Social pain is associated with altered developmental trajectories of connectivity among the triple network model of psychopathology

Abstract

Childhood and adolescence are sensitive periods for the refinement of increasingly complex executive and social functions. A particularly important skill is the ability to navigate and interpret interpersonal relationships, which is reflected in part by the maturation of distributed resting networks. However, the relationships between negative social perceptions in youth and long-term alterations in between-network connectivity are limited. To partially address this gap, we utilized longitudinal resting-state functional magnetic resonance imaging (N = 93) with social pain measures from the NIH Toolbox Emotion Battery to determine the links between negative social perceptions on the trajectory of connectivity between the salience, frontoparietal, and default mode networks in the triple network model of psychopathology. Higher scores of perceived hostility, but not perceived rejection, tended to increase functional connectivity between the salience and both frontoparietal and default mode networks over time. These results suggest that more direct forms of threat (hostility) may be more impactful than rejection (limited desired social interactions), highlighting the importance of a dimensional approach to understanding developmental trajectories. While these connectivity changes align with several aberrant connectivity signatures observed across mental health disorders, these phenotypes are not pathognomonic of psychopathology and may reflect adaptive mechanisms in the context of social adversity.

Keywords: fMRI; longitudinal development; social distress; structural equation modelling.

Figure 1.

Data distribution for negative social perception measures and conceptual path diagram. Latent growth curve modelling was used to estimate the impact of perceived hostility and rejection on baseline connectivity and longitudinal changes in connectivity, while controlling for age and sex. T-score distributions for perceived hostility and perceived rejection are shown in the raincloud plots, with T-scores > 60 considered to be potentially problematic.

Figure 2.

Relationship between the slope of connectivity change between the salience and frontoparietal networks and perceived hostility T-scores. The scatterplot shows individual T-scores of perceived hostility, as well as the estimated slope of connectivity change (following adjustment by regressing out the effect of sex and age in the latent growth curve model). The salience network is shown in blue (darker) and the frontoparietal network is shown in red (lighter colored areas). T-scores >60 are considered to be ‘potentially problematic’.

Figure 3.

Relationship between the slope of connectivity change between the salience and default mode networks and perceived hostility T-scores. The scatterplot shows individual T-scores of perceived hostility, as well as the estimated slope of connectivity change, which were adjusted by regressing out the effect of sex and age in the latent growth curve model. The salience network is shown in blue (darker) and the default mode network is shown in green (lighter colored areas). T-scores >60 are considered to be ‘potentially problematic’.