Association Between Obesity-Mediated Atrial Fibrillation and Therapy With Sodium Channel Blocker Antiarrhythmic Drugs | Atrial Fibrillation | JAMA Cardiology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Patient Flow Diagram of the Antiarrhythmic Drug (AAD) Study Cohort
Patient Flow Diagram of the Antiarrhythmic Drug (AAD) Study Cohort

AF indicates atrial fibrillation.

Figure 2.  Mice Characteristics and Response to Pacing-Induced Atrial Fibrillation (AF)
Mice Characteristics and Response to Pacing-Induced Atrial Fibrillation (AF)

A, Mouse weights in control and diet-induced obese cohorts. B, Atrial electrogram showing sinus rhythm at baseline and pacing-induced AF in mouse with obesity. C, Burden (duration) of pacing-induced AF among control mice and those with obesity. D, Incidence of AF in the obese vs nonobese group. E, AF reduction (%) in mice with obesity after treatment with flecainide and sotalol.

Table 1.  Clinical Characteristics of Patients Treated With Class I and III Antiarrhythmic Drugs for Symptomatic Atrial Fibrillation
Clinical Characteristics of Patients Treated With Class I and III Antiarrhythmic Drugs for Symptomatic Atrial Fibrillation
Table 2.  Multivariate Logistic Regression Analysis of Failure to Respond to Class I and III AADs for Symptomatic Atrial Fibrillation
Multivariate Logistic Regression Analysis of Failure to Respond to Class I and III AADs for Symptomatic Atrial Fibrillation
1.
Benjamin  EJ, Wolf  PA, D’Agostino  RB, Silbershatz  H, Kannel  WB, Levy  D.  Impact of atrial fibrillation on the risk of death: the Framingham Heart Study.  Circulation. 1998;98(10):946-952. doi:10.1161/01.CIR.98.10.946PubMedGoogle ScholarCrossref
2.
Flegal  KM, Carroll  MD, Kit  BK, Ogden  CL.  Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010.  JAMA. 2012;307(5):491-497. doi:10.1001/jama.2012.39PubMedGoogle ScholarCrossref
3.
Chatterjee  NA, Giulianini  F, Geelhoed  B,  et al.  Genetic obesity and the risk of atrial fibrillation: causal estimates from mendelian randomization.  Circulation. 2017;135(8):741-754. doi:10.1161/CIRCULATIONAHA.116.024921PubMedGoogle ScholarCrossref
4.
January  CT, Wann  LS, Alpert  JS,  et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines.  2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society.  J Am Coll Cardiol. 2014;64(21):e1-e76. doi:10.1016/j.jacc.2014.03.022PubMedGoogle ScholarCrossref
5.
Duytschaever  M, Demolder  A, Phlips  T,  et al.  Pulmonary vein isolation with vs without continued antiarrhythmic drug treatment in subjects with recurrent atrial fibrillation (POWDER AF): results from a multicentre randomized trial.  Eur Heart J. 2018;39(16):1429-1437. doi:10.1093/eurheartj/ehx666PubMedGoogle ScholarCrossref
6.
Darghosian  L, Free  M, Li  J,  et al.  Effect of omega-three polyunsaturated fatty acids on inflammation, oxidative stress, and recurrence of atrial fibrillation.  Am J Cardiol. 2015;115(2):196-201. doi:10.1016/j.amjcard.2014.10.022PubMedGoogle ScholarCrossref
7.
Lieve  KV, Verkerk  AO, Podliesna  S,  et al.  Gain-of-function mutation in SCN5A causes ventricular arrhythmias and early onset atrial fibrillation.  Int J Cardiol. 2017;236:187-193. doi:10.1016/j.ijcard.2017.01.113PubMedGoogle ScholarCrossref
8.
Musa  H, Kline  CF, Sturm  AC,  et al.  SCN5A variant that blocks fibroblast growth factor homologous factor regulation causes human arrhythmia.  Proc Natl Acad Sci U S A. 2015;112(40):12528-12533. doi:10.1073/pnas.1516430112PubMedGoogle ScholarCrossref
9.
Darbar  D, Kannankeril  PJ, Donahue  BS,  et al.  Cardiac sodium channel (SCN5A) variants associated with atrial fibrillation.  Circulation. 2008;117(15):1927-1935. doi:10.1161/CIRCULATIONAHA.107.757955PubMedGoogle ScholarCrossref
10.
Ellinor  PT, Nam  EG, Shea  MA, Milan  DJ, Ruskin  JN, MacRae  CA.  Cardiac sodium channel mutation in atrial fibrillation.  Heart Rhythm. 2008;5(1):99-105. doi:10.1016/j.hrthm.2007.09.015PubMedGoogle ScholarCrossref
11.
Makiyama  T, Akao  M, Shizuta  S,  et al.  A novel SCN5A gain-of-function mutation M1875T associated with familial atrial fibrillation.  J Am Coll Cardiol. 2008;52(16):1326-1334. doi:10.1016/j.jacc.2008.07.013PubMedGoogle ScholarCrossref
12.
Yoshida  K, Solomon  DH, Kim  SC.  Active-comparator design and new-user design in observational studies.  Nat Rev Rheumatol. 2015;11(7):437-441. doi:10.1038/nrrheum.2015.30PubMedGoogle ScholarCrossref
13.
Chalazan  B, Mol  D, Sridhar  A,  et al.  Genetic modulation of atrial fibrillation risk in a Hispanic/Latino cohort.  PLoS One. 2018;13(4):e0194480. doi:10.1371/journal.pone.0194480PubMedGoogle Scholar
14.
Alzahrani  Z, Ornelas-Loredo  A, Darbar  SD,  et al.  Association between family history and early-onset atrial fibrillation across racial and ethnic groups.  JAMA Netw Open. 2018;1(5):e182497. doi:10.1001/jamanetworkopen.2018.2497PubMedGoogle Scholar
15.
Nagueh  SF, Smiseth  OA, Appleton  CP,  et al.  Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.  J Am Soc Echocardiogr. 2016;29(4):277-314. doi:10.1016/j.echo.2016.01.011PubMedGoogle ScholarCrossref
16.
do Carmo  JM, da Silva  AA, Rushing  JS, Pace  B, Hall  JE.  Differential control of metabolic and cardiovascular functions by melanocortin-4 receptors in proopiomelanocortin neurons.  Am J Physiol Regul Integr Comp Physiol. 2013;305(4):R359-R368. doi:10.1152/ajpregu.00518.2012PubMedGoogle ScholarCrossref
17.
Li  XF, Lytton  J.  An essential role for the K+-dependent Na+/Ca2+-exchanger, NCKX4, in melanocortin-4-receptor–dependent satiety.  J Biol Chem. 2014;289(37):25445-25459. doi:10.1074/jbc.M114.564450PubMedGoogle ScholarCrossref
18.
Banach  M, Popławska  M, Borowicz-Reutt  KK.  Sotalol enhances the anticonvulsant action of valproate and diphenylhydantoin in the mouse maximal electroshock model.  Pharmacol Rep. 2017;69(6):1173-1177. doi:10.1016/j.pharep.2017.05.005PubMedGoogle ScholarCrossref
19.
Martin  CA, Zhang  Y, Grace  AA, Huang  CL.  In vivo studies of Scn5a+/− mice modeling Brugada syndrome demonstrate both conduction and repolarization abnormalities.  J Electrocardiol. 2010;43(5):433-439. doi:10.1016/j.jelectrocard.2010.05.015PubMedGoogle ScholarCrossref
20.
Kim  S, Popkin  BM.  Commentary: understanding the epidemiology of overweight and obesity—a real global public health concern.  Int J Epidemiol. 2006;35(1):60-67. doi:10.1093/ije/dyi255PubMedGoogle ScholarCrossref
21.
Hayashi  K, Konno  T, Tada  H,  et al.  Functional characterization of rare variants implicated in susceptibility to lone atrial fibrillation.  Circ Arrhythm Electrophysiol. 2015;8(5):1095-1104. doi:10.1161/CIRCEP.114.002519PubMedGoogle ScholarCrossref
22.
Kim  GE, Ross  JL, Xie  C,  et al.  LKB1 deletion causes early changes in atrial channel expression and electrophysiology prior to atrial fibrillation.  Cardiovasc Res. 2015;108(1):197-208. doi:10.1093/cvr/cvv212PubMedGoogle ScholarCrossref
23.
Lin  CS, Pan  CH.  Regulatory mechanisms of atrial fibrotic remodeling in atrial fibrillation.  Cell Mol Life Sci. 2008;65(10):1489-1508. doi:10.1007/s00018-008-7408-8PubMedGoogle ScholarCrossref
24.
Van Wagoner  DR.  Oxidative stress and inflammation in atrial fibrillation: role in pathogenesis and potential as a therapeutic target.  J Cardiovasc Pharmacol. 2008;52(4):306-313. doi:10.1097/FJC.0b013e31817f9398PubMedGoogle ScholarCrossref
25.
Darbar  D, Roden  DM.  Symptomatic burden as an endpoint to evaluate interventions in patients with atrial fibrillation.  Heart Rhythm. 2005;2(5):544-549. doi:10.1016/j.hrthm.2005.01.028PubMedGoogle ScholarCrossref
26.
Martin  DT, Bersohn  MM, Waldo  AL,  et al; IMPACT Investigators.  Randomized trial of atrial arrhythmia monitoring to guide anticoagulation in patients with implanted defibrillator and cardiac resynchronization devices.  Eur Heart J. 2015;36(26):1660-1668. doi:10.1093/eurheartj/ehv115PubMedGoogle ScholarCrossref
27.
Israel  CW, Grönefeld  G, Ehrlich  JR, Li  YG, Hohnloser  SH.  Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care.  J Am Coll Cardiol. 2004;43(1):47-52. doi:10.1016/j.jacc.2003.08.027PubMedGoogle ScholarCrossref
28.
Euler  DE, Friedman  PA.  Atrial arrhythmia burden as an endpoint in clinical trials: is it the best surrogate? lessons from a multicenter defibrillator trial.  Card Electrophysiol Rev. 2003;7(4):355-358. doi:10.1023/B:CEPR.0000023138.85821.63PubMedGoogle ScholarCrossref
29.
Calkins  H, Kuck  KH, Cappato  R,  et al; Heart Rhythm Society Task Force on Catheter and Surgical Ablation of Atrial Fibrillation.  2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design.  Heart Rhythm. 2012;9(4):632-696.e21, e621. doi:10.1016/j.hrthm.2011.12.016PubMedGoogle ScholarCrossref
30.
Inohara  T, Shrader  P, Pieper  K,  et al.  Association of atrial fibrillation clinical phenotypes with treatment patterns and outcomes: a multicenter registry study.  JAMA Cardiol. 2018;3(1):54-63. doi:10.1001/jamacardio.2017.4665PubMedGoogle ScholarCrossref
31.
Darbar  D, Motsinger  AA, Ritchie  MD, Gainer  JV, Roden  DM.  Polymorphism modulates symptomatic response to antiarrhythmic drug therapy in patients with lone atrial fibrillation.  Heart Rhythm. 2007;4(6):743-749. doi:10.1016/j.hrthm.2007.02.006PubMedGoogle ScholarCrossref
32.
Parvez  B, Chopra  N, Rowan  S,  et al.  A common β1-adrenergic receptor polymorphism predicts favorable response to rate-control therapy in atrial fibrillation.  J Am Coll Cardiol. 2012;59(1):49-56. doi:10.1016/j.jacc.2011.08.061PubMedGoogle ScholarCrossref
33.
Echouffo-Tcheugui  JB, Shrader  P, Thomas  L,  et al.  Care patterns and outcomes in atrial fibrillation patients with and without diabetes: ORBIT-AF Registry.  J Am Coll Cardiol. 2017;70(11):1325-1335. doi:10.1016/j.jacc.2017.07.755PubMedGoogle ScholarCrossref
34.
Chalazan  B, Dickerman  D, Sridhar  A,  et al.  Relation of body mass index to symptom burden in patients with atrial fibrillation.  Am J Cardiol. 2018;122(2):235-241. doi:10.1016/j.amjcard.2018.04.011PubMedGoogle ScholarCrossref
35.
McCauley  MD, Darbar  D.  Race and socioeconomic status regulate lifetime risk of atrial fibrillation.  Circ Arrhythm Electrophysiol. 2018;11(7):e006584. doi:10.1161/CIRCEP.118.006584PubMedGoogle Scholar
Original Investigation
November 27, 2019

Association Between Obesity-Mediated Atrial Fibrillation and Therapy With Sodium Channel Blocker Antiarrhythmic Drugs

Author Affiliations
  • 1Division of Cardiology, Department of Medicine, University of Illinois at Chicago
  • 2Division of Epidemiology and Biostatistics, University of Illinois at Chicago
  • 3Department of Medicine, Jesse Brown Veterans Affairs Medical Center, Chicago, Illinois
JAMA Cardiol. 2020;5(1):57-64. doi:10.1001/jamacardio.2019.4513
Key Points

Question  Does obesity mediate response to sodium channel vs potassium channel blocker antiarrhythmic drugs in patients with atrial fibrillation and in mice with diet-induced obesity?

Findings  In this cohort study of 311 patients, those with obesity had greater recurrence (30%) of atrial fibrillation compared with those who were not obese who received sodium channel blocker antiarrhythmic drugs (6%); significant factors associated with failure to respond to antiarrhythmic drugs were use of sodium channel blockers, obesity, female sex, and hyperthyroidism. Mice with obesity showed reduced association of flecainide acetate in suppressing pacing-induced atrial fibrillation vs sotalol hydrochloride.

Meaning  Possible reduced response to sodium channel blocker antiarrhythmic drugs in suppressing atrial fibrillation in patients with obesity has important clinical implications.

Abstract

Importance  The association between obesity, an established risk factor for atrial fibrillation (AF), and response to antiarrhythmic drugs (AADs) remains unclear.

Objective  To test the hypothesis that obesity differentially mediates response to AADs in patients with symptomatic AF and in mice with diet-induced obesity (DIO) and pacing induced AF.

Design, Setting, and Participants  An observational cohort study was conducted including 311 patients enrolled in a clinical-genetic registry. Mice fed a high-fat diet for 10 weeks were also evaluated. The study was conducted from January 1, 2018, to June 2, 2019.

Main Outcomes and Measures  Symptomatic response was defined as continuation of the same AAD for at least 3 months. Nonresponse was defined as discontinuation of the AAD within 3 months of initiation because of poor symptomatic control of AF necessitating alternative rhythm control therapy. Outcome measures in DIO mice were pacing-induced AF and suppression of AF after 2 weeks of treatment with flecainide acetate or sotalol hydrochloride.

Results  A total of 311 patients (mean [SD] age, 65 [12] years; 120 women [38.6%]) met the entry criteria and were treated with a class I or III AAD for symptomatic AF. Nonresponse to class I AADs in patients with obesity was less than in those without obesity (30% [obese] vs 6% [nonobese]; difference, 0.24; 95% CI, 0.11-0.37; P = .001). Both groups had similar symptomatic response to a potassium channel blocker AAD. On multivariate analysis, obesity, AAD class (class I vs III AAD [obese] odds ratio [OR], 4.54; 95% Wald CI, 1.84-11.20; P = .001), female vs male sex (OR, 2.31; 95% Wald CI, 1.07-4.99; P = .03), and hyperthyroidism (OR, 4.95; 95% Wald CI, 1.23-20.00; P = .02) were significant indicators of the probability of failure to respond to AADs. Pacing induced AF in 100% of DIO mice vs 30% (P < .001) in controls. Furthermore, DIO mice showed a greater reduction in AF burden when treated with sotalol compared with flecainide (85% vs 25%; P < .01).

Conclusions and Relevance  Results suggest that obesity differentially mediates response to AADs in patients and in mice with AF, possibly reducing the therapeutic effectiveness of sodium channel blockers. These findings may have implications for the management of AF in patients with obesity.

Introduction

Atrial fibrillation (AF), the most common arrhythmia worldwide, is associated with significant morbidity and increased mortality.1 A major factor in the increase in incidence and prevalence of AF is the obesity epidemic. Although two-thirds of US adults are overweight or obese (body mass index [BMI] >30 [calculated as weight in kilograms divided by height in meters squared]),2 a causal inference between BMI and incident AF was reported only recently in a mendelian randomization study.3

Despite advances in catheter-based therapy, antiarrhythmic drugs (AADs) are still commonly used to treat symptomatic AF.4,5 However, response to membrane-active drugs is variable, with approximately 50% of patients experiencing symptomatic recurrence of AF within 6 months.6 This variability is due to heterogeneity of the underlying substrate and failure to select mechanism-based therapies. Although the underlying pathophysiologic factors by which obesity causes AF are poorly understood, emerging evidence supports modulation of the cardiac sodium channel, Nav1.5, as one potential mechanism. An increase or decrease in Nav1.5 expression in animal models is associated with AF implicating sodium channel remodeling and obesity-induced AF.7,8 Furthermore, mutations in SCN5A, encoding Nav1.5, are associated with familial AF.9-11 However, the role of the cardiac sodium channel in modulating response to AADs, especially class I agents (sodium channel blockers), in patients with obesity with AF remains unclear. We hypothesized that obesity differentially mediates response to antiarrhythmic therapy in patients with symptomatic AF and in mice with diet-induced obesity (DIO).

Methods
Study Design and Population

An observational cohort study was conducted to assess the association between obesity, an established risk factor for AF, and response to AADs. Observational studies assessing AAD therapy are limited by confounding by indication.12 We used a prevalent disease with new drugs (new user design) including only patients who began receiving the AAD during the study period so that pretreatment characteristics can be assessed and response can be described. Patients were prospectively enrolled in the University of Illinois at Chicago AF Registry, a clinical and genetic biorepository.13,14 Inclusion criteria included age older than 18 years, a documented history of AF, and attempted maintenance of sinus rhythm with a class I or III AAD. Detailed medical, drug, and family histories were obtained in all patients at enrollment, and patients completed a questionnaire. All patients experiencing symptoms after beginning AAD therapy for AF were given a 30-day event recorder to look for AF, and 12-lead electrocardiograms (ECGs) were performed at every clinic visit. At enrollment, an echocardiogram was performed in all patients. This study was conducted from January 1, 2018, to June 1, 2019. Written informed consent was obtained under a University of Illinois at Chicago Institutional Review Board–approved protocol. The participants did not receive financial compensation.

Definitions

Hypertension was defined by a history and/or the presence of therapy. Criteria for coronary artery disease included a history of myocardial infarction, angina, previous coronary bypass surgery or percutaneous coronary intervention, and drug treatment. Congestive heart failure was defined by a history and/or drug treatment. Obstructive sleep apnea was diagnosed by a sleep study or continuous positive airway pressure therapy. Left atrial and left ventricular measurements were made according to the American Society of Echocardiography.15 Obesity was defined as BMI of 30 or greater and nonobesity as BMI of less than 30. Both QRS and QTc interval duration were measured at baseline and at maximum doses of sodium channel blocker or potassium channel blocker AADs. The maximum daily dose of the AAD used at the end of the study was also recorded.

Response to AAD Therapy in Patients With AF

Response to therapy was defined prospectively as symptomatic response when the patient continued to receive the same class I or III drug at maximum tolerated doses for at least 3 months. Nonresponse was defined as discontinuation of the AAD at maximum tolerated doses within 3 months of initiation because of symptomatic recurrence of AF necessitating a change to another therapy. Symptomatic AF recurrence was assessed by 12-lead ECG, Holter monitor, and/or a 30-day event recorder.

Mouse Model of Obesity

Animal studies were performed according to protocols approved by the University of Illinois at Chicago Institutional Animal Care and Use Committee. Mice with DIO were bred on a C57BL/6J genetic background and were fed a high-fat diet (60% fat; Teklad 06414, Envigo) starting at age 8 weeks for 10 weeks. Wild-type (lean) controls were fed a regular mouse diet. Mice were considered obese when their weight was more than 30 g.16,17 Female mice fed a high-fat diet for 10 weeks failed to reach the 30-g obesity threshold. Thus, our studies were conducted only in male mice.

Mice were anesthetized with isoflurane, 1% to 2%, in 95% oxygen. Transesophageal (TE) pacing was performed using a 1.1-F octapolar catheter (Millar Instruments) inserted into the esophagus to the level of the heart. Adequate catheter positioning was confirmed by TE pacing at 800 μV. Electrocardiographic channels were amplified (0.1 mV/cm) and filtered between 0.05 and 400 Hz.18 A computer-based data acquisition system (Emka Technologies) was used to record a 3-lead body surface ECG and 4 intraesophageal bipolar electrograms. Transesophageal pacing was performed using 2-millisecond current pulses delivered by an external stimulator (STG3008-FA; Multichannel Systems). Standard clinical electrophysiologic pacing protocols were used to induce AF as previously described.19 Briefly, atrial burst pacing was performed with 300 cycles of 2-millisecond bursts at 800 μV and at a cycle length of 50, 40, 30, 25, 20, and 15 milliseconds. Next, the Verheule pacing protocol was performed with 2-second burst cycles at cycle length of 40 milliseconds, and then decreasing in each successive burst with a 2-millisecond decrement to a cycle length of 20 milliseconds.20 This procedure was followed by right atrial burst pacing with eight 50-millisecond and four 30-millisecond-cycle-length trains repeated several times, up to a maximum 1-minute limit of total stimulation. Series of bursts was repeated twice. Atrial fibrillation was defined as a period of rapid, irregular atrial rhythm lasting at least 1 second. If 1 or more bursts in the 2 series of bursts evoked an AF episode, AF was considered to be inducible in that animal; otherwise, AF was considered to be noninducible.

Statistical Analysis

Categorical variables are expressed as counts and percentages, and continuous variables as mean (SD). Bivariate analyses of continuous and categorical variables were performed using a 2-sample t test or analysis of variance, and 2 categorical variables were compared using the χ2 test for independence and Fisher exact test to compare AF incidence in control and DIO mice. A Mann-Whitney test was performed for AF burden. To assess the percentage of AF reduction with flecainide acetate, a class I AAD, vs sotalol hydrochloride, a class III AAD, a 2-tailed t test was performed, with a statistical significance level of P < .05.

We developed a multivariate logistic regression model using 311 nonmissing observations that included AAD class, obesity, age, sex, race/ethnicity (4 groups), diabetes, hypertension, coronary artery disease, obstructive sleep apnea, ejection fraction, chronic obstructive pulmonary disease, hyperthyroidism, and the interaction between obesity and AAD class as covariates. These variables were selected based on their known association with the development of AF and likelihood of modulating response to class I or III AADs. A backward selection procedure was then applied to derive a parsimonious model retaining covariates (AAD class, obesity, female sex, hyperthyroidism) that were significant (P < .05) indicators of the probability of failure to respond to AADs on univariate analysis (Wald test). Other variable selection methods, including forward and stepwise, selected the same subset of covariates. We also performed a logistic regression analysis based on the numeric BMI centered on 30 to assess the nonlinearity of the association of BMI with response to class I AADs. Sensitivity analysis similar to pairwise comparison, focusing on response to the most commonly used class III AADs, was performed. Statistical software SAS, version 9.4 (SAS Institute Inc), was used to conduct the analysis.

Results
Baseline Characteristics

A flowchart describing the cohort is shown in Figure 1. After exclusionary criteria were applied, 311 patients treated with a class I or III AAD for symptomatic AF were eligible for the study and enrolled over a 36-month period. Patients were excluded from this analysis if no trial of AAD therapy was pursued, assessment of response to therapy was inadequate or incomplete, and class I and III drugs were discontinued due to adverse effects. The baseline clinical characteristics of patients with and without obesity who were treated with a class I or III AAD are reported in Table 1. The mean (SD) age of the cohort was 65 (12) years, and 120 were women (38.6%); 168 patients (54.0%) were obese and 143 patients (46.0%) were not obese. Patients with obesity were more likely to have diabetes (58 [34.5%] vs 30 [21.0%]), hypertension (142 [84.5%] vs 103 [72.0%]), obstructive sleep apnea (55 [32.7%] vs 10 [7.0%]), and coronary artery disease (37 [22.0%] vs 14 [9.8%]).

Response to AADs

The clinical characteristics of patient responses to class I and III AADs for symptomatic AF are reported in eTable 1 in the Supplement, and eTable 2 in the Supplement reports the specific AADs prescribed. Nonresponse to class I AADs in patients with obesity was less than in those without obesity (30% [obese] vs 6% [nonobese]; difference, 0.24; 95% CI, 0.11-0.37; P = .001). At least 1 AAD failed to suppress symptomatic AF in 36 of 311 patients (11.6%). Of 64 patients with AF and obesity treated with class I AADs, 19 patients (29.7%) had symptomatic recurrence of AF captured by long-term monitoring (event recorder, Holter monitor, device) in 18 patients and by ECG in 1 patient; 14 of the 19 patients (73.7%) were switched to a class III AAD and were maintained on the drug for at least 6 months, and 5 of these patients (35.7%) had successful AF ablations. Only 3 of 51 patients (5.9%) without obesity had symptomatic recurrence of AF recorded by long-term monitoring when treated with class I AADs (eTable 3 in the Supplement; Figure 1). Of the 9 patients with obesity who failed to respond to a class III AAD (recorded by long-term monitoring), treatment in 5 patients (55.6%) was switched to amiodarone, which was maintained for at least 6 months with no documented recurrence, and 4 of these patients had successful AF ablations. Of the 5 patients without obesity who did not respond to class III AADs, 4 patients (80.0%) had successful AF ablations and 1 individual (20.0%) was switched to amiodarone. This patient maintained sinus rhythm during the study period and had no symptomatic recurrence of AF documented by an event recorder.

The multivariate logistic regression model with 12 covariates and an interaction between AAD class and obesity was statistically significant (likelihood ratio test statistic = 40.60; degrees of freedom = 14; and P < .001). However, only 5 covariates (AAD class, obesity, sex, hypertension, hyperthyroidism) were significantly associated with failure to respond to AADs (eTable 4A and eTable 4B in the Supplement). We included the interaction to illustrate the higher odds of failure of class I vs III AADs in individuals with obesity. Body mass index as a continuous variable was not associated with failure to respond to class I AADs (adjusted odds ratio [OR] for a 1-unit increase in BMI: 1.03; 95% CI, 0.99-1.07; P = .17). However, when we stratified BMI by 2.5-unit increments, there was a graded, nonlinear increase in the adjusted odds ratio supporting reduced effectiveness of class I AADs in patients with obesity. On multivariate analysis, obesity, AAD class, female sex, and hyperthyroidism were significant indicators of the probability of failure to respond to AADs (class I vs III AAD [obese]: OR, 4.54; 95% Wald CI, 1.84-11.2; P = .001; female vs male: OR, 2.31; 95% Wald CI, 1.07-4.99; P = .03; hyperthyroidism: OR, 4.95; 95% Wald CI, 1.23-20.00; P = .02) (Table 2).

Obesity was significant in the multivariable model but BMI was not. To investigate this further, we performed logistic regression analysis showing that a combination of nonlinearity, large variability, and small sample size at extreme BMI values was likely responsible for why BMI was not associated with failure to respond to class I AADs but obesity was (eResults and eTable 5 in the Supplement). Furthermore, eFigure 1 in the Supplement shows that BMI has a protective or no effect on the binary outcome with a BMI less than 30 and the opposite effect with a BMI 30 or above. It is quite likely that the dichotomized BMI using a cutoff of 30 for obese and nonobese groups reduced the association between nonlinearity and high variability at extreme BMI values, causing the mean BMI for the 2 groups to be significant in the multivariate model. Sensitivity analysis showed that there was no significant difference in failure to respond to flecainide vs other class I AADs and amiodarone vs other class III AADs in patients with obesity (eTable 6A and eTable 6B in the Supplement).

Given our findings of a differential response to class I and III AADs in patients with AF and obesity, we sought to replicate these findings in a DIO mouse model. The mean (SD) weight of the 10 DIO mice was 34.8 (2.7) g vs 24.6 (2.7) g for the control mice (P < .001) (Figure 2A). To evaluate the arrhythmogenic phenotype of DIO mice, we performed TE atrial pacing, which is commonly used to experimentally induce AF (Figure 2B).19 The DIO mice showed increased mean AF burden vs controls (280 [228] vs 16.5 [34] seconds; P < .01) (Figure 2C) and an AF incidence of 100% vs 30% in controls (Figure 2C). eFigure 2 in the Supplement shows the weights of individual DIO mice treated with sotalol or flecainide and the burden of pacing-induced AF.

Next, mice were randomly treated with an intraperitoneal injection of either flecainide, 20 mg/kg, or sotalol, 10 mg/kg, for 2 weeks.18,19 After treatment, mice underwent pacing with the same TE atrial pacing protocol to determine inducibility of AF, with the reduction in AF burden quantified as a percentage. We found that, similar to patients with AF and obesity, DIO mice showed a greater reduction in AF burden when treated with sotalol compared with flecainide (85% vs 25%; P < .01) (Figure 2D). eTable 7 in the Supplement reports AF burden before and after AAD administration in individual DIO mice. Collectively, our clinical and mouse data support our hypothesis that obesity not only mediates response to AADs for AF, but that there may be a differential response to sodium channel vs potassium channel blocker AADs.

Discussion

Obesity is an established risk factor for AF, but the underlying mechanisms remain unclear. Emerging evidence suggests that reduced cardiac sodium channel expression may be a factor. Our results suggest that obesity mediates response to antiarrhythmic therapy in patients with symptomatic AF with a differential reduction in therapeutic effectiveness of sodium channel vs potassium channel blocker AADs. Our clinical observation was supported by mouse studies that showed increased AF burden in DIO mice and reduced therapeutic effectiveness of flecainide vs sotalol in suppressing pacing-induced AF. Our findings suggest implications for the management of AF in patients with obesity.

Although a causal inference between BMI and incident AF was recently reported, the underlying molecular mechanisms by which obesity is associated with AF remain unclear.3 This uncertainty may be owing to the acquired, genetic, and metabolic components associated with the complex obesity phenotype.20 One potential mechanism for obesity-mediated AF is modulation of the cardiac sodium channel. In animal models, an increase or decrease in Nav1.5 expression has been associated with AF.7,8,21,22 In addition, gain- and loss-of-function mutations in SCN5A have been linked with AF in humans.9-11 Although the precise mechanisms by which alterations in Nav1.5 expression contribute to obesity-mediated AF are poorly understood, cumulative evidence from animal and human studies suggests that oxidative stress plays a central role.23,24 Thus, increased oxidative stress may represent one common pathway by which obesity mediates AF risk.

Clinical Implications

Our findings may have clinical implications for the treatment of AF in patients with obesity because class I AADs, which are frequently used to treat symptomatic AF, block the cardiac sodium channel. Emerging evidence supports obesity downregulating Nav1.5 and the sodium current.7,8 Thus, prescribing class I AADs in patients with obesity and symptomatic AF will further decrease the sodium current and may be proarrhythmic, paradoxically worsening AF incidence and/or maintenance.

Limitations

Our study has a number of limitations. First, our findings need to be replicated in larger prospective cohorts. Although the clinical analysis was retrospective, response to AAD therapy was defined a priori and we used a prevalent disease with new drugs (new user design). Second, this is not a comparative study of AAD drug efficacy. We do not attempt to make this direct comparison because there is no randomization of AADs. eTable 1 in the Supplement reports worse response of class I AADs in treating symptomatic AF. However, a potential inference cannot be made in part owing to the observational nature of our study and because the clinical indications for the membrane-active drugs are patient-specific and cannot be randomized. Third, because patients with obesity had a greater burden of AF risk factors, we cannot completely eliminate residual confounding. However, we not only adjusted for several baseline covariates but also performed a backward elimination approach to select the most significant factors (AAD class, obesity, sex, and hyperthyroidism) of nonresponse to AADs. Fourth, although the number of patients who failed to respond to class I AADs was limited, our mouse data supported our hypothesis that obesity differentially mediates response to sodium channel blockers. Fifth, symptoms do not reliably measure AF burden.25,26 However, AF burden is not yet routinely used as a clinical or study end point, but advances in monitoring technology that allow continuous recording of cardiac rhythm have now made it possible to rigorously measure AF burden as a surrogate end point to assess the efficacy of therapy.27 Continuous recording of cardiac rhythm is a more robust metric to assess response because it is less subject to investigator bias and does not have the sampling error associated with episodic monitoring or reliance on patient symptoms.28,29 Sixth, the apparently high therapeutic effectiveness of AADs in our study cohort is in part owing to the exclusion of patients who were intolerant of the medication, use of amiodarone treatment in a large number of patients, and consideration of obesity as a modulator of AAD response. The seventh limitation relates to selection bias and lack of generalizability of our findings occur with use of a clinical-genetic registry. Data from large AF registries,30-33 including ours,34,35 have provided insights into patterns of care and outcomes. There is no selection bias when enrolling patients in the University of Illinois at Chicago AF Registry. Consecutive patients presenting with new-onset AF are approached by multilingual research coordinators to participate in the registry. Although our recruitment rate is high (>95%), concern is voiced about the genetic aspects of the study. However, in our registry, patients are given the option to opt out and participate only in the clinical registry. Thus, we believe the findings of our study are representative of a diverse and urban population and thus generalizable to the US population. Eighth, mouse models may not fully reproduce the electrophysiologic phenotype of obesity-mediated AF in humans. However, our finding that there was a reduced association of flecainide vs sotalol with suppressing pacing-induced AF in DIO mice appears to support our clinical observations.

Conclusions

Obesity is a significant risk factor for AF in susceptible patients. Our study not only suggests that obesity mediates response to AADs for AF, but also that there is reduced therapeutic effectiveness of sodium channel compared with potassium channel blocker AADs. Our findings may have important implications for the management of AF in patients with obesity.

Back to top
Article Information

Accepted for Publication: September 13, 2019.

Corresponding Author: Dawood Darbar, MBChB, MD, Division of Cardiology, University of Illinois at Chicago, 840 S Wood St, 920S (MC 715), Chicago, IL 60612 (darbar@uic.edu).

Published Online: November 27, 2019. doi:10.1001/jamacardio.2019.4513

Author Contributions: Ms Ornelas-Loredo, Dr Kany, and Mr Abraham contributed equally to the study. Drs Konda and D. Darbar had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Kany, Abraham, Alamar, Hong, Konda, D. Darbar.

Acquisition, analysis, or interpretation of data: Ornelas-Loredo, Kany, Abraham, Alzahrani, F. A. Darbar, Sridhar, Ahmed, Alamar, Menon, Zhang, Chen, Konda.

Drafting of the manuscript: Ornelas-Loredo, Abraham, F. A. Darbar, Zhang.

Critical revision of the manuscript for important intellectual content: Kany, Abraham, Alzahrani, F. Darbar, Sridhar, Ahmed, Alamar, Menon, Chen, Konda, Hong, D. Darbar.

Statistical analysis: Ornelas-Loredo, Kany, Alzahrani, F. Darbar, Sridhar, Alamar, Menon, Konda.

Obtained funding: D. Darbar.

Administrative, technical, or material support: Ornelas-Loredo, F. Darbar, Ahmed, Alamar, Zhang, Chen, Hong.

Supervision: Konda, D. Darbar.

Conflict of Interest Disclosures: None reported.

Funding/Support: This work was supported in part by Merit Review Award I01 BX004268 of the US Department of Veterans Affairs Biomedical Laboratory Research and Development Service and NIH grant T32 HL139439.

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Benjamin  EJ, Wolf  PA, D’Agostino  RB, Silbershatz  H, Kannel  WB, Levy  D.  Impact of atrial fibrillation on the risk of death: the Framingham Heart Study.  Circulation. 1998;98(10):946-952. doi:10.1161/01.CIR.98.10.946PubMedGoogle ScholarCrossref
2.
Flegal  KM, Carroll  MD, Kit  BK, Ogden  CL.  Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010.  JAMA. 2012;307(5):491-497. doi:10.1001/jama.2012.39PubMedGoogle ScholarCrossref
3.
Chatterjee  NA, Giulianini  F, Geelhoed  B,  et al.  Genetic obesity and the risk of atrial fibrillation: causal estimates from mendelian randomization.  Circulation. 2017;135(8):741-754. doi:10.1161/CIRCULATIONAHA.116.024921PubMedGoogle ScholarCrossref
4.
January  CT, Wann  LS, Alpert  JS,  et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines.  2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society.  J Am Coll Cardiol. 2014;64(21):e1-e76. doi:10.1016/j.jacc.2014.03.022PubMedGoogle ScholarCrossref
5.
Duytschaever  M, Demolder  A, Phlips  T,  et al.  Pulmonary vein isolation with vs without continued antiarrhythmic drug treatment in subjects with recurrent atrial fibrillation (POWDER AF): results from a multicentre randomized trial.  Eur Heart J. 2018;39(16):1429-1437. doi:10.1093/eurheartj/ehx666PubMedGoogle ScholarCrossref
6.
Darghosian  L, Free  M, Li  J,  et al.  Effect of omega-three polyunsaturated fatty acids on inflammation, oxidative stress, and recurrence of atrial fibrillation.  Am J Cardiol. 2015;115(2):196-201. doi:10.1016/j.amjcard.2014.10.022PubMedGoogle ScholarCrossref
7.
Lieve  KV, Verkerk  AO, Podliesna  S,  et al.  Gain-of-function mutation in SCN5A causes ventricular arrhythmias and early onset atrial fibrillation.  Int J Cardiol. 2017;236:187-193. doi:10.1016/j.ijcard.2017.01.113PubMedGoogle ScholarCrossref
8.
Musa  H, Kline  CF, Sturm  AC,  et al.  SCN5A variant that blocks fibroblast growth factor homologous factor regulation causes human arrhythmia.  Proc Natl Acad Sci U S A. 2015;112(40):12528-12533. doi:10.1073/pnas.1516430112PubMedGoogle ScholarCrossref
9.
Darbar  D, Kannankeril  PJ, Donahue  BS,  et al.  Cardiac sodium channel (SCN5A) variants associated with atrial fibrillation.  Circulation. 2008;117(15):1927-1935. doi:10.1161/CIRCULATIONAHA.107.757955PubMedGoogle ScholarCrossref
10.
Ellinor  PT, Nam  EG, Shea  MA, Milan  DJ, Ruskin  JN, MacRae  CA.  Cardiac sodium channel mutation in atrial fibrillation.  Heart Rhythm. 2008;5(1):99-105. doi:10.1016/j.hrthm.2007.09.015PubMedGoogle ScholarCrossref
11.
Makiyama  T, Akao  M, Shizuta  S,  et al.  A novel SCN5A gain-of-function mutation M1875T associated with familial atrial fibrillation.  J Am Coll Cardiol. 2008;52(16):1326-1334. doi:10.1016/j.jacc.2008.07.013PubMedGoogle ScholarCrossref
12.
Yoshida  K, Solomon  DH, Kim  SC.  Active-comparator design and new-user design in observational studies.  Nat Rev Rheumatol. 2015;11(7):437-441. doi:10.1038/nrrheum.2015.30PubMedGoogle ScholarCrossref
13.
Chalazan  B, Mol  D, Sridhar  A,  et al.  Genetic modulation of atrial fibrillation risk in a Hispanic/Latino cohort.  PLoS One. 2018;13(4):e0194480. doi:10.1371/journal.pone.0194480PubMedGoogle Scholar
14.
Alzahrani  Z, Ornelas-Loredo  A, Darbar  SD,  et al.  Association between family history and early-onset atrial fibrillation across racial and ethnic groups.  JAMA Netw Open. 2018;1(5):e182497. doi:10.1001/jamanetworkopen.2018.2497PubMedGoogle Scholar
15.
Nagueh  SF, Smiseth  OA, Appleton  CP,  et al.  Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.  J Am Soc Echocardiogr. 2016;29(4):277-314. doi:10.1016/j.echo.2016.01.011PubMedGoogle ScholarCrossref
16.
do Carmo  JM, da Silva  AA, Rushing  JS, Pace  B, Hall  JE.  Differential control of metabolic and cardiovascular functions by melanocortin-4 receptors in proopiomelanocortin neurons.  Am J Physiol Regul Integr Comp Physiol. 2013;305(4):R359-R368. doi:10.1152/ajpregu.00518.2012PubMedGoogle ScholarCrossref
17.
Li  XF, Lytton  J.  An essential role for the K+-dependent Na+/Ca2+-exchanger, NCKX4, in melanocortin-4-receptor–dependent satiety.  J Biol Chem. 2014;289(37):25445-25459. doi:10.1074/jbc.M114.564450PubMedGoogle ScholarCrossref
18.
Banach  M, Popławska  M, Borowicz-Reutt  KK.  Sotalol enhances the anticonvulsant action of valproate and diphenylhydantoin in the mouse maximal electroshock model.  Pharmacol Rep. 2017;69(6):1173-1177. doi:10.1016/j.pharep.2017.05.005PubMedGoogle ScholarCrossref
19.
Martin  CA, Zhang  Y, Grace  AA, Huang  CL.  In vivo studies of Scn5a+/− mice modeling Brugada syndrome demonstrate both conduction and repolarization abnormalities.  J Electrocardiol. 2010;43(5):433-439. doi:10.1016/j.jelectrocard.2010.05.015PubMedGoogle ScholarCrossref
20.
Kim  S, Popkin  BM.  Commentary: understanding the epidemiology of overweight and obesity—a real global public health concern.  Int J Epidemiol. 2006;35(1):60-67. doi:10.1093/ije/dyi255PubMedGoogle ScholarCrossref
21.
Hayashi  K, Konno  T, Tada  H,  et al.  Functional characterization of rare variants implicated in susceptibility to lone atrial fibrillation.  Circ Arrhythm Electrophysiol. 2015;8(5):1095-1104. doi:10.1161/CIRCEP.114.002519PubMedGoogle ScholarCrossref
22.
Kim  GE, Ross  JL, Xie  C,  et al.  LKB1 deletion causes early changes in atrial channel expression and electrophysiology prior to atrial fibrillation.  Cardiovasc Res. 2015;108(1):197-208. doi:10.1093/cvr/cvv212PubMedGoogle ScholarCrossref
23.
Lin  CS, Pan  CH.  Regulatory mechanisms of atrial fibrotic remodeling in atrial fibrillation.  Cell Mol Life Sci. 2008;65(10):1489-1508. doi:10.1007/s00018-008-7408-8PubMedGoogle ScholarCrossref
24.
Van Wagoner  DR.  Oxidative stress and inflammation in atrial fibrillation: role in pathogenesis and potential as a therapeutic target.  J Cardiovasc Pharmacol. 2008;52(4):306-313. doi:10.1097/FJC.0b013e31817f9398PubMedGoogle ScholarCrossref
25.
Darbar  D, Roden  DM.  Symptomatic burden as an endpoint to evaluate interventions in patients with atrial fibrillation.  Heart Rhythm. 2005;2(5):544-549. doi:10.1016/j.hrthm.2005.01.028PubMedGoogle ScholarCrossref
26.
Martin  DT, Bersohn  MM, Waldo  AL,  et al; IMPACT Investigators.  Randomized trial of atrial arrhythmia monitoring to guide anticoagulation in patients with implanted defibrillator and cardiac resynchronization devices.  Eur Heart J. 2015;36(26):1660-1668. doi:10.1093/eurheartj/ehv115PubMedGoogle ScholarCrossref
27.
Israel  CW, Grönefeld  G, Ehrlich  JR, Li  YG, Hohnloser  SH.  Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care.  J Am Coll Cardiol. 2004;43(1):47-52. doi:10.1016/j.jacc.2003.08.027PubMedGoogle ScholarCrossref
28.
Euler  DE, Friedman  PA.  Atrial arrhythmia burden as an endpoint in clinical trials: is it the best surrogate? lessons from a multicenter defibrillator trial.  Card Electrophysiol Rev. 2003;7(4):355-358. doi:10.1023/B:CEPR.0000023138.85821.63PubMedGoogle ScholarCrossref
29.
Calkins  H, Kuck  KH, Cappato  R,  et al; Heart Rhythm Society Task Force on Catheter and Surgical Ablation of Atrial Fibrillation.  2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design.  Heart Rhythm. 2012;9(4):632-696.e21, e621. doi:10.1016/j.hrthm.2011.12.016PubMedGoogle ScholarCrossref
30.
Inohara  T, Shrader  P, Pieper  K,  et al.  Association of atrial fibrillation clinical phenotypes with treatment patterns and outcomes: a multicenter registry study.  JAMA Cardiol. 2018;3(1):54-63. doi:10.1001/jamacardio.2017.4665PubMedGoogle ScholarCrossref
31.
Darbar  D, Motsinger  AA, Ritchie  MD, Gainer  JV, Roden  DM.  Polymorphism modulates symptomatic response to antiarrhythmic drug therapy in patients with lone atrial fibrillation.  Heart Rhythm. 2007;4(6):743-749. doi:10.1016/j.hrthm.2007.02.006PubMedGoogle ScholarCrossref
32.
Parvez  B, Chopra  N, Rowan  S,  et al.  A common β1-adrenergic receptor polymorphism predicts favorable response to rate-control therapy in atrial fibrillation.  J Am Coll Cardiol. 2012;59(1):49-56. doi:10.1016/j.jacc.2011.08.061PubMedGoogle ScholarCrossref
33.
Echouffo-Tcheugui  JB, Shrader  P, Thomas  L,  et al.  Care patterns and outcomes in atrial fibrillation patients with and without diabetes: ORBIT-AF Registry.  J Am Coll Cardiol. 2017;70(11):1325-1335. doi:10.1016/j.jacc.2017.07.755PubMedGoogle ScholarCrossref
34.
Chalazan  B, Dickerman  D, Sridhar  A,  et al.  Relation of body mass index to symptom burden in patients with atrial fibrillation.  Am J Cardiol. 2018;122(2):235-241. doi:10.1016/j.amjcard.2018.04.011PubMedGoogle ScholarCrossref
35.
McCauley  MD, Darbar  D.  Race and socioeconomic status regulate lifetime risk of atrial fibrillation.  Circ Arrhythm Electrophysiol. 2018;11(7):e006584. doi:10.1161/CIRCEP.118.006584PubMedGoogle Scholar
×