Fig 1.
Correlodendogram with pearson correlations among bacterial translocation, intestinal permeability, villus structure, metabolites, and growth variables.
Color intensity and size of the circle are proportional to the correlation coefficients. Significant correlations are shown following correction for multiple testing. FABP, fatty acid binding protein; LAZ, length-for-age Z score; IGF-1, insulin like growth factor 1; IGFBP-3, insulin like growth factor binding protein 3; LPS, lipopolysaccharide; LBP, lipopolysaccharide binding protein; MUAC, mid-upper arm circumference; VH, villus height; VPMM, villus perimeter or epithelial surface area; VW, villus weight; WAZ, weight-for-age Z score; WLZ, weight-for-length Z score.
Table 1.
Characteristics of the participants.
Fig 2.
Mucosal abnormalities in biopsies from hospitalized children.
The biopsies showed a range of abnormalities including (A) moderate villus blunting or (B) total villus atrophy.
Fig 3.
Dendogram and heatmap representation of unsupervised hierarchical clustering (HCA) of the metabonome for all children sorted according to the villus height (VH).
Each column corresponds to a single children and each row corresponds to a specific metabolites. Only metabolites identified to correlate with VH through and OPLS model were used for metabolite clustering. Metabolite z-score transformation was performed on the levels of each metabolite across samples, with blue denotating a lower and red a higher level compared to the mean. Metabolites were clustered using correlation distance and average linkage. DMA, dimethylamine; 2-HIB, 2−hydroxyisobutyrate; 4-HPP, 4-hydroxypurate; 4-HPA, 4-hydroxyphenylacetate; 3-IS, 3−indoxyl sulfate; NAG, N-acetylglycoprotein; PAG, phenylacetylglutamine; TMA, trimethylamine.