THE ARISING · Longevity & Healthspan
Bone Density Fell in GLP-1 Trials With the Most Weight Loss
In the two randomized trials where people lost the most weight on a GLP-1 drug, bone mineral density at the spine and hip went down, according to a new systematic review. The authors say the evidence is too thin to draw firm conclusions.
The review, published October 23 in Archives of Osteoporosis, is a systematic review of eight randomized controlled trials with 493 adults in total. A systematic review gathers all the trials that meet set criteria and reports what they found. All participants had overweight or obesity. This one did not pool the results, because the four drugs studied differ in how much weight they take off. Instead it reported each drug on its own. The authors rated the certainty of the evidence as very low for every drug.
Bone mineral density, or BMD, is a measure of how much mineral is packed into bone. It was measured by DXA, a low-dose X-ray scan. Lower density is generally a concern because it can mean weaker bone, though these trials did not measure fractures. The semaglutide trial had 64 participants and the largest weight loss of any trial: 6.8 kg more than placebo. It found lower lumbar spine BMD (a difference of 0.018 g/cm² versus placebo) and lower whole-body BMD (a difference of 0.02 g/cm²). It also found a higher level of C-terminal telopeptide, or CTX, a blood marker of bone breakdown. The review rated this trial at low risk of bias.
Liraglutide was tested at two doses, with different results. At 3.0 mg daily, the weight-management dose, the one trial in adults with obesity and no diabetes found 6.7 kg more weight loss than placebo. Lumbar spine BMD was lower by 0.019 g/cm² and total hip BMD by 0.013 g/cm². At 1.8 mg daily, the diabetes dose, weight loss was 3.8 kg and no site changed materially.
Exenatide, the drug with the most trials (five), gave inconsistent results. At the lumbar spine, one trial found a gain of 0.03 g/cm², while another found a change of −0.04 g/cm² with a wide margin of uncertainty that included no change. Semaglutide and dulaglutide were each tested in one trial, and liraglutide in two.
There are important limits. Five of the eight trials enrolled adults with type 2 diabetes rather than obesity alone. Three trials were at low risk of bias, and five raised some concerns, mainly about how participants were randomized and how results were reported. No drug is backed by more than one low-risk trial. The review's search ran through July 2026 and included only trials lasting at least 12 weeks.
What the authors do say is that the matching result in the two biggest weight-loss trials deserves its own study, rather than being assumed to apply to every drug in the class. Whether the bone changes reflect weight loss itself, the drugs, or both is not answered by this review. Fracture risk, the outcome that matters most for older adults, was not assessed.
The review does not overlap with The Arising's October 15 report on menopause hormone therapy and semaglutide weight loss, which was an observational finding about weight. This review is about bone, and it adds a different question: how bone fares as weight comes off.
The authors call for adequately powered trials of individual drugs with fracture as an endpoint. That is the evidence to watch for next.