Rotator Cuff Tear — What You Need to Know
Author: Dr. Alexandru Grecu — Senior Orthopedic and Trauma Surgeon
Published: March 26, 2026 · Reviewed: 2 mai 2026
What Is the Rotator Cuff?
The rotator cuff is a group of four muscles whose tendons form a "cuff" around the humeral head. The supraspinatus passes over the joint and starts the lifting of the arm. The infraspinatus and the teres minor attach at the back and rotate the arm outwards. The subscapularis attaches at the front and rotates the arm inwards. Together they do something simple to describe and hard to replace: they keep the humeral head centred in the socket so that the large muscles, the deltoid first of all, can work.
That is also where the vulnerability comes from. The shoulder has no deep socket, as the hip does. Stability comes from the tendons, and the tendons work with every movement above the head. The supraspinatus fails first for two reasons that add up. It runs through the narrowest place, the space between the humeral head and the acromion, where it is rubbed and compressed when the arm is raised. And close to its insertion on the humerus it has a zone with a poorer blood supply, which recovers less well after repeated microdamage.
A tear means that the tendon fibres have detached from the bone, either partially or through the full thickness. It does not automatically mean pain and it does not automatically mean surgery. Those two statements look contradictory only until the explanation below.
Degenerative vs. Traumatic
A degenerative tear is a process, not an accident. The tendon thins over years, fibre by fibre. It starts as irritation, continues as a partial tear on the deep surface and becomes, in some patients, a full-thickness tear. There is no day on which it happened. Patients often say "it has bothered me for a long time, but for a month now I cannot manage". What changed in that month is not necessarily the tendon but the balance: the shoulder could no longer compensate.
This is where the observation that changes the discussion comes in: many degenerative tears do not hurt. They are found by chance, on an MRI done for something else, or in the shoulder opposite the painful one. Follow-up over time of asymptomatic degenerative tears shows that some stay silent for years while others become painful. It cannot be said in advance, for an individual patient, which group they will be in.
A traumatic tear is an event. A fall on the arm, a sudden pull when you grab something to avoid falling, a jerk while lifting a weight. The pain appears at that moment, and the weakness is immediate and clear: the arm no longer lifts, or no longer stays up. In an active patient with a good shoulder until the day of the accident, this situation follows a different logic from wear.
Why the difference matters. Because the decision is not made on the image, it is made on the story of the shoulder. In a degenerative tear the image shows a state that probably existed before the pain, and treatment begins without surgery. In a traumatic tear on a previously healthy tendon, repair has a clearer purpose and the time window matters more, because a detached tendon retracts and the muscle degrades. The same image, two different decisions, because the question is not "what can be seen" but "what happened and what can you no longer do".
Symptoms
Night pain on the affected shoulder is the most commonly reported symptom. During the day the arm is supported and distraction is plentiful. At night, lying down, the capsule tightens and lying on the painful side becomes impossible. Patients move their pillow, sleep in an armchair, wake several times. Broken sleep is often the real reason for coming to the doctor.
Pain when lifting the arm, felt on the outer side of the shoulder and often running down towards the middle of the arm, typically appears in a particular sector of the movement. The gestures that provoke it are ordinary: reaching the top shelf, fastening the seatbelt, putting on a coat.
Weakness on lifting is the sign that separates irritation from a tear. It is not pain stopping the movement but strength that is missing. In large tears the "drop arm" phenomenon appears: you lift the arm with the help of the other hand, and when you let go it falls. Sometimes the patient compensates by shrugging from the shoulder blade, with a movement pattern visibly different from the healthy side.
What separates a tear from capsulitis. In a tear, someone else can move your arm further than you can: passive motion is preserved, active motion is limited. In adhesive capsulitis (frozen shoulder) the shoulder is blocked both ways — neither you nor the examiner obtain the range, especially in external rotation. This distinction is made in the consulting room, in a few minutes, and it changes the treatment completely.
What separates a tear from osteoarthritis. Glenohumeral osteoarthritis gives deep, diffuse pain, with grinding on rotation and slowly increasing stiffness, and the X-ray shows a narrowed joint space. A cuff tear gives pain with a specific gesture and weakness, with an X-ray that is often normal at first. The two can overlap in older patients, and when the damage is advanced, shoulder arthroplasty is discussed, including the reverse design, which was conceived precisely for a shoulder without a functioning cuff.
Diagnosis
The clinical examination tests each tendon separately. For the supraspinatus, the Jobe test: you hold the arms out in front, slightly to the side, with the thumbs down, while the examiner presses down — pain or giving way points to the supraspinatus. Also for it, the drop-arm test: the arm raised passively to the horizontal cannot be lowered under control but falls. For the infraspinatus, external rotation against resistance with the elbow against the body: if strength is missing, the tendon at the back is involved. For the subscapularis, the lift-off test: with the back of the hand on the back at waist level, you try to lift the hand away; being unable to shows a lesion at the front. To these are added the subacromial impingement manoeuvres, which reproduce the pain on lifting with the arm rotated inwards.
No single test makes the diagnosis on its own. Their value lies in the combination and in the comparison with the other shoulder. A patient with night pain, a positive Jobe test and reduced strength in external rotation has, until proven otherwise, rotator cuff disease.
The X-ray does not show the tendon, but it is taken anyway: it excludes osteoarthritis, shows calcifications and may show the humeral head riding high, an indirect sign of an old and large tear.
Ultrasound shows the tendon directly and, unlike MRI, shows it in motion. In experienced hands it says whether the tear exists, whether it is partial or full-thickness, how large the defect is and whether there is fluid in the bursa. It is quick, both shoulders can be examined in the same session and it can be repeated. In a patient with a clear clinical picture in whom treatment without surgery will be started anyway, ultrasound is usually enough.
MRI becomes useful when surgery is being discussed. It shows how far the tendon has retracted, how far the muscle has degraded, whether several tendons are involved and whether there are associated lesions — labrum, long head of the biceps. These details do not change conservative treatment, but they change what can and cannot be repaired.
Treatment Without Surgery, Honestly
In degenerative tears, physiotherapy is the first line. Not as a way of waiting, but as treatment in its own right. The logic is mechanical: if one tendon no longer pulls, the others can be trained to keep the humeral head centred, and the shoulder blade can be taught to rotate so that it leaves room for the tendon under the acromion. A shoulder with a tear can work well if the rest of the mechanism works correctly.
The most cited evidence comes from the multicentre MOON cohort, which followed patients with atraumatic full-thickness tears treated with a structured physiotherapy programme. Many of them improved in pain and function and chose not to have surgery during follow-up. That is a descriptive result, not a guarantee: it shows that the non-surgical path is a real option in this type of tear, not that it works for everyone.
The programme follows a few practical rules. Mobility is worked first, then control of the shoulder blade, then the strength of the rotators, at low ranges below shoulder level. Overhead exercises and pressing are avoided at the beginning. Work is frequent and short, not rare and long. And pain that increases and stays increased after the session means the dose was too high.
The corticosteroid injection has a limited, short-term role. It reduces inflammation in the bursa and may ease pain enough for you to do the physiotherapy — that is its main purpose. It does not repair the tendon. It is not repeated indefinitely, because repeated exposure of the tendon to corticosteroid does it no good. When surgery is likely in the near future anyway, the timing of the injection is discussed, for exactly this reason.
What conservative treatment does NOT do: it does not glue the torn tendon back. Fibres detached from bone do not reattach through exercise. What can change is pain and function, and for many patients that is exactly what they asked for. When biological options such as PRP injections are discussed, the same limit applies and has to be stated beforehand, not afterwards.
When Surgery Is Considered, and What the Evidence Says
The indications I use are three, and none of them is "a tear was seen on MRI". First: a traumatic tear in an active patient with a good shoulder before the accident. Second: marked weakness with clear loss of function — the arm no longer lifts, you can no longer work above your head, you can no longer put your hand behind your back. Third: failure of properly conducted conservative treatment, with physiotherapy actually done, not merely recommended.
Now the part a surgeon has to write about the operation he performs. The 2019 Cochrane review of surgery for rotator cuff tears finds that repair adds little or nothing compared with conservative treatment in degenerative tears, with low to moderate certainty of the evidence. Also from 2019, the Cochrane review of subacromial decompression shows that this procedure, done on its own, brings no important benefit over placebo or over treatment without surgery. These are uncomfortable conclusions, but they are the data we have, and I state them at the consultation.
What these reviews do not say. They do not say that surgery helps nobody. The included studies mostly recruited patients with degenerative tears, often older, with poor tendon quality — not the young patient with a traumatic tear and an arm that no longer lifts. Where good studies are missing, the evidence is missing too, which is not the same as evidence that it does not work. Honesty means telling you which category you are in and how solid the basis for the decision is.
In practice, surgery is done arthroscopically: the tendon is released, the insertion area is prepared and the tendon is reattached to the bone with anchors. When the tendon has been retracted for a long time and the muscle has degraded, repair may no longer be possible, and the discussion moves to other solutions, including a reverse replacement in an older patient with associated osteoarthritis. That is why the decision should not be postponed indefinitely when there is a repairable traumatic tear.
Recovery After Cuff Repair
Surgery puts the tendon back on the bone. The healing is done by the body, and the tendon reintegrates into the bone slowly, more slowly than skin or muscle heals. That is why rehabilitation is run in stages, and the stages are not skipped. Exact timings are set according to the type of tear, the quality of the tendon and the fixation used, so what follows describes the sequence, not the calendar.
The protection stage. The arm stays in a sling, at night as well. The fingers, wrist and elbow are moved so that they do not stiffen. The shoulder is not moved actively. The single aim of this stage is that the repair is not loaded.
The passive motion stage. The physiotherapist, or you with the healthy hand, moves the arm without the operated muscles pulling. Range is regained, stiffness is prevented, and the tendon stays protected. This is the stage in which patients most often go wrong, because the pain has eased and the temptation to use the arm is strong.
The active motion stage. The muscles begin to work, without weights, at low ranges first. Control of the shoulder blade is restored. Everyday movements return gradually.
The strength stage. Progressive resistance is added, then the specific movements of work or sport. Returning to activities that load the shoulder above the head comes last, once strength and control have been restored.
Signs that call for a review without waiting for the next appointment: pain that increases suddenly after an uncontrolled movement, with the feeling that "something gave way"; loss of a movement you had regained; fever, redness or discharge at the incisions; persistent numbness of the hand; sudden swelling of the arm.
The Key Message
Not every tear requires surgery — but every tear requires an evaluation. The image describes the tendon. The decision is made on what you can no longer do, on how the tear appeared and on what you have already tried.
Frequently Asked Questions
Does a rotator cuff tear heal on its own?
A torn tendon does not reattach by itself. The pain, however, can settle, and the shoulder can work well if the remaining muscles take over the load. That is why there are patients with a tear on imaging and a useful shoulder. What heals is the symptom, not the tendon.
If I do not have surgery, will the tear get bigger?
It may enlarge, but not in everyone and not predictably. Follow-up of degenerative tears shows that some stay stable and silent while others progress and become painful. That is why the decision is not made once: it is reassessed at follow-up, with clinical examination and, when needed, imaging.
Is ultrasound enough?
Often yes. In experienced hands, ultrasound shows whether the tendon is torn, how large the defect is and how the shoulder moves in real time. MRI is added when surgery is being discussed, when labral or other lesions are suspected, or when the ultrasound is inconclusive.
Does an injection fix the tear?
No. A corticosteroid injection reduces inflammation and may ease pain in the short term, which sometimes makes physiotherapy possible. It does not repair the tendon and it is not repeated indefinitely. If pain returns immediately after every injection, that is a sign the problem needs a different approach.
Can I sleep on the shoulder after surgery?
Not in the first weeks. Sleep is usually on the back or on the healthy side, with the arm supported, and the sling is worn as instructed. Moving back to sleeping on the operated shoulder is gradual, once night pain has gone and active motion has resumed — it is decided at follow-up.
For context, see the dedicated page on shoulder arthroplasty, and if you want to go deeper, read aboutthe causes of shoulder pain and aboutadhesive capsulitis (frozen shoulder).
Have questions or need a consultation?
Scientific References
Karjalainen TV, Jain NB, Heikkinen J et al. — Surgery for rotator cuff tears. Cochrane Database Syst Rev. 2019;12:CD013502.
For rotator cuff repair: the Cochrane review finds that surgery adds little or nothing compared with conservative treatment in degenerative tears, with low to moderate certainty of the evidence.
Karjalainen TV, Jain NB, Page CM et al. — Subacromial decompression surgery for rotator cuff disease. Cochrane Database Syst Rev. 2019;1:CD005619.
For isolated subacromial decompression: it brings no important benefit over placebo or over treatment without surgery in rotator cuff disease.
Kuhn JE, Dunn WR, Sanders R et al. — Effectiveness of physical therapy in treating atraumatic full-thickness rotator cuff tears: a multicenter prospective cohort study. J Shoulder Elbow Surg. 2013;22(10):1371-9.
For atraumatic full-thickness cuff tears: in the MOON cohort, a structured physiotherapy programme improved pain and function in many patients, and few went on to surgery during follow-up.
Torchia MT, Sefko JA, Steger-May K et al. — Evaluation of survivorship of asymptomatic degenerative rotator cuff tears in patients 65 years and younger. J Shoulder Elbow Surg. 2023;32(9):1857-1865.
For painless degenerative tears: follow-up over time shows that some remain silent while others become symptomatic — which is why imaging alone does not dictate surgery.
