GLP-1 agonists trim lean mass, but spare bone and joints

Design

Systematic review and meta-analysis (RCTs and real-world evidence), GRADE-rated

Population

60 studies, over 1.2 million individuals on GLP-1 receptor agonists

Primary outcome

Bone, muscle and joint outcomes

Effect

Lean mass standardised mean difference about -0.5 (95% CI -0.8 to -0.23); no effect on bone density, fractures or osteoarthritis

A systematic review and meta-analysis pooled 60 studies (46 randomised trials, 13 real-world and one pharmacovigilance study, over 1.2 million people) on the musculoskeletal effects of GLP-1 receptor agonists, spanning semaglutide, liraglutide, tirzepatide and others.

There was no effect on bone mineral density or fractures, and no change in osteoarthritis pain, function or stiffness. There was, however, a consistent fall in lean body mass / fat-free mass (standardised mean difference about -0.5), robust across sensitivity analyses and driven mainly by liraglutide and semaglutide against placebo. The certainty was low, and the loss appeared largely to accompany weight loss rather than to be a separate toxic effect.

For prescribing, the practical point is to protect muscle while people lose weight on these drugs: encourage resistance exercise and adequate protein, and be alert in older or frail patients where lean-mass loss matters most. There is no signal to justify bone-density monitoring or joint concern on these grounds.

• GLP-1 receptor agonists showed no effect on bone density or fractures.
• They did not change osteoarthritis pain, function or stiffness.
• Lean body mass fell consistently (standardised mean difference about -0.5), mainly with liraglutide and semaglutide.
• Counsel resistance exercise and adequate protein during GLP-1-driven weight loss.
• Pay particular attention to older or frail patients, where muscle loss matters most.

Why it matters

It answers a common question about these fast-spreading drugs: the muscle concern is real but modest, while feared bone and joint harms are not seen.

Don't overread it

Low-certainty evidence with high heterogeneity; the lean-mass loss largely accompanies weight loss and has not been shown to impair measured muscle function.

The statistics, in plain English

A standardised mean difference near -0.5 is a moderate effect, but high heterogeneity (I-squared 88%) and low certainty mean the size is imprecise; because it tracks weight loss, it may not reflect a direct drug toxicity.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.