Introduction: Geographical variation is an important factor contributing to differences in gut microbiome composition as different regions have distinct lifestyles. Breastfeeding practices and dietary preferences are some of the most important influences in the early life gut microbiome composition. This is correlated with the physical health and incidence of illness in infants.

Objectives: To study differences in microbiota composition in healthy infants belonging to different geographical regions

Methods: Open-access 16S rRNA sequencing taxonomic data of healthy infants (based on anthropometry and reported illnesses) from various studies was obtained from Microbiome DB and the AFRIBIOTA consortium. The final curated dataset includes 302 participants from 4 studies conducted across 4 regions (USA, South America, Asia and Africa). Statistical analyses were performed using R. Relative abundance, alpha (Shannon index) and beta diversity (Bray-Curtis distance) metrics were estimated. Statistical significance in beta diversity was studied using principal coordinate analysis (PCoA) and PERMANOVA. Linear mixed effects model (MaAsLin2 package) was used to test differences in alpha diversity and relative abundances between regions, with age and region as fixed effects, and study as random effect.

Results: The analysis included 302 children across the four regions, with a median age of 7 months. Phylum Actinobacteriota (particularly genus Bifidobacterium) had overwhelming preponderance in South America. African subjects had higher relative abundance of genus Prevotella (16.6%) while USA had higher abundance of Bacteroides, both belonging to the phylum Bacteroidetes.

Africa and South America showed distinct clustering in principal coordinate analysis using Bray-Curtis distance as the measure, while PERMANOVA model, adjusted for age and study effects showed that 11.7% of the variation was explained by region (p value =0.001).

Linear mixed effects models showed significantly higher relative abundance of Verrucomicrobiota in the USA when compared to Africa (Coeff=2.82, FDR<0.0001). Campilobacterota was lower in South

Asia (Coeff =-2.24, FDR<0.005), and South America (Coeff =-2.93, FDR<0.005) when compared to Africa.

Discussion: Increased abundance of certain genera may correspond to the dietary practices in the region. Genus Prevotella has been shown to be associated with plant-based diets, which are a staple in sub-Saharan Africa. Considering that breastmilk and other milk products/substitutes would dominate the diet in infants, all regions show a higher relative abundance of lactic acid bacteria (particularly genera Bifidobacterium and Lactobacillus), but South America had a significantly higher abundance, possibly indicative of a higher preference for milk-based diets. Breastmilk composition which is in turn is shaped by maternal diet, amongst other factors, and diet of infants seem to be major influences. Further, certain phyla comprise of disease-causing pathogens, which had higher relative abundance in Africa. Studying these variations across regions, which begin at a very early stage in life, is essential.

Conclusion: Relative abundance, alpha diversity and beta diversity show that significant differences are present across regions. Dietary practices and disease states maybe key contributors. Further analyses are needed to comprehensively look at other important factors contributing to the differences, and robust curated databases including both compositional data and associated factors are required.