Fracture‐Centred Skeletal Reporting Profiles of Glucose‐Lowering Drug Classes in FAERS and JADER : Cross‐Database and Regional Reporting‐Context Analyses

AIMS: To characterise fracture-centred skeletal adverse-event reporting across nine glucose-lowering drug classes and test whether regional differences persist after separating Japan from non-Japan Asian reporting contexts. MATERIALS AND METHODS: We analysed FAERS and JADER reports, with fracture as the primary endpoint and reporting frequency per 10 000 mapped suspect reports. FAERS older age was harmonised to ≥ 60 years, with ≥ 65 and ≥ 70-year sensitivity analyses. Regional analyses separated Western, Japan, and non-Japan Asia; class-specific reporting odds were adjusted for age, sex, calendar year, and reporter type. Disproportionality analyses were supplementary. RESULTS: The study included 15 473 015 FAERS and 1 037 161 JADER reports, including 202 881 and 10 678 fractures. Using final deduplicated FAERS denominators, fracture reporting was highest for insulin (126.08/10000), DPP-4 inhibitors (111.03) and sulfonylureas (103.76), whereas thiazolidinediones (152.54), SGLT2 inhibitors (115.14) and GLP-1 receptor agonists (112.15) were most prominent in JADER. Age was unavailable in 44.9% of unique FAERS suspect reports, but findings were stable across age thresholds. Japan accounted for 83.0%-92.9% of total-Asian fractures for thiazolidinediones, GLP-1 receptor agonists, SGLT2 inhibitors and DPP-4 inhibitors; corresponding non-Japan Asian RFRs were not clearly elevated. Insulin remained higher after excluding Japan (RFR 1.96, 95% CI 1.57-2.45). No class met FDR-positive fracture disproportionality criteria. CONCLUSIONS: Skeletal reporting profiles varied by drug class, database, and reporting context. Apparent total-Asian elevations for several classes were predominantly Japan-driven and should not be interpreted as patient-level Asian fracture risk.