Abstract
Deficits in the recognition of emotional information may contribute to the social dysfunction that characterizes depression. Emotion recognition studies in depression, however, have predominately used facial stimuli - a notable limitation as emotion is also inferred from body movement. The present study used a large set of point-light display (PLD) videos, which are minimalistic representations of dynamic whole-body motion, to assess emotion recognition abilities in depression. First, a sample was collected via MTurk (N = 101) to establish normative emotion category ratings (Happy, Sad, Afraid, Angry, Neutral) for each video. A separate sample of healthy controls (N = 83) and lifetime MDD (N = 142) rated the emotion of each video. Test-retest stability of body affect ratings were also assessed for a subset of participants. There were no significant group differences on accuracy for any emotion category, but individuals with greater depressive symptoms exhibited lower accuracy for neutral body affect relative to those with lower depressive symptoms. This effect was not accounted for by co-occurring anxiety symptoms, highlighting the specificity of the effects to depression. Additionally, body affect ratings showed significant test-retest stability over six months for all emotions. In sum, individuals with greater depressive severity may misinterpret neutral body affect as representing an emotional expression. Future studies should clarify the role of impaired body affect recognition in broader social impairments and course of illness.