Cortical surface area, thalamic nuclei volume, and associations with neurocognitive functioning in 6-13-year-old children born late preterm and full term

Background: Late preterm (LP) children (34 0/7 - 36 6/7 weeks of gestation) are at risk for neurocognitive challenges, yet the neuroanatomic and functional disruptions underlying these developmental differences remain poorly understood. Purpose or Hypothesis: Investigate (1) regional differences in cortical surface area and thalamic nuclei volume and (2) structural variations that may explain heterogeneity in neurocognitive functioning in LP versus full term (FT) school-aged children, using two complementary approaches: a vertex-wise analysis of cortical surface area and a region-of-interest volumetric analysis of thalamic nuclei. Study Type: Secondary analysis of a prospective observational cohort study. Population or Subjects: 48 LP and 72 FT children aged 6-13 years. Field Strength/Sequence: 3T; Sagittal T1-weighted MP-RAGE and axial T2-weighted 3D SPACE. Assessment: FreeSurfer 7.4.1 was used to parcellate cortical regions (Destrieux atlas) and segment thalamic nuclei (25 nuclei). Neurocognitive functioning was assessed across four domains: hyperactivity/inattention (Pediatric Behavior Scale), processing speed (WISC-IV), working memory (WISC-III), and visual-spatial perception (Benton JLO). Statistical Tests: Independent samples t-tests for group comparisons; vertex-wise GLMs for cortical surface analyses with false discovery rate (FDR) correction; and multiple linear regression for thalamic nuclei analyses with FDR-corrected q < 0.05. Results: Vertex-wise analyses revealed regional differences in cortical surface area between LP and FT children in both hemispheres, along with associations between surface area in several cortical regions and hyperactivity/inattention, working memory, and visual-spatial perception. In contrast, using an ROI approach, thalamic nuclei volumes did not differ between groups and were not associated with any neurocognitive domains. Conclusion: LP birth may be associated with subtle, region-specific bidirectional left and right hemispheric cortical surface area differences rather than widespread structural alterations in middle childhood, with regional macrostructural measures associated with individual differences in neurocognitive functioning.