Neurobehavioral Profiles in Term, Healthy Infants: Examining Associations with Maternal Adversity, Perinatal Events, and Neonatal Outcomes

Background: The NICU Network Neurobehavioral Scale (NNNS) assesses neonatal neurobehavioral performance through examinations and observations. The tool was initially developed for use in drug-exposed or preterm infants. This study aims to examine NNNS scores in a sample of healthy, term (>=37 weeks) infants and assess the association between NNNS scores and maternal adversity, perinatal and delivery characteristics, and infant brain volumes.

Methods: A prospective, longitudinal study recruited pregnant mothers in their first trimester between 2017 and 2020. Their infants underwent an NNNS assessment between two to five weeks after birth. The NNNS consists of 115-items clustered into thirteen summary scores based on measures of alertness, activity, tone, reflexes, responses to stimuli and signs of stress/abstinence. Using latent profile analysis (LPA), infants were categorized into phenotypic profiles based on their distribution of scores. Maternal demographic and prenatal clinical characteristics, early life adversity (measured by social advantage and psychosocial stress scores derived from structural equations modeling), and perinatal events were prospectively collected. Infants underwent high-quality brain MRI scans the same day as the NNNS assessment. Total and regional brain volumes were measured on T2-weighted MRI data using an automated segmentation algorithm. Differences in variables between profiles were assessed by ANOVA/Kruskal-Wallis test for continuous variables, Chi-Square/Fisher exact test for categorical variables, or multivariate models for brain volumes.

Results: After applying exclusion criteria, 296 mother/infant pairs were included. Infant gestational age ranged from 37.0-41.6 weeks, and postmenstrual age at neonatal assessment ranged from 39.0-44.9 weeks. LPA generated three distinct profiles, labeled ‘hypotonic’ (49 infants, 16.6%), ‘fussy’ (113 infants, 38.2%), and ‘regulated’ (134 infants, 45.3%). Infants in the fussy profile had extreme values on individual summary scores, including lowest scores on attention, self-regulation, and lethargy, with highest scores on handling, arousal, and excitability. These infants had younger gestational age (p=0.039), smaller birthweight (p=0.021), and smaller head circumference (p=0.002) compared to infants in the regulated profile. Infants in the fussy profile had smaller total brain volume (p=0.028) and cerebral white matter volume (p=0.015) compared to infants in the hypotonic profile, with no difference in cortical gray matter volume (p=0.093). Infants in the hypotonic profile were more likely to have prenatal marijuana exposure (p=0.039). No differences were found between profiles in maternal demographics, social advantage or psychosocial stress during pregnancy, or perinatal/delivery characteristics.

Conclusion: In a sample of healthy, term infants, atypical NNNS profiles were associated with prenatal marijuana use, younger gestational age, and smaller birth measures of weight, head circumference, total absolute brain volume, and cerebral white matter volume. Failing to confirm our hypotheses, no measures of maternal adversity or prenatal events were different between profiles. Long-term follow-up of these infants is essential to determine if identified profiles as newborns are associated with later outcomes in infancy and childhood.