Infant feeding practices and the gut microbiome: impact on infant growth and development in South Africa

Abstract

In Africa, undernutrition is seen in approximately 45% of morbidity associated with the under five years' mortality rate. Children who are undernourished are at risk of dying, delayed or stunted growth and impaired brain development. Even if interventions are created, the negative effects of early poor nutrition seem to persist into adulthood, and warrant further research. There is strong support that the influence of microbial communities within the infant gut (gut microbiome) is significant in the utilisation of food nutrients for optimal growth. Alterations in the gut microbiome (dysbiosis) are associated with age, feeding practices and maternal host genetics. Maternal genotypes, specifically α(1,2)-fucosyltransferase 2 (FUT2) rs601338 genotypes that determine secretor status, have been shown to alter breast milk composition. Little is known about these associations, the role of these variants, and their influence on the infant gut microbiome in non Western populations. The aim of this pilot study was three-fold: (1) to characterize the gut microbiome profile of South African infants living in an urban environment (Soweto) over time; (2) to assess whether the infant gut microbiome profiles are associated with exclusive breastfeeding, mixed feeding, and formula feeding practices; and, (3) to investigate the influence of maternal FUT2 genotypes (secretor status) on the infant gut microbiome. We recruited 46 mother-infant pairs via a collaboration with Wits Developmental Pathways for Health Research Unit. We then collected biological samples (infant stool sample and maternal saliva) from participants following informed consent. Infant stool samples were collected “at recruitment” (12-21 weeks/Time0), six months (Time6), nine months (Time9), and twelve months (Time12). We performed genotyping (TaqMan assay) of the FUT2 gene mutation using DNA extracted from maternal saliva samples. We generated 16S rRNA sequencing data (via the sequencing facility at the National Institute for Communicable Diseases, Sandringham, Johannesburg) using DNA extracted from infant stool samples. Metagenomic analysis was performed in order to characterise the diversity and composition of microbiota present at different time points in infant stool samples using the QIIME2 pipeline, R and LEfSe. Association between infant gut taxonomic profiles at given time points (ages of the infants), feeding practice and maternal secretor status were investigated. A total of 46 maternal saliva and 125 infant stool samples were collected. In this study, 83% of mothers participated in some form of breastfeeding – either exclusively or in combination with formula. Alpha diversity measures of Faith’s Phylogenetic Diversity revealed significant differences were specifically for Time12: versus Time0 (P/q-value=0.44x10-3 /1.00x10-3 ); versus Time6 (P/q value=0.01/0.02); and versus Time9 (P/q-value=0.31x10-3 /1.00x10-3 ). Pielou’s Evenness revealed significant differences were specifically for the Time0: versus Time6 (P/q-value=0.01/0.03); and VI versus Time9 (P/q-value=2.00x10-3 /0.02). There was also a higher diversity index in male samples compared to females for both alpha diversity measures. There were no distinct clusters when beta diversity was assessed using principle coordinates analysis. Beta diversity measures computed using PERMANOVA revealed statistically significant values for feeding practice (P=0.003) and time point analyses (P=0.001). Taxonomic analysis revealed that the greatest biodiversity is observed in exclusively breastfed infants and the least, in formula fed infants. Additionally, the ratio of the composition of the gut microbiome can be affected by the administration of medication at recruitment and nine months. This is between stages where feeding practices change and, the gut microbiome is dynamic and increases in diversity. Bacteria from the phylum Firmicutes make up the majority composition of the infant gut microbiome, and Bifidobacteria from the phylum Actinobacteria is uniquely associated with increased abundance in breast milk-fed infant microbiomes. Genotyping results revealed 79.1% of mothers were secretors, and 20.9% were non secretors. Secretor status was not significantly associated with gut microbiome diversity - potentially due to the small sample size that excluded formula fed infants. We were successful in using 16S rRNA sequencing and bioinformatic analysis to characterise the gut microbiome of clinically healthy infants from urban Soweto, South Africa. Future studies with a larger sample size in different geographical regions representing different communities within South Africa, would generate the necessary statistical power and enable the generalisability of these results to the South African population. The inclusion of infants adequately representing malnourished or stunted phenotypes, is necessary for a greater understanding of the relationship of this phenotype to the gut microbiome profile. This could address how the state of dysbiosis in those infants relates to these phenotypes, and would further describe the contributing factors that result in these poor health outcomes