Early and late cortical reactivity to passively viewed emotional faces in pediatric bipolar disorder

Background: We studied emotional information processing in youth with pediatric bipolar disorder (pBD)using the late positive potential (LPP), assessing automatic allocation of attentional resources to emotionally salient stimuli, and the occipital P1, assessing early sensory processing. Methods: Participants were 20 youth with pBD and 26 healthy controls (HC). Participants passively viewed faces with a fearful, neutral or happy expressions. Group differences were tested with general linear models. P1 was included to examine modulating effects on LPP. We calculated Bayes factor (BF)values to express strength of evidence for choosing one hypothesis over another. Results: A significant emotion by group interaction for LPP amplitude was associated with a larger amplitude for happy faces for pBD than HC (F[1,40]= 6.04, p =.018); this was not modulated by P1 amplitude or latency. P1 amplitude did not differ between groups, although P1 peaked earlier for HC (F[1,40]= 5.45, p =.025). BF for LPP was 2.93, suggesting moderate evidence favoring H1. BF for P1 latency of 14.58 suggests strong evidence favoring H1. Limitations: Inclusion of children and adolescents prohibited careful control for neurodevelopmental effects. Conclusions: Larger LPP amplitude for happy faces without change in P1 suggests enhanced automatic allocation of attentional resources to positive information in pBD. Delayed P1 latency in pBD suggests slower early processing of emotional information.