Rotator cuff tear: surgery or exercise?

What is the rotator cuff, and why do tears occur?

The rotator cuff is four muscle–tendon groups that surround the head of the shoulder and let us lift and rotate the arm (the most commonly torn is the supraspinatus tendon on top). Tears occur by two main routes:

• Degenerative (atraumatic): develops over years with wear of the tendon, often without a clear accident. It is a problem of middle–older age and is very common.
• Traumatic: occurs after a specific event such as a fall, a sudden pull of the arm, or a dislocation, often in a tendon that was healthy until that moment. Most full-thickness tears in young patients are of this group.

This distinction is the backbone of treatment; because the evidence points in a different direction for the two groups.

Seeing a tear on MRI = not the cause of pain

Perhaps the most important fact on this subject is this: with age, a significant portion of people who have no shoulder pain at all also show a rotator cuff tear on MRI or ultrasound.

*Combined approximate values from different population screenings (ultrasound/MRI); rates vary from study to study.

In the classic study by Sher and colleagues, a tear (full or partial) was detected on MRI in 34% of volunteers with no complaints at all; over the age of 60 this rate rose to 54%. An important detail: this 54% is the sum of full and partial tears — in the same age group the full-thickness tear rate is about 28%. A screening of a village in Japan found the frequency of tears in the community to be around 20% and showed that painless tears were about twice as common as painful ones. The practical result is this: seeing a tear on an MRI taken for shoulder pain in middle–older age does not prove that the source of the pain is that tear. Pain can also come from other causes such as impingement, tendon disease, frozen shoulder, or referred pain from the neck. That is why an imaging report does not replace a careful history and examination; the treatment decision is made according to the patient, not the report.

Evidence in degenerative tears: starting with exercise is reasonable

• Finnish randomized trial (Kukkonen et al.): 180 shoulders over age 55 with small supraspinatus tears and no history of trauma were divided into three arms: physiotherapy alone; acromioplasty (bone shaving) + physiotherapy; tendon repair + acromioplasty + physiotherapy. There was no significant difference between the groups at years 1 and 2. In follow-up beyond 5 years, clinical outcomes remained similar and surgery was seen not to prevent joint degeneration; the authors' conclusion is that in these patients conservative treatment is a suitable first option.
• Cochrane review (2019): a combined analysis of randomized trials comparing repair surgery with non-operative treatment in full-thickness tears concluded that surgery provides no clinically important added benefit in pain, function, and quality of life at one year (certainty of evidence moderate for pain, low for some function and quality-of-life outcomes).
• In patient series followed non-operatively, it has been reported that with an exercise program the majority of patients remained in good condition over the years without surgery.

Exercise therapy is not "wait-and-see": it is a structured program, usually lasting at least 12 weeks, built on scapular control and gradual strengthening of the intact portions of the cuff and the surrounding muscles. The aim is not to "glue" the tear but to make the shoulder work painlessly and functionally despite the tear — which is achieved in most degenerative tears.

The other side of the coin: data favoring surgery in the long term

To be honest: the evidence picture is not one color. In the study by Moosmayer and colleagues, which randomized 103 patients with small–medium (under 3 cm) tears to repair versus physiotherapy, the early small difference widened in favor of repair at the 10- and 15-year follow-up (about a 10-point difference in shoulder score at 10 years). Also, a portion of the patients in the physiotherapy arm crossed over to surgery during follow-up. These findings suggest that especially in patients with a long life expectancy, active and relatively young, repairing small–medium tears may be meaningful in the long term. By contrast, no such difference was seen in the older group with smaller tears in the Finnish study. That is, the evidence does not allow us to say "no degenerative tear is ever operated on"; both paths can be defended according to the patient profile.

When does surgery come to the fore?

In the surgical decision, the size and location of the tear, the degree of tendon retraction, the degree of muscle fatty change, the patient's age, job/sport goal, and expectations are weighed together.

Does a tear progress? Why is follow-up important?

Degenerative tears are not static. In studies where painless tears were followed over the years (Keener et al.), about half of full-thickness tears enlarged within 5 years and pain developed over time in a significant portion of the shoulders followed. What is interesting is this: enlargement and pain do not always go hand in hand — some enlarging tears remain painless, while some non-enlarging tears can become painful. The practical messages:

• Starting with exercise is not "closing the door"; the surgical option remains open in most patients.
• But follow-up should not be without a plan: if symptoms worsen, if loss of power is added, or if control imaging shows marked enlargement/muscle deterioration, the decision is reviewed.
• In a young patient and in a traumatic tear, evaluation should not be delayed, and an orthopedic opinion should be obtained early.

The decision is shared

There is no single "right for everyone" answer on this subject; the evidence supports choosing together with the patient between two legitimate paths. In a patient in their 60s, without trauma, with a small tear and a daily-life goal, starting with exercise rests on strong evidence; in a 45-year-old patient who cannot lift the arm after a fall, early repair comes to the fore. In the wide gray zone between the two, the characteristics of the tear, the patient's goals, and preferences determine the decision. Remember: this article provides general information; it does not replace an examination and an individualized assessment.

For clinicians

The first split of the decision tree is etiology and chronicity: in an atraumatic–degenerative tear (especially >55–60 years, <3 cm, supraspinatus-limited in the sagittal plane) the first step is 3–6 months of structured exercise (Kukkonen 1–2 years and >5 years; Cochrane 2019 — moderate certainty for pain, low for some function/quality-of-life outcomes). In an acute traumatic full-thickness tear — especially in a young patient and with marked loss of power — prompt evaluation and early orthopedic opinion are recommended; there are observational data that structural and functional outcomes are better with repair done before retraction and Goutallier stage progress, but a precise week/month limit has not been shown with high-quality evidence — timing is individualized to tear size, retraction, loss of power, and patient characteristics. Moosmayer's 10–15 year data reported a difference widening over time in favor of primary repair in small–medium tears (Constant ~9–10 points), and cases in the physiotherapy arm that crossed over to secondary surgery and became irreparable; in a young–active patient this data must be part of the shared-decision conversation. Because of the prevalence of asymptomatic tears (>25–50% over age 60), an indication should not be made without correlating the imaging finding with the symptom; as pain generators, subacromial pain syndrome, biceps pathology, glenohumeral OA, and cervical radiculopathy should be ruled out. In full-thickness tears followed conservatively, serial imaging is not routine; imaging is reasonable in clinical deterioration, new loss of power, a young patient, or cases with a high risk of progression. Progression (>5 mm enlargement), new pseudoparalysis, or deterioration in muscle quality lowers the surgical threshold. The AAOS rotator cuff guideline updated in 2025 reports moderate-level evidence that, in healed repairs, surgery gives better outcomes than non-operative treatment, and does not recommend routinely adding acromioplasty to small–medium full-thickness tear repair. Massive–irreparable tear and cuff tear arthropathy are a separate algorithm (partial repair, tendon transfer, superior capsule reconstruction, reverse shoulder arthroplasty) and are outside the scope of this article.

Frequently asked questions

Related articles

• Shoulder impingement syndrome: why the name is changing, and why exercise comes first
• Recurrent shoulder dislocation: what happens after the first dislocation, and when does surgery come up?
• Frozen shoulder (adhesive capsulitis): does it resolve on its own, and what helps in treatment?

Scientific references

This information is for general guidance only; it does not replace an examination, a diagnosis, or your physician's individual advice. Please consult a physician for your symptoms. About the author: Doç. Dr. Oğuz Yüksel — Academic Profile.