Normative modeling of brain development has gained traction for quantifying individual deviations in maturation. The brain age gap (BAG), the difference between predicted age from MRI features and chronological age, offers a potential individualized normative metric of neurodevelopment. However, consistent patterns across psychiatric disorders remain elusive, and no studies have examined whether BAG can predict developmental trajectories within an inclusive continuous model of youth’s cognition and behavior. Using longitudinal data from the Adolescent Brain Cognitive Development Study (ages 9-15, n=9,074), we built 8 region-specific brain age models using volumes, thicknesses, and surface areas of parcels from the Brainnetome adolescent atlas. We derived psychiatric diagnoses from a parental questionnaire. Multivariate linear regression was used to assess case-control differences and cross-sectional continuous cognitive/behavioral profiles. We modeled cognitive/behavioral trajectories using a multivariate joint latent-class mixed model and assessed the relationship with BAG values using multinomial logistic regression. Children with ADHD showed negative BAG across multiple regions (Cohen’s d : -0.12 to -0.08), while subcortical BAG emerged as a transdiagnostic indicator of younger-appearing brains ( d : -0.07, p fdr = 0.024). Older-appearing brains were associated with a high-cognition, low-symptom profile, whereas younger-appearing brains were associated with a low-cognition profile. Three developmental trajectories were identified: stable, towards externalizing behaviors, or internalizing behaviors. Widespread positive BAG predicted evolution towards internalizing behaviors, but reduced the likelihood of the externalizing trajectory. Integrating BAG with continuous cognitive and behavioral profiles yielded a plausible framework for early identification of atypical trajectories, potentially contributing to personalized medicine in psychiatry.
Gagnon et al. (Sun,) studied this question.