Frozen Shoulder and Diabetes: Why the Link Is So Strong

A shoulder that hurts at night and slowly stops moving is unpleasant for anyone. If you also have diabetes, the odds of it happening to you are much higher, the arc tends to run longer, and a few treatment decisions work differently than they would for someone else.

That last part gets missed often. People with diabetes are handed the same generic frozen shoulder advice as everyone else, including advice about injections that carries a consequence specific to them. The exercises are broadly the same. The timeline, the escalation decisions, and the injection conversation are where diabetes changes things.

This post covers the diabetes side specifically. If you want the general picture of what the three stages feel like and what a physiotherapist does in each one, the staged breakdown of frozen shoulder covers that ground in detail.

How much more likely is frozen shoulder with diabetes

The clearest figures come from a meta-analysis by Zreik, Malik and Charalambous in the Muscles, Ligaments and Tendons Journal, published in 2016. Pooling 13 studies, it put the prevalence of adhesive capsulitis among people with diabetes at 13.4% (95% CI 10.2 to 17.2%). Against controls, people with diabetes were 5 times more likely to have it (95% CI 3.2 to 7.7).

The reverse direction is worth knowing too. Across five studies, the same meta-analysis found diabetes in 30% of people presenting with adhesive capsulitis (95% CI 24 to 37%). Roughly one in three people arriving at a clinic with a frozen shoulder has diabetes, and a portion of those have not been diagnosed yet.

For comparison, a 2024 narrative review in the Journal of Clinical Medicine puts the overall prevalence of adhesive capsulitis in the general population at 2 to 5%, with most people aged 40 to 60 at diagnosis.

A few further details from that same review:

• Type 1 and type 2 both carry the risk. The 2016 meta-analysis found no significant difference between the two types, though the confidence intervals around each were wide.
• Insulin use versus oral medication showed no significant difference in how often adhesive capsulitis appeared.
• Duration of diabetes matters. Onset clustered in the chronic phase, at 32.3% for a diabetes duration of 1 to 5 years and 33.8% for 5 to 10 years.
• A newer systematic review puts the increase in risk at 3.69 times, which is lower than the 2016 figure and still a large effect.

If you have diabetes and a shoulder that has been stiffening for two months, this is the background probability your clinician should have in mind.

Why high blood sugar stiffens a joint capsule

The joint capsule is a sleeve of connective tissue around the shoulder, and it is mostly collagen. Frozen shoulder is that sleeve becoming inflamed, thickened, and contracted.

The 2024 review names advanced glycation end products as "the most significant influencing factor." These are compounds that build up when blood sugar stays high. They cause collagen cross-linking, which makes the capsule stiffer and more resistant to being broken down and remodelled. The review describes how their accumulation "accelerates the cross-linking of collagen and increases resistance to proteolysis, thereby promoting fibroblast proliferation."

Two other pathways sit alongside that one. High blood sugar drives chronic release of inflammatory messengers including tumour necrosis factor alpha and interleukin-6, which the review links to collagen and extracellular matrix building up into fibrosis. Increased fat cells release interleukin-13, which is thought to contribute to fibrosis in the synovium and connective tissue.

The practical version: the same tissue chemistry that makes a capsule contract is being fed continuously by high blood sugar. That is why blood sugar control belongs in the treatment plan alongside the exercises, and why the review states plainly that any treatment for adhesive capsulitis in diabetes "must be accompanied by proper blood sugar control." That side is managed by your family physician or endocrinologist, not by a physiotherapist.

Does diabetes make recovery slower

The honest answer is that it often does, and you should plan for it rather than be surprised by it.

The 2024 review reports that symptoms of adhesive capsulitis in people with diabetes "are often more severe, requiring prolonged conservative treatment, and may require surgical treatment more often than non-diabetic patients."

It goes further on the surgical end. Arthroscopic capsular release improves range of motion and pain in people with diabetes, and "the prognosis following ACR remains worse for diabetic patients, particularly in terms of internal rotation and forward flexion." Internal rotation is the movement behind your back. It is the one people notice when reaching a back pocket or doing up a bra strap.

Frozen shoulder is described in that review as often self-limiting over about 1 to 3 years. With diabetes, expect to sit toward the longer part of that range rather than the shorter part.

What that means for how you approach it:

• Start earlier rather than waiting it out. Time is the thing in shortest supply here.
• Expect the plan to be measured in months. If you want a sense of what a typical course of care looks like, how long physiotherapy usually takes sets out the general expectations.
• Judge progress against your own trend line, since comparing your pace to a friend without diabetes will be discouraging and will not tell you anything useful.
• Keep the metabolic side active. Blood sugar control is part of the shoulder plan, even though it is managed elsewhere.

What the treatment evidence honestly shows

This is where the marketing and the evidence part ways, so it is worth being direct about each option.

Exercise and manual therapy. A 2014 Cochrane review by Page and colleagues pooled 32 trials with 1,836 participants. Its conclusion was that "the best available data show that a combination of manual therapy and exercise may not be as effective as glucocorticoid injection in the short-term." In the trial behind that comparison, 46% of the manual therapy and exercise group reported treatment success against 77% of the injection group (RR 0.6, 95% CI 0.44 to 0.83), with pain improving 26 points less on a 100-point scale (95% CI 15 to 37). The review rated that moderate quality evidence. Adverse event rates were similar between the two groups, at 56% and 53%.

That finding surprises people, and it deserves care rather than dismissal. It compares short-term outcomes. It does not say exercise is pointless, and it does not tell you how the two compare over the full arc of a condition that runs for years. Exercise remains the backbone of care because it maintains the function you have, keeps the surrounding muscle from wasting, and rebuilds strength as range returns. A clinician who tells you exercise alone will beat an injection for short-term pain is overstating the evidence.

Corticosteroid injection. For short-term pain and range, this is the option with the strongest supporting evidence, and it is described in the 2024 review as reasonable first-line therapy for people with diabetes. The same review notes outcomes are "generally less favorable compared to those of non-diabetic patients." There is also a consequence specific to diabetes, covered in the next section.

Hydrodilatation. This involves injecting fluid into the joint to stretch the capsule from the inside, usually with a steroid. It is an established option offered by some radiologists and physicians in BC. The trial evidence comparing it against steroid injection alone is mixed, and it is a reasonable thing to ask about when injection and exercise have not moved things.

Surgery. Arthroscopic capsular release and manipulation under anaesthetic are for shoulders that have stalled after a genuine trial of conservative care. The 2024 review calls capsular release "the most effective invasive treatment" while noting the worse prognosis in diabetes described above. Surgery is a decision made with an orthopaedic surgeon.

The injection decision is different when you have diabetes

A corticosteroid injection raises blood glucose. This matters enough to plan around.

A 2025 paper in Medicina by Singer and colleagues reviewing corticosteroid use in musculoskeletal procedures reports that glucose typically peaks 1 to 3 days after injection, and that in people with diabetes the effects are more pronounced, "lasting for several days to weeks." One retrospective study it cites found the largest rises in people whose pre-procedure HbA1c was above 7% and who were also on insulin, with higher HbA1c before injection associated with greater glucose disturbance afterward.

The authors recommend close glucose monitoring after any such injection, and that corticosteroid injections be avoided in people who are poorly controlled, which they define as a fasting glucose above 200 to 250.

None of that makes injection the wrong choice. It makes it a decision to take with the physician who manages your diabetes, with a plan for monitoring in the days afterward. Practical steps: know your recent HbA1c before the appointment, tell the injecting physician you have diabetes even if it is on the form, plan to check your glucose more often for about a week, and ask in advance what reading should prompt a call.

How to decide when to escalate

Escalation is a conversation with a physician, and a physiotherapist's role is to tell you honestly when the conservative plan has stopped earning its place. The signals worth acting on:

• No meaningful change in range after a genuine trial. That means consistent work over 8 to 12 weeks at an appropriate dose, not a programme abandoned after two flares.
• Pain that is still preventing sleep after several weeks of treatment. Sleep loss compounds everything else, and it is one of the better reasons to consider injection earlier rather than later.
• Function that is falling rather than holding. Losing the ability to dress, drive, or work is a stronger argument for escalation than the range numbers alone.
• A picture that does not fit. Frozen shoulder has a characteristic pattern of stiffness in all directions, including when someone else moves your arm for you. Weakness without that stiffness points elsewhere, and the difference between impingement and rotator cuff problems is a common source of confusion.

The reason escalation deserves more attention in diabetes is the longer expected course. Waiting another six months to see what happens is a bigger cost when your baseline timeline is already toward the long end.

What a first visit looks like

Your physiotherapist will take a history first: when the pain started, whether anything triggered it, how it behaves at night, what you can no longer do, and how long each stage has lasted. If you have diabetes, expect questions about how long you have had it, how it is currently managed, and roughly where your control sits.

The examination checks how far the shoulder moves when you move it and how far it moves when the therapist moves it for you. Frozen shoulder restricts both, which is what separates it from most other shoulder problems. External rotation, meaning turning your forearm outward with your elbow at your side, is usually the most restricted direction.

From there you get an honest estimate of where you are in the arc, a small home programme you can sustain, and a clear statement of what would trigger a referral back to your physician. Expect the first programme to be short. Two or three exercises done regularly beat a long list dropped after a week.

Physiotherapy is a direct-access profession in British Columbia, regulated by the College of Health and Care Professionals of BC, so no physician referral is needed to book an assessment. Some extended-health plans require one for reimbursement, so check with your insurer before your first visit. You can book physiotherapy at Launch Rehab without a diagnosis in hand.

When to see a doctor instead

Some shoulder presentations need medical assessment rather than a physiotherapy appointment. See a physician promptly for a shoulder that follows a significant fall or trauma, for sudden weakness where you cannot lift the arm at all, or for pain with fever, redness, or feeling unwell, which can indicate infection. Shoulder pain with chest pain, jaw pain, shortness of breath, or sweating needs emergency care, since the heart can refer pain to the shoulder. New neurological symptoms in the arm, including numbness, pins and needles, or a hand that is losing grip, should be assessed medically.

One more, specific to this topic. If you have a frozen shoulder and have never been screened for diabetes, raise it with your family doctor. Given that 30% of people presenting with adhesive capsulitis have diabetes, a stiffening shoulder with no clear trigger is a reasonable prompt for a blood test.

This article is general information, not personal medical advice. A regulated practitioner can confirm whether the patterns described apply to you.

Frequently asked questions

How much more common is frozen shoulder in people with diabetes?

A 2016 meta-analysis in the Muscles, Ligaments and Tendons Journal pooled 13 studies and found adhesive capsulitis in 13.4% of people with diabetes (95% CI 10.2 to 17.2%), with an odds ratio of 5 against controls (95% CI 3.2 to 7.7). A more recent systematic review cited in a 2024 review puts the increase at 3.69 times. For context, prevalence in the general population is around 2 to 5%.

Why does diabetes cause frozen shoulder?

High blood sugar drives the build-up of advanced glycation end products, which cross-link collagen in the joint capsule and make it stiffer and harder to remodel. A 2024 review in the Journal of Clinical Medicine calls this the most significant influencing factor. High blood sugar also promotes chronic release of inflammatory messengers such as tumour necrosis factor alpha and interleukin-6, which contribute to fibrosis in the capsule.

Does frozen shoulder take longer to recover if I have diabetes?

Often, yes. The 2024 review reports that symptoms in people with diabetes are often more severe, need prolonged conservative treatment, and come to surgery more often than in people without diabetes. Frozen shoulder is generally described as self-limiting over about 1 to 3 years, and with diabetes it is sensible to plan for the longer end of that range.

Will controlling my blood sugar help my shoulder?

Blood sugar control is part of the plan, and the 2024 review states that treatment for adhesive capsulitis in diabetes must be accompanied by proper blood sugar control. The mechanism supports it, since the collagen cross-linking that stiffens the capsule is driven by sustained high glucose. How much improvement it produces in an already-established frozen shoulder has not been well quantified, so treat it as one necessary part of care rather than a fix on its own.

Should I get a cortisone injection if I have diabetes?

It can be a reasonable option, and it requires a plan. A 2025 review in Medicina reports that glucose usually peaks 1 to 3 days after injection and that the effect can last several days to weeks in people with diabetes, with the largest rises in those with an HbA1c above 7% who are on insulin. The authors advise close glucose monitoring afterward and avoiding injection in people who are poorly controlled. Discuss it with the physician who manages your diabetes before booking.

Is exercise or injection better for frozen shoulder?

For short-term pain and function, the evidence favours injection. A 2014 Cochrane review found that manual therapy and exercise may not be as effective as glucocorticoid injection in the short term, with 46% reporting treatment success against 77% for injection. Exercise still matters over the longer arc, because it maintains the function you have and rebuilds strength as range returns. Many people end up using both.

What is hydrodilatation and should I consider it?

Hydrodilatation involves injecting fluid, usually with a steroid, into the shoulder joint to stretch the contracted capsule from the inside. It is offered by some radiologists and physicians in BC. The trial evidence comparing it against steroid injection alone is mixed, so it is best treated as a reasonable option to discuss when injection and exercise have not produced change, rather than a first step.

When should I consider surgery for a frozen shoulder?

Surgery is considered when a genuine trial of conservative care has stalled, meaning consistent treatment over months without meaningful change in range or function. The 2024 review describes arthroscopic capsular release as the most effective invasive treatment while noting the prognosis stays worse for people with diabetes, particularly for internal rotation and forward flexion. That decision belongs with an orthopaedic surgeon.

Can frozen shoulder happen in both shoulders?

Frozen shoulder affecting the second shoulder is reported more often than the same shoulder freezing twice. Having diabetes is one of the factors associated with a more severe overall picture. If a second shoulder begins stiffening, it is worth assessing early rather than waiting, since the first one will have shown you how long the arc can run.

Could a frozen shoulder mean I have undiagnosed diabetes?

It is a reasonable prompt to get checked. The 2016 meta-analysis found diabetes in 30% of people presenting with adhesive capsulitis across five studies (95% CI 24 to 37%). A shoulder that stiffens with no clear injury behind it is a sensible reason to ask your family physician about a blood test, particularly if you have other risk factors.

Will physiotherapy make my frozen shoulder worse?

Treatment matched to the wrong phase can aggravate it, which is why assessment comes before a programme. Early on, when pain dominates, forceful end-range stretching tends to provoke symptoms rather than restore movement. Once stiffness dominates and pain has settled, graded loading and range work become the useful tools. Your therapist will set the dose and adjust it based on how you respond over the following days.

Do I need a referral to see a physiotherapist in BC?

No. Physiotherapy is a direct-access profession in British Columbia, regulated by the College of Health and Care Professionals of BC, so you can book an assessment without a physician's referral and without a confirmed diagnosis. Some extended-health insurance plans require a referral for reimbursement even though the profession does not, so confirm with your insurer before your first visit.

Sources

• Zreik, Malik and Charalambous, "Adhesive capsulitis of the shoulder and diabetes: a meta-analysis of prevalence," Muscles, Ligaments and Tendons Journal, 2016
• Chen and Chen, "A Narrative Review of Adhesive Capsulitis with Diabetes," Journal of Clinical Medicine, 2024
• Page et al., "Manual therapy and exercise for adhesive capsulitis (frozen shoulder)," Cochrane Database of Systematic Reviews, 2014
• Singer et al., "Corticosteroid Use in Musculoskeletal and Neuraxial Interventions: Effects on Glycemic Control," Medicina, 2025
• College of Health and Care Professionals of BC, physiotherapist scope of practice