Abstract
IMPACT: This study demonstrates that neonatal sepsis, blood transfusion, combined antibiotic use, asphyxia, and maternal gestational diabetes (GDM) are independent risk factors for NEC. By utilizing a 1:1 matched case-control design, we effectively eliminated the confounding effects of gestational age and birth weight, revealing critical perinatal drivers of the disease. The resulting prediction model (AUC = 0.937) provides a robust tool for early screening. These findings underscore the urgent need for enhanced antibiotic stewardship and rigorous maternal GDM management to reduce NEC incidence and improve outcomes in high-risk preterm infants. BACKGROUND: Identifying independent clinical risk factors for necrotizing enterocolitis (NEC) facilitates early screening of high-risk preterm infants and enables the development of targeted prevention strategies.This retrospective study explored the influencing factors of NEC and applied logistic regression analysis to identify the key contributors to NEC onset, aiming to provide evidence for clinical diagnosis and treatment. METHODS: A 1:1 matched case-control study was conducted, involving 172 preterm infants with NEC and 172 controls matched by GA (±1 week), BW (±100 g), and admission time. Univariate and multivariable logistic regression analyses were performed to identify independent risk factors. RESULTS: A total of 172 preterm infants with NEC (study group) and 172 matched non-NEC preterm infants (control group) were enrolled in this retrospective case-control study. Univariate analysis revealed significant differences between the two groups in preterm premature rupture of membranes (PPROM), gestational diabetes mellitus (GDM), combined antibiotic use, blood product transfusion, neonatal sepsis, and neonatal asphyxia (all P < 0.05). Multivariate logistic regression analysis identified independent risk factors for NEC in preterm infants, including neonatal sepsis (OR = 2.522, 95%CI: 1.403-4.538, P = 0.002), blood transfusion (OR = 2.18, 95%CI: 1.245-3.809, P = 0.008), neonatal asphyxia (OR = 1.887, 95%CI: 1.085-3.274, P = 0.024), GDM (OR = 1.824, 95%CI: 1.102-3.020, P = 0.018), and combined antibiotic use (OR = 1.976, 95%CI: 1.305-2.972, P = 0.001). The Hosmer-Lemeshow test confirmed good fit of the regression model (χ (2) = 13.152, P = 0.071). Receiver operating characteristic (ROC) curve analysis showed that the risk prediction model constructed with these independent factors had an area under the curve (AUC) of 0.937 (95%CI: 0.915-0.960), with a sensitivity of 93.8% and specificity of 79.5%. CONCLUSION: Sepsis, blood transfusion, combined antibiotics, asphyxia, and maternal GDM are key independent predictors of NEC. Clinicians should implement targeted monitoring and stewardship, particularly regarding antibiotic use and GDM management, to mitigate NEC risk.