The method · Something concrete, starting tonight
An investigation into you, run with rigor.
A photo of your plate, a word about how you feel. Digestio crosses the two, exposure window after exposure window, and learns what agrees with you, what doesn’t sit well, and at what dose. The same rigor sheds light on your metabolism, your macros and your weight. And when nothing is clear, Digestio says nothing.
Method inspired by published trials · references at the bottom of the page
Exposure windows
POSTERIOR PROBABILITY · D+12
Milk → bloating · 0.93
“Probable” · 5 exposed windows, 4 control windows
The principle
Your days with, your days without.
Universal diets often fail because your digestion isn’t an average. Digestio compares your days with a food against your days without, and measures what changes. Medicine has used this logic since the 1980s for personalized decisions — Digestio applies it to your plate, with no treatment aim whatsoever.
- Your days with the food are compared with your days without — never you with an average.
- Each food and each family (gluten, lactose, FODMAP, histamine…) is tracked separately.
- Nothing is imposed: Digestio suggests, you decide what to test.
Why not an average?
Two randomized trials on FODMAP elimination show a real effect on irritable bowel syndrome — but also that the response varies widely from one person to another. A population statistic tells you what works “on average”; it doesn’t tell you what works for you. That’s exactly the gap Digestio fills: a reading built on your meals, not on other people’s.
Ref. Halmos 2014 · Böhn 2015 · Staudacher 2017 — see References.
The math
A probability that updates, not a verdict.
Rather than a binary “yes / no,” Digestio reasons in posterior probability (Bayes): the chance that a meal → symptom link is real, revised with each new data point. The same logic is used, in research, to combine trials run on a single person.
01
We start from a cautious prior
At the outset, Digestio assumes no link exists. No food is “guilty” by default.
02
Each day updates the probability
With each exposed then control window, Bayes' theorem re-estimates the probability that the link is real — upward if it's confirmed, downward otherwise.
03
The discovery only shows if it holds
It only appears once the posterior probability has become credible. An isolated coincidence is never enough.
How long?
~8 windows
of exposure are usually enough to make a discovery clear — about 14 days of regular tracking.
An “exposure window” = the period following a meal containing the tracked food. You need enough windows with and without for a difference not to be down to chance.
Exposure windows
Why it takes a little patience.
A single episode proves nothing: maybe you ate something else, slept badly, or simply had a bad day. It’s the repetition of the pattern — discomfort when the food is present, calm when it isn’t — that pushes the probability up.
The more often the food recurs in your diet, the faster the windows pile up and the sooner the discovery arrives. The rarer it is, the longer it takes — Digestio owns that and asserts nothing until it’s solid.
The confidence scale
Plausible, probable, clear. And sometimes, nothing.
The posterior probability reads on a simple scale. When nothing is clear, Digestio says nothing — no false discovery. It’s the same scale shown on the homepage.
A “probable” or “clear” level is an invitation to test a food and observe — never an instruction.
Posterior probability
Between 0.60 and 0.75, the link is still too uncertain to present: Digestio keeps observing in silence.
Metabolism & weight
The same observation principle, applied to your weight.
This reading isn’t only for your gut. It also sheds light on a question many people ask: why isn’t my weight following ? Because metabolism adapts and not all foods are equal, Digestio observes your calories, your macros and your tolerances — to understand, not to promise.
A metabolism that adapts
Your energy expenditure isn't a fixed “calories in − calories out” equation. It adjusts to what you eat and to your weight history — after major weight loss, resting expenditure stays durably lowered (research by K. Hall, NIH). Digestio reads your own responses rather than applying an average formula.
Macro breakdown, not just the total
At equal presented calories, an ultra-processed diet leads to spontaneously eating ~500 kcal/day more (NIH randomized trial). The share of protein, carbs and fat — and food quality — weighs as much as the day's number. Digestio shows each day's breakdown, captured from a photo or a simple voice note.
Your intolerances and your weight
Wondering whether your intolerances are keeping you from losing weight? Bloating, water retention and discomfort can blur what the scale says. Digestio doesn't decide for you: it helps you see clearly between what doesn't sit well, how you feel and your weight.
Ref. Hall 2019 (ultra-processed) · Fothergill & Hall 2016 (metabolic adaptation) — see References. Digestio promises no weight loss.
The limits, said plainly
What the method does not do.
It's not a diagnosis
Digestio is not a medical device. It identifies no illness, allergy or intolerance.
Your data doesn't leave
Compiled for you, never sold, never used to train third-party models.
It's not a health promise
No guaranteed result, on symptoms or on weight. For any health question, talk to a professional.
Method questions
The substance, no jargon.
Why compare my days with each other, rather than with an average?
Because your body isn't an average. Digestio compares your days “exposed” to a food with your days “not exposed,” and measures what changes. Medicine has used this logic since the 1980s for personalized decisions — adapted here to the plate, with no therapeutic aim.
What is a posterior probability?
It's the probability that a meal → symptom link is real, updated with each new day of data (Bayes' theorem). The more consistent windows you accumulate, the sharper that probability gets — upward if the link holds, downward if it fades.
Why about 14 days before a first discovery?
A clear discovery usually needs ~8 exposure windows: enough days with the food and enough days without, so the difference isn't down to chance. With a varied diet, that's roughly two weeks of regular tracking.
What about metabolism and weight?
The same observation logic applies: Digestio links your calories, your macro breakdown and your tolerances to how your weight changes. Because metabolism adapts (research by K. Hall, NIH), it observes your own responses rather than applying an average formula — to understand, never to promise weight loss.
What does the confidence scale mean?
Below 0.60, Digestio shows nothing. Between 0.75 and 0.90 a discovery is “plausible,” between 0.90 and 0.97 “probable,” above 0.97 “clear.” Digestio would rather show nothing than display a false certainty.
Does Digestio make a diagnosis?
No. Digestio is not a medical device and diagnoses no illness. It surfaces statistical correlations on your own data to help you test an elimination. For any health question, talk to a professional.
References
What the method rests on.
Our method draws on the trials known as “N-of-1,” where each person serves as their own reference. The work below — published and peer-reviewed — covers those trials, their Bayesian analysis, elimination diets, and metabolism and weight (research by Kevin Hall, NIH). It illuminates the method; it constitutes neither proof of Digestio’s effectiveness nor medical advice.
Guyatt G, Sackett D, Taylor DW, Chong J, Roberts R, Pugsley S
Determining optimal therapy — randomized trials in individual patients
N Engl J Med · 1986 · 314(14):889-92
Founding paper of the N-of-1 trial: testing a hypothesis on a single individual, against themselves.
Duan N, Kravitz RL, Schmid CH
Single-patient (n-of-1) trials: a pragmatic clinical decision methodology for patient-centered comparative effectiveness research
J Clin Epidemiol · 2013 · 66(8 Suppl):S21-8
Modern methodological framework for N-of-1, designed for personalized decisions.
Zucker DR, Schmid CH, McIntosh MW, D'Agostino RB, Selker HP, Lau J
Combining single patient (N-of-1) trials to estimate population treatment effects and to evaluate individual patient responses to treatment
J Clin Epidemiol · 1997 · 50(4):401-10
Hierarchical Bayesian model to combine individual trials — the logic behind our posterior probabilities.
Halmos EP, Power VA, Shepherd SJ, Gibson PR, Muir JG
A diet low in FODMAPs reduces symptoms of irritable bowel syndrome
Gastroenterology · 2014 · 146(1):67-75.e5
Controlled crossover trial: lowering FODMAPs reduces irritable bowel syndrome symptoms.
Böhn L, Störsrud S, Liljebo T, Collin L, Lindfors P, Törnblom H, Simrén M
Diet low in FODMAPs reduces symptoms of irritable bowel syndrome as well as traditional dietary advice: a randomized controlled trial
Gastroenterology · 2015 · 149(6):1399-1407.e2
Randomized trial: targeted elimination works, but the effect varies from one person to another.
Staudacher HM, Lomer MCE, Farquharson FM, Louis P, Fava F, Franciosi E, Scholz M, Tuohy KM, Lindsay JO, Irving PM, Whelan K
A Diet Low in FODMAPs Reduces Symptoms in Patients With Irritable Bowel Syndrome and A Probiotic Restores Bifidobacterium Species: A Randomized Controlled Trial
Gastroenterology · 2017 · 153(4):936-947
Randomized controlled trial confirming the value — and the limits — of guided elimination.
Hall KD, Ayuketah A, Brychta R, Cai H, Cassimatis T, Chen KY, Chung ST, Costa E, Courville A, Darcey V, et al.
Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake
Cell Metab · 2019 · 30(1):67-77.e3
Inpatient randomized trial (NIH): at equal presented calories and macros, an ultra-processed diet leads to spontaneously eating ~500 kcal/day more. Food quality matters, not just total calories.
Fothergill E, Guo J, Howard L, Kerns JC, Knuth ND, Brychta R, Chen KY, Skarulis MC, Walter M, Walter PJ, Hall KD
Persistent metabolic adaptation 6 years after “The Biggest Loser” competition
Obesity (Silver Spring) · 2016 · 24(8):1612-1619
Metabolic adaptation: after major weight loss, resting energy expenditure stays durably lowered. Metabolism is not a fixed equation — it adjusts to your history.
The method is built for your data. Give it something to observe.
Five minutes to get started, about two weeks of regular tracking for a first clear discovery.