Abstract
Abstract
Background: Tuberculosis is a major infectious disease whose effects may extend beyond the lungs to influence microbial communities at other body sites. Although studies have reported alterations in the respiratory and gut microbiomes of people with TB, evidence from sub-Saharan Africa remains limited, and the extent to which these alterations differ between body sites is not fully understood. This study characterized sputum and stool microbiomes among Ugandan adults with pulmonary TB and those without TB.
Methods: We conducted a cross-sectional study involving 200 participants, comprising 150 adults with pulmonary TB and 50 contacts, at Mulago National Referral Hospital in Kampala, Uganda. Sputum and stool samples were profiled using 16S rRNA gene sequencing. Following quality control and contaminant removal, within-sample diversity was assessed using multiple alpha diversity indices. Beta diversity was evaluated using distance-based methods, with group differences assessed by permutational multivariate analysis of variance (PERMANOVA) and homogeneity of dispersion tests. Differentially abundant taxa were identified using ANCOM-BC2.
Results: After quality control, 135 sputum and 151 stool samples were retained for downstream analysis. In sputum, microbial richness estimators Chao1 and ACE were lower among TB cases than controls, whereas differences in overall community composition were not statistically significant after correction for multiple testing. In stool, several richness and diversity measures differed between groups, and beta-diversity analyses detected differences in community composition. However, TB cases exhibited greater within-group dispersion, and the sequencing platform contributed substantially to variation in stool microbial profiles. In multivariable analyses, the sequencing platform was the only factor independently associated with stool community composition after correction for multiple testing. Differential-abundance analyses identified body site-specific taxonomic associations with TB, including differences involving Streptococcus, Parvimonas, and Actinobacillus in sputum, and Dorea, Lachnoclostridium, and Turicibacter in stool.
Conclusion: Pulmonary TB was associated with differences in microbial richness and the abundance of specific taxa in both sputum and stool, but evidence for overall community composition differences varied by body site and was influenced by dispersion and sequencing platform. These findings highlight the importance of examining both respiratory and gut microbial communities in TB research. Longitudinal studies using standardized sequencing methods are needed to clarify how these patterns change during treatment and whether they have diagnostic or clinical relevance.