A systematic review and meta-analysis published in Bone reports that GLP-1 receptor agonists reduced lumbar spine bone mineral density slightly and increased a marker of bone breakdown compared with placebo, while leaving hip measurements unchanged. The pooled evidence comes almost entirely from older drugs, not the high-dose obesity agents dominating the market.

GLP-1 receptor agonists are the drug class that includes exenatide, dulaglutide and liraglutide, along with newer entries such as semaglutide (sold as Ozempic and Wegovy). They mimic a gut hormone that influences insulin release and appetite, and the open question here is what happens to the skeleton when body weight falls.

Key results

Across 11 trials pooled for meta-analysis, with one more handled narratively, treatment lasted 12 to 52 weeks. Lumbar spine BMD fell by a weighted mean difference of -0.02 g/cm2 versus placebo (3 trials, N = 222), rated moderate certainty, which the authors describe as small and borderline.

The bone resorption marker CTX rose by 0.05 micrograms per liter (6 trials, N = 439), though certainty was low and heterogeneity high. Femoral neck BMD, total hip BMD and the bone formation marker P1NP showed no significant differences. Serum hs-CRP, a systemic inflammation marker, fell by 0.49 mg/L (4 trials, N = 340).

The gap in the evidence base

This is the part worth sitting with. The authors identified zero eligible placebo-controlled trials of semaglutide 2.4 mg/week or tirzepatide, the two agents responsible for most of the weight loss happening in the real world right now.

One trial of semaglutide 1.0 mg/week was included, and removing it did not change the direction or significance of any pooled estimate. Everything else is first-generation agents producing moderate weight loss, which is not the exposure most current patients have.

Limitations

Sample sizes are small: the lumbar spine finding rests on 222 participants across three trials. Certainty ratings ranged from moderate down to very low, and heterogeneity exceeded 68 percent for several hip and femoral neck estimates.

The authors state plainly that these findings cannot be extrapolated to higher-potency anti-obesity therapies, and call for dedicated long-term skeletal safety trials.

Bottom line

The signal reported here is modest and confined to the spine, with hip density preserved and inflammation lower. What the field still does not have is placebo-controlled skeletal data on the drugs people are actually taking for obesity.