Abstracts

Rationale: Assessment of heart rate variability (HRV) via spectral analysis can used to measure the autonomic cardiac function. One can separately measure the sympathetic and parasympathetic components of the spectrum. Reduced HRV has been shown to predict increased mortality and risk of sudden death, including patients with epilepsy. Previous studies have typically looked at 5 minute Epochs over a 24 hour period or an isolated 1 hour period. In this study we wanted to ascertain whether the data collected from one 5 minute epoch of EEG data during a standardized time could be used to differentiate the autonomic cardiac function between patients with epileptiform abnormalities on EEG and those who do not. Methods: We retrospectively reviewed the baseline portion of 39 patient monitored in our adult epilepsy monitoring unit along with the heart rate data which is recorded with the EEG. We chose to review a five minute portion of the ECG during the baseline without activating procedures. The R-R intervals were then analyzed and the heart rate variability was determined via time domain measures and frequency domain analysis.Results: Of the 39 patients analyzed, 20 patients had no epileptiform abnormalities on their EEG monitoring and 19 patients had epileptiform abnormalities or captured seizures on their EEG recording. Of the patients with epileptiform abnormalities on their EEG, 6 of those patients were presurgical evaluations who have medically refractory epilepsy. When the group of patients with no epileptiform abnormalities on their EEG monitoring is compared to the group of patients with epileptiform abnormalities, we see the only statistically significant differences between the groups low frequency (LF) components (p=0.05). When we compare those patient who are assumed to have more refractory epilepsy (i.e. those patients in the presurgical group), to the patients with no epileptiform abnormalities on their EEG monitoring, we continue to see statistically significant differences in LF components (p=0.03). We also see statistical significant differences in R-R variability (SDNN) (p=0.05).Conclusions: Using spectral analysis of heart rate variability, we were able to confirm that there are some components of heart rate variability (LF) which are statistically different in patients without epileptiform abnormalities on EEG and those with epileptiform abnormalities in a short 5 minute epoch. Furthermore, in patients who are assumed to have refractory epilepsy, there continues to be differences from those patients with no epileptiform abnormalities in LF, but we also start to differences in SDNN. We think that this study helps to demonstrate that measurable differences can be seen even in short term recordings. Typically a routine EEG is recorded over 30 minutes. Normally there is at least 5 minutes during that recording when no activating procedures are done. This might offers us a new way of measuring HRV in a larger population without specialized equipment and ultimately help in better ascertaining which epilepsy patient are at increased risk for sudden death.