S70. PROFILES OF SOCIAL COGNITION AND METACOGNITION IN FIRST-EPISODE PSYCHOSIS: A LATENT PROFILE ANALYSIS

BACKGROUND: Social cognition and metacognition have emerged as cornerstones of research and treatment of schizophrenia. Both constructs are known to influence the onset of psychosis, to predict functional outcome and to be associated with symptoms. However, whether the deficits in first-episode psychosis are homogeneous or group in patterns remains to be studied. This study aimed to analyze patterns of social cognitive and metacognitive variables in a sample of subjects with first-episode psychosis. METHODS: We recruited 192 subjects with first-episode psychosis from ten public mental-health services in Spain. We collected: demographic information, measures of functioning, performance in social cognition (the Faces Test, IPSAQ and the Hinting Task), a battery of metacognitive tasks (BCIS, and the Beads Task) and a neuropsychological assessment. We performed a Latent Profile Analysis (LPA) with the metacognitive and social-cognitive variables. The variable importance was assessed via a classification tree (CART) and the mean differences among the resulting groups for clinical, neuropsychological and functioning variables were calculated with ANOVA and Kruskal-Wallis tests. RESULTS: Our sample was comprised of 192 (62 women) with first-episode psychosis. The mean age of the sample was 27.93(1.39). The mean PANSS total score of the sample was 58.48 (17.79). The average GAF score was 58.93 (12.25). We included 174 cases with complete social-cognitive and metacognitive data in the cluster analysis. We identified three type-VEE clusters (i.e. ellipsoidal clusters with equal shape and orientation) according to BIC (BIC=-3600.651). The 85-15 condition of the Beads Task and the Hinting Task emerged as the most important variables in determining the clustering structure. The first cluster (60.9%) was characterized by average scores in most of the metacognitive and social cognitive variables, but the presence of the jumping to conclusions bias. The second cluster (5.7%) was characterized by low self-reflectiveness, presence of personalizing bias and an excessive number of trials in the beads task. The third cluster (33.5%) was characterized by average scores in all metacognitive tasks but low scores in the social cognitive tasks. DISCUSSION: We found three clusters in a large sample of subjects with first-episode psychosis. Our results indicate that the three groups differ in the proneness to present deficits in specific domains. Furthermore, in our sample, patients may not exhibit a homogeneous deficit in all social-cognitive and metacognitive variables. Instead, the impairment may be particularly prominent in either social-cognitive or metacognitive variables. Subjects in different clusters may present differences in their clinical characteristics, what could be relevant in the treatment. Therefore, with further research, a thorough assessment of social cognition and metacognition may help personalize the treatment according to the person’s subtype of the deficit.