Abstracts

Rationale: Seizures oftentimes do not occur randomly but present at certain peaks and troughs throughout the day. The propensity to have a seizure at certain times and/or during wake/sleep may be influenced by individual risk factors. We aim to identify risk factors that predict timing of seizures. Methods: We retrospectively reviewed the charts of 955 patients that underwent video-EEG in our center. Patients ≤21 years with recorded seizures were included. Seizures were evaluated based on EEG localization and their occurrence in 3-hour time bins, during the day (6AM-6PM) or night, and out of wakefulness/sleep. Patients were classified according to age: infants (≤3 years), children (3-12 years), and adolescents (>12-21 years). An analysis utilizing generalized estimating equations with a negative binomial distribution was performed. Models were expanded to test for differences in the main effects by age and adjusted for potential confounders (gender, semiology, MRI lesion and medications). Results: A total of 390 patients (188 girls, mean age 9.2 years, SD 6.0) had 1,754 recorded seizures. Generalized seizures (109 patients, 490 seizures) occurred more out of wakefulness (p<0.001) and more during the day (p<0.001). Infants had more seizures between 6AM-3PM (p<0.05). Children had more seizures between 6-9AM and 12-3PM (p<0.05). Adolescents had more seizures between 6-9AM and 6-9PM (p<0.05). Modeling revealed a greater occurrence of seizures at night with increasing age (p=0.046). Temporal lobe seizures (62 patients, 271 seizures) occurred overall more out of wakefulness (p=0.03). This was specifically observed in children (p<0.05) and adolescents (p<0.01). In infants, seizures were more common between 3-6AM (p<0.05). In the >12 age group, seizures occurred more between 12-3PM (p<0.01). There was no evidence of sleep/wake or day/night seizure variation based on age or age groups in temporal lobe seizures. Frontal lobe seizures (41 patients, 184 seizures) occurred more out of wakefulness in infants (P<0.05) and more out of sleep in adolescents (p<0.0001). Children had more seizures between 9PM-12AM (p<0.01) and 12-3AM (p<0.05) and adolescents had more seizures between 6AM-12PM (p<0.05). Adolescents were 3.6 times more likely to have seizures during sleep compared to other children with frontal lobe seizures (95% confidence interval: 1.8-7.2). Modeling revealed a gradual increase in seizures during sleep with each unit increase in age (p=0.02). This was not affected by potential confounders. All occipital lobe seizures (2 patients, 13 seizures) occurred during wakefulness. Parietal lobe seizures (11 patients, 50 seizures) did not occur in a specific pattern in relation to the sleep/wake or day/night state. Conclusions: Seizures occur in different sleep/wake and diurnal patterns depending on patient age and EEG location. These findings are suggestive of maturational changes in circadian rhythmicity that may alter seizure susceptibility in different age groups. Results may assist in age related and EEG localization based prediction of times of greatest seizure propensity.