Context-Dependent Association Between Serum 25-Hydroxyvitamin D and Romosozumab Bone Mineral Density Response: A Stratified Analysis by Renal Function Category and Prior Treatment History in a Real-World Japanese Cohort

Background/Objectives: Serum 25-hydroxyvitamin D (25OHD) is a key determinant of calcium-phosphorus homeostasis and bone metabolism; however, its role as a modifier of the bone mineral density (BMD) response to romosozumab, a dual-action anabolic agent for osteoporosis, remains poorly characterized, particularly across different levels of renal function. This study investigated whether the renal function category and prior treatment history modified the association between baseline 25OHD and romosozumab BMD response. Methods: We conducted a retrospective cohort study of 315 consecutive Japanese patients treated with romosozumab (210 mg subcutaneously monthly for 12 months) at Mutsu General Hospital, Aomori, Japan (2019–2025; IRB approval RO7-5; date of approval: 20 January 2026). Patients were stratified by eGFR-based renal function category: preserved renal function (eGFR ≥ 60 mL/min/1.73 m2, n = 199) and moderately reduced renal function (eGFR 30–59 mL/min/1.73 m2, n = 86). Patients with severely reduced renal function (eGFR 15–29, n = 11) were excluded from the comparative analyses. Spearman rank correlations (Rs) were computed between baseline 25OHD and (i) baseline TRACP-5b and (ii) 12-month lumbar spine BMD changes. Mediation analysis was performed to examine TRACP-5b as a potential mediator. Results: In the preserved renal function group, baseline 25OHD was significantly and inversely correlated with TRACP-5b (Rs = −0.246, p = 0.0007). This correlation was absent in the moderately reduced renal function group (Rs = +0.036, p = 0.74), and the interaction was statistically significant (z = −2.38, p = 0.017). Among treatment-experienced patients, lower 25OHD levels were correlated with greater LS-BMD response (Rs = −0.197, p = 0.036), whereas no such correlation was observed in treatment-naïve patients (Rs = −0.009, p = 0.902). Mediation analysis did not identify TRACP-5b as a significant mediator. Conclusions: The association between serum 25OHD and romosozumab BMD response appears to be context-dependent across renal function categories and prior treatment history. These findings are hypothesis-generating and require prospective validation before they can be applied to clinical practice.