The proarrhythmic conundrum of alcohol intake

Abstract

The arrhythmogenic potential of alcohol consumption that leads to cardiac arrhythmia development includes the induction of both atrial and ventricular arrhythmias, with atrial fibrillation (AF) being the commonest alcohol-related arrhythmia, even with low/moderate alcohol consumption. Arrhythmias occur both with acute and chronic alcohol use. The “Holiday Heart Syndrome” relates to the occurrence of AF, most commonly following weekend or public holiday binge drinking; however, other arrhythmias may also occur, including other supraventricular arrhythmias, and occasionally even frequent ventricular premature beats and a rare occurrence of ventricular tachycardia. Arrhythmias in individuals with alcohol use disorder, in addition to AF, may comprise ventricular arrhythmias (VAs) that may be potentially fatal leading to cardiac arrest. The effects of alcohol on triggering VAs appear to be dose-dependent, observed more commonly in heavy drinkers, both in healthy individuals and patients with underlying structural heart disease, including ischemic heart disease and alcoholic cardiomyopathy. Men appear to be affected at higher dosages of alcohol, while women can suffer from arrhythmias at lower dosages. On the other hand, low to moderate consumption of alcohol may confer some protection from serious VAs and cardiac arrest (J- or U-curve phenomenon); however, abstinence is the optimal strategy. These issues as they relate to alcohol-induced proarrhythmia are herein reviewed, with the large studies and meta-analyses tabulated and the arrhythmogenic mechanisms pictorially illustrated.

Introduction

Alcohol is a social risk factor related to lifestyle habits that has been shown to contribute, in addition to liver and other organ toxicity, to deleterious cardiovascular (CV) effects, noted especially after heavy and/or binge alcohol consumption [1]. The CV toxicity incurred by alcohol use is reviewed elsewhere [1]. The focus of this review is the arrhythmogenic potential of alcohol consumption that leads to both supraventricular (SVAs) and ventricular arrhythmias (VAs) (Table 1) [2, 3]. Most data about the adverse effects of alcohol, particularly its proarrhythmic effects, relate to heavy drinking and concern patients with alcoholism [3, 4]. The effect of social drinking on cardiac arrhythmias in non-alcoholic individuals needs further scrutiny, although the data on alcohol consumption promoting atrial fibrillation (AF) are most compelling [5, 6]. Despite a robust association of alcohol consumption with both SVAs and VAs, the preponderance of evidence concerns its association with AF [5, 7].

“Holiday Heart Syndrome” relates to the occurrence of AF following weekend or public holiday binge-drinking, but other arrhythmias may also occur, including other SVAs, e.g., atrial flutter (AFlu) or atrial premature beats (APCs), and occasionally frequent ventricular premature beats (VPCs) and a rare occurrence of ventricular tachycardia (VT); the “syndrome” was initially described in individuals who were heavy habitual and/or binge drinkers without apparent structural heart disease (SHD) [2]. In addition to the arrhythmogenic effects of acute alcohol drinking, chronic alcohol use, especially when heavy and habitual, may cause both SVAs, most commonly AF, and VAs that may be potentially fatal [1].

Issues relating to the arrhythmogenicity of alcohol consumption are herein reviewed, relevant large studies and meta-analyses are tabulated and the arrhythmogenic mechanisms are pictorially illustrated.

Alcohol enhances vulnerability to AF and AFlu in patients without SHD. Among 47,949 participants (mean age, 56 years) in the Danish Diet, Cancer, and Health Study (alcohol consumption/day 28.2 ± 25 g in men and 13.9 ± 15 g in women), over a mean of 5.7 years, AF/AFlu developed in 556 subjects [8]. After adjusting for known risk factors, there was a modest increase in risk of AF/AFlu by increasing alcohol consumption in men. When using the lowest quintile of alcohol consumption in men as a reference, the adjusted hazard ratios (HRs) in men in quintiles 2, 3, 4, and 5 were 1.04, 1.44, 1.25, and 1.46, respectively (P for trend, 0.04). The HRs in women were non-significant. The authors concluded that consumption of alcohol was associated with an increased risk of AF/AFlu in men and not in women.

A retrospective study reviewed 64 cases with idiopathic acute AF and 64 age- and sex-matched controls; 62% of cases and 33% of controls were heavy alcohol users [4]. The Copenhagen City Heart Study (2005), examining the association between alcohol use and incident AF among 16,415 women and men, showed no association with risk of AF throughout the moderate range; however, consumption of ≥35 drinks/week among men was associated with an HR of 1.45; few women consumed this amount of alcohol [9].

A recent community-based cohort study explored the association between alcohol consumption (median 3 g/day) and incident AF among 107,845 individuals (median age 47.8 years, 48.3% men) [6]. Over a median of 13.9 years, 5854 individuals developed AF. Alcohol consumption was non-linearly and positively associated with incident AF (HR for one drink/day of 12 g: 1.16, P<0.001).

A population-based cohort study of 109,230 health examination participants in northern Sweden, whereby AF was diagnosed in 5230 individuals, showed that among men, AF risk increased over quartiles of weekly alcohol consumption (P for trend 0.001) [7]. Men with alcohol consumption in the highest quartile (≥4.83 standard drinks/week-SDW) had an HR of 1.21 for AF compared to men in the lowest quartile (<0.90 SDW). However, among women, AF risk was not significantly associated with alcohol consumption. Another large study of 47,002 Norwegian participants showed an adjusted HR for AF of 1.38 when comparing participants consuming >7 drinks per week with abstainers; on the other hand, there was no association detected at <1 drink per day for women and <2 drinks per day for men in the absence of risky (binge or problem) drinking [10].

Regarding women, among 34,715 initially healthy women (>45 years, free of AF at baseline) participating in the Women's Health Study, over a median of 12.4 years, 653 cases of incident AF were confirmed [11]. Compared with nondrinking women, women consuming ≥2 drinks/d had an absolute risk increase of AF of 0.66 events/1000 person-years (HR 1.60).

Finally, a large prospective study of 19,634 individuals (50% male; age 19–74 years) followed for a period of 7.0 ± 2.8 years recorded 199 cardiac events, including new-onset AF (n = 160) [12]. The incidence of new-onset AF was higher in drinkers (HR 2.21; P<0.001). There was a dose-dependent increase in the risk of AF according to the amount of alcohol consumed, and the risk increased more abruptly in men than in women. The risk of AF was highest in frequent binge drinkers (HR 3.15; P<0.001), compared to infrequent light drinkers.

Meta-Analyses. The results of relevant large studies and meta-analyses of studies concerning the effect of alcohol on AF are summarized in Table 2. A meta-analysis of 6 studies and 8 independent datasets (4767 patients/63124 controls) showed a consistent dose–response relationship between the amount of alcohol consumed daily and the probability of new-onset AF [13]. A prospective study (79,019 men and women, free from AF at baseline) and a dose-response meta-analysis (7 prospective studies, including 12,554 AF cases) indicated that alcohol consumption, even at moderate intakes, is a risk factor for AF [14].

A meta-analysis of 14 studies comprising 130,820 participants and 7558 AF cases indicated that the pooled estimate of AF for the highest vs the lowest alcohol intake was 1.51 [15]. The pooled estimate for an increment of 10 g/d alcohol intake was 1.08 (R²=0.43, P<0.001). The AF risk increased with increasing levels of alcohol consumption. The authors concluded that not consuming alcohol at all is the most favorable behavior for avoiding AF rather than moderate alcohol consumption.

A meta-analysis of 9 prospective studies (N = 249,496) indicated that high levels of alcohol intake were associated with an increased incident AF risk (HR 1.34, P<0.001) [16]. Moderate levels of alcohol intake were associated with a heightened AF risk in males (HR 1.26, P = 0.02) but not in females (HR 1.03, P = 0.74). Low alcohol intake, of up to 1 drink/day, was not associated with AF.

There appears to be a link between heavy alcohol intake and all types of SVAs. In an analysis of 4131 participants of the community-based KORA-S4 Study, the investigators showed an association of chronic alcohol consumption with sinus tachycardia (OR 1.03; P = 0.006) [17]. A large prospective study of 1322 individuals who reported daily intake of ≥6 drinks (6+) and 2644 lighter drinkers who reported taking alcohol at least monthly but, on average, <1drink/day (<1), showed that comparing 6+ to <1 drinkers, the relative risk was at least doubled for each of the following arrhythmias: AF (1.1% vs 0.5%, relative risk-RR 2.3), AFlu (0.6% vs 0.2%, RR 3.0), supraventricular tachycardia (SVT) (0.4% vs 0.1%, RR 5.0) and APCs (3.3% vs 1.3%, RR 3.0) [18].

Although AF is the most commonly reported and better studied alcohol-related arrhythmia, VAs and sudden cardiac death (SCD) are also reported, albeit less frequently [2, 19]. The effects of alcohol on triggering VAs are dose-dependent, observed both in healthy subjects and patients with underlying SHD [19–21]. Men are affected at higher dosages of alcohol, while women can suffer from arrhythmias at lower dosages. On the other hand, low/moderate consumption of alcohol may confer some protection from serious VAs and SCD (J-curve phenomenon, see discussion below).

A Japanese observational study (FACT-1) examined the occurrence of VPCs (303 individuals or 1.7%) and its relation to lifestyle habits in 17,641 office-workers, who had an electrocardiogram during a regular medical checkup and had no known heart disease [22]. The occurrence of regularly drinking alcohol (P = 0.009, OR 1.7) and insomnia (P = 0.006) were significantly higher in the VPC group.

A case was recently described of a 36-year-old male who, with no risk factors for coronary artery disease (CAD) and with a structurally normal heart, experienced two episodes of cardiac arrest 5 years apart, with both events occurring after significant alcohol consumption [23]. Alcohol intake can lead to potassium and magnesium deficiency, which can prolong the QT interval and lead to polymorphic VT in the form of torsade-des-pointes (TdP) [24–27]. Another case report of alcohol-associated cardiac arrest concerned a 51-year-old lady with long-term history of alcoholism who sustained a cardiac arrest episode associated with protracted hypokalemia [28].

Alcohol intoxication may also be a possible trigger of the Brugada syndrome (BS) [29, 30]. In a case series of 19 patients with BS or early repolarization syndrome, presenting with cardiac arrest or syncope, the events occurred after alcohol consumption in 4 of them [31].

An older study (1993) examined the relation between QT-intervals and mortality in 69 patients with alcoholic liver disease (without evidence of SHD), in comparison with 40 healthy non-drinking controls matched for age and sex [32]. Patients were abstinent for at least 7 days to exclude acute effects of alcohol. Maximum QT-intervals were longer in alcoholics than in controls (QTc 450 vs 439, P = 0.016), not explained by variations in electrolytes. QT-intervals were prolonged in the 14 patients who died compared with survivors (QTc 471 vs 446, P = 0.007).

Case reports of VT/VF precipitated by alcohol ingestion, especially in individuals with no SHD or risk factors for CAD [23, 28], raise concern that this scenario is under-recognized given the fact that many of these cases are incorrectly attributed to minor coronary lesions or classified as “idiopathic.”

A pathology series of 7 individuals out of 1292 (0.5%) concerned SCD occurring in chronic alcohol users [33]. The only findings at post mortem were fatty liver and a negative or low blood alcohol. The authors suggested that alcohol-associated arrhythmic death or sudden unexpected death in alcohol misuse (SUDAM) is an under-recognized entity, as many of these deaths are misattributed to other causes (see further discussion below).

In patients with risk factors for or with underlying CAD, heavy alcohol consumption may be a potential trigger for SCD. A cross-over study of 309 SCD victims (253 men; median age at death, 57 years) with each of them having 2.8 risk factors for CAD, showed that the estimated relative risk of dying within 2 h after alcohol consumption was 3.00 [34]. A prospective study of a cohort of 7735 men, aged 40–59 at screening and followed-up for 8 years, indicated that there were 217 deaths from CAD of which 117 (54%) were classified as sudden [19]. Heavy drinkers (>6 drinks/d) showed the highest incidence rate of SCD irrespective of the presence or degree of pre-existing CAD, with heavy drinkers showing an increase of >60% compared with occasional or light drinkers. Two Chicago epidemiological studies comprising 1233 and 1899 white men, respectively, showed that problem-drinkers had significantly higher 15-year mortality rates from all causes, CV, and SCD in the first study; while heavy drinkers had increased 10-year mortality rates both for all causes and non-CV causes in the second study [35].

Sudden Unexpected Death in Alcohol Misuse. The term sudden unexpected death in alcohol misuse (SUDAM) is ascribed to sudden death occurring in heavy alcohol drinkers and individuals with alcohol use disorder (AUD) in whom no autopsy findings other than histological changes of fatty liver are detected [33, 36]. Among several potential causes of such deaths, a malignant VA may be suspected which is unmasked by persistent heavy alcohol use. When compared with deaths due to SCD, the SUDAM group are older and have a 10‐fold higher incidence of fatty liver disease [37]. Among the suspected causes of SUDAM, QT interval prolongation and/or electrolyte changes incurred by heavy alcohol consumption, other substance use and comorbid psychiatric conditions managed with QT‐prolonging drugs (e.g. tricyclic antidepressants, selective serotonin reuptake inhibitors, lithium, methadone, and antipsychotics) may be implicated [32]. Finally, life-threatening VAs may also be associated with underlying alcoholic cardiomyopathy (ACM), or related to other underlying SHD [21].

A cross-sectional US Nationwide cohort study of 114,958 inpatients (age, 15–54 years), primarily managed for a variety of SVAs and VAs and further grouped by comorbid AUD (n = 11,214/~10%), indicated that mortality risk increased with age, as elders (45–54 years) had a two-fold higher risk, whereas men had a lower risk (odds ratio-OR 0.8) of inpatient death [38]. Comorbid CAD (OR 4.5) and diabetes (OR 1.4) increased mortality risk in arrhythmia inpatients, as did AUD (OR 1.7). The authors concluded that AUD is an independent risk factor for mortality in arrhythmia inpatients, and it is elevated by 72% in such patients.

With regards to individuals with AUD, the AUD Identification Test (AUDIT), a 10-item questionnaire, was developed as a screening instrument for hazardous alcohol consumption, derived from a study comprising 1888 individuals attending primary health care facilities [39]. After excluding non-drinkers and known alcoholics, 18% of individuals had a hazardous level of alcohol intake and 23% had experienced at least one alcohol-related problem in the previous year [40]. These individuals have a higher risk of arrhythmia and sudden death compared to the control group. In a cross-sectional study of 48 individuals diagnosed with AUD (mean age 44.35 ± 10.24 years), all measured ECG parameters, including QT and QTc interval dispersion, were pathologically elevated (all P = 0.000) compared to a control group (n = 49) of similar mean age (40.90 ± 13.45 years) [41]. There was a significant difference between the groups based on smoking status (P = 0.000). The mean period of alcohol use was 20.71±12.04 years, and the alcohol intake was 5.88±1.65 units/d.

Some investigators have suggested that patients with SUDAM have different characteristics compared to patients with typical SCD. A retrospective cross-sectional study of 62 cases of SUDAM and 41 cases of SCD indicated that the SUDAM group were older than the SCD group (mean age 35.8 years vs 27.7 years; P = 0.0002), had a higher incidence of significant psychiatric illness (19.7% vs 2.4%), while post mortem liver examination was more likely to reveal heavier and more fatty liver changes (24.2% vs 2.4%) [37]. The investigators stress the fact that SCD, unlike SUDAM, is often associated with heritable channelopathies that may affect surviving family members, and thus differentiating between the two may assist families likely to benefit from screening for these mutations, thus preventing further SCD.

J- or U-Curve. As mentioned, individuals who consume large amounts of alcohol (>5 drinks/d) have increased risks of VAs and SCD [19, 35]. However, the relationship is less clear for drinkers of light-to-moderate amounts. A population-based study examined the association between survival after an acute cardiac event and history of smoking and alcohol consumption among patients aged 25–69 years who suffered myocardial infarction or SCD [42]. Among 10,170 events, 2504 resulted in death within 28 days. Individuals who consumed >8 alcoholic drinks/day on >2 days/week (OR=1.93) and former moderate-to-heavy drinkers (OR=4.59) were more likely to die than people who were nondrinkers.

According to the Physicians Health Study, among 21,537 males who were free of CV disease over 12 years, 141 SCDs were confirmed [43] (Table 2). After control for multiple confounders, men who consumed 2–4 drinks/week (RR=0.40; P = 0.004) or 5–6 drinks/week (RR=0.21; P = 0.002) at baseline had significantly reduced risks of SCD compared with those who rarely or never consumed alcohol. The relationship for SCD was U-shaped (P = 0.002), with the risk approaching unity at ≥2 drinks/d. In contrast, the relationship of alcohol intake and non-sudden CV death was L-shaped or linear (P for trend=0.02).

The Women's Health Initiative Observational Study examined the association between long-term alcohol and caffeine intakes and risk of SCD in 93,676 postmenopausal women [44]. Intake of 5–15 g alcohol/d (~1 drink/d) was associated with a non-significantly reduced risk of SCD compared with 0.1–5 g/d of baseline intake (adjusted HR: 0.64), of cumulative average intake (HR: 0.69), and of most recent intake (HR: 0.58). The authors concluded that ~1 drink/day (or 5.1–15 g/d) may be associated with a reduced risk of SCD.

The Nurses' Health Study, comprising 85,067 women free of chronic disease at baseline, found a U-shaped association between alcohol intake and risk of SCD, with the lowest risk among women who drank 5.0–14.9 g/day of alcohol (P for quadratic trend=0.02) [45]. Compared to abstainers, the relative risk for SCD was 0.79 for former drinkers, 0.77 for 0.1–4.9 g/day, 0.64 for 5.0–14.9 g/day, 0.68 for 15.0–29.9 g/day, and 1.15 for ≥30 g/day. In contrast, the relationship of alcohol intake and nonfatal and fatal CAD was more linear (P for linear trend <0.001).

Cases of alcohol-induced atrioventricular (AV) block are very rare in the absence of alcoholic cardiomyopathy (ACM) [46, 47]. Apart from isolated case reports of alcohol-induced AV block [46, 48], a review described 8 cases of second- and third-degree AV block following acute alcohol intoxication in otherwise healthy individuals [47]. In all cases, patients reverted to sinus rhythm upon becoming sober.

According to the prospective CODE-AF registry [49], among 9411 patients with AF, categorized into 4 groups according to the amount of alcohol consumption-abstainer/rare, light (<100 g/week), moderate (100–200 g/week), and heavy (≥200 g/week), over 17.4 ± 7.3 months, heavy alcohol consumption conferred an increased risk of adverse outcomes (adjusted HR-aHR 1.32) compared with abstention from alcohol use. Adverse events included ischemic stroke, transient ischemic attack (TIA), systemic embolic event, AF hospitalization, death, or major bleeding. More specifically, in the heavy group, the primary composite outcome (a composite of thromboembolic events and AF-related hospitalization) (aHR 1.48), secondary composite outcome including primary composite outcome and death and major bleeding (aHR 1.42), ischemic stroke/TIA or systemic embolic event (aHR 2.27), and AF hospitalization (aHR 1.39) showed significantly increased risks compared with those in the abstainer-rare group. However, no significantly increased risk for adverse outcomes was observed in the light (aHR 0.88) and moderate (aHR 0.91) alcohol consumption groups. Thus, a J-shaped relationship was seen between continuous dose of alcohol consumption and the risk of composite adverse events of AF.

Acute excessive alcohol consumption alters cardiac electrophysiology and can provoke arrhythmias. Electrophysiologic studies of alcohol effects conducted in 14 patients (two with dilated cardiomyopathy) with a history of arrhythmias and alcohol consumption, showed induction of nonsustained VT in 1, nonsustained AF in 1, and paired ventricular responses in 1 patient [48]. After ingestion of 90 ml of 80-proof whiskey, 10 of the 14 patients developed sustained or nonsustained SVAs and VAs.

According to a prospective study of 3028 participants who received a smartphone-based ECG and breath-alcohol-concentration (BAC) measurements during the 2015 Munich Octoberfest (mean age 34.4 ± 13.3 years, 29% women; mean BAC 0.85 ± 0.54 g/kg), alcohol consumption was associated with arrhythmias in 30.5% (sinus tachycardia 25.9%; other arrhythmias 5.4%) [17]. BAC was significantly associated with arrhythmias overall (OR per 1-unit change 1.75; P < 0.001) and sinus tachycardia in particular (OR 1.96; P < 0.001). Sinus arrhythmia reflecting autonomic tone status was significantly reduced under the influence of alcohol.

A study investigating the effects of alcohol intoxication (8.4 ± 3.1 drinks) and hangover with rhythm monitoring and cardiac magnetic resonance imaging (MRI) among 50 patients (age 49±15 years, 40% paroxysmal AF) indicated that heart rate increased from 72±10 to 80±13 bpm during consumption (P<0.001) [50]. The hangover period was characterized by higher daily median APC count (50 vs baseline 43; P = 0.04) and reduced heart rate variability (P = 0.007). There was evidence of heightened parasympathetic activity post-hangover (P = 0.04). Three patients developed AF 11, 29 and 34 h post-binge. Cardiac MRI (2.7 ± 0.7 days post-binge) demonstrated a decrease in left atrial emptying fraction (57.9 ± 8.5 to 53.5 ± 6.7%; P = 0.003).

As mentioned, a recent community-based cohort study exploring the association between alcohol consumption and risk of AF among 107,845 individuals, determined that even light/moderate alcohol consumption may increase the AF risk (HR 1.16 for one drink/d, P<0.001) [6]. Another population-based cohort study, in addition to heavy drinking, also showed an association between moderate alcohol consumption and risk of AF in men, albeit not in women [7].

However, there are also reports claiming that low/moderate alcohol consumption, particularly wine consumption, is not associated with increased risk of AF, like the PREDIMED trial (N = 6077, HR 0.92–1.15 compared to non-drinkers) [51], and the HUNT study (N = 47,002; higher risk of AF with heavier alcohol intake, but no association at <1 drink/day for women and <2 drinks/day for men) [10]. Finally, it is suggested that frequent drinking and amount of alcohol consumed per week are significant risk factors for new-onset AF, while the amount of alcohol consumed per each drinking session is not an independent risk factor [52]. Thus, curtailing the frequency of alcohol use even if the amount is low might prevent AF.

Lack of a protective effect. Although a CV protective effect conferred by light/moderate alcohol drinking (“French paradox”) has been claimed for other CV diseases (CAD or heart failure-HF), this has not been the case for AF [1]. A population-based study of 22,824 AF- or HF-free subjects (48% men, age ≥35 years), showed that, over a median of 8.2 years, drinking alcohol in the range of 1–4 drinks/day was associated with a lower risk for HF (a 22% maximum risk reduction at 20 g/day); however, no association of alcohol consumption with onset of AF was observed [53].

Several mechanisms may be implicated in the genesis of arrhythmias in individuals with alcohol consumption (Fig. 1). The hyperadrenergic state of drinking and withdrawal may play a role, as may electrolyte abnormalities, impaired autonomic heart rate control, repolarization abnormalities with prolonged QT-intervals and worsening of myocardial ischemia or even sleep apnea [3]. As mentioned, alcohol-induced or associated with other reasons (e.g., malnutrition, vomiting, diarrhea, and dehydration) hypokalemia and hypomagnesemia can prolong the QT-interval and lead to polymorphic VT (TdP) that may degenerate into VF [24], [25], [26], [27]. Importantly, subclinical myocardial injury from chronic heavy alcohol use with ensuing fibrosis may contribute to conduction delays and thus form the substrate for reentrant arrhythmias [54].

It has been reported that stress-activated c-Jun N-terminal kinase (JNK) signaling contributes to AF development [55]. Binge alcohol-exposure may lead to activated JNK in the atria, specifically JNK2, which may account for the induction of AF. Thus, binge-alcohol activates JNK2, which subsequently phosphorylates the calmodulin kinase II (CaMKII) protein, a proarrhythmic molecule known to enhance CaMKII-driven sarcoplasmic reticulum Ca²⁺ mishandling [56]. Sarcoplasmic reticulum Ca²⁺ leak induced by activated CaMKII leads to cellular arrhythmias via enhanced early and delayed afterdepolarizations [57]. CaMKII inhibition eliminates binge alcohol-evoked arrhythmic activities. These results reveal a previously unrecognized form of alcohol-driven kinase-on-kinase proarrhythmic crosstalk [56]. Atrial JNK2 function represents a potential novel therapeutic target to prevent and/or manage AF.

Acute alcohol consumption, particularly when excessive, may alter cardiac electrophysiology and can provoke arrhythmias. A recent study, using a multiscale computational model, showed that ethanol has concentration-dependent electrophysiological effects [54]. Ethanol acutely modulates numerous targets in cardiomyocytes, including ion-channels, Ca²⁺-handling proteins and gap-junctions. Ethanol-induced gap-junction remodeling is an important determinant of ethanol-induced reentrant SVAs and VAs with high ethanol concentrations promoting both the inducibility and stability of reentrant arrhythmias, particularly in the presence of disease-associated remodeling. High ethanol concentrations have been shown to abbreviate atrial action potential duration (APD) and prolong ventricular APD [54]. High ethanol concentrations promote reentry in tissue simulations. The proarrhythmic effects of alcohol were cell-type dependent, different in atria and ventricle, and influenced by disease-related gap-junction remodeling, but also by the degree, type, and patterns of structural remodeling (fibrosis) of the underlying cardiac pathology. Thus, proarrhythmia relates to interactions between a preexisting substrate and acute changes in cardiac electrophysiology, is concentration-dependent and differs between atria and ventricles, and between healthy and diseased hearts; the proarrhythmic effects may be modulated by other co-factors, including autonomic nervous influences [54].

Indeed, acute ingestion of alcohol may activate the sympathetic nervous system (SNS) and increase circulating catecholamines leading to sinus tachycardia either with or without (postural) hypotension [58, 59]. The hypotensive effect may be enhanced by the impairment of vasoconstriction or the vasodilatory effect of short-term alcohol consumption [60] and aggravated by alcohol-induced diuresis and its attendant hypovolemia [61]. Others suggest that the acute vasodilatory effect of alcohol and the ensuing hypotension may be the triggers to SNS activation [62]. Furthermore, alcohol suppresses the baroreceptor reflex, which together with the systemic vasodilation acts to potentiate orthostatic hypotension that may induce syncope after drinking in susceptible individuals [63, 64]. On the other hand, high concentrations of alcohol constrict most blood vessels; alcohol also acts to enhance the vasoconstriction caused by catecholamines and vasopressin and inhibits endothelium-dependent vasodilation; mechanisms which may account for the hypertensive effect of chronic alcohol use [64]. All these autonomic and hemodynamic changes are also potential arrhythmogenic triggers. As mentioned, a recent study using serial cardiac MRI pre- and post-binge with continuous Holter monitoring in 50 patients indicated that binge drinking is associated with sympathetic activation followed by a 'rebound' heightened parasympathetic activity and atrial mechanical dysfunction which may explain the propensity and temporal association between binge drinking and AF [50].

Particular mention should also be made for the associated comorbidities in individuals with AUD enhancing their risk for VAs and SCD. These individuals have a higher prevalence of hypertension, obesity, diabetes, CAD, depression, and other substance abuse, compared to non-alcoholics [38]. These comorbidities may contribute to hyperadrenergic state, impaired vagal activity and baroreceptor sensitivity, and/or provide the substrate for fatal arrhythmias, especially when combined with electrolyte disturbances that might be present in alcoholics resulting in QT prolongation, TdP and cardiac arrest.

A large US study examining the national trends of arrhythmia hospitalizations and comorbid AUD, included 570,556 arrhythmia inpatients (age, 15–54 years), of whom 55,730 inpatients had comorbid AUD (~86% males; 68% older-age adults, 45–54 years) [65]. There was a significantly increasing trend in arrhythmia inpatients with AUD with comorbid diabetes, hypertension, and obesity over the five-year period. In-hospital mortality had a variable trend from 1.1% in 2010 to 1.3% in 2014.

Younger AF age of onset is associated with a higher genetic burden of AF. Genome-wide association studies (GWAS) have identified several AF-associated genes [66, 67]; titin is the gene most commonly associated with mutations in individuals with AF [68]; many of the putative AF genes act via cardiac structural remodeling, potentially in the form of an 'atrial cardiomyopathy', either during fetal heart development or as a response to stress in the adult heart [66]. Similarly, other GWAS have identified genetic correlations with alcohol consumption and AUD [69]. There is a genetic distinction between GWAS results for alcohol consumption (non-pathological drinking behavior) and alcohol dependence (pathological drinking behavior), albeit with partial overlap; GWAS single-nucleotide polymorphisms (SNPs) confirm the importance of functional variants affecting alcohol metabolism to the risk of alcohol dependence [70]. In search of a genetic relationship of alcohol consumption/dependence with other disorders, such as AF, GWAS have suggested a polygenic risk, indicating that individuals with alcohol intake above an acceptable range are at much higher risk for AF; however, certain individuals have an increased risk of AF even with low alcohol consumption, e.g., ~1 drink/day (see below) [71].

The association between alcohol intake, polygenic predisposition to AF, and incident AF was assessed in a prospective cohort study in the UK Biobank [71]. Among 376,776 participants (47.5% male, mean age 56.9 years), 6293 developed AF during a median of 6.9 years. Alcohol consumption was associated with AF (HR, 1.10 for intake above an acceptable range:>98 g/week for women or >196 g/week for men; HR, 1.04 per 100 g/week). An AF polygenic risk score was associated with AF (HR, 1.38 per SD). In models including both alcohol and the AF polygenic risk score, each remained associated with AF. The 5-year cumulative risk of AF for individuals with alcohol intake above an acceptable range and in the highest decile of polygenic risk was 2.33%, compared with 0.69% for those with alcohol intake within an acceptable range and in the lowest decile of polygenic risk.

A multicenter controlled trial examined the effect of abstinence from alcohol on secondary prevention of AF among 140 adults with history of AF (85% men; age, 62±9 years) who consumed ≥10 drinks/week (with 1 drink containing ~12 g of alcohol), and were in sinus rhythm at baseline [72]. The study group were randomly assigned in a 1:1 ratio to either abstain from alcohol (n = 70) or continue their usual alcohol consumption (n = 70). Over 6 months, patients in the abstinence group reduced their weekly alcohol intake by 87.5%, and patients in the control group by 19.5%. After a 2-week blanking period, AF recurred in 53% in the abstinence group and in 73% in the control group. The abstinence group had a longer period before AF recurrence than the control group (HR, 0.55; P = 0.005). The AF burden over 6 months was significantly lower in the abstinence group (median percentage of time in AF, 0.5% vs. 1.2%; P = 0.01). The authors concluded that abstinence from alcohol reduced arrhythmia recurrences in regular drinkers with AF.

In the Atherosclerosis Risk in Communities Study (ARIC) (N = 15,222; 2886 or 19% being former drinkers; 1631 incident AF cases over a mean of 19.7 years), after adjustment for potential confounders, every decade abstinent from alcohol was associated with a ~20% lower rate of incident AF [73].

A retrospective observational study (N = 282) compared the clinical characteristics and long-term arrhythmic events in patients with ACM (94 individuals) and idiopathic dilated cardiomyopathy (IDCM) (188 individuals) [20]. Over a median of 38 months, 42 patients died and 79 underwent heart transplantation: 31 (33%) with ACM vs 90 (48%) with IDCM (P = 0.017); 37 (13%) patients suffered malignant VAs: 18 (19%) ACM vs 20 (11%) IDCM (P = 0.048). On multivariate analysis, left bundle branch block (LBBB) (OR 2.4; P = 0.015) and alcoholic etiology (OR 2.3; P = 0.026) were the only independent predictors of malignant VAs. No malignant VAs were recorded during follow-up in ACM patients if left ventricular ejection fraction had increased or remained ≥40%.

According to a cross-sectional study using the US Nationwide Inpatient Sample database (2007–2014), among 75,430 hospitalizations of adults with ACM, 48% had arrhythmias [21]. Individuals with arrhythmias tended to be older (56.9 vs 53.2-year-old) and male (89.5% vs 81.9%). The most prevalent arrhythmias were AF/AFlu (31.5%), followed by VT (7.9%). Excluding cardiac arrest, arrhythmias were not associated with increased in-hospital mortality.