Exclusive breastfeeding is associated with the gut microbiome maturation in infants according to delivery mode

Abstract

Exclusive breastfeeding (EBF) plays a crucial role in infant gut microbiome assembly and development. However, few studies have investigated the effects of EBF in restoring a perturbed microbiome. In this study, we applied whole metagenomic sequencing to assess the gut microbiome assembly in 525 Brazilian infants from 3 to 9 months of age of the Germina Cohort, demonstrating the early determinants of microbial taxonomy and function modulation. Our analysis shows that EBF alters the relative abundance of genes related to the microbiome taxonomy and function, with effects varying by delivery mode. EBF alters the pattern of carbohydrates, lipid metabolism, and cell structure pathways depending on the delivery mode. The microbiome age is closer to chronological infant age in EBF than in non-EBF infants, meaning a lower microbiome maturation index (MMI). Using a complementary machine learning approach, we show that Escherichia coli, Ruminococcus gnavus, and Clostridium neonatale, as well as vitamin K and o-antigen pathways contribute strongly to EBF prediction. Moreover, EBF influences the microbiome maturation in early life, toward a microbiome age more similar to the chronological infant's age.

Keywords: Early-life gut microbiome; exclusive breastfeeding; microbiome functional pathways.

Figure 1.

Compositional and functional profiles of the gut microbiome in 966 samples from infants aged 3–9 months. PCoA of Bray-Curtis distances illustrating a) microbial species; b) functional pathways. The analysis is colored by delivery mode and further stratified by breastfeeding status (EBF = exclusive breastfeeding; non-EBF = non-exclusive breastfeeding) and time points. Statistical significance was assessed using PERMANOVA (two-sided).

Figure 2.

Alpha diversity and taxonomic composition of the gut microbiome in 966 samples from infants aged 3–9 months. a) Chao1 index (alpha diversity) and b) Heatmap of the relative abundance of the top 20 bacterial species, stratified by time points and colored according to infant groups. Statistical comparisons between groups were performed using ANOVA, followed by Tukey’s honest significant difference post-hoc test for pairwise comparisons. Significance levels are denoted as ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05. Infant groups: Vag_EBF = vaginally delivered-EBF; CS_EBF = C-section delivered-EBF; Vag_nonEBF = vaginally delivered-non-EBF; CS_nonEBF = C-section delivered-non-EBF.

Figure 3.

Effect of exclusive breastfeeding on the taxonomic and functional profiles of the gut microbiome in vaginally and C-section delivered infants. a) differentially abundant species at 3–4 months and; b) differentially abundant functional pathways at 3–4 months. The plots depict unique and shared species or pathways for each delivery mode, with colors representing the coefficient value of the significant associations with EBF (q < 0.10). Unique features are specific to either vaginal or C-section delivery, while shared features are common to both delivery modes.

Figure 4.

Top ten species attributed to carbohydrate pathways significantly associated with EBF at 3–4 months of age. a) C-section delivered infants and b) vaginally delivered infants. The plots display the top ten species attributed to carbohydrate metabolism pathways that show a significant association with exclusive breastfeeding. Blue and red triangles indicate pathways that increase and decrease in abundance at 3–4 months of age, respectively.

Figure 5.

Contribution of each EBF predictors in a random forest classifier. a) top-ranked species predicting EBF at each time point and delivery mode. b) top-ranked functional pathways predicting EBF at each time point and delivery mode. The X-axis represents SHAP values, and the Y-axis lists species/functional pathways. Each dot represents an infant, with colors indicating the relative abundance of the species/pathways. Only statistically significant models (p ≤ 0.05) and features (p ≤ 0.01) are included for clarity and conciseness. Thus, functional pathways at time points 5–9 months are excluded from the figure.

Figure 6.

Compositional and functional microbiome trajectories by infant groups. Microbiome trajectory mean and 95% confidence interval of a) species and; b) functional pathways are shown for infants grouped by delivery mode and EBF status. The X-axis represents infants' chronological age at sample collection, and the Y-axis represents the microbiome maturation index. The zoomed-in portion of the plot emphasizes differences in trajectories at each time point. The tables summarize performance metrics, including mean absolute error (MAE), R² values, and linear p-values (k, n), for comparisons between infant groups. Infant groups: Vag_EBF = vaginally delivered-EBF; CS_EBF = C-section delivered-EBF; Vag_nonEBF = vaginally delivered-non-EBF; CS_nonEBF = C-section delivered-non-EBF.