Maternal and Early Neonatal Serum Cytokine and Endothelin-1 Concentrations in Preeclampsia: A Single-Center...

Background/Objectives: Preeclampsia (PE) may be associated with maternal endothelial activation and altered early neonatal inflammation. The primary objective was to compare maternal and early neonatal serum interleukin 2 (IL2), IL6, IL8, tumor necrosis factor α (TNFα), and endothelin 1 (ET1) concentrations between PE and control mother-neonate pairs. Methods: This secondary report from a single-center observational cohort included 31 women with PE (34 neonates) and 45 control women (47 neonates). Biomarker concentrations were log-transformed, and PE-control differences were estimated using linear models, with results expressed as geometric mean ratios (GMRs). Neonatal models used pregnancy-clustered standard errors to account for twins. The Benjamini-Hochberg procedure was applied to control the false discovery rate (FDR). Additional neonatal models adjusted for gestational age at birth served as sensitivity analyses. Selected biomarker-outcome associations were explored post hoc using univariable Firth bias-reduced logistic regression. Results: Maternal ET1 was higher in PE (GMR = 1.79, 95% confidence interval-CI-1.53-2.09; q < 0.001). Neonatal IL2 (GMR = 1.37, 95% CI 1.11-1.71; q = 0.016) and TNFα (GMR = 1.49, 95% CI 1.22-1.82; q < 0.001) were higher in PE-exposed neonates. After gestational age at birth adjustment, the TNFα difference persisted (GMR = 1.33, 95% CI 1.14-1.55; q = 0.003), whereas IL2 was attenuated (GMR = 1.20, 95% CI 0.99-1.45; q = 0.164). None of the exploratory Firth associations remained significant after FDR correction. Thrombocytopenia at birth occurred in 5/33 PE-exposed and 1/47 control neonates; intraventricular hemorrhage occurred in 2/34 and 2/47, respectively. Conclusions: PE was associated with between-group differences in maternal endothelial and early neonatal inflammatory biomarkers, with maternal ET1 and neonatal TNFα as the most robust signals. Complication-related findings remain hypothesis-generating and do not support predictive cut-offs, risk stratification, or causal inference.