Older adults with type 2 diabetes who started taking GLP-1 receptor agonists, the drug class behind semaglutide and similar medications, broke bones 11 percent more often than peers prescribed other common diabetes treatments. The finding comes from a study of 46,177 patients age 65 and older tracked through Israel’s Clalit Health Services database between 2018 and 2022. With tens of millions of older adults now using these drugs worldwide, the result sharpens a question that prescribers and patients have largely set aside: whether the rapid weight loss these medications produce weakens bones in people who can least afford a fracture.
Why fracture risk in older GLP-1 users demands attention now
The study, published in The Journal of Clinical Endocrinology and Metabolism, compared 11,257 adults who began GLP-1 receptor agonists with 34,920 who started SGLT-2 inhibitors or DPP-4 inhibitors instead. After adjusting for confounders, the GLP-1 group had a hazard ratio for fragility fractures of 1.11, with a 95 percent confidence interval of 1.01 to 1.21. That interval barely clears statistical significance, but the population size and the real-world clinical-record design give the signal weight that smaller trials have not been able to provide, as detailed in the Clalit analysis.
The timing matters because GLP-1 prescriptions have surged among older patients since 2020, driven by cardiovascular benefit data and broadening insurance coverage. Older adults already face elevated fracture risk from age-related bone loss, reduced muscle mass, and higher fall rates. Rapid weight loss, which GLP-1 drugs reliably produce, strips away mechanical loading on bones and can reduce lean muscle mass, a condition known as sarcopenia. In younger patients, that trade-off may be manageable. In someone over 75, a hip fracture carries roughly a 20 to 30 percent one-year mortality rate, making even a modest increase in fracture incidence a serious clinical concern.
A separate randomized phase 2 trial tested once-weekly semaglutide in adults with increased fracture risk and tracked bone-turnover biomarkers and bone mineral density endpoints. In that study, investigators followed changes in markers such as P1NP and CTX as well as dual-energy X-ray absorptiometry scans, but they did not assess clinical fractures as an outcome, leaving uncertainty about how laboratory shifts translate into real-world harms. The trial, described in detail in the semaglutide fracture-risk trial, had a shorter duration and smaller sample, so it could not fully capture what happens to bones over years of treatment in very old adults. The Clalit study fills part of that gap by counting actual fractures in a large, older population over a four-year window.
Clalit cohort data and the fracture signal from FDA labeling
The Clalit Health Services dataset draws from one of the largest integrated health systems in Israel, covering roughly half the country’s population. Its electronic medical records allowed researchers to identify new users of each drug class and follow them for fracture events documented by treating physicians, not self-reported surveys. The investigators excluded people with prior use of the study drugs and adjusted for age, sex, comorbidities, and prior fracture history, aiming to approximate the conditions of a randomized comparison while retaining the scale of a national cohort. According to the journal’s article record, the manuscript was received in December 2025, accepted in early February 2026, and corrected by late February 2026, underscoring how recent these findings are.
The 11 percent increase is not the first fracture signal linked to this drug class. U.S. prescribing information for Wegovy, the branded semaglutide injection approved for weight management, already discloses that in a cardiovascular outcomes trial, hip and pelvis fractures occurred in 1.0 percent of Wegovy-treated females compared with 0.2 percent on placebo. Among patients age 75 and older, the gap was wider: 2.4 percent on Wegovy versus 0.6 percent on placebo, according to the FDA-approved label. Those numbers appear in the drug’s official prescribing information but have not triggered a formal boxed warning or a dedicated fracture-risk section of the kind the FDA issued for canagliflozin, an SGLT-2 inhibitor, after bone mineral density concerns emerged.
An earlier U.S. study using Medicare claims data compared fracture rates across GLP-1 receptor agonists, SGLT-2 inhibitors, and DPP-4 inhibitors in older adults with type 2 diabetes and reported similar nontraumatic fracture risk across groups. In that analysis, which reviewed hundreds of thousands of beneficiaries, GLP-1 drugs did not appear to raise fracture risk relative to the other two classes. Those results, described in the Medicare claims study, helped reassure clinicians that GLP-1 therapy was bone-neutral in older patients. The Clalit study’s larger sample of GLP-1 initiators and its different healthcare setting now challenge that conclusion, at least for older adults in routine clinical practice.
Gaps in bone density data and what patients should watch
Several questions remain open. The Clalit study did not release patient-level bone mineral density measurements or fall data, so researchers cannot yet say whether GLP-1 drugs directly weaken bone tissue, increase falls through effects on muscle and balance, or simply unmask underlying fragility by accelerating weight loss. Without imaging data, it is also unclear whether particular skeletal sites, such as the hip or vertebrae, are disproportionately affected. The modest hazard ratio suggests that any single mechanism is likely subtle, but at a population level, even small shifts can translate into thousands of extra fractures.
Another unresolved issue is how fracture risk evolves over time. GLP-1 drugs often cause the most rapid weight loss in the first six to 12 months, followed by a plateau. If the primary driver is loss of mechanical loading on bone, fracture risk might spike early and then stabilize. If the drugs alter bone remodeling pathways directly, risk could accumulate with longer exposure. The Clalit analysis followed people for up to four years but did not break out hazard ratios by duration of therapy, leaving clinicians to extrapolate from general principles of bone biology and the more granular biomarker data from smaller trials.
For patients and prescribers, the immediate question is not whether GLP-1 drugs should be abandoned in older adults, but how to use them more safely. For many people with type 2 diabetes, especially those with obesity and established cardiovascular disease, GLP-1 receptor agonists lower heart attack and stroke risk, improve glycemic control, and support sustained weight loss. Those benefits can themselves reduce fracture risk indirectly by improving mobility, lowering fall risk, and decreasing the likelihood of debilitating vascular events. The challenge is to preserve those gains while minimizing harm to the skeleton.
Experts who treat osteoporosis and diabetes increasingly recommend a few pragmatic steps. Before starting a GLP-1 drug in someone over 65, clinicians can review prior fractures, check for height loss or back pain that might signal vertebral compression, and consider ordering a bone density scan in patients with multiple risk factors. During treatment, monitoring weight trajectory, muscle strength, and balance can help identify those who may be losing lean mass too quickly. Simple interventions such as resistance training, adequate protein intake, and ensuring sufficient calcium and vitamin D can support bone and muscle health, even if they do not fully offset drug-related changes.
Patients already taking GLP-1 medications should not stop them abruptly without medical advice, particularly if the drugs are controlling blood sugar or heart disease risk. Instead, they can ask their clinicians whether their fracture risk profile warrants additional evaluation or preventive therapy, such as bisphosphonates, if osteoporosis is documented. For frailer older adults, especially those with prior hip or vertebral fractures, the new data may tilt the balance toward using alternative diabetes drugs when possible, or at least toward lower GLP-1 doses and slower titration to limit rapid weight loss.
Regulators and guideline committees will likely wait for confirmatory analyses before issuing broad recommendations. Replicating the Clalit findings in other large datasets, ideally with more detailed information on bone density and falls, will be essential. In the meantime, the emerging fracture signal is a reminder that powerful metabolic drugs can have far-reaching effects beyond blood sugar and body weight. For older adults with type 2 diabetes, protecting bone health may need to move higher on the checklist when GLP-1 therapy is on the table.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
