DOACs associated with lower risk of fracture than warfarin in atrial fibrillation

Patients taking apixaban, dabigatran, or rivaroxaban had lower rates of hip and vertebral fractures than those taking warfarin, with no difference seen between the direct oral anticoagulants (DOACs).


Direct oral anticoagulants (DOAC) were associated with lower risk of osteoporotic fracture compared with warfarin in patients with atrial fibrillation, with no difference found between the DOACs, according to a retrospective study.

To compare the risk for osteoporotic fracture between anticoagulants, researchers used a database of the Hong Kong Hospital Authority to study patients who were newly diagnosed with atrial fibrillation between 2010 and 2017 and were prescribed warfarin or a DOAC (apixaban, dabigatran, or rivaroxaban). Osteoporotic hip and vertebral fractures were compared using propensity score-weighted cumulative incidence differences (CIDs) among 23,515 patients (3,241 apixaban users, 6,867 dabigatran users, 3,866 rivaroxaban users, and 9,541 warfarin users). Findings were published May 19 by Annals of Internal Medicine.

Over a median follow-up of 423 days, there were 401 fractures (crude event number [weighted rate per 100 patient-years]: 53 [0.82] for apixaban; 95 [0.76] for dabigatran; 57 [0.67] for rivaroxaban; and 196 [1.11] for warfarin). All of the DOACs were associated with a lower risk for fracture than warfarin (CIDs, −0.88% [95% CI, −1.66% to −0.21%] for apixaban, 0.81% [95% CI, −1.34% to −0.23%] for dabigatran, and −1.13% [95% CI, −1.67% to −0.53%] for rivaroxaban). There were no significant differences in head-to-head comparisons between DOACs at 24 months. The CIDs were −0.06% (95% CI, −0.69% to 0.49%) for apixaban versus dabigatran, −0.32% (95% CI, −0.84% to 0.18%) for rivaroxaban versus dabigatran, and −0.25% (95% CI, −0.86% to 0.40%) for rivaroxaban versus apixaban.

The authors noted that these findings may help inform the benefit-risk assessment when choosing anticoagulants. They acknowledged that the study had limited power to compare between DOACs and could only rule out more than a twofold higher or a 50% lower relative risk for osteoporotic fractures between individual DOACs. In addition, they noted, the risk for osteoporotic fractures is often overlooked during the selection of an oral anticoagulant for patients with atrial fibrillation, even though surgery is often required to treat a fracture and perioperative management of anticoagulation is difficult.

“Given the supportive evidence from experimental settings, findings from our study using clinical data, and the indirect evidence provided by the previous meta-analysis of randomized controlled trials, there exists a compelling case for evaluating whether the risk for osteoporotic fractures should be considered at the point of prescribing an oral anticoagulant to minimize fracture risk,” they concluded.