Gut and respiratory tract microbiota in children younger than 12 months hospitalized for bronchiolitis compared with healthy children: can we predict the severity and medium-term respiratory outcome?

Abstract

Growing evidence indicates that gut and respiratory microbiota have a potential key effect on bronchiolitis, mainly caused by respiratory syncytial virus (RSV). This was a prospective study of 96 infants comparing infants with bronchiolitis (n = 57, both RSV and non-RSV associated) to a control group (n = 39). Gut (feces) and respiratory [nasopharyngeal aspirate (NPA)] microbial profiles were analyzed by 16S rRNA amplicon sequencing, and respiratory viruses were identified by PCR. Clinical data of the acute episode and follow-up during the first year after infection were recorded. Pairwise comparisons showed significant differences in the gut (R² = 0.0639, P = 0.006) and NPA (R² = 0.0803, P = 0.006) microbiota between cases and controls. A significantly lower gut microbial richness and an increase in the NPA microbial diversity (mainly due to an increase in Haemophilus, Streptococcus, and Neisseria) were observed in the infants with bronchiolitis, in those with the most severe symptoms, and in those who subsequently developed recurrent wheezing episodes after discharge. In NPA, the higher microbial richness differed significantly between the control group and the non-RSV bronchiolitis group (P = 0.01) and between the control group and the RSV bronchiolitis group (P = 0.001). In the gut, the richness differed significantly between the control group and the non-RSV group (P = 0.01) and between the control group and the RSV bronchiolitis group (P = 0.001), with higher diversity in the RSV group. A distinct respiratory and intestinal microbial pattern was observed in infants with bronchiolitis compared with controls. The presence of RSV was a main factor for dysbiosis. Lower gut microbial richness and increased respiratory microbial diversity were associated with respiratory morbidity during follow-up.

Importance: Both the intestinal and respiratory microbiota of children with bronchiolitis, especially those with respiratory syncytial virus infection, are altered and differ from that of healthy children. The microbiota pattern in the acute episode could identify those children who will later have other respiratory episodes in the first year of life. Preventive measures could be adopted for this group of infants.

Keywords: gut microbiota; infants; nasopharyngeal microbiota; respiratory syncytial virus; virus.

Fig 1

Patients’ and samples’ flow chart.

Fig 2

Taxonomy and alpha and beta diversity from nasopharyngeal aspirate and gut microbiota in the patients with bronchiolitis and in the controls. (A) Shows the amplicon sequence variants shared by all the study groups. (B and C) Principal coordinates analysis (PCoA) plot based on Bray–Curtis dissimilarities of nasopharyngeal (B) and gut (C) microbiota composition of samples from all patients included in the study. P-value corresponds to the Adonis2 PERMANOVA test. (D) Boxplot showing richness and Simpson and Shannon diversity indices according to the bronchiolitis and control groups in the NPA and gut microbiota of these patients. (E and F) Relative abundance of ASVs according to bronchiolitis in NPA (E) and gut (F).

Fig 3

Statistically significant taxonomic change in bronchiolitis vs control samples in nasopharyngeal aspirate and gut. (A and D) Linear discriminant analysis (LDA) effect size in NPA (A) and gut (D). (B and E) Triplot of RDA showing the distribution of NPA (B) and gut (E) samples with reference to bacterial genera and explanatory variables. The ellipses are drawn containing 75% of each group of samples from each study group and colored accordingly. The arrows indicate the direction and strength (length) of the explanatory variables. The red names correspond to the higher abundance of each bacterial genus. (C and F) Performance of random forest (RF) classifier models. RF models were used for distinguishing bronchiolitis cases from controls in NPA samples (C) and bronchiolitis cases from controls in gut samples (F) through the representation of the mean decrease in the Gini index of each variable.

Fig 4

Effects of the co-infection variable on alpha and beta diversity and on taxonomy in the patients with bronchiolitis. (A and D) Boxplot showing richness, Simpson, and Shannon diversity indexes in NPA (A) and gut (D) microbiota with respect to the co-infection. (B and E) Principal coordinates analysis (PCoA) plot based on Bray–Curtis dissimilarities of nasopharyngeal (A) and gut (B) microbiota composition of samples from all study patients. P-value corresponds to the Adonis PERMANOVA test. (C and F) Linear discriminant analysis (LDA) effect size in NPA (A) and gut (D).

Fig 5

Effects of the respiratory syncytial virus infection on alpha and beta diversity and on taxonomy in the patients with bronchiolitis. (A and D) Alpha diversity boxplot (richness, Simpson, and Shannon diversity indexes) in NPA (A) and gut (D) microbiota. (B and E) Principal coordinates analysis (PCoA) plot based on Bray–Curtis dissimilarities in nasopharyngeal (A) and gut (B) microbiota composition of samples from all study patients. P-value corresponds to the Adonis PERMANOVA test. (C and F) Linear discriminant analysis (LDA) effect size in NPA (A) and gut (D).