Abstract
BACKGROUND: This cohort study aimed to estimate incidence rates of femoral shaft fracture in patients who were treated with antiresorptive drugs. METHODS: We used data from the National Database of Health Insurance Claims of Japan from April 2009 and October 2016. All patients with new use of an antiresorptive drug, prescription-free period of ≥3 months, and no prior femoral fractures were included. Femoral shaft fractures were identified using a validated definition based on International Classification of Diseases, 10(th) revision (ICD-10) codes. Incidence rate ratios were estimated using Poisson regression, with adjustment for sex, age, and the Charlson Comorbidity Index. RESULTS: We identified 7,958,655 patients (women: 88.4%; age ≥75 years: 51.2%). Femoral shaft fractures were identified in 22,604 patients. Incidence rates per 100,000 person-years were 74.8 for women, 30.1 for men, 30.1 for patients aged ≤64 years, 47.7 for patients aged 65-74 years, and 99.0 for patients aged ≥75 years. Adjusted incidence rate ratios in patients taking versus not taking each type of antiresorptive drug were 1.00 (95% confidence interval [CI], 0.98-1.03) for bisphosphonates, 0.46 (95% CI, 0.44-0.48) for selective estrogen receptor modulators, 0.24 (95% CI, 0.18-0.32) for estrogens, 0.75 (95% CI, 0.71-0.79) for calcitonins, and 0.93 (95% CI, 0.84-1.03) for denosumab. The adjusted incidence rate ratio for alendronate was 1.18 (95% CI, 1.14-1.22). CONCLUSION: The incidence rates of femoral shaft fracture varied across patients treated with different antiresorptive drugs. Further research on a specific antiresorptive drug can increase understanding of the risk of femoral shaft fracture.