Antipsychotic Effects on Longitudinal Cognitive Functioning in First-Episode Psychosis: A randomised, triple-blind, placebo-controlled study

Abstract

Objective: Cognitive impairment occurs in antipsychotic-naive first-episode psychosis (FEP), but antipsychotics confound interpretation of the longitudinal course of cognition. The primary aim was to disentangle the effects of illness from antipsychotics on cognition over the first 6-months of FEP treatment. Methods: Randomised, triple-blind placebo-controlled trial (Staged Treatment and Acceptability Guidelines in Early Psychosis; STAGES), where cognition was a secondary outcome. Antipsychotic-naive FEP patients were allocated to receive risperidone/paliperidone (N=38) or placebo (N=40) in addition to intensive psychosocial therapy for 6-months. A healthy control group (N=42) was also recruited. A cognitive battery assessing attention, working memory, processing speed, verbal fluency, cognitive control and verbal paired-associate learning and memory was administered at baseline and 6-months. Twelve- and 24-month follow-up was also conducted. Results: Over the 6-month trial period, cognitive performance remained stable (working memory, verbal fluency) or improved (attention, processing speed, cognitive control), with no group-by-time interaction evident. The exception was for verbal paired-associate learning and memory, where a significant group-by-time interaction was observed. The placebo and healthy control groups improved, and the medication group deteriorated on immediate paired-associate recall (p=0.039) and delayed cued recall (p=0.005); effect sizes were medium-to-large. Findings were similar when only trial completers were included in the analysis. Conclusions: Risperidone/paliperidone may cause progression of memory impairment in the early months of FEP. Replication is needed in confirmatory trials. The findings support the need for careful consideration of the risks and benefits of various antipsychotics and the importance of accounting for their cognitive effects in longitudinal research.