407 - Neonatal Correlates of Autism Diagnosis in a High-Risk Infant Follow Up Program Partnered with an Autism Center of Excellence

Poster Number: 407
Publication Number: 407.380

Background: Infants who require NICU admission are at high-risk for developmental challenges and are diagnosed with autism spectrum disorder (ASD) at significantly higher rates than in the general population. Early identification of ASD results in improved outcomes, but few specific clinical characteristics beyond prematurity have been identified to allow for targeted surveillance in high-risk infant follow-up (HRIF) programs.

Objective: To describe early clinical characteristics, associations with developmental disorders, and ASD severity in a HRIF.

Design/Methods: This retrospective cohort study included all infants with confirmed diagnoses of ASD in our regional HRIF (7/2021-8/2023). Data collected included gestational age at birth (EGA), bronchopulmonary dysplasia (BPD, < 32 weeks EGA & requiring >28 days of oxygen), severe brain insults (intraventricular hemorrhage Grade III-IV, periventricular leukomalacia, moderate-severe hypoxic ischemic encephalopathy), necrotizing enterocolitis, and prolonged NICU sedation (>14 days of continuous or scheduled opioids); severity of ASD (restrictive/repetitive behaviors and social/communication), and presence of other developmental disorders or delays. Results are presented using group level statistics and ordinal regression models to examine associations between perinatal variables and severity of ASD features, controlling for EGA.

Results: Infant characteristics are found in Table 1; 92 infants diagnosed with ASD were included in this study with 88% born preterm. Mean EGA was 29.46 weeks and mean post menstrual age (PMA) at ASD diagnosis was 37 months. Frequent neonatal comorbidities included severe brain injury (29.3%), BPD (57%) and prolonged sedation (29.3%). Other developmental disorders, such as Cerebral Palsy and vision or hearing impairments, frequently co-existed with ASD (18.5% and 32.6% respectively). Presence of BPD, but not severe brain insult (or type), was associated with increased level of ASD severity in social communication (R=0.236, p=0.04). BPD and sedation (but not severe brain injury or type) were together associated with increased level of ASD severity in restrictive/repetitive behaviors (R=0.294, p=0.046). Adjusting for EGA did not improve predictive models

Conclusion(s): In our HRIF population of children with ASD, the presence of BPD was most associated with severity of ASD features, beyond degree of prematurity or brain injury. Further studies will need to confirm to what extent BPD contributes to evolution of ASD, and if targeted early interventions can mitigate impacts on development of contingent social communication.