How Toleah scores.

A traffic-light is only as good as the rule behind it. This page explains transparently how Toleah's scoring engine turns an ingredient list into a personalised number — and where that method has limits.

Scoring logic under scientific review

The engine in numbers

400+
curated triggers
8
intolerances
7
enzyme profiles
3
severity levels

The principle in one sentence

Every scan starts at one hundred points. Deductions come from trigger ingredients, weighted by their position in the recipe, multiplied by your severity, corrected by your enzyme intake. The result is a score between 0 and 100, mapped to one of four traffic-light stages.

How the engine scores — step by step

  • Over 400 trigger definitions

    The engine currently knows over 400 researched trigger definitions across eight intolerances, compiled from specialist guidelines and food-science literature. Each entry carries synonyms, multilingual aliases and a severity grade (high, medium, low). Example histamine: "salami" (high), "aged hard cheese" (high), "yeast" (medium). Trigger lists are versioned and traceable.

  • Three severity levels

    Your profile holds a severity per intolerance: mild (factor 0.5×), moderate (1.0×), severe (1.5×). This multiplier applies to every deduction for the relevant intolerance. So Toleah grades the same salami differently for someone with mild histamine intolerance than for someone with severe — because clinical reality demands it. Gluten/coeliac uses a special rule (Catassi 2007): even "mild" is treated as 1.0× because trace amounts trigger an autoimmune response.

  • Seven enzyme profiles

    When your profile indicates that you take an enzyme supplement for an intolerance, Toleah adjusts your score conservatively upward. Toleah does not recommend taking any enzyme — that decision is between you and your doctor. The correction factor in the scoring model is a computational aid for personalisation, not a dosage recommendation.

  • Position weighting

    Ingredients are weighted by position: the first ingredient gets factor 1.3×, the last 0.7×. So a trigger at the top of the list pulls the score harder than one at the bottom — because ingredient lists are sorted by quantity.

  • Traces & allergens

    For most intolerances, trace declarations are handled moderately. For coeliac disease the engine applies the full penalty — because even minute gluten amounts can trigger an autoimmune reaction.

  • The four traffic-light stages

    Green (80–100) means well tolerated. Yellow (60–79) with consideration. Orange (30–59) with caution. Red (0–29) unsuitable. Each stage combines colour, icon and label — so the traffic-light is unambiguous even without colour perception.

Methodology & sources

The rulesets rest on guidelines from German and European expert bodies (including DGAKI for histamine, DGKJ for lactose, DZG for coeliac), on clinical practice with conservative severity interpretation aligned with dietetics standards, and on a planned external peer review. Trigger lists and weightings are versioned; every change is traceable.

What Toleah is not

Toleah does not replace medical diagnosis or dietetic counselling. The engine works with general rulesets — individual exceptions, comorbidities or pregnancy are not modelled. If you have symptoms, you belong under medical care. Toleah can structure the time between appointments, not replace the appointments.