Shifting management paradigms in gastrointestinal disorders associated with gluten

Gluten has been linked to two human conditions – coeliac disease (CD) and non-coeliac gluten sensitivity (NCGS). The role of gluten in causing CD is unequivocally proven, and there is a pressing need for robust prospective data that clarify the natural history of a condition that is increasing in incidence, but reducing in symptom severity. In contrast, it is unclear whether gluten alone can cause gastrointestinal symptoms in patients without CD, and there has been a paucity of blinded randomised and placebo controlled trials dissecting the role of gluten in human disease. This thesis aims to address these important areas. Two prospective studies in CD are presented in this work. In the first, a cohort of patients with newly diagnosed coeliac disease was reassessed at diagnosis and again at one and five years after commencing a gluten-free diet (GFD) with respect to their symptoms, intestinal healing, body composition and routine pathology testing. Response at the mucosal level was noted in the vast majority by 5 years, but coeliac antibodies correlated poorly with these outcomes. Weight gain was greatest in those with an average BMI compared to obese patients whilst skeletal muscle mass increased irrespective of baseline weight. Bone mass increased, but only in those with reduced BMD at diagnosis. An increased HLA-DQ2 and/or DQ8 dose was associated with a more severe clinical phenotype at diagnosis, but it held poor predictive value for long-term outcomes. Abnormal liver function tests occurred at any time during follow-up but overall tended to improve with treatment, whilst mild neutropenia is described as a new association. Point-of-care testing for coeliac antibodies appeared to have little place in the follow-up of CD. In the second prospective study, the issue of ‘brain fog’ was explored in patients with newly diagnosed CD in a pilot study in which patients were cognitively assessed regularly over one year. Cognition was found to improve in concert with endoscopic and serological markers of CD. NCGS was examined initially in a pilot parallel-group, double-blind, randomised, placebo-controlled trial of FODMAP-deplete wheat protein in a group of 37 patients, in which wheat protein was associated with greater gastrointestinal symptoms than placebo. However, a subsequent two trials utilizing gold-standard methodology for identifying a food intolerance – adequately-powered, double blind, placebo-controlled crossover rechallenge trials where all food was provided – found no evidence for wheat-protein-specific induction of symptoms in patients in a low FODMAP background. Lastly, an analysis of surveys received from applicants to the last two randomised trials identified short-comings in the diagnosis of NCGS that included inadequate exclusion of CD and ongoing moderate-to severe symptoms despite adherence to a GFD. In conclusion, these studies have provided important data that inform management of CD and NCGS. The positive effects of a GFD for multiple outcomes in patients with newly-diagnosed CD underline the clinical value of dietary adherence even in the apparently asymptomatic. In contrast, the use of gluten-free diet in those with gastrointestinal symptoms but without CD was not supported and raises issues regarding the entity of NCGS.