Frozen Shoulder in Menopause: Why It Peaks After 45

Frozen shoulder is a condition where the connective tissue capsule surrounding the shoulder joint becomes inflamed, thickened, and contracted — physically shrinking around the joint until the arm cannot move through its normal range. The clinical name, adhesive capsulitis, describes what is happening: the capsule adheres and stiffens.

The defining feature, and the one that distinguishes it from almost everything else that hurts a shoulder, is loss of passive range of motion. If someone else lifts your arm for you and it still will not go, that is capsular restriction. With a rotator cuff tear or tendonitis, you often cannot lift the arm yourself but someone else can move it further. This distinction is the core of the diagnosis and is usually made on examination alone, without imaging.

External rotation is typically the most restricted direction. The classic real-world tests are reaching behind you to fasten a bra, reaching into a back pocket, reaching up to a high shelf, or putting on a coat. Sleeping on the affected side becomes impossible for most people, and night pain is often what finally prompts a medical visit.

The condition affects roughly 2 to 5% of the general population, but that figure disguises how concentrated it is. It clusters in women — most estimates put it at two to four times more common than in men — and peaks between ages 40 and 60. Both shoulders are affected in a meaningful minority of cases, though usually not at the same time.

What makes it particularly frustrating is that it often starts without any injury at all. Many women describe simply noticing one day that they could not reach behind them, with no fall, no strain, nothing to point to.

The age distribution is the clue that has driven a decade of research interest: the peak incidence window for frozen shoulder overlaps almost exactly with the menopause transition, and the female predominance is hard to explain by mechanical factors alone.

The biological rationale is reasonable. Estrogen receptors have been identified in tendon, ligament, and joint capsule tissue. Estrogen influences collagen synthesis and turnover, and collagen is the structural protein that gives connective tissue its balance of strength and flexibility. It also modulates inflammatory signaling and fibroblast activity — fibroblasts being the cells that lay down the fibrous tissue that contracts the capsule in frozen shoulder.

When estrogen falls, the working model is that collagen becomes more disorganized and less elastic, inflammatory signaling in connective tissue rises, and the tissue's capacity to remodel after minor insults declines. That is a plausible setup for a capsule that inflames and then fibroses rather than settling.

This fits a broader midlife pattern that many women recognize: frozen shoulder does not usually arrive alone. It often appears alongside generalized joint aches, tendinopathies (particularly at the hip, elbow, and Achilles), plantar fasciitis, and trigger finger. Some clinicians group these under the umbrella of musculoskeletal syndrome of menopause, a term that has gained traction precisely because so many women present with several of these at once in their late forties.

An important honesty note: the estrogen link is well-supported mechanistically and strongly suggested epidemiologically, but there is no large randomized trial showing that hormone therapy prevents or treats frozen shoulder. Observational studies have produced mixed signals. It is a compelling explanation for why this happens now, not a prescription.

If your joint symptoms are more diffuse than one shoulder, it is worth reading how estrogen loss affects tissue more broadly in [menopause skin and collagen loss after 45](/blog/menopause-skin-collagen-loss-what-happens-after-45), which describes the same collagen mechanism in a different tissue.

Frozen shoulder is one of the few musculoskeletal conditions with a genuinely predictable arc, which is useful — knowing which phase you are in tells you what treatment will and will not help.

Phase 1: Freezing (roughly 2 to 9 months). This is the painful, inflammatory phase. Pain comes first and stiffness follows. Night pain is at its worst here, and range of motion is progressively lost. This is the phase where aggressive stretching hurts and can prolong things, and it is also the phase where anti-inflammatory treatment does the most good. Most women push through this phase assuming it will resolve on its own — this is the window where getting seen matters most.

Phase 2: Frozen (roughly 4 to 12 months). Pain settles substantially but stiffness is at maximum. The capsule has fibrosed. This is the phase where structured mobility work earns its keep, because the tissue now needs remodeling rather than calming. Daily function is hardest here even though it hurts less.

Phase 3: Thawing (roughly 5 to 24 months). Range gradually returns. Recovery is usually good but not always total — a meaningful minority retain some restriction in external rotation long-term, though most find it functionally unimportant.

Total duration is commonly quoted as 12 to 30 months. The old teaching that it is fully self-limiting and always resolves completely has been challenged by longer follow-up studies, which is part of why the wait-and-see approach has fallen out of favor.

One thing worth flagging: the recovery timeline here is longer than most midlife women expect, and it interacts with everything else. If you are already managing fatigue, disrupted sleep, and reduced exercise capacity, a two-year shoulder problem is a significant load. Our piece on [why menopause requires more recovery between sessions](/blog/menopause-exercise-recovery-why-you-need-more-rest-days) is relevant to how you plan training around it.

The best-supported approach is a corticosteroid injection during the painful freezing phase, combined with physiotherapy — and the evidence for this combination is stronger than for either alone. Multiple trials and systematic reviews have found that intra-articular corticosteroid injection produces meaningful short- and medium-term improvement in pain and range of motion, with the largest benefit when given early rather than after the capsule has fully fibrosed.

Physiotherapy is the backbone, but it has to be phase-appropriate. In the freezing phase, aggressive end-range stretching provokes the inflammatory process and can genuinely make things worse — a common and demoralizing experience for women told to "work through it." In the frozen phase, the tissue needs sustained, progressive loading and end-range work. Getting this sequence right is the difference between physiotherapy that helps and physiotherapy that hurts.

Hydrodilatation — injecting fluid under pressure to distend the capsule, usually under ultrasound guidance — is a reasonable next step when conservative care has stalled after several months. It is less invasive than surgery and has a solid evidence base.

Manipulation under anaesthesia or arthroscopic capsular release are reserved for cases that have not improved after six to twelve months of appropriate conservative care. They work, but they are not first-line.

What has weak evidence: oral NSAIDs alone (helpful for pain, not for the course), and simply waiting.

One genuinely important step: get screened for diabetes and thyroid disease. Diabetes is the strongest known medical association with frozen shoulder — prevalence in people with diabetes is several times that of the general population — and thyroid disorders are also over-represented. Both are common in midlife and both are easy to check. Thyroid symptoms overlap heavily with perimenopause, which we untangle in [thyroid or perimenopause: how to tell the difference](/blog/thyroid-or-perimenopause-overlapping-symptoms-what-to-test).

If you are on a GLP-1, there is a relevant upside: better glycemic control is worth pursuing for shoulder tissue as well as everything else. And for anyone weighing hormone therapy for other reasons, joint aches are one of the symptoms many women report improving — see [HRT side effects in the first three months](/blog/hrt-side-effects-first-3-months-whats-normal) for what starting it realistically looks like.

There is no proven prevention protocol, but there are sensible risk-reduction steps — and they matter, because a meaningful minority of people develop frozen shoulder in the opposite shoulder within a few years.

Control blood sugar. This is the strongest lever available, given how tightly diabetes is associated with frozen shoulder. If you have prediabetes or type 2 diabetes, glycemic control is shoulder care.

Keep the shoulder moving, especially after any period of immobilization. Frozen shoulder frequently follows a period of not moving the arm — after a fracture, surgery, mastectomy, or even a prolonged painful episode. If you are ever in a sling or recovering from upper-body surgery, ask specifically about early, safe range-of-motion work.

Maintain overall connective tissue load. Resistance training preserves tendon and ligament quality along with muscle and bone. This is the same argument that underpins [protein intake in menopause](/blog/protein-in-menopause-how-much-you-actually-need) and impact work for bone — connective tissue responds to being used.

Address symptoms early on the other side. If your second shoulder starts aching or losing external rotation, that is the moment to act, not to hope. Catching it in the freezing phase is the whole ballgame.

What about hormone therapy? There is a reasonable mechanistic case and many women anecdotally report improvement in joint and tendon symptoms after starting HRT, but the evidence specific to frozen shoulder is not there. It should not be started for this indication alone, though if you are already considering it for vasomotor symptoms or bone health, generalized musculoskeletal symptoms are a legitimate part of that conversation.

Finally: be persistent about being taken seriously. Midlife women with musculoskeletal complaints are frequently told it is age, stress, or overuse. Loss of passive external rotation is an objective finding, not a subjective complaint, and it takes about thirty seconds to check.