L’altération de la cognition sociale associée à la schizophrénie, dont fait partie la perception des émotions faciales, a fait l’objet d’une récente méta-analyse qui a révélé son impact majeur sur l’insertion sociale et professionnelle. Or l’utilisation de programmes de remédiation cognitive portant sur le traitement des émotions faciales est freinée par un manque d’outil permettant d’identifier et de quantifier les troubles pour un sujet donné. Le test de reconnaissance des émotions faciales (TREF) permet d’évaluer la capacité à reconnaître 6 émotions universelles (joie, colère, tristesse, peur, dégoût et mépris), présentées à neuf intensités d’expression s’étalant de 20 % à 100 %. Les résultats de l’étude comparative réalisée auprès d’un échantillon de 64 sujets témoins sains et de 45 personnes souffrant de schizophrénie montrent, de manière cohérente avec les données de la littérature, une diminution significative des performances de reconnaissance des émotions faciales dans la schizophrénie. Des scores seuils sont proposés afin de permettre l’utilisation du TREF pour l’identification et la spécification de la présence d’un déficit de reconnaissance des émotions faciales chez un sujet donné, et permettre son orientation vers une prise en charge spécifique.

The impairment of social cognition, including facial affects recognition, is a well-established trait in schizophrenia, and specific cognitive remediation programs focusing on facial affects recognition have been developed by different teams worldwide. However, even though social cognitive impairments have been confirmed, previous studies have also shown heterogeneity of the results between different subjects. Therefore, assessment of personal abilities should be measured individually before proposing such programs.

Most research teams apply tasks based on facial affects recognition by Ekman et al. or Gur et al. However, these tasks are not easily applicable in a clinical exercise. Here, we present the Facial Emotions Recognition Test (TREF), which is designed to identify facial affects recognition impairments in a clinical practice. The test is composed of 54 photos and evaluates abilities in the recognition of six universal emotions (joy, anger, sadness, fear, disgust and contempt). Each of these emotions is represented with colored photos of 4 different models (two men and two women) at nine intensity levels from 20 to 100 %. Each photo is presented during 10seconds; no time limit for responding is applied.

The present study compared the scores of the TREF test in a sample of healthy controls (64 subjects) and people with stabilized schizophrenia (45 subjects) according to the DSM IV-TR criteria. We analysed global scores for all emotions, as well as sub scores for each emotion between these two groups, taking into account gender differences. Our results were coherent with previous findings. Applying TREF, we confirmed an impairment in facial affects recognition in schizophrenia by showing significant differences between the two groups in their global results (76.45 % for healthy controls versus 61.28 % for people with schizophrenia), as well as in sub scores for each emotion except for joy. Scores for women were significantly higher than for men in the population without psychiatric diagnosis. The study also allowed the identification of cut-off scores; results below 2 standard deviations of the healthy control average (61.57 %) pointed to a facial affect recognition deficit.

The TREF appears to be a useful tool to identify facial affects recognition impairment in schizophrenia. Neuropsychologists, who have tried this task, have positive feedback. The TREF is easy to use (duration of about 15minutes), easy to apply in subjects with attentional difficulties, and tests facial affects recognition at ecological intensity levels. These results have to be confirmed in the future with larger sample sizes, and in comparison with other tasks, evaluating the facial affects recognition processes.