GLP-1 medicines and musculoskeletal care: what the evidence says about muscle, bone and joints

"Are you on one of the weight-loss jabs?"

A few years ago, that question would have been an unusual one in an MSK case history. Now it's worth asking. NICE has recommended semaglutide (Wegovy, TA875) and tirzepatide (Mounjaro, TA1026) for managing overweight and obesity in adults, NHS England has published phased implementation guidance for tirzepatide, and some people are prescribed these medicines privately. The same drug classes are also used for type 2 diabetes. That means some of the people we see for back pain, knee pain or a stiff shoulder may be losing weight on a medicine we won't know about unless we ask.

This article looks at what the current evidence says about how GLP-1 medicines relate to muscle, bone and joints, what is and isn't known about exercise alongside them, which symptoms in an MSK clinic deserve a second look, and how to talk to the prescriber. It's written for chiropractors, physiotherapists, osteopaths, sports therapists and other clinicians who treat musculoskeletal problems. A quick note on terms: "GLP-1 medicines" here includes semaglutide, liraglutide and similar drugs, plus tirzepatide, which acts on GLP-1 and GIP receptors.

One boundary up front. Starting, stopping or changing a dose is a decision for the prescriber and the patient, not for us. Our job is to ask, to notice, to adapt the plan and to refer or write when something needs a medical look.

Muscle: real losses, uncertain meaning

Any substantial weight loss, whether from diet, surgery or medicines, usually includes some loss of lean tissue as well as fat. "Lean mass" on a body scan is not the same thing as muscle, because it also includes water and other non-fat tissue, so these figures are an imperfect guide to strength.

A DXA substudy of the SURMOUNT-1 tirzepatide trial (Look and colleagues, Diabetes, Obesity and Metabolism, 2025) is a useful reference point. It included 160 of the trial's 2,539 participants (73% female, mean BMI 38). Over 72 weeks, body weight fell by 21.3% on tirzepatide and 5.3% on placebo. Fat mass fell by 33.9% against 8.2%, and lean mass by 10.9% against 2.6%. Roughly three quarters of the weight lost was fat and about a quarter was lean mass, in both the tirzepatide and placebo groups. Several authors were employees of the manufacturer, Eli Lilly, and the substudy was a small subset of a larger trial.

A 2026 systematic review in the Annals of Internal Medicine pulled together 35 primary studies of incretin-based weight loss. The studies were small and short (median duration 26 weeks), and the amount of muscle-related loss varied widely. In about two thirds of the drug groups, muscle-related loss exceeded the benchmarks the reviewers set in advance (about a quarter of total weight lost for DXA or bioimpedance lean measures). It also exceeded that benchmark in nearly half of the non-drug weight-loss groups, so this is not unique to medicines. Importantly, no study reported objective physical function outcomes. A 2026 meta-analysis of musculoskeletal outcomes with GLP-1 medicines in Drugs (60 studies, including 46 randomised trials) also found lower lean or fat-free mass with these medicines. The authors rated the certainty of evidence as low, found the changes appeared largely related to the weight loss itself, and said it remains uncertain whether they translate into meaningful impairments in muscle function or physical performance.

What this means in clinic: lean tissue loss is real, its size varies a lot between people and studies, and nobody yet knows how often it matters for strength or function. That uncertainty is a reason to measure and to protect strength, rather than a reason to assume harm or to dismiss it.

Bone: not much signal, but not much data

The same 2026 meta-analysis in Drugs found no effect of GLP-1 medicines on bone mineral density or fractures at any site in its main analyses. A separate 2026 systematic review in Archives of Osteoporosis (eight randomised trials, 493 adults with overweight or obesity) was more cautious. The evidence was sparse and the certainty was very low for every agent, and five of the eight trials were in adults with type 2 diabetes. The two trials with the largest weight loss, one of semaglutide and one of liraglutide 3.0 mg, both reported lower lumbar spine bone density, and the liraglutide trial also reported lower total hip density. The authors said this deserves dedicated study rather than assumptions across the whole drug class.

A 2025 joint advisory in the journal Obesity (Mozaffarian and colleagues, on behalf of four US lifestyle, nutrition and obesity societies) points out that bone loss with weight reduction depends on factors such as starting weight, age, sex, activity, how much energy intake is cut, protein intake and how fast the weight comes off, with older people and women losing more. The UK Wegovy patient leaflet lists hip fractures among its rare side effects (up to 1 in 1,000 people).

What this means in clinic: there's no clear evidence of a bone problem across the class, and also no strong evidence of safety for older people or people already at risk of fracture. It's reasonable to keep bone health in mind in older patients, in people with osteoporosis risk factors and in anyone losing weight fast, and to include load-bearing exercise in the plan.

Joints: knee osteoarthritis is the clearest benefit

The best-known trial for joints is STEP 9 (Bliddal and colleagues, New England Journal of Medicine, 2024). It enrolled 407 adults with obesity and painful knee osteoarthritis (mean age 56, mean BMI 40.3, 81.6% women). Everyone also received counselling on physical activity and a reduced-calorie diet. At 68 weeks, body weight had fallen 13.7% with once-weekly semaglutide (2.4 mg) against 3.2% with placebo. The WOMAC pain score improved by 41.7 points against 27.5, and SF-36 physical function improved by 12.0 points against 6.5. Adverse events led to permanent discontinuation in 6.7% of the semaglutide group and 3.0% of the placebo group, with gastrointestinal disorders the most common reason. The trial was funded by Novo Nordisk.

Pooled results are less clear-cut. The 2026 meta-analysis in Drugs found no significant change in WOMAC pain, function or stiffness in its joint models. Taken together, it looks as though weight loss on these medicines can reduce knee pain in people with obesity and knee osteoarthritis, but the participants in STEP 9 were a specific group (high BMI, mostly women), the trial was industry-funded, and the wider evidence for joints is thinner than for weight itself.

What this means in clinic: don't promise that a weight-loss medicine will fix someone's knee. Weight loss may help pain, and exercise and strength work remain a core part of osteoarthritis care alongside it.

Exercise alongside these medicines: what's consensus and what's trial evidence

This is where it's easy to overstate things, so it helps to separate the types of evidence.

Expert advice and consensus (not drug-specific trials):

• The 2025 Obesity joint advisory says structured strength (resistance) training, or resistance plus aerobic training, is well established for helping preserve lean mass during weight reduction, while aerobic exercise alone has a smaller effect. It also says clinicians should make clear that higher protein intake on its own is likely not enough to preserve muscle without structured resistance training. It recommends a baseline assessment that includes muscle strength, function and body composition. It adds that higher protein targets, such as 1.2 to 1.6 g/kg/day, have been proposed during active weight loss. Those numbers are for the patient, their prescriber and a dietitian to settle, not something for us to prescribe.
• A 2025 perspective in Physical Therapy (Mulcahy and colleagues) recommends that every care plan for a patient on a GLP-1 medicine includes both aerobic and resistance training, with general recommendations of 150 to 300 minutes of moderate-intensity activity and 2 to 3 resistance sessions per week. It notes that people may need fewer sessions per week or a longer ramp-up, and says more studies are needed to define best practice.

Randomised trial evidence:

• In a 2021 trial in the New England Journal of Medicine (Lundgren and colleagues), 195 adults with obesity who had lost weight on a low-calorie diet were randomised for a year to exercise, liraglutide 3.0 mg, both or placebo. The exercise programme was moderate-to-vigorous intensity. The combination group had the best result for body weight (9.5 kg lower than placebo) and body-fat percentage (down 3.9 percentage points, about twice the fall with either exercise or liraglutide alone). Only the combination improved glycated haemoglobin, insulin sensitivity and cardiorespiratory fitness. That trial was about weight-loss maintenance and fitness, not about preserving muscle.
• Trials aimed squarely at muscle preservation are on the way. The LEAN-PREP protocol (BMJ Open, 2026) describes a planned 6-month randomised trial of home-based resistance exercise and protein in 232 adults starting semaglutide or tirzepatide, with MRI-measured quadriceps size as the primary outcome. That's a protocol, not results.

What this means in clinic: progressive resistance training is a sensible, low-risk part of care for people on these medicines and fits with the wider evidence and expert advice, but we should be honest that direct trial evidence on muscle and function in this group is still emerging.

Symptoms worth a second look in an MSK clinic

The UK Wegovy patient leaflet (last revised August 2026 on the version we checked) and MHRA safety advice give us a practical list. Other products have their own leaflets, so check the specific medicine.

• Acute pancreatitis. The Wegovy leaflet lists it as uncommon (up to 1 in 100) and advises stopping the medicine and seeking urgent medical help for severe, persistent abdominal pain. In January 2026 the MHRA strengthened warnings across all GLP-1 and GLP-1/GIP medicines, including rare reports of necrotising and fatal pancreatitis. Between 2007 and October 2025 it had received 1,296 Yellow Card reports of pancreatitis, 19 of them fatal and 24 necrotising, set against roughly 25.4 million packs dispensed in the past five years. The advice is to be alert for severe and persistent abdominal pain that may radiate to the back, which may come with nausea and vomiting, and to seek urgent medical attention. The MHRA also notes that privately prescribed GLP-1 medicines may not appear in the medical record, so enquire about their use if a patient presents with these symptoms. Back pain that is not mechanical in behaviour, with abdominal pain or vomiting, is a good reason to ask.
• Kidney or bladder stones. The Wegovy leaflet lists these as uncommon (up to 1 in 100). Signs can include back or lower abdominal pain, difficulty passing urine or a change in urine colour. Gallstones are listed as common (up to 1 in 10).
• Dehydration and kidney problems. Nausea, vomiting and diarrhoea can cause dehydration, which the leaflet says matters especially for people with kidney problems.
• Dizziness on standing. Dizziness is listed as common, and low blood pressure and feeling dizzy or lightheaded on standing or sitting up as uncommon. Dizziness is mainly reported in the first three months of treatment. It's worth taking a patient up from the couch slowly, and thinking about this before standing balance or sit-to-stand testing.
• Low blood sugar. The leaflet lists hypoglycaemia as common in people with diabetes, more likely if they also take a sulfonylurea or insulin, and as uncommon in people without diabetes. Warning signs include cold sweat, pale skin, headache, fast heartbeat, feeling shaky, anxious or confused, and difficulty concentrating.
• Bowel obstruction. Listed with unknown frequency: severe constipation with stomach ache, bloating and vomiting.
• Fatigue and headache. Both are listed as very common, usually settle over time, and can muddy the picture of an MSK complaint or a rehabilitation response.

Putting it into practice

None of this is a formal guideline for MSK clinicians, so treat it as a practical checklist drawn from the sources above.

1. Ask, and record it. Which medicine, who prescribes it (NHS service, GP or a private provider), when it started and whether the dose has changed recently.
2. Screen for side effects that change what you do today. Ask about nausea, vomiting, diarrhoea, how much they're eating and drinking, dizziness and falls, and any new abdominal, flank or back pain that doesn't behave mechanically.
3. Get a baseline. The advisory recommends assessing muscle strength and function. Pick something you can repeat reliably, such as a sit-to-stand count, grip strength or a simple lower-limb strength test, and re-test over time.
4. Build strength into the plan. Progressive resistance work, with graded loading, adapted to how the person feels on the day. Ask them to speak to their prescriber or a dietitian about protein and calorie intake rather than giving your own targets.
5. Don't overpromise. Be careful about attributing pain relief, weakness or fatigue to the medicine or to the rehab. Several things are changing at once.
6. Write to the GP or prescriber when you need to. For example, if you're concerned about symptoms that could be adverse effects, or about falls, weight loss that seems too fast or a patient who has lost strength. Describe what you found and ask for review; leave the medicine decision to them.
7. Know when it's urgent. Severe, persistent abdominal pain, with or without vomiting, needs same-day medical attention rather than a manual therapy plan.

Key points

1. Ask about GLP-1 medicines, including private prescriptions, as part of the history.
2. Lean mass falls with weight loss on these medicines (about a quarter of the weight lost in the SURMOUNT-1 substudy), but whether that harms strength or function is unproven.
3. Evidence on bone is sparse: no overall effect in a large meta-analysis, but some trials with large weight loss showed lower bone density.
4. Weight loss on semaglutide improved knee osteoarthritis pain in an industry-funded trial of people with obesity. Pooled joint results are less clear.
5. Resistance training is the consistent recommendation from expert advice, and direct trial evidence in this group is still emerging.
6. Keep pancreatitis, gallstones, kidney stones, dehydration, dizziness on standing and hypoglycaemia in mind as red flags and adverse effects.
7. Stay in scope: describe, document and refer, and leave medicine decisions to the prescriber.

Musculoskeletal assessment and rehabilitation for people taking these medicines is covered in ManualCPD's 1 CPD hour course, Rehab Reimagined: Musculoskeletal Care in the GLP-1 Era. It equips healthcare professionals with essential knowledge of the physiological effects of GLP-1 receptor agonists on musculoskeletal health, so they can adapt patient assessment and tailor rehabilitation strategies, and it builds clinical reasoning through practical insights into history taking, physical examination modifications and collaborative care. ManualCPD offers online, on-demand CPD for chiropractors, physiotherapists, osteopaths and sports therapists, broken into manageable chunks you can work through between patients, at lunch or after clinic. Every ManualCPD course is accredited by The CPD Group (Provider 788442), and you receive a CPD certificate on completion.