Association between glucagon-like peptide-1 receptor agonists and femur fracture risk in type 2 diabetes: A large-scale target trial emulation.

Abstract

Type 2 diabetes mellitus (T2DM) paradoxically increases fracture risk despite normal or elevated bone mineral density. While glucagon-like peptide-1 receptor agonists (GLP-1 RAs) show promise in bone health, evidence remains controversial. This large-scale real-world study evaluated the primary outcome of femur fracture risk associated with GLP-1 RAs versus dipeptidyl peptidase-4 inhibitors (DPP-4is) in T2DM patients using a rigorous target trial emulation approach. This retrospective cohort study adopted a large-scale target trial emulation framework with a new-user, active-comparator design using data from the TriNetX US Collaborative Network. Adults (≥18 years) with T2DM identified between 2018 and 2022 (n = 3,620,983) were classified as GLP-1 RA (n = 491,936) or DPP-4i (n = 345,484) users at treatment initiation. Propensity score matching (PSM) was performed to emulate random assignment and balance baseline characteristics. Cox proportional-hazards models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for bone fractures during a maximum 5-year follow-up. After 1:1 PSM, each cohort included 172,381 patients (mean age 59 years; 51% male) with a mean follow-up of 40 months. GLP-1 RA users demonstrated significantly lower femoral fracture risk compared to DPP-4i users (HR 0.91, 95% CI 0.85-0.98), with consistent protective effects across all sensitivity analyses. Similar associations were observed when comparing GLP-1 RAs with most other antidiabetic medications. GLP-1 RA use was associated with a modest reduction in femur fracture risk. However, given the observational design, possible residual confounding, and multiple testing, these findings are exploratory and require prospective confirmation before clinical application.