Most epilepsies appear to have some diurnal influence, and seizure cycles are patient-specific, robust, and more widespread than previously understood, according to a study published in The Lancet Neurology.

Researchers in this large, retrospective, cohort study used analytic techniques from circular statistics to quantify the prevalence, frequency, and strength of seizure cycles in patients with epilepsy. They collected data from the NeuroVista study (Melbourne, VIC, Australia) and SeizureTracker (United States). Patients from NeuroVista were selected for analysis based on having intractable focal epilepsy and at least 30 clinical seizures. As participants in SeizureTracker are self-selected, patients were chosen for analysis based on having a minimum of 100 seizures. Investigators determined the presence of cycles over multiple time scales using R value (the mean resulting length), and they also determined circular uniformity using the Hodges-Ajne test and the Rayleigh test paired with Monte-Carlo simulations for confirmation.

Of the 1118 SeizureTracker patients, 891 (80%) showed 24-hour circadian modulation of seizure rates, compared with 11 of the 12 (92%) NeuroVista study patients. Seventy-seven to 233 (7% to 21%) of SeizureTracker patients showed strong circaseptan rhythms with a clear 7-day time period, and 151 to 247 (14% to 22%) had significant cycles that lasted more than 3 weeks. In the NeuroVista study, 1 patient out of the total 12 participants had an exact 1-week cycle, 2 others had approximate 1-week cycles, and 2 others had 2-week cycles. Seizure cycles were equally common amongst men and women, with peak seizure rates evenly distributed over the days of the week.

Related Articles

The study investigators explained that the seizure times obtained from patients’ self-reported diaries backed up long-term electrocorticography recordings, which “suggests that diaries can provide a valuable clinical tool to detect and monitor repeating patterns of seizures. The ubiquity of seizure cycles indicates that this is an important clinical phenomenon that affects most patients. The ability to identify unique cyclic patterns in individuals could lead to new opportunities in seizure forecasting tools, as well as improved treatment.”


Karoly PJ, Goldenholz DM, Freestone DR, et al. Circadian and circaseptan rhythms in human epilepsy: a retrospective cohort studyLancet Neurol. 2018:S1474-4422(18)30274-30276.