A few years ago I had several runs of paroxysmal atrial fibrillation. After I eventually admitted it could be a problem and a trip to our local cardiology team, I figured that alcohol use could be the trigger and going tee-total was infinitely preferrable to ablation (much to the surprise and disappointment of the cardiologist).
However, after a year of zero alcohol I had a further episode, and this made me realise there was another potential trigger: caffeine.
So I went caffeine-free more than a decade ago, and despite a beer or glass of wine here and there, no further AF. Now when I see patients with atrial fibrillation as well as alcohol consumption I always ask about their caffeine intake and if it’s high, suggest they cut it down or ideally out.
But does this actually help or are my own cognitive biases depriving my patients of their usual morning pick me up?
Caffeine is commonly reported as a trigger for atrial fibrillation by patients – one questionnaire-based study published in HeathRhythm journal in 2019 found that 28% of 1295 participants with symptomatic AF identified caffeine as a trigger (not far behind alcohol at 74%).
But caffeine as a pro-arrhythmic has been disputed in the medical community for several decades – a concept born in the 70s and 80s, more recent data has contradicted this idea.
In 2006 a paper in the Journal of Electrocardiology found that caffeine administered intravenously at various different doses “may result in an unexpected reduction in the propensity for AF in healthy individuals and in those with a predisposition for AF”. However, this study was conducted in dogs, so it’s difficult to know how to compare this against getting a double shot latte at a Costa drive-thru. The authors also failed to report how long a walk the poor canines required after treatment.
More recent studies have, however, painted a reassuring picture. In 2023 a prospective RCT published in the NEJM (the catchily titled Coffee and Real-time Atrial and Ventricular Ectopy or CRAVE study…) showing that consumption of caffeinated coffee did not result in significantly more ‘premature atrial contractions’ than avoidance of coffee in 100 people wearing continuous recording ECG devices.
Which brings us to a paper published just a few weeks ago in JAMA: Caffeinated Coffee Consumption or Abstinence to Reduce Atrial Fibrillation – The DECAF randomized clinical trial. This paper asked the question “does drinking caffeinated coffee have a beneficial, detrimental or neutral effect on the risk of recurrent AF episodes”.
To figure this out the authors conducted a RCT including 200 patients (average age 69, 71% male) with persistent AF undergoing cardioversion and randomised them to either continue regular coffee consumption (at least 1 cup a day) or coffee and caffeine abstinence for 6 months.
Surprisingly, AF or atrial flutter recurrence was less in the coffee group (47%) compared to the abstinence group (64%), a statistically significant difference.
Why it might be beneficial isn’t certain. The authors suggest it could be due to blockade of AF-inducing adenosine receptors, the anti-inflammatory properties of coffee, or the catecholaminergic effects during vagally-driven AF (I once flipped into AF after vomiting due to a migraine, hours before having to present a Hot Topics course…).
What is clear is that for most patients with a diagnosis of atrial fibrillation, avoidance of caffeinated drinks is not required.
One caveat the authors point out is that the average consumption of coffee in the active group was 1 cup a day – we can’t extrapolate a positive dose effect or that higher levels are safe. So Doris can enjoy her cappuccino in the morning, but should probably leave the cans of Red Bull alone. And definitely don’t mix them with vodka. Now that really could bring on AF…

You can quickly add CPD to your account by writing a reflective note about the NB Blog – Atrial Fibrillation – is coffee the culprit? post you've read.
Log in to your NB Dashboard and use the 'Add Reflective Note' button at the bottom of a blog entry to add your note.