| Sumario: | Background: Individuals at clinical high risk (CHR) for psychosis have become a major focus for research designed to explore early predictors of transition to full psychosis. Characterizing differences in neurocognitive (NC) functioning between psychosis converters (CHR-C) and non-converters (CHR-NC) might contribute to the identification of specific NC predictors of psychosis onset. Therefore, the aim of the present meta-analysis was to compare the baseline NC performance between CHR-C and CHR-NC.Method: PubMed (MEDLINE), Web of Science, Embase and reference lists were searched for studies reporting baseline cognitive data of CHR-C and CHR-NC. Included NC tests were classified within the MATRICS - Measurement and Treatment Research to Improve Cognition in Schizophrenia - cognitive domains.Results: Of 95 studies assessed for eligibility, 9 studies comprising 583 CHR subjects (N CHR-C=195, N CHR-NC=388) met all the inclusion criteria. CHR-C performed significantly worse compared to CHR-NC on 2 MATRICS domains namely working memory (ES=-0.29, 95% CI=-0.53 to -0.05) and visual learning (ES=-0.40, 95% CI=-0.68 to -0.13). For the remaining 4 domains (processing speed, attention/vigilance, verbal learning, reasoning/problem solving) no significant differences between CHR-C and CHR-NC were observed.Conclusion: Based on the current meta-analytic data we might conclude that it is possible to differentiate between CHR-C and CHR-NC with respect to working memory and visual learning. The addition of visual learning and working memory tasks to psychosis regression models might contribute to the predictive power of these models.
|
|---|