Acute Alcohol Consumption Raises Risk for Atrial Fibrillation

John M. Mandrola, MD; Gregory M. Marcus, MD, MAS


May 20, 2021

This transcript has been edited for clarity.

John M. Mandrola, MD: This is John Mandrola, from | Medscape Cardiology. I'm pleased to be here with Dr Gregory Marcus from the University of California, San Francisco (UCSF), who presented a very interesting study today as part of the late-breaking clinical trials at the virtual American College of Cardiology meeting. Greg, welcome.

Gregory M. Marcus, MD, MAS: Thank you.

Mandrola: Your study looked at the effects of alcohol intake on discrete atrial fibrillation (AF) events. Tell us what you found.

Marcus: We determined participants' alcohol consumption in real time by having them press a button on their ECG monitor every time they had a standard alcoholic drink, and by having them wear a transdermal alcohol sensor. From this, we found that alcohol consumption and increasing alcohol concentration was indeed associated with a higher risk of a discrete AF event happening just a few hours later.

Mandrola: And these were patients who had AF?

Marcus: Correct. All of these patients already had paroxysmal AF.

A Novel Investigation of Alcohol's Acute Effects

Mandrola: It's pretty well known that alcohol has a strong association with AF, both from observational data as well as an elegant randomized trial published in 2020 in The New England Journal of Medicine concluding that alcohol abstinence reduced AF. What was the motivation behind your experiment and how does it differ from this earlier research?

Marcus: The great majority of the evidence shows that chronic alcohol consumption seems to heighten the risk for a diagnosis of AF. That important randomized trial you noted looked at abstinence as a predictor of AF burden. However, no previous study has looked at the acute relationship, which is really what our patients usually tell us about — that when they drink alcohol it seems to act as an acute trigger for heightened risk of a discrete AF event.

The other relevant background question that is even broader than alcohol is, why does AF happen when it happens? Is it completely random, or is there some way to actually influence whether it happened? And is there a modifiable aspect that's under the control of patients that might influence the risk of a discrete AF event?

We really honed in on these temporal relationships that are generally difficult to study because they require continuous ambulatory monitoring. Therefore, the hypothesis we sought to test that has been elusive to previous research was whether discrete consumption of alcohol would be associated with a near-term or nearly immediate and discrete AF episode.

Mandrola: All of us who take care of these patients have likely warned them that if they drink, they're at risk of getting these episodes. Is that sufficient?

Marcus: One of the issues is that alcohol is so ubiquitous. And we know that patients often will ascribe various triggers to their events. In fact, we've published data where we simply ask patients, "What do you think triggers your AF?" and we get a panoply of answers. This begs the question, is alcohol just so commonly consumed that it gets attributed as a cause of their AF without there truly being a relationship?

In this study, we were able to follow the same people and look over time to see when they consumed alcohol vs when they didn't. This was a nice way to address confounding, because each person served as their own control. Indeed, we found that when they consumed alcohol, temporally there was an AF episode that was associated shortly thereafter.

Dose-Response Relationship

Mandrola: How much of a risk for AF was there and how strong was the correlation?

Marcus: When a participant indicated that they consumed one drink of alcohol, there was a twofold greater risk of AF for that individual. Interestingly, there did seem to be a dose-response relationship. So, if they had two or more drinks, the heightened risk was 3.5-fold greater. Similarly, data obtained from the alcohol sensor showed that, essentially, the more alcohol consumed, the higher the risk.

A common question is whether there's a threshold effect. Is there an amount of alcohol that is safe? We couldn't identify such a threshold effect, although there's the important caveat that this was not a general population we were studying, but rather people who had a previous diagnosis of AF.

Mandrola: I must admit to some ignorance. I didn't know that there was a blood measurement of alcohol concentration, almost like you could use with hemoglobin A1c in studies of diabetes. And I also didn't know about this transdermal device that participants wore.

Marcus: The blood test you're referring to is called phosphatidylethanol. It's fascinating in that it's only formed when one consumes ethanol; it's not naturally formed. Using this means we don't have to worry about any false positives. On the other hand, it may not be super sensitive. It's been primarily used in research, more than anything else, to detect binge drinking. We did find a good correlation between the level of phosphatidylethanol and how many times our participants push the button, giving us good evidence that pushing the button really did reflect alcohol events.

The alcohol sensor we use is something called a SCRAM device, which is routinely used in law enforcement for people with DUIs and such. In fact, that brought up an issue with trying to convince our patients to participate because they were worried about the stigma of wearing those alcohol sensors on their ankle. We tried to get around that by putting a big UCSF Cardiology sticker on it so they could tell people they're participating in a study and hopefully avoid that stigma.

Cutting Alcohol and Quality of Life

Mandrola: What's the main critical takeaway in terms of translating this to expand our knowledge of AF?

Marcus: I think there are two big takeaways. The big picture is, we think these data demonstrate that an AF event is happening not due to chance alone. There are determining factors — in this case, alcohol. Clearly, alcohol doesn't explain every single AF episode, but it perhaps opens up this field to start thinking about what are the environmental influences. That may increase the chance of understanding how many factors are really under the control of a patient, which would be really empowering for them to understand.

The second is more specific to alcohol, which is the growing evidence that not only is chronic alcohol consumption a risk factor for eventually developing AF, but there also appears to be an immediate relationship. That can be, in some ways, more convincing. It's been shown in other kinds of behavioral intervention research (eg, with tobacco) that when there is an immediate consequence to a behavior, that can often be a more compelling reason to avoid it.

So, for our patients with AF, and especially those that drink alcohol, I think this adds to the repertoire by showing that there's very good evidence now indicating that there's a causal relationship. And if you really want to do everything you can to avoid AF, you probably should abstain from alcohol or at least minimize consumption.

Mandrola: I'm hearing the voice of my friend, the Italian electrophysiologist Luigi Di Biase, saying, "Mandrola, for Italians, enjoying life is important and they can't give up their wine, so we should just ablate them." You're one of the preeminent alcohol researchers in the field; how would you answer Luigi?

Marcus: It's a very fair comment, and it would be naive of me to simply say, well, just avoid alcohol. I'm always a little reticent to give that answer for precisely the reason that it is important that we acknowledge quality of life. In fact, many of the reasons we treat arrhythmias at all is primarily to help with quality of life. We can't ignore the fact that alcohol — a good glass of wine — can be important to people's quality of life as well.

That then brings us to the next question: Is ablation sufficient to render one no longer prone to alcohol-induced AF? We need to figure that out or whether we can achieve the same thing with an antiarrhythmic.

There's also this interesting question as to the potential benefits of light alcohol consumption. The data are very consistent that binge drinking, alcohol abuse, and those things heighten risk for heart attack, heart failure, and all kinds of other problems. There's really no question there.

I think we can also be very comfortable counseling our patients to avoid, for example, having more than two drinks in 24 hours. But when it comes to that moderate alcohol consumption, I think the jury remains out. We need to contribute to that with more science and, for example, to perform randomized trials in this field.

Alcohol and the Heart: Questions Remain

Mandrola: Earlier this year, you published a study showing that acute alcohol can change pulmonary vein refractory periods. Can you speculate on whether this relationship is causal? It sure seems that way. And if so, does that give us any clues about AF pathophysiology?

Marcus: My main motivation to conduct this research is that alcohol is such a commonly consumed substance, and yet there's so little we know about it. If we could truly understand why this happens, then we might learn something new about AF that would reveal some relevant therapeutic target.

We demonstrated in a randomized trial that alcohol seems to shorten the refractory period within the pulmonary vein. That's an electrophysiologic phenomenon we would expect to render the atria more prone to fibrillate. We still don't understand precisely why that happens. We know from some animal models that there could be some effect on potassium channels, for example, but that hasn't been shown in humans.

I think the other interesting insight from the current study is that the relationship appears to have a delayed effect. That fits very well also with what our patients tell us, that there seem to be several hours between alcohol consumption and pace of AF onset. That may point to more of an autonomic effect — for example, a parasympathetic effect — that occurs not during the acute phase of alcohol consumption, or it could be the effect of a metabolite. That's a rich area for research that we still haven't fully figured out.

It is quite amazing to think that the great majority of Americans consume alcohol on a regular basis, and yet we still don't really understand what it's doing to our heart, even as we're drinking it.

Mandrola: Greg, congratulations on finishing your experiment and your excellent presentation. I'm really grateful for you being with us.

Marcus: Thank you so much, John. It's been my pleasure.

Follow | Medscape Cardiology on Twitter

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: