About
Who writes this
One person's reflux log, analyzed carefully and published in full. That includes the analyses that fell apart on closer inspection.
01 — The short version
Nayts is a pen name. I write under one because I'd rather publish my own medical data than attach my name to it permanently, and those turned out to be the only two options. The name stays consistent. Every analysis here is written by the same person, and always will be.
I am not a doctor, a dietitian, or a gastroenterologist. Nothing here is medical advice, and I have no training that would qualify me to give it.
An engineer who works with data professionally. The statistics on this site (logistic regression, multiple-comparison corrections, confounder checks) are methods I use in my day job, applied to a log I kept myself. That is the entire basis for taking anything here seriously.
02 — What this project is
Since early 2026 I've logged every day: meals, symptom events, sleep, stress, and how much I moved. As of the most recent analysis that is 179 days, 564 meals, and 95 symptom events. Every day is recorded, including the uneventful ones. A symptom diary that only gets filled in on bad days can't tell you anything about the good ones.
Then I analyze it and write up what I find. The methodology page documents exactly what gets logged and the judgment calls involved; the template is the same sheet I use, free, if you want to run this on yourself.
03 — What this project is not
It is an n-of-1 study: sample size one. My results describe my body. They are not evidence about yours, and a correlation in my log is not proof of cause even for me. Where I cite published research, I'm using it as context for what I found, not as a claim that the literature and my log agree.
If you have reflux symptoms, see a doctor. Persistent heartburn can indicate conditions this site is entirely unequipped to detect.
04 — How I try not to fool myself
Self-tracking is unusually easy to get wrong, because you already believe something before you start counting. These are the rules I hold myself to:
- Null results get published. My first analysis went looking for trigger foods and found nothing. That was the finding, so that was the post.
- So do my own mistakes. The strongest food effect I ever measured turned out to be confounded, and I said so at length rather than quietly dropping it.
- Numbers come with their denominators. Effect sizes are reported with the sample they came from and the corrections applied, so you can judge whether they're worth anything.
- Claims are separated from evidence. Where a statement rests on published research it's cited; where it rests only on my log, it says so.
- Corrections are made in public. If an analysis turns out to be wrong, the page gets updated and dated, not deleted.
05 — Corrections and questions
If you spot a statistical error, a misread study, or a claim that outruns its evidence, I want to know. That's the failure mode this project is most exposed to. Write to support@gerdanalytics.com.