A223 THE GEM PROJECT_ ASSOCIATIONS OF DIETARY PATTERNS WITH MICROBIOME AND FECAL CALPROTECTIN IN HEALTHY FIRST-DEGREE RELATIVES OF CROHN’S DISEASE PATIENTS
THE GEM PROJECT: ASSOCIATIONS OF DIETARY PATTERNS WITH
MICROBIOME AND FECAL CALPROTECTIN IN HEALTHY FIRST- DEGREE RELATIVES OF CROHN‟S DISEASE PATIENTS
G. Sasson1, J. Raygoza Garay2, W. Turpin1, N. Power3, M. Smith3, D.S. Guttman1, T.
Research Consortium4, K. Croitoru3
1. University of Toronto, Toronto, ON, Canada; 2. Medicine, University of Toronto,
Toronto, ON, Canada; 3. Mount Sinai Hospital, Toronto, ON, Canada; 4. The CCC GEM Project, Toronto, ON, Canada
Background: Crohn‘s disease (CD) is thought to be due to an interaction between
environmental factors and the gut microbiome that activates an immune response in genetically susceptible hosts. Epidemiologic studies suggest diet is an important variable in CD development; however, little is known about the mechanism by which diet contributes to pathogenesis. It has recently been shown that diet plays a substantial role in shaping microbiome composition (MC). We hypothesize that specific diet patterns are associated with differences in MC that may be related to CD risk. Aims: To characterize associations between diet patterns with MC and fecal calprotectin (FC) in healthy first-degree relatives (FDR‘s) of CD patients in the Genetic, Environmental, Microbial (GEM) Project. Methods: A validated food frequency questionnaire (FFQ) was used to assess diet for North American FDR‘s at recruitment. Each question was summarized as a score based on weekly consumption frequency. The Dirichlet method of unsupervised clustering was used to generate dominant diet clusters. Diet-microbiome associations were assessed using the two-part microbiome model. Stool microbiota at recruitment was characterized by 16s RNA sequencing of V4 region using MiSeq platform. Baseline FC was measured by BUHLMANN ELISA test. Results: 2766 FDR‘s had FFQ‘s at recruitment; mean age 18.52 years, 53% female. The Dirichlet method identified 4 clusters, some of which resembled known diet patterns: Superbowl (mainly organ meats, non-red meat, beer, spirits), High Carbohydrate (HC), Mediterranean (MD) and Western (WD) diets. HC was associated with increased relative abundance of V. veillonella (P=2.68E-4). Both HC and MD were associated with decreased abundance of E. klebsiella (P=2.64E-5 and P=1.02E-5 respectively). WD was associated with decreased abundance of L. dorea (P=5.74E-5), a genus considered high risk for CD. A per question analysis demonstrated significant associations between several taxa and individual foods. Diet clusters were then correlated with FC, and a decrease in FC was observed with MD (estimate -16.96, P=0.012). There were no associations with FC in a per question analysis. Conclusions: Dominant dietary patterns and certain individual foods are associated with specific gut MC. As well, MD is inversely associated with FC and therefore has potential use as an intervention for lowering inflammation. Understanding relationships between diet, MC and FC in individuals at high risk for CD would be beneficial in defining new dietary strategies in predictably modulating future risk of CD.