Frozen shoulder — the honest timeline, and how to change it.

Root Physical Therapy

Clinical Note — No. 09

Shoulder

Frozen shoulder — the honest timeline, and how to change it.

A shoulder that hurts for months, then stiffens for months more. You'll be told it "resolves on its own." That's mostly true, eventually — but the timeline runs one to three years, a real minority never fully recover, and what you do in each phase changes the course.

Georgetown, Seattle10 min readEvidence-based

Written by the Doctors of Physical Therapy at Root Physical Therapy — on-site inside Root Strength, Georgetown. Frozen shoulder is one of the most misunderstood conditions we treat — and one where the wrong advice at the wrong phase genuinely costs people months. If your shoulder is stiffening, book an assessment — no referral needed in Washington.

Frozen shoulder tends to arrive with two pieces of advice, and they contradict each other. One camp says don't worry, it's self-limiting — it thaws on its own, just wait. The other proposes a procedure: manipulation under anesthesia, or surgery to release the capsule.

Both contain some truth, and both, offered alone, fail the person hearing them. The honest version is this: frozen shoulder follows a long, phased course; the "wait" can be one to three years; a meaningful minority never fully recover without doing anything about it; and the largest trial ever run on the condition found that the least invasive path kept pace with surgery. Here's the whole picture.

One

What's actually happening in there

Your shoulder joint is wrapped in a capsule — a soft-tissue envelope that normally has generous slack, which is what lets the shoulder be the most mobile joint you own. In frozen shoulder (the clinical name is adhesive capsulitis), that capsule becomes inflamed, then progressively fibrotic: it thickens, scars, and contracts. The envelope shrinks around the joint.

That's why the signature of this condition is different from most shoulder problems: the motion is lost even when someone else moves your arm for you. With a rotator cuff issue, the joint itself usually still travels — it's painful or weak, but passively movable. A frozen shoulder is mechanically blocked, and the tell clinicians look for first is loss of passive external rotation — the reaching-back motion, as when you reach for a seatbelt or the back seat. No scan needed to make the diagnosis; the exam is the diagnosis.

Who Gets It

Figure 01

2–5%

of the general population — most often ages 40–60, women more than men

10–30%

of people with diabetes — the strongest risk factor, with longer, more stubborn courses

~20%

go on to develop it in the other shoulder — worth knowing, not fearing

It can start on its own (primary) or after a trigger — surgery, injury, or a period of immobilization (secondary). Thyroid conditions also raise the risk. Often, though, it arrives without any story at all — which is part of what makes it so unnerving.

Two

The three phases — with real numbers

Frozen shoulder moves through overlapping phases, and almost everything about managing it well comes down to knowing which one you're in.

The Course, Phase by Phase

Figure 02

I

Freezing — the painful phase

Roughly 2–9 months

Pain dominates: constant ache, sharp with movement, and — the complaint that brings most people in — it wrecks sleep, especially lying on that side. Stiffness is building underneath, but pain is the loudest signal. This is the inflammatory phase, and it's the window where anti-inflammatory treatment does its best work.

II

Frozen — the stiff phase

Roughly 4–14 months

Pain eases; stiffness takes over. Reaching overhead, behind your back, into a jacket sleeve — all blocked. This is capsular contracture, not weakness, and no amount of forcing changes fibrotic tissue faster than it's willing to change. Function is the casualty here, and working within the range you have keeps the rest of the shoulder healthy.

III

Thawing — the recovery phase

Roughly 5–24 months

Motion gradually returns. This is where the payoff for staying engaged comes: a shoulder that was kept moving and strong through the middle phase recovers into a usable shoulder faster — and one that spent a year in a sling of caution has two problems to fix instead of one.

Total course: commonly 1–3 years, with wide individual variation. Phases overlap rather than switching cleanly, and diabetes tends to stretch every number above. The point of the ranges isn't precision — it's calibration. This condition is measured in months, and knowing that upfront is half the battle.

The treatment doesn't fight the phases. It matches them.

Three

The "it goes away on its own" problem

Where the reassurance overreaches — stated plainly

Frozen shoulder is commonly, but not entirely correctly, called self-limiting. Most people do improve substantially with time. But long-term follow-up studies complicate the comforting version: up to 40% of patients report persistent symptoms beyond three years, and roughly 7–15% are left with some degree of lasting functional loss. "It usually thaws" and "you can safely do nothing" are not the same sentence.

The practical conclusion isn't panic — it's that the passive-waiting plan has a real failure rate, and the things that improve the odds (pain control in the right window, movement matched to phase, strength maintained throughout) are low-risk and worth doing.

Four

What the evidence supports

First: in the painful phase, a corticosteroid injection is the biggest single lever. Reviews and network meta-analyses consistently find that an intra-articular steroid injection, given early — while inflammation is still the driver — produces the fastest improvement in pain and function, especially when it's combined with an exercise program rather than used alone. It doesn't shorten the biology to zero, but it can buy back months of sleep and function during the worst stretch. That's a physician-delivered treatment we coordinate around, not a PT one — and the timing is the point: its advantage fades once the shoulder moves from inflamed to fibrotic.

Second: the invasive options didn't beat the conservative one. The UK FROST trial — the largest randomized trial ever conducted on frozen shoulder, across 35 hospitals — compared manipulation under anesthesia, arthroscopic capsular release, and early structured physiotherapy with a steroid injection.

UK FROST — Three Treatments, One Verdict

Lancet · 2020

I

Manipulation under anesthesia

The shoulder is moved through range while you're under, tearing the contracted capsule loose. Outcomes at 12 months: no better than physiotherapy on patient-reported pain and function.

II

Arthroscopic capsular release

Keyhole surgery to cut the contracted capsule. At most a marginal benefit over physiotherapy — below the threshold of clinical importance — and it carried the most serious adverse events of the three arms.

III

Early structured physiotherapy + steroid injection

Twelve weeks of structured, progressive PT with an injection. Kept pace with both procedures — without an anesthetic, an operating room, or the recovery from either.

None of the three was superior at 12 months. Procedures still have a place — a small share of shoulders stay refractory and get referred on, and that's a legitimate, informed decision. But the trial's message for the newly diagnosed is clear: the conservative path is not the consolation prize. It's the evidence-backed starting line.

Five

The protocol — what PT actually does

Here's the course of care at Root PT. Five pillars — sequenced, as always, by what the assessment finds, and re-sequenced as the shoulder changes phase under us.

The Five Pillars

Figure 03

I

Confirm the diagnosis, stage the phase

Passive range testing — external rotation above all — separates a true frozen shoulder from the rotator cuff problems it gets mistaken for, and the pain-versus-stiffness balance tells us the phase. The phase sets the plan. Getting this wrong is how people end up aggressively stretching an inflamed shoulder or babying a thawing one.

II

Freezing phase — calm it, protect sleep, coordinate the injection

Pain management is the treatment here: gentle motion in comfortable ranges, positioning strategies so you can actually sleep, activity modification that keeps you living — and coordination with your physician on a corticosteroid injection while the window for it is open. What we don't do: force range. Aggressive stretching into an inflamed capsule reliably makes this phase louder and longer.

III

Frozen phase — mobility work with a realistic contract

Progressive, sustained, tolerable stretching and joint mobilization to maintain and gradually expand range — working to the edge of stiffness, not through pain. The honest contract: we are keeping the joint as mobile as the biology allows and setting up the thaw, not bullying fibrotic tissue into surrender.

IV

Keep the whole system strong throughout

The capsule is frozen; the muscles don't have to be. Scapular control, rotator cuff work within available range, and — critically — the rest of your training continues. Legs, trunk, the other arm, your conditioning. A year is too long to put a whole body on hold for one joint, and the on-site gym makes not-stopping practical.

V

Thawing phase — chase the returning range

As motion comes back, we load it: progressive strengthening through the new range, then a graded return to overhead work, lifting, sport, and everything the shoulder stopped doing. Range you don't strengthen is range you'll guard — this phase is what turns a thawed shoulder back into a trusted one.

Expect this to be a months-long relationship, not a six-visit package — with visit frequency that flexes by phase: closer support early and late, lighter-touch check-ins through the long middle.

The reframe that helps most

The total duration is partly out of your hands. What the months are like is not. Whether you sleep, whether you keep training, whether the rest of your shoulder stays strong, whether the thaw returns you a shoulder you trust — those are the parts treatment actually moves, and over a course this long, they're most of what matters.

If you have diabetes

Diabetes is the strongest risk factor for frozen shoulder — and courses tend to run longer and more stubborn, with a higher chance of the other shoulder joining later. That's a reason to act earlier, not to despair: the same phase-matched approach applies, the injection conversation with your physician is worth having promptly, and a stiffening shoulder is also a data point worth mentioning at your next diabetes review.

See a physician promptly if you have

Frozen shoulder is a clinical diagnosis with a known course — but these point somewhere else and need medical evaluation first:

Shoulder stiffness after significant trauma — a fall or accident

Visible deformity, or an arm you cannot move at all

A hot, red, swollen joint — especially with fever

Numbness, tingling, or weakness spreading into the arm or hand

A history of cancer with new, unexplained shoulder pain

Severe night pain with unexplained weight loss

Six

The bottom line

Frozen shoulder is a contracted joint capsule moving through a long, phased course — painful first, stiff second, recovering third. It usually improves, but "usually" hides a real minority who don't, which is why passive waiting isn't a plan. The evidence-backed one: a steroid injection early if the painful phase warrants it, physical therapy matched to each phase, strength maintained throughout — and procedures held in reserve, knowing the largest trial ever run found they couldn't beat that approach anyway.

The Short Version

• Frozen shoulder is a contracted joint capsule — the tell is losing motion even when someone else moves your arm, external rotation first.
• Three phases: freezing (pain, ~2–9 months), frozen (stiffness, ~4–14 months), thawing (recovery, ~5–24 months). Total: commonly 1–3 years.
• "Self-limiting" oversells it. Up to 40% have symptoms beyond 3 years; 7–15% keep some lasting loss. Waiting passively has a failure rate.
• Early steroid injection is the biggest painful-phase lever — best combined with exercise, and time-sensitive.
• UK FROST, the largest trial ever run: surgery and manipulation were no better than structured PT plus injection at 12 months — and carried the complications.
• Match treatment to phase. Calm the freezing shoulder; mobilize the frozen one; load the thawing one. Forcing an inflamed capsule backfires.
• Diabetes means longer, more stubborn courses — a reason to start earlier, not wait longer.

This article provides general educational information and is not medical advice or a substitute for individual assessment. Shoulder stiffness has several causes, and any red-flag symptom warrants prompt medical evaluation. Decisions about injections or procedures should be made with your physician.

Start by finding out what phase you're in.

The plan depends on the phase, and the phase takes one exam to determine. The Doctors of Physical Therapy at Root PT see patients on-site at Root Strength, Georgetown. No referral required in Washington, most major insurance accepted — and we verify your benefits before scheduling, so there are no surprises.

Book an Assessment

Keep Reading — Root Physical Therapy

Hurt vs Harm: Why Pain Doesn't Always Mean Damage Do You Actually Need an MRI? What Actually Happens at Your First PT Appointment Shoulder Pain — Why It Happens and How to Fix It (Root Strength)

Sources

1. Rangan A, Brealey SD, Keding A, et al. Management of adults with primary frozen shoulder in secondary care (UK FROST): a multicentre, pragmatic, three-arm, superiority randomised clinical trial. The Lancet. 2020;396(10256):977–989.
2. Hand C, Clipsham K, Rees JL, Carr AJ. Long-term outcome of frozen shoulder. Journal of Shoulder and Elbow Surgery. 2008;17(2):231–236.
3. Challoumas D, Biddle M, McLean M, Millar NL. Comparison of treatments for frozen shoulder: a systematic review and network meta-analysis. JAMA Network Open. 2020;3(12):e2029581.
4. Zreik NH, Malik RA, Charalambous CP. Adhesive capsulitis of the shoulder and diabetes: a meta-analysis of prevalence. Muscles, Ligaments and Tendons Journal. 2016;6(1):26–34.
5. Ramirez J. Adhesive capsulitis: diagnosis and management. American Family Physician. 2019;99(5):297–300.
6. Kelley MJ, Shaffer MA, Kuhn JE, et al. Shoulder pain and mobility deficits: adhesive capsulitis — clinical practice guidelines. Journal of Orthopaedic & Sports Physical Therapy. 2013;43(5):A1–A31.
7. Lewis J. Frozen shoulder contracture syndrome — aetiology, diagnosis and management. Manual Therapy. 2015;20(1):2–9.
8. St. Angelo JM, Taqi M, Fabiano SE. Adhesive capsulitis. In: StatPearls. StatPearls Publishing; updated 2025.