Drinking alcohol increases the likelihood an episode of atrial fibrillation (AF) will occur within a few hours, and the more alcohol consumed, the higher the risk, a new study shows.
Any alcoholic drink captured by real-time recordings was associated with more than twofold greater odds of an AF event in the subsequent 4 hours (odds ratio [OR], 2.26; 95% CI, 1.50 – 3.40), and two or more drinks were linked to more than a threefold higher risk (OR, 3.58; 95% CI, 1.63 – 7.89).
“The highest risk occurred at 3 to 4 hours after alcohol consumption, which fit very well with what our patients told us at the beginning, but the significant effects lasted until almost 9 hours,” said Gregory M. Marcus, MD, University of California, San Francisco.
“The probability a particular atrial fibrillation event will occur is not simply due to random chance” but, rather, “alcohol consumption is a modifiable exposure that may empower patients to influence their individual risk of an atrial fibrillation event,” he said in a late-breaking clinical trial session at the American College of Cardiology 2021 Scientific Session (ACC.21).
Chronic alcohol consumption is a well-established risk factor for the initial diagnosis of AF, but studies, as well as patients, frequently describe an acute relationship, sometimes referred to as the “holiday heart,” where alcohol consumption triggers a discrete AF event shortly thereafter, he noted.
A recent study by the investigators provides some biologic plausibility, demonstrating that AF patients infused with alcohol exhibited a substantial reduction in the pulmonary vein atrial effective refractory period, a finding that would be expected to render their atria more prone to fibrillate. Studies have also shown that advising drinkers with AF to abstain from alcohol results in fewer AF episodes.
HOLIDAY Monitors
For the HOLIDAY (How Alcohol Induces Atrial Tachyarrhythmias) Monitors study, 100 paroxysmal AF patients who consumed at least one alcoholic drink a month wore a continuous ECG monitor (LifeWatch or Zio patches) and a transdermal ankle alcohol monitor (SCRAM) for 4 weeks.
Patients (mean age, 64 years; 79% male) were asked to press an activator button on the ECG monitor every time they had an alcoholic drink, providing a time stamp for each. Finger-stick blood spots were also collected for phosphatidylethanol testing (PEth) to corroborate drinking events.
Most patients were White (85%), about half had hypertension, a third had other cardiovascular morbidities, and antiarrhythmic drug use was rare. The ECG monitor was worn for a median of 27 days.
Recordings revealed patients consumed a median of 19 drinks (interquartile range [IQR], 10 – 38) on a median of 12 different days (IQR, 7 – 21), and 56 patients had at least one AF episode on a median of 5 different days (IQR, 2.5 – 12.5).
PEth results were significantly correlated with the number of real-time recorded drinks and with transdermal alcohol-detected events.
Every 0.1% increase in the inferred peak blood alcohol concentration over the previous 12 hours was associated with a 38% increase in the odds of an AF episode (1.38; 95% CI, 1.04 – 1.83).
“We were not able to identify any threshold effects, such as a particular number of drinks that were required or particular alcohol concentration, but rather the relationships appeared to be fairly linear,” he said.
Although the case-crossover design should mitigate against confounding as patients served as both case and control subjects, it’s not possible to exclude a concomitant behavior, such as smoking or caffeine, as the true causal factor, Marcus said. Poor sleep can result from drinking and trigger AF, but in that case would serve more as a mediator than a confounder.
A “Clear Nail in the Coffin”
“This is just a fantastic study design and really addresses this problem head on because for many years we kind of debated back and forth with observational studies and all the limitations that come with it. But this is, I think, a very clear nail in the coffin,” Dhanunjaya Lakkireddy, MBBS, chair of the ACC Electrophysiology Council, said during a discussion of the results.
He asked about the pathophysiology behind the observations and the potential interplay of the associated sympathetic surge and hypoxia, sleep apnea exacerbation, and other things oftentimes connected with alcohol. “I wonder if there was a time stamp on the correlation between when they consumed alcohol and did they go to bed 3 or 4 hours after they started noticing these episodes of afib. Any further light on that?”
Marcus said not yet, but that the data should provide information, for example, about changes in heart rate variability to help inform about autonomic tone and how that may have mediated these relationships.
“The majority of the afib episodes did occur at night, interestingly, which would be consistent with that idea,” he said. “Alcohol is thought to have this fairly near-term sympathetic effect and more of a parasympathetic effect over time and the fact that there does seem to be this delay may suggest that the parasympathetic tone is especially important.”
During a press conference, Lakkireddy highlighted the paradox faced by clinicians. Upon hearing the results, he said, “a patient emailed saying ‘On the one hand, you guys are telling us that a glass of wine keeps the cardiologists away and now you’re coming out and telling us that drinking alcohol is going to take me one step closer to an electrophysiologist.’ Well, what’s this paradox? How are we going to really answer this impact of alcohol on plumbing versus the electrical system of the heart?”
Marcus replied that the results provide a “good opportunity actually to educate the public about the differences in terms of arrhythmias versus heart attacks,” which are often conflated, but that more research in this area is needed.
“We don’t know what the net risk benefit is for those who drink in moderation. I think that’s frankly begging for a true randomized trial,” he said.
Marcus said he frequently gets asked these questions in his own practice and for patients with AF, “it’s now fairly clear we should recommend, if you really want to do everything you can that’s under your power to reduce the chance you’ll suffer an event, I would cut out alcohol.”
“For someone without afib, and especially someone who may be at risk for coronary disease, I’d tell them, the honest answer is we don’t know.”
Anne Curtis, MD, University of Buffalo, New York, said “the results may be very disappointing to a lot of people who enjoy a glass of wine or beer at various times.” Still there was sufficient drinking in the study to test the hypothesis and “between correlating the episodes of drinking and the AF episodes and showing the time course, it really is an impressive and tight correlation.”
“So, very impressed with the findings and I do think this is very helpful in the whole armamentarium of lifestyle modification for preventing atrial fibrillation,” she said.
The study was funded by the National Institute on Alcohol Abuse and Alcoholism. Marcus reports receiving research grants from the National Institutes of Health, Patient-Centered Outcomes Research Institute, Tobacco-Related Disease Research Program, Medtronic, Eight Sleep, and Baylis; consulting for InCarda Therapeutics and Johnson & Johnson; and equity in InCardia Therapeutics. Lakkireddy reported having no relevant financial disclosures. Curtis reported consultant fees/honoraria from Abbott, Janssen Pharmaceuticals, Medtronic, Milestone Pharmaceuticals, and Sanofi Aventis; and serving on a data safety monitoring board for Medtronic.
American College of Cardiology 2021 Scientific Session (ACC.21): Abstract 411-14. Presented May 17, 2021.
Follow Patrice Wendling on Twitter: @pwendl. Follow ACC.21 coverage from theheart.org | Medscape Cardiology on Twitter and Facebook.