Body mass index from the RE-LY trial: further evidence of the obesity paradox

Abstract

Abstract Background The obesity paradox has been reported in 3 post-hoc analyses evaluating the direct oral anticoagulants (DOAC) against warfarin (W): apixaban (ARISTOTLE), rivaroxaban (ROCKET), and edoxaban (ENGAGE-AF). Purpose To evaluate the obesity paradox in a post-hoc analysis of the RE-LY trial, comparing dabigatran 110 mg BID (D110), 150 mg BID (D150), and W by body mass index (BMI). Methods Baseline characteristics were evaluated using World Health Organization (WHO) criteria of overweight and obese (BMI ≥25 kg/m2) and under and normal weight (BMI <25 kg/m2). Stroke and systemic embolism, ischemic stroke, major bleeding, mortality, and intracranial hemorrhage were evaluated using BMI as a continuous variable and by the WHO criteria using a cox proportional hazard model. Results BMI was available in 99.9% of patients randomized; 74% had a BMI ≥25. At baseline, patients with a BMI ≥25 were younger (70.9 vs 73.1, p<0.001) and had fewer prior strokes (11.5% vs 15.6%, p<0.001), but higher mean creatinine clearance (78.3 vs 57.0, p<0.001) and rates of diabetes (25.8% vs 16.1%, p<0.001) (Table 1). Independent of drug assignment, patients with a BMI ≥25 had lower rates of stroke and systemic embolism (HR 0.65 [95% CI 0.54–0.79], p<0.001), ischemic stroke (0.75 [95% CI 0.60, 0.94], p=0.01), major bleeding (HR 0.79 [95% CI 0.69,0.89], p<0.001), mortality (HR 0.60 [95% CI 0.53, 0.67], p<0.001) and intracranial hemorrhage (HR 0.53 [95% CI 0.38, 0.73], p<0.001) compared to those with a BMI <25. Using BMI as a continuous variable combining all outcomes at 3 years, endpoint rates declined as BMI approached 25 in all 3 treatment groups. The exceptions were intracranial hemorrhage for D110 and D150 and stroke in D150 patients, where rates were low independent of BMI. No significant interaction between BMI and treatment was observed in individual outcomes except for the D110 vs. D150 comparison for major bleeding, in favor of D110 for patients with BMI ≥25 (HR 0.77 [95% CI 0.65, 0.91] and HR 1.12 [95% CI 0.86, 1.47], interaction p=0.0190). Conclusions In RE-LY, independent of drug assignment, patients with a higher BMI had improved outcomes, demonstrating the obesity paradox. As BMI increased towards 25, outcome rates improved except for intracranial hemorrhage rates for both D110 and D150 and ischemic stroke rates for D150, which were low independent of BMI. Patients treated with D110 with a BMI ≥25 kg/m2 had significantly lower rates of bleeding compared to D150. Funding Acknowledgement Type of funding sources: Foundation. Main funding source(s): Sharpe Strumia Foundation