(CSPOT09) Evaluating the Association of Denosumab versus Bisphosphonates on Diabetes Risk: A Nationwide Cohort Study and Meta-Analysis

Background: Previous preclinical studies have suggested that denosumab, a widely used anti-osteoporotic drug, may positively impact glucose metabolism by improving insulin sensitivity and glucose tolerance. However, clinical evidence demonstrating the association between denosumab use and the risk of diabetes mellitus (DM) remains limited.

Objectives: This study aimed to assess the risk of DM in patients treated with denosumab compared to those treated with bisphosphonates.

Methods: A nationwide population-based cohort study was conducted using Taiwan’s National Health Insurance Research Database. Patients aged 50 years and older, diagnosed with osteoporotic fractures, and treated with either denosumab or bisphosphonates between 2012 and 2020 in Taiwan, were included. Patients were categorized into denosumab and bisphosphonate groups based on their initial treatment according to prescription records. The primary outcome was newly diagnosed DM requiring antidiabetic drug treatment. Stabilized inverse probability of treatment weighting was used to balance baseline characteristics between groups and mitigate potential confounders, such as age, sex, income, comorbidities, Charlson comorbidity index, medications, and medical utilization. Cox proportional hazards models estimated the hazard ratios (HRs) for DM development. Additionally, a meta-analysis of the current evidence was conducted to strengthen our findings, using a random-effects model for data synthesis.

Results: The cohort study included 17,156 denosumab users and 28,213 bisphosphonate users. Denosumab use was not significantly associated with a lower risk of developing DM compared to bisphosphonate use (HR = 0.93, 95% CI: 0.83-1.05), consistent across age ( < 65 years or ≥ 65 years) and sex subgroups. The meta-analysis, which pooled data from five studies (three post-hoc analyses of RCTs using adverse event data, one cohort study, and our present cohort study) involving a total of 76,385 individuals, also showed a non-significant association between denosumab use and a lower risk of DM (HR = 0.84, 95% CI: 0.66-1.08) compared with bisphosphonate use. However, moderate between-study heterogeneity was observed, as indicated by an I-square value of 34.4%.

Conclusions: This cohort study and meta-analysis suggest that denosumab use is not significantly associated with the risk of DM compared to bisphosphonate use. Nonetheless, the observed between-study heterogeneity, along with the limited number of studies currently available, underscores the need for further research to evaluate this issue more comprehensively and to confirm our findings.