To investigate the relationship between systemic inflammatory mediators and early neurodevelopmental outcomes in term infants following invasive bacterial infection. This retrospective cohort study analyzed neonates admitted to a single center from March 2016 to February 2019. Clinical data from birth to 120 days were reviewed. Fifty term infants with clinically diagnosed invasive bacterial infection (the vast majority with culture confirmation) constituted the exposure group, matched with fifty term infants without any inflammatory history. Neurodevelopment was assessed at 18–22 months using the Bayley Scales of Infant and Toddler Development-IV (BSID-IV) as the primary comprehensive tool and the Ages and Stages Questionnaire (ASQ) to derive a developmental quotient (DQ) and incorporate parent-reported measures, providing a multi-faceted evaluation. Serum levels of IL-6, TNF-α, IL-1β, MMP-2, and MMP-9 were measured during the acute infection phase. The infection group demonstrated significantly lower BSID-IV composite scores (82.19 ± 1.55 vs. 89.59 ± 1.52, P < 0.001) and ASQ-derived DQs (80.25 ± 4.85 vs. 89.14 ± 5.83, P < 0.001) compared to controls. These infants also exhibited markedly elevated levels of all inflammatory mediators (all P < 0.001), with IL-6 showing the most pronounced elevation (157.89 ± 11.17 pg/mL vs. 5.16 ± 0.58 pg/mL). Furthermore, a history of invasive infection remained a significant independent predictor of poorer developmental outcomes in multivariable analysis (β = -6.92 for BSID, P < 0.001). All inflammatory markers showed significant but weak-to-moderate inverse correlations with neurodevelopmental scores. In term infants, invasive bacterial infection and its associated acute systemic inflammatory response are strongly associated with adverse neurodevelopmental outcomes at 18–22 months, independent of key demographic confounders. Integrating early cytokine profiling with routine developmental screening may aid in the early recognition of infants at potential risk, though further research is needed to determine if these associations are modifiable.
Xu et al. (Mon,) studied this question.