SEATTLE — Use of GLP-1 receptor agonists (GLP-1s) was associated with a reduction in fragility fractures among older women with type 2 diabetes (T2D) in a retrospective analysis of data from a global health network. The results suggest that GLP-1 therapy could contribute to bone health.
Previous studies have produced mixed results about GLP-1 therapy and bone health, according to Jawaad Chaudhry, who presented the new study at the American Society for Bone and Mineral Research (ASBMR) 2025 Annual Meeting. GLP-1 receptor knockout mice have shown increased bone resorption and reduced calcitonin expression, and clinical symptoms include osteopenia and bone fragility, while zebrafish scale models show GLP-1s lead to increased bone formation. Meta-analyses have also shown a lower risk for fractures associated with GLP-1s than with other diabetes therapies, according to Chaudhry.
“These findings have varied across studies, so we set out to evaluate in a large real-world cohort of older women with type 2 diabetes our central question: Do [GLP-1s] have a protective effect on bone? More specifically, women age 55-89 years old with type 2 diabetes compared with peers on alternative therapies?” said Chaudhry, who is a medical student at Penn State College of Medicine, Hershey, Pennsylvania.
The researchers analyzed data from the TriNetX Global Collaborative Network, which is a global database with deidentified patient records. They used International Classification of Diseases, 10th Revision, and Current Procedural Terminology codes to identify patients with T2D who were prescribed a GLP-1 (lixisenatide, liraglutide, exenatide, tirzepatide, semaglutide, or dulaglutide) during 2005-2025 and propensity score matched these patients to those without exposure to GLP-1s. A 2-year follow-up cohort included 351,571 patients exposed to GLP-1s and an equal number of control individuals. A 5-year cohort had 351,597 patients and the same number of control individuals.
In the 2-year group, there were reduced risks for vertebral wedge compression fractures (relative risk [RR], 0.836; 95% CI, 0.781-0.895), subtrochanteric fractures (RR, 0.61; 95% CI, 0.474-0.776), intertrochanteric fractures (RR, 0.765; 95% CI, 0.67-0.874), femoral neck fractures (RR, 0.66; 95% CI, 0.601-0.725), distal radius fractures (RR, 0.786; 95% CI, 0.724-0.854), and periprosthetic fractures (RR, 0.746; 95% CI, 0.645-0.864).
There was a similar pattern in 5-year fracture risk, with reduced risks for vertebral wedge compression fractures (RR, 0.746; 95% CI, 0.712-0.781), subtrochanteric fractures (RR, 0.594; 95% CI, 0.492-0.717), intertrochanteric fractures (RR, 0.618; 95% CI, 0.566-0.675), femoral neck fractures (RR, 0.572; 95% CI, 0.538-0.608), distal radius fractures (RR, 0.674; 95% CI, 0.637-0.713), and periprosthetic fractures (RR, 0.716; 95% CI, 0.636-0.806).
Could GLP-1-Induced Changes in Medication Use Confound the Findings?
During the Q&A session after the talk, an audience member pointed out a key potential confounder, which is that the weight loss associated with GLP-1 use could lead to patients reducing the use of other medications such as cholesterol meds or nonsteroidal anti-inflammatory drugs. “How do you know that this [fracture risk reduction] was due to GLP-1s and not due to all of the other changes that were induced, including many of those other drugs that we know have adverse [effects on bone] quality? This is not to negate the fact that this is a great study, but it has to be answered. Was it really the drugs or was it really the weight loss and the changes associated with the weight loss that helped these patients?” said the questioner, who asked to remain anonymous.
Chaudhry acknowledged the limitation but cited other evidence. “There are several studies out there citing intentional weight loss in diabetic patients actually increases the risk for fracture. So I’m curious to see what future studies in our own analysis will produce,” he said.
Also during the Q&A session, Richard Eastell, MD, wondered if there were different effects of drug generations. He said that early GLP-1 drugs such as liraglutide may have had a neutral or beneficial effect on bone, but later, more potent drugs might lead to accelerated bone loss. One study that he co-authored showed no effect of semaglutide on the bone formation biomarker plasma procollagen type 1 N-terminal propeptide, but there was an increase in bone resorption. “So I just wonder whether or not what we’re seeing here, if you’re looking over time, is dominated by the earlier drugs,” said Eastell, who is a professor of bone metabolism at the University of Sheffield, Sheffield, England.
Chaudhry reported having no relevant financial disclosures. Eastell reported having consulted for Samsung, Curatech, and Sandoz.
Jim Kling is a writer based in Bellingham, Washington.
