Abstracts

Rationale: A well-established bidirectional relationship exists between sleep and epilepsy. Inadequate and dysfunctional sleep in patients with epilepsy (PWE) may be a marker of poor seizure control [1]. PWE tend to have less efficient sleep and notably shorter rapid eye movement sleep (REM). Seizures are more likely to arise from sleep transitions and non-rapid eye movement (NREM) sleep than REM [1]. It has been hypothesized that REM may have a suppressive effect on seizures due to cortical desynchronization [2]. Fluctuations in REM duration and REM latency in relation to seizures may provide an avenue to better understand individual seizure incidence and risk.

Methods: A retrospective analysis of the sleep record of 41 patients (mean age, 35.3 ± 10.7 years; mean duration of epilepsy, 17.1 ± 11.5 years; 20 male) with drug-resistant epilepsy was conducted. Patients were admitted to the NIH Clinical Center for epilepsy monitoring and presurgical evaluation. The spectrograms derived from Fast-Fourier (FFT) analysis of electroencephalogram (EEG) data were used to guide sleep-staging of multi-day EEG data. Sleep and seizure variables of interest were recorded and calculated for every 24-hour period. Seizure information and other clinical epilepsy features were recorded for each patient. Linear mixed models were utilized to compare seizure occurrence and sleep measures of interest.

Results: Seizures that occurred the preceding day were related to a significant increase (p=0.002) in REM latency, with the seizure group demonstrating an estimated latency of 137.5 (SE, 10.5) minutes and the no-seizure group showing a latency of 101.2 (SE, 8.3) minutes. Preceding day seizures were also associated with a decrease in REM quantity, with an estimated decrease of 15.5 (SE, 5.4) minutes (p=0.004). Same night seizures were associated with a decrease in REM quantity by 18.6 (SE, 7.1) minutes on average (p=0.008) and an increase (p=0.01) in NREM:REM ratio. We also found that on nights preceding daytime seizure activity, there was a significant decrease in REM quantity (p=0.02), with the seizure group demonstrating an estimated REM quantity of 73.5 (SE, 5.5) minutes and the no-seizure group showing a REM quantity of 86.1 (SE, 4.3) minutes.

Conclusions: A significant decrease in REM quantity was identified in relation to all three times of seizure occurrence-during the preceding, same-night, and the following day. The results are significant because they reveal the potential for identifying REM patterns in concert with seizure occurrence. If determined reliable, these measures may become a realistic clinical tool for patients and physicians to use in longitudinal seizure tracking._x000D_ _x000D_ References:_x000D_ 1. Moore JL, Carvalho DZ, St Louis EK, Bazil C. Sleep and epilepsy: a focused review of pathophysiology, clinical syndromes, co-morbidities, and therapy. Neurotherapeutics. 2021;18(1):170-180._x000D_ 2. Frauscher B, von Ellenrieder N, Dubeau F, Gotman J. EEG desynchronization during phasic REM sleep suppresses interictal epileptic activity in humans. Epilepsia. 2016;57(6):879-888.

Funding: Funding was provided through the NIH Intramural Research Program.