Stunted childhood growth is associated with decompartmentalization of the gastrointestinal tract and overgrowth of oropharyngeal taxa

Abstract

Linear growth delay (stunting) affects roughly 155 million children under the age of 5 years worldwide. Treatment has been limited by a lack of understanding of the underlying pathophysiological mechanisms. Stunting is most likely associated with changes in the microbial community of the small intestine, a compartment vital for digestion and nutrient absorption. Efforts to better understand the pathophysiology have been hampered by difficulty of access to small intestinal fluids. Here, we describe the microbial community found in the upper gastrointestinal tract of stunted children aged 2-5 y living in sub-Saharan Africa. We studied 46 duodenal and 57 gastric samples from stunted children, as well as 404 fecal samples from stunted and nonstunted children living in Bangui, Central African Republic, and in Antananarivo, Madagascar, using 16S Illumina Amplicon sequencing and semiquantitative culture methods. The vast majority of the stunted children showed small intestinal bacterial overgrowth dominated by bacteria that normally reside in the oropharyngeal cavity. There was an overrepresentation of oral bacteria in fecal samples of stunted children, opening the way for developing noninvasive diagnostic markers. In addition, Escherichia coli/Shigella sp. and Campylobacter sp. were found to be more prevalent in stunted children, while Clostridia, well-known butyrate producers, were reduced. Our data suggest that stunting is associated with a microbiome "decompartmentalization" of the gastrointestinal tract characterized by an increased presence of oropharyngeal bacteria from the stomach to the colon, hence challenging the current view of stunting arising solely as a consequence of small intestine overstimulation through recurrent infections by enteric pathogens.

Keywords: decompartmentalization; microbiota; oropharyngeal taxa; stunting; sub-Saharan Africa.

Fig. 1.

Overall microbiota composition of gastrointestinal samples. (A) PCoA plot of the Bray–Curtis dissimilarity index of the samples rarefied to 10,000 sequences. Sample count: gastric aspirates: n = 50; duodenal aspirates: n = 46; feces: n = 343. Gastric samples are colored in red, duodenal samples in green, and fecal samples in blue. PERMANOVA test based on sample type yielded a P value of 2 × 10⁻⁴. (B) Venn-diagram depicting the number of taxa constituting the core microbiota of gastric (red), duodenal (green), and fecal (blue) samples. The core microbiota was defined as taxa with a relative abundance of at least 0.01% in each sample and a prevalence of 90% in the corresponding sample group. The number of shared core taxa between the different compartments is indicated in the overlapping regions. (C) Richness (as measured by the Chao1 index) according to nutritional status in gastric (Left, Student’s t test, P = 0.60), duodenal (Center, Student’s t test, P = 0.53), and fecal (Right, ANOVA, P = 0.70) samples. Fecal samples compared: 200 from nonstunted, 80 from moderately stunted, and 63 from severely stunted children. Duodenal samples compared: 25 from moderately stunted and 21 from severely stunted children. Gastric samples compared: 30 from moderately stunted and 27 from severely stunted children. Analysis performed on samples rarefied to 10,000 sequences. MCM, moderately stunted; MCS, severely stunted; NN, nonstunted.

Fig. 2.

Relative abundance of the 10 most-abundant genera in the duodenum as well as genera cultured in the samples. The analysis was performed on 12 samples from Madagascar and 34 samples from CAR. Data are stratified on country of origin of the duodenal samples. The color code for the different genera is given on the right. Genera highlighted in yellow are also identified by culture techniques.

Fig. 3.

Taxa with differential relative abundance in fecal samples of stunted children compared with nonstunted controls (DeSeq2 analysis). Analysis was performed on 404 fecal samples, of which 236 were nonstunted and 168 were stunted. Taxa are ordered according to log₂ fold-change. Each individual OTU affected for a given genus is indicated with a dot. Dots are colored according to their class according to the color code given on the right of the graph. Only taxa with at least log₂ fold-change and which are significantly differentially abundant (LRT, P ≤ 0.01) are indicated on the graph/in the table. Differentially abundant OTU are grouped according to genera. Genera with species mainly found in the oropharyngeal cavities are indicated with a green star. Genera, which represent enteropathogens, are indicated with a yellow star.

Fig. 4.

Schema of the decompartimentalization of the gastrointestinal tract observed in the context of stunting. (Left) Gastrointestinal compartmentalization is indicated as described in the literature for nonstunted children. (Right) The observed decompartimentalization with associated changes in stunted children are depicted.