Vitamin B12 status and long-term fracture risk in a multicenter propensity score–matched cohort

Abstract Vitamin B12 status may be associated with fracture risk, but prior evidence is inconsistent, and abnormal concentrations may partly reflect broader nutritional status, comorbidity, or healthcare utilization. We examined whether laboratory-defined vitamin B12 deficiency was associated with subsequent incident fracture among adults aged ≥ 50 years undergoing vitamin B12 testing. Using electronic health record data from the TriNetX Global Collaborative Network, we conducted a retrospective multicenter cohort study with a 1-year landmark design. Patients with deficiency (≤ 199 pg/mL) were compared with those with normal concentrations (300–900 pg/mL) using 1:1 propensity score matching. After matching, 115,735 patients were included in each group. Vitamin B12 deficiency was associated with a higher hazard of incident fracture over the full analytic period (HR, 1.33; 95% CI, 1.29–1.38). Because the proportional hazards assumption was violated, a 1–5-year sensitivity analysis showed a similar association (HR, 1.37; 95% CI, 1.32–1.42) without evidence of non-proportionality. Exploratory secondary analyses showed directionally similar associations for fracture-, fall-, and osteoporosis-related outcomes. These findings apply to patients who underwent vitamin B12 testing and remained event-free through the landmark period and do not establish vitamin B12 deficiency as a causal or predictive fracture marker.