Coffee and atrial fibrillation: what DECAF found
4 min read
The debate over coffee and atrial fibrillation now has a randomized trial behind it, but the finding is narrower than many headlines suggest. DECAF didn't test whether coffee protects everyone. It studied people who already drank coffee and had returned to sinus rhythm after successful cardioversion.
Short answer: in that specific group, participants assigned to drink at least one cup of caffeinated coffee daily had fewer clinically detected recurrences than those assigned to abstain. The result doesn't establish prevention for the public, a treatment for AF, or universal safety.
The DECAF result and the 39% figure
The DECAF randomized clinical trial in JAMA enrolled 200 adults after successful cardioversion, with 100 in each group. During six months, atrial fibrillation or flutter recurred in 47 of 100 people assigned to coffee and 64 of 100 assigned to abstinence.
That's an observed difference of 17 percentage points. The time-to-first-recurrence analysis produced a hazard ratio of 0.61, with a 95% confidence interval of 0.42 to 0.89, according to the same JAMA report. The quoted 39% describes an estimated lower relative hazard during follow-up. It isn't a 39-point gap or 39 people “protected.”
Fussy wording? Maybe. But it changes what the number says.
Who was actually studied?
Participants had persistent atrial fibrillation, or atrial flutter with a history of AF. They had also consumed at least one cup of coffee a day at some point in the previous five years. Randomization happened only after cardioversion succeeded and sinus rhythm was maintained, as the trial report explains.
The coffee group was encouraged to drink at least one caffeinated cup daily. The abstinence group was asked to avoid caffeinated coffee, decaffeinated coffee, and other caffeine-containing products. The trial therefore can't cleanly separate caffeine from other compounds in coffee.
Ordinary habits got in the way too: only 69% of the abstinence group reported no coffee during follow-up, according to JAMA.
What DECAF doesn't show
DECAF doesn't address a first AF episode in someone with no history of the condition. It didn't test people who never drink coffee, several daily espressos, energy drinks, or decaf alone. Coffee wasn't tested as a treatment.
The trial was open-label. Recurrence meant a clinically detected episode lasting at least 30 seconds, without standardized continuous monitoring for everyone, according to JAMA. Two hundred people followed for six months can't rule out rare harms. An American College of Cardiology discussion says patients with AF may not need blanket avoidance of caffeinated coffee. That's expert commentary, not individual medical advice.
Palpitations add another wrinkle. Feeling a racing or pounding heart after coffee isn't the same as documenting AF. In I-STOP-AFib, reported in JAMA Cardiology, caffeine exposure didn't clearly raise self-reported AF episodes on average. That doesn't make a person's symptoms imaginary, and an article can't diagnose their cause.
Why the paper has two dates
The study was published online on November 9, 2025, then appeared in the JAMA issue dated January 27, 2026. It isn't a newly published October 2026 study.
DECAF challenges automatic coffee avoidance after cardioversion. It doesn't turn a cup into a prescription. Any change in coffee intake, especially alongside symptoms or medication, belongs in a conversation with the clinician who knows the case.
Frequently Asked Questions
Does coffee treat atrial fibrillation?
No. DECAF compared two instructions after cardioversion and measured recurrence. It didn't test coffee as medicine or replace treatment and clinical follow-up.
Who does the DECAF result apply to?
Current or recent coffee drinkers with persistent AF, or flutter plus a history of AF, randomized after successful cardioversion. The general public wasn't studied.
Why is the relative figure different from the observed gap?
The first comes from a time-to-event analysis. The second simply compares cumulative proportions in the two groups. They describe the same trial from different angles.
Should someone stop drinking coffee after palpitations?
DECAF doesn't answer that individual question. New, severe, or persistent symptoms need clinical assessment rather than a decision based on this article.
