Reduced rapid eye movement sleep proportion is associated with anxiety and depressive symptoms in newly diagnosed epilepsy

06/08/2026

AI Summary

Lower REM sleep proportion in newly diagnosed epilepsy correlates with greater depressive symptoms and reduced quality of life.
Depressive symptoms partially mediate the association between REM proportion and quality of life, accounting for roughly 52.5% of the total effect.
Anxiety partially mediates REM-depression but does not directly mediate REM-quality of life; REM shows little independent longitudinal QOL effect after adjustment for baseline depression.

Epilepsia. 2026 Aug 5. doi: 10.1002/epi.70413. Online ahead of print.

ABSTRACT

OBJECTIVE: This study was undertaken to examine the associations between rapid eye movement (REM) sleep proportion, depressive and anxiety symptoms, and quality of life (QOL) in newly diagnosed epilepsy (NDE).

METHODS: A total of 104 treatment-naive people with NDE and 50 age-/gender-matched healthy controls were enrolled. Assessments included polysomnography and the Chinese version of the Neurological Disorders Depression Inventory for Epilepsy (C-NDDI-E), Generalized Anxiety Disorder-7, and Quality of Life in Epilepsy Inventory-31 (QOLIE-31). Eighty-eight patients completed 12-month QOL follow-up. Cross-sectional analyses included Spearman correlation and mediation (PROCESS Model 4, 5000 bootstraps). Longitudinal analyses used logistic and linear regression with sequentially added covariates, and least absolute shrinkage and selection operator for variable selection.

RESULTS: Baseline REM sleep correlated negatively with C-NDDI-E (r_s = -.512, p < .001) and positively with QOLIE-31 (r_s = .394, p < .001). C-NDDI-E correlated negatively with QOLIE-31 (r_s = -.461, p < .001). Depressive symptoms partially mediated the REM-QOL association (indirect effect = .253, 52.5% of total effect). Anxiety did not mediate REM-QOL but partially mediated REM-depression. Longitudinally, the REM-12-month QOL association was significant after adjusting for age and gender (odds ratio [OR] = 1.113, p = .005), but not after additionally adjusting for baseline C-NDDI-E (OR = .968, p = .542). No independent linear association remained (B = .087, 95% confidence interval = -.119 to .292, p = .405).

SIGNIFICANCE: Reduced REM sleep is associated with increased depressive severity in NDE. Depressive symptoms mediate both cross-sectional and prospective REM-QOL associations, with little independent effect of REM sleep on QOL. Anxiety does not directly mediate REM-QOL but partially mediates REM-depression.

PMID:42555138 | DOI:10.1002/epi.70413

Document this CPD