Sensitivity to Social Threat in Adolescent Girls: Linking Neural, Eyetracking, Questionnaire, and Ecological Momentary Assessments

• In Event: From Brain to Behavior to Daily Life: Cross-Level Assessments of Emotional Response to Peer Interactions

Abstract

Sensitivity to social threat is a risk factor for anxiety and depression during adolescence, a time in which the salience of peer feedback increases dramatically. Sensitivity to potential threat is a key construct in the National Institute of Mental Health’s RDoC initiative, but little attention has been given to the measurement of social threat in adolescence, especially across different units of analysis and using ecologically valid approaches. We combined four different measures of social threat sensitivity (amygdala reactivity to negative social feedback, eye-tracking bias toward social threat, and parent- and adolescent-reports of threat sensitivity). We used a psychoneurometric approach (e.g. Patrick, Durbin, & Moser, 2012) to explore whether sensitivity to social threat across these multiple levels of analysis reflected a unified construct, and whether this construct predicted real-world affect and feelings of connectedness during peer interactions.

Participants were 129 girls ages 11-13 (Mage=12.28, SDage=.80; 65% Caucasian) oversampled for shy/fearful temperament. Sensitivity to social threat was measured via (a) neural activation in sub-regions of the amygdala to social punishment during an fMRI Social Incentive Delay task (Fig 1) in which adolescents received performance feedback via a frowning or smiling photo from a virtual peer they believed was watching them, (b) eyetracking indices of attention bias to angry faces on the Dot Probe Task, and (c) child- and parent-report of rejection sensitivity, loneliness, and peer victimization. Ecological momentary assessment (EMA) over 16 days captured negative affect (NA) related to peer events and perceived social disconnectedness with peers in daily life.

Confirmatory factor analyses (Fig 2) supported a method-specific four-factor model (χ2(58)= 71.288, p = .112, RMSEA = .042; CFI = .97; TLI = .96; SRMR = .054), indicating that measures were correlated within level of analysis but largely uncorrelated across levels. A latent eyetracking factor indicating attentional bias toward angry faces was weakly associated (r=.25) with the adolescent report factor. Both the child report (β=.31) and attention bias (β=.24) factors predicted increased peer-related NA in daily life. The amygdala reactivity factor was not correlated with other factors, but predicted social disconnectedness with peers in daily life (β=.28), as did the adolescent report factor (β=.43). The parent report factor was not correlated with any other variables.

Findings suggest that measures of sensitivity to social threat across different levels of analysis are not measuring a single construct, highlighting measurement challenges confronting efforts to quantify developmentally sensitive constructs across methods, such as the NIMH RDoC initiative. However, both amygdala response to simulated social punishment and attentional bias toward threatening faces predicted real-world experiences with peers, supporting the ecological validity of these tasks and suggesting that vigilance toward potentially threatening social information could be a mechanism through which vulnerable youth come to experience their peer interactions more negatively. While brain, behavioral (eyetracking), and adolescent-report all contributed unique variance to understanding adolescents' daily peer experience, parent report was unrelated to the other factors or to daily experience, suggesting that researchers need to go beyond parent report to understand how adolescents experience and process social threat.

Authors

• Jennifer S. Silk, University of Pittsburgh
  Presenting Author

• Aleksandra Kaurin, University of Pittsburgh
  Non-Presenting Author

• Stefanie Sequeira, University of Pittsburgh
  Non-Presenting Author

• Kirsten M.P. McKone, University of Pittsburgh
  Non-Presenting Author

• Aidan Wright, University of Pittsburgh
  Non-Presenting Author

• Cecile D Ladouceur, University of Pittsburgh
  Non-Presenting Author