Article

How to Find Your Own AFib Triggers (and Why Generic Lists Don’t Work)

Published 5 August 2026 · 9 minute read · by Bryan, developer of My AFib Companion

Search “AFib triggers” and you get the same list every time.

Caffeine. Alcohol. Poor sleep. Stress. Exercise. Dehydration. A large meal. Cold. Heat. Illness.

You have probably read that list three times already, held it up against your own last episode, and found that none of it fit. Or worse — that all of it fit. You’d had coffee, you were stressed, you slept badly, and you still have no idea which one mattered.

The question underneath this is usually not academic. It’s if I don’t know what set this off, I can’t stop it happening again. That is a reasonable thing to want.

So this is about how to go looking properly: what it costs in time, what a real answer looks like, and why the honest ending is often “I still don’t know.”

The list isn’t wrong. It’s just not about you.

That list is a summary of what large numbers of people with AFib have reported. That’s all it is. It’s a vocabulary — a set of things worth writing down — not a set of findings about your heart.

I’ll be straight about this. My own app ships almost exactly that list: eleven checkboxes, and I didn’t invent them either. They’re there because they’re what people most often think to record, and tapping a checkbox is faster than typing at 3am. That’s what they’re good for. They are a starting vocabulary, not a verdict on your case.

“Alcohol is a common trigger” is a sentence about a group. Translated honestly, it means that across a lot of people who were surveyed or studied, alcohol came up often. Whether it comes up for you is a separate question, and no list can answer it, because no list has ever met you.

There’s a second problem with the lists, and it’s quieter.

Things that are easy to name make the list. Things that are hard to name don’t. A second glass of wine is easy to notice and easy to write down. A restless night four days ago, a virus you never quite got, a dose taken two hours later than usual, a room that was too warm — those are hard to notice and harder still to connect to anything. Published lists lean toward the memorable. The thing that matters most for you might simply never have made anybody’s list.

Why your memory is the wrong tool for this

Ask most people what set off their last episode and they’ll have an answer in five seconds.

That answer is usually unreliable, and not because they’re careless.

Memory keeps the striking thing. You will remember the night you had two glasses of wine and went into AFib at midnight. You will not remember — because there is nothing to remember — the eleven other nights you had two glasses of wine and slept straight through. One of those is a story. The other eleven are not anything.

So when you search your memory for a pattern, you are searching a filing cabinet where only the bad days got filed.

Here is the same point as plain arithmetic, because this is the idea the whole article rests on.

Say you had six episodes in three months, and you’d had coffee before five of them. Coffee looks guilty. Five out of six.

Now add the number that’s missing. You drink coffee every morning, so that’s about ninety coffee days in those three months, and five of them had an episode. Your episode rate on coffee days is roughly one in eighteen. Your episode rate on non-coffee days is — nothing. You don’t have any non-coffee days. The comparison can’t be made.

Coffee isn’t guilty. Coffee is just always there.

You cannot find a pattern without the boring days. The days nothing happened are not padding around the real data. They are the comparison. Without them, anything you do daily looks suspicious and nothing can ever be cleared.

In practice that means one habit: write a line on the good days too. Thirty seconds. Slept fine, two coffees, no alcohol, ordinary day, nothing happened. It feels pointless as you write it. It is the entry that does the work three months later.

How to actually test one thing

Say you want to look at something specific. Here’s the method, starting with the least satisfying rule.

Don’t change anything yet. This is observation, not an experiment on yourself. Cutting something out, adding something in, or shifting how you take a medication is a conversation with your cardiologist first — including the ones that sound harmless. Begin by watching what you already do.

Pick one thing. One. Whatever you’re already suspicious of. Testing three at once produces a result you can’t read, because the three overlap.

Write your definition down before you start. What counts as “poor sleep”? Under six hours? Waking more than twice? Pick one, and put it at the top of the page. If you decide what counts after you’ve seen the data, you will pick the definition that fits what you already believe. Everybody does this, and writing it down first is the only real defence against it.

Decide the time window in advance as well. Are you looking for episodes the same evening? The next day? Within 48 hours? A trigger with no time limit will always find a match somewhere.

Give it enough weeks. Four to six is a floor, not a target. And the real constraint isn’t the calendar — it’s how often you have episodes. If you have one a month, six weeks of watching contains one or two of them. That’s an anecdote with a start date, not a test. Some people would need six months to say anything at all. That’s slow, and I’d rather say so now than have you quit in week three thinking you’d done it wrong.

Then compare two rates, not two stories. Episodes per day with the thing, against episodes per day without it. Both numbers, or neither.

Even a real pattern might not mean what you think

Say it comes out clean: episodes are more likely on the nights you slept badly. Three different things could be true, and your log cannot tell them apart.

It could run backwards. Your heart may have been doing something before you consciously noticed, and that’s what made the night restless. The bad sleep would be a symptom, not a cause.

Something else could be driving both. A hard week means less sleep, more coffee, more alcohol, meals at odd hours and a missed dose. Those move together. Pull on one and the whole knot comes with it.

It could be chance. Check ten candidate triggers against a handful of episodes and one of them will look meaningful. That’s arithmetic, not insight.

None of that makes the exercise pointless. It makes the result a question you bring to your cardiologist rather than a conclusion you act on by yourself. “Eight of my last ten episodes were on nights I slept under five hours — does that mean anything to you?” is one of the more useful sentences available to you in a short appointment.

The usual suspects, plus one that belongs

If you want the checklist, this is the standard set:

caffeine · alcohol · poor sleep · stress · exercise · dehydration · large meals · cold · heat · illness · unknown

That last one matters more than the other ten.

Leave yourself a way to record “I don’t know.” If every entry demands a cause, you will supply one — and by the fourth time you’ve written “stress” because nothing else was available, you’ll believe stress is your trigger. You will have manufactured a pattern out of an empty field.

“Unknown” is an honest data point. A log with a lot of them is telling you something real: that your episodes so far aren’t tracking with anything you’ve thought to measure. That’s worth knowing, and it’s worth saying out loud at your next appointment.

When the answer is “no pattern”

This happens often. It may well happen to you.

Plenty of people track carefully for a year and find nothing that holds up. Episodes turn up in good weeks and bad ones, after wine and after water, at 3am and at 3pm.

That is a result. It is not a failure of your record-keeping, and it doesn’t mean you did it wrong.

The log still earned its keep, because trigger-hunting was never the only job it had. You can walk into the appointment and say how many episodes you’ve had since last time, how long they ran, what they felt like, and what’s changed. That’s what the appointment actually needs. The trigger question was the bonus, and sometimes the bonus doesn’t arrive.

Knowing you don’t have an obvious trigger is useful on its own. It’s one fewer thing to police every day, and one fewer reason to blame yourself for a bad week.

About tracking it

You can do all of this on paper. One row a day, a few columns, the boring days included. A pocket notebook or the Notes app is genuinely enough — and if you’ll be more consistent with a pen, use the pen. Consistency beats every other consideration here.

The arithmetic further up is why I built My AFib Companion: a thirty-second check-in on the days nothing happens, and then it compares how often episodes landed on the days you logged a factor — alcohol, caffeine, poor sleep, stress, low mood, low hydration — against the days you didn’t. Same limits as everything above — it’s arithmetic on your own log, not proof of anything, and it needs months of entries before the numbers say much. It’s $9.99/month or $59.99/year after a 30-day free trial, with no free tier after that.

The notebook does the same job for nothing. Start there if you’d rather.

---

---