Gluten-free diet

A gluten-free diet removes wheat, barley, and rye. It is the required treatment for coeliac disease, an autoimmune condition that can present with IBS-like symptoms, which is why the ACG suggests serologic testing for coeliac disease in people with IBS and diarrhoea before any dietary change PMID 33315591. Many people without coeliac disease also report feeling better off gluten, a pattern labelled non-coeliac gluten sensitivity. The complication is that wheat carries both gluten and fructans, a FODMAP, so symptom relief after cutting wheat does not by itself show that gluten was the problem.

Controlled studies have tried to separate the two. At the Mayo Clinic, 45 patients with IBS-D were randomised to four weeks of a gluten-containing or gluten-free diet; those eating gluten had more bowel movements per day and higher small-bowel permeability, most markedly in carriers of the coeliac-associated HLA-DQ2/8 genes, while transit and biopsy histology did not differ PMID 23357715. A Monash University crossover study took 37 people with self-reported gluten sensitivity and IBS, improved their symptoms first with a low FODMAP diet, then re-challenged them blind with high-gluten, low-gluten, or whey; symptoms worsened to a similar degree on every challenge, gluten-specific effects were seen in only 8% of participants, and no biomarker changed PMID 23648697. A Norwegian crossover challenge of 59 people on self-instituted gluten-free diets, using muesli bars containing gluten, fructans, or neither, reported higher overall and bloating scores after fructans than after gluten, and no difference between gluten and placebo PMID 29102613.

Pooled evidence is thin. The 2018 systematic review that informed the ACG guideline found only two randomised trials of a gluten-free diet, with 111 participants, both of which enrolled people who had already responded to the diet and then re-exposed some of them to gluten; the pooled relative risk of symptoms persisting was 0.42 with a confidence interval from 0.11 to 1.55, not statistically significant, and the authors judged the evidence insufficient to recommend the diet PMID 30046155. The BSG reached the same conclusion and issued a weak recommendation against it, while noting that its popularity means further trials are warranted PMID 33903147.

Points clinicians commonly raise: coeliac testing is done while still eating gluten, because results become unreliable once gluten is removed; for most people with IBS the evidence suggests fructans and other FODMAPs are the more likely trigger, which a supervised low FODMAP trial can test directly; a long-term gluten-free diet has a nutritional and social cost that is justified in coeliac disease but has not been shown to be justified in IBS; and the BSG mentions preliminary data that people with serum antigliadin antibodies may be a subgroup who respond PMID 33903147.

Talk to a clinician before starting or stopping anything.