You Reached for the Seatbelt and Your Arm Stopped

Body · The Reading Room, Vol. 50 · 10 min read · Why the old “it clears up in a year or two” advice has been formally challenged

Based on the published research of
Jocelyn R. Wittstein, MD — orthopedic surgeon, Duke Health, lead researcher connecting estrogen loss to frozen shoulder
Faculty page, Duke Department of Orthopaedic Surgery · Duke Health, hormone therapy and shoulder pain · ResearchGate

Frozen shoulder in menopause peaks between 40 and 60, and the idea that it clears up on its own is now contested. What helps, and the two blood tests to ask for.

You reached back for the seatbelt and your arm stopped. Not painfully, the first time. It just did not go where you sent it, and you had to turn your whole body to get the belt across.

Then it was the bra clasp. Then the coat sleeve, which you now put on one particular arm first without deciding to. Then the night you woke at two because you had rolled onto that side.

Somewhere in there you decided you had slept on it wrong. That was four months ago.

The short answer: this pattern has a name, adhesive capsulitis, and it is not a muscle problem. The joint capsule itself thickens and contracts. It affects 2% to 5% of people, and most of those diagnosed are women between 40 and 60.¹ The part worth knowing before you wait any longer: the old reassurance that it resolves on its own within a year or two has been formally challenged, and the best available evidence says most of the improvement happens early, not late.¹

The skim version

• The clinical term is adhesive capsulitis. Say that word and the appointment changes.
• The giveaway is external rotation. Elbow at your side, forearm turning outward. It goes first, and it is stuck whether you move it or someone else does.²
• Most patients diagnosed are women aged 40 to 60, and the mean age of onset is around 55.¹²
• Three randomised trials with long-term data found most improvement in pain and movement happens early in the course, not in the final phase.¹ Waiting spends the useful window.
• It is strongly linked to diabetes and thyroid disease. Around 30% of people with adhesive capsulitis have diabetes.¹ Two blood tests belong in that appointment.
• The menopause connection is real but the HRT evidence is thinner than the headlines. The numbers are below.

In this article: Is it frozen shoulder · Why now · What else it could be · Will it pass · When to escalate · What to say · The one-minute check · Questions

How do I know if it is frozen shoulder and not a pulled muscle?

By whether someone else can move it. With a strained muscle or an irritated tendon, you cannot lift the arm yourself but another person can move it for you. With adhesive capsulitis the shoulder is stuck both ways. The first movement lost is external rotation: elbow tucked at your side, forearm turning outward. That is the cardinal finding.¹

The reason is anatomical rather than muscular. The shoulder sits inside a capsule, a sleeve of connective tissue around the joint. In adhesive capsulitis that sleeve becomes inflamed and then fibrotic. The coracohumeral ligament thickens, the capsule contracts, and the space the joint moves in physically shrinks.² Nothing is torn. The container got smaller.

Which is why the exercises you found for shoulder impingement did nothing, and why the four months of assuming you slept on it wrong were four months of a condition that was quietly progressing.

What else could this be, and how do I tell?

Why did this show up around menopause?

Because oestrogen is not only a reproductive hormone. It maintains connective tissue integrity, reduces inflammation, and supports bone and muscle.³ The shoulder capsule is connective tissue. The decade when oestrogen falls and the decade when adhesive capsulitis peaks are the same decade, and researchers have stopped treating that as coincidence.

Jocelyn Wittstein, the orthopaedic surgeon who has led much of this work, put it plainly: “A disease that has such a strong predilection for women is unlikely to be truly idiopathic.”⁴

She is also an author on the 2024 review that named the musculoskeletal syndrome of menopause, the wider cluster of aching hands, hips and shoulders that arrives in the same years and that so often comes back with a normal scan. If your shoulder is not the only joint that has changed, those two things are probably one thing.

“A disease that has such a strong predilection for women is unlikely to be truly idiopathic.” — Jocelyn Wittstein, MD

The claim you have already read, and what the study actually found

You will have seen that women on hormone therapy are around half as likely to develop frozen shoulder. That comes from a real study, and it is worth knowing precisely what it says.

So the honest position is this. The association is real enough that a major medical centre thinks it should drive further investigation. It is not established enough to treat as proven, and it says nothing at all about reversing a frozen shoulder you already have.

The practical instruction, so this does not sit in your head as one more unresolved thing: do not start hormone therapy for your shoulder. If you are already weighing it for other symptoms, add the shoulder to the list you bring, and go in knowing what it does and does not treat. Then treat the shoulder on its own terms, starting this month.

If the reason you have not booked the appointment is that it did not seem bad enough to bother anyone with, the Quiet Audit is ten minutes on everything else you have filed under that. Start here.

Will frozen shoulder just go away on its own?

Sometimes, but not reliably, and the timeline you were probably given is under revision. The traditional picture of full resolution without treatment over one to two years has, in the words of the review in American Family Physician, been challenged by evidence of functional limitation persisting for years.¹ Roughly 10% to 20% are left with lasting stiffness even in the more optimistic estimates.²

Here is the finding that should change what you do this week rather than next spring. Three randomised controlled trials with long-term data showed that most improvement in pain and range of motion occurs early, not late.¹

Read that alongside the standard advice to wait it out and the problem becomes obvious. The window in which treatment does the most work is the window you are being asked to sit through.

It is also true, and I would rather say it than have you find it elsewhere, that the single most effective treatment is genuinely uncertain. Specialists disagree.¹ That is not a reason to do nothing. It is a reason to start the clock deliberately instead of by default, because the one option with clear evidence against it is another six months of hoping.

So: write down the date you first noticed the movement was gone. Then treat six to twelve weeks of treatment without meaningful improvement as the point at which you ask for an orthopaedic referral, because that is the threshold the guidance itself uses.¹

When does this need urgent attention rather than a routine appointment?

What should I actually say at the appointment?

Two sentences. The first names the condition and the specific movement, which is what separates this from general shoulder pain. The second asks for the two tests that the guidance says to consider, and which are easy to leave out when the complaint sounds musculoskeletal.

That second sentence is not a formality, and it is the part of this article most likely to matter to somebody. A 2016 meta-analysis found people with diabetes were five times more likely to have adhesive capsulitis, and estimated that 30% of people with adhesive capsulitis have diabetes (95% CI, 24% to 37%). A 2017 case-control study found hypothyroidism in 27.2% of the frozen shoulder group against 10.7% of controls.¹

The guidance is explicit that a fasting glucose or HbA1c, or a thyroid-stimulating hormone measurement, may be considered in patients with adhesive capsulitis.¹ Occasionally the shoulder is what finds the diabetes. It is a cheap pair of tests to have asked for and an expensive pair to have skipped.

The advice that quietly makes it worse

Rest it until it settles. This is the instinct, it is correct for an injury, and it is close to the worst available option here. Prolonged immobilisation is itself a recognised cause of adhesive capsulitis, which is why it develops so often after a period in a sling.² A shoulder that is protected because it hurts gets stiffer, and the stiffness is the disease.

The opposite error is worth naming too. Forcing the range in the early painful phase does not speed anything up. What has evidence behind it is unglamorous: physiotherapy and a corticosteroid injection combined produce greater improvement than physiotherapy alone.¹ Not one or the other, which is usually how the choice gets offered.

And the third thing, which is not medical. Waiting until it is bad enough to justify the appointment. By the time a shoulder is bad enough to feel justified, you are typically well past the phase where the intervention does the most good.

The one-minute check, and the date to write down

Stand up. Tuck both elbows against your sides and bend them to ninety degrees, forearms pointing straight ahead. Now turn both forearms outward, as if opening a book, keeping the elbows pinned to your ribs.

Watch the two sides. If one stops noticeably short of the other, that is loss of external rotation, and it is the single most useful observation you can hand a clinician. Have someone gently try to push that forearm a little further. If it will not go for them either, you have the passive restriction as well.

Then write two things on your phone: the date you first noticed it, and roughly how far short that side stops. The date is what converts vague waiting into a six-to-twelve-week decision point you can actually act on.

Questions women actually ask about this

Will it show up on a scan?

Not necessarily, and it does not need to. Adhesive capsulitis is a clinical diagnosis made from the examination.¹ An X-ray may be done to rule out other things, and MRI can show a thickened coracohumeral ligament with high specificity, but imaging is meant to exclude other pathology rather than to confirm this one.¹² A normal scan does not mean nothing is wrong with your shoulder.

Why is it so much worse at night?

Night pain is characteristic of the early phase and is often what finally gets it taken seriously.² It is worth knowing that shoulder pain and the kind of waking that is already happening in these years compound each other. Treating the shoulder is also a sleep intervention.

How long does each phase actually last?

The painful freezing phase typically runs two to nine months. The frozen phase, when the pain eases but the stiffness dominates, runs four to twelve months. Then a thawing phase.² The wide ranges are honest, not evasive, and they are why a fixed timeline should not be the basis of your plan.

I have diabetes. Does that change things?

Yes, and it is worth knowing early. People with diabetes or thyroid dysfunction tend to have a longer and more severe course and a less favourable outlook, and longer diabetes duration is associated with poorer outcomes.² That argues for getting seen sooner rather than later, not for resignation.

Does this end in surgery?

For most people, no. Referral to an orthopaedic surgeon is considered where there has been minimal improvement after six to twelve weeks of non-surgical treatment.¹ Manipulation under anaesthesia and arthroscopic capsular release both exist and generally do well, though manipulation carries real risks including fracture and tendon rupture, which is a reasonable thing to ask about directly.²

Informational, not medical advice. Fever, night sweats or unexplained weight loss alongside shoulder pain, or new weakness or numbness in the arm or hand, point somewhere other than adhesive capsulitis and need your own clinician promptly rather than at your convenience.

Where this evidence thins out

The hormone-therapy-and-shoulder finding is the single weakest link in this article’s evidence, and it is worth restating plainly: Duke’s own researchers describe the difference as not statistically significant, likely due to sample size.³ The connective-tissue mechanism connecting estrogen loss to capsule thickening is biologically plausible and consistent with the wider musculoskeletal syndrome of menopause literature, but a definitive, adequately powered trial establishing hormone therapy as protective against frozen shoulder does not yet exist. The diabetes and thyroid associations, by contrast, rest on larger, more consistent evidence and should be treated with more confidence.¹

References

1. Ramirez, J. (2019). Adhesive capsulitis: diagnosis and management. American Family Physician, 99(5), 297–300.
2. Li, D., St Angelo, J. M., & Taqi, M. (2025). Adhesive capsulitis (frozen shoulder). In StatPearls. StatPearls Publishing.
3. Duke Health. (2022). Hormone therapy appears to reduce risk of shoulder pain in older women. Duke University Health System.
4. The North American Menopause Society. (2022). Hormone therapy may help prevent shoulder pain and loss of motion in menopausal women.

Written by Nora Whitfield for Blue Leaf Journal. Updated: September 6, 2026.

You have almost certainly already adapted. You dress in a particular order, you reach with the other arm, you have stopped putting anything you use on the top shelf. Adaptation is what competent women do, and it is also what keeps a treatable thing off the list for another season.

A shoulder capsule is connective tissue, and connective tissue runs on estrogen exactly the way brain tissue does. If a lost word has been keeping you up as often as the shoulder has, the mechanism is worth knowing — it is not the decline it feels like.

The seatbelt still catches, most mornings. That is information, not a life sentence, and the window that matters most is the one you are in right now.