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

Design

Systematic review and meta-analysis of 60 studies (46 randomised trials); GRADE-assessed

Population

Over 1.2 million people on GLP-1 receptor agonists for diabetes or obesity

Primary outcome

Bone density and fractures, muscle mass, and osteoarthritis symptom scores

Effect

No effect on bone or joint symptoms; lean-mass reduction (standardised mean difference −0.52, −0.80 to −0.23)

With GLP-1 receptor agonists now used widely for diabetes and obesity, their effect on the musculoskeletal system matters to anyone managing bone and joint disease. This meta-analysis pooled 60 studies and over 1.2 million people across several agents.

There was no effect on bone mineral density or fractures when the best-adjusted estimates were used, and no change in osteoarthritis symptom scores. The consistent finding was a reduction in lean body mass, or fat-free mass, seen across sensitivity analyses and driven mainly by semaglutide and liraglutide.

For practice, the reassurance is that these drugs do not appear to harm bone or worsen joint symptoms. The caveat is the muscle signal: the lean-mass loss largely tracks weight loss and its functional importance is unproven, but it is a reason to pair GLP-1 therapy with resistance exercise and adequate protein, particularly in older or frailer patients at risk of sarcopenia.

• No effect on bone mineral density or fractures using the most-adjusted estimates.
• A consistent reduction in lean/fat-free mass (standardised mean difference −0.52, 95% CI −0.80 to −0.23), mainly with semaglutide and liraglutide.
• No change in osteoarthritis pain, function or stiffness (WOMAC).
• Pair GLP-1 therapy with resistance exercise and adequate protein in patients at risk of muscle loss.

Why it matters

It answers two common worries — fractures and joint symptoms — reassuringly, while flagging muscle loss as the effect to manage.

Don't overread it

The lean-mass loss is largely a consequence of weight loss, and whether it translates into weaker muscles or worse function is not established.

The statistics, in plain English

A standardised mean difference of −0.52 is a moderate reduction in lean mass, and the interval staying below zero makes it a real signal; but very high heterogeneity (I² 88%) and low certainty mean the exact size is soft, and the loss mostly reflects overall weight loss.

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.