Class-specific and cumulative effects of psychotropic medication on sleep architecture: a polysomnographic retrospective study.
Abstract
Objective
This study aims to investigate the impact of psychotropic medication interactions on sleep architecture. With the growing use of psychotropics, including antidepressants, antipsychotics, and mood stabilizers, this research seeks to understand how these drugs and their cumulative burden may affect key sleep parameters. MATERIAL AND
Methods
This retrospective study analyzed 60 non-psychotropic medication users, as well as 61 psychotropic medication users who underwent hospital polysomnography and met all inclusion criteria. Sociodemographic and polysomnographic data were obtained from a sleep laboratory database, between January 2018 and December 2021. Statistical analysis included normality assessment and appropriate parametric or non-parametric tests for group comparisons. Multivariable general linear models adjusted for age, sex, and BMI with HC3 robust errors evaluated associations. Cumulative psychotropic burden was analyzed as a continuous predictor.
Results
Medicated individuals exhibited reduced sleep efficiency (p=.001), prolonged sleep onset and REM latency (p=.001), increased light sleep, and fewer sleep cycles. Multivariable analyses showed antidepressants markedly increased REM latency (p=.001) and periodic limb movements (p=.015), while anticonvulsants enhanced slow-wave sleep (p=.013) and antipsychotics reduced arousals (p=.004). A dose-response relationship was observed, with greater psychotropic burden independently associated with longer REM latency (p=.029) and reduced arousal index (p=.001).
Conclusion
Psychotropic medications alter sleep architecture in recognizably class-specific ways, and these effects compound as the number of concurrent agents grows. Clinically, this argues for treating total psychotropic burden as a relevant variable in the interpretation of polysomnography, and underscores the need for prospective studies that account for dosage, serum levels, and cognitive outcomes.