Phenotype-Specific Associations Between Cortical Age-Related Trajectory Features and the Sensorimotor–Association Axis from Age 6 to 15

Background/Objectives: Cortical development in children and adolescents shows marked regional heterogeneity, but whether age-related trajectory properties are organized consistently along an externally defined sensorimotor–association (S–A) axis across morphometric phenotypes remains unclear. This study aimed to characterize relative net change, trajectory nonlinearity, and timing centroids across cortical phenotypes and determine how these properties relate to the sensorimotor–association (S–A) axis. Methods: This cross-sectional study included structural MRI data from 936 screened clinical participants aged 6–18 years. Normative age trajectories were modeled for 68 Desikan–Killiany cortical regions using cortical thickness (CT), surface area (SA), cortical volume (CV), and Morphometric INverse Divergence (MIND)-derived nodal degree. Scanner was included as a fixed effect in the model. Relative change (RC), nonlinearity index (NLI), and timing centroid (TC) were extracted from sex-averaged conditional median trajectories within the 6–15-year analysis window. Associations with the S–A axis were tested using Spearman correlations and 10,000 hemisphere-preserving spatial spin permutations, with a two-sided per-test p < 0.05 criterion. Participant-level bootstrap resampling quantified estimation uncertainty. Results: In fitted trajectories, CT, CV, and MIND nodal degree showed net decreases in most cortical regions, whereas SA showed predominantly net increases. Regional distributions of NLI and TC also differed across phenotypes. Of 12 predefined S–A association tests, seven met the two-sided Spin criterion. Toward the association-cortex end, TC increased for CT, CV, and MIND nodal degree, RC increased for SA and MIND nodal degree, and NLI increased for SA but decreased for MIND nodal degree. Conclusions: Cortical age-related trajectories show phenotype- and property-specific spatial associations with the external S–A axis. Concordant gradients of relative net change can coexist with opposite gradients of trajectory nonlinearity, indicating that a common cortical hierarchy is expressed through different age-related properties across cortical phenotypes.