Social cognitive impairment in individuals at clinical high risk for psychosis: a systematic review and meta-analysis.
Abstract
Background
Social cognition is a clinically important domain in psychotic disorders and may already be altered during the clinical high-risk (CHR) stage. However, the magnitude of impairment across different social cognitive domains and its relationship with later transition to psychosis remain uncertain.
Methods
We conducted a systematic review and meta-analysis of studies identified through searches of PubMed, Embase, and Ovid for articles published from January 1996 to July 2025. Eligible studies compared individuals at CHR for psychosis with healthy controls (HCs) using previously described or commonly used measures of social cognition. Random-effects models were applied to calculate standardized mean differences (SMDs) and 95% confidence intervals (CIs).
Results
Seventeen studies involving 2019 CHR individuals and 1039 HCs were included. Compared with HCs, CHR individuals showed poorer overall social cognitive performance (SMD = -0.46, 95% CI: -0.68 to -0.24), although between-study heterogeneity was substantial (I2 = 81.0%). Domain-specific analyses indicated significant deficits in theory of mind (SMD = -0.37, 95% CI: -0.48 to -0.27; I2 = 0%) and emotion perception (SMD = -0.34, 95% CI: -0.53 to -0.14; I2 = 59.6%). These impairments were generally consistent across task modalities. Exploratory longitudinal evidence linking poorer emotion perception to psychosis transition remains sparse, and the findings are highly heterogeneous.
Conclusions
Individuals at CHR for psychosis exhibit measurable impairments in social cognition, particularly in theory of mind and emotion perception. Current evidence does not sufficiently support social cognitive deficits as robust predictors of psychosis transition. Further longitudinal studies using better validated and harmonized assessments are needed to determine their prognostic value.