In Reply We thank Uzoigwe and colleagues for their careful consideration of our article,1 which reported that among patients with atrial fibrillation (AF), fracture risk was lower among users of direct oral anticoagulants (DOACs), particularly apixaban, compared with users of warfarin. Uzoigwe and colleagues suggest an alternate explanation for our findings: that confounding by use of proton pump inhibitors (PPIs) may underlie the differences we observed. This hypothesis is intriguing, given the high prevalence of PPI prescriptions among older adults2,3 combined with evidence linking PPIs to fracture risk.4 We had not previously explicitly considered PPIs; however, confounding by PPI use should have been mitigated through our rigorous approach to minimize confounding, particularly via the use of high-dimensional propensity scores, which incorporated all available information on pharmacy prescription fills, inpatient and outpatient diagnostic codes, and inpatient and outpatient procedure codes.