Abstract
BACKGROUND: Current evidence shows that systemic dexamethasone administration starting after the first week of age reduces bronchopulmonary dysplasia for very preterm (VPT) infants, but its neurological effects remain obscure. Using resting-state functional magnetic resonance imaging (rs-fMRI), we assessed the changes in functional network connectivity (FNC) in very preterm infants treated with late systemic dexamethasone (≥7 days of age). METHODS: VPT infants (GA ≤ 32 weeks) who needed to rely on mechanical ventilation for more than 7 days but fewer than 14 days to maintain vital signs were included in the study. The cohort was divided into two groups according to whether they were given systemic dexamethasone. In addition, 26 healthy term infants were recruited as controls. At term-equivalent age (TEA), rs-fMRI and 3D-T1 data from eligible infants were acquired with a 3.0-T MRI scanner. After the MRI data were preprocessed, group-level independent component analysis (ICA), a technique used for blind source separation, was used to identify the components of resting-state networks (RSNs). Then, the functional connectivity between components and RSNs was compared among different groups. Upon follow-up at 3 months of corrected age, the neurodevelopmental outcomes of enrolled infants were assessed with the Bayley Scales of Infant Development-Chinese Revision (BSID-CR), and the Motor Development Index (MDI) and Psychomotor Development Index (PDI) were measured. Finally, the correlations between resting-state FNC and BSID scores were analysed. RESULTS: Ultimately, 59 infants were included in the final analysis, including 19 preterm infants who received dexamethasone, 20 who did not, and 20 healthy term infants as controls. Based on their data, 11 components were identified, belonging to 5 RSNs: the visual network (VN), the dorsal attention network (DAN), the auditory network (AN), the primary sensorimotor network (SMN), and the default-mode network (DMN). Compared with the term infants, the preterm infants showed significantly weakened functional connectivity between the DAN and VN, as well as the VN and AN (P < 0.05). Among preterm infants, those who were given dexamethasone showed significantly stronger functional connectivity between the DAN and VN, as well as the DMN and AN (P < 0.05), than those who were not. The correlation analysis demonstrated that the connectivity values between the DAN and VN and between the VN and AN were positively correlated with the MDI (r = 0.432, P＜0.001, and r = 0.479, P＜0.001, respectively) and the PDI (r = 0.436, P＜0.001 and r = 0.516, P＜0.001, respectively). CONCLUSIONS: Our investigation uncovers a noteworthy link between the administration of late systemic dexamethasone (≥7 days of age) in VPT infants and distinct improvements in FNC. Furthermore, the observed positive correlation between inter-network connectivity and scores on the BSID-CR implies a plausible neuroprotective aspect of this therapeutic approach in this specific group of children.