Associations Between Resting-State Functional Connectivity and a Hierarchical Dimensional Structure of Psychopathology in Middle Childhood

Abstract

Background: Previous research from the Adolescent Brain Cognitive Development (ABCD) study delineated and validated a hierarchical 5-factor structure with a general psychopathology ('p') factor at the apex and five specific factors (internalizing, somatoform, detachment, neurodevelopmental, externalizing) using parent-reported child symptoms. The current study is the first examining associations between dimensions from a hierarchical structure and resting state functional connectivity (RSFC) networks.Methods: Using 9-11-year-old children from the ABCD baseline sample, we compared the variance explained by each hierarchy level (p-factor, 2-factor, 3-factor, 4-factor, and 5-factor models) in RSFC. Analyses were first conducted in a discovery dataset (n=3790) with significant associations examined in a replication dataset (n=3791). Results:The current study found associations between p-factor and lower connectivity within default mode network (DMN), although stronger effects emerged for the neurodevelopmental factor. Neurodevelopmental impairments were related to variation in RSFC networks associated with attention to internal states and external stimuli. These networks included within DMN, DMN with cingulo-opercular (CON) and 'Other' (Unassigned) networks, CON with ventral attention and 'Other' network, and dorsal attention with 'Other' network. Results held when accounting for parental psychopathology. Conclusion:The hierarchical structure of psychopathology showed replicable links to RSFC alterations in middle childhood. The p-factor had minimal association with altered connectivity, while the specific neurodevelopmental dimension showed robust

Widespread attenuating changes in brain connectivity associated with the general factor of psychopathology in 9- and 10-year olds

“…This model is suggestively reflected in our findings, where we found altered connectivity in networks associated with spontaneous thought (DMN) [52,53] and bottom-up attention (VAN) [54], as well as altered connectivity in control networks (FPN, CO, DAN) involved in response control [14,55]. Notably, other recent studies of the P factor examining functional connectivity also found alterations in networks involved in bottom-up processing, including DMN [16], visual network [15], and somatomotor network [18]. Altered functional connectivity in core control networks (i.e., FPN and DAN), however, appears to be a relatively new finding reported here, and our results thus provide firmer grounding for the dysregulation model of the P factor.…”

“…The current study examines functional-connectivity patterns across the whole connectome associated with the P factor in a sample of 11,875 9-and 10-year olds in the Adolescent Brain and Cognitive Development (ABCD) national consortium study, Release 2.0.1 [31]. Recently, our group constructed and validated a P factor model [32] in ABCD from the Child Behavior Checklist (CBCL) parent report [33] using bifactor modeling, and performed additional analyses that support a P factor structural model of psychopathology in this sample [32,34], see also similar work by [16,35]. For the present study, we produced resting-state connectomes for 6593 youth who met stringent neuroimaging quality-control standards.…”

“…These findings raise the possibility that the P factor involves alterations in networks involved in the generation of spontaneous thought (default mode network, DMN) and bottomup attention (ventral attention network, VAN), as well as networks involved in cognitive control (e.g., frontoparietal network, FPN, dorsal attention network, DAN; cingulo-opercular network, CO) [14]. Recent studies found altered functional-connectivity patterns associated with transdiagnostic dimensions [15][16][17][18][19][20][21], but findings have been mixed and await further clarification.…”

An exploration of dimensions of early adversity and the development of functional brain network connectivity during adolescence: Implications for trajectories of internalizing symptoms

“…The SAL encompasses anterior cingulate and ventral anterior insular regions and is a broad network with proposed roles in awareness, salience detection, and autonomic processing (Seeley, 2019). Several studies have found that internalizing symptoms in adolescents are associated with altered connectivity in the FPN (Pan et al, 2020), DMN (Kaiser & Pizzagalli, 2015;Karcher et al, 2020), and SAL (Ordaz et al, 2016), as well as between these networks; for example, DMN-FPN connectivity has been found to be higher, and FPN-SAL connectivity lower, in youth with more severe internalizing symptoms (see review by Chahal et al, 2020 for further information about within-and between-network connectivity differences). However, most studies examining associations between adversity and neurocircuitry have focused on fronto-limbic (mainly prefrontal-amygdala) functional connectivity.…”

Common and Disorders-Specific Cortical Thickness Alterations in Internalizing, Externalizing and Thought Disorders over a 2-year Period in the Preadolescents of the ABCD Study

“…Based on accumulating evidence the HiTOP proposes that single mental disorders could be classified into three broad diagnostic families: externalizing disorders (inattention, aggressive and disruptive behavior), internalizing disorders (depression, anxiety and fear) and thought disorders (delusions, hallucinations and obsessions). Accordingly, a growing literature [13][14][15][16] has focused on mapping neural correlates for general psychopathology ('p factor'), reflecting an overarching susceptibility to any mental disorder [17,18], and several specific low-order broad diagnostic families (e.g., externalizing, internalizing and thought disorders). This hierarchical framework of mental disorders removes unclear boundaries between single mental disorders by grouping disorders with related symptoms into broad diagnostic families, which may further contribute to determining the underlying pathological dimensions on the neural, genetic and phenotype level [19,20].…”