This is a replication dataset for the submitted manuscript titled: "Plasma lipopolysaccharide levels, microbiota and biomarkers of enteric dysfunction predict mortality in acutely ill children in sub-Saharan Africa and South Asia.
Sub-Saharan Africa and South Asia account for a disproportionately high share of global under-five mortality. Acute infections, including those caused by gram-negative bacteria, are common contributors to childhood hospitalization in these regions. Malnutrition and presumed environmental enteric dysfunction (EED) are often associated with these circumstances. Lipopolysaccharide (LPS), a component of the outer membrane of gram-negative bacteria, triggers host immune responses by engaging with toll-like receptor 4 (TLR4), leading to the release of inflammatory cytokines such as TNF and IL-1β. While hyper-activation of this pathway can result in conditions like sepsis, chronic exposure to LPS may promote immune tolerance. Additionally, LPS-TLR4 interactions play a crucial role in maintaining gut homeostasis.
Although LPS is used as a biomarker for microbial translocation in conditions like EED, its association with mortality in low-resource settings has yet to be fully explored. Previous research has examined the relationship between systemic LPS levels and mortality in diseases such as sepsis and HIV, but studies focusing on environmental EED are scarce. Furthermore, chronic inflammation—a hallmark of EED—remains poorly understood in this setting.
This analysis of LPS was conducted using demographic, clinical, anthropometric, outcome, proteomic, and entero-pathogen data from the CHAIN nested case-cohort (NCC) study to investigate the link between systemic LPS and mortality in a context marked by prevalent malnutrition and presumed EED. The study included 889 children across nine CHAIN sites. To address selection bias and enhance generalizability, inverse proportional weighting was applied where appropriate. The analysis utilized various methods, including group comparison statistics, survival analyses, correlation matrices, regression models, pathway analysis, and deconvolution of single-cell transcriptomic datasets.
