Beyond hostility: exploring facial emotion recognition biases in youths with conduct disorder.

Abstract

Facial emotion recognition (FER) biases refer to systematic tendencies to recognize specific emotions when processing facial expressions. In youths with conduct disorder (CD), who are characterized by highly impairing antisocial behavior, research on FER biases has focused on hostile attribution biases. This work has shown that youths with CD perceive ambiguous social cues as angry. However, youths with CD may not only show biases towards anger, which is why we investigated FER biases in youths with CD towards the six basic emotions. Within the European FemNAT-CD study, we analyzed data from 610 youths with CD (60% female) and 818 typically developing controls (TDCs; 68% female), aged 9 to 18 years (M = 14.1, SD = 2.41 years). FER biases were assessed using the Emotion Hexagon Task by showing morphed emotional expressions and asking participants to choose the predominant emotion. Biases were calculated as tendency towards an emotion shown at 0%, 10%, 30%, or 50% intensity. Our findings from hierarchical linear modelling indicate that youths with CD exhibited stronger FER biases than TDCs across all emotions, meaning that they misclassified each emotion more often. However, this difference varied by intensity, with youths with CD displaying weaker biases at higher intensity levels and a smaller increase in bias with increasing intensity level. Our findings indicate that youths with CD not only show a hostile attribution bias but rather misclassify emotions as predominant when they are present at low intensity, regardless of type of emotion.