Author + information
- Received May 27, 2018
- Revision received July 13, 2018
- Accepted July 20, 2018
- Published online November 19, 2018.
- Aleksandr Voskoboinik, MBBSa,b,c,
- Benedict T. Costello, MBBSb,c,
- Elana Kalman, MBBSc,
- Sandeep Prabhu, MBBSa,b,c,
- Hariharan Sugumar, MBBSa,b,c,
- Geoff Wong, MBBSa,
- Chrishan Nalliah, MBBSa,
- Liang-Han Ling, MBBS, PhDb,c,
- Alex McLellan, MBBS, PhDa,b,c,
- Thushan Hettige, MBBSc,
- Fabian Springer, MBBSb,d,
- Andre La Gerche, MBBS, PhDb,
- Jonathan M. Kalman, MBBS, PhDa,e,
- Andrew J. Taylor, MBBS, PhDb,c and
- Peter M. Kistler, MBBS, PhDb,c,e,∗ ()
- aDepartment of Cardiology, Royal Melbourne Hospital, Melbourne, Australia
- bBaker Heart & Diabetes Institute, Melbourne, Australia
- cHeart Centre, The Alfred Hospital, Melbourne, Australia
- dDepartment of Diagnostic and Interventional Radiology, University Hospital Tübingen, Tübingen, Germany
- eDepartment of Medicine, University of Melbourne, Melbourne, Australia
- ↵∗Address for correspondence:
Prof. Peter Kistler, Heart Centre, Alfred Hospital, Commercial Road, Melbourne NA, Australia.
Objectives This study sought to determine the impact of regular alcohol consumption on left atrial (LA) mechanical and reservoir function.
Background Earlier studies suggest that regular alcohol intake is associated with increased atrial fibrillation (AF) and LA dilatation.
Methods This study prospectively enrolled 160 patients with paroxysmal or persistent AF to undergo 3-T cardiac magnetic resonance (CMR) imaging in sinus rhythm. Patients self-reported alcohol consumption in standard drinks (∼12 g alcohol) per week over the preceding 12 months and were categorized into 4 groups: 1) lifelong nondrinkers; 2) mild drinkers (3 to 10 standard drinks/week); 3) moderate drinkers (11 to 20 standard drinks/week); 4) heavy drinkers (>20 standard drinks/week). Permanent AF and cardiomyopathy were excluded. On CMR, maximum LA volume (LAmax) and minimum LA volume (LAmin), global LA emptying fraction (LAEF) as (LAmax − LAmin) / LAmax, and LA reservoir function as (LAmax − LAmin) / LAmin were calculated.
Results Regular alcohol consumption (mean 15.8 ± 6.9 standard drinks/week, n = 120) was associated with larger LA size (LA volume index 50 ± 13 ml/m2 vs. 43 ± 12 ml/m2; p = 0.005), reduction in LAEF (40 ± 14% vs. 52 ± 15%; p < 0.001), and reduction in reservoir function (77 ± 48% vs. 119 ± 63%; p < 0.001) compared with lifelong nondrinkers (n = 40). There were progressive dose-related impairments in LAEF (mild 45.4 ± 13.5% vs. moderate 39.1 ± 14.7% vs. heavy drinkers 35.6 ± 12.6%; p < 0.01) and reservoir function (mild 95.8 ± 55.6% vs. moderate 74.8 ± 47.1% vs. heavy drinkers 61.7 ± 34.4%; p < 0.01). Predictors of atrial mechanical dysfunction included weekly alcohol intake (p = 0.001), older age (p = 0.018), and persistent AF (p = 0.016), but not binge drinking or beverage type.
Conclusions In patients with AF, habitual alcohol consumption is associated with significantly increased LA size and atrial mechanical dysfunction compared with nondrinkers.
Alcohol consumption has emerged as an important risk factor for atrial fibrillation (AF) (1). In a meta-analysis of 7 studies involving 859,420 patients and 12,554 AF cases, AF incidence increased by 8% for each additional standard drink per day (2). Population-based studies have demonstrated an association between regular alcohol consumption and left atrial (LA) enlargement in a dose-related manner (3). LA size is a well-established determinant of AF recurrence. The direct contribution of alcohol to LA dilatation and AF risk has been estimated to be ∼ 24%; however, the contribution of AF itself and associated conditions such as hypertension, obesity, and sleep apnea can be difficult to untangle (4). Daily drinking has been linked with atrial fibrosis in a dose-dependent manner (5).
Moreover, excessive alcohol consumption is a well-established cause of dilated cardiomyopathy. The putative mechanisms responsible for alcohol-mediated cardiac chamber enlargement include oxidative stress, inflammation, and apoptosis (6). We hypothesized that the thinner-walled atrium may be more vulnerable than the ventricle to the potentially deleterious effects of “moderate” alcohol consumption and an important factor in AF pathogenesis. Moreover, atrial mechanical dysfunction on echocardiography has been correlated with a higher risk of LA thrombus formation (7) and may provide incremental diagnostic information over CHA2DS2-VASc (congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex) score for predicting stroke risk in patients with AF (8). To date there have been limited studies evaluating the effect of regular alcohol consumption on atrial mechanical function. We sought to evaluate the impact of alcohol consumption on LA size and mechanical function by using 3-T cardiac magnetic resonance (CMR) imaging in patients with AF.
Patient selection and study design
This is a single-center observational study conducted from April 2016 to May 2018. We aimed to prospectively recruit 160 outpatients with paroxysmal or persistent AF and a rhythm-control strategy to undergo 3-T CMR in sinus rhythm. AF type (paroxysmal or persistent) and burden were determined according to the patients’ AF profile and the percentage of time spent in AF, respectively, in the 12 months before CMR. Patients self-reported their average alcohol consumption in standard drinks (∼12 g alcohol) per week (standard drinks/week) over the preceding 12 months, and patients consuming ≥3 standard drinks/week were categorized as regular drinkers. Patients also categorized their beverage of choice as wine, beer, or spirits. Regular drinkers were subdivided into 3 groups on the basis of quantity consumed, as follows: mild (3 to 10 standard drinks/week), moderate (11 to 20 standard drinks/week), or heavy (>20 standard drinks/week) drinkers. Lifelong nondrinkers were used as control subjects. We aimed to recruit 40 consecutive patients in each of the 4 groups.
Exclusion criteria were as follows: 1) occasional drinking, defined as >1 consecutive month of nondrinking during the preceding 12 months; 2) former drinking; 3) permanent AF; 4) alcoholic liver cirrhosis; 5) significant renal impairment (estimated glomerular filtration rate <30 ml/min); and 6) significant known structural heart disease (left ventricular [LV] ejection fraction [LVEF] <40% or previous myocardial infarction with regional wall motion abnormality).
Our CMR scanning and analysis protocol has been described previously (9). All CMR examinations were performed on a 3-T scanner (Magnetom Prisma, Siemens Healthineers, Erlangen, Germany). All post-processing of images was performed using a dedicated CMR analysis workflow of the Syngo.via software package (Siemens Healthineers).
Cine sequences were acquired in the 4-chamber, 2-chamber, 3-chamber, and short-axis views by using an electrocardiogram-gated balanced steady-state free precision (SSFP) sequence in expiration, extending from the pulmonary veins to the LV apex (4-mm slice thickness, no gap). LV mass, end-diastolic volume, end-systolic volume, and LVEF were analyzed on commercially available post-processing software, with LV volumetric analysis performed using the summation of disks method. Papillary muscles were regarded as part of the ventricular cavity. Measurements were indexed to body surface area.
Because the focus of the scan was the left atrium, radiographers optimized the scan to minimize LA foreshortening where possible and image the LA length perpendicular to the mitral annular plane. Maximum LA volume (LAmax) and minimum LA volume (LAmin) were measured in sinus rhythm by using the biplane area-length method, as shown in Figures 1A to 1D, with the pulmonary veins and LA appendage carefully excluded at their junctions with the LA. LA length was measured from the midpoint of the mitral annulus to the midpoint of the superior LA wall. Global LA emptying fraction (LAEF) was (LAmax − LAmin) / LAmax. LA reservoir function was (LAmax − LAmin) / LAmin.
CMR measurements were performed by 2 cardiologists, and interobserver variability for 100 consecutive atrial measurements was calculated using Pearson’s correlation coefficient, with r = 0.841.
The primary outcome measure was a comparison of LAEF between drinkers and nondrinkers, with LA volume indexed (LAVI) and LA reservoir function used as secondary endpoints. On the basis of previous studies looking at comparisons of electrical remodeling between drinkers and nondrinkers, we calculated that to detect a minimum absolute difference of 10%, we would need to enroll ∼100 patients to provide a power of 0.8 at an alpha value of 0.05.
The Shapiro-Wilk test was performed to confirm normal distribution of data, and Student’s t-test was then performed. All continuous data are summarized as mean ± SD or median, where appropriate. Comparisons of the clinical characteristics among groups were performed using a chi-square or Fisher exact test. The Mann-Whitney U test was used for continuous variables where normal distribution was not present.
Multiple linear regression was performed to determine multivariate clinical predictors of atrial mechanical dysfunction, by using LAEF as a continuous dependent variable. Data analysis was performed using Statistical Package for the Social Sciences for Windows (SPSS version 23, IBM Corp., Armonk, New York). A p value <0.05 was considered statistically significant. The study was prospectively approved by Alfred Health Human Research Ethics Committee, and all patients provided written informed consent to undergo CMR for the purposes of this research study.
In total, 160 participants with a history of AF underwent CMR in sinus rhythm between April 2016 and May 2018. Baseline characteristics of lifelong nondrinkers (n = 40) and regular drinkers (n = 120) are shown in Table 1. Patients in both groups were predominantly male and did not differ significantly with respect to age, AF type and duration, weight, and medical comorbidities. Mean and median alcohol intake was 15.8 ± 6.9 and 15 standard drinks per week (∼180 g), respectively, for 25.1 ± 10.3 years.
Regular drinkers were classified as mild (3 to 10 standard drinks/week; n = 40), moderate (11 to 20 standard drinks/week; n = 40), or heavy drinkers (>20 standard drinks/week; n = 40), as shown in Table 2. Most patients in both drinking groups were wine or beer drinkers, and a significant proportion consumed both beverage types on a regular basis. A few patients in both drinking groups were frequent binge drinkers (as defined by ≥5 standard drinks in a 2-h sitting more than monthly).
CMR findings are presented in Table 3. The majority of patients had preserved LV systolic function, with normal LV volumes and mass. However, regular long-standing drinkers had significantly larger atria (right atrial area 25.3 ± 5.9 cm2 vs. 22.7 ± 4.8cm2; p = 0.02; LAVI 50 ± 13 ml/m2 vs. 43 ± 12 ml/m2; p = 0.005), with impaired mechanical function (LAEF in sinus rhythm 40 ± 14% vs. 52 ± 15%; p < 0.001) and reservoir function (77 ± 48% vs. 119 ± 63%; p < 0.001) compared with lifelong nondrinkers.
There were significant dose-related reductions of both LAEF (mild 45.4 ± 13.5% vs. moderate 39.1 ± 14.7% vs. heavy drinkers 35.6 ± 12.6%; p < 0.01) and reservoir function (mild 95.8 ± 55.6% vs. moderate 74.8 ± 47.1% vs. heavy drinkers 61.7 ± 34.4%; p < 0.01), as shown in Figure 2. LAVI was not significantly different between moderate and heavy drinkers (50.5 ± 15.6 ml/m2 vs. 51.2 ± 9.8; p = 0.82) although both of these groups had significantly larger atria compared with nondrinkers (42.9 ± 12.0 ml/m2), as shown in Figure 3. Type of alcoholic beverage consumed (beer, wine, and/or spirits) did not significantly affect LAVI (p = 0.637), LAEF (p = 0.335), or reservoir function (p = 0.234).
Multiple linear regression was undertaken to determine multivariate clinical predictors of atrial mechanical dysfunction (on the basis of LAEF). Significant predictors included number of standard drinks per week (p = 0.001), older age (p = 0.018), and persistent AF (p = 0.016). Sex (p = 0.105), hypertension (p = 0.363), diabetes mellitus (p = 0.140), history of binge drinking (p = 0.195), time since first AF episode (p = 0.549), LVEF (p = 0.240), and CHA2DS2-VASc score (p = 0.121) were not statistically significant.
This cross-sectional observational study of 160 patients with a history of AF aimed to determine the impact of regular alcohol consumption on atrial and ventricular parameters by using high-definition CMR. The main findings were as follows:
1. Regular alcohol drinkers had significantly larger right and left atria with impaired mechanical and reservoir function (Figure 4) compared with nondrinkers, despite similar ventricular parameters.
2. A dose-related effect of alcohol consumption was demonstrated, with more severe atrial mechanical dysfunction associated with higher alcohol intake.
3. Habitual regular alcohol consumption, but not binge drinking or alcohol beverage type, was associated with atrial mechanical dysfunction.
4. Regular alcohol consumption, even at mild levels, was a significant predictor of impaired LA emptying fraction.
As a burgeoning epidemic in Western countries, AF is increasingly seen as a lifestyle-related condition. Although binge drinking is a well-recognized acute precipitant of acute AF (‘holiday heart syndrome’ ), long-term habitual alcohol consumption is increasingly appreciated as a risk factor for incident AF (11), even at levels as low as 1 standard drink per day (12,13). In the present study, even “mild” drinkers (average intake ∼7 drinks/week) had impaired mechanical function, whereas “moderate” alcohol drinkers (average ∼16 drinks/week) also had significantly dilated atria compared with nondrinkers. The putative mechanisms are beyond the scope of this clinical study but may include direct myopathic and fibrotic effects on the thinner-walled atria. Consistent with previous studies, lower levels of alcohol intake are less likely to affect ventricular size and function, with “alcoholic cardiomyopathy” predominantly described in persons consuming >80 g/day (∼45 drinks/week) (14).
The deleterious downstream structural effects of heavy alcohol consumption on the ventricle are well established; however, low to moderate alcohol intake is generally considered cardioprotective in patients with dyslipidemia and coronary disease (15,16) . The impacts of alcohol on the atrium are only beginning to emerge. In a population-based echocardiographic study, McManus et al. (4) observed a 0.16-mm (95% confidence interval: 0.10 to 0.21) increase in LA dimension for each 10 g alcohol intake per day and estimated that 24% of the incident AF risk was related to alcohol-mediated LA enlargement. Singh et al. (17) similarly reported an increase in echocardiographic LA volume index with relatively modest levels of alcohol intake (adjusted odds ratio: 1.81; 95% confidence interval: 1.13 to 2.90; p = 0.01). The present study used high-definition 3-T CMR volumetric analysis to determine dose-related effects of regular alcohol intake on atrial dysfunction.
Mechanisms of alcohol toxicity on cardiac function
Postulated mechanisms for direct cardiotoxic effects of alcohol include inflammation, oxidative stress (18), apoptosis (19), mitochondrial dysfunction (20), deranged fatty acid metabolism (21), and accelerated protein degradation (22,23). Although these effects have been predominantly studied in the ventricle, recent studies suggest that the much thinner-walled atria may be more vulnerable to myopathic effects at moderate doses sustained over a long period. Animal studies suggest that even 1 week of excessive alcohol consumption may cause a significant reduction in sodium and calcium atrial current densities (24). Adverse atrial electrical remodeling ensues with increasing consumption and culminates in fibrosis. In patients undergoing AF ablation, the degree of atrial fibrosis, as characterized by regional low voltage, increased by 10% in the left atrium for each standard drink consumed (5). In patients with a history of atrial flutter who were undergoing electrophysiological study, atrial effective refractory periods were significantly shorter in drinkers (≥1 drink/day) compared with nondrinkers (25).
As the most commonly consumed “drug” in the United States (26), alcohol’s effects on the atrium in the ever-growing group of patients with AF have important implications for prognosis. In a study of 122 patients undergoing pulmonary vein isolation for paroxysmal AF, nondrinkers had the lowest recurrence rates (81%), followed by moderate (69%) and then heavy drinkers (35%) (5). Our findings of impaired atrial mechanical function and larger LA size in regular drinkers may also explain why alcohol consumption in patients with AF was associated with higher stroke rates in a large cohort study (27), even after adjustment for established clinical risk factors.
Recent studies suggest that eradication of “alcohol abuse” may result in 73,000 fewer AF cases in the United States (28). Although alcohol abstinence may be appropriate in patients with AF, there is insufficient evidence to recommend complete abstinence in the broader group of cardiac patients. Whereas some studies suggest a U-shaped relationship between alcohol consumption and all-cause mortality (29), others argue that beneficial associations of low-intensity alcohol are largely attributable to inappropriate reference group selection and poor adjustment for confounders (30). Nevertheless, in patients with AF, a reduction in alcohol intake to <3 standard drinks/week, among other cardiometabolic risk factor reduction measures, has been shown to reduce AF recurrence rates (31). This study supports the notion that regular alcohol intake at even modest doses may be associated with adverse LA mechanical remodeling, and this may explain the vulnerability to AF recurrence in these patients.
Because this was an observational study, we are able to demonstrate only association, rather than causation. Moreover, atrial parameters before the development of AF were unavailable in these patients. Alcohol intake was self-reported, and the actual amount consumed may be higher and subject to recall and misclassification bias. Although regular drinkers had evidence of atrial mechanical dysfunction compared with lifelong nondrinkers, the clinical significance of these changes with respect to clinical endpoints such as stroke or mortality is unknown. Further studies are needed to determine whether these changes are reversible with abstinence from alcohol.
In patients with AF, moderate alcohol consumption is associated with significantly increased LA size and impaired atrial mechanical function. This may in part explain the propensity for AF recurrence and stroke risk in patients with AF who regularly consume moderate amounts of alcohol.
COMPETENCY IN MEDICAL KNOWLEDGE: Regular alcohol consumption results in dose-related biatrial dilatation, mechanical dysfunction, and impaired reservoir function compared with nondrinkers. In patients with a history of AF, excessive alcohol consumption is associated with adverse atrial remodeling and may represent a potentially modifiable risk factor for AF progression.
TRANSLATIONAL OUTLOOK: Although this study adds to the growing evidence base that alcohol adversely affects atrial substrate, the mechanisms of toxicity from habitual consumption, particularly at moderate levels, remain poorly understood. Further research is required to investigate the effect of abstinence from alcohol on atrial size, mechanical function, and AF burden.
This research is supported in part by the Victorian Government’s Operational Infrastructure Funding. Dr. Voskoboinik is supported by co-funded National Health and Medical Research Council (NHMRC)/National Heart Foundation (NHF) post-graduate scholarships and Baker Institute Bright Sparks scholarships. Dr. Jonathan Kalman is supported by an NHMRC practitioner fellowship. Dr. Prabhu has received fellowship support from Abbott and Boston Scientific. Dr. Sugumar is supported by an NHMRC/NHF post-graduate scholarship; and by the Cardiac Society of Australia and New Zealand. Dr. McLellan is supported by a Heart Foundation post-doctoral fellowship. Dr. La Gerche is supported by an NHMRC/NHF Future Leader Fellowship. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Michael R. Gold, MD, served as Guest Editor for this paper.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex
- cardiac magnetic resonance
- left atrial
- left atrial emptying fraction
- maximum left atrial volume
- minimum left atrial volume
- left atrial volume indexed
- left ventricular
- left ventricular ejection fraction
- Received May 27, 2018.
- Revision received July 13, 2018.
- Accepted July 20, 2018.
- 2018 American College of Cardiology Foundation
- Voskoboinik A.,
- Prabhu S.,
- Ling L.H.,
- Kalman J.M.,
- Kistler P.M.
- Larsson S.C.,
- Drca N.,
- Wolk A.
- Gonçalves A.,
- Jhund P.S.,
- Claggett B.,
- et al.
- McManus D.D.,
- Yin X.,
- Gladstone R.,
- et al.
- Qiao Y.,
- Shi R.,
- Hou B.,
- et al.
- Steiner J.L.,
- Lang C.H.
- Kim D.,
- Shim C.Y.,
- Hong G.R.,
- et al.
- Costello B.T.,
- Springer F.,
- Hare J.L.,
- et al.
- Gallagher C.,
- Hendriks J.M.L.,
- Elliott A.D.,
- et al.
- Larsson S.C.,
- Drca N.,
- Wolk A.
- Kodama S.,
- Saito K.,
- Tanaka S.,
- et al.
- Rehm J.,
- Hasan O.S.M.,
- Imtiaz S.,
- Neufeld M.
- Xi B.,
- Veeranki S.P.,
- Zhao M.,
- Ma C.,
- Yan Y.,
- Mi J.
- Piano M.R.
- Tsiplenkova V.G.,
- Vikhert A.M.,
- Cherpachenko N.M.
- Hasin D.S.,
- Stinson F.S.,
- Ogburn E.,
- Grant B.F.
- Overvad T.F.,
- Rasmussen L.H.,
- Skjøth F.,
- et al.
- Whitman I.R.,
- Agarwal V.,
- Nah G.,
- et al.
- Knott C.S.,
- Coombs N.,
- Stamatakis E.,
- Biddulph J.P.
- Pathak R.K.,
- Middeldorp M.E.,
- Lau D.H.,
- et al.