All-cause mortality with current and past use of antidepressants or benzodiazepines after major osteoporotic and hip fracture

In the first year, after an osteoporotic fracture of a hip, forearm, upper arm, or spine, the dispensing rates of antidepressants and benzodiazepines increased significantly. After those fractures, recent and past use of antidepressants and benzodiazepines was associated with increased all-cause mortality; current use was not associated with mortality risk. INTRODUCTION: It remains unclear to what extent use of antidepressants and benzodiazepines is associated with mortality risk after a major osteoporotic fracture (MOF). We aimed to study the cumulative use of antidepressants and benzodiazepines during the year after MOF or hip fracture (HF) and whether the use was associated with mortality. METHODS: A cohort study was performed within the Dutch PHARMO Database Network including all patients aged 65+ with a first record of MOF (hip, humerus, forearm, and clinical vertebral fracture) between 2002 and 2011. Data were analyzed using Cox regression models, adjusted for comorbidities, and concomitant medication use and broken down to index fracture type. RESULTS: A total of 4854 patients sustained a first MOF, of whom 1766 patients sustained a HF. Mean follow-up was 4.6 years, divided in 30-day periods. The cumulative antidepressant and benzodiazepine use during the first year after MOF increased from 10.6 to 14.7% and from 24.0 to 31.4%, respectively. Recent (31-92 days before each follow-up period) and past use (> 92 days before) of antidepressants and benzodiazepines after MOF or HF was associated with an increased all-cause mortality risk but current use (< 30 days before) was not. CONCLUSION: There is a considerable increase in dispensing rate of antidepressants and benzodiazepines in the first year after a MOF. Recent and past use of these medications was associated with all-cause mortality. The finding that current use was not associated with mortality should be further explored and may probably be explained by the healthy survivor's bias