Biomarker Dynamics and Associations with Organ Dysfunction in Two Independent Neonatal and Adult Sepsis Cohorts: A Retrospective Two-Center Study

Background and Objectives: Sepsis-associated organ dysfunction is accompanied by complex inflammatory, metabolic, hematological, and renal alterations. However, the relationships between routinely available biomarkers and organ dysfunction remain incompletely characterized across different clinical populations. This study aimed to investigate baseline and longitudinal biomarker–organ dysfunction associations in two independent neonatal and adult sepsis cohorts recruited from different hospitals. Cross-cohort observations were intended to be descriptive, without attributing differences to age-related biological mechanisms. Materials and Methods: This exploratory, retrospective, two-center observational study included 120 neonates and 120 adults with sepsis who were consecutively identified between January 2021 and December 2025. Organ dysfunction was assessed using the neonatal Sequential Organ Failure Assessment (nSOFA) score in neonates and the Sequential Organ Failure Assessment (SOFA) score in adults. Baseline and longitudinal associations involving C-reactive protein, lactate, leukocyte subsets, neutrophil-to-lymphocyte ratio (NLR), platelet count, and creatinine were evaluated separately within each cohort using unadjusted Spearman’s rank correlations. Longitudinal changes were assessed from baseline to day 3 in neonates and from admission to 48 h in adults. Results: In unadjusted analyses, baseline creatinine was positively associated with organ dysfunction in neonates and adults, while platelet counts were inversely associated with nSOFA and SOFA. After adjustment for gestational and postnatal age, neonatal creatinine and lactate remained associated with nSOFA, whereas platelet count did not. After adjustment for CKD, adult creatinine and NLR remained associated with SOFA, whereas platelet count did not. In longitudinal analyses adjusted for baseline severity and baseline biomarker value, only the neonatal Δlactate–ΔnSOFA association remained below the FDR threshold in the adjusted longitudinal analyses. Conclusions: Several principal baseline associations persisted after targeted covariate adjustment, but the platelet associations did not. Only the neonatal lactate–nSOFA association remained statistically significant in adjusted longitudinal analyses. The remaining reported associations are unadjusted and should not be interpreted as causal or as evidence of independent prognostic value.