Two-Month Cognitive Changes Enhance Prediction of Non-Remission in Clinical High-Risk Individuals.

Abstract

Background: Longitudinal changes in cognitive function may be crucial in predicting clinical outcomes in clinical high risk (CHR) individuals. This study aims to investigate the predictive value of baseline cognitive impairment and short-term cognitive changes for non-remission and conversion to psychosis in individuals at CHR for psychosis, compared with healthy controls (HC).

Methods: This study employed a multiple-group prospective design with a 3-year follow-up. CHR individuals and HCs were assessed at baseline and at a 2-month follow-up. Neuropsychological performance was evaluated using the Chinese version of the Measurement and Treatment Research to Improve Cognition in Schizophrenia Consensus Cognitive Battery.

Results: The study included 310 CHR individuals and 93 HCs. Significant improvements in predicting non-remission in CHR individuals were observed when incorporating cognitive changes over 2 months (AUC for baseline cognition, 0.690; AUC for changes, 0.819; Z=3.365, p<0.001). Key predictors included the Revised Hopkins Verbal Learning Test (HVLT-R; β=0.083, p=0.003), Wechsler Memory Scale-III spatial span (WMS-3; β=0.330, p<0.001), and Revised Brief Visuospatial Memory Test (BVMT-R; β=0.127, p<0.001). Conversely, predicting conversion to psychosis showed no significant difference between baseline and 2-month cognitive changes (AUC for baseline cognition, 0.667; AUC for changes, 0.666; Z=0.021, p=0.242).

Conclusions: The findings underscore the importance of dynamic cognitive monitoring in CHR individuals. Short-term cognitive changes significantly enhance the prediction of non-remission but do not add predictive value for conversion to psychosis beyond baseline assessments. Specific cognitive domains, such as verbal learning and working memory, are particularly valuable for predicting clinical outcomes.