From Breath to Brain: NICU Respiratory Interventions and Bedside Brain Signal Entropy Predict Later Autism Risk

Area of Science:

• Neuroscience
• Neonatology
• Developmental Pediatrics

Background:

• Premature infants face risks like hypoxia and inflammation, increasing susceptibility to neurodevelopmental disorders such as Autism Spectrum Disorder (ASD).
• Early identification of infants at risk for ASD is crucial for timely intervention and improved outcomes.
• Developing accurate, non-invasive methods for early ASD risk assessment in neonates is a significant clinical need.

Purpose of the Study:

• To investigate the relationship between respiratory support burden, brain signal variability, inflammation, and the risk of developing ASD in premature infants.
• To identify non-invasive biomarkers for predicting ASD risk in early infancy.

Main Methods:

• Electroencephalography (EEG) was used to measure brain signal entropy in social and nonsocial resting states.
• Saliva samples were collected to assess inflammatory markers (cytokines).
• A novel Prognostic Respiratory Intensity Scoring Metric (PRISM) was calculated to quantify respiratory support needs.

Main Results:

• Higher PRISM scores correlated with increased brain signal entropy in the nonsocial resting state, particularly in female infants, potentially mediated by cytokines.
• This association was not significant in the social resting state, especially for male infants.
• A predictive model combining nonsocial resting state brain signal entropy, sex, and PRISM scores accurately identified infants at risk for ASD (88% accuracy).

Conclusions:

• Non-invasive measures, including EEG and PRISM scores, can effectively identify premature infants at high risk for ASD.
• These findings support the potential for early ASD risk stratification before hospital discharge, enabling prompt intervention.
• Understanding sex-specific differences in brain activity and inflammation may offer insights into ASD etiology.