Anxiety Symptoms in Preschool Children Born Very Preterm: Associations with Cognition and Neonatal Striatal Volumes

Background/Objectives: Children born very preterm (VP) are at increased risk of later emotional and cognitive difficulties, including anxiety symptoms during childhood. Altered early brain development, particularly within subcortical circuits involved in emotional regulation, may contribute to this vulnerability. This study aimed to assess anxiety symptoms in preschool-aged children born VP, examine their relationship with cognitive performance, and determine whether neonatal brain volumes at term-equivalent age (TEA) were associated with later anxiety symptoms. We also explored whether cognition mediated the association between neonatal striatal volumes and anxiety. Methods: We conducted a longitudinal cohort study of infants born at ≤32 weeks of gestation and/or with a birth weight ≤1500 g admitted to a tertiary neonatal intensive care unit between 2018 and 2021. At 4-6 years of age, anxiety symptoms were assessed using the Child Behavior Checklist (CBCL) Anxiety Problems subscale, and cognitive performance was evaluated with the Wechsler Preschool and Primary Scale of Intelligence-IV (WPPSI-IV). Neonatal magnetic resonance imaging performed at TEA was used to obtain regional brain volumetric measures. Associations were analyzed using adjusted linear regression, interaction-based volumetric modeling, path analysis, and receiver operating characteristic analysis. Results: Ninety-five children were included, and 24 (25.3%) showed clinically relevant anxiety symptoms according to the CBCL Anxiety Problems subscale. Higher WPPSI-IV scores were associated with lower anxiety scores (β = -0.183; p = 0.042). The best-fitting MRI model included caudate volume, putamen volume, and their interaction, with a significant association between the putamen-caudate interaction and anxiety symptoms (β = -17.807; p < 0.001). In the path model, both cognition and the putamen-caudate interaction were directly associated with anxiety, whereas the indirect effect through cognition was not significant. The final MRI model showed acceptable discrimination for clinically relevant anxiety (AUC = 0.796). Conclusions: Anxiety symptoms were frequent in preschool-aged children born VP and were associated with lower cognitive performance. Neonatal striatal volumetric organization, particularly the interaction between the putamen and caudate volumes, was independently associated with later anxiety symptoms, suggesting that cognitive and early neural factors may contribute to anxiety risk through parallel rather than mediated pathways.