PouchWise

PouchWise / How it works

How PouchWise scores food

Every score is a modelled estimate — our best, transparent guess at how gentle or harsh a food is likely to be on a sensitive gut. It is not a clinical measurement, and it is not medical advice.

We're independent. We're not doctors making clinical claims, and we don't pretend the science is more precise than it is. These numbers are our opinion of how foods may affect gut comfort, built on published research plus reasoned assumptions where the evidence runs out. Use them as a starting point for your own experimentation — not as a verdict.

Why a score, not a single “good / bad” label

The diet research for sensitive guts — J-pouches, ostomies, IBD, IBS — is repeatedly described as limited, patient-specific and partly empirical. There is no validated table that says "10 g of onion produces X cc of gas" for a specific person. What the evidence does support is that the same food can behave very differently depending on which symptom you care about: a food can be low in gas but high in urgency, or low in urgency but high in coarse residue.

A single "avoid this" label hides exactly the distinction that matters most. So PouchWise scores every food on five separate axes instead of one.

The five axes

GGasrisk — higher is worse

Fermentation, bloating and wind (and odour) risk.

Driven by: FODMAP load, rapidly fermentable fibre, sulphur/odour burden, concentration.

AAgitationrisk — higher is worse

Irritation, burning and stimulant-driven urgency.

Driven by: Capsaicin, acidity, caffeine, alcohol, high-fat stimulation.

SStool-looseningrisk — higher is worse

Faster, looser, more urgent output.

Driven by: Osmotic sugar/polyol load, free-water delivery, fat/bile effect, rapid transit.

PParticle loadrisk — higher is worse

Coarse, undigested residue — and, with a narrow outlet, blockage risk.

Driven by: Rawness, peels, seeds, fibrous structure, insoluble roughage.

BBindingprotective — higher is better

Protective — the tendency to thicken and calm output.

Driven by: Binding starch (rice, potato, banana), gel-forming fibre (psyllium, oats).

G, A, S and P are risk scores — higher means harsher. B (Binding) is protective — it works the other way and subtracts from the overall score. Each axis is scored 0–100 inside the engine and shown to you on a friendlier 0.0–10.0 scale.

How a food becomes a score

  1. 1. Underlying factors

    Each food is described by a set of normalised inputs — things like fermentable-carbohydrate load, capsaicin heat, binding starch, gel-forming fibre, frying-fat load. These are estimated from the food's USDA nutrient profile plus Monash-style FODMAP category knowledge, and each carries a confidence grade.

  2. 2. Preparation changes the score

    The same food in a different form is scored separately. Boiling and draining leaches water-soluble fructans out; powdering concentrates them; garlic-infused oil carries the flavour without most of the FODMAP burden. We treat preparation as a score adjustment — never as a blanket "cooked = safe" claim, because the research is clear that cooking-based reduction is real but variable.

  3. 3. Portion, against a reference serving

    Risk is scaled against a sensible reference serving (vegetables ~75 g, cooked starches and proteins ~100 g, fresh aromatics ~15 g, powdered spices a teaspoon). This is what stops onion powder from looking harmless just because nobody eats 100 g of it — and why a sauce can score low yet still warrant "don't pour it on".

  4. 4. The roll-up formula

    The five axes combine into one overall “gut load”, weighting gas and loose/urgent output most heavily, with binding pulling the number back down:

    Overall = 0.35·Gas + 0.20·Agitation + 0.30·Stool-loosening
          + 0.15·Particle − 0.25·Binding

    For meals, small synergy rules nudge the total: a bump when two high-FODMAP foods or spicy-plus-acidic foods land together, and a discount when a strong binder like rice, potato or banana is on the plate. Every weight and rule is an editable product heuristic, not a clinical constant.

We grade our own confidence

Because a lot of this is informed estimation, every food carries a confidence grade so the model is honest about what it actually knows:

  • A — backed by direct measured composition or explicit clinical guidance.
  • B — strong mechanism plus food-composition inference.
  • C — expert-curated inference.
  • D — anecdotal or pattern-based.

The first large pass of the database was drafted with AI assistance grounded in the nutrient data and evidence summaries, then capped at C/D until a human reviews it. Those rows are starting points to be checked, not final answers.

Why our numbers may differ from other tools

A low-FODMAP app, a dietitian's handout and PouchWise can all rate the same food differently — and that's expected. Most FODMAP tools answer "how fermentable is this carbohydrate?" PouchWise tries to answer a broader, messier question: "how is a normal serving of this likely to affect gut comfort and output overall?" That folds in fat, capsaicin, caffeine, coarse residue and binding — not just FODMAPs.

Where we've had to make a judgement call, we've made one, and we'll tell you it's a judgement call rather than dress it up as settled science. Sometimes the honest answer is simply: this is how we decided to weight it. You're free to disagree and use the tool however suits you.

The research we draw on

The model is built on FODMAP food-composition work, J-pouch and ostomy patient guidance, IBD and IBS dietary research, gut-gas physiology, and baseline nutrient data — plus primary studies on FODMAP restriction, capsaicin, coffee and the odour of wind. The full write-up lives in our internal research notes; the main public sources are below.

Links go to the source organisations. We're not affiliated with any of them, and a link is not their endorsement of PouchWise.

What PouchWise is not

  • • Not a gas-volume meter. Real gas output varies with your microbiome, the rest of the meal, and how adapted your gut is.
  • • Not personalised medical advice — even after you personalise the scores.
  • • Not a claim that any "low" food is safe for you, or any "high" food must be avoided. It's a transparent, evidence-informed place to start.
  • • Not a substitute for getting worsening symptoms checked. Pouchitis, obstruction, dehydration, bile-acid issues and flares are medical problems, not diet problems.
Eating green isn't the whole diet →Who it's for & how to tailor it →FAQ →