Abstract:Objective: Traumatic brain injury (TBI) is a prevalent cause of long-term morbidity in children and adolescents and can lead to persistent difficulties with social and behavioral function. TBI may impact brain structures that support social cognition, social perception, and day-to-day social interactions—termed the social brain network (SBN). We examined differences in links among the SBN and regions of interest from other neural networks thought to support social outcomes, i.e., the default mode network (DMN) and salience network (SN). Furthermore, we examined how differences in co-activation among the SBN and these other key networks were associated with ratings of social and day-to-day adaptive outcomes. Participants and Methods: Participants included children and adolescents with moderate to severe TBI (msTBI; n=11, Mage=11.78, 6 male), complicated-mild TBI (cmTBI; n=12, Mage=12.59, 9 male), and orthopedic injury (OI; n=22, Mage=11.69, 15 male). Participants underwent resting-state functional MRI on a 3Tesla Siemens Prisma scanner. Parents rated their child’s social and adaptive function on the Child Behavior Checklist (CBCL) and Adaptive Behavior Assessment System-Third Edition (ABAS-3). Resting-state connectivity was assessed using the CONN Toolbox, including preprocessing, denoising, and alignment to the participants’ processed T1 MPRAGE sequence followed by seed-to-voxel analysis using a SBN mask and targeted regions of interest within the DMN and SN. Individual-level r-to-z correlations were extracted from resulting clusters of co-activation with the SBN mask and exported into SPSSv28.0 for integration with behavioral data. Results: One-way ANOVAs used to examine group differences in social and adaptive outcome revealed significant group differences in CBCL Social Competence (F =4.49, p =.019) and all composite scores on the ABAS-3 (F s=3.78 to 5.17, p s=.031 to .010). In each domain, children with msTBI were rated as having elevated difficulties relative to cmTBI or OI, whereas cmTBI and OI groups did not differ. Connectivity also differed significantly between groups, with children with OI demonstrating greater connectivity between the SBN and the anterior cingulate cortex of the SN (t =5.19, p (FDR)<.0001) and posterior cingulate cortex of the DMN (f =4.30, p (FDR)<.001) than children with msTBI. Children with cmTBI also showed greater connectivity between the SBN and left temporal pole of the DMN (t =7.45, p (FDR)<.000001) than children with msTBI. Degree of connectivity between the SBN and posterior cingulate was significantly positively correlated across all domains of adaptive function (r s=.451 to .504, p s=.010 to .003), whereas degree of connectivity between the SBN and left temporal pole was strongly positively related to Social Competence ( a =.633, p =.006) and conceptual adaptive skills on the ABAS (A =.437, p =.037). Conclusions: Our findings provide insights into the neural substrates of social and adaptive morbidity after pediatric TBI, particularly msTBI, by linking alterations in connectivity among the SBN, DMN, and SN with measures of social and adaptive outcome. While the posterior cingulate was broadly associated with adaptive outcome, the temporal pole was particularly strongly associated with social competence. This may reflect the diverse functions and high degree of interconnectivity of the posterior cingulate, which contributes to various cognitive and attentional processes, relative to the strong amygdala/limbic connections of the temporal pole.

Bayesian Varying-Effects Vector Autoregressive Models for Inference of Brain Connectivity Networks and Covariate Effects in Pediatric Traumatic Brain Injury

Effective connectivity in the default mode network after paediatric traumatic brain injury

Bayesian Modeling of Effective and Functional Brain Connectivity using Hierarchical Vector Autoregressions

Bayesian modelling of effective and functional brain connectivity using hierarchical vector autoregressions

Semi-parametric Bayes Regression with Network Valued Covariates

Altered Gray Matter Structural Covariance Networks at Both Acute and Chronic Stages of Mild Traumatic Brain Injury

High-dimensional multisubject time series transition matrix inference with application to brain connectivity analysis

Bayesian pathway analysis over brain network mediators for survival data

Network analysis of structural MRI predicts executive function in paediatric traumatic brain injury

A Predictor-Informed Multi-Subject Bayesian Approach for Dynamic Functional Connectivity

Bayesian Time-Varying Tensor Vector Autoregressive Models for Dynamic Effective Connectivity

Bayesian Joint Modeling of Multiple Brain Functional Networks

Longitudinal Analysis of Brain Recovery after Mild Traumatic Brain Injury Based on Groupwise Consistent Brain Network Clusters.

Covariate-Dependent Clustering of Undirected Networks with Brain-Imaging Data

Bayesian mixed model inference for genetic association under related samples with brain network phenotype

Temporal Concatenated Sparse Coding of Resting State fMRI Data Reveal Network Interaction Changes in mTBI.

Learning Brain Connectivity in Social Cognition with Dynamic Network Regression

Evaluating the effects of high-throughput structural neuroimaging predictors on whole-brain functional connectome outcomes via network-based vector-on-matrix regression

Bayesian longitudinal tensor response regression for modeling neuroplasticity

Exploring the link among injury severity, white matter connectivity and psychosocial outcomes in pediatric TBI: a probabilistic tractography approach

1 Social Brain Network Connectivity Relates to Social and Adaptive Outcomes Following Pediatric Traumatic Brain Injury