Research fellow Institute of Clinical Pharmacy and Pharmaceutical Sciences (CPPS), College of Medicine, National Cheng Kung University, Tainan, Taiwan., Taiwan (Republic of China)
Background: Few studies suggested denosumab may provide benefits for impaired glucose tolerance by reducing fasting plasma glucose and improving insulin resistance. However, the association has hitherto not been evaluated by a large population.
Objectives: To compare the incidence of diabetes in patients with osteoporosis who persisting vs. non-persisted to denosumab.
Methods: We designed a cohort study using National Health Insurance Database (NHID) in Taiwan. We included patients aged 18 and older who were exposed to denosumab for osteoporosis between 2012-2019. The denosumab regimen for osteoporosis patients lasts for six months, thus we determined the lag-time period to be 225 days following the first dose of denosumab (6-month regimen plus 45 days). The index date was established as being 225 days following the first dose. Patients who received the second dose during the lag-time interval were assigned to the persistent group; if not, they were assigned to the non-persistent group. We defined the primary outcome as the use of diabetes medications. We applied 1-year washout period before the index date and excluded patients who had a record of diabetes or anti-diabetic medications to ensure the event was incident. We conducted the as-treated analysis and followed the patients from the index date until the first occurrence of the outcomes, discontinuation, end of data source, or death whichever came first. We performed propensity score matching to balance the baseline covariates, including comorbidities, medications, inflammation status, and healthcare utilization. We performed Cox models to generate HR and 95%CIs to compare the incidence of diabetes between the persistent and non-persistent groups. We also conducted ITT analysis for sensitivity analysis. We considered the outcome of the seizure as a negative control for this study.
Results: We included 10592 patients in the persistent and 6365 patients in the non-persistent group. The mean age and proportion of females were similar between the two groups (74.8 years and 84.8% in the persistent group and 76.8 years and 80.2% in the non-persistent group). The incidence rate of the main outcome is 4.2 per 100 person-years (PYs) in the persistent group and 5.15 per 100 PYs in the non-persistent group. Persistent users had a lower risk of using anti-diabetes agents than non-persistent users (HR: 0.79, 95% CI: 0.69-0.89). The result remained consistent in the sensitivity analysis (HR: 0.9 95% CI: 0.81-0.99). We didn’t observe significant differences in the negative control analysis (HR: 1.00, 95% CI: 0.92-1.09).
Conclusions: The findings suggested the incidence of diabetes was lower in patients persisting vs. non-persisted to denosumab, suggesting the possible favorable effect to the impaired glucose tolerance.
