Impact of early events and lifestyle on the gut microbiota and metabolic phenotypes in young school-age children

Abstract

Background: The gut microbiota evolves from birth and is in early life influenced by events such as birth mode, type of infant feeding, and maternal and infant antibiotics use. However, we still have a gap in our understanding of gut microbiota development in older children, and to what extent early events and pre-school lifestyle modulate the composition of the gut microbiota, and how this impinges on whole body metabolic regulation in school-age children.

Results: Taking advantage of the KOALA Birth Cohort Study, a long-term prospective birth cohort in the Netherlands with extensive collection of high-quality host metadata, we applied shotgun metagenomics sequencing and systematically investigated the gut microbiota of children at 6-9 years of age. We demonstrated an overall adult-like gut microbiota in the 281 Dutch school-age children and identified 3 enterotypes dominated by the genera Bacteroides, Prevotella, and Bifidobacterium, respectively. Importantly, we found that breastfeeding duration in early life and pre-school dietary lifestyle correlated with the composition and functional competences of the gut microbiota in the children at school age. The correlations between pre-school dietary lifestyle and metabolic phenotypes exhibited a striking enterotype dependency. Thus, an inverse correlation between high dietary fiber consumption and low plasma insulin levels was only observed in individuals with the Bacteroides and Prevotella enterotypes, but not in Bifidobacterium enterotype individuals in whom the gut microbiota displayed overall lower microbial gene richness, alpha-diversity, functional potential for complex carbohydrate fermentation, and butyrate and succinate production. High total fat consumption and elevated plasma free fatty acid levels in the Bifidobacterium enterotype are associated with the co-occurrence of Streptococcus.

Conclusions: Our work highlights the persistent effects of breastfeeding duration and pre-school dietary lifestyle in affecting the gut microbiota in school-age children and reveals distinct compositional and functional potential in children according to enterotypes. The findings underscore enterotype-specific links between the host metabolic phenotypes and dietary patterns, emphasizing the importance of microbiome-based stratification when investigating metabolic responses to diets. Future diet intervention studies are clearly warranted to examine gut microbe-diet-host relationships to promote knowledge-based recommendations in relation to improving metabolic health in children.

Keywords: Enterotype; Gut microbiota; Metabolic phenotypes; School-age children.

Fig. 1

Comparison between early school-age Dutch children and adults. a Categories of phenotypic data collected within the KOALA cohort. b Box plot showing the gene-based α-diversity (Shannon index) in early school-age Dutch children (n = 281) and healthy Dutch adults (n = 62). c Genus-based principal component analysis (PCA) of children and adults. d Box plots of intra- and inter-group beta diversity based on genus profiles in children and adults. The “Intra-children” and “Intra-adults” indicate the genus-based beta diversity in children and adults, and the “Inter-groups” indicates the genus-based beta diversity between children and adults (***P < 0.001; Wilcoxon rank-sum test). e Box plots displaying the ten most abundant genera among children and adults. Genera indicated with red font are enriched in children, and genera in blue are enriched in adults. Boxes are ordered according to median relative abundance of genus in children

Fig. 2

Stratification of early school-age children into three enterotypes based on their gut microbiome. a Scatter plot representing the three enterotypes identified using Dirichlet multinomial mixtures (DMM)-based clustering among early school-age Dutch children. MDS multidimensional scaling. b Genus abundance box plots showing the main contributors of each enterotype. c Correlations between enterotype enriched species, with the log10-transformed relative abundance of each species indicated by the circle area. Only the top 10 most abundantly enriched species in each enterotype are displayed. Red line indicates positive correlation and gray line indicates negative correlation (SparCC, pseudo P < 0.01)

Fig. 3

Multiple early events and pre-school lifestyle associated with the school-age gut microbiota. a PCA showing the multivariate variation of children and the major contributions of different factors to PC1 and PC2. A total of 18 factors including early events and pre-school lifestyle (Additional file 1: Table S1) were subjected to PCA, and those factors with component scores for PC1 or PC2 ≥ 0.2 were shown as major contributors. Box plots showing the overall distribution of PC1 and PC2 scores within each enterotype (#P<0.05; Wilcoxon rank-sum test). b Significant major contributors in PCA between enterotypes (#P < 0.05, Wilcoxon rank-sum test; *P < 0.05, Dunn’s post-hoc test). The details of statistical results for all factors are shown in Additional file 1: Table S11. c PERMANOVA of the influence of single-factor early events and pre-school lifestyle on the gut microbial gene profile in the entire cohort and within each enterotype (#P < 0.05; * adjusted P < 0.05)

Fig. 4

Enterotype-associated differences in potential for butyrate, succinate, and propionate production. a Pathway showing the genes involved in final biosynthetic steps from crotonyl-CoA to butyrate, including bcd (butyryl-CoA dehydrogenase, K00248), ptb genes (phosphate butyryltransferase, K00634), buk genes (butyrate kinase, K00929), and but (butyryl CoA:acetate CoA transferase, K01034). b The relative abundance of genes involved in butyrate production within each enterotype. c The relative abundance of genes involved in succinate production (succinate dehydrogenase complex (K00239, K00240 and K00241) within each enterotype. d Mean relative abundance of bcd genes (K00248) listed according to annotated bacterial species within each enterotypes. e Mean relative abundance of sdhA genes (K00239) listed according to annotated bacterial species within each enterotypes. Dunn’s post-hoc test; **P < 0.01, ***P < 0.001