“Shoulder Impingement” Is a Description, Not a Diagnosis

Impingement was a coherent idea in 1972: a bony spur catching a tendon, correctable surgically. Fifty years of research has substantially undermined that model, and the term outlived it.

What the word describes

Pain in an arc of shoulder elevation, typically between about sixty and one hundred and twenty degrees, often with pain reaching overhead or behind.

That is a symptom pattern. It says where it hurts and in what range. It does not say why.

Calling it a diagnosis is the equivalent of diagnosing a cough. It is real, it is describable, and it does not tell you what to do.

Why the original model fell apart

The theory was that the acromion — the bony roof over the shoulder — catches the cuff tendon underneath, wearing it. Shave the bone, resolve the problem.

Then the trials arrived. Subacromial decompression surgery, compared against placebo surgery, produced no meaningful advantage. Both groups improved. The bone removal was not what helped.

Meanwhile, imaging studies found acromial shapes and cuff changes in enormous numbers of people with no shoulder pain at all.

Two findings, both robust, both pointing the same way: the structural story does not explain the symptoms.

What is usually going on instead

A rotator cuff tendon that is being asked to do more than it can currently tolerate. That is the short version and it covers most cases.

The load may have gone up — a new sport, a house painted, a training block. Or capacity may have gone down — a period of inactivity, illness, sleep deprivation, age-related change.

Either way the gap is the problem, and the gap can be closed from both sides.

Why 'subacromial pain syndrome' is the better term

Because it says where the pain is and admits what is not known, which is honest.

It also stops implying a mechanical catching that needs correcting, which is the belief that leads people to rest, avoid overhead work, and get weaker.

Language shapes behavior here more than most places. People told their tendon is being ground down by bone behave very differently from people told their tendon needs building up.

What the assessment should measure

Strength, tested against resistance in specific positions and compared with the other side. External rotation, internal rotation, abduction and the empty-can position.

How many repetitions you can manage before the shoulder gives up, which tells you about endurance rather than peak force and is often the more relevant number.

Range in every direction, active and passive, because a restriction of passive range changes the diagnosis entirely.

Scapular control, thoracic mobility, and the neck, because shoulder pain that originates in the neck is common and responds to nothing done at the shoulder.

What treatment actually looks like

Progressive loading of the rotator cuff, which is the intervention with the best evidence and the least glamour.

That means resistance work three to four times a week, at a load that produces some symptom during the exercise and settles within twenty-four hours. Discomfort during loading is acceptable and expected.

Twelve weeks is a realistic course. Improvement is usually noticeable within four.

Manual therapy and dry needling have a place in reducing symptoms enough that loading becomes possible. They are the on-ramp, not the road.

The rest instruction, and why it backfires

Tendons respond to load. Removing load reduces capacity, and reduced capacity widens the gap that caused the problem.

So a shoulder rested for six weeks typically feels better while resting and returns to the same pain on the first day back, plus six weeks of lost strength.

Relative rest — reducing the provocative volume while continuing to load in tolerable ranges — is a different instruction and a much better one.

Where injections fit

A subacromial corticosteroid injection reliably reduces pain in the short term. Over six months to a year the advantage over exercise largely disappears.

It is useful when pain is preventing any loading at all, as a way of opening a window for rehabilitation.

It is not useful as a standalone treatment, and repeated injections into a tendon region are worth avoiding.

Where surgery fits

Rarely, for this presentation. The placebo-controlled trials of decompression are about as clear as surgical evidence gets.

Full-thickness traumatic tears in younger people are a genuinely different situation with a genuine surgical case.

For the ordinary painful arc in a forty-five-year-old who paints ceilings on weekends, loading is the answer and it works.

The scapula, and how much it matters

Scapular movement changes in people with shoulder pain, which is well documented. Whether those changes cause the pain or result from it is much less clear.

That distinction is not academic. If the scapular pattern is a consequence of pain, drilling scapular exercises for months treats a symptom while the cuff stays weak.

The practical position: include scapular and upper back work because it is easy and it does no harm, and do not let it displace the cuff loading that the evidence actually supports.

Thoracic mobility belongs in the same category. Useful, quick, and not the main event.

Where the neck comes into it

A meaningful share of shoulder pain is referred from the cervical spine, and it presents convincingly as shoulder pain with a painful arc.

The clue is usually that shoulder tests are inconsistent, that resisted strength testing does not reproduce the symptom, and that neck movements do.

A shoulder assessment that skips the neck will treat that patient for twelve weeks without changing anything, which is a common and avoidable outcome.

It takes five minutes to screen and it changes the plan entirely when it is positive.

Why the diagnosis you are given changes your outcome

People told their shoulder is being mechanically damaged move less, avoid more, and report worse function at follow-up than people given a loading explanation for the same presentation.

That is not a reason to be dishonest about findings. It is a reason to be accurate about what they mean.

A tendon that is not currently strong enough is a fixable problem with a known method. A tendon being ground away by bone is a helpless situation. Only one of those descriptions matches the evidence, and it happens to be the useful one.

What to ask the person who gave you the label

What is the strength deficit, and compared with what? A number means an assessment happened.

What is the plan, and what will be measured in six weeks?

Is the neck involved? A shoulder examination that did not include the neck is incomplete.

And if imaging has been done: does the finding explain the symptom, given how common that finding is in people without pain?

What you can start this week

Isometric external rotation: elbow at the side, push out into a doorframe, hold thirty seconds, five times. Comfortable for most irritable shoulders and often reduces pain immediately.

Banded external and internal rotation, three sets of twelve, at a resistance that is genuinely hard by the last few.

Scapular rows and a wall slide for the upper back.

Keep using the arm. Modify what provokes it sharply, continue everything else.

Overhead athletes and overhead workers

Throwers, swimmers, tennis players, painters and electricians all present with this and all need the plan built around the demand rather than around the label.

Volume is usually the first thing to look at. A pitcher whose innings doubled, a swimmer who added a session, a painter who took a ceiling contract — the load went up faster than the tendon adapted.

The rehabilitation therefore has to end where the demand is: loaded overhead work, at speed where speed matters, before returning to full participation.

Stopping at pain-free daily activity leaves an athlete underprepared and the problem recurs on the second week back. That is not bad luck, it is an incomplete program.

A note on how long it takes

Tendons are slow. Four weeks to notice, twelve to consolidate, and longer if the shoulder has been painful for a year before you started.

The most common reason loading programs fail is that they stop at week five, when things feel better and before anything has been rebuilt.

The second most common is that the load was never hard enough to be a stimulus. Three sets of twelve with a light band is a warm-up, not a program.

A useful rule while you build: the shoulder may be sore during the work and for a few hours afterward, and it should be back to its usual baseline by the following morning. Soreness that is still elevated at twenty-four hours means the last session was too much, and the answer is to reduce the load a little rather than to stop altogether.

That single rule lets most people self-manage the progression between appointments, which is where the majority of the work happens anyway.