Skip to main content

Comparative fracture risk during osteoporosis drug holidays after long-term risedronate versus alendronate therapy: a propensity score-matched cohort study

Hayes KN, Brown KA, Cheung AM, Kim SA, Juurlink DN, Cadarette SM. Ann Intern Med. 2022; 175(3):335-43. Epub 2022 Jan 11. DOI:

Background — An osteoporosis drug holiday is recommended for most patients after 3 to 5 years of therapy. Risedronate has a shorter half-life than alendronate, and thus the residual length of fracture protection may be shorter.

Objective — To examine the comparative risks of drug holidays after long-term (≥3 years) risedronate versus alendronate therapy.

Design — Population-based, matched, cohort study.

Setting — Province-wide health care administrative databases providing comprehensive coverage to Ontario residents aged 65 years or older between November 2000 and March 2020.

Patients — Persons aged 66 years or older who had long-term risedronate therapy and a drug holiday were matched 1:1 on propensity score to those who had long-term alendronate therapy and a drug holiday.

Measurements — The primary outcome was hip fracture within 3 years after a 120-day ascertainment period. Secondary analyses included shorter follow-up and sex-specific estimates. Cox proportional hazards models were used to estimate hazard ratios (HRs) for fracture risk between groups.

Results — A total of 25 077 propensity score-matched pairs were eligible (mean age, 81 years; 81% women). Hip fracture rates were higher among risedronate than alendronate drug holidays (12.4 and 10.6 events, respectively, per 1000 patient-years; HR, 1.18 [95% CI, 1.04 to 1.34]; 915 total hip fractures). The association was attenuated when any fracture was included as the outcome (HR, 1.07 [CI, 1.00 to 1.16]) and with shorter drug holidays (1 year: HR, 1.03 [CI, 0.85 to 1.24]; 2 years: HR, 1.14 [CI, 0.96 to 1.32]).

Limitation — Analyses were limited to health care administrative data (potential unmeasured confounding), and some secondary analyses contained few events.

Conclusion — Drug holidays after long-term therapy with risedronate were associated with a small increase in risk for hip fracture compared with alendronate drug holidays. Future research should examine how best to mitigate this risk.