General Neural Correlates of Executive Impairment in Childhood Psychopathology
Poster No:
666
Submission Type:
Abstract Submission
Authors:
Adam Kaminski1, Hua Xie2, Brylee Hawkins1, Laura Campos2, Madison Berl2, Lauren Kenworthy2, Chandan Vaidya1
Institutions:
1Georgetown University, Washington, DC, 2Children's National Hospital, Washington, DC
First Author:
Co-Author(s):
Introduction:
Childhood psychopathology is a worsening public health crisis leading to negative life outcomes, including self-harm and suicide. A latent general factor of psychopathology, termed "p-factor", has gained traction given high rates of comorbidity and shared variance in symptoms, and there is debate about its functional underpinnings. Difficulty in controlling impulsivity and distractibility, termed executive control, as early as 3 years old predicts p-factor (Moffitt et al., 2011), which has led to the hypothesis that it reflects executive impairment. Here, we test this hypothesis by predicting that p-factor is related to dysfunction in functional connectivity (FC) network integration during the execution of demanding tasks, a theorized general feature of executive control (Menon & D'Esposito, 2022). We tested this prediction with a two-pronged approach, first identifying "hub" regions defined by high FC network integration across 3 executive control tasks and testing the strength of their between-network connections; and second, identifying regions most predictive of task performance defined by connectome-based predictive modeling (CPM) and testing their FC network integration. This strategy enabled us to first test association to p-factor of network integration broadly, and then to focus on regions specifically related to task behavior.
Methods:
We included 204 children [53 F/149 M/2 NC; mean age (SD)=11 years (1.7)] with varied diagnoses (e.g., attention deficit disorder [n=80]; autism spectrum disorders [n=91]). Principal component analysis on parent-reported Child Behavior Checklist was used to define p-factor. For participants with high quality fMRI data on 3 tasks (n=79), tapping interference suppression and flexibility, working memory, and response inhibition, we examined FC connectomes reflecting a general executive control state (TR=2000ms, TE=30ms, 256x256mm FOV, 64x64 acquisition matrix, 90-degree flip angle; preprocessing with fMRIprep 22.0.1 [Esteban et al., 2019]). First, we selected regions in the top 5% for FC network integration, operationalized with the graph theory metric participation coefficient, in the group average connectome. We then took the weighted sum of FC for only between-network connections of these "hub" regions and tested for association with p-factor in a multiple linear regression. Second, we applied CPM to identify connections predictive of a latent general factor of in-scanner task performance. We then measured FC network integration, again operationalized with participation coefficient, of regions in the top 5% for predictive connections and tested for association with p-factor in a multiple linear regression.
Results:
Our first approach yielded 16 regions in the top 5% for participation coefficient, highlighting executive networks as well as subcortical areas, and implicating motor control. Strength of between-network FC of two regions significantly predicted lower p-factor: right posterior middle (R2=0.37, F(16,62)=2.25; B=-0.0041, p<0.05) and superior (B=-0.0047, p<0.05) frontal gyrus. Our second approach yielded 22 regions in the top 5% for predictive connections in the CPM. We repeated CPM 1,000 times with 10-fold cross validation (mean r=0.25, permutation p=0.02). Connections selected a maximum of 10,000 times (10 folds * 1,000 repetitions) were predictive of task impairment (r=-0.5, p<0.001), highlighting executive networks as well as the default mode network. Participation coefficient of one region in left posterior superior frontal gyrus significantly predicted lower p-factor (R2=0.26, F(22,56)=0.87; B=-0.49, p<0.05).
Conclusions:
Between-network connectivity of portions of bilateral dorsolateral prefrontal cortex, associated with executive and motor control, explained individual variance in p-factor in two complementary analyses. Identification of such a neurobehavioral mechanism underlying p-factor may lead to novel intervention targets.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 2
Keywords:
Attention Deficit Disorder
Autism
Cognition
FUNCTIONAL MRI
PEDIATRIC
Psychiatric Disorders
1|2Indicates the priority used for review
Provide references using author date format
Menon, V., & D’Esposito, M. (2022), 'The role of PFC networks in cognitive control and executive function', Neuropsychopharmacology, 47(1), 90-103
Moffitt T.E., (2011), 'A gradient of childhood self-control predicts health, wealth, and public safety', Proceedings of the National Academy of Sciences, 108: 2693–2698