Rotator Cuff & Shoulder Pain
A tear on the scan is not automatically the reason your shoulder hurts
What it is
Shoulder pain is a symptom, not a diagnosis, and the rotator cuff is only one of its causes. The cuff is a group of four muscles and their tendons that hold the ball of the shoulder into its shallow socket and drive most of its movement, and it can be painful without being torn (tendinopathy), partly torn, or fully torn. But a painful shoulder can equally be adhesive capsulitis — 'frozen shoulder', where the joint capsule itself becomes inflamed and contracted, producing a distinctive loss of movement even when someone else moves the arm for you. It can be glenohumeral osteoarthritis, or pain from the small acromioclavicular joint on top of the shoulder. It can be instability. It can be a nerve root in the neck referring pain down into the shoulder, which is a spine problem wearing a shoulder disguise. And it can be pain referred from somewhere that is not musculoskeletal at all. These have different natural histories and different treatments, which is why 'shoulder pain' and 'rotator cuff tear' are not synonyms — and why the honest starting point is that telling them apart is harder than it sounds.
Why it matters
The central problem in this area is that the pictures are unreliable narrators. Rotator cuff abnormalities are extraordinarily common in people whose shoulders do not hurt: in a pooled analysis, they were present in 9.7% of people aged 20 or younger and rose to 62% of those aged 80 and over, and the prevalence climbed with age at a similar rate regardless of whether the shoulder had symptoms. The authors' conclusion is the sentence that should govern how any shoulder MRI is read — that cuff degeneration is common enough to be considered a normal part of human ageing, and that this makes it genuinely difficult to know when a finding is new or is the cause of the symptoms. So a scan showing a tear tells you a tear is there. It does not tell you that the tear is why it hurts, and it certainly does not tell you it needs repairing. The second problem follows from the first: when imaging findings are treated as causes, operations follow — and the best trials of the commonest shoulder operation found it performed no better than a placebo operation. This page exists to hold both of those facts steady without tipping into the opposite error, which is dismissing a shoulder that genuinely needs looking at.
What BioSignal knows about treating this
What works for Rotator Cuff & Shoulder Pain
BioSignal’s clinical summary, most important first.
- Time and activity modification — most non-traumatic shoulder pain improves without anything invasive
- Exercise-based rehabilitation — the reasonable first-line, though the placebo-controlled evidence is thinner than its popularity implies
- Analgesia and NSAIDs — for symptom control, not repair
- Corticosteroid injection — small, transient relief at 4-8 weeks and no better than placebo by 3 months; useful to open a window, not a treatment plan
- Surgical repair — for traumatic full-thickness tears with genuine weakness, and where a specialist judges it worthwhile; not an automatic consequence of a tear on a scan
- Subacromial decompression — questioned by its own trials; no better than placebo surgery for subacromial pain
- Specific treatment for the specific diagnosis — frozen shoulder, arthritis and radiculopathy are managed differently and are not cuff problems
Signal Records relevant to this condition
Interventions and contributing factors — some of these records describe a cause rather than a cure. The rating shown is BioSignal’s confidence in that Signal Record, not a claim about how well it treats this condition. Open any of them for the full evidence.
New to this? Read these first
How this usually unfolds
- Recognize risk factors
- Get diagnosed
- Track key biomarkers
- Lifestyle first
- Evidence-based treatment
- Long-term monitoring
Who is at risk
- Increasing age — the dominant factor, and the reason imaging findings are so common
- Overhead work or repetitive overhead loading
- Previous shoulder injury or dislocation
- Diabetes — a well-established risk factor for frozen shoulder specifically
- Thyroid disease (associated with frozen shoulder)
- Smoking
- Prolonged immobility of the arm — the usual route into a frozen shoulder
How it's diagnosed
The diagnosis is built from the story and the examination, and both are less precise than people expect. History does a lot of the work: pain that came on after a specific injury behaves differently from pain that crept up over months, and sudden weakness after a fall is a different proposition from a gradually stiff, aching shoulder. Examination narrows it — the single most useful distinction is between a shoulder that will not move because it hurts and one that will not move even when the examiner moves it, which points toward frozen shoulder or arthritis rather than the cuff. But the specific 'impingement' tests are weaker than their reputation: in meta-analysis, the Neer test had a sensitivity of 72% and a specificity of 60%, the Hawkins-Kennedy test 79% and 59%, and the painful arc 53% and 76%. Specificities around 60% mean these tests label a great many people positive who do not have the thing being tested for, so they are useful as part of a pattern and misleading as a verdict. Imaging is for specific questions, not for reassurance: ultrasound or MRI genuinely help when there has been significant trauma, when weakness suggests a full-thickness tear that might need repairing, when surgery is actually being considered, or when the diagnosis is unclear after a proper assessment. Scanning an ordinary aching shoulder early tends to find age-related changes that were always going to be there and to attach them to the pain.
Key biomarkers
Lifestyle
Explore this condition across BioSignal
Related Foundations
Related Signal Records
Frequently asked questions
My MRI shows a rotator cuff tear — do I need surgery?
Usually not, and the reasoning behind that is stronger than most people are told. Start with what a scan means. In a pooled analysis of people WITHOUT shoulder symptoms, rotator cuff abnormalities were found in 9.7% of those aged 20 and under and in 62% of those aged 80 and over — and the rise with age was similar whether or not the shoulder hurt. The authors concluded that cuff degeneration is common enough to be regarded as a normal part of ageing, and that this makes it hard to tell when a finding is actually causing symptoms. So a tear on your scan may be the reason your shoulder hurts, or it may be a finding you have had for a decade without noticing. Then look at what surgery achieves. In a randomised trial of non-traumatic cuff tears comparing physiotherapy alone, acromioplasty plus physiotherapy, and cuff repair plus physiotherapy, there was no significant difference in clinical outcome between the three at two years, and the authors concluded conservative treatment is a reasonable initial option — while honestly noting that tears may progress in unrepaired shoulders, which is why follow-up matters. The picture changes for a traumatic tear with genuine new weakness, especially in a younger person: that is the situation where early surgical opinion is appropriate. The point is not that surgery never helps. It is that 'the scan shows a tear' is not by itself the reason to have it.
Does the operation for shoulder impingement work?
The best evidence says essentially no, and this is one of the clearest results in orthopaedics. Arthroscopic subacromial decompression was for years among the most commonly performed shoulder operations. Then it was tested against a placebo operation. In the CSAW trial, 313 patients were randomised to decompression surgery, to arthroscopy alone — a sham procedure with no decompression — or to no treatment. At six months, there was no difference between real surgery and the sham (Oxford Shoulder Score 32.7 versus 34.2). Both surgical groups did slightly better than no treatment at all, but the difference was not clinically important, and the authors noted it might reflect placebo effects or the post-operative physiotherapy rather than the operation. The Finnish FIMPACT trial reached the same conclusion, and its five-year follow-up found no differences that exceeded the minimally important difference. The authors of CSAW wrote that their findings question the value of the operation and that this should be communicated to patients. That is why the underlying 'impingement' story — that a bone spur is mechanically rubbing the tendon and needs shaving away — has fallen out of favour as an explanation: when the mechanical fix was tested properly, it did not outperform pretending to do it.
Do steroid injections help?
A little, briefly, and not as much as their popularity suggests. A meta-analysis in rotator cuff tendinosis found that corticosteroid injection did not reduce pain more than a placebo injection at three months. There was a small transient benefit at the four-to-eight-week mark — a standardised mean difference of 0.52 — and the authors calculated that at least five people must be injected for one person's pain to be transiently reduced to no more than mild. Multiple injections were not more effective than a single one. Their conclusion was that steroid injections provide, at best, minimal transient relief in a small number of people. That does not make an injection pointless: a few weeks of reduced pain can be genuinely worth it if it lets you sleep or lets you engage with rehabilitation you otherwise could not. It does mean an injection is a window rather than a treatment, that repeat injections are not a strategy, and that anyone offering a course of them should be asked what the plan is for when they stop working.
Is my shoulder pain a rotator cuff problem, or something else?
Often something else, and this is worth being specific about because the treatments diverge. Frozen shoulder (adhesive capsulitis) has a distinctive signature: the shoulder becomes stiff as well as painful, and crucially it will not move even when someone else moves it for you — a cuff problem usually permits passive movement. It is strongly associated with diabetes, which is a link worth knowing if you have one. Glenohumeral arthritis produces stiffness and grinding, and shows on a plain X-ray. Acromioclavicular pain sits pinpoint on the top of the shoulder rather than down the outside. Cervical radiculopathy — a nerve root irritated in the neck — refers pain into the shoulder and arm, and is the one people most often self-diagnose as a shoulder problem; if there is pain past the elbow, or tingling, numbness, or weakness in the hand, the neck is a likelier culprit than the cuff. And some shoulder pain is not musculoskeletal at all. The examination that separates these takes minutes and does not require a scanner, which is a good reason to be assessed before being imaged.
Is it caused by bad posture, or a bone spur?
Both of these are told with far more confidence than the evidence supports. The bone spur story is the impingement model — the idea that a hook of bone mechanically abrades the tendon, so removing it should fix the pain. It is a tidy mechanical explanation, and it was tested: shaving the bone away performed no better than a sham operation in randomised trials, which is difficult to reconcile with the bone being the problem. Spurs are also common in people with no pain, in the same way tears are. Posture is a similar story: it is widely blamed, intuitively appealing, and not well supported as a specific cause of shoulder pain — postural variation is enormous among people who never develop it. This does not mean how you use your shoulder is irrelevant. Sustained overhead loading matters, and how much and how suddenly you load a tendon matters. But those are questions of load and time, not of a bone shape you were told to be worried about or a spine position you were told to correct. BioSignal states this plainly because both stories are widely used to sell treatment.
What actually helps, then?
Time, load management, and — with an honesty caveat — exercise. Most non-traumatic shoulder pain improves. Exercise-based rehabilitation is the reasonable first-line, and it is what guidelines and clinicians recommend, but BioSignal is going to be straight about the evidence behind it: a Cochrane review identified 60 eligible trials and found that only ONE compared a combination of manual therapy and exercise reflective of common practice against placebo — it was judged high quality, and it found no clinically important differences in any outcome. The same review suggested the effects of manual therapy and exercise may be similar to those of glucocorticoid injection and of arthroscopic decompression, on low-quality evidence. Read that alongside the surgery and injection trials and a consistent picture emerges: in subacromial shoulder pain, nothing available is dramatically effective, most people improve anyway, and the invasive options do not beat the non-invasive ones. That argues for starting with the least invasive, least harmful option and giving it real time — not because we have proof it is powerful, but because everything else has been tested against placebo and has not earned its risks. That is a less satisfying answer than 'do these three exercises', and it is the accurate one.
When should shoulder pain be seen urgently?
When it is not really a shoulder problem, or when something structural has just happened. Get urgent care for shoulder pain that comes with chest pain, breathlessness, sweating, nausea, or pain spreading to the jaw or arm — pain referred from the heart can present in the shoulder, and this is the one presentation where assuming a musculoskeletal cause is the dangerous error. Seek prompt assessment after significant trauma, if the shoulder is visibly deformed, if you cannot lift the arm at all after an injury, or if you suspect a fracture or dislocation. A hot, red, swollen joint with fever needs same-day assessment — a joint infection is an emergency. New numbness, a cold or pale hand, or significant weakness needs urgent review. Otherwise most shoulder pain is not urgent, and the more common error is the opposite one: an early scan, an incidental finding, and a treatment aimed at a picture rather than a person.
Evidence summary
Shoulder pain is a symptom with several distinct causes — rotator cuff tendinopathy and tears, subacromial pain, adhesive capsulitis, glenohumeral and acromioclavicular arthritis, instability, cervical radiculopathy and referred pain — and the diagnostic categories genuinely overlap. Imaging is a poor arbiter of causation: pooled data show rotator cuff abnormalities in 9.7% of asymptomatic people aged 20 and under rising to 62% at 80 and over, increasing with age at a similar rate irrespective of symptoms, leading the authors to conclude cuff degeneration is a common aspect of normal ageing and that it is difficult to establish when a finding causes symptoms. Physical examination tests are correspondingly limited (Neer sensitivity 72%/specificity 60%; Hawkins-Kennedy 79%/59%; painful arc 53%/76%), useful as pattern rather than verdict. The mechanical impingement model has not survived testing: in CSAW (313 patients), arthroscopic subacromial decompression was no better than placebo arthroscopy at six months, both surgical arms exceeded no treatment only by a margin that was not clinically important, and the authors concluded the findings question the operation's value; FIMPACT's five-year follow-up found no differences exceeding the minimally important difference. For non-traumatic cuff tears, a randomised trial found no significant difference at two years between physiotherapy alone, acromioplasty plus physiotherapy, and repair plus physiotherapy, supporting conservative treatment as a reasonable initial option while noting unrepaired tears may progress. Corticosteroid injection gives small transient relief at 4-8 weeks (SMD 0.52) with no benefit over placebo by three months, NNT at least 5, and no advantage from repeat injections. Exercise-based rehabilitation is first-line by convention and consensus rather than by strong placebo-controlled evidence: a Cochrane review of 60 trials found only one comparing common-practice manual therapy and exercise against placebo, which found no clinically important difference, with effects possibly similar to injection and decompression on low-quality evidence. The consistent reading is that no available treatment for subacromial shoulder pain is dramatically effective, most cases improve, and invasive options do not outperform non-invasive ones — which argues for the least invasive first. Traumatic full-thickness tears with new weakness are a distinct situation warranting early specialist opinion. Diabetes is an established risk factor for adhesive capsulitis. This page covers the shoulder; neck pain, cervical radiculopathy, generalised arthritis, elbow pain and tendinopathy as a tissue disorder are separate objects, and shoulder pain with cardiac features is a guarded emergency.
References & sources
- Teunis T, Lubberts B, Reilly BT, Ring D. A systematic review and pooled analysis of the prevalence of rotator cuff disease with increasing age. J Shoulder Elbow Surg 2014;23(12):1913-1921 (PMID 25441568; DOI 10.1016/j.jse.2014.08.001)
- Beard DJ, Rees JL, Cook JA, et al. (CSAW Study Group). Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. Lancet 2018;391(10118):329-338 (PMID 29169668; DOI 10.1016/S0140-6736(17)32457-1)
- Paavola M, Kanto K, Ranstam J, et al. (FIMPACT Investigators). Subacromial decompression versus diagnostic arthroscopy for shoulder impingement: a 5-year follow-up of a randomised, placebo surgery controlled clinical trial. Br J Sports Med 2021;55(2):99-107 (PMID 33020137; DOI 10.1136/bjsports-2020-102216)
- Kukkonen J, Joukainen A, Lehtinen J, et al. Treatment of nontraumatic rotator cuff tears: a randomized controlled trial with two years of clinical and imaging follow-up. J Bone Joint Surg Am 2015;97(21):1729-1737 (PMID 26537160; DOI 10.2106/JBJS.N.01051)
- Mohamadi A, Chan JJ, Claessen FMAP, Ring D, Chen NC. Corticosteroid injections give small and transient pain relief in rotator cuff tendinosis: a meta-analysis. Clin Orthop Relat Res 2017;475(1):232-243 (PMID 27469590; DOI 10.1007/s11999-016-5002-1)
- Hegedus EJ, Goode AP, Cook CE, et al. Which physical examination tests provide clinicians with the most value when examining the shoulder? Update of a systematic review with meta-analysis of individual tests. Br J Sports Med 2012;46(14):964-978 (PMID 22773322; DOI 10.1136/bjsports-2012-091066)
- Page MJ, Green S, McBain B, et al. Manual therapy and exercise for rotator cuff disease. Cochrane Database Syst Rev 2016;(6):CD012224 (PMID 27283590; DOI 10.1002/14651858.CD012224)
- Dyer BP, Rathod-Mistry T, Burton C, van der Windt D, Bucknall M. Diabetes as a risk factor for the onset of frozen shoulder: a systematic review and meta-analysis. BMJ Open 2023;13(1):e062377 (PMID 36599641; DOI 10.1136/bmjopen-2022-062377)
Educational information — not medical advice
Is this condition page clear, accurate, and useful? Your feedback shapes what we review next.
Give feedback