulder and anterior shoulder instability.
Measuring the Interval: Coracohumeral Distance and Coracoid Overlap
Coracohumeral distance (CHD). The shortest distance between the coracoid tip and the lesser tuberosity or humeral head, measured on axial (and sagittal-oblique) MRI or CT. Published control shoulders average around 10.4 mm on MRI, but the normal range is wide, roughly 7 to 14 mm, so no single normal value applies to everyone. A CHD below 6 mm is the usual threshold for a narrowed space.
A static number for a dynamic problem. The interval can also narrow dynamically, so the provocative position and the symptoms matter as much as the static measurement.
Coracoid overlap. How far the coracoid tip projects laterally beyond the glenoid on axial imaging. A narrow CHD with increased overlap defines a tight space, and overlap is widely quoted, but in the bilateral study below it did not differ between sides in either group, so it did not discriminate and should not be leaned on.
Management
Conservative first. Activity and biomechanical modification, physiotherapy, and a corticosteroid injection into the subcoracoid space, which is also diagnostic.
Coracoplasty. Arthroscopic subcoracoid decompression resects the posterolateral coracoid to widen the interval. It is reserved for recalcitrant, genuinely subcoracoid pain and is often combined with repair of any subscapularis tear. In selected patients coracoplasty reliably increases the CHD, reduces coracoid overlap, and improves anterior pain and subscapularis strength.
Take the diagnosis seriously as a cause of genuine anterior shoulder pain while remaining critical about its role in subscapularis tears: attributing a tear solely to a 'tight' subcoracoid interval can lead to an unnecessary or insufficient operation. Confirm that the pain is truly subcoracoid (anterior, provocative position, relieved by a subcoracoid injection) before offering coracoplasty; it is the clinical picture, not the millimetres, that justifies the operation. Address the subscapularis on its own merits.
The Causation Controversy
The case against. Subcoracoid impingement is classically implicated in upper-border subscapularis tears. Yet comparative MRI studies have found no significant difference in CHD or coracoid overlap between shoulders with isolated subscapularis tears and the contralateral healthy shoulder, with coracoacromial-ligament degeneration (a subacromial process) present in most. On that evidence a tight interval is not proven to cause the tear, and subacromial/coracoacromial factors are implicated instead.
Reconciling the evidence. The apparent conflict is less contradictory than it first appears. Two distinctions do most of the reconciling: which tear, and what is measured.
Distinction One: Degenerative Versus Traumatic Tears
A study of 83 arthroscopically confirmed subscapularis tears against 20 controls with an intact subscapularis measured the coracohumeral distance on axial MRI.
- Mean CHD
- 8.6 +/- 2.0 mm
- Range
- 4.0 to 13.2 mm
- Mean CHD
- 10.2 +/- 2.0 mm
- Range
- 6.6 to 16.2 mm
- Mean CHD
- 10.4 +/- 1.8 mm
- Range
- 6.8 to 14.0 mm
What it showed. The degenerative group had a significantly smaller interval than either the traumatic group or the controls (p = 0.0003), while the traumatic tears were indistinguishable from controls (p = 0.79). The caveat that matters more than the threshold: a CHD below 6 mm occurred only in patients with degenerative tears.
Why it reconciles. The subcoracoid space is narrowed in the tears you would expect it to be narrowed in, the attritional ones, and normal in tears caused by an injury, which is exactly what a mechanical impingement hypothesis predicts. A study that pools both mechanisms together will find no difference.
Distinction Two: Absolute Value Versus Side-to-Side Discrepancy
A prospective study of 213 patients undergoing arthroscopic cuff repair measured the interval bilaterally on CT, comparing 72 with subscapularis tears against 141 without. In the tear group the affected shoulder had a significantly smaller CHD than the patient's own opposite shoulder (95% CI 6.1-7.2 mm versus 7.2-8.0 mm, p < 0.0001); in the control group there was no such side-to-side difference.
The discriminating measurement. Bilateral CHD discrepancy was the best predictor of a subscapularis tear, with an area under the curve of 0.882; a cut-off of 0.5 mm gave a sensitivity of 76.4% and a specificity of 99.3%. A single absolute CHD is weak because the normal value varies between people, and the patient's own contralateral shoulder is the better reference. That is why comparisons of absolute values across different patients have been inconclusive while a within-patient comparison performs well.
What an asymmetric interval means. Even a genuinely asymmetric interval identifies a mechanical association, not the source of the pain. Imaging is supportive rather than diagnostic.
A "narrow-looking" coracohumeral distance in isolation means little. Measure the contralateral shoulder and compare. A finding of asymmetric narrowing in a patient with anterior pain warrants specific intra-operative inspection of the subscapularis.
Mnemonics & Memory Aids
CORACOID
Hook:CORACOID: narrowed interval + Overlap, Roller-wringer on subscap, Anterior pain, measure CHD, On provocative position, Iatrogenic/long coracoid, Decompress if needed.
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“What is subcoracoid impingement, and how would you assess a patient you suspect has it?”
“How would you manage subcoracoid impingement, and what does the evidence say about coracoplasty?”
Definition & mechanism
- Anterior shoulder pain from a narrowed coracohumeral interval (coracoid tip to lesser tuberosity)
- Subscapularis compressed between coracoid and humeral head ('roller-wringer')
- Quantified by coracohumeral distance (CHD) on MRI
Causes
- Long or laterally placed coracoid (congenital/developmental)
- Iatrogenic/post-surgical (coracoid transfer - Latarjet/Bristow; glenoid osteotomy)
- Dynamic narrowing in provocative positions
Assessment
- Provocative position: forward flexion + adduction + internal rotation
- Coracoid tenderness; subscapularis tests (belly-press, lift-off)
- MRI CHD/coracoid overlap; diagnostic subcoracoid injection
Controversy & management
- Implicated in subscapularis tears but not proven causal - imaging is supportive, not diagnostic
- Conservative first (modification, physio, injection)
- Arthroscopic coracoplasty for recalcitrant cases (+/- subscapularis repair)
The numbers that matter
- Controls average ~10.4 mm on MRI; normal range is wide, roughly 7-14 mm
- Degenerative subscap tears 8.6 mm vs traumatic 10.2 mm (traumatic = same as controls)
- CHD under 6 mm occurred ONLY in degenerative tears
- Side-to-side CHD discrepancy is the discriminator: AUC 0.882; 0.5 mm = 76% sens, 99% spec
- Coracoid overlap did NOT differ between sides in 213 patients - do not lean on it
- Asymmetric narrowing warrants specific intra-operative subscapularis inspection
Evidence & Key Studies
Effect of arthroscopic coracoplasty on subscapularis strength in subcoracoid impingement
- Subcoracoid impingement causes anterior shoulder pain; arthroscopic subcoracoid decompression (coracoplasty) is the preferred treatment in recalcitrant cases.
- Coracoplasty increased the coracohumeral distance and reduced coracoid overlap, with improvement in anterior pain, internal rotation, ASES score and subscapularis strength.
- The improvement in subscapularis strength was inversely related to the postoperative coracohumeral distance, indicating a mechanical effect of the coracoid on the subscapularis.
Is it subcoracoid or subacromial impingement that tears the subscapularis? An MRI comparison
- In patients with isolated subscapularis tears, coracohumeral distance and coracoid overlap did not differ significantly between the operated and the contralateral healthy shoulder.
- Coracoacromial-ligament degeneration (a subacromial process) was present in 75% of patients.
- These findings question whether subcoracoid narrowing is the true cause of isolated subscapularis tears, implicating subacromial factors instead.
The coracohumeral distance in shoulders with traumatic and degenerative subscapularis tendon tears
- Axial MRI coracohumeral distance in 83 arthroscopically confirmed subscapularis tears (44 degenerative, 39 traumatic) against 20 controls with an intact subscapularis.
- Degenerative tears had a significantly smaller CHD (8.6 +/- 2.0 mm) than traumatic tears (10.2 +/- 2.0 mm) or controls (10.4 +/- 1.8 mm), p = 0.0003; traumatic tears did not differ from controls (p = 0.79).
- A CHD of less than 6 mm occurred only in patients with degenerative subscapularis tears - so the narrowed interval tracks the attritional mechanism specifically.
Bilateral coracohumeral distance discrepancy is associated with subscapularis tear in rotator cuff rupture patients
- Bilateral CT measurement in 213 consecutive patients undergoing arthroscopic cuff repair, 72 with and 141 without a subscapularis tear.
- In the tear group the affected shoulder had a smaller CHD than the patient's own contralateral shoulder (95% CI 6.1-7.2 versus 7.2-8.0 mm, p < 0.0001); no such side-to-side difference existed in the control group, and coracoid overlap did not differ between sides in either group.
- Bilateral CHD discrepancy was the best predictor of a subscapularis tear (AUC 0.882); a 0.5 mm cut-off gave 76.4% sensitivity and 99.3% specificity, and its presence warrants specific intra-operative inspection of the subscapularis.
The description of subcoracoid impingement as a cause of anterior shoulder pain and the effect of coracoplasty - widening the interval, reducing overlap and improving subscapularis strength, with that improvement inversely related to the postoperative CHD - come from the cited Acan study. The finding that CHD and coracoid overlap do not differ from the healthy side in isolated subscapularis tears, with coracoacromial ligament degeneration in 75%, is Cetinkaya.
The reconciliation of that apparent conflict draws on two further sources. The separation by tear mechanism - degenerative 8.6 mm against traumatic 10.2 mm and controls 10.4 mm, with a CHD under 6 mm occurring only in degenerative tears - is Balke, a cohort of 83 tears and 20 controls. The bilateral measurements, the finding that coracoid overlap does not differ between sides, and the discriminating performance of side-to-side CHD discrepancy (AUC 0.882, 0.5 mm cut-off, sensitivity 76.4%, specificity 99.3%) are Zhu, in 213 patients measured on CT rather than MRI, which is worth noting because absolute values are not interchangeable between the two modalities. Reading those two alongside Cetinkaya is what suggests that pooling tear mechanisms, and comparing absolute values between different people rather than between a patient's own shoulders, explains the inconclusive results - that synthesis is an inference from the studies read together rather than a conclusion any one of them states.
No verified figure was identified for how much coracoid may safely be resected at coracoplasty before the conjoint tendon origin is compromised, so no amount is given. The coracohumeral-interval anatomy, the roller-wringer mechanism, the provocative position and the conservative-then-coracoplasty pathway are standard, well-established teaching. (See also Subscapularis Tears, Subacromial Impingement and External Impingement.)