Abstracts

Rationale: Circadian patterns have been reported in patients with seizures. We evaluated the relationship between seizure occurrence and diurnal and nocturnal changes of temperature and electrodermal activity (EDA) in pediatric patients with epilepsy.Methods: Retrospective study of pediatric patients who were admitted to the video EEG monitoring units at Boston Children’s Hospital and underwent continuous measurement of skin surface temperature, electrodermal activity, and accelerometry. Patients who had an IV line on forearms underwent recordings at the ankles. A visual analysis of the acceleration line allowed identification of the periods with and without forearm movement, which were regarded as periods of wake and sleep, respectively.Results: Video-EEG was performed on average for a duration of 3.30 ± 2.24 days in 33 children (mean age 10.85 ± 7.62years), including 14 (42.42%) girls. Twenty children displayed seizures after the first wake-sleep cycle. Mean temperature (Celsius) and EDA (µS) during wakefulness and sleep of the first wake-sleep cycle (Tw, Ts, EDAw and EDAs ), and the ratios Ts/Tw and EDAs/EDAw for each case were correlated with seizure occurrence over the following days. The means ± S.E. for children with and without seizures after the first day of wristband monitoring were Ts=33.78±1.55 and 34.68±1.58, Tw=32.74±1.64 and 31.35±1.56, Ts/Tw=1.03±0.43 and 1.11±0.46, EDAs=2.01±1.37 and 1.99±2.52, EDAw=1.79±2.18 and 0.69±0.60, and EDAs/EDAw=1.81±1.42 and 4.10±3,23, respectively. The most prominent differences were observed for the Tw(p=0.02), and the ratios Ts/Tw(p<0.001) and EDAs/EDAw(p=0.009).Conclusions: The ratios of mean temperature and EDA during sleep and the preceding wake period may be related to seizures over the following days in children with epilepsy. Analyses of additional confounding factors in larger numbers are in progress. These findings may have potential implications for seizure prediction algorithms, for the management of children with refractory epilepsy, and the understanding of SUDEP-pathophysiology using temperature and EDA as biomarkers of autonomic nervous function.