Selective neural sensitivity to familial threat in adolescents with weak family bonds

Abstract

Familial stressors, such as weak familial connectedness, are associated with the development of maladaptive threat processing, yet little is known regarding how weak familial bonds impinge on biological mechanisms of threat processing. The present study leveraged multivoxel pattern analysis of fMRI data to compare the neural encoding of familial and nonfamilial threatening and non-threatening stimuli in adolescents who endorsed varying levels of connectedness to their families. Adolescents (N = 22, Mage = 14.38 years) reporting lower family connectedness 1 year earlier showed elevated sensitivity to familial threat, but not to nonfamilial threat in a neural network associated with threat processing, comprising left and right amygdala, and right inferior and middle temporal gyri. Results suggest that a learning history about one's social environment may shape neural mechanisms of threat processing by sensitizing them to risk-relevant stimuli. Such findings advance our understanding of how familial stressors contribute to disordered threat processing in adolescence.

Keywords: MVPA; Threat; amygdala; fMRI; family stressor.

Figure 1.

(A) Represents how we conceptualize social threat encoding as a multidimensional process, comprising two dimensions (non-exhaustive) of familiarity and valence. Holding valence constant, familial threat is always more familiar. Similarly, when holding familiarity constant, familial threat tends to elicit greater negative valence. However, the difference in either dimension may be minimal, which may be why it is difficult to distinguish these closely related cognitive representations using traditional univariate fMRI analyses. This conceptual structure may be implemented in neural systems associated with threat encoding as depicted in (B). Assume the 2 voxels in the axes comprise the entire amygdala. If one used traditional voxel-wise univariate analysis of fMRI data to test the hypothesis that voxels in the amygdala distinguish familial from nonfamilial threat, the mean voxel-wise activations for a given population for familial (red circles) and non- familial (yellow circles) stimuli may be insignificantly different from each other. Conversely, an MVPA algorithm that takes into account the distributed pattern of activity across voxels could estimate the red-dotted line that best distinguishes these types of stimuli.

Figure 2.

Mother vs. Stranger Go No-Go fMRI task.

Figure 3.

Family connectedness and Sensitivity to unpredictable, familial threat in the threat network. The red-dotted line represents chance levels (50% accurate] for a 2-class classifier.