Very Preterm Children Gut Microbiota Comparison at the Neonatal Period of 1 Month and 3.5 Years of Life

Abstract

Prematurity is a risk factor for dysbiosis of the gut microbiota due to particular birth conditions and frequent prolonged hospitalization of neonates. Although gut microbiota colonization after birth and its establishment during the hospitalization period have been studied in preterm infants, data on gut microbiota following discharge, particularly during early childhood, are scarce. The present study investigated the relationship between gut microbiota at 1 month after birth (hospitalization period) and 3.5 years of age in 159 preterm children belonging to the French EPIFLORE prospective observational cohort study. Analysis using bacterial 16S rRNA gene sequencing showed that the gut microbiota of preterm neonates at 1 month was highly variable and characterized by six distinct enterotypes. In contrast, the gut microbiota of the same children at 3.5 years of age showed less variability, with only two discrete enterotypes. An absence of association between enterotypes at 1 month and 3.5 years of age was observed. While the alpha diversity of gut microbiota significantly increased between 1 month and 3.5 years of age, for both alpha and beta diversities, there was no correlation between the 1-month and 3.5-years time points. Comparison at 3.5 years between children born either preterm (n = 159) or full-term (n = 200) showed no differences in terms of enterotypes, but preterm children harbored a lower Shannon diversity index and a different overall composition of microbiota than full-term children. This study suggests that the characteristics of the early gut microbiota of preterm children are not predictive of the microbial community composition at 3.5 years of age. However, the impact of gestational age is still noticeable on the gut microbiota up to 3.5 years of age.

Keywords: DOHaD; children; enterotypes; gut microbiota; prematurity.

FIGURE 1

Gut microbiota composition and diversity of the 141 preterm infants at 1 month after birth. (A) Clustering based on the genus taxonomic profiles. Biplot arrows indicate the top six genera that drive the sample to different locations on the plot. (B) The boxplots represent the relative abundance of the top 10 genera distributed among the five enterotypes. The sixth enterotype is not represented because it is constituted by infants without sample amplification (low bacterial load). The boxplots show the smallest and largest values, 25 and 75% quartiles, the median, and outliers. (C) Boxplots of alpha diversity (top) assessed by Chao1 and Shannon index, and pairwise beta diversity dissimilarities (bottom) assessed by Bray–Curtis and Unifrac distances within enterotypes based on genus taxonomic profiles (*p < 0.05, ****p < 0.001, ns = p > 0.05, Reference group = Ent3); Ent, Enterotype.

FIGURE 2

Gut microbiota composition and diversity of the 159 preterm children at 3.5 years of age. (A) Clustering based on the genus taxonomic profile. Biplot arrows indicate the top four genera that drive samples to different locations on the plot. (B) The boxplots represent the relative abundance of the top 10 genera distributed among the P_ and B_type enterotypes. The boxplots show the smallest and largest values, 25 and 75% quartiles, the median, and outliers. (C) Boxplot of alpha diversity assessed by Chao1 estimate based on the genus taxonomic profile. (D) Inter-individual dissimilarity of the genus community based on the Bray–Curtis distance and represented by the heat map and boxplot. Each cell in the heat map, ordered according to B_ and P_type enterotypes, represents the dissimilarity level as per the color scale beside the box plot (*p < 0.05, ****p < 0.0001).

FIGURE 3

Clustering of the gut microbiota of preterm (n = 159) and full-term (n = 200) children at 3.5 years of age. (A) The boxplots represent the relative abundance of the top 10 genera distributed between the B_ and P_types among the full-term children. The boxplots show the smallest and largest values, 25 and 75% quartiles, the median, and outliers. (B) Gut microbiota clustering of the 159 preterm and 200 full-term children at 3.5 years of age based on the genus taxonomic profile. Biplot arrows indicate the top four genera that drive samples to different locations on the plot. (C) Distribution of enterotypes among full-term and preterm children. The distribution is expressed as the proportion of each enterotype among each group of children. (D) Boxplot of alpha diversity assessed by Shannon index based on the genus taxonomic profile between preterm and full-term children (*p < 0.05). The boxplots show the smallest and largest values, 25 and 75% quartiles, the median, and outliers.

FIGURE 4

Comparison of the gut microbiota of the same preterm children at 1 month and 3.5 years of age. (A) Boxplots of alpha (top) and beta diversities (bottom) based on both Chao1 and Shannon index and both Bray–Curtis and Unifrac dissimilarity distances, respectively, at genus taxonomic level in the 141 preterm children with sequencing data at two time points. (B) Distribution of B_ and P_type enterotypes among the six enterotypes at 1 month in 159 preterm children with a fecal sample at two time points. The values are expressed as proportions of each enterotype at 3.5 years of age among each enterotype at 1 month. (C) Boxplots of Shannon diversity based on genus level at 3.5 years of age according to enterotype at 1 month (*p < 0.05, ns = p > 0.05; Reference group = Ent3, Ent, Enterotype). The boxplots show the smallest and largest values, 25 and 75% quartiles, the median, and outliers. ****p < 0.0001.