Effect of gluten and wheat on symptoms and behaviours in adults with irritable bowel syndrome: a single-centre, randomised, double-blind, sham-controlled crossover trial

Abstract

Background

Many patients with irritable bowel syndrome (IBS) believe gluten or wheat triggers their symptoms. We compared symptomatic responses to wheat and gluten with gluten-free sham challenge in patients with IBS who previously perceived benefit from a gluten-free diet.

Methods

We conducted this randomised, double-blind, sham-controlled crossover study at McMaster University Medical Centre, ON, Canada. Eligible participants were adults aged 18 years or older who met Rome IV criteria for IBS and had previously self-reported improvement on a gluten-free diet, which was implemented for at least 3 weeks before enrolment. Eligible participants were randomly assigned (1:1:1:1:1:1) to receive one of six sequences of wheat, gluten, and sham (containing gluten and wheat free flour) in three periods of 7 days, separated by 14-day washout periods. Randomisation was done using the randomizeBE package in R and the cereal bars were designed to have the same appearance, taste, and smell to maintain blinding. The primary outcome was worsening of IBS symptoms of at least 50 points on the IBS Symptom Severity Score (IBS-SSS) after dietary challenges. Outcome and safety analyses were done in all patients who completed all three challenges. This trial was registered with ClinicalTrials.gov, NCT03664531.

Findings

Between Nov 15, 2018, and June 19, 2023, we assessed 101 people for eligibility. 72 people were excluded due to ineligibility (n=15), refusing screening (n=42), and refusing participation (n=15). 29 participants were enrolled and randomly assigned to wheat–gluten–sham (n=5), wheat–sham–gluten (n=5), gluten–wheat–sham (n=5), gluten–sham–wheat (n=5), sham–wheat–gluten (n=5), and sham–gluten–wheat (n=4). One participant in the wheat–sham–gluten group completed the first challenge but withdrew without providing a reason. In the 28 patients completing the study, there were no statistically significant differences in the proportion of participants with a worsening of IBS-SSS of at least 50 points after wheat (11 [39%] of 28 participants, risk difference vs sham 0·11; 95% CI –0·16 to 0·35) or gluten (ten participants [36%], 0·07; –0·19 to 0·32) versus sham (eight participants [29%]). Adverse events were reported in 26 (93%) of 28 patients after wheat, 26 patients (93%) after gluten, and 26 patients (93%) after sham. Study emergent adverse events were similar between challenges (five [18%] of 28 participants after wheat, five [18%] after gluten, and seven [25%] after sham). No patients reported severe adverse events.

Interpretation

IBS patients with self-perceived gluten sensitivity reacted similarly to gluten, wheat, and sham challenges. These findings suggest that expectations played a major role in symptom generation, and that only some of these patients could benefit from gluten or wheat restriction. Identifying this subset of patients while destigmatising wheat and gluten in the remaining ones should be considered for effective management of patients with IBS.

Funding

Canadian Digestive Health Foundation, Society for the Study of Celiac Disease.