Reduced REM sleep: a potential biomarker for epilepsy - a retrospective case-control study

Area of Science:

• Neurology
• Sleep Medicine
• Epilepsy Diagnosis

Background:

• Diagnosing epilepsy can be difficult when seizures or interictal epileptic discharges (IEDs) are not detected.
• Sleep disturbances are common in individuals with epilepsy, with reduced REM sleep percentage (REM%) noted post-seizure.

Purpose of the Study:

• To investigate differences in REM% between individuals diagnosed with epilepsy and those with differential diagnoses.
• To evaluate the predictive value of REM% in epilepsy diagnosis.

Main Methods:

• A retrospective case-control study involving 128 participants undergoing video-EEG monitoring.
• REM% was assessed on the first and last nights of monitoring in an epilepsy monitoring unit (EMU).
• Logistic regression models were used to analyze the predictive performance of REM%.

Main Results:

• The epilepsy group exhibited a significantly lower REM% (12.2%) compared to the control group (17.2%).
• REM% predictive models showed good performance (AUC 0.791), comparable to IED and seizure-based models (AUC 0.853).
• A combined model incorporating REM%, IEDs, and seizures demonstrated superior diagnostic accuracy (AUC 0.933).

Conclusions:

• Reduced REM% is a significant finding in individuals with epilepsy.
• REM% assessment offers a valuable tool to enhance diagnostic accuracy, particularly in ambiguous cases lacking clear EEG or clinical seizure evidence.
• Further prospective, multicentric trials are recommended to validate REM% as a diagnostic biomarker.