Temporal distribution of partial seizures: comparison of an animal model with human partial epilepsy.

Seizures do not often strike randomly but may occur in circadian patterns. We compared daily times of partial seizures determined by continuous electroencephalography among patients with mesial temporal lobe epilepsy (MTLE; n = 64), those with extratemporal lobe (XTLE; n = 26) or lesional temporal lobe epilepsy (LTLE; n = 8), and a rat model similar to MTLE in which rats become epileptic after electrically induced limbic status epilepticus (postlimbic status [PLS]; n = 20). Rats were maintained on a 12-hour light/dark cycle with lights on at 0700 hours. The distributions of seizures were fitted by cosinor analysis to determine time of peak seizure incidence +/- 95% confidence interval (95% CI). The mean fraction +/- SD of seizures recorded during light was 63 +/- 17% in PLS animals and 60 +/- 21% in humans. Peak incidence of seizures for PLS rats (547 seizures) was 1645 (95% CI = 1448,1830) and for MTLE subjects (774 seizures) was 1500 (95% CI = 1324,1636). Seizures from XTLE (465 seizures) and LTLE (48 seizures) did not fit a cosinor model and occurred no more frequently during light than dark. In conclusion, limbic seizures in humans and PLS rats occur more often during light than dark and have similar cosinor daily distributions. The chronological similarity between human MTLE and PLS rat epilepsy suggests that limbic seizure occurrence has a relation to the circadian regulatory system.