Autonomic cardiac control in patients with epilepsy : spectral analysis of heart rate variability

The heart is affected by the central nervous system via sympathetic and para-sympathetic efferents from autonomic centres in the brain stem. By assessing the beat to beat variation of the RR-intervals of the heart, i.e. the heart rate variability (HRV) it is possible to separately analyse sympathetic and parasympathetic effects on the heart. With spectral analysis of HRV the variability is separated into different bands, a high frequency band (HF) reflecting mainly parasympathetic control and a low frequency band (LF) reflecting sympathetic control. The total frequency power (TP) reflects all cyclic variations in the recorded period. A reduced HRV is a well documented prognostic factor for sudden death in many conditions and might also be of relevance for sudden death in epilepsy (SUDEP). The risk of sudden death is 24 times higher in people with epilepsy compared to the general population, with highest incidence among candidates for epilepsy surgery and particularly those with a poor surgical outcome (continuing seizures after surgery). Frequent tonic-clonic seizures, polytherapy with antiepileptic drugs and high serum concentrations of the antiepileptic drug carbamazepine (CBZ) have been identified as a risk factor for SUDEP. Our underlying hypothesis for the present project is that persons with epilepsy may have a disturbed autonomic heart control, and that such alterations might be a factor making them susceptible to SUDEP. The general aim was to study heart control with spectral analysis of HRV, based on digital ECG recordings, on patients with epilepsy of different stages and to analyze effects on HRV of different therapeutic interventions. We investigated HRV in patients with newly diagnosed epilepsy (NDE). We found no effect of epilepsy per se but treatment with CBZ significantly reduced LF, HF and TP. We also assessed HRV before epilepsy surgery in patients with refractory temporal lobe epilepsy (TLE) and found that the patients with poor surgical outcome, one year after surgery had reduced HRV already before surgery in contrast to the patients who became seizure free after surgery. This difference might be linked to the higher risk of SUDEP reported in patients with poor surgical outcome. HRV was also assessed after epilepsy surgery in the same patients but we found no effect on HRV by TLE surgery. We also investigated circadian variation of HRV in patients with NDE and refractory TLE and assessed the effect of drug treatment and surgery. In the NDE patients, treatment with CBZ decreased the night/day ratio of TP. Temporal lobe resection in the patients with refractory TLE did not change the night/day ratio. Taken together our results indicate that there is an impaired autonomic cardiac control, reflected in decreased HRV, preferentially among the type of epilepsy patients that have been considered to have a higher risk of SUDEP