What's Actually Going On
Two nerves, two different problems, one presenting complaint.
The suprascapular nerve runs to the two muscles on the back of the shoulder blade and passes through two narrow notches on the way. Squeezed at the upper notch, both muscles go weak. Squeezed at the lower notch, only the lower one does. It supplies no skin anywhere, which is why this condition produces no numbness and gets read as a tendon problem.
The commonest structural cause is a fluid-filled cyst at the lower notch, found in 42% of surgically managed cases. It is not really a nerve problem at source: a tear in the joint's cartilage rim acts as a one-way valve, joint fluid escapes into the notch, and the cyst grows and presses on the nerve. Close the valve and the cyst loses its supply, which is why repairing the rim turns out to matter more than chasing the cyst.
The competing explanation is stretch rather than squeeze. In overhead athletes the majority view is repeated traction through extreme reaching positions, and one cadaveric study through the pitching motion found five possible points of injury rather than one. These two explanations predict opposite treatments, which is the heart of the debate below.
The axillary nerve runs to the deltoid and teres minor through a gap called the quadrilateral space, alongside an artery. Fibrous bands are the usual explanation, with genuine lumps less common. This one does produce numbness, over the outside of the shoulder, in a patch that does not follow the usual nerve-root map. The artery in the same gap is a separate problem with a much harder consequence, because it can clot and throw fragments down into the hand.
How to Identify It
The single question that does most of the work is which muscles are weak.
| Pattern | Where the problem is |
|---|
| Both back-of-blade muscles weak, no numbness | Suprascapular nerve at the upper notch |
| Only the lower one weak, no numbness | Suprascapular nerve at the lower notch. Image the cartilage rim |
| Deltoid weak, teres minor wasted, numbness over the outer shoulder, tenderness in the gap | Axillary nerve in the quadrilateral space |
| Wasting on a scan, no matching symptoms | Probably an incidental finding. Treat the patient, not the report |
How good are the tests? Nobody has published an answer, for any of them.
- Side-to-side strength testing, muscle by muscle No published accuracy data
- Tenderness over the quadrilateral space with the arm raised and turned out No published accuracy data
- Ultrasound-guided anaesthetic block into the quadrilateral space No published accuracy data
Across 107 papers and five separate database searches, not one clinical test for either condition carries a published figure for how well it catches the condition or rules it out. The tests above are listed because they tell you where the problem is, not because their accuracy is known. The block is the closest thing to a reference standard the condition has, and it has never been formally measured.
Nerve conduction testing is the practical standard, with one caveat worth carrying: among patients already selected for surgery, only 85% had test results consistent with the diagnosis.
The Debate
Does decompression actually work?
Momaya 2018 · Memon 2018 · 534 patients pooled
92% return to sport. Symptom resolution in 97%. Complication rates under 4%. Consistently excellent across 61 studies.
VS
Sachinis 2021 · Gerber 2020 · the only randomised comparisons
Adding nerve release changed nothing. 89.5% of controls recovered against 83.3% of released patients. The second trial tested decompressing a normal nerve as an explicit null.
Follow the randomised trials. Every study in both systematic reviews is uncontrolled, and patients were operated after a mean of 19 months of failed conservative care, so the operated group is the residue of everyone who did not recover on their own. Do not quote 92% return-to-sport to a patient as a treatment effect.
Does the arteriogram confirm the diagnosis?
Cahill 1983, the founding series
A positive arteriogram with the arm raised confirms the syndrome and selects who gets surgery.
VS
Mochizuki 1994
The same occlusion appeared in 80% of asymptomatic volunteers. The authors concluded the test has no diagnostic value.
Eleven years and a control group. The 1983 series had no asymptomatic comparison, and the symptoms come from the nerve rather than the artery, so an arterial finding was never the right test for a nerve complaint. It stays appropriate when the question is a clot in someone with cold fingers.
Squeeze or stretch?
The entrapment model
The nerve is pinched at a specific notch, so release the ligament over it.
VS
Drez 1976 · Ringel 1990 · Challoumas 2017
It is a traction injury, not an entrapment, with five possible injury points across the throwing motion and repeated stretch as the majority mechanism.
Let the imaging decide. A structural cause settles it toward squeeze. Its absence does not settle it toward anything, and defaults you to managing load and position.
No clinical guideline exists for either condition. No NICE, APTA, BOA, EULAR, ACR or JOSPT guidance appeared anywhere in 107 retrieved papers as of August 2026. Every recommendation on this page rests on primary studies and reviews, which is why the debates above are study against study rather than guideline against trial.
Honest Limitations
The surgical literature selects its own success
What the research shows: 92% return to sport and near-ideal function scores across 275 decompressed patients.
The real-world gap: those patients were operated after a mean of 19.0 months of symptoms. Everyone who recovered in month three is missing from the denominator. Pooling the survivors of a long conservative trial and reporting their outcome describes the operated remainder, not the condition.
Clinical adjustment: quote those numbers as outcomes after surgery in people who had already failed prolonged conservative care, because that is exactly what they are.
The conservative arm has a duration and no content
What the research shows: at least six months of non-surgical management, including medication, activity modification and physical therapy, before considering surgery.
The real-world gap: no exercise, dose, set, rep, load, frequency or progression rule appears anywhere in the retrieved literature for either condition. The whole physical therapy evidence base is one case report. Every clinician following that recommendation is inventing a protocol.
Clinical adjustment: say so out loud, then run the six months against an objective checkpoint rather than a symptom report, because the alternative is the observed 19-month drift.
Cadaveric strain is not diagnostic accuracy
What the research shows: five nerve-tensioning test positions raised strain in the suprascapular nerve by a mean of 22.70%, and one loaded the far portion significantly more than the rest.
The real-world gap: that establishes the tests tension the nerve in a cadaver. It establishes nothing about whether a positive result in a living patient means anything, and it is precisely the kind of finding that gets quoted as validation of a test.
Clinical adjustment: use these tests to reproduce symptoms and steer treatment. Do not record a positive as confirmation, and do not tell the patient it confirms anything.
The Nuance
The most important thing on this page is not a treatment. It is that the findings used to make both diagnoses are common in shoulders that are completely fine.
- 3% of 2,563 consecutive routine shoulder scans showed isolated wasting of the teres minor. Not one of those 61 people had a history of the syndrome, and not one had a lesion in the space.
- 80% of asymptomatic volunteers showed the artery closing off when the arm was raised, which is the finding the founding diagnostic criterion was built on.
- 14 of 16 normal cadaveric shoulders contained the fibrous bands blamed for the compression, in a study run precisely because the original describers had not seen them.
- Muscle wasting at the back of the shoulder blade is markedly more common in overhead athletes than in everyone else, with no convincing link to either pain or performance, and healthy pitchers develop measurable nerve slowing as a season goes on.
Order the scan in someone with symptoms and you will find a criterion met. What nobody can tell you is whether it was met before the symptoms started. The wasting, the slowing, the artery closing and the fibrous bands are findings of the sport as much as findings of the disease.
Surgery against conservative care, honestly: there is no usable conservative success rate to compare against. The conservative evidence is four cases from 1976 that all recovered, one documented spontaneous resolution, and one case report. The surgical numbers look superb and have no comparison group. So the honest position is narrow: operate on a lesion you can see, not on a diagnosis you have inferred. For everyone else the six-month trial is the recommendation, its content is undefined, and the thing that most changes the outcome is not which exercises you pick but whether you check objectively at twelve weeks instead of drifting to month nineteen.
