Shoulder pain after 60 is usually not a torn rotator cuff worth operating on

The rotator cuff is four muscles — supraspinatus, infraspinatus, teres minor, subscapularis — whose tendons wrap the head of the humerus and hold it centred in a shallow socket while the larger muscles move the arm. The shoulder trades stability for range, which is why it does remarkable things and why it goes wrong.

Here is the finding that should shape how any shoulder scan is interpreted. Studies imaging asymptomatic shoulders find rotator cuff tears in roughly a quarter of people over sixty, and more than half of those over eighty. In people with a painful shoulder on one side, tears are commonly present on the pain-free side too.

So a tear on a scan is a normal age-related finding as often as it is a diagnosis. It may explain the pain. It may be an incidental discovery in a shoulder that hurts for another reason entirely.

The things that actually cause shoulder pain at this age

Rotator cuff related shoulder pain — the current preferred umbrella term, replacing "impingement", which described a mechanism that has not held up well. Pain on reaching overhead, on lying on that side at night, and weakness on resisted movement.

Frozen shoulder (adhesive capsulitis) — a genuinely distinct condition, and the one most often misdiagnosed. The distinguishing feature is loss of passive range: someone else moving your arm cannot get it there either. It is associated with diabetes and thyroid disease, follows a long course of freezing, frozen and thawing phases over one to three years, and generally resolves — though often with some residual restriction.

Glenohumeral osteoarthritis, less common than at the hip or knee but not rare, producing stiffness and crepitus.

Acromioclavicular joint arthritis, pain localised to the top of the shoulder and worse on reaching across the body.

Referred pain from the cervical spine, which is common and frequently missed. Neck-derived shoulder pain often extends past the elbow, comes with neck stiffness, and changes with neck position.

What the surgical trials showed

Subacromial decompression was among the most commonly performed orthopaedic operations for shoulder pain. Then two well-conducted trials tested it against a placebo procedure.

CSAW, published in 2018, randomised patients to decompression, to arthroscopy without decompression, and to no treatment. Both surgical groups improved similarly, and the difference from no treatment was small and of uncertain clinical importance. The Finnish FIMPACT trial reached a comparable conclusion.

The rate of the operation has fallen substantially in several countries as a result, and guidelines now advise against it for this indication. It remains one of the clearer examples of a widely performed procedure not surviving a placebo-controlled test.

Cuff repair is a different operation and the evidence is more nuanced. For acute traumatic tears in younger, active patients, repair is well supported. For chronic degenerative tears in older adults, trials comparing repair with structured physiotherapy have generally found similar outcomes at one and two years — with the caveat that tears may enlarge over time in some patients, and that a subset does better with surgery.

Exercise as first-line, and what it involves

The evidence for structured exercise in rotator cuff related shoulder pain is reasonably good, and it outperforms doing nothing by a comfortable margin.

The programme has three components. Cuff strengthening — external and internal rotation against a band, with the elbow at the side, progressing in load. Scapular control — rows, retraction work, wall slides, because the shoulder blade is the platform the arm moves from and poor control of it changes everything downstream. Progressive loading into range, gradually reintroducing the overhead positions that hurt.

It takes time. Twelve weeks minimum before judging, and tendon adaptation is slow. The commonest reason people conclude physiotherapy failed is that they did six weeks of it.

Injections, briefly

Subacromial corticosteroid injection provides short-term pain relief — meaningful for a few weeks, diminishing thereafter, with little evidence of long-term benefit. Its legitimate use is to reduce pain enough to allow rehabilitation to start. Repeated injections into tendon are associated with tendon weakening and are best avoided.

For frozen shoulder specifically, intra-articular steroid injection in the painful freezing phase has better evidence and can substantially shorten the miserable part.

When to get it looked at promptly

Most shoulder pain can be managed conservatively for six weeks before imaging. Some cannot.

Sudden weakness after a fall or a lifting injury, particularly an inability to lift the arm at all, suggests an acute tear and is one of the situations where early surgical opinion matters. Trauma with deformity suggests fracture or dislocation. Any mass, unexplained weight loss, fever, or night pain that is unremitting and not positional needs assessment for other causes. Pain radiating below the elbow with numbness points at the neck.

The unglamorous conclusion

Shoulders take a long time. A degenerative cuff problem at sixty-five is typically a six-month project with bands and patience, not a six-week one, and the scan that names it will very likely also name findings that were there ten years ago and never hurt.