Association of Body Mass Index With Care and Outcomes in Patients With Atrial Fibrillation: Results From the ORBIT-AF Registry.

OBJECTIVES: This study sought to determine the association between body mass index (BMI) and clinical outcomes among patients with prevalent atrial fibrillation (AF). BACKGROUND: Higher BMI is an independent risk factor for incident AF. However, its impact on management strategies and clinical outcomes among patients with prevalent AF is unclear. METHODS: Patients with AF enrolled in the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) registry from June 2010 through August 2011 were stratified into BMI-based categories as normal weight, overweight, class I obese, class II obese, and class III obese. Unadjusted and adjusted Cox frailty models were constructed to assess the association of BMI with clinical outcomes over a 2-year follow-up. RESULTS: We evaluated 9,606 patients with AF (42% women; 78% overweight/obese) from 174 ORBIT participating practices in the United States. Higher BMI patients were younger and had a greater prevalence of diabetes, hypertension, and obstructive sleep apnea (OSA). Use of anticoagulation and rhythm control strategies was significantly greater among higher BMI patients. Rates for all-cause mortality and thromboembolic events decreased in a near linear fashion across increasing BMI categories (p < 0.001). After multivariable adjustment, higher BMI was associated with lower risk for all-cause mortality with lowest risk among class I obese patients (hazard ratio [HR]: 0.65; 95% CI: 0.54 to 0.78); reference: normal weight). For every 5-kg/m2 increase in BMI, the odds of risk-adjusted mortality were 7% lower. In contrast, BMI was not associated with adjusted risk for thromboembolic events and AF progression. CONCLUSIONS: Although AF patients with higher BMI were significantly younger, higher BMI in AF patients was associated with similar or better clinical outcomes.

Duke Scholars

Author Laine Elliott Thomas Biostatistics & Bioinformatics, Division of Biostatistics

Author Jonathan Paul Piccini Sr. Medicine, Cardiology