Neural processing correlates of assaultive violence exposure and PTSD symptoms during implicit threat processing: a network-level analysis among adolescent girls

Abstract

Assaultive violence exposure during childhood is a significant risk factor for posttraumatic stress disorder (PTSD). The purpose of the present study was to characterize the relationships of assault and PTSD severity with the organization of large-scale networks identified during emotion processing. Adolescent girls aged 12-16 with (N=15) and without (N=15) histories of assault underwent functional magnetic resonance imaging (fMRI) while engaged in a task that presented images of fearful or neutral facial expressions. Independent component analysis (ICA) identified a frontocingulate network, a frontoparietal network, and a default mode network. Assault exposure was associated with significantly greater activation of the frontocingulate network for fear versus neutral faces. Within the frontocingulate network, Posttraumatic stress disorder (PTSD) severity was associated with weakened functional connectivity between the left amygdala and the perigenual anterior cingulate. Within the frontoparietal network, assaulted girls demonstrated weakened connectivity of the premotor cortex with the right middle frontal gyrus. Within the default mode network, assault exposure and PTSD severity were associated with strengthening functional connectivity of the parahippocampus with the medial and lateral prefrontal cortex, respectively. Individual differences in functional connections within the frontocingulate network and frontoparietal network among the assaulted group were strongly associated with caregiver-rated family disengagement. These results demonstrate associations between assault and PTSD symptoms with the functional organization of large-scale frontoparietal, frontocingulate, and default mode networks during emotion processing. The relationship with caregiver-rated family disengagement suggests the impact of family support on the neural processing correlates of assault and PTSD symptoms.

Keywords: Adolescence; Early life trauma; Emotion processing; Neuroimaging; PTSD.

Figure 1

Axial, sagittal, and coronal slices of frontocingulate, frontoparietal, and default mode networks identified through independent component analysis and matched to spatial templates identified through automated term-based meta-analysis (Yarkoni et al., 2011) (see supplemental Figure 1).

Figure 2

Functional connectivity indices between nodes in the frontocingulate network that varied as a function of assault exposure (top) and PTSD symptoms severity within the assaulted group (bottom). pgACC = perigenual anterior cingulate; dACC = dorsal anterior cingulate; L Amy = left amygdala; R Amy = right amygdala; Pre-SMA = pre-supplementary area; L Caudate = left caudate; R Caudate = right caudate; L AI = left anterior insula; R AI = right anterior insula; L vAI = left ventral anterior insula; r vAI = right ventral anterior insula; L Parietal = left parietal; R Parietal = right parietal; R MFG = right middle frontal gyrus; L MFG = left middle frontal gyrus.

Figure 3

Functional connectivity indices between nodes in the frontoparietal network that varied as a function of assault exposure (top) and PTSD symptoms severity within the assaulted group (bottom). R premotor = right premotor; L Parietal = left parietal, R Parietal = right parietal; R Pos Middle Frontal = right posterior middle frontal gyrus; R striatum = right striatum; R Ant Middle Frontal = right anterior middle frontal gyrus; L Ant Middle Frontal = left anterior middle frontal gyrus; mPFC = medial prefrontal cortex; L Pos Middle Frontal = left posterior middle frontal gyrus; dmPFC = dorsomedial prefrontal cortex; L premotor = left premotor.

Figure 4

Functional connectivity indices between nodes in the default mode network that varied as a function of assault exposure (top) and PTSD symptoms severity within the assaulted group (bottom). vmPFC = ventral medial prefrontal cortex; R superior frontal = right superior frontal gyrus; L superior frontal = left superior frontal gyrus;

Figure 5

Model performance when using the functional connections associated with PTSD severity in the frontocingulate network to predict family disengagement just among the assaulted participants (left), and when using the functional connections associated with PTSD severity in the frontoparietal network to predict family disengagement just among the assaulted participants (right). The scatter plots depict model-predicted family disengagement (y axis) as a function of observed family disengagement (x axis), with optimal performance indicated via linear fit.