Atrial fibrillation Healthy Living

Alcohol and the risk of atrial fibrillation — it’s different than you may think

Life overflows with choices. And consequences. You can choose to smoke and then not be surprised to die a difficult death from cancer. You can choose to eat more calories than you burn and then buy bigger belts. You can choose to “need” more stuff and then, surely, remain unfulfilled.

The more we learn about atrial fibrillation the more it looks like a consequence of our choices. The notion that AF is (mostly) not a mysterious fluke of nature but a result of simple lifestyle choices represents a big shift in thinking, which, like most things in medicine, will take some time to become mainstream. It’s very early in the process of understanding that AF may be unnecessary.

One of the choices we make in our daily life is whether to drink alcohol. And if we decide to drink, we also decide the dosage. These choices, too, come with consequences. And one of them may be to disrupt the heart’s rhythm.

It has long been known that excessive alcohol intake can lead to the “holiday heart,” a disorder in which AF occurs and then passes. Fast (intuitive) thinking holds that it was just too much alcohol at one time, and, no worries, just don’t do that dose again and all will be well.

New studies, however, tell a different story. In recent years, researchers who study populations have found linear (not u-shaped) relationships of alcohol intake and AF incidence. What this means is that one drink daily increases the risk of AF a little. Two drinks daily increases AF risk a little more. And so on. This is different than the old idea that mild-to-moderate alcohol intake is protective (u-shaped curve).

A new study published this week in the Journal of the American College of Cardiology strengthens the idea that alcohol intake increases the risk of having AF in a dose-dependent way. Although the study was observational, the Swedish researchers provided compelling evidence that alcohol intake is indeed an important lifestyle choice, one that comes with potential consequences for the heart’s electrical system. (Recall that the heart beats 100,000 times daily. And that once AF starts, it sometimes doesn’t just stop so easily.)

In my column over at Trials and Fibrillations on, I summarized this study, and then expanded on four thoughts that came to mind when thinking about alcohol and AF risk. The areas I wrote about included the biology, statistics, ethics and philosophy of alcohol use.

If you want to read more, the title and link to the post is here: Alcohol and AF: More Data, More Questions, and Some Philosophy.


(Medscape requires e-mail registration, but it is free.)

11 replies on “Alcohol and the risk of atrial fibrillation — it’s different than you may think”

Those nice California reds from 2010. One glass of a new varietal to complement a fine dinner. Modest joy in small doses – and it’s good for you.

Wait! It’s not!

This is like fat now. And caffeine. Researchers declare that it’s bad for you. No, wrong, it’s good for you. Stop. It’s all bad! …Good??

How does one eat well?

Statins, beta blockers. Yes or no?

What are we to make of all this? Evidence based medicine is only as good as the quality of the evidence.

We don’t REALLY know yet? After all this research and TIME? When will everything be turned upside down again?

Since less than 1% of the population has afib,(and most afib has an identifiable cause, including age) and about 33% of the population has heart disease (for which 2 drinks or less a day is touted to be protective by the medical community), this seems to be a “no brainer”. In absolute numbers, drinking alcohol in moderation has a more beneficial effect than risk for most people. At least it did a few months ago. How many people (in absolute numbers) really get afib because of moderate alcohol consumption? How many have lowered their ASCVD risk?

Compare that to 1) anticoagulants (with side effects) for those afibbers with CHA2DS2Vasc scores of one and 2) statins for men over 65 with no ASCVD risk factors other than age who are being put on statins (with side effects) because their 10 year risk is over 7.5%, and it looks pretty good as a drug when taken as prescribed.

I don’t drink. But maybe alcohol should be covered under Obamacare. Beer for those with afib risk. Wine or hard liquor for those without.

I guess we don’t know enough about afib yet. It’s not like it is the most common arrhythmia or anything. Oh, wait a minute, it is? Looks like the medical community is caught with its scrub pants down again.

I have lone AFIB that is pretty “active” (lots of episodes every week/month, each lasting a day or two). I’m a moderate alcohol drinker. No other real associated conditions or AFIB risk factors. I tried cutting out alcohol completely and there was no change in the amount or pattern of my AFIB. But I only tried this experiment for about a week or two.

For some, alcohol is a definite trigger of more than just a short lived burst of afib.

Everyone is different as to what is a “safe” level of consumption. Long term consumption of alcohol can cause modification, or remodeling of the heart tissue and electrical system. Quitting for a short time to see if afib goes away does not give the heart time to remodel and give a true answer.

Just my 1 cents worth. If I were a medical pro, it would be 2 cents, but for now a penny for my thoughts is all I can ask for.

The linked study’s own data set reports a not even statistically significant 7% increase in the relative risk of developing AF at a consumption level of 7-14 drinks per week, including people who binged on weekends and teetotaled on weekdays. (A meta-analysis of other data sets claims a 17% increase in relative risk at that level; of course, we all know that some participants will have underestimated their consumption.) There was a just barely significant 14% increase in relative risk for those who drank up to 21 drinks a week, which most of us would consider a bit much, and a significant 39% increase above that level. For moderate drinkers, this is hardly a risk worth worrying about. Since the absolute incidence in the entire population was less than 1% per year, if the 7% relative increase in risk for moderate (<= 14 drinks/week) drinkers had been statistically supported, it would still mean that well under 1 per thousand of them per year had been affected. If that were the rate of a significant but nonlethal side effect from a popular drug or procedure, doctors would tell patients not to worry their little heads about it.

Now, one could come back and argue that those who were spared heart attacks as a result of moderate drinking probably outweighed those who may have experienced AF, that the longer life expectancies seen in so many studies surely outweigh tiny increases in AF incidence, that correlations with reduced dementia, a more common and more fearsome condition, might well outweigh correlations with increased AF, etc. But all such arguments implicitly accept the healthist assumption that what is most important in life is to extend one's existence – except by drinking, natch – or, alternatively, to preserve at all costs the function of whatever body part the guy you're arguing with happens to specialize in. It would imply that the best, if not the only adequate, reason to have a glass of wine is that you can marshal evidence that the statistical benefits to physical health outweigh any possible statistical health risks. Social benefits, emotional benefits, sensory pleasure – all of those should take a back seat to the obsession with any perceived threat to physical health.

But this is of course not how anyone, except for a few rather sad and pathetic people who spend way too much time in doctors' offices and scanners, looks at most aspects of life. You enjoy extreme cycling. It's surely more likely to cause you physical harm than physical benefit; you've written about suffering a badly broken wrist and a concussion from bicycle crashes. Should you feel obliged to quit? No, because there's more to life than keeping your body as pristine as possible so you'll last a good long time in the nursing home.

Anecdotally, I’ve seen alcohol trigger episodes of AF in my husband, who’s soon to have his 3rd ablation, this time for atrial flutter. Separately, a recent study suggests that any alcohol consumption can have deleterious cardiac consequences, even in those with no issues. So much for the hallowed notion that a daily glass or two of wine is beneficial (granted, the study’s methodology was a bit dodgy).

Otherwise, for evidence-based dietary and lifestyle advice, there’s always the Ornish program.

Jeff: I don’t think much of that study you linked. It was based on the fact that certain people have gene variant ADH1B that makes them not want to drink too much. They found that those people have slightly less heart disease. It could be that the gene lowers heart disease risk also. It could be that less/no alcohol cuts heart disease risk for only those carrying that gene. The study didn’t even look at alcohol consumption. I wish I had a job doing medical studies… or maybe an ancient alien theorist. Seems like there is a lot of latitude in both fields.

It’s amazing to me how much press that very fishy study is getting and how uncritically it is embraced by those who are inclined to favor its claims. One of the commenters on the website noted the catastrophic statistical error underlying the study’s main claim by saying that just because overall people with the fast Adh genotype drink less than others, you cannot assume that within subgroups that were selected specifically because they reported certain levels of drinking, those with the fast genotype drank less within each subgroup, including the lightest-drinking. No effort was made to show that.

If you scrounge carefully through the full paper, indeed people with the fast Adh genotype had less heart disease. However, comparable to every other epidemiological study, the incidence rate was 13.67% for teetotalers, 9.92% for light drinkers and 9.54% for moderate drinkers – and light-drinking people with the fast genotype had no extra reduction compared to moderate drinkers. How anyone can report these data then claim with a straight face that they have just proven that no matter how little you drink, Less is always better – all the way down to zero – is truly beyond me. All the yap about Mendelian randomization, which this study really did not do, is designed to make us ignore those actual data in favor of the unproven claims that (1) heart disease was lower in fast-Adh moderate drinkers than slow-Adh moderate drinkers because the former just MUST have drunk less, but that (2) fast-Adh light drinkers didn’t have an even lower incidence, and fast-Adh teetotalers had a much higher incidence, because the benefit of drinking ever-less was counterbalanced by the lack of the secret protective effect that is strongly correlated with regular drinking yet has nothing to do with the beverage.

(What might that effect be? Surely someone should offer a plausible hypothesis and test it before decreeing that it must exist. If it does exist, since it has not been identified yet is so commonplace and so powerful as to make drinkers live significantly longer and healthier lives than nondrinkers despite the harmfulness of every drop of wine they consume, the rational thing to do is to enjoy whatever beverages you like in the reasonable assumption that you are likely to encounter and benefit from the secret correlated protective effect.)

I’d enjoy your ancient aliens research, Pete. At least it would be a break from the constantly conflicting advice about how to live today.

Let me know when you publish. I’m on board for at least one copy.

Comments are closed.