Abstracts

Rationale: Memory and sleep complaints are very common among epilepsy patients including those older than 50. Memories once formed need to be consolidated, a process largely dependent on sleep. In the current study, we tried to analyze sleep physiology and predictors of memory consolidation in a cohort of older patients with epilepsy. Methods: We prospectively recruited patients with epilepsy older than 50 undergoing ambulatory EEG from the Brigham and Women’s epilepsy clinic. Patients were asked to perform a declarative memory task consisting of recalling the location of at least 5 of 15 pairs of colored pictures on a 5 x 6 grid prior to their study. They were also asked to wear two wrist electrodermal activity (EDA) sensors (E4, Empatica SRL) during their ambulatory EEG study allowing continuous monitoring of sympathetic activity. Their recall of the picture pairs was reassessed upon their return. Their first night of sleep was then scored by a sleep technologist. For the EDA analysis, the number of EDA peaks was assessed as well as the number and duration of sleep storm events. An EDA storm epoch was defined as a 30 second epoch of EDA during sleep containing a minimum of 3 significant EDA peaks. Storm epochs positioned adjacent or within 10 minutes of each other were clustered together into one storm event. Controls with no neurologic history were also recruited to perform the study. Results: A total of 30 subjects were prospectively recruited, including 10 controls. Mean age of the 20 (50% female) epilepsy patients was 64.5 years (range: 50-91) vs. 61.6 years (range: 51-84) for the 10 controls (40% female).  The epilepsy localization was 75% temporal (9/12 left lateralized) and 20% generalized.  An analysis of the first night of sleep revealed no difference in total sleep duration (6.6 hours vs. 6.7 hours), or different sleep stages. Seventeen subjects completed the memory task, with the 24-hour retention rate for the 8 epilepsy patients at 60% as compared to 63% for the 9 controls, with no difference in the number of trials needed to encode the 5 pairs of pictures. EDA data were available for 16 subjects and a trend for lower storm event numbers in epilepsy vs. control subjects (Left 7.2 vs. 11.7, Right 10.5 vs. 11), total duration of storm events (Left 4670 sec vs. 6234 sec, Right 4301 sec vs. 6917 sec), and EDA peak counts (Left 478 vs. 721, Right 479 vs. 631). In a subset analysis of 14 subjects with data for retention rates and EDA measures, there was a strong correlation between retention rates and the number of sleep storm events on the left (?= +0.84, p=0.0006). Conclusions: As compared to a control population, older patients with epilepsy exhibited similar total sleep duration and sleep stages as well as similar retention rates on a declarative memory task. There was a strong correlation for higher retention rates and a higher number of EDA sleep storm events suggesting that the EDA storm events could represent a marker of “sleep stability” translating into better memory consolidation. Funding: This research was supported by the A.J. Trustey Epilepsy Research Endowed Fund.