Curr Neurol Neurosci Rep (2011) 11:443–450 DOI 10.1007/s11910-011-0201-3

Biofeedback and Epilepsy Yoko Nagai

Published online: 5 April 2011 # Springer Science+Business Media, LLC 2011

Abstract Biofeedback is a noninvasive behavioral treatment that enables a patient to gain volitional control over a physiological process. As a treatment for epilepsy, biofeedback interventions were explored from as early as the 1970s, concentrating on sensory motor rhythm (SMR) as a neurophysiologic parameter. Whereas SMR biofeedback aims to modulate frequency components of the electroencephalography (EEG), slow cortical potential (SCP) biofeedback (which was introduced in the 1990s) focuses on the regulation of the amplitude of cortical potential changes (DC shift). In its application to epilepsy, biofeedback using galvanic skin response (GSR), an electrodermal measure of sympathetic activity, is a relatively new cost-effective methodology. The present article first reviews biofeedback using SMR and SCP, for which efficacy and neural mechanisms are relatively well characterized. Then recent data regarding promising applications of GSR biofeedback will be introduced and discussed in detail. Keywords Biofeedback . Volitional control . Sensory motor rhythm (SMR) . Slow cortical potentials (SCPs) . Galvanic skin response (GSR)

Introduction Antiepileptic drugs are the mainstay in the management of epilepsy. However, despite optimal drug therapy, approximately 30% of patients continue to have seizures. BehavY. Nagai (*) Brighton and Sussex Medical School, University of Sussex, Falmer, Brighton BN1 9RR, UK e-mail: [email protected]

ioral interventions such as biofeedback have become increasingly available over the past three decades and considered as good adjunctive treatments for drug-resistant epilepsy. Biofeedback enables the explicit display of bodily information or physiological processes that are normally hidden from conscious awareness or control. In this way, biofeedback facilitates the learning of techniques for the volitional regulation of automatic bodily responses. A chosen physiological parameter (e.g., galvanic skin response [GSR] or electrocardiogram) is measured and monitored by attaching a sensor such as a surface electrode to the body. Sensory feedback is generated typically through changes in a visual display or sometimes auditory tones. Through this biofeedback, one can learn how to control volitionally “inner” bodily functions. The targeted bodily function could be one of many physiological parameters to which we do not usually have conscious access (e.g., autonomic processes such as heart beat, blood pressure, or sympathetic skin responses [mainly sweat gland function, indexed electrically through electrodermal measures such GSR] or electroencephalography [EEG] signals [signatures of synchronized intracranial neural activity]). Through biofeedback and associated training methods, people gain awareness of their previously hidden physiological responses, which they can eventually learn how to control intentionally. In epilepsy, biofeedback parameters that have proved most useful include EEG measures of cortical activity, both different EEG frequencies and specific cortical potentials (i.e., neurofeedback), and perhaps less intuitively, peripheral autonomic responses such as respiration and GSR. The general mechanism underlying the efficacy of biofeedback is through enhancement of behavioral selfregulation: control is facilitated by closing the feedback loop. When this is applied to epilepsy, biofeedback has much potential, allowing patients to develop and implement

444

countermeasures to mitigate and terminate the occurrence of seizures. In general, and unlike drug treatments, side effects of biofeedback treatment are rare and acquired biofeedback techniques can be practiced at home. The results of clinical trials in epilepsy have been mostly encouraging, although the sample sizes of currently published studies are usually small.

Biofeedback Treatment Using Central Parameter Sensory Motor Rhythm Biofeedback One early research study of biofeedback treatments in epilepsy was reported in the 1960s. Sterman et al. [1] observed a distinct cortical rhythm (12–14 Hz) appearing over the sensory motor area that was prominently related to motor stillness in the cat. This rhythm was named as sensory motor rhythm (SMR). The same research team later succeeded using operant conditioning to train cats to voluntarily increase this rhythm and enhance motor stillness, eventually suppressing experimentally induced epileptic activity. Following these proof-of-principle studies, the application of SMR biofeedback was extended to humans, and a number of studies by Sterman et al. investigated effects on reducing seizure frequency in patients with epilepsy. The results of these earlier studies are mostly encouraging, although patient sample sizes are relatively small. Clinical reports of SMR biofeedback were predominant in the 1970s and 1980s and the technique has been largely superseded. However, a recent review on the effect of SMR biofeedback for epilepsy and related metaanalyses has renewed interest into the efficacy of biofeedback approaches, including SMR, as a complementary treatment for epilepsy. Sterman [2] conducted a metaanalysis of 24 studies including 243 patients. This indicated that 82% of patients showed more than 50% seizure reduction after SMR biofeedback training. Tan et al. [3] also independently assessed EEG biofeedback studies conducted between 1970 and 2005 for the treatment of epilepsy. Of 63 studies published, they evaluated 10 studies that were most stringently controlled, nine of which used SMR biofeedback (one of them was slow cortical potential [SCP] biofeedback). Seizure reduction was observed in 79% of patients who were involved in the study. It is noteworthy that in most biofeedback studies of this nature, epileptic patients who participate in biofeedback training are typically drug resistant. The durations and protocols for SMR biofeedback training have varied across different studies, but are mostly 60- to 90-min sessions, conducted one to three times per week over a period of 6–24 weeks. A range of EEG frequencies has been tested as feedback parameters for SMR biofeedback (mainly on electrode

Curr Neurol Neurosci Rep (2011) 11:443–450

site C3). For some studies, a combination of SMR enhancement with suppression of frontal slow wave has been used to elicit a successful reduction of seizure frequency (Table 1 [4–26]). SMR is the most commonly studied biofeedback approach targeting EEG frequency in the treatment of epilepsy. However, recent investigations introduced another EEG biofeedback protocol aiming to normalize abnormal focal EEG activity (e.g., slow waves [27]). The neural mechanisms underlying the effect of SMR biofeedback on epilepsy are reasonably well established [28•]. Intercranial recordings in animals suggest that the source of SMR rhythms is the ventrobasal nucleus (nVB) of the thalamus, which relays afferent somatosensory information [29]. During SMR biofeedback, the pattern of firing within nVB changes from a fast nonrhythmic state to systematic rhythmic state. This functional change in nVB activity is related to the suppression of somatosensory information and reduction of muscular tension, ultimately impacting on thalamocortical information flow (for more details see [28•]). Reverberating neural activity within this thalamocortical loop is responsible for generating the SMR, and volitional enhancement of this cortical rhythm using biofeedback may modulate cortical overexcitement to increase in seizure threshold. A recent neuroimaging study demonstrated enhanced activation of anterior cingulate, caudate nucleus, and substantia nigra during performance of a selective attention task in children with attention deficit hyperactivity disorder after about 13 weeks of SMR biofeedback training [30], suggesting increased control over hyperactivity achieved through enhanced engagement of motor control regions. However, although this illustrates a tonic sustained effect of SMR biofeedback, it is uncertain if the same neural mechanisms underlie the effect of SMR biofeedback training in reducing seizure frequency in patients with epilepsy. EEG biofeedback is still far from main stream as a treatment for epilepsy. However, recent advances in computer technology now enable the production of improved biofeedback systems that can use quantitative EEG signals in biofeedback (neurofeedback). An expansion of this treatment option is therefore anticipated. SCP Biofeedback In SMR biofeedback, the EEG parameter was the frequency of a specific cortical rhythm. Another type of EEG biofeedback targets the modulation of DC potential changes: SCPs. Experimentally, the SCP is a type of event-related potential (i.e., evoked response) that usually shows a surface negativity in amplitude and is generated over several seconds after the eliciting event. It is generally agreed that generation of the SCP is related to depolarization of apical dendrites in pyramidal cells and thus indicates

Curr Neurol Neurosci Rep (2011) 11:443–450

445

Table 1 Examples of biofeedback treatment of epilepsy Study

Type of biofeedback

Patients

Results

Sterman and Friar [4] Sterman et al. [5] Sterman and Macdonald [6] Finley et al. [7]

Enhancement of 12–14 Hz

Case report of generalized seizures 4 patients with mixed seizure types 8 patients with mixed seizure types

Reduction in seizures from 1.97/mo to 1.67/mo. Three months of seizure free just after the session Significant seizure reduction in 4 patients

Seifert and Luber [8] Kaplan [9]

Enhancement of 12–16 Hz Enhancement or suppression of 6–9 Hz, 12–15 Hz, and 18–23 Hz Enhancement of 11–13 Hz, suppression of epileptic discharge Enhancement of 12–14 Hz Enhancement of 12–14 Hz Enhancement of 6–12 Hz

Case report of patient with frequent epileptic seizures 6 patients 2 patients for 12–14 Hz 3 patients for 6–12 Hz

Lubar and Bahler [10] Cabral and Scott [11] Quy et al. [12]

Enhancement of 12–14 Hz

8 patients

Enhancement of SMR, desensitization, relaxation Enhancement of 12–16 Hz, suppression of 8–10 Hz, random feedback

3 patients

Lubar et al. [13] Whitsett et al. [14]

Enhancement of 12–15 Hz or 11–19 Hz, suppression of 3–8 Hz Suppression of 3–7 Hz, enhancement of 12–15 Hz, simultaneous suppression of 3–7 Hz, and enhancement of 11–19 Hz

8 patients 8 patients

Fried et al. [15] Helmstaedter et al. [16]

End-tidal CO2

18 patients

Paroxysmal cortical activity

Tozzo et al. [17] Andrews and Schonfeld [18] Rockstroh et al. [19] Swingle [20]

SMR

Case report of patient with complex partial seizures 6 patients

Holzapfel et al. [21]

Regulation of SCP, behavioral therapy

Kotchoubey et al. [22– 24]

Regulation of SCP

3 patients

Alpha band as a part of complex treatment

Regulation of SCP Theta band/SMR ratio

25 patients (data of 18 patients were available) 3 patients Case report of patient with secondary generalized tonicclonic seizure 12 patients (1996)

18 patients (1997) 27 patients (1999)

6/8 patients showed significant reduction in their seizures. Maximum reduction with the enhancement of 12–15 Hz and 18–23 Hz Reduction in seizures by a factor of 10 in clinical seizure rate and significant reduction in percentage of epileptiform discharges 5/6 patients showed reduction in seizure frequency. EEG normalization in several patients No reduction in seizures with biofeedback with 12–14 Hz; however; 2/3 patients showed seizure reduction with biofeedback of 6–12 Hz. Biofeedback setting might have provided nonspecific but new relaxation technique 2/8 patients became seizure free. The other patients had reduced numbers of seizures. Reduction in seizures (all three patients) Significant reduction in seizures in all three patients, but this was not related to any one particular training condition; supported Kaplan’s (1975) observation. 5/8 showed decrease in their mean monthly seizure rate 4/8 patients showed changes in their nocturnal paroxysmal activity. Suppression of 4–7 Hz and enhancement of 8–11 Hz were related to seizure reduction. 10/18 patients showed EEG normalization and reduction in seizures Successful control of seizures

3/6 reduced seizures with biofeedback. 2/6 reduced seizures with relaxation. Successful seizure control in 83% of patients. The results do not attribute success only to alpha band biofeedback 6/18 patients became seizure free. 7/18 patients had reduced seizures Reduction in seizures. Seizure reduction is related to the reduction in theta/SMR ratio Reduction in seizures from 12/week to 7.5/week

Significant reduction in averaged seizures from 9.25/week baseline to 4.33/week post-training phase (1996). Significant reduction in averaged seizures from 2.7/ week baseline to 2.3/week post-training (1997). Significant reduction in averaged seizures from 4.0/ week baseline to 2.9/week during the 6-mo followup period (1999).

446

Curr Neurol Neurosci Rep (2011) 11:443–450

Table 1 (continued) Study

Type of biofeedback

Patients

Results

Kotchoubey et al. [25]

Regulation of SCP

34 patients in SCP biofeedback 11 patients in respiration biofeedback 7 patients in medication group 18 patients (10 biofeedback active and 8 sham control)

SCP biofeedback and medication group significantly decreased seizure frequency

Comparison with respiration Biofeedback and medication Nagai et al. [26]

Galvanic skin response biofeedback, randomized controlled study

6/10 patients in biofeedback group showed more than 50% seizure reduction after 1 mo of biofeedback training. No significant effect in sham control group

EEG electroencephalography; SCP slow cortical potential; SMR sensory motor rhythm.

cortical excitation [31]. There is an association between seizure activity in epilepsy and “spontaneous” slow negative cortical potentials. Chatrian et al. [32] reported that in human, paroxysmal ictal activity is consistently accompanied by a negative DC potential shift. In animal models of epilepsy, a localized cortical negative potential is produced by chemical induction of focal epilepsy [33]. Similarly, parenteral injection of pentylenetetrazol to induce generalized tonic-clonic seizures in mice leads to a negative DC shift with superimposed fast EEG waves [34]. Thus, because SCPs indicate cortical excitability and are associated with epileptogenesis, self-regulation of SCPs to suppress negative potential shift was explored as a means of reducing seizure frequencies in patients with epilepsy. Rockstroh et al. [19] first reported a successful application of SCP biofeedback treatment on drug-resistant patients with epilepsy. Out of 18 patients, six became seizure free, seven reduced seizures, and five patients showed no obvious changes. The results of this clinical trial were subsequently replicated [25, 35, 36]. The variation in treatment duration and protocol across studies are less for SCP biofeedback (measured at Cz) compared with SMR biofeedback. A standard SCP biofeedback treatment is a 60- to 90-min session that includes more than 100 times of SCP biofeedback trials. The duration of the treatment seems to be 5 weeks to 6 months depending on SCP biofeedback alone or combination with behavioral intervention. Transfer sessions (in which patients practice the acquired regulatory skill in the absence of biofeedback) are also typically included. The underlying neural mechanism through which SCP biofeedback can diminish seizure frequency is formulated in terms of a reduction of thalamocortical excitation, hence seizure propagation through a positive cortical potential shift serving as a competitive distracter to epileptic or excitatory activity. At least one neuroimaging study identified deactivation of the paracentral lobule, the superior parietal lobe, the frontal lobe, and the thalamus in the patients with epilepsy who were trained to successfully perform positive

potential shift [37•]. The results are broadly consistent with similar observations in normal subjects [38] with the exception of signal reduction within the thalamus of patients with epilepsy. Ultimately, these findings suggest that voluntary control of cortical positivity can lead to the conditioned inhibition of widespread neural activity.

Biofeedback Treatment Using Peripheral Parameters GSR Biofeedback Although publications on biofeedback treatment for epilepsy are dominated by neurobiofeedback, peripheral biofeedback approaches remain relevant and potentially as powerful as EEG-based approaches. GSR is an electrical measure of centrally mediated changes in the skin, reflecting autonomic nervous influence on sweat gland function. GSR is a measurement of sympathetic tone. Sweat gland innervation is efferent and the effector synapse within the sweat gland is uniquely cholinergic (in contrast with the majority of sympathetic effector synapses, which are adrenergic or noradrenergic). Thus, GSR does not directly reflect tonic circulating levels of adrenaline and noradrenaline. Sympathetic neural activity is related to emotional and cognitive states and thus GSR is widely used as a sensitive index of autonomic bodily arousal related to emotion and attention [39]. Nagai et al. [40] investigated the relationship between peripheral and central state of arousal using, respectively, the GSR and the SCP, which was previously described as a measurement of central arousal. It was found that there is an inverse relationship between GSR and SCP amplitude, such that increases in peripheral sympathetic activity were associated with reductions of SCP amplitude (EEG index of cortical neural excitation, see above). In other words, an increase in peripheral GSR activity may diminish cortical excitation. This empiric observation provided the basis for a GSR biofeedback treatment of epilepsy: First a clinical random-

Curr Neurol Neurosci Rep (2011) 11:443–450

ized controlled trial was conducted with 18 patients with drug-resistant epilepsy [26]. During the treatment phase, patients attended a total of 12 sessions (3 sessions/week). Each session lasted 30 min, during which patients assigned to the active treatment group received true biofeedback, and patients assigned to the sham control group received noncontingent biofeedback. All participants were instructed to actively attend to and change the animation on the computer screen by increasing their level of alertness. The results showed a significant reduction in seizure frequency that was not experienced by the control patients who received sham biofeedback. In the “biofeedback-active” group of patients, a month of GSR biofeedback training was associated with a significant decrease in seizure frequency. Out of 10 patients in the biofeedback group, six patients showed more than a 50% seizure reduction compared with their seizure frequency before the treatment (Fig. 1). The important factor in reducing seizure frequency was an ability of patients to learn biofeedback performance, in other words controlling and enhancing sympathetic GSR arousal level using biofeedback. The performance of biofeedback was correlated positively with subsequent seizure reduction. The theoretic rationale behind this novel approach to the treatment of epilepsy is founded on the fact that negative cortical potentials reflect cortical arousal, which in turn is related to the abnormal cortical activity in epilepsy. Notably, reduction in SCP amplitude is associated with the occurrence of fewer seizures [25, 35, 36]. Most importantly, there is an inverse relationship between the amplitude of a particular SCP (the contingent negative variation; CNV) and peripheral sympathetic arousal that predicated this therapeutic application of GSR biofeedback [40]. A volitional increase in sympathetic arousal using GSR biofeedback elicited a reduction in CNV amplitude. A comprehensive understanding of the precise neural mechanisms awaits further research. The reduction in CNV by biofeedback arousal is likely to be mediated by the regulation of sensory inputs to the cortex via the thalamus, especially the flow of sensory information via thalamocortical circuits, which in turn impact on cortical excitability to modulate seizure threshold. Data from our neuroimaging study of CNV, using a combination of functional MRI (fMRI) and EEG, are consistent with this view and suggest critical involvement of the thalamus and midline dorsal cingulate cortex in generation of the CNV (the experimental SCP) [41]. It is noteworthy that the role of the cingulate cortex was highlighted by earlier neuroimaging studies of autonomic biofeedback, and this region, with related limbic and paralimbic structures, is directly involved in the control of GSR activity and in integrating bodily arousal responses with sensory processing [42, 43]. Thus, it is highly likely

447

that interactions between central and peripheral arousal mediated at the level of anterior cingulate cortex impact on thalamocortical regulation and perhaps underlie the therapeutic effect of GSR biofeedback in reducing seizures in patients with epilepsy. Nagai et al. [44•] also investigated changes in SCP in patients with epilepsy before and after the GSR biofeedback treatment using the CNV paradigm. One month of GSR biofeedback treatment significantly reduced baseline CNV in patients with epilepsy and this phenomena was more prominent in the early CNV wave: CNV is composed of two separable components, an early wave, which is related to orienting response, and a later wave, which is related to motor preparation and expectancy. It is speculated that the decrease in the early wave GSR biofeedback training of patients may be related to a reduction in neural responses to internal and external stimuli that potentially trigger seizure occurrence. Interestingly, a neuroimaging study also identified activity within the ventromedial prefrontal cortex (VMPFC) and the orbitofrontal cortex (OFC) as related to tonic shifts in GSR level during biofeedback where GSR level (i.e., sympathetic arousal) was negatively correlated with activity within these areas [45]. This broad area of prefrontal cortex is part of a set of intercorrelated regions (including medial parietal cortex) termed the default mode network, which is active in awake relaxed states and deactivates with increase in attention and internal and external stimulus [46]. These areas are also related to loss of consciousness during epileptic seizures [47]. The volitional control of neural activity within VMPFC and OFC acquired through GSR biofeedback training may contribute to enhanced management of seizure occurrence in patients with epilepsy by influencing information exchange across distributed brain regions.

Future of Biofeedback Treatment on Epilepsy Awareness of the availability of behavioral interventions has been growing. This has been fuelled in part by growing concern over the long-term side effects of medications on neurologic and psychological health. Thus, biofeedback is increasingly practiced for various clinical conditions beyond epilepsy. Biofeedback gained a reputation only as one style of relaxation therapy. However, recent research on biofeedback including the GSR biofeedback treatment for epilepsy shows a developing diversity in treatment protocols and associated increased complexity in the targeted neural mechanisms believed to underpin reductions in symptoms. As a treatment of epilepsy, biofeedback has been currently constrained to drug-resistant patients. However, this application may usefully expand to help patients with a low tolerance for medication and to children. Exponential advances in computer technology

448

Curr Neurol Neurosci Rep (2011) 11:443–450

Fig 1 Example of individual seizure frequency changes with biofeedback treatment (galvanic skin response biofeedback). Individual seizure frequency change over 28 weeks: patients who showed

response rate greater than 50%. Vertical line indicates actual number of patients’ seizures. (Adapted from Nagai et al. [44•])

increasingly enable more sophisticated and low-cost biofeedback systems. EEG biofeedback (neurofeedback) is widely practiced in the United States and Europe and regular training

courses are available for the therapist. Although much cheaper to implement and probably equally as effective, the clinical use of GSR biofeedback is currently limited compared with

Curr Neurol Neurosci Rep (2011) 11:443–450

neurofeedback. Nevertheless, the easier operation of the equipment and potential for much cheaper individual access is anticipated to lead to growing demand for GSR biofeedback. Despite this, there are some problems to be overcome. First, a large variety of GSR biofeedback equipment is available and the manner in the different machines implement built-in biofeedback is highly variable. The previous study showed apparent linear correlations between reduction of seizure frequency and improvement of performance in patients’ GSR biofeedback technique [25]. It is very important that the equipment provide a good feedback signal and that the patients are properly instructed to obtain maximum effect of the treatment. Standardization of GSR biofeedback equipment and training of biofeedback therapists would be necessary for good clinical outcome. Lastly, I would like to point out current problems of research in biofeedback treatment. As in other behavioral interventions, biofeedback treatment generally takes 1 month to half a year with training visits twice or three times a week. It thus places a high demand on both the patient’s and therapist’s time and commitment. This makes it harder to conduct a wider clinical trial compared with drug trials. The placebo effect can be a target of criticism with any behavioral intervention because it is usually difficult to set up a double-blinded randomized controlled trial in such studies. It is unavoidable that the therapist is aware of which intervention he/she is giving to the patients; however, an alternative procedure may be used to allocate independent assessors to measure study outcomes. Thus, there is more need for data on behavioral aspects of biofeedback treatment. Furthermore, the overall (presently reported) number and sample sizes of biofeedback trials are very small compared with drug trials. It is thus not often possible to identify what type of epilepsy and seizures are more responsive to biofeedback treatment. Finally, biofeedback is currently considered as a “side-effect free” treatment. However, more detailed monitoring in a wider clinical trial will be necessary to clarify the potential longer-term effects of effective biofeedback treatments. Overall, there is a need for more research and better availability of biofeedback interventions as adjunctive strategies for the management of patients with epilepsy.

Conclusions Biofeedback treatment for epilepsy remains on the periphery of management protocols. Although clinical results, accumulated over the past few decades are small, so far findings are promising; it is a fact that there are some patients who certainly obtain a beneficial effect. Although the exact neural mechanisms through which biofeedback strategies reduce seizure frequency are opaque, all types of biofeedback imply influence on thalamocortical regulation with inferred con-

449

sequences of thalamocortical excitability. It is expected that wider clinical trials will help elucidate the overall effects and detailed neural mechanisms of biofeedback on epilepsy, thereby facilitating easier access to nondrug and complementary strategies for optimal management of epilepsy.

Acknowledgment Y. Nagai’s work was supported by the Bial Foundation, the Raymond Way Fund, and Ultrasis Plc. Disclosure Ultrasis Plc owns a patent arising from their support of Y. Nagai’s work, from which Y. Nagai may potentially benefit.

References Papers of particular interest, published recently, have been highlighted as: • Of importance 1. Sterman MB, Wyrwicka W, Roth S. Electrophysiological correlates and neural substrates of alimentary behavior in the cat. Ann NY Acad Sci. 1969;157:723–39. 2. Sterman MB. Basic concepts and clinical findings in the treatment of seizure disorders with EEG operant conditioning. Clin Electroencephalogr. 2000;31:45–55. 3. Tan G, Thornby J, Hammond DC, et al. Meta-analysis of EEG biofeedback in treating epilepsy. Clin EEG Neurosci. 2009;40:173–9. 4. Sterman MB, Friar L. Suppression of seizures in an epileptic following sensorimotor EEG feedback training. Electroencepalogr Clin Neurophysiol. 1972;33:89–95. 5. Sterman MB, Macdonald LR, Stone RK. Biofeedback training of the sensorimotor electroencephalogram rhythm in man: Effects on epilepsy. Epilepsia. 1974;15:395–416. 6. Sterman B, Macdonald LR. Effect of central cortical EEG feedback training on incidence of poorly controlled Seizures. Epilepsia. 1978;19:207–22. 7. Finley WW, Smith HA, Etherton MD. Reduction of seizures and normalization of the EEG in severe epileptic following sensorimotor biofeedback training: Preliminary study. Biol Psychol. 1975;2:189–203. 8. Seifert AR, Lubar JF. Reduction of epileptic seizures through EEG biofeedback training. Biol Psychol. 1975;3:157–84. 9. Kaplan BJ. Biofeedback in epileptics: Equivocal relationship of reinforced EEG frequency to seizure reduction. Epilepsia. 1975;16:477–85. 10. Lubar JF, Bahler WW. Behavioural management of epileptic seizures following EEG biofeedback training of the sensory motor rhythm. Biofeedback Self Regul. 1976;1:77–104. 11. Cabral RJ, Scott DF. Effects of two desensitization techniques, biofeedback and relaxation, on intractable epilepsy: follow-up study. J Neurol Neurosurg Psychiatry. 1976;39(5):504–7. 12. Quy RJ, Hutt JS, Forrest S. Sensorymotor rhythm feedback training and epilepsy: some methodological and conceptual issues. Biol Psychol. 1979;9:129–49. 13. Lubar JF, Shabsin HS, Natelson SE, Holder GS, Whitsett SF, Pamplin WE, et al. EEG operant conditioning in intractable epileptics. Arch Neurol. 1981;38(11):700–4. 14. Whitsett SF, Lubar JF, Holder GS, Pamplin WE, Shabsin HS. A double-blind investigation of the relationship between seizure activity and the sleep EEG following EEG biofeedback training. Biofeedback Self Regul. 1982;7(2):193–209.

450 15. Fried R, Rubin SR, Carlton RM, Fox MC. Behavioral control of intractable idiopathic seizures: I. Self-regulation of end-tidal carbon dioxide. Psychosom Med. 1984;46(4):315–31. 16. Helmstaedter C, Riedel R, Stefan H. Treatment of drug-resistant focal epilepsy using visual orientation activity, EEG- EMG-ZElektroenzephalogr-VErwandte-Geb 1992; 92–95 17. Tozzo CA, Elfner LF, May JG. EEG biofeedback and relaxation training in the control of epileptic seizures. Int J Psychophysiol. 1988;6(3):185–94. 18. Andrews DJ, Schonfeld WH. Predictive factors for controlling seizures using a behavioural approach. Seizure. 1992;1:111–6. 19. Rockstroh B, Elbert T, Birbaumer N, Wolf P, Düchting-Röth A, Reker M, et al. Cortical self-regulation in patients with epilepsies. Epilepsy Res. 1993;14(1):63–72. 20. Swingle PG. Neurofeedback treatment of pseudoseizure disorder. Biol Psychiatry. 1998;44(11):1196–9. 21. Holzapfel S, Strehl U, Kotchoubey B, Birbaumer N. Behavioral psychophysiological intervention in a mentally retarded epileptic patient with brain lesion. Appl Psychophysiol Biofeedback. 1998;23(3):189–202. 22. Kotchoubey B, Streho U, Blankenhorn V, Frescher W, Birbaumer N. Negative potential shifts and the prediction of the outcome of neurofeedback therapy in epilepsy. Clin Neurophysiol. 1999;110:683–6. 23. Kotchoubey B, Blankenhorn V, Frosher W, Streho U, Birbaumer N. Stability of cortical self regulation in epilepsy patients. Neuroreport. 1997;8:1867–70. 24. Kotchoubey B, Schneider D, Schleichert H, Strehl U, Uhlmann C, Blankenhorn V, et al. Self regulation of slow cortical potentials in epilepsy: a retrial with analysis of influencing factors. Epilepsy Res. 1996;25:269–76. 25. Kotchoubey B, Strehl U, Uhlmann C, et al. Modification of slow cortical potentials in patients with refractory epilepsy: acontrolled outcome study. Epilepsia. 2001;42:406–16. 26. Nagai Y, Goldstein LH, Fenwick PB, et al. Clinical efficacy of galvanic skin response biofeedback training in reducing seizures in adult epilepsy: a preliminary randomized controlled study. Epilepsy Behav. 2004;5:216–23. 27. Walker JE, Kozlowski GP. Neurofeedback treatment of epilepsy. Child Adolesc Psychiatr Clin N Am. 2005;14:163–76. 28. • Sterman MB, Egner T. Foundation and practice of neurofeedback for the treatment of epilepsy. Appl Psychophysiol Biofeedback 2008; 31:21–35. This article summarizes an overview of SMR biofeedback as an application of treatment for epilepsy. The review refers to background rationales, basic and clinical research including neural mechanisms and clinical practice of SMR biofeedback. 29. Howe RC, Sterman MB. Cortical-subcortical EEG correlates of suppressed motor behavior during sleep and waking in the cat. Electroencephalogr Clin Neurophysiol. 1972;32:681–95. 30. Lévesque J, Beauregard M, Mensour B. Effect of neurofeedback training on the neural substrates of selective attention in children with attention-deficit/hyperactivity disorder: a functional magnetic resonance imaging study. Neurosci Lett. 2006;394:216–21. 31. Birbaumer N. Effects of hyperventilation on EEG frequency and slow cortical potentials in relation to an anticonvulsant and epilepsy. J Psychophysiol. 1989;3:147–54.

Curr Neurol Neurosci Rep (2011) 11:443–450 32. Chatrian GE, Somasundaram M, Tassinari CA. DC changes recorded transcranially during “typical” three per second spike and wave discharges in man. Epilepsia. 1968;9:185–209. 33. Casper H, Speckman EJ. Cerebral pO2m pCO2 and pH : Changes during convulsive activity and their significant for spontaneous arrest of seizures. Epilepsia 1972; 699–725. 34. Speckmann EJ, Walden J. Anti-epileptic effects of organic calcium channel blockers in animal experiments. In: Schwartzkroin P, editor. Epielpsy: Models mechanisms and concepts. Cambridge University press; 1993; p. 462–486. 35. Kotchoubey B, Schneider D, Schleichert H, et al. Self-regulation of slow cortical potentials in epilepsy: a retrial with analysis of influencing factors. Epilepsy Res. 1996;25:269–76. 36. Strehl U, Kotchoubey B, Trevorrow T, et al. Predictors of seizure reduction after self-regulation of slow cortical potentials as a treatment of drug-resistant epilepsy. Epilepsy Behav. 2005;6:156– 66. 37. • Strehl U, Trevorrow T, Veit R et al. Deactivation of brain areas during self-regulation of slow cortical potentials in seizure patients. Appl Psychophysiol Biofeedback 2006; 31:85–94. This is one recent article on SCP biofeedback on epilepsy. The study explores neural mechanisms of effect of SCP biofeedback in reducing seizures using fMRI. 38. Hinterberger T, Veit R, Strehl U, et al. Brain areas activated in fMRI during self-regulation of slow cortical potentials (SCPs). Exp Brain Res. 2003;152:113–22. 39. Venables PH, Christie MJ. Electrodermal activity. In: Martin I, Venables PH, editors. Techniques in Psychophysiology. John Willey & Sons Ltd; 1980. p. 3–67. 40. Nagai Y, Goldstein LH, Critchley HD, et al. Influence of sympathetic autonomic arousal on contingent negative variation: implications for a therapeutic behavioural intervention in epilepsy. Epilepsy Res. 2004;58:185–93. 41. Nagai Y, Critchley HD, Featherstone E, et al. Brain activity relating to the Contingent Negative Variation (CNV): fMRI investigation. Neuroimage. 2004;21:1232–41. 42. Critchley HD, Elliot R, Mathias CJ, et al. Neural activity relating to the generation and representation of galvanic skin conductance response: a functional magnetic imaging study. J Neurosci. 2000;20:3033–40. 43. Critchley HD, Melmed RN, Featherstone E, et al. Brain activity during biofeedback relaxation: a functional neuroimaging investigation. Brain. 2001;124:1003–12. 44. • Nagai Y, Critchley HD, Rothwell JC et al. Changes in cortical potential associated with modulation of peripheral sympathetic activity in patients with epilepsy. Psychosom Med. 2009; 71:84– 92. This article introduces a theoretical background of an effect of GSR biofeedback on epilepsy. The study investigated the effect of peripheral GSR biofeedback training on central cortical activity changes using EEG in patients with epilepsy. Proposed neural mechanisms are discussed. 45. Nagai Y, Critchley HD, Featherstone E, et al. Brain activity relating to the Contingent Negative Variation (CNV): fMRI investigation. Neuroimage. 2004;21:1232–41. 46. Raichle ME, MacLeod AM, Snyder AX, et al. A default mode of brain function. Proc Natl Acad Sci USA. 2001;16:676–82. 47. Cavanna AE, Monaco F. Brain mechanisms of altered conscious states during epileptic seizures. Nat Rev Neurol. 2009;5:267–76.