Fig. 1
Obesity disrupts the mucosal-luminal interface (MLI) microbiome of non-IBD controls with shifts that parallel those in Crohn’s disease (CD). A Alpha diversity metrics (Chao1 and Shannon indices) are shown for the sigmoid and cecum MLI microbiome of non-IBD controls subdivided into three BMI categories. Significance was determined by ANOVA adjusting for gender and age with post hoc Tukey. *p<0.05, **p<0.01. B Principal coordinates analysis (PCoA) plot based on Bray-Curtis dissimilarity visualizing the cecal and sigmoid microbiome of controls subdivided into three BMI categories (indicated by color). Each dot represents one subject; ellipses represent 95% confidence intervals. p-values calculated by multivariate Adonis adjusting for gender and age. C Taxonomic summary at the phylum level of the cecal and sigmoid colon MLI microbiome of controls. *q<0.05, **q<0.01, ***q<10−4 for obese vs. BMI <25 in DESeq2 models adjusting for gender. D Differential ASVs in obese vs. normal weight controls in the cecal MLI microbiome from DESeq2 models controlling for gender and age. Effect size is represented as the log2 fold change between CD and control. Size of each dot is proportional to normalized relative abundance. Only ASVs with a mean normalized relative abundance greater than 10−5 are included. Color corresponds to phylum. ASVs are grouped into genera; those without an assigned genus are represented as unclassified members of families (f) or orders (o). E CD dysbiosis index, representing the log ratio of taxa enriched in CD to taxa depleted in CD, is shown for CD patients and controls stratified into three BMI categories. Significance determined by Kruskal Wallis with post hoc Dunn test corrected by Benjamini-Hochberg. *p<0.05, ***p<0.001