REM sleep in schizophrenia: a systematic review and meta-analysis

Abstract

A systematic review and meta-analysis of REM studies in schizophrenia were conducted using published articles researched in major databases within the period from January 1, 1955, to March 15, 2025. REM sleep parameters such as REM time, REM percentage, REM latency and REM density of drug-naive, drug-free and treated schizophrenic patients were analyzed and, where available, compared with case-control data of healthy controls and depressed patients. One hundred ninety studies were identified in the systematic review. Ninety-two case-control studies with 1845 schizophrenic patients, 1269 healthy controls and 262 depressed patients were included in the meta-analyses. The primary outcome was the standard mean difference. Data were fitted with a random-effects model. Publication bias assessment was checked by Egger's Regression and funnel plot asymmetry. Compared to healthy controls, REM time and REM latency were found respectively reduced and shortened in both drug-naive and drug-free schizophrenics, while REM density was found increased in drug-free and treated schizophrenics. No significant difference, compared to healthy control, emerged for REM percentages either in drug-naive, drug-free and treated schizophrenics. Reduced REM latency in drug-free and drug-naive schizophrenics was like that of depressed patients. Antipsychotic treatment appeared to have a normalizing effect on REM time and REM latency but not on REM density: increased density in both drug-free and treated schizophrenics suggest that this variable might be controlled by different mechanisms and/or neurotransmitters than those controlling REM occurrence.