Abstracts

Rationale: Sleep can affect seizure frequency and presentation, and epilepsy can disrupt the sleep-wake cycle and sleep architecture. There is some data on relationship of sleep to specific epilepsy types, but no data on the relationship of seizures with generalized semiology to diurnal rhythms, especially in children. The aim of this study was to evaluate the relationship of sleep/wake and day/night patterns to generalized EEG seizures. Methods: Charts of 1044 consecutive pediatric epilepsy patients undergoing Video-EEG monitoring (V-EEG) over five years were reviewed: 967 patients were excluded due to focal epilepsy (556), non-epileptic recorded events (217), missing data (125), age over 21 years (59) and no recorded events or seizures (10). Seizure semiology recorded during V-EEG was classified according to the ILAE seizure semiology terminology, and presence of generalized onset of seizures on EEG, and analyzed based on occurrence during day (6am to 6pm) or night and on their relationship to wakefulness and sleep, with calculated occurrence in 3-hour time blocks throughout 24 hours. Statistical analysis analysis was performed with binomial testing. Results: Three-hundred-and-sixteen generalized seizures were analyzed in 77 children. Mean age was 6.4 years 5.4 (range 0-20 years), including 50.6% girls. Tonic and tonic-clonic seizures were more frequently seen in sleep, whereas all other semiologically generalized seizure types occurred more frequently out of wakefulness. Generalized clonic seizures had two peaks: (6-9am) and (noon-3pm) in wakefulness. Absence seizures occurred predominantly in wakefulness, (9am-noon & 6pm-midnight). Atonic seizures occurred predominantly in wakefulness (noon-6pm). Myoclonic seizures occurred in wakefulness (6am-noon). Epileptic spasms had two peaks: (6-9am and 3pm-6pm) in wakefulness. Conclusions: Sleep and wakefulness, as well as time of day/night, are important considerations in proper characterization of generalized seizure-types. Characterization of individual diurnal seizure pattern may offer exciting new possibilities, including sleep/wake pattern variability, EEG or Video-EEG scheduling, and differential (day/night) medication dosing.