Article Title

Bisphosphonate Drug Holiday and Fracture Risk

Publication Date



osteoporosis, fracture risk


Background/Aims: Among women with ≥ 3 years exposure to bisphosphonates (BPs), we compared the incidence of fragility fractures in those who discontinued BPs for ≥ 12 months (drug holiday) to those who continued to use BPs (persistent use).

Methods: This retrospective cohort study included women aged ≥ 45 years who initiated BP use from four Kaiser Permanente regions between January 1, 1998, and December 31, 2009. Drug holiday was defined as ≥ 12 months with BP use at 0% adherence. Persistent use status required ongoing use at ≥ 50% adherence. The primary outcome of interest was the first occurrence of an incident clinical osteoporosis-related fragility fracture, identified from the electronic medical record (EMR) via ICD-9-CM codes. All subjects were followed until fracture, disenrollment from the health plan, death, or December 31, 2012. From the EMR, we collected information on the following potential confounders and effect modifiers: race/ethnicity, age, body mass index, comorbidities, history of previous fragility fracture, lowest T-score prior to cohort entry, fall risk, 10-year fracture risk, and prior/concomitant use of bone-active medications. Persistent users and drug holiday subjects were compared with regard to several demographic and clinical characteristics. Time-varying Cox proportional hazards models were used to compare osteoporosis-related fracture incidence between the two groups.

Results: The cohort of 28,620 women, observed for 111,997 person-years, included 17,123 (59.8%) persistent BP users and 11,497 (40.2%) drug holiday subjects. The drug holiday group had fewer comorbidities, higher baseline T-scores, and lower fracture and fall risk scores. A total of 3,571 osteoporosis-related fractures were observed. The unadjusted rate ratio (RR) for any osteoporosis-related fractures for drug holiday compared to persistent use was 0.87 (95% confidence interval [CI]: 0.81–0.94), but was 1.0 (95% CI: 0.9–1.2) for hip fractures only. The time-varying models suggested no differences in fracture risk (hazard ratio [HR]: 0.90, 95% CI: 0.80–1.00) after adjustment for baseline fall and fracture risk, comorbidities and other bone-active medication use. Similarly, no difference in hip fracture risk was observed (HR: 0.84, 95% CI: 0.68–1.03).

Discussion: Women who undertake a holiday from BP use are not at greater risk of osteoporosis-related fragility fractures, nor hip fractures specifically, than are women who continue to use BPs persistently.