Functional brain connectivity from the fetal period through infancy: A narrative review of resting-state fMRI studies.

The interval from the fetal period through infancy is critical for the formation and reorganization of large-scale functional brain networks. Resting-state functional magnetic resonance imaging (rs-fMRI) requires no task performance and provides a means of characterizing whole-brain functional organization during this interval. This narrative review synthesizes human rs-fMRI evidence on typical functional network development, prematurity-related differences, associations with maternal and early caregiving environments, and links between early functional connectivity and subsequent neurodevelopmental outcomes. Overall, early functional network development shows a degree of hierarchical organization: primary sensorimotor systems show relatively stable organization earlier, whereas higher-order association systems undergo more prolonged refinement. Local specialization and cross-network integration overlap, with patterns varying by connection distance, brain region, and individual. We further examine associations of prematurity, maternal psychological and social environments, metabolic and immune status, substance and environmental chemical exposures, neonatal intensive care unit (NICU) experiences, and early sensory and social interactions with functional network differences. We also summarize prospective evidence relating fetal, neonatal, and infant connectivity features to later motor, language, cognitive, and socioemotional outcomes, as well as autism spectrum disorder-related risk measures. These findings suggest that early functional connectivity may contain information about developmental vulnerability before behavioral or clinical features become apparent. Finally, we discuss methodological challenges related to scan state (sleep or wakefulness), head motion, brain parcellation, and analytical pipelines, and consider how multisite longitudinal studies, standardized data acquisition, and independent-cohort validation could improve reproducibility and support future clinical translation.