In blinded rechallenge trials, fructans — not gluten — produced the symptoms

When people who believe they react to gluten are given gluten without knowing it, the reaction mostly does not appear. In the largest blinded rechallenge study of its kind, symptoms were worse after a fructan challenge than after a gluten challenge, and there was no difference in symptom scores between gluten and placebo at all [s1].

That does not mean the symptoms are imagined. It means the substance being blamed is probably not the one responsible.

The Oslo crossover trial

Researchers at Oslo University Hospital recruited 59 people who had put themselves on a gluten-free diet and in whom coeliac disease had been excluded [s1]. Between October 2014 and May 2016, participants were randomly assigned to seven days of a diet containing gluten (5.7 g), fructans (2.1 g), or placebo, concealed inside muesli bars [s1]. After a washout of at least seven days — long enough for symptoms from the previous challenge to resolve — each participant crossed over into a different group until all three challenges had been completed [s1]. Symptoms were measured with the irritable bowel syndrome version of the Gastrointestinal Symptom Rating Scale [s1].

Fructans are a fermentable carbohydrate found in wheat alongside gluten — and in onions, garlic and a long list of other foods. They belong to the FODMAP group.

Overall GSRS-IBS scores differed significantly across the three challenges: 33.1 ± 13.3 after gluten, 38.6 ± 12.3 after fructan, and 34.3 ± 13.9 after placebo (P = .04) [s1]. The bloating subscale followed the same pattern: 9.3 ± 3.5, 11.6 ± 3.5 and 10.1 ± 3.7 respectively (P = .004) [s1].

The comparisons that matter are the pairwise ones. Overall scores were significantly higher after fructan than after gluten (P = .049), as were bloating scores (P = .003) [s1]. And between gluten and placebo, there was no difference in GSRS-IBS scores [s1].

Counting individuals rather than means gives the same answer from a different angle. Thirteen participants had their highest overall score after gluten, 24 after fructan, and 22 after placebo [s1]. Twenty-two people out of 59 felt worst on the challenge that contained nothing.

The earlier Australian trial that found the same absence

A double-blind crossover trial published in 2013 had already tested this in 37 people with self-reported non-coeliac gluten sensitivity and IBS by Rome III criteria, coeliac disease excluded [s2]. Participants spent two weeks on a diet with reduced FODMAP content, then one week on a high-gluten diet (16 g gluten a day), a low-gluten diet (2 g gluten plus 14 g whey protein), or a control diet (16 g whey protein), with washouts of at least two weeks [s2].

Gastrointestinal symptoms improved consistently and significantly in all participants during the period of reduced FODMAP intake — and then worsened to a similar degree whether the diet that followed contained gluten or whey protein [s2]. Gluten-specific effects were observed in only 8% of participants, and there were no diet-specific changes in any of the serum and faecal markers of intestinal inflammation, injury or immune activation the researchers measured [s2].

Twenty-two of the participants then crossed over into a shorter three-day rechallenge with gluten (16 g a day), whey (16 g a day) or no additional protein. Symptoms increased by similar amounts in all groups, and the gluten-specific gastrointestinal effects were not reproduced [s2]. The authors also report an order effect — the sequence in which challenges were given influenced results, which is itself a warning about interpreting unblinded self-experiments [s2].

Why the distinction is practically useful

If the trigger in wheat is fructan rather than gluten, then a gluten-free diet is an approximate and expensive way to solve the problem — it removes wheat, which removes fructans as a side effect, while leaving fructans in onion, garlic and legumes untouched. That would explain the common experience of someone going gluten-free, improving partially, and never getting all the way there.

It also matters for what gets excluded before any of this applies. Both trials enrolled only people in whom coeliac disease had been ruled out [s1] [s2]. Coeliac disease is a distinct autoimmune condition with real intestinal damage, and testing for it is unreliable once someone has already stopped eating gluten — which is the order in which most people do it.

The 2025 European consensus on functional bloating puts the clinical position bluntly. There is insufficient evidence to recommend a gluten-free diet in patients with functional bloating and abdominal distension unless they have coeliac disease, a statement reached with 90% expert agreement on low-quality evidence [s3].

What this does not settle

Neither trial tested large numbers of people, and neither ran long. The Oslo study's own design gave one week per challenge, which cannot detect anything slower [s1]. The 2013 trial's finding that gluten-specific effects appeared in 8% of participants is not zero, and a small genuinely gluten-reactive subgroup remains possible [s2]. Both studies enrolled people who had already self-selected onto gluten-free diets, which is a particular population.

What the evidence does support is a reordering. The strongest blinded human data attributes the symptoms to a fermentable carbohydrate rather than to gluten [s1], a second blinded trial could not reproduce a gluten effect at all [s2], and the relevant European consensus does not recommend gluten avoidance outside coeliac disease [s3]. The popular explanation and the tested one have come apart.

This article is informational and is not medical advice. Coeliac disease requires testing while gluten is still being eaten, and decisions about dietary restriction belong with a reader and a clinician.

Sources

• Fructan, Rather Than Gluten, Induces Symptoms in Patients With Self-Reported Non-Celiac Gluten Sensitivity — Gastroenterology, 2017-11-02
• No effects of gluten in patients with self-reported non-celiac gluten sensitivity after dietary reduction of fermentable, poorly absorbed, short-chain carbohydrates — Gastroenterology, 2013-05-04
• European Consensus on Functional Bloating and Abdominal Distension — An ESNM/UEG Recommendations for Clinical Management — United European Gastroenterology Journal, 2025-08-22