N=1: Bite the Bullet

slimemoldtimemold, ,

Previously in this series:
N=1: Introduction
N=1: Single-Subject Research
N=1: Hidden Variables and Superstition
N=1: Why the Gender Gap in Chronic Illness? 
N=1: Symptom vs. Syndrome
N=1: Latency and Half-Life
N=1: n of Small
N=1: Dr. Garcia’s Queasy Irradiated Rats

When it comes to chronic illnesses, most people try to find ways to avoid the pain. This is because pain bad, and no pain, good. 

But we worry that avoiding pain is good in the short term but bad in the long term; penny wise and pound foolish. 

If you’re worried that pizza makes you bloated, it’s good common sense to try to avoid pizza. But it’s bad science. Past a certain point, avoiding pizza tells you nothing. To learn something more, you have to bite the bullet.

Better to wait for a day when you feel great, no bloating at all, and then intentionally go and eat pizza, and see what happens. Or each afternoon you feel good, flip a coin, and eat pizza when it comes up heads. Do this a couple of times. If you do this, you should be able to see if pizza is really a reliable trigger for your bloating.

There are two reasons to do this. The first is — what is it about pizza that makes you bloated? If you can show that pizza is a trigger, you can start doing empirical splits. Buy the same pizza and flip a coin. If it’s heads, scrape off the cheese and tomato and just eat the bread. If it’s tails, toss the bread and just eat the cheese and tomato scraped off into a bowl. Which makes you bloated? If it’s the cheese and tomato, separate them and do the same thing.

The second is that a lot of the time, we suspect you’ll find that pizza (or your personal equivalent) is not a trigger at all. It’s all too easy to think you see a pattern where there’s none to be found, and we tend to see food triggers even when food doesn’t matter at all. You could have been enjoying pizza this whole time. That seems worth knowing.

5 thoughts on “N=1: Bite the Bullet”

  1. Very good call! Staying too high level (“pizza makes me bloated”) also prevents you from finding the real cause.
    E.g. many people say “stay away from dairy” or “stay away from fat” or “stay away from processed food.”
    But most people have a completely different reaction to cheese vs. dairy, butter vs. soybean oil, or deep-fried doritos vs. canned sardines/air dried beef jerky.

    Like

    Reply

  2. Unfortunately, maybe it’s the pills you’re taking, or the nasal spray you’re huffing, or the formaldehyde leaking out of your modern house, or the pollution in the air that’s destroying your immune/digestive system. Or maybe it’s just that your pizza was made by a guy who didn’t wash his hands that day. It’s impossible to know.

    If the “universal diet” could have been found, it already would have been found. But real people and real bodies and real environments don’t work that way. All anyone can say is “sometimes, this works, for me.”

    Like

    Reply
  3. Jameverywhere says:

    I have celiac disease.

    Forums for people with celiac disease can be pretty wild in this way. No matter what the foodstuff is, someone in the forum will claim it’s cross-contaminated at because it made them sick, and someone else will say they eat it all the time with no symptoms.

    There’s also always people with large lists of foods they cannot eat… I mean, like, horrifically restricted diets. And when you’re still feeling bad yourself, it’s easy to think that maybe the only answer is to cut more things out too.

    Diet can absolutely make a huge health difference, too, which makes it even harder to navigate. I know my lingering symptoms are related to gut bacteria somehow. But trying to fix dysbiosis is still shooting arrows in the dark, unfortunately.

    Like

    Reply
  4. Robin Green says:

    Can anyone recommend an app that helps in some way with running N=1 studies on yourself?

    I’m dreaming of something that will let me put in 10 variables, randomise them all for me every day (or whatever time period I choose), let me enter the data, allow me to flake out and not comply with the study design some days, and finally use an AI-driven statistics whizz to analyse the results.

    Don’t know if such a thing exists. If not, maybe I should make it?

    Like

    Reply

Leave a comment