Functional brain network organization and multidomain resilience to neighborhood disadvantage in youth

Abstract

Though youth living in disadvantaged neighborhoods experience greater risk for poor behavioral and mental health outcomes, many go on to show resilience in the face of adversity. A few recent studies have identified neural markers of resilience in cognitive and affective brain networks, yet the broader network organization supporting resilience in youth remains unknown, particularly in relation to neighborhood disadvantage. Moreover, most studies have defined resilience as the absence of psychopathology, which does not consider growing evidence that resilience also includes positive outcomes across multiple domains (e.g., social, academic). We examined associations between brain network organization and multiple resilience domains in a sample of 708 twins (7-19 years old) recruited from neighborhoods with above-average poverty levels. Graph analysis on functional connectivity data from resting-state and task-based functional magnetic resonance imaging was used to characterize features of intrinsic whole-brain and network-level organization, from which we explored associations with resilience in three domains: psychological, social, and academic. Fewer connections between a brain network involved in self-referential processing (i.e., default mode network) and the subcortical system were associated with greater social resilience. Further, greater whole-brain functional integration (i.e., efficiency) was associated with better psychological resilience among youth with relatively lower levels of cumulative adversity exposure. Alternatively, lower whole-brain efficiency and higher whole-brain robustness to disruption (i.e., assortativity) were associated with greater psychological and social resilience among youth with relatively higher levels of cumulative adversity. These findings advance support for multidimensional resilience models and reveal distinct neural mechanisms supporting resilience to neighborhood disadvantage across specific domains in youth. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Figure 1. Neuroresilience Methods

Note. Panel A: Multidimensional Resilience Model. All factor loadings are from the standardized solution. Based on modification indices, the residual variances for parent and youth activities report, and parent activities report and social competency were allowed to co-vary. Standard error of factor loadings are represented in parentheses. All factor loadings and covariances were significant at p <.001. Root mean square error of approximation [RMSEA] = 0.05, comparative fit index, [CFI] = 0.93, standardized root mean residual [SRMR] = 0.05. Child-reported forms include the CYRM, SWLS, and YSR. Parent-reported forms include the CBCL. Teacher-reported forms include the TRF. A follow-up exploratory factor analysis with the two CYRM subscales (personal and relational resilience) further showed that both facets of the CRYM loaded onto the psychological resilience latent factor (Loadings: CYRM_Personal = .80, CYRM_Relational = .81). Panel B: Brain networks of interest. Networks were parcellated using Gordon (cortical) and Tian (subcortical) functional atlases, and visualized using BrainNetViewer (Xia et al., 2013).

Figure 2. Cumulative Risk Exposure Moderates the Association Between Whole-Brain Organization and Multiple Resilience Domains

Note. Panel A: Higher whole-brain global efficiency is associated with greater psychological resilience in youth experiencing less cumulative risk, while lower global efficiency is associated with greater psychological resilience in youth experiencing more cumulative risk (b* = −0.132, p_FDR = .031; Simple slopes: p_−1SD = .053, p_+1SD = .035). Panel B: Higher whole-brain robustness to disruption (i.e., assortativity) is associated with higher psychological resilience in youth with greater cumulative risk exposure (b* = 0.144, p_FDR = .007; Simple slopes: p_+1SD = .001). Panel C: Higher whole-brain robustness to disruption (i.e., assortativity) is associated with higher social resilience in youth with greater cumulative risk exposure (b* = 0.148, p_FDR = .007; Simple slopes: p_+1SD = .008). Simple slopes significance values are positioned near their respective group. Interaction effects are significant within regions shaded in gray. For visualization purposes, plots were generated using the interactions package in R, which does not adjust standard errors for family clustering and depicts only participants with fMRI scan data (n = 559). p < .05*, p < .01**, p < .001***