Modulation of Cardiac Arrhythmogenesis by Epicardial Adipose Tissue

Authors, Journal, Affiliations, Type, DOI

• Authors: Auriane C. Ernault (MSc), Veronique M.F. Meijborg (PhD), Ruben Coronel (MD, PhD)
• Journal: Journal of the American College of Cardiology (JACC State-of-the-Art Review)
• Volume/Issue: Vol. 78, No. 17, 2021 (October 26, 2021): pp. 1730–1745
• Affiliations: Department of Experimental Cardiology, Amsterdam UMC; INSERM/Aix-Marseille Université; Université de Bordeaux (IHU LIRYC)
• Type: JACC State-of-the-Art Narrative Review
• DOI: https://doi.org/10.1016/j.jacc.2021.08.037

Overview

This JACC state-of-the-art review synthesises how epicardial adipose tissue (EAT) — the visceral fat depot between the myocardium and visceral pericardium — promotes cardiac arrhythmias through three interlocking mechanisms. First, adipose tissue infiltrates the myocardium creating anatomical obstacles to impulse propagation ("zigzag" conduction, fractionated electrograms, re-entry). Second, adipocytes may electrically couple to cardiomyocytes via Cx43 gap junctions, altering resting membrane potential and excitability. Third, EAT secretes a rich repertoire of adipokines and extracellular vesicles that prolong action potential duration, downregulate connexins, and drive fibrosis — all creating substrate for arrhythmias. ECG correlates of elevated EAT volume span P-wave prolongation, PR interval, QRS duration, fragmented QRS, and Tpeak–Tend interval.

Keywords

Arrhythmias, cardiac electrophysiology, cardiovascular diseases, epicardial adipose tissue, obesity

Key Takeaways

EAT Anatomy and Origin

• EAT is visceral fat located between the myocardium and the epicardium, enclosed within the pericardial sac and sharing coronary blood supply with the heart. It is predominantly distributed along AV and interventricular grooves, around the atria, and along coronary arteries.
• No fascia separates EAT from the underlying myocardium — enabling direct paracrine crosstalk between adipocytes and cardiomyocytes.
• EAT originates from splanchnopleuric mesoderm via epicardial progenitor cell epithelial-to-mesenchymal transition; distinct from paracardial adipose tissue (PAT) which derives from thoracic mesenchyme and is outside the pericardium.
• EAT volume increases during the first 40 years, then depends on BMI rather than age. Average EAT = ~20% of total heart weight.
• Rodent models lack significant EAT and are not suitable for cardiac adiposity research; PAT is not an equivalent substitute.

EAT Role in Cardiac Metabolism

• EAT provides free fatty acids (FFAs) — the primary metabolic substrate of cardiomyocytes (60–90% of energy). Higher FFA uptake than subcutaneous fat; serves as a local buffering mechanism against lipotoxicity.
• Lipid overload overwhelms mitochondrial oxidative capacity → accumulation of toxic ceramides → mitochondrial and ER dysfunction, calcium dysregulation, ROS production.
• ROS-induced Ca²⁺ overload → delayed afterdepolarizations (DADs) and triggered activity. ROS also alter Cx43-mediated electrical coupling → conduction slowing.
• Cardiac lipid droplets (LDs) act as a buffer; when LD capacity is saturated or dysfunctional → lipotoxicity, excess ROS, arrhythmias. Lipid storage diseases are associated with lethal arrhythmias.

EAT and ECG Parameters

Atrial Conduction

• P-wave duration: Positively associated with EAT thickness/volume across multiple studies (Framingham Heart Study n=1,946; Jhuo n=100; Fernandes-Cardoso morbidly obese n=40). Also associated with interatrial conduction block (Jhuo). P-wave fragmentation (marker of heterogeneous conduction) correlates with infiltrated adipose tissue in the atrial septum.
• P-wave dispersion: Associated with EAT thickness in healthy subjects (Quisi, Çiçek) — implies anisotropic sinus impulse propagation facilitating re-entry.
• PR interval: EAT volume linearly correlates with longer PR interval in 4 studies (Friedman, Jhuo, Hung, Chi). Patients with highest EAT volume had PR 10–16 ms longer. Prolonged PR independently predicts AF, HF, and mortality.

Ventricular Conduction

• QRS duration: Strongly associated with EAT volume (Chi, n=3,087; Hung, n=287); highest vs lowest EAT → 6.7 ms longer QRS. Prolonged QRS reflects slowed ventricular conduction or hypertrophy.
• Fragmented QRS (fQRS): Associated with increased EAT in both hypertensive patients (Bekar, n=114; EAT >4.5 mm predicted fQRS with 75% sensitivity/58% specificity) and healthy subjects (Yaman, n=308). fQRS reflects heterogeneous anisotropic ventricular conduction → substrate for re-entry.

Ventricular Repolarisation

• QTc: Evidence is conflicting — one study reports negative association (Quisi) while others show no relationship. Effect of EAT on ventricular repolarisation remains unclear.
• Tpeak–Tend (TpTe) interval: Increased in subjects with higher EAT (Kaplan, n=90), along with increased QT dispersion — together indicating repolarisation heterogeneity and predicting SCD/mortality.

Structural Cross Talk: Adipose Tissue Infiltration

• Two patterns of myocardial fat infiltration: (1) thin compact cords of adipocytes from the epicardium penetrating between cardiac bundles; (2) fibro-fatty infiltration — adipocytes surrounded by dense fibrotic areas (suggests inflammatory remodeling).
• Atrial fibro-fatty replacement is more extensive in persistent than paroxysmal AF (Haemers, Platonov). AF induces increased fibrosis of atrial fatty infiltrates in sheep (Haemers).
• Computer modelling (De Coster): fibrosis is more arrhythmogenic than adipose tissue per unit of non-conductive tissue; adipose + fibrosis synergistically create substrate.
• Adipocyte infiltration creates anatomical obstacles → "zigzag" path of activation wavefront (as in myocardial infarction) → activation delay and electrogram fractionation → re-entry. In sheep MI model: intramyocardial adiposity → myocardial discontinuity, decreased conduction velocity, reduced electrogram amplitude, increased arrhythmia risk, and lateralisation of Cx43.
• In CAD patients: local slower conduction, electrogram fractionation, and lateralisation of Cx40 at sites with high local EAT volume.

Structural Cross Talk: Intercellular Coupling

• Both adipocytes (white adipose tissue) and cardiomyocytes express Cx43; adipocytes from mouse white adipose tissue are functionally coupled to each other via Cx43. Whether EAT adipocytes couple to cardiomyocytes remains unproven.
• Cardiomyocyte RMP ≈ −90 mV; white adipocyte membrane potential ≈ −30 mV; fibroblasts/macrophages −10 to −30 mV. If EAT cells couple to cardiomyocytes via gap junctions:
  • Slight depolarisation (−75 to −65 mV): increased excitability ("superexcitability")
  • Further depolarisation (−65 to −55 mV): partial Na⁺ channel inactivation → decreased upstroke velocity → reduced conduction velocity, increased risk of block
  • Extreme depolarisation: complete Na⁺ channel inactivation; propagation dependent on ICa,L only (very slow)
• Electrical coupling between cardiomyocytes and fibroblasts demonstrated in rabbit hearts (Camelliti). Myofibroblasts co-cultured with cardiomyocytes induce spontaneous electrical activity via diastolic depolarisation (Miragoli).

Paracrine Cross Talk: EAT Secretome

• The EAT secretome includes: soluble factors (growth factors, cytokines, bioactive lipids) and extracellular vesicles (EVs carrying proteins, lipids, mRNA, miRNA, lncRNA, circRNA).
• Epicardial fat-derived EVs from AF patients contain pro-inflammatory/profibrotic cytokines and profibrotic microRNA; incubation with hiPSC-CM sheets shortened APD80 and induced sustained rotors (Shaihov-Teper).
• EAT secretome prolongs APD in rabbit atrial cardiomyocytes (longer APD90, less negative RMP; changes in INaLate, ICaL, IK1, Ito) and in H9c2 cells (reduced IKr). 24-hour sheep EAT incubation of hiPSC-CMs extends field potential duration.
• In silico study: secretome-induced cardiomyocyte remodeling created more complex spiral wave dynamics in a 2D monolayer and human atria model, suggesting increased arrhythmia complexity.

Adipokines Involved in Electrical Remodeling

• TNF-α: Reduces Ito (transient outward K⁺) in rat ventricular myocytes → APD prolongation; reduces Cx40 expression; slows conduction velocity in guinea pig hearts; induces excessive MMP secretion → ECM remodeling.
• IL-1β: Decreases Ito → APD prolongation in rat hearts; prolongs field potential duration in hiPSC-CMs; increases spontaneous diastolic SR Ca²⁺ release → triggered activity risk; downregulates/degrades Cx43 in post-MI mouse models.
• IL-6: Modulates L-type calcium current (dose- and duration-dependent); downregulates SERCA2a → Ca²⁺ handling dysregulation; reduces Cx40 and Cx43 in HL-1 mouse atrial myocytes.
• FABP4 (fatty acid-binding protein-4): Reduces intracellular systolic peak Ca²⁺ in rat cardiomyocytes.

Profibrotic Adipokines (Table 2 of paper)

• Activin A (TGF-β superfamily): Highly present in EAT secretome; induces collagen 1 deposition in rat atria.
• Visfatin: Promotes cardiac fibroblast proliferation and collagen synthesis in rats.
• TGF-β1: Induces fibroblast-to-myofibroblast phenotype switch → ECM production, fibroblast proliferation, interstitial fibrosis (overexpression in mice → fibrosis).
• Leptin: Secreted by EAT adipocytes; stimulates collagen I synthesis in cardiac myofibroblasts → fibrosis.
• MCP-1 (CCL2): Secreted by EAT adipocytes and immune cells; enhances myofibroblast activation, fibroblast proliferation, and collagen expression.
• IL-6: IL-6 infusion in rats → increased collagen volume fraction.
• MMPs (MMP-1, -2, -8, -9): Secreted at high levels by EAT; TNF-α induces excessive MMP secretion → ECM remodeling.

Conclusions (as stated in paper)

• EAT creates an anatomical obstacle delaying cardiac activation; heterogeneous fat distribution over atrial myocardium and cellular uncoupling cause heterogeneous conduction slowing → re-entry.
• Cx43 expressed by both adipocytes and cardiomyocytes; electrotonic coupling between EAT and cardiomyocytes remains unproven but hypothetically exacerbates arrhythmogenicity.
• Adipokines from EAT induce electrical remodeling and promote fibrosis — key for arrhythmia substrate.
• EAT accumulation may lead to lipotoxicity contributing to arrhythmogenesis; this mechanism requires further research.
• EAT represents a potential target for improved risk assessment and prevention of cardiac arrhythmias.

Limitations of the Document

• Narrative review — no systematic review methodology or meta-analysis; susceptible to selection bias in cited literature.
• Most mechanistic evidence relies on animal models (rabbit, rat, sheep, mouse, guinea pig) and in vitro systems; direct human EAT-cardiomyocyte electrophysiology data are sparse.
• Electrotonic coupling between EAT adipocytes and cardiomyocytes is entirely hypothetical — no in vivo proof in humans.
• Specific secretome components responsible for APD changes remain unidentified.
• Rodent models lack substantial EAT — animal data may not translate to human pathophysiology.
• ECG association studies in Table 1 are mostly observational, small (n=40–3,087), cross-sectional, with heterogeneous fat measurement methods (echocardiographic EAT thickness vs CT volumetry).
• No RCT or interventional data on EAT reduction and arrhythmia outcomes.

Key Concepts Mentioned

• concepts/Epicardial-Adipose-Tissue-Arrhythmogenesis — central topic; all three arrhythmogenic pathways
• concepts/Visceral-Adiposity — EAT as the cardiac ectopic fat depot
• concepts/Electrical-Remodeling — adipokine-mediated ion channel and gap junction remodeling
• concepts/Cardiac-Action-Potential — APD prolongation, altered RMP, Na⁺ channel inactivation
• concepts/Torsades-de-Pointes — EAT-mediated APD prolongation and repolarisation heterogeneity as risk factor
• concepts/Cardiac-Repolarization — Tpeak–Tend and QT dispersion correlates of EAT

Key Entities Mentioned

• entities/Atrial-Fibrillation — EAT volume positively associated with AF incidence, duration, recurrence; fibro-fatty infiltration substrate
• entities/Obesity — primary driver of EAT accumulation; 2× SCD risk, 29% AF risk per 5-BMI-unit increment

Wiki Pages Updated

• wiki/sources/epi-adipose-arrhythmia-jacc-2021.md — created (this file)
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — new concept entry added
• wiki/concepts/Epicardial-Adipose-Tissue-Arrhythmogenesis.md — created
• wiki/concepts/Visceral-Adiposity.md — updated with mechanistic detail
• wiki/entities/Atrial-Fibrillation.md — updated with EAT substrate