Gluten

What is it?

Gluten is the protein complex in wheat, rye, barley and their crosses. Three distinct conditions involve it, and they are not variants of one thing: coeliac disease, which is autoimmune, needs strict lifelong avoidance and is diagnosed by biopsy; wheat allergy, which is IgE-mediated; and non-coeliac gluten sensitivity, symptoms without the coeliac markers.

What "gluten-free" legally means

Commission Implementing Regulation (EU) No 828/2014 sets two tiers. A food may be labelled gluten-free at not more than 20 mg of gluten per kilogram, and very low gluten at not more than 100 mg/kg. Those are conditions a food must meet to carry the words, not measurements of anything: the regulation sets a ceiling and publishes no gluten figure for any food. It also defines the term — gluten is a protein fraction from wheat, rye, barley, OATS or their crossbred varieties, and wheat means any Triticum species.

Oats are in the legal definition and not in the practice

Both texts name oats. Regulation (EU) No 1169/2011 lists them in Annex II among the cereals containing gluten, and 828/2014 includes them in its definition. AFDIAG, the French coeliac association, writes that uncontaminated oats would be edible by almost all coeliacs — and keeps them prohibited all the same. The law and the clinical practice do not say the same thing here, and the gap is the answer: oats are excluded by default, certified uncontaminated oats are a separate question to settle with a clinician.

The exemptions are what a food name never tells you

Annex II is a labelling rule — it says what must be declared, never how much is there — and the part that carries the weight is its exceptions. Wheat-based glucose syrups including dextrose are exempt, as are wheat-based maltodextrins, glucose syrups from barley, and cereals used to make distillates for spirit drinks. An ingredient can therefore come from wheat and not have to be declared as a gluten cereal. No amount of label-reading recovers that; only knowing the exemption does.

Drinks split on one process

Coeliac UK states it in a sentence: all spirits are distilled, and distillation removes any trace of gluten, so spirits are suitable whatever grain they started from. Beers, lagers, stouts and ales contain varying amounts and are not. The distinction worth carrying into a shop is between a beer labelled gluten free, which must be at 20 ppm or less, and a beer sold as gluten removed, which is a different claim. Coeliac UK publishes no measurement of any food, so what it gives is guidance and a rule of thumb, not a figure.

Hidden sources, and one thing that has been tested

Soy sauce, beer, malt, some sauces and seasonings, and oats unless certified. The third condition above has been tested directly: Biesiekierski (2013) put 37 people with self-reported gluten sensitivity on a reduced-FODMAP diet first, then challenged them double-blind with 16 g of gluten a day, 2 g, or whey protein alone. No gluten-specific effect survived the FODMAP reduction. That is a trial in 37 people with self-reported sensitivity and irritable bowel syndrome — not a composition study, and not about coeliac disease, for which avoidance is not in question. It does not settle what those people react to; it does mean the gluten in a wheat food is not the only candidate.

Source: Commission Implementing Regulation (EU) No 828/2014, Article 2 and Annex, for the 20 mg/kg and 100 mg/kg tiers and for the definition of gluten — which includes oats; Regulation (EU) No 1169/2011, Annex II, for the cereals that must be declared and for the exemptions, wheat-based glucose syrups among them — a labelling rule that says what must be declared and never how much is there; AFDIAG (France), for the three prohibited cereals and for keeping oats out while noting that uncontaminated oats would be edible by almost all coeliacs; Coeliac UK, for distillation removing gluten from spirits and for the gluten-free versus gluten removed distinction in beer — guidance, with no per-food measurement; Biesiekierski J.R. et al., Gastroenterology 145 (2013) 320-328, double-blind cross-over in 37 people with self-reported sensitivity after a FODMAP reduction

Frequently asked questions