Generalization of learned fear and associated neurobiology in children exposed to early caregiving adversity

• In Event: Biological embedding of early adversity: new insights on the effects on neurodevelopment and threat processing

Abstract

Learning to distinguish safe from unsafe cues is critical for survival, particularly in dangerous environments. When safe cues are confused as threatening, it can lead to the generalization of fear learning. Fear generalization has been attributed to deficient pattern discrimination both behaviorally and within the hippocampus (Lange et al., 2017; Lecei & van Winkel, 2019), and is highest in childhood, declining with increasing age as discrimination abilities improve (Schiele et al., 2016). Importantly, fear generalization may be exacerbated by environmental disruptions, a prediction supported by findings in emerging adults (Lange et. al., 2019) and work indicating that the development of hippocampal regions implicated in fear generalization is modified by early adversity (Teicher et al., 2016). Examining the influence of adversity on fear generalization behavior may provide important insight into how the early environment shapes the development of fear learning processes.
In the present study, we investigated fear generalization behavior in children who experienced heterogeneous early adverse caregiving experiences (i.e. institutionalization, foster care, kinship care, temporary parental separations) (N = 190, Age = 7-14 years). Fear generalization behavior was characterized using a modified behavioral faces task (Schiele et al., 2016), in which participants provided trial-by-trial ratings of neutral faces either paired with an aversive US (CS+), not paired (CS-) or morphed neutral faces containing varying CS+/CS- ratios. Prior to the behavioral task, participants underwent fMRI scanning, during which they participated in a 5-minute passive viewing task, which involved viewing fearful and neutral emotional facial stimuli.
Multilevel logistic regression analysis demonstrated an adversity group x face stimulus interaction (β̂ = -1.27, 95% CI [-2.50, -.10]), such that children who experienced foster care attributed greater fear to CS- morphs (i.e., blended with CS+) relative to non-adversity-exposed comparisons. Planned analyses include conducting representational similarity analyses (RSA) of functional magnetic resonance imaging (fMRI) to examine whether the similarity of hippocampal patterns of activity in response to fearful versus neutral faces predicts generalization behavior. We will compute representational similarity by correlating within-subject patterns of hippocampal activity for all pairs of neutral and fear facial stimuli from a passive viewing emotional faces task to create similarity matrices. We will then calculate the difference between pattern similarity for faces of the same emotion versus faces of different emotions as an index of hippocampal pattern discrimination, and assess the association between this discrimination index and fear generalization behavior.
Overall, these behavioral results indicate that specific kinds of adversity exposure (i.e., foster care) are associated with greater generalization of learned fear in children. We anticipate that foster care exposed children will show decreased hippocampal pattern discrimination between fear and neutral faces relative to the comparison group, and this will be associated with increased fear appraisal of morphed stimuli, providing novel insight into the neural underpinnings of childhood fear generalization behavior. This work may have important implications for understanding the mechanism underlying certain forms of psychopathology, like anxiety, which has been linked to both greater fear generalization behavior (Dunsmoor et al., 2015) and foster care experience (Turney & Wildeman, 2016).

Authors

• Andrea Fields, Columbia University
  Presenting Author

• Paul A Bloom, Columbia University
  Non-Presenting Author

• Ayumi Tachida, Columbia University
  Non-Presenting Author

• Anna Vannucci, Columbia University
  Non-Presenting Author

• Ian J Douglas, University of Texas at Austin
  Non-Presenting Author

• Nicolas L Camacho, Columbia University
  Non-Presenting Author

• Lisa Gibson, Columbia University
  Non-Presenting Author

• Syntia Hadis, Columbia University
  Non-Presenting Author

• Tricia Choy, University of California - Riverside
  Non-Presenting Author

• Charlotte Heleniak, University Of Washington
  Non-Presenting Author

• Nim Tottenham, Columbia University
  Non-Presenting Author