# Afib Cases

> One person's atrial fibrillation, measured and written down. Five years of Apple Watch
> data, 112 ECGs, 169 irregular-rhythm alerts and 1,316 nights, with the dates and the
> methods attached. A single case, not evidence about any population, and not medical advice.

The case: male, 55, endurance swimmer, occasional drinker, not overweight, paroxysmal
atrial fibrillation diagnosed September 2026. Written by the patient, who is not a doctor.

Every page states where its numbers come from, carries its findings as a plain-text block
meant to be copied whole, and carries the analysis as an instruction that can be run against
another person's own data.

## Cases

- [Fifty-four without a verdict](https://afibcases.com/fifty-four-without-a-verdict): of 112
  Apple Watch ECGs, 54 returned no verdict; 21 showed atrial fibrillation never reported to
  anyone.
- [Dinner predicts the night](https://afibcases.com/dinner-predicts-the-night): across 1,285
  nights, the hour of the last meal moved the overnight heart rate more than body weight or
  medication.
- [Same month, four years apart](https://afibcases.com/same-month-four-years-apart): an
  attempt to separate a heart-rate-slowing drug from the body it was given to, using a single
  subject, and why it cannot be done.
- [One record, two watches](https://afibcases.com/one-record-two-watches): Apple's "up to 24
  times more often" claim for heart-rate-variability sampling, measured on one person's own
  record; the five-fold jump came from a software build, not the new hardware.

## Method

- [Reading your own heart data](https://afibcases.com/reading-your-own-heart-data): how every
  number on this site was extracted from an encrypted iPhone backup and counted.