Fig. 1: Fecal microbiota composition and diversity associated with peanut oral immunotherapy outcomes and peanut allergy severity.

a Schematic overview of IMPACT trial fecal microbiome study⁹. Numbers represent the number of fecal samples in each dataset and time point. b No significant difference in α-diversity (Faith’s phylogenetic diversity) was observed over the course of the IMPACT trial between Peanut Oral Immunotherapy (POIT) and placebo arms (β = −0.84, SE = 1.40, df = 166, t = −0.60, P = 0.55, two-sided linear mixed-effects model). Data are presented as mean values ± SEM. c No significant difference in α-diversity (Faith’s phylogenetic diversity) was observed over the course of the IMPACT trial between the POIT outcome groups (D+R−, D−R−, and D+R+), after adjustment for age using a two-sided linear mixed-effects model with random intercepts for subject (D+R− vs. D+R+: β = 2.58, t = 1.59, P = 0.115 and D−R− vs. D+R+: β = 2.69, t = 1.26, P = 0.210). d However, at baseline, prior to POIT initiation, the D+R+ group exhibited significantly lower phylogenetic diversity compared to either the D+R− and D−R− groups. Wilcoxon signed-rank test (two-sided, n = 47; D+R+ = 16, D+R− = 23, D−R− = 8). Data are presented as mean values ± SEM. Boxplots show the median (center line), 25th and 75th percentiles (box bounds), and whiskers extend to values within 1.5× the interquartile range. e Although, the longitudinal fecal microbiota composition (unweighted Unifrac distance matrix) was similar between POIT and placebo arms (two-sided linear mixed-effect model, n = 90 subjects and 263 samples. P = 0.80 for Axis 1, and P = 0.78 for Axis 2). f Significantly different fecal microbiota composition was observed within the POIT outcome groups (two-sided linear mixed-effect model, n = 57 subjects and 190 samples, Axis 1: D+R+ vs. D+R−, p = 0.01 and D+R+ vs. D−R−, p = 0.004). g D+R+ group exhibited significantly different fecal microbiota composition compared to both D+R− (p = 0.01) and D−R− (p = 0.004) groups throughout the trial (two-sided linear mixed-effect model). Data are presented as mean values ± SEM. Boxplots show the median (center line), 25th and 75th percentiles (box bounds), and whiskers extend to values within 1.5× the interquartile range. h Baseline gut bacterial phylogenetic diversity positively correlates with baseline total IgE, i peanut-specific IgE, and j Ara h 2-specific IgE levels, respectively. Participants who enrolled in the IMPACT trial but did not complete the trial were included as 16S rRNA sequencing and clinical serum IgE levels were available (two-sided Pearson correlation, P < 0.05, adjusted for age). k Baseline differentially abundant bacterial taxa between children who achieved remission (n = 16) versus no remission (n = 31). Two-sided linear mixed-effect model (P.FDR < 0.05, adjusted for age). l Baseline Peanut-IgE associated bacterial taxa (n = 47). Two-sided linear mixed-effect model (P.FDR < 0.05). Estimate represents the predicted value of the effect size or relationship derived from the statistical model, reflecting the magnitude and direction of the association. EoB: End of Buildup, MM: Mid Maintenance, EoT: End of Treatment, and EoA: End of Avoidance. D+R+: Desensitized and Remission, D+R−: Desensitized no Remission, D−R−: Not desensitized and no Remission. LME: Linear mixed-effect model. Source data are provided as a Source Data file.