This article brought forth some interesting observations.  First, a major caveat – this study was done in mice.  Mice with a relevant mutation in the Scn1a gene have proven to be a useful model of Dravet, but one must never directly extrapolate any results to humans.  Dividing the day into three 8-hour periods, it was found that the number of SUDEPs that occurred between 6:00 PM and 2:00 AM was the highest – that is, 3.6-fold greater than those between 10:00 AM-6:00 PM, and 2.4-fold greater than those between 2:00 AM-10:00 AM.  Interestingly, mice are nocturnal, so this higher rate of SUDEP was not correlated with more extensive periods of sleep.  It was also found that seizure frequency increased significantly 1–2 days prior to death.

One subset of mice was put on a ketogenic diet, and this treatment reduced mortality rates, from 56% to 14% over the period of observation.   Perhaps surprisingly, no significant differences in seizure frequency or severity were observed between the ketogenic and the control diet groups.

Teran FA, Kim Y, Crotts MS, Bravo E, Emaus KJ and Richerson GB. Time of Day and a Ketogenic Diet Influence Susceptibility to SUDEP in Scn1aR1407X/+ Mice. Front. Neurol. 2019 10:278. doi: 10.3389/fneur.2019.00278.