Shoulder tendinopathy is overload of the rotator cuff tendons, most often supraspinatus. It is treated with a progressive loading programme over three to six months, not with rest. A UK placebo controlled trial found the standard operation for it no better than sham surgery.
The word tendinitis, still on most British health pages, describes an inflammation that in a chronic tendon is largely absent. What is actually happening is a failed adaptation to load, and that changes everything about the treatment.
Four tendons, and which one is yours
Supraspinatus, the usual suspect
Runs across the top of the joint and lifts the arm away from the body. Painful between 60 and 120 degrees of lifting, and sore to lie on at night.
Subscapularis, at the front
Rotates the arm inwards. Hurts reaching behind your back, doing up a coat, or pulling a wallet from a back pocket.
Infraspinatus and teres minor
Rotate the arm outwards. Weakness here shows up reaching sideways for a seatbelt or turning a stiff tap.
The rotator cuff is four small muscles wrapping the head of the humerus. Their job is not power, it is centring: they hold the ball in the socket while the big muscles move the arm. When one tendon loses capacity, that centring degrades and the whole shoulder starts complaining.
Supraspinatus accounts for most cases. The classic pattern is a painful arc, comfortable at the start of the movement, sore between roughly 60 and 120 degrees, then easier again at the top, plus pain lying on that side at night that wakes people at three in the morning.
Impingement, the older term, described the tendon being pinched under the bony arch of the shoulder. It is still the title on the NHS page, but many British shoulder services have moved to rotator cuff related shoulder pain, because tendinopathy, partial tears and bursitis behave alike and respond to the same management.
Why calling it tendonitis is misleading
Where a tendon sits on the continuum
swollen but intact
days to weeks
disorganised fibres
weeks to months
tissue partly lost
months to years
In 2002 a group of tendon researchers published a paper in the British Medical Journal with a blunt title: time to abandon the tendinitis myth. Biopsies of chronically painful tendons show disorganised collagen, increased ground substance and new blood vessels growing in, but very few inflammatory cells.
Jill Cook and Craig Purdam later set out the continuum model shown above. A reactive tendon is swollen and irritable but structurally intact and reverses quickly. A degenerative tendon has lost tissue in places and will not return to normal, although the surrounding healthy tendon can be strengthened enough that the shoulder works well.
This matters practically because it explains the two commonest mistakes. Loading a reactive tendon too hard keeps it inflamed. Protecting a degenerative tendon indefinitely leaves it weaker every month, and it is the second mistake that dominates in shoulders.
What the loading programme actually looks like
How a loading programme is actually built
Settle the flare, without stopping
Cut the aggravating movement for a week or two. Complete rest is not the aim: an unloaded tendon loses capacity fast and comes back weaker.
Isometric holds
Push into a wall or hold a light weight still for around 40 seconds. Loads the tendon with almost no movement and often reduces pain for a few hours afterwards.
Slow, heavy resistance
Rows, external rotation and press movements taken slowly under real load, three seconds up and three seconds down. This is the phase that changes the tendon rather than the symptoms.
Add speed and range
Once heavy slow work is comfortable, faster and end range movements return, including overhead work if your sport or job needs it.
Keep two sessions a week
Tendon capacity fades when loading stops. The people who relapse are almost always the ones who stopped the programme the week the pain went.
Notice what is not in that list. There is no ultrasound therapy, no rest sling, no six weeks of doing nothing. The evidence for progressive resistance in tendinopathy is much stronger than for any passive treatment, and the shoulder responds to it just as the Achilles and the patellar tendon do.
Heavy slow resistance deserves a specific mention. Trials comparing it with the older eccentric only protocols find broadly similar results with better adherence, because a slow controlled lift is easier to fit into a normal gym session than a prescriptive drop and lower routine.
Expect the pain to lag behind the progress. A tendon that is objectively stronger can still be sore, and the useful marker is what it feels like the next morning rather than during the set. Pain settling within 24 hours means the load was right.
What happens at the physiotherapy assessment
A physiotherapist starts with the story, because in this condition the history carries most of the diagnostic weight. When did it begin, what changed in the weeks before, which movements catch, and how are your nights. Then comes the physical examination: active range of motion first, then passive range, then resisted tests in each direction of the cuff.
The pattern they are looking for is a shoulder with near full passive motion, restricted or painful active motion, and pain on resisted testing in a specific direction. If passive range is also badly limited in every direction, the picture shifts towards frozen shoulder, which is a different condition managed differently.
Strength testing separates tendinopathy from a substantial tear. A tendon that hurts but holds is a loading problem. A tendon that cannot hold the arm against gravity at all suggests a full thickness tear, and those patients do get imaged and do get a surgical opinion.
Other conditions that look like it
Several conditions share the same postcode, and telling them apart changes what you should do. These are the ones that come up most often in British shoulder clinics.
- Frozen shoulder. Passive movement is lost as well as active, and reaching for a seatbelt becomes physically impossible rather than merely painful. Strongly associated with diabetes.
- Acromioclavicular joint problems. Pain sits on the top of the shoulder, at the point you would put a finger on, and is worse reaching across the body. Often follows an old injury to that joint.
- Calcific tendinopathy. A deposit of calcium in the tendon producing sudden, severe pain out of proportion to the activity. Visible on X-ray, and the natural history is often self limiting.
- Referred neck pain. Pain spreading below the elbow, with pins and needles, points to the neck rather than the cuff, and shoulder movements may be entirely comfortable.
- Subacromial pain syndrome after a fall. Injuries in this area frequently mix bruising, bursitis and tendon degeneration, which is why post traumatic shoulders take longer to settle than gradual onset ones.
Injections, scans and the operation
A corticosteroid injection reliably reduces pain for a few weeks. It also weakens tendon tissue with repetition, so British shoulder services use it to open a window for rehabilitation rather than as a treatment in itself. One injection alongside a loading programme is reasonable. Three injections and no programme is how shoulders end up worse. The trade-off has been measured most precisely at the elbow, where the injected group was doing best at six weeks and worst at a year.
Scans are less useful than people expect. Rotator cuff abnormalities are found in a large proportion of shoulders that have never hurt, and their prevalence rises steeply with age, so a scan will frequently show something that predates your symptoms entirely. Imaging earns its place when there is real weakness suggesting a substantial tear, or when a properly delivered exercise programme has failed.
Then there is the operation. Arthroscopic subacromial decompression, shaving bone from under the arch to give the tendon room, was one of the most frequently performed shoulder procedures in the country. The CSAW trial tested it properly: 32 UK hospitals, 51 surgeons, three groups including a placebo arthroscopy where the surgeon looked inside and removed nothing.
Both surgical groups improved more than the no treatment group, and the difference between real decompression and placebo decompression was not clinically important. In other words, most of the benefit was not coming from the bone that was removed. That finding, from British surgeons in British hospitals, is why the pathway now expects a documented course of exercise first.
The things that slow a tendon down
Tendon adaptation is a metabolic process, not purely a mechanical one, and several everyday factors blunt it. Smoking reduces tendon blood supply and healing. Poorly controlled diabetes is strongly associated with rotator cuff problems. Raised cholesterol shows the same association, and the link appears to run through the tendon tissue itself rather than through activity levels.
Sleep is the underrated one. Shoulder tendinopathy disturbs sleep, and disturbed sleep raises pain sensitivity, which is a loop worth breaking deliberately. Sleeping on the other side, with a pillow under the painful arm to stop it dropping across the body, is a simple change that helps a lot of people within a week.
The general inflammatory background is worth attending to as well, since it influences how readily any connective tissue repairs, and it can be measured rather than guessed at. If you are adding something on top of the loading programme, judge a bioavailable anti-inflammatory formula on how much of the active compound is absorbed, and treat a supplement aimed at tendon pain as a support to the exercise rather than an alternative to it. Nothing swallowed rebuilds tendon capacity. Only load does that.
When to get it looked at urgently
- Sudden loss of power after a fall or a wrench, especially if you cannot hold the arm out to the side.
- An arm that drops when you let go of it after someone lifts it for you.
- Shoulder pain with fever, or a hot swollen joint.
- Pain spreading down the arm with numbness, tingling or weakness in the hand.
- Shoulder pain that is unrelated to movement and does not change position by position, which occasionally reflects something outside the joint.
Frequently asked questions
How do you treat shoulder tendinopathy?
With a progressive loading programme, guided by a physiotherapist. Relative rest for a short period, then isometric holds, then slow heavy resistance work, then a return to speed and full range. Anti-inflammatories and injections manage symptoms but do not change the tendon. In the UK this is the pathway the NHS uses, often starting with a supported self care programme before face to face physiotherapy.
Does shoulder tendinopathy ever go away?
Yes, in most people, but slowly. Three to six months of consistent loading is the usual timeframe, and tendons that have been painful for years take longer than tendons that flared last month. What does not work is waiting for it to settle by itself: an unloaded tendon loses capacity, and the shoulder becomes easier to irritate rather than harder.
What not to do with shoulder tendonitis?
Do not rest it completely, do not push through sharp pain into a full overhead programme, and do not stop the exercises the week the pain improves. Avoid repeated cortisone injections, which give short term relief at the cost of tendon quality. And do not sleep on the painful side if you can avoid it, because sustained compression of the tendon overnight is a common reason mornings are worse.
What triggers tendinopathy?
A change in load rather than load itself. A decorating weekend after a sedentary winter, a new gym programme, a job change involving overhead work, a fortnight of heavy gardening. The tendon is not weak because you used it, it is sore because the demand rose faster than the tissue adapted. Age, smoking, diabetes and high cholesterol all reduce how fast tendons adapt.
Is shoulder impingement the same as rotator cuff tendinopathy?
They overlap so much that UK services have largely abandoned the distinction. Impingement described a mechanical pinching under the bony arch, but the evidence that pinching is the primary problem did not hold up. Many NHS shoulder services now use the term rotator cuff related shoulder pain, which covers tendinopathy, partial tears and bursitis together because they behave and respond alike.
Do I need a scan for shoulder tendinopathy?
Usually not at first. Rotator cuff changes are extremely common on scans of people with no pain at all, so an ultrasound or MRI often finds something that was already there and had nothing to do with your symptoms. Imaging is useful when there is significant weakness suggesting a large tear, when symptoms are not improving after a proper loading programme, or when surgery is genuinely being considered.
Does surgery work for shoulder tendinopathy?
Far less often than was assumed. The CSAW trial, run across 32 UK hospitals, compared arthroscopic subacromial decompression with a placebo arthroscopy and with no treatment. Both procedures beat doing nothing, but the difference between real surgery and placebo surgery was not clinically important. That result is a large part of why NHS commissioners now expect a documented course of exercise before this operation is offered.
What to take away
Work out which movement provokes it: lifting sideways, reaching behind your back, or rotating outwards. That single observation tells a physiotherapist more than an ultrasound will, it costs nothing, and it takes about two minutes standing in front of a mirror this evening.
Then commit to twelve weeks. Not six, and not until the pain goes. The shoulders that come back are the ones that kept loading after they felt better.
And if an operation is raised, ask one question: has a proper loading programme been tried and documented? The largest British trial of the standard procedure found it no better than a placebo operation, which makes that question entirely reasonable to put to a surgeon.
Sources
Khan K.M., Cook J.L., Kannus P. et al. (2002). Time to abandon the “tendinitis” myth. British Medical Journal, 324(7338), 626-627. DOI: 10.1136/bmj.324.7338.626
Cook J.L., Purdam C.R. (2009). Is tendon pathology a continuum? A pathology model to explain the clinical presentation of load-induced tendinopathy. British Journal of Sports Medicine, 43(6), 409-416. DOI: 10.1136/bjsm.2008.051193
Beard D.J., Rees J.L., Cook J.A. et al. (2018). Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. The Lancet, 391(10118), 329-338. DOI: 10.1016/S0140-6736(17)32457-1
Beyer R., Kongsgaard M., Hougs Kjaer B. et al. (2015). Heavy slow resistance versus eccentric training as treatment for Achilles tendinopathy. American Journal of Sports Medicine, 43(7), 1704-1711. DOI: 10.1177/0363546515584760
Sivrika A.P., Papadamou E., Kypraios G. et al. (2023). Comparability of the effectiveness of different types of exercise in the treatment of Achilles tendinopathy: a systematic review. Healthcare, 11(16), 2268. DOI: 10.3390/healthcare11162268
