Dissecting human cortical similarity networks across the lifespan

Abstract

The human cortex exhibits remarkable morphometric similarity between regions; however, the form and extent of lifespan network remodeling remain unknown. Here, we show the spatiotemporal maturation of morphometric brain networks, using multimodal neuroimaging data from 33,937 healthy participants aged 0–80 years. Global architecture matures from birth to early adulthood through enhanced modularity and small worldness. Early development features cytoarchitecturally distinct remodeling: sensory cortices exhibit increased morphometric differentiation, paralimbic cortices show increased morphometric similarity, and association cortices retain stable hub roles. Morphology-function coupling peaks in early adolescence and then decreases, supporting protracted functional maturation. These growth patterns of morphometric networks are correlated with gene expression related to synaptic signaling, neurodevelopment, and metabolism. Normative models based on morphometric networks identify person-specific, connectivity-phenotypic deviations in 1,202 patients with brain disorders. These data provide a blueprint for elucidating the principle of cortical network reconfiguration and a benchmark for quantifying interindividual network variations.