Obesity paradox in atrial fibrillation: a distracting reality or fictitious finding?

This editorial refers to ‘The ‘obesity paradox’ in atrial fibrillation: observations from the ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial’†, by R.K. Sandhu et al., on page 2869. Atrial fibrillation (AF) is the most commonly encountered arrhythmia, with increasing prevalence worldwide. The rising tide of AF has been heavily linked to obesity, which is also a contemporary global epidemic. A recent meta-analysis of cohort studies estimated a 29% increased incident AF risk with every 5-unit increment in body mass index (BMI).1 Further, obesity has been shown to contribute to progression of paroxysmal to persistent AF.2 Obesity leads to a large array of risk factors that are known contributors to AF development such as hypertension, diabetes mellitus, sleep apnoea, dyslipidaemia, and increased pericardial fat. Recent studies on the obese substrate for AF have unveiled unique adipose tissue infiltration from the epicardial adiposity together with increased interstitial fibrosis contributing to atrial conduction abnormalities and increased AF propensity.3–5 Despite the well-established causal association between obesity and AF, the study by Sandhu et al. published in this issue of the journal provided further surprising findings regarding the ‘obesity paradox’ whereby both higher BMI (>25 kg/m2) and waist circumference (>102 cm for men, >88 cm for women) were associated with lower all-cause mortality and stroke.6 These data came from a large cohort of 17 913 AF patients with at least one risk factor for stroke, who were randomized to either apixaban or warfarin anticoagulation in the ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial with median follow-up duration of 1.8 years. At baseline, significant differences in patient characteristics were present, with younger age, better renal function, more males, and lower CHA2DS2-VASc and HAS-BLED scores evident in those with higher BMI despite a higher burden of hypertension, diabetes, heart failure, and coronary artery disease, as well as larger left atrial dimensions in comparison with those in the normal BMI category. Further, as a result of the increased co-morbidities, medication use was significantly higher in the higher BMI patients, including aspirin, calcium channel blockers, beta-blockers, angiotensin-converting enzyme (ACE) inhibitor/angiotensin receptor blocker, lipid-lowering agent, and amiodarone. Of note, usage of digoxin, other antiarrhythmic drugs, and rhythm control strategies such as catheter ablation or cardioversion was not presented. The ‘obesity paradox’ presented in this report is susceptible to many residual confounding factors given the high number of non-matched baseline characteristics among the normal and higher BMI patients despite statistical adjustments undertaken by the authors. Of the list of differences mentioned earlier, one key concern is the older age in the normal BMI patients. Mortality data from a contemporary study of anticoagulated AF patients in ROCKET-AF suggested age ≥75 years as one of the strongest predictors for all-cause mortality (hazard ratio of 1.69).7 Interestingly, other studies that reported on the ‘obesity paradox’ in AF subjects also had significantly younger patients in the higher BMI category.8,9 Strikingly, the median age of higher BMI subjects in the only study that did not observe this ‘obesity paradox’ was not lower than in the normal BMI group.10 In addition, this study was one of only two (from a total of six) studies whereby the CHA2DS2-VASc score was not lower in the higher BMI category.10 Further studies utilizing age-matched cohorts are therefore needed to delineate this seeming paradox further. Other major residual confounders to the ‘obesity paradox’ that are present in similar studies may include: greater use of ACE inhibitors/angiotensin receptor blockers and lipid-lowering agents, as well as superior renal function in the higher BMI subjects. Importantly, one must not forget the likelihood of unmeasured confounding factors of this ‘obesity paradox’. For example, using only baseline BMI measures would not capture the dynamic nature of weight changes that may influence survival and outcome. Further, subjects in the normal or lower BMI category may have non-purposeful weight loss due to other serious medical conditions that may have an impact on survival. Similarly, subjects with normal BMI may have different underlying causes for their AF as opposed to the metabolic causes seen in the obese subjects that may confer differential response to therapy and outcome. Additionally, the often forgotten risk factor in cardiorespiratory fitness may also attenuate the mortality paradox in obese patients with coronary heart disease and heart failure.11 This is in keeping with our recent work demonstrating that cardiorespiratory fitness has a large impact on AF burden in both overweight and obese individuals.12 The signal of an ‘obesity paradox’ in AF may be one fictitious finding yet to be proven otherwise. However, one must not allow this to be a distraction towards our fight against obesity, a primary modifiable risk factor for so many other cardiovascular conditions including AF. Specifically, purposeful weight loss has been shown to be highly beneficial in reversing atrial remodelling, reducing AF symptom burden and severity as well as improving sinus rhythm maintenance in both randomized and prospective observational cohorts over long-term follow up.13–15 Further work is required to determine if these benefits seen with weight reduction will translate to a hard endpoint such as reduced mortality. Taken together, the ‘obesity paradox’ in AF should be approached with caution and a new paradigm of care beyond rate and rhythm management of AF should be actively pursued in the form of aggressive lifestyle modification strategies to facilitate purposeful weight reduction, increased physical activity, and improved cardiorespiratory fitness in overweight and obese AF individuals (Figure 1).5 The left panel highlights the impact of progressive increase in weight with the suggested possible mortality paradox. The right panel shows the benefit of weight loss and provides the targets for aggressive risk factor management that has been shown to be beneficial in patients with obesity and AF. AF, atrial fibrillation; AHI, apnoea–hypopnoea index; BMI, body mass index; CPAP, continuous positive airway pressure; HbA1c, glycated haemoglobin. D.H.L. is supported by a Robert J. Craig Lectureship from the University of Adelaide. M.E.M. is supported by a Robert J. Craig Scholarship from the University of Adelaide. P.S. is supported by a Practitioner Fellowships from the National Health and Medical Research Council of Australia and by the National Heart Foundation of Australia.