Abstracts

Rationale: The second stage of human non-rapid eye movement (NREM) sleep is characterized by two major electroencephalographic (EEG) elements: the K-complex (KC) and the sleep spindle. Although the KC was first described by Loomis et al. 70 years ago, its functional role is still a debatable issue, but is considered by some to be an arousal reaction. Similarly, the function of the sleep spindle also remains undefined, but there are data supporting its role as an arousal inhibitor and thus a sleep preservation mechanism. There are also data supporting a role of spindles in memory consolidation during sleep. A recent paper by Kokkinos and Kostopoulos has shown that sleep spindles are blocked upon spontaneous K-complex coincidence and resume as higher frequency spindles afterwards in normal volunteers. This effect has been explained by a brief arousing stimulus triggering a KC and at the same time blocking spindles by a brainstem inhibitory action at the reticular nucleus (RN) level, upon the depolarizing rebound of which a new spindle rhythm of higher frequency is produced. In this study we wished to determine whether this phenomenon persisted or was altered in patients with epilepsy, and if this could shed additional light on epileptic mechanisms. Methods: Routine EEG studies recorded over a 20 month period were included in this study. We included patients followed at our institution with a definite diagnosis of epilepsy. A sleep spindle had to precede and follow a KC to be analyzed and the KC had to have a minimum duration of 500 msec. Superimposed spindle and KCs followed or preceded immediately by similar patterns (KC/spindle trains) were not analyzed. The frequency of the sleep spindle was calculated with the filter bandpass set at 6-17 Hz. Statistical analysis was performed using the Student's t-test. Results: A total of 350 patient recordings were analyzed. Six patients only were found to have the required superimposed sleep spindle and KC appearance. Only four of these patients had a final diagnosis of epilepsy (with 11 sleep spindles and coincident KC recorded in total from all four patients). Two of these patients had generalized and two had focal onset epilepsy. The mean frequency of the sleep spindles before the KC was 13.568 ± 1.072 standard deviations and 12.182 ± 1.068 after the KC. The difference was found to be significant (t=2.896, 0.01