Objective High bile acid serum concentrations have been implicated in cardiac disease, particularly in arrhythmias. Most data originate from in vitro studies and animal models. We tested the hypotheses that (1) high bile acid concentrations are arrhythmogenic in adult human myocardium, (2) serum bile acid concentrations and composition are altered in patients with atrial fibrillation (AF) and (3) the therapeutically used ursodeoxycholic acid has different effects than other potentially toxic bile acids.

Methods and Results Multicellular human atrial preparations (‘trabeculae’) were exposed to primary bile acids and the incidence of arrhythmic events was assessed. Bile acid concentrations were measured in serum samples from 250 patients and their association with AF and ECG parameters analysed. Additionally, we conducted electrophysiological studies in murine myocytes.

Taurocholic acid (TCA) concentration-dependently induced arrhythmias in atrial trabeculae (14/28 at 300 µM TCA, p<0.01) while ursodeoxycholic acid did not. Patients with AF had significantly decreased serum levels of ursodeoxycholic acid conjugates and increased levels of non-ursodeoxycholic bile acids. In isolated myocytes, TCA depolarised the resting membrane potential, enhanced Na⁺/Ca²⁺ exchanger (NCX) tail current density and induced afterdepolarisations. Inhibition of NCX prevented arrhythmias in atrial trabeculae.

Conclusions High TCA concentrations induce arrhythmias in adult human atria while ursodeoxycholic acid does not. AF is associated with higher serum levels of non-ursodeoxycholic bile acid conjugates and low levels of ursodeoxycholic acid conjugates. These data suggest that higher levels of toxic (arrhythmogenic) and low levels of protective bile acids create a milieu with a decreased arrhythmic threshold and thus may facilitate arrhythmic events.