TL;DR: Frozen shoulder (adhesive capsulitis) is not a rotator cuff problem and it is not something you have simply let seize up. It runs a recognisable course through three stages — a painful freezing phase lasting roughly 2 to 9 months, a stiff frozen phase of about 4 to 12 months, and a thawing phase — and it is usually self-limiting, resolving in one to three years. The tell on examination is lost passive external rotation: someone else cannot move the shoulder either. Diabetes and thyroid disease both make it more likely and the course longer.
What Is Actually Going On
The shoulder sits inside a capsule of connective tissue. In adhesive capsulitis that capsule becomes inflamed and then progressively thickens and contracts, physically reducing the space the joint has to move in. That is a different mechanism from a torn tendon, an impinged structure, or a stiff shoulder from disuse — and it is why the treatment logic differs.
The distinguishing examination finding is the loss of passive range, especially external rotation. If you cannot lift your arm but a clinician can move it for you, the limit is pain, weakness or a tendon problem. If the clinician cannot move it either, the capsule is the constraint — and external rotation is typically the most affected direction.
It also has a risk profile. People with diabetes are markedly more likely to develop it — one meta-analysis put the odds at around five times the control group — and both diabetes and thyroid dysfunction are associated with a longer, more severe course and a less favourable prognosis. If you have either and your shoulder is stiffening, that is worth flagging early.
The Three Stages, and What Each One Needs
Stage 1 — freezing (painful), roughly 2 to 9 months
Diffuse, disabling shoulder pain that is characteristically worse at night, alongside range of motion that is quietly disappearing. This is the stage where most people are misdiagnosed, because the pain dominates and the stiffness has not yet become obvious.
What treatment should be doing: controlling pain, protecting sleep, and maintaining whatever range you have — gently. This is the stage where aggressive stretching backfires. Pushing hard into an irritable, inflamed capsule tends to increase pain and, with it, the guarding that accelerates the loss.
Stage 2 — frozen (stiff), roughly 4 to 12 months
Progressive limitation in every plane of movement, but the pain gradually settles. This is the strange stage: your shoulder is at its most useless and at its least painful. Reaching behind you, fastening a seatbelt, or putting on a jacket becomes the daily reminder.
What treatment should be doing: now the balance shifts toward restoring motion. Because the tissue is less irritable, it tolerates — and needs — more stretching and mobilisation than it did in stage 1. This is where structured physical therapy earns its place, and where a generic exercise handout does not.
Stage 3 — thawing, gradual
Pain reduces further and motion gradually returns. Recovery is rarely a straight line and people frequently have some residual end-range limitation that is unnoticeable in daily life.
What treatment should be doing: converting recovered range into usable strength. A shoulder that can now reach overhead but has spent a year not doing so needs loading, or the function lags well behind the range.
Why the Stage Matters More Than the Diagnosis
This is the practical point of the whole article. The same intervention can be right in one stage and counterproductive in another. Vigorous stretching in the freezing phase commonly makes things worse. Cautious, protective handling in the frozen phase wastes the window where motion is most recoverable.
So “I did physical therapy for my frozen shoulder and it did not help” is often really “I did stage-2 therapy during stage 1.” A programme that is not periodically reassessed as the shoulder moves through the stages is not really a programme for this condition.
What This Is Not
The confusion that costs the most time is with rotator cuff problems. A cuff issue typically shows painful, weak active movement with passive range preserved; adhesive capsulitis shows both restricted. They are treated differently, and a shoulder programme built for the wrong one will not deliver.
Stiffness after a period of immobilisation — a sling, a cast, a surgery — can look similar but generally recovers faster and responds differently. Referred pain from the neck can also masquerade as shoulder pain; that one is usually distinguishable on examination and is covered in our neck and shoulder guide.
Being Evaluated
An examination that distinguishes active from passive range takes minutes and reframes the next year. At our shoulder clinic in Addison, TX, Rowena Calleja, PT assesses which stage you are in and builds the programme for that stage, reassessing as it changes; Charles Ortega, PA-C evaluates the medical side, including the diabetes and thyroid associations that are worth screening for when a frozen shoulder appears out of nowhere.
Call (214) 233-3094 or request an appointment.
How long does frozen shoulder last?
It is usually self-limiting and typically resolves within one to three years, though the overall timeline can extend beyond that. The painful freezing phase generally lasts about 2 to 9 months and the stiff frozen phase about 4 to 12 months, followed by a gradual thawing phase. Diabetes and thyroid dysfunction are both associated with a longer and more severe course.
How do I know if it's frozen shoulder or a rotator cuff tear?
The distinguishing feature is passive range of motion. With a rotator cuff problem, movement is painful or weak when you do it yourself but a clinician can usually still move the shoulder through range. With adhesive capsulitis, the capsule itself is the restriction, so passive movement is limited too — external rotation most of all. An examination settles it quickly.
Should I stretch a frozen shoulder?
It depends entirely on the stage. During the painful freezing phase, aggressive stretching typically increases pain and guarding; the goal there is pain control and maintaining the range you have. Once the shoulder enters the less irritable frozen phase, more assertive stretching and mobilisation become appropriate and important. This is why the programme needs reassessing rather than repeating.
Does diabetes cause frozen shoulder?
There is a strong association: a meta-analysis found people with diabetes were around five times more likely than controls to have adhesive capsulitis, and both diabetes and thyroid dysfunction predict a longer, more severe course. If you have either and a shoulder that is stiffening, mention it early — it changes what your clinician expects and how they plan.
