Longitudinal Fecal Short-Chain Fatty Acid Trajectories in Very Preterm Infants with Early-Onset Neonatal Sepsis: A Pilot Study

Background: Early-onset neonatal sepsis (EONS), defined as systemic infection occurring within the first 72 hours of life, remains a major cause of morbidity and mortality in preterm infants. Increasing evidence indicates that the gut may play an active role in systemic inflammation, yet the temporal behaviour of fecal short-chain fatty acids (SCFAs) during EONS has not been characterised. SCFAs and branched-chain fatty acids (BCFAs) are key microbial metabolites involved in epithelial maturation and immune regulation and may provide a non-invasive window into early inflammatory vulnerability. Methods: This pilot prospective longitudinal cohort study enrolled preterm infants (⩽ 32 weeks’ gestation) originally identified as at high risk for necrotizing enterocolitis (NEC) and subsequently stratified into EONS and non-sepsis groups. Serial stool samples were collected at predefined timepoints (TPs; TP1 ≈ 3 days of life [DoL], TP2 ≈ 7 DoL, TP3 ≈ 14 DoL, TP4 ≈ 21 DoL, and TP5 ≈ 28 DoL). Samples were analysed using gas chromatography–mass spectrometry (GC–MS) to quantify a panel of 12 SCFAs, including BCFAs and medium-chain fatty acids (MCFAs). Both absolute concentrations and relative fractions were evaluated, with emphasis on ratio-based metrics (e.g., acetic/propionic acid ratio) and timepoint-specific group contrasts, complemented by partial least squares discriminant analysis (PLS–DA). Results: At the earliest sampling window (TP1), infants with EONS exhibited a broad redistribution of the fecal SCFA pool. The median relative fraction of acetic acid was significantly lower in EONS (86.6% vs. 94.5% in non-sepsis), while several non-acetate components—including propionic, valeric, and branched-chain acids—were relatively enriched. Acetate-to-non-acetate ratios were markedly reduced in EONS (e.g., acetic/propionic and acetic/isobutyric ratios), indicating an early shift away from acetate dominance. PLS–DA at TP1 demonstrated partial separation between groups, with acetic-acid depletion and non-acetate enrichment among the strongest contributors to discrimination. By later TPs, these early differences narrowed to a small subset of BCFA-related ratios and largely attenuated by the end of the first month. Conclusions: In this pilot cohort of preterm infants, EONS was associated with early, structured alterations in fecal SCFA profiles, characterised by reduced acetic-acid dominance and relative enrichment of non-acetate acids. Dynamic, ratio-based assessment proved more informative than absolute concentrations alone, revealing transient intestinal metabolic signatures accompanying systemic infection. These findings provide the first longitudinal evidence of gut metabolic involvement in EONS and lay the groundwork for larger, multi-centre studies integrating SCFA trajectories with microbiome and immune profiling to refine early risk stratification for systemic infection in high-risk neonatal populations.