Fig. 3: Bacterial microbiota change with age and supplementation.

a Relative abundances of bacterial phyla in 12 (top) and 24 (bottom) month old children based on 16S data. b Rarefaction curves comparing mean species richness by age, micronutrient supplementation, nutritional status, and place of residence (site). Shaded regions represent standard errors and the dotted lines denote the read depth at which significance was tested. c Non-metric multidimensional scaling of bacterial compositions based on Bray–Curtis dissimilarities. Samples are colored by age, and ellipses represent 95% confidence intervals. Arrows indicate the direction of increasing abundance of the top 10 bacterial OTUs significantly correlated with the ordination axes, scaled by their strength of correlation (r2). d Mean DESeq2-transformed abundance of Actinobacteria and Firmicutes in samples (n = 80) grouped by age and treatment. e Compositional variance among samples grouped by supplementation arm and age, based on weighted UniFrac dissimilarities and measured as distances to the centroid. Significance was tested using two-way ANOVA with the Tukey HSD post-hoc test (n = 80, 12 months CTL vs MNP p = 0.0032, CTL vs MNP + Zn p = 0.0029) f Venn diagram shows numbers of bacterial taxa significantly increased or decreased in abundance, as indicated by arrows, in supplemented groups relative to the control. Numbers in brackets refer to taxa at 12 and 24 months of age. Select taxa are listed. g Normalized abundance of Escherichia–Shigella and Bifidobacterium OTUs across supplementation arms at 12 months (n = 80). Significant differences were determined using the Wald test in DESeq2 and Benjamini–Hochberg multiple testing correction (Escherichia–Shigella CTL vs MNP p = 2.6e-11, MNP vs MNP + Zn p = 7.1e-17; Bifidobacterium CTL vs MNP p = 0.029, MNP vs MNP + Zn p = 0.0064). Boxplots in d, e, and g represent medians and interquartile ranges (IQR), and whiskers demarcate 1.5x IQR. Asterisks in (e and g) are *p < 0.05, **p < 0.01, ***p < 0.001, ns not significant.