An emerging problem in clinical practice is how to manage the growing number of patients who experience symptoms related to the ingestion of gluten-containing foods gluten and in whom celiac disease has been ruled out. These patients most frequently report gastrointestinal symptoms such as diarrhea, abdominal discomfort or pain, bloating, flatulence, but also extraintestinal symptoms, including headache, lethargy, attention-deficit/hyperactivity disorder, ataxia, or recurrent oral ulceration. This heterogeneous syndrome, which has been reported to improve or even disappear after gluten withdrawal and to relapse after gluten challenge, is called non-celiac gluten sensitivity.1 The concept of nonceliac gluten sensitivity is not completely new. Besides several sporadic single cases reported, more than 30 years ago an oral, double-blind, placebo-controlled, cross-over gluten challenge trial showed that six out of eight adult non-celiac patients, who suffered from intestinal symptoms caused by the ingestion of gluten-containing food, were affected by gluten-sensitive diarrhea. Over the last few years, an intense debate about the existence and the prevalence of nonceliac gluten sensitivity has emerged, as shown by the considerable increase in internet forums of discussion on this topic and in availability of gluten-free food. A definition of nonceliac gluten sensitivity based on definitive scientific evidence does not still exist, and the clinical trials conducted so far in order to fill this knowledge gap are burdened by a number of biases. In a cross-over trial of subjects with suspected NCGS, the severity of overall (intestinal plus extraintestinal) symptoms increased significantly during 1 week of intake of small amounts of gluten (daily 4.375 grams), compared with placebo. Among the 59 participants in this trial, the Investigatotors identified only three true gluten-sensitive patients, defined as having a delta overall score -calculated by subtracting the weekly overall score under placebo from that under gluten- higher than the mean delta overall score plus 2 standard deviations. However, these results should be cautiously interpreted due to the lack of a control group of non-gluten-sensitive subjects.

On this basis, the Investigators will conduct a randomized, dose-finding, double-blind, placebo-controlled, cross-over gluten challenge trial aimed at comparing the effects of a daily dose of 8.4 grams of gluten with those of a daily dose of 6.0 or 4.2 grams of gluten on a cohort of subjects with nonceliac gluten sensitivity versus healthy volunteers.

The Investigators will consecutively enroll adult patients referred to outpatients' Clinics of 21 Centres because of persistence of relevant intestinal and extraintestinal symptoms believed by them to be caused by the ingestion of food containing even low doses of gluten. In their clinical management patients in gluten-containing diet since at least two months will undergo ad hoc tests including serum determination of IgA anti-tissue transglutaminase and anti-endomysial antibodies, IgG anti-gliadin antibodies (AGA), total IgA and IgE, wheat-specific IgE, upper endoscopy with collection of multiple duodenal biopsies, HLA genotyping, fecal calprotectin and lactose breath test. The Investigators will investigate serum AGA IgG since they have been detected in some patients with "suspected" nonceliac gluten sensitivity. Patients with celiac disease, wheat allergy, intolerance to further food source of fermentable oligo- and disaccharides, monosaccharides and polyols, lactose intolerance, Helicobacter pylori infection, giardiasis, inflammatory bowel disease, cirrhosis, excessive alcohol intake, intake of nonsteroidal anti-inflammatory agents, use of systemic immunosuppressant medication, poorly controlled psychiatric disease, pregnancy and those unable to give written informed consent will be excluded. The Investigators will also recruit as control group 420 age- and sex-matched healthy volunteers, undergoing IgG AGA and all negative for anti-tissue transglutaminase antibodies. Healthy volunteers will not be affected by any comorbidity and food intolerance.

Each Centre will enroll 20 patients with suspected nonceliac gluten sensitivity and 20 healthy volunteers. Recruitment of patients and, at greater extent, healthy volunteers should not be an issue as the 21 Centres involved in this trial are well-known for diagnosis and treatment of gluten-related disorders. The Investigators will consider drop-outs from the study all the patients who will fail to attend even a single appointment or test in the trial. Moreover, compliance with the study treatment will be assessed by unused capsule counts at T2 and T4. Compliance with the gluten-free diet at each time point will be assessed through a validated questionnaire. To exclude the possibility of patients being able to discriminate between gluten capsules and placebo capsules, we preliminarily tested capsule recognition in a group of 20 healthy volunteers. On two different days, healthy volunteers were asked to take capsules and then clearly state if the capsules contained gluten or not. Healthy volunteers were not able to differentiate the appearance and taste of gluten capsules from those of placebo capsules. Assuming a within-patient comparison and a SD of the overall score of 40, 70 subjects per arm would be needed in order to achieve a power of 80%, at a 2-sided 5% significance level, if the true difference is 15.

Inclusion Criteria:

• persistence of relevant intestinal and extraintestinal symptoms believed by them to be caused by the ingestion of food containing even low doses of gluten

Exclusion Criteria:

• celiac disease
• wheat allergy
• intolerance to further food source of fermentable oligo- and disaccharides, monosaccharides and polyols
• lactose intolerance
• Helicobacter pylori infection
• giardiasis
• inflammatory bowel disease
• cirrhosis
• excessive alcohol intake
• intake of nonsteroidal anti-inflammatory agents
• use of systemic immunosuppressant medication
• poorly controlled psychiatric disease
• pregnancy
• patients unable to give written informed consent will be excluded