A Cross-Lagged Panel Analysis of Cortisol Levels and Internalizing Behaviors in Children Born Very Preterm Across Early Childhood: Associations Differ for Boys and Girls at Age 1.5 Years

Abstract

Children born very preterm (≤32 weeks’ gestation) are exposed to considerable stress in the neonatal period that, in turn, is associated long-term with altered physiological stress reactivity and regulation, as well as increased internalizing (anxiety and depressive) behaviors. Whether cortisol levels are related to evolving internalizing behaviors in this population has not been evaluated to our knowledge. The present study investigated the association between cortisol reactivity to a cognitive assessment in a novel clinic environment and parent-reported internalizing behaviors both concurrently and across ages in children born very preterm and examined whether relationships differed by biological sex at birth. Total cortisol output (AUCg) and reactivity (AUCi) were calculated from saliva assayed across age-appropriate cognitive tasks, and parents reported on their child's behavior at ages 1.5, 3, and 4.5 years. Valid cortisol data at one or more assessment points were available from 174 participants. Random-intercept cross-lagged panel models showed no longitudinal relationships between internalizing behaviors and cortisol output (AUCg, AUCi). Follow-up multilevel models revealed that the relationship between cortisol AUCg and internalizing behaviors was specific to girls at age 1.5 years. Findings highlight the importance of examining sex differences in biobehavioral relationships across development. Future research should consider factors that may attenuate these relationships across development.