Abstract

Depression is a common mental health disorder, with 12 of the UK population diagnosed at any one time. We assessed whether there are cues to depressive symptoms within the static, non-expressive face, and if other socially relevant impressions might be made by these cues. Composite "average" face images were created from students scoring high and low on self-report measures of depressive symptoms, capturing potential correlations between facial appearance and symptoms of depression. These were then used in a warping procedure, creating two versions of individual faces, one warped towards the high symptom composite, and the other towards the low. In Experiment 1, we first found observers were able to identify images representing high and low symptom occurrence at levels significantly greater than chance. Secondly, we collected observer impressions of the two versions of each face. The faces reflecting high levels of depressive symptoms were picked as less socially desirable over a broad range of personality trait estimates compared to low symptom images. In Experiment 2, we replicated the key finding that the static face contains cues to levels of depression symptoms, using composites created from a new database of student photos and depression inventory scores.