Clinical Comparison of the Impact of Two Dietary Interventions on Abdominal Pain and Intestinal Microbiome Caused by Simple Carbohydrate Intolerance.

This is a prospective cross over study. Participants who meet the study entry criterion, will be randomized to a specific dietary intervention: low FODMAP for a 3-week period or Limited Added Sugar for 3 weeks. After the initial study phase, participants will enter a 3 week break followed by a crossover to the other diet phase.

Each participant will be randomized to either the low-FODMAP diet group (LFD) or the low added sugar diet group (LAS) for 3 weeks. All participants will be assigned to a nutrition intervention consisting of 2 one-hour counseling sessions (pre and post diet initiation) plus one 30 min telephone session at week 1 and 2 of the intervention for both diet groups. Macronutrients will not be restricted and both study diets will be ad libitum. Diet interventions will include administration of a 24 hour food recall and completion of the Comprehensive Nutrition Assessment Questionnaire in addition to comprehensive education of the assigned diet.

Investigators will collect stool samples prior to intervention and post each diet intervention to assess initial microbiome composition as well as post intervention changes. Symptom data will be collected via an online survey link consisting of abdominal pain related questionnaires. Measures for child and parent will be requested at the time enrollment of the study and at the end of each diet phase.

Animal studies indicate microbiome adjustment to an increase in dietary carbohydrate pool. It is difficult to capture this change in humans as the microbiome shift may not be evident by actual change in certain strains but by genomic adjustment. Investigators propose a metagenomic approach in combination with metabolomics to identify microbiome end products present as a result of changes in dietary carbohydrate intake by dietary elimination.

Investigators will characterize shifts in metagenomic composition in response to dietary shifts using multivariate methods and linear mixed effects models. To address our hypothesis that FODMAP and low added sugar diets will affect microbial gene content in similar ways, investigators will compare metagenomic shifts relative to a baseline diet using multivariate models that control for subject, age, sex, diet order and potential confounding factors in PerMANOVA and distance-based redundancy analyses. Investigators will have 95% power to detect medium shifts in metabolic profiles among dietary groups (Cohen's D = 0.520). Investigators will use mixed effect linear models to measure enrichment for specific metabolic pathways and diet-driven shifts in specific metabolites over time while controlling for subject, age, sex, diet order, symptoms and potential confounding factors. In these repeated measures design, we will have 95% power to detect a small to medium effect of diet on a given metabolite (Cohen's D = 0.371). The Benhamini-Hochberg procedure will be applied to decrease the false discovery rate when running multiple tests (Yoav et al) It is also of interest to analyze how microbial gene content and metabolic profiles covary. Investigators will apply qualitative genetic methods of covariance matrix estimation to determine how metabolite profiles can be predicted by metagenomic content.