The proposal responds to RFA-MH-16-510 by focusing on the domain of ?Cognitive Systems? and constructs ?cognitive control? and ?working memory? and integrating units of analysis ?brain circuit? and ?behavior?. These constructs are subsumed under executive function (EF), the ability to voluntarily constrain thoughts and actions in the service of goals. Among pediatric psychiatric categories, EF deficits define Attention Deficit Hyperactivity Disorder (ADHD) and are comorbid with a variety disorders, including Autism Spectrum Disorders, disruptive behavior disorders, mood and anxiety disorders, Tourette's/tics, and learning disabilities. Across these disorders, EF deficits limit adaptive functioning and success of behavioral intervention. Ameliorating EF deficits is a challenge, however, because current EF nosology falls short of capturing heterogeneity within and across disorders. The primary challenge then is identifying the dimensions of EF that capture the specific nature of impairment across disorders. Most past approaches utilize dimension- reducing methods that are sensitive to shared variance, but exclude unique variance. Here, we address this challenge through novel data-driven generation of behavioral profile-based EF dimensions derived from graph theory community-detection (following [1, 2]), applied to common clinical parent-report measures (ADHD Rating scale, inattention, hyperactivity/impulsivity, 8 Behavior Rating Inventory of Executive Function subdomains, Child Behavior Checklist internalizing, and externalizing). Community-detection applied to N=322 (8-13 yrs; IQ>70; no ?medical? diagnosis) presenting at Children's National Medical Center neuropsychology clinics identified three EF profiles distinguished by deficits and relative strengths: 1) poor working memory; good flexibility and inhibition; 2) poor inhibition; good working memory; 3) poor flexibility and emotion regulation; good working memory. We will recruit from this growing cohort to examine:
Aim 1 ? seek replication by testing a new larger cohort with support vector machine classification trained on preliminary data.
Aim 2 - characterize functional networks distinguishing the 3 profiles, by group comparison and dimensional analysis. Task-based functional connectivity will test hypothesis about specific circuits distinguishing the novel EF dimensions using fMRI during: 1) N-back working memory; 2) Response inhibition; and 3) Adaptive socio-emotional cognitive control. Task-free resting-state fMRI will test hypothesis about large-scale network interaction differences between EF dimensions.
Aim 3 - test the hypothesis that the novel EF dimensions are associated with specific domains of adaptive function, mediated by specific functional networks. Results will: 1) provide neurobiologically validated EF dimensions for re- conceptualizing pediatric psychiatric nosology, and 2) identify treatment targets and increase precision in measuring treatment effects ? i.e., who should receive what treatment and how to best measure response and outcome, both of which are essential to the success of a personalized approach to clinical practice.
Problems with executive function, the ability to voluntarily constrain thoughts and actions in the service of goals are ubiquitous across psychiatric childhood disorders, but heterogeneity within and across diagnostic categories make them intractable, posing a major challenge for their amelioration. Using a novel data-driven method, we will generate dimensions of executive function that cut across disorders and identify underlying neural circuitry with task-based and task-free functional magnetic resonance imaging. Results will provide neurobiologically validated behavioral dimensions necessary for revising diagnostic criteria and aiding decisions about who should receive what treatment and how to best measure response, both of which are essential to the success of a personalized approach to clinical practice.
