Different as night and day: Patterns of isolated seizures, clusters, and status epilepticus.

Abstract

Using approximations based on presumed U.S. time zones, we characterized day and nighttime seizure patterns in a patient-reported database, Seizure Tracker. A total of 632 995 seizures (9698 patients) were classified into 4 categories: isolated seizure event (ISE), cluster without status epilepticus (CWOS), cluster including status epilepticus (CIS), and status epilepticus (SE). We used a multinomial mixed-effects logistic regression model to calculate odds ratios (ORs) to determine night/day ratios for the difference between seizure patterns: ISE versus SE, ISE versus CWOS, ISE versus CIS, and CWOS versus CIS. Ranges of OR values were reported across cluster definitions. In adults, ISE was more likely at night compared to CWOS (OR=1.49, 95% adjusted confidence interval [CI] = 1.36-1.63) and to CIS (OR=1.61, 95% adjusted CI = 1.34-1.88). The ORs for ISE versus SE and CWOS versus SE were not significantly different regardless of cluster definition. In children, ISE was less likely at night compared to SE (OR=0.85, 95% adjusted CI = 0.79-0.91). ISE was more likely at night compared to CWOS (OR=1.35, 95% adjusted CI = 1.26-1.44) and CIS (OR=1.65, 95% adjusted CI = 1.44-1.86). CWOS was more likely during the night compared to CIS (OR=1.22, 95% adjusted CI = 1.05-1.39). With the exception of SE in children, our data suggest that more severe patterns favor daytime. This suggests distinct day/night preferences for different seizure patterns in children and adults.