Abstract
Although contemporary models of face processing suggest that emotion recognition and identity processing dissociate, relying on different neural and cognitive mechanisms, selective deficits in facial emotion processing are not often observed. These are evident, however, in individuals with alexithymia, a subclinical trait characterized by difficulties identifying and describing one's own feelings. Our previous research indicated that, in Autism Spectrum Conditions (ASC), levels of alexithymia, and not autistic symptom severity, predicted participants' ability to recognize others' facial emotion. Alexithymia was unrelated, in individuals with and without ASC, to ability to recognize facial identity. Alexithymia is known to be associated with multiple psychological disorders, but whether it has identical underlying causes, or manifests in the same way, across clinical groups is not yet understood. The present study sought to determine whether alexithymia is associated with selective emotion recognition deficits in individuals with eating disorders and schizophrenia, in the same way as in ASC. Eighty participants with an eating disorder, schizophrenia, or no psychological disorder viewed facial images, presented for 800 ms and obscured by visual noise. Following the presentation of each face, they were prompted to attribute either emotion or identity to the face, in the form of a two-alternative forced choice question e.g. 'Anger: yes or no?'. The level of visual noise was adjusted using an adaptive staircase procedure, in order to determine each participant's tolerance for noise when making identity and emotion judgments. In both populations, alexithymia was found to be predictive of emotion recognition ability, despite being unrelated to identity recognition ability. These results suggest that varying degrees of co-occuring alexithymia might contribute to the equivocal findings of emotion recognition deficits in these conditions. Reporting, and controlling for, levels of co-occuring alexithymia should therefore be regarded as routine practice in future studies of emotion recognition in clinical populations.
Meeting abstract presented at VSS 2014