Is elevated pre-ictal heart rate associated with secondary generalization in partial epilepsy?

Article Outline

• Abstract
• 1. Introduction
• 2. Methods
  • 2.1. Subjects
  • 2.2. Data recording and analysis
• 3. Results
• 4. Discussion
• Acknowledgements
• References
• Copyright

Abstract 

Background

People with epilepsy are at risk for sudden unexpected death. Cardiac arrhythmia is one possible mechanism. We have studied seizure-related changes in cardiac rhythm.

Methods

Video-EEG and ECG from 38 patients with epileptic seizures during long-term monitoring for investigation of partial epilepsy with ictal impairment of consciousness were obtained. Seizures were classified as either complex partial or secondarily generalized. Inter-ictal, pre-ictal, ictal and post-ictal heart rate was calculated for the first recorded seizure.

Results

Heart rate during the pre-ictal period was higher (p=0.016) in patients with secondarily generalized seizures (n=11) compared to patients with complex partial seizures (n=27). Heart rate was also elevated during and after generalized seizures (p<0.015). Inter-ictal heart rate was not different in patients with secondary generalization compared to patients with partial seizures.

Conclusion

We report elevated heart rate prior to partial seizure onset in those attacks which become secondarily generalized compared to seizures which remain localized. The finding may be relevant for the understanding of sudden death in epilepsy.

Keywords: Autonomic nervous system, Temporal lobe, SUDEP, Seizure prediction

Back to Article Outline

1. Introduction 

Various mechanisms may be responsible for sudden unexpected death in epilepsy (SUDEP). Cardiac arrhythmia is one of them.¹ Generalized seizures are a risk factor for SUDEP², ³ and ECG abnormalities are reported to be more frequent during generalized compared to non-generalized partial seizures.⁴ However, it is not known whether SUDEP primarily is related to the ictal event itself, to its consequences or to a combined effect of both.

During a seizure, the sympathetic nervous system is activated as a stress response, and heart rate and blood pressure increase more during generalized seizures compared to seizures which remain partial.⁵ It is also a fact that mental stress alone increases norephineprine levels and facilitates ventricular arrhythmias in patients with a preceding history of ventricular tachycardia.⁶ Whether generalized seizures and ECG abnormalities during seizures are two independent risk factors for SUDEP may thus be difficult to ascertain as ECG abnormalities may be induced by seizure-related sympathetic activity.

Autonomous manifestations may in different ways be related to epileptic attacks. They may be part of prodromes and the ictus itself and they may occur as consequences of seizure activity. Furthermore, emotional stress is the most frequently reported seizure-precipitating factor.⁷

Why some partial onset seizures generalize and others do not is not well understood. More insight into the mechanisms of generalization of partial seizures may improve treatment, and possibly also contribute to a reduced number of SUDEPs. The aim of this study was to compare seizure-related changes in cardiac rhythm in partial seizures with and without generalization.

Back to Article Outline

2. Methods 

2.1. Subjects 

A total of 254 patients were investigated for epilepsy with long-term (>24h) continuous EEG monitoring with video from January 2004 to February 2008. Patients with evident partial seizures with ictal impairment of consciousness during monitoring were included (40 patients). Two subjects were excluded due to technical problems. Thirty-four patients used more than one antiepileptic drug, and drug tapering was performed in 20 patients in order to induce seizures. Twelve patients had symptomatic epilepsy, four patients had autosomal dominant nocturnal frontal lobe epilepsy, and 22 patients had probable symptomatic epilepsy. If more than one seizure occurred during registration, the first was used for classification and analysis. Seizures were classified as complex partial seizures (CPS); n=27) or secondarily generalized tonic–clonic seizures (SGTC; n=11). Seizure onset was defined as either (1) start of clinical seizure symptoms (as recorded by the patient) or signs (observed by video), usually associated with a definite change in background EEG activity amplitude and/or frequency (often with associated EMG-activity) or (2) onset of an electrographic seizure pattern.⁸ Seizure offset was defined as the end of the video-encephalographic clinical seizure. Patients had a seizure pattern and/or ictal/post-ictal slowing suggesting that seizures originated from the frontal lobes (14), temporal lobes (22), parietal lobes (1), or were multifocal (1). Distinct laterality of seizure onset could be determined in only 29 patients (left: 11/right: 18). The study group consisted of 27 women and 11 men, mean (SD) age 33 (13) years, with first seizure at average 12 (10) years with no difference in age between men and women (t(36)=−1.36, p=0.36). In the 12 patients with symptomatic epilepsies, nine had focal parenchymal brain MRI abnormalities, all corresponding to the localization of seizure onset. Inter-ictal epileptiform abnormalities were present in 30 patients with a localization corresponding to the seizure origin in 21. Patient characteristics are presented in Table 1.

Table 1. Patient characteristics.

Gender	Age (years)	Epilepsy duration (years)	Lateralisation	Localization	Syndrome	Seizure frequency	Heart rate
Seizure type							Pre-ictall	Ictal
Complex partial (n=27)
aFemale	52	≥20	Left	Frontal	Idiopatic	>1 per month	66	109
Female	57	≥20	Right	Temporal	Cryptogenic	>1 per week	81	104
aFemale	33	10–20	Right	Temporal	Symptomatic	>1 per week	69	82
Female	51	≥20	Right	Temporal	Cryptogenic	>1 per month	51	153
Female	13	10–20	No	Multifocal	Symptomatic	>1 per week	121	162
Female	19	10–20	No	Frontal	Cryptogenic	>1 per week	59	104
aFemale	38	≥20	Right	Temporal	Cryptogenic	>1 per week	60	60
aMale	27	≥20	Left	Frontal	Symptomatic	>1 per week	60	144
aMale	54	10–20	No	Frontal	Cryptogenic	>1 per week	54	89
aMale	19	5–10	Right	Temporal	Cryptogenic	>1 per month	51	110
aMale	37	≥20	Right	Temporal	Symptomatic	>1 per month	102	103
aFemale	33	10–20	Right	Frontal	Symptomatic	>1 per month	62	99
Female	30	≥20	No	Frontal	Idiopatic	>1 per week	68	95
aMale	20	10–20	Right	Temporal	Cryptogenic	>1 per month	68	79
Female	21	5–10	Right	Temporal	Symptomatic	>1 per week	63	65
Female	28	≥20	No	Frontal	Idiopatic	>1 per week	50	65
Female	9	5–10	Left	Temporal	Symptomatic	>1 per week	68	98
aFemale	33	5–10	No	Temporal	Cryptogenic	>1 per month	85	138
aFemale	45	≥20	Left	Temporal	Cryptogenic	>1 per week	86	91
aFemale	24	10–20	No	Frontal	Cryptogenic	>1 per week	74	91
Female	24	10–20	Left	Temporal	Cryptogenic	>1 per week	92	b
Male	50	≥20	No	Frontal	Idiopatic	>1 per week	64	59
Female	33	≥20	Right	Frontal	Cryptogenic	>1 per week	64	114
aMale	21	5–10	Left	Frontal	Cryptogenic	>1 per week	70	b
aFemale	22	5–10	Right	Temporal	Cryptogenic	>1 per month	67	112
aFemale	44	≥20	Right	Frontal	Symptomatic	>1 per week	69	72
aFemale	28	10–20	Left	Temporal	Cryptogenic	>1 per week	75	99
Secondarily generalized (n=11)
Female	23	10–20	No	Frontal	Cryptogenic	>1 per week	76	165
aFemale	68	≥20	Right	Temporal	Cryptogenic	>1 per week	136	130
Male	33	≥20	Right	Temporal	Cryptogenic	>1 per week	109	132
Male	24	≥20	Right	Temporal	Cryptogenic	>1 per week	158	140
aFemale	41	≥20	Right	Parietal	Symptomatic	>1 per week	96	132
Male	47	≥20	Right	Temporal	Cryptogenic	1-11 per year	52	103
Female	39	≥20	Left	Temporal	Symptomatic	>1 per month	56	78
aFemale	21	5–10	Left	Temporal	Cryptogenic	>1 per month	84	127
Male	40	<5	Left	Temporal	Cryptogenic	>1 per month	102	b
Female	29	≥20	Left	Frontal	Symptomatic	>1 per week	77	169
aFemale	36	10–20	Right	Temporal	Symptomatic	>1 per week	52	92

Sixty-eight percent of subjects had seizures once a week or more (CPS/SGTC: 70/64%); 29% had seizures more than monthly but less than weekly (CPS/SGTC: 30/27%) and three percent (one subject) had seizures more than once a year but less than monthly (10% of SGTC subjects; see also Table 1), and seizure frequency was not different between groups (Pearson's Chi-square test, χ (1)<0.001, p=0.98).

Simple partial seizure onset (aura) was a common phenomenon in 20 patients (Table 1: CPS/SGTC: 76/44%). Seven patients (CPS/SGTC: 18/19%) had seizures only during sleep and two patients (both CPS) could not report the presence of habitual auras due to cognitive deficits. Aura symptoms and consciousness were not systematically recorded during monitoring and hence only video-encephalographic seizure onset could be accurately assessed in the present study.

2.2. Data recording and analysis 

Digital EEG (Nervus 3.0 with M40 amplifier) was recorded with a common reference at a sampling frequency of 256Hz. Twenty-one scalp electrodes were placed according to the 10/20 international system, although ear references were replaced by anterior temporal electrodes T1 and T2. Eye movements were recorded by horizontal (T1–T2) and vertical (Fp2-Right cheek) bipolar channels. Single lead ECG was recorded (sampling frequency 256Hz, low-frequency filter 0.5Hz, high-frequency filter 70Hz) with active electrodes across the upper part of the chest and exported for further analysis. Heart rate was calculated (Chart Pro; ADInstruments Pty Ltd) based on the 30s with as little artefacts as possible from the 2min before seizure onset (pre-ictally), during seizure, and from the 2min after seizure offset, as well as inter-ictally. Inter-ictal recordings were obtained from periods during the same examination with the subjects in the same state as during the start of the seizure (e.g. during sleep if seizure started during sleep) at least 12h after a seizure and at least 1h before a seizure. Each ECG signal was manually inspected for artefacts and the RR interval detections were manually edited if necessary. Ictal ECG was disturbed by artifacts in one SGTC patient. In two CPS patients with attack durations <30s ictal ECG could not be analysed. In five SGTC patients with early generalization the ECG-analysis had to be performed later in the attacks (63–363s after onset).

All reported variables are distributed normally (One-Sample Kolmogorov–Smirnov). One-way ANOVA was used for between-groups comparison of heart rate. General linear models (multiple univariate ANOVAs with pre-ictal heart rate as dependent variable and seizure type as fixed factor) were applied in order to test for possible confounders by sequentially applying the variables: gender, age, use of carbamazepine, polytherapy, drug tapering, duration of epilepsy, seizure frequency, lateralisation (left vs. right), region of epileptic focus (frontal vs. temporal), or sleep/awake state as random factors or covariates. Student's t-test was used for comparison age between men and women. Pearson's Chi-square test was used for comparison of seizure frequency and for the occurrence of habitual aura between groups. Analysis was performed using SPSS version 15. Two-sided p-values <0.05 are reported as significant.

Back to Article Outline

3. Results 

Heart rate increased with more than 10beats/min from the pre-ictal to the ictal period in 74% of the patients (CPS 74%, GTC 73%; see also Table 1).

Heart rate in the pre-ictal period was higher in patients with SGTC compared to patients with CPS. Heart rate was also elevated during and after generalized seizures, but not inter-ictally (Table 2 and Fig. 1).

Table 2. Seizure-related heart rate.

Epoch	Seizure types		Test statisticsa
	Complex partial Mean (SD)a	Secondarily generalized Mean (SD)
Inter-ictal (beats/min)	65.6 (8.6)	65.4 (4.6)	F(1,9)=0.003, p=0.96
Pre-ictal (beats/min)	70.3 (15.9)	90.7 (34.3)	F(1,36)=6.3, p=0.016
Ictal (beats/min)	99.9 (27.6)	126.9 (29.1)	F(1,33)=6.7, p=0.015
Post-ictal (beats/min)	87.9 (26.3)	121.3 (24.3)	F(1,34)=12.1, p=0.001

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Seizure-related heart rates. *p<0.016.

No interaction with gender, age, use of carbamazepine, polytherapy, drug tapering, duration of epilepsy, seizure frequency, lateralisation (left vs. right), region of epileptic focus (frontal vs. temporal), or sleep/awake status when seizure started was found, and neither factors contributed significantly when added to the model. However, drug tapering contributed with a trend when included as a random factor in the ANOVA model (F(1,34)=111.2, p=0.060).

There was a trend towards increased frequency of habitual aura in the patients with CPS compared to SGTC (Pearson Chi-square=3.7, p=0.056). There was no difference in pre-ictal heart rate when comparing patients with and without habitual auras (F(1,28)=0.60, p=0.50; patients with seizures during sleep and patients with cognitive deficits excluded).

Moreover, subjects with frontal lobe seizures had higher inter-ictal heart rate and a trend towards lower pre-ictal heart rate compared to subjects with temporal lobe seizures (Table 3).

Table 3. Seizure-related heart rate in frontal vs. temporal lobe seizures.

Epoch	Seizure onset		Test statisticsa
	Frontal Mean (SD)	Temporal Mean (SD)
Inter-ictal (beats/min)	71.1 (3.5)	61.7 (6.5)	F(1,9)=6.8, p=0.031
Pre-ictal (beats/min)	65.2 (7.8)	80.3 (27.9)	F(1,35)=3.9, p=0.057
Ictal (beats/min)	105.7 (35.0)	104.9 (25.8)	F(1,32)=0.007, p=0.93
Post-ictal (beats/min)	92.7 (27.3)	97.1 (30.4)	F(1,33)=0.182, p=0.67

Back to Article Outline

4. Discussion 

We found increased pre-ictal heart rate in subjects with SGTC compared to those with seizures which remained localized. Sinus tachycardia during generalized seizures is not surprising, but a pre-ictal difference between CPS and SGTC has to our knowledge not been reported before. Furthermore, the findings indicate that increased heart rate is an integral part of the ictal event which may manifest before seizure onset, particularly in SGTC, and not merely a secondary effect of convulsions. This is in line with experiments performed half a century ago with pentylenetetrazol-induced seizures in patients with temporal lobe epilepsy which demonstrated that heart rate and skin resistance may change before seizures are clinically or electroencephalographically evident.⁹ Later, Opherk et al.⁴ have reported higher heart rate during seizures and more potentially serious ECG abnormalities in patients with generalized seizures compared to patients with partial seizures (with or without impairment of consciousness), but did apparently not investigate the time period immediately before seizure onset. Novak et al.¹⁰ investigated heart rate variability using a method with high spectral time resolution (time–frequency mapping) before, during and after partial seizures in 12 patients with temporal lobe epilepsy and found a reduction in respiration-dependent heart rate variability (suggestive of parasympathetic withdrawal) starting approximately 30s before the seizure was evident in the EEG. In contrast to these and the present findings, Delamont et al.¹¹ found elevated parasympathetic/vagal activity in the pre-ictal period in 10 patients with secondarily generalized partial epilepsy. However, they did not report the heart rate explicitly, and due to methodological limitations (e.g. an unequal number of seizures per patient were studied) it is difficult to compare the results directly with our study. However, the percentage of patients with increase in heart exceeding 10beats/min was similar in the present study and the study of Zijlmans et al, although the compared periods were somewhat differently defined.¹²

The immediate and intuitive explanation for the increased pre-ictal heart rate is that the seizure activity involves central autonomic regions (e.g. insula, amygdala, hypothalamus) before the seizure is evident. Accordingly, increased heart rate before the emergence of EEG changes appears to be more common in seizures arising in the mesial temporal lobe compared to the lateral¹³, ¹⁴ or in extratemporal seizures.¹⁵ However, one study reports that pre-ictal heart rate changes are more frequent in lateral than in mesial temporal lobe seizures.¹⁶ We could not assess the effects of different seizure onset localization within the temporal lobe in the present study, but frontal vs. temporal lobe onset did not influence the main finding. However, the number of generalized frontal lobe seizures in this study was small. Our study contrasts the finding of the two abovementioned studies from Vienna reporting no significant difference between CPS and SGTC regarding pre-ictal heart rate changes.¹⁵, ¹⁶

One may point out that increased heart rate pre-ictally may be related to aura symptoms. Regrettably, accurate identification of auras in the form of simple partial onset without objective ictal signs could not be obtained during the present monitoring. However, we found a trend towards less habitual auras for SGTC patients compared to CPS patients. This finding makes it less likely that aura symptoms explain the increased pre-ictal HR for SGTC patients.

Whether or not secondary generalization may occur in a partial seizure may be dependent on various pre-ictal factors. For example, motor cortical excitability as measured with transcranial magnetic stimulation, is reported to be increased pre-ictally contralateral to the epileptic foci in SGTC, but not in partial seizures without generalization.¹⁷ It can be hypothezised that increased heart rate prior to generalization is a sign of a more widespread limbic or brain stem neuronal process which increases heart rate and also increases the neuronal excitability in the area and networks around the primary epileptic focus. Brain stem noradrenergic neurons are known to modulate the excitability of spinal motoneurons¹⁸ and one may speculate whether similar mechanisms exist for cortical neurons. On the contrary, most animal studies indicate that the noradrenergic system has an anticonvulsive effect.¹⁹, ²⁰ For example, one animal study indicates that the noradrenergic fibres from locus coerulus are necessary for the anticonvulsive effect of vagus nerve stimulation.²¹ It is accordingly more probable that increased pre-ictal heart rate reflects a “top-down” process. Patients with primarily more instability and activation in cortical networks surrounding the seizure onset zone, which probably involve limbic areas, will simultaneously increase sympathetic tone and increase the risk for generalization of the seizure.

The present findings may be of relevance for the understanding of SUDEP, but apart from seizure generalization there was no significant interaction with various known risk factors.², ³ The trend towards more elevated pre-ictal heart rate in patients undergoing drug tapering is interesting, as non-adherence to treatment is among the risk factors. It is known that abrupt antiepileptic drug withdrawal can be associated with increased sympathetic tone during sleep and the occurrence of cardiac arrhythmias.²², ²³ Reducing sympathetic or increasing vagal activity may possibly be a rational approach for preventing SUDEP. The potential of beta-blockers has been discussed by Nei et al. who reported higher heart rate during generalized seizures among patients who later died of SUDEP.²⁴ Information from measurements of sympathovagal activity (e.g. heart rate) may also prove to be useful for the development of seizure prediction algorithms,²⁵, ²⁶ particularly to predict generalization.

The study has some limitations. The short pre-ictal epoch did not allow appropriate standardized heart rate variability analysis. Furthermore, the ECG should ideally be sampled with standardized 12-lead setup in order to study changes related to sympathovagal modulation of e.g. QT-interval changes before, during and after seizures.²⁷ The findings should be replicated in independent samples, preferably with paired design comparing CPS with SGTC in the same subjects. Furthermore, extended ECG recordings in a larger number of patients could possibly give more clues as to whether pre-ictal autonomic changes are associated with prodromes or seizure triggering mechanisms, or if they represent early ictal autonomic activation. Nevertheless, our study suggests that autonomic changes prior to ictal onset are more pronounced in seizures which become secondarily generalized compared to seizures which remain localized.

Back to Article Outline

Acknowledgements 

We thank Anne Grete Eggen for skilful technical assistance and John Wilson for valuable comments on the manuscript.

Back to Article Outline

References 

1. Schuele SU. Effects of seizures on cardiac function. Journal of Clinical Neurophysiology. 2009;26:302–308
   • View In Article
2. Tomson T, Nashef L, Ryvlin P. Sudden unexpected death in epilepsy: current knowledge and future directions. Lancet Neurology. 2008;7:1021–1031
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (183 KB)
   • CrossRef
3. Nashef L, Hindocha N, Makoff A. Risk factors in sudden death in epilepsy (SUDEP): the quest for mechanisms. Epilepsia. 2007;48:859–871
   • View In Article
   • MEDLINE
   • CrossRef
4. Opherk C, Coromilas J, Hirsch LJ. Heart rate and EKG changes in 102 seizures: analysis of influencing factors. Epilepsy Research. 2002;52:117–127
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (177 KB)
   • CrossRef
5. Wannamaker BB. Autonomic nervous system and epilepsy. Epilepsia. 1985;26(Suppl. 1):S31–39
   • View In Article
   • CrossRef
6. Lampert R, Jain D, Burg MM, Batsford WP, McPherson CA. Destabilizing effects of mental stress on ventricular arrhythmias in patients with implantable cardioverter-defibrillators. Circulation. 2000;101:158–164
   • View In Article
7. Nakken KO, Solaas MH, Kjeldsen MJ, Friis ML, Pellock JM, Corey LA. Which seizure-precipitating factors do patients with epilepsy most frequently report?. Epilepsy Behavior. 2005;6:85–89
   • View In Article
8. Chatrian G, Bergamini L, Dondey M, Klass D, Lennox-Buchtal M, Petersén I. A glossary of terms most commonly used by clinical electroencephalographers. Electroencephalography and Clinical Neurophysiology. 1974;37:538–548
   • View In Article
   • Full-Text PDF (1052 KB)
   • CrossRef
9. Van Buren JM, Ajmone-Marsan C. A correlation of autonomic and EEG components in temporal lobe epilepsy. Archives of Neurology. 1960;3:683–703
   • View In Article
   • MEDLINE
10. Novak V, Reeves AL, Novak P, Low PA, Sharbrough FW. Time-frequency mapping of R-R interval during complex partial seizures of temporal lobe origin. Journal of the Autonomic Nervous System. 1999;77:195–202
    • View In Article
    • MEDLINE
    • CrossRef
11. Delamont RS, Julu PO, Jamal GA. Changes in a measure of cardiac vagal activity before and after epileptic seizures. Epilepsy Research. 1999;35:87–94
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (89 KB)
    • CrossRef
12. Zijlmans M, Flanagan D, Gotman J. Heart rate changes and ECG abnormalities during epileptic seizures: prevalence and definition of an objective clinical sign. Epilepsia. 2002;43:847–854
    • View In Article
    • MEDLINE
    • CrossRef
13. Di Gennaro G, Quarato PP, Sebastiano F, Esposito V, Onorati P, Grammaldo LG, et al. Ictal heart rate increase precedes EEG discharge in drug-resistant mesial temporal lobe seizures. Clinical Neurophysiology. 2004;115:1169–1177
    • View In Article
14. Mayer H, Benninger F, Urak L, Plattner B, Geldner J, Feucht M. EKG abnormalities in children and adolescents with symptomatic temporal lobe epilepsy. Neurology. 2004;63:324–328
    • View In Article
15. Leutmezer F, Schernthaner C, Lurger S, Potzelberger K, Baumgartner C. Electrocardiographic changes at the onset of epileptic seizures. Epilepsia. 2003;44:348–354
    • View In Article
    • MEDLINE
    • CrossRef
16. Schernthaner C, Lindinger G, Potzelberger K, Zeiler K, Baumgartner C. Autonomic epilepsy—the influence of epileptic discharges on heart rate and rhythm. Wiener Klinische Wochenschrift. 1999;111:392–401
    • View In Article
    • MEDLINE
17. Badawy R, Macdonell R, Jackson G, Berkovic S. The peri-ictal state: cortical excitability changes within 24h of a seizure. Brain. 2009;132:1013–1021
    • View In Article
    • CrossRef
18. Heckman CJ, Johnson M, Mottram C, Schuster J. Persistent inward currents in spinal motoneurons and their influence on human motoneuron firing patterns. Neuroscientist. 2008;14:264–275
    • View In Article
    • CrossRef
19. Weinshenker D, Szot P. The role of catecholamines in seizure susceptibility: new results using genetically engineered mice. Pharmacology and Therapeutics. 2002;94:213–233
    • View In Article
20. Giorgi FS, Pizzanelli C, Biagioni F, Murri L, Fornai F. The role of norepinephrine in epilepsy: from the bench to the bedside. Neuroscience and Biobehavioral Reviews. 2004;28:507–524
    • View In Article
    • MEDLINE
    • CrossRef
21. Krahl SE, Clark KB, Smith DC, Browning RA. Locus coeruleus lesions suppress the seizure-attenuating effects of vagus nerve stimulation. Epilepsia. 1998;39:709–714
    • View In Article
    • MEDLINE
    • CrossRef
22. Kenneback G, Ericson M, Tomson T, Bergfeldt L. Changes in arrhythmia profile and heart rate variability during abrupt withdrawal of antiepileptic drugs. Implications for sudden death. Seizure. 1997;6:369–375
    • View In Article
    • Abstract
    • Full-Text PDF (687 KB)
    • CrossRef
23. Hennessy MJ, Tighe MG, Binnie CD, Nashef L. Sudden withdrawal of carbamazepine increases cardiac sympathetic activity in sleep. Neurology. 2001;57:1650–1654
    • View In Article
    • MEDLINE
24. Nei M, Ho RT, Abou-Khalil BW, Drislane FW, Liporace J, Romeo A, et al. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia. 2004;45:338–345
    • View In Article
    • MEDLINE
    • CrossRef
25. Lehnertz K, Mormann F, Osterhage H, Muller A, Prusseit J, Chernihovskyi A, et al. State-of-the-art of seizure prediction. Journal of Clinical Neurophysiology. 2007;24:147–153
    • View In Article
26. Kerem DH, Geva AB. Forecasting epilepsy from the heart rate signal. Medical and Biological Engineering and Computing. 2005;43:230–239
    • View In Article
27. Frederiks J, Swenne CA, Kors JA, van Herpen G, Maan AC, Levert JV, et al. Within-subject electrocardiographic differences at equal heart rates: role of the autonomic nervous system. Pflügers Archiv European Journal of Physiology. 2001;441:717–724
    • View In Article
    • MEDLINE
    • CrossRef