Brain Wide Associations-Polyneuro Risk Scores to characterize heterogeneity in ADHD - Project Summary/Abstract Attention-Deficit/Hyperactivity Disorder (ADHD) is a complex neurodevelopmental condition with deficits in cognitive control, emotional regulation, and attention. Although ADHD is widely recognized as a disorder of self-regulation across interconnected domains, a consistent mechanistic explanation remains elusive. This gap exists partly because research has often focused too narrowly on specific brain regions instead of looking at how the whole brain works together, even as growing evidence suggests that psychiatric symptoms usually arise from distributed dysfunction across the brain. Recent large-scale studies support this broader view, but reproducible results require harmonized methods and large, well-characterized datasets, resources that have only recently become available. This project seeks to test whether brain-behavior associations relevant to ADHD and its cognitive and emotional dimensions are best modeled using globally distributed network profiles. The central hypothesis is that ADHD arises from deficits across multiple domains—such as general ability, executive function, and emotion regulation—each supported by distributed, rather than focal, patterns of brain network dysfunction. This framework mirrors polygenic models in genomics, where aggregate effects from numerous loci better explain complex traits than individual variants. The primary goal is to derive and validate PolyNeuro Risk Scores (PNRS) for General Ability (GA), Executive Function (EF), and Emotion Dysregulation (ED) using functional connectivity data from large developmental cohorts. The discovery dataset is an ADHD-enriched, deeply phenotyped cohort (Oregon ADHD-1000). Validation will leverage large, publicly available neurodevelopmental samples: the Adolescent Brain Cognitive Development Study, the Healthy Brain Network, and the Human Connectome Project in Development. This project addresses critical gaps in ADHD research by integrating harmonized imaging, theory-driven constructs, and meta-matching to project individualized PNRS across cohorts. The four aims are: (1) Derive and share PNRS for GA, EF, and ED using harmonized, regularized modeling; (2) Test whether brain-behavior associations for ED are focal or globally distributed; (3) Assess whether combining PNRS across domains improves prediction of ADHD symptoms; and (4) Identify PNRS-based subgroups of ADHD and validate them using behavioral and clinical data. In the applicants' opinion, this work supports NIMH priorities by advancing interpretable, reproducible neural markers of psychiatric risk and heterogeneity. Public release of all modeling pipelines, harmonized scores, and documentation will support broad adoption and accelerate progress in developmental neuroscience and precision psychiatry.
