Abstract
OBJECTIVES: To evaluate the association between fragility fracture and interruption of bisphosphonate prescription after three and five years of initial prescription, and to explore the incidence of atypical femoral fracture and osteonecrosis of jaw after three and five years of continued prescription of bisphosphonates. DESIGN: Nested case-control and cohort studies. SETTING: Clinical Practice Research Datalink Aurum, an anonymised longitudinal database of NHS primary care electronic health records, linked to hospital admission and mortality records from Hospital Episode Statistics and Office for National Statistics databases, respectively, 1 January 1997 to 31 December 2022. PARTICIPANTS: Adults aged ≥18 years on the date of the first prescription of alendronate or risedronate, and with primary care data linked to Hospital Episode Statistics and Office for National Statistics databases. Participants with a medication possession ratio of ≥67% in each of the first three and five years were followed up from the three or five year time points to the earliest of fragility fracture, death, end of study, two years, or last data availability. Cases were those who had a fragility fracture and were matched with up to four control participants. MAIN OUTCOME MEASURES: Incidence of fragility fracture, atypical femoral fracture, and osteonecrosis of jaw were main outcomes. RESULTS: Data for 26 809 and 13 408 participants prescribed bisphosphonates for three and five years, respectively, were included. Fragility fracture was not associated with interruption of bisphosphonate prescription or intermittent prescription compared with continuous prescription in those previously prescribed bisphosphonates for three years (adjusted odds ratio 1.02, 95% confidence interval (CI) 0.85 to 1.24 and 1.13, 0.93 to 1.38, respectively) or five years (adjusted odds ratio 0.92, 0.75 to 1.13 and 0.87, 0.67 to 1.13, respectively). Time since last prescription was not associated with fragility fracture. After three and five years of bisphosphonate prescription, the incidence of atypical femoral fractures was 1.93 (95% CI 1.55 to 2.41) and 2.50 (1.92 to 3.26) per 1000 person years, respectively, and the incidence of osteonecrosis of jaw was 0.07 (0.02 to 0.21) and 0.00 per 1000 person years, respectively, within two years. CONCLUSION: Fragility fracture was not associated with interruption of bisphosphonate prescription for up to two years, after either three or five years of prescription. The incidence of atypical femoral fracture or osteonecrosis of jaw was uncommon and rare respectively.