Comparison of Outcomes in Patients With Atrial Fibrillation Under Oral Anticoagulation Therapy Analyzed by Body Weight (<60, 60 to 100, and >100 kg)

There is limited knowledge regarding the efficacy and safety of fixed-dose oral anticoagulants in overweight patients because of the possible increased risk of embolism and hemorrhage. This study aimed to evaluate embolic, hemorrhagic, and mortality events in anticoagulated patients, administered both antivitamin K and direct oral anticoagulants based on the body weight (<60 kg, 60 to 100 kg and >100 kg). A retrospective registry-based cohort study including all consecutive patients with a diagnosis of atrial fibrillation between January 2014 and January 2018 in the health area of Vigo (Galicia, Spain) was used (CardioCHUVI‐AF registry; ClinicalTrials.gov identifier: NCT04364516). The final cohort comprised 11,821 AF patients. The cohort was classified into 3 categories: low body weight ([LBW], <60 kg, 924 patients); middle body weight (60 to 100 kg, 9,546 patients); and high body weight ([HBW], >100 kg, 958 patients). Outcomes were predicted using the Fine and Gray model and Cox proportional hazards model when appropriate. Middle body weight was the reference group. No association was found between the weight and major bleeding in the univariate analyses: LBW with a sub-distribution hazard ratio (sHR) of 1.13 (95% confidence interval [CI] 0.92 to 1.41), and HBW with an sHR of 1.02 (95% CI 0.83 to 1.26). Stroke/systemic embolism events occurred in 817 patients (6.6%). In the univariate analyses, we found an association between weight and risk of stroke/systemic embolism: LBW sHR 1.37 (95% CI 1.09 to 1.72), and HBW sHR 0.66 (95% CI 0.49 to 0.89) but no association was found in the multivariable model. The same situation was observed with all-cause death: in the univariable model, LBW presented a hazard ratio of 1.48 (95% CI 1.31 to 1.68) and the HBW group presented a hazard ratio of 0.53 (95%CI 0.44 to 0.63) whereas no significant association was found in the multivariable model. We conclude that in our registry, extreme weights were not related to more events during follow-up.