Abstract
BACKGROUND: Although theorists have posited that adolescent depression is characterized by emotion-processing biases (greater propensity to identify sad than happy facial expressions), findings have been mixed. Additionally, the neural correlates associated with putative emotion-processing biases remain largely unknown. Our aim was to identify emotion-processing biases in depressed adolescents and examine neural abnormalities related to these biases using high-density resting EEG and source localization. METHODS: Healthy (n = 36) and depressed (n = 23) female adolescents, aged 13-18 years, completed a facial recognition task in which they identified happy, sad, fear, and angry expressions across intensities from 10% (low) to 100% (high). Additionally, 128-channel resting (i.e., task-free) EEG was recorded and analyzed using a distributed source localization technique (low-resolution electromagnetic tomography (LORETA)). Given research implicating the dorsolateral prefrontal cortex (DLPFC) in depression and emotion processing, analyses focused on this region. RESULTS: Relative to healthy youth, depressed adolescents were more accurate for sad and less accurate for happy, particularly low-intensity happy faces. No differences emerged for fearful or angry facial expressions. Further, LORETA analyses revealed greater theta and alpha current density (i.e., reduced brain activity) in depressed versus healthy adolescents, particularly in the left DLPFC (BA9/BA46). Theta and alpha current density were positively correlated, and greater current density predicted reduced accuracy for happy faces. CONCLUSION: Depressed female adolescents were characterized by emotion-processing biases in favor of sad emotions and reduced recognition of happiness, especially when cues of happiness were subtle. Blunted recognition of happy was associated with left DLPFC resting hypoactivity.