EEG Biofeedback: A Generalized Approach to Neuroregulation

By Siegfried Othmer, Susan F. Othmer, and David A. Kaiser

To appear in "APPLIED NEUROPHYSIOLOGY
& BRAIN BIOFEEDBACK"
Edited by Rob Kall, Joe Kamiya, and Gary Schwartz

Page 10 of 13

Attention Deficit Disorder-Combined Type
ADHD of the combined subtype has been ranked of slightly greater difficulty than the inattentive and the impulsive subtypes. And even though we are presenting issues in the order of difficulty as shown in Table 2, it is appropriate to take up the matter of ADHD here. The issues are just slightly more complex, and require a somewhat more complicated protocol management. The additional complexity is partially attributed to the comorbidities of ADHD previously mentioned. The prominence of significant comorbidities in the clinical population makes research problematic in that setting.

Protocols may involve training at C3 and C4 (and, historically, Cz) on the sensorimotor strip, with both SMR and beta reward frequencies. They may also involve frontal training at Fz or Fpz, as well as parietal training at P4 or Pz. Left hemisphere and frontal training are more likely to involve the higher frequencies (nominally 15-18 Hz), whereas Pz and right-side training are more likely to involve the lower frequencies (nominally 12-15 Hz). This is consistent with a lower degree of localization in right-hemisphere functions. It is also consistent with current theories of activation and arousal, as previously discussed (the Tucker- Williamson and Malone, Kershner, and Swanson models). This is consistent with the strategy that has emerged in EEG training, namely high frequency training for improved control of activation on the left hemisphere (sometimes with a frontal bias with bipolar montage), combined with lower frequency training on the right hemisphere (sometimes with a parietal bias with bipolar montage).

Migraines

One of the remarkable findings of the past few years is that migraine headaches respond readily to EEG biofeedback training. Efficacy has also been demonstrated for migraines with conventional biofeedback, but there seems to be particular merit in training the brain directly for this vulnerability. Ongoing migraines can sometimes be aborted, or more typically significantly lessened in severity, in thirty minutes of training with the appropriate protocol. In the case of migraines, there are two principal protocols. Choice of the wrong one may often lead to increased migraine pain within a matter of minutes, which motivates a change of course. It is also found that migraines will move from one place in the head to another in response to the training. It may be advantageous to respond to the movement of pain locus for the most effective training.

These prompt responses to the training are concrete evidence that the training is having a specific effect. However, these are the least interesting effects. If the training is pursued long-term, then a propensity toward migraines can be arrested relatively permanently. On the order of twenty to forty training sessions may be required to achieve this objective (absent complicating issues). Moreover, such an outcome is highly predictable. Migraines are extraordinarily responsive to this training. Follow-up data indicate that these gains may be held for several years (that is, for as long as follow-up has been conducted). Barring the happenstance of further trauma, the effects seem permanent. Also, the training efficacy does not appear to depend a great deal on what kind of migraine one is dealing with, classic or common. It is interesting to speculate how this might occur.

Migraines can be seen as a particular form of collective activity of neuronal populations. It is fundamentally a matter of the brain rather than of the vasculature. After all, migraines can be triggered by light stimulation. It is assumed that the effect of such stimulation is on neuronal systems, not on the vascular system. Hence, what happens to the vasculature is consequence, not cause. The light stimulation of a vulnerable brain is assumed to unleash a cascade of collective activity that alters neuromodulator function (serotonin in particular) at the brainstem level. The time constants of such changes in neuromodulator function may be long, but not long enough to account for the duration of migraines. The latter requires some kind of self-reinforcement of the adverse state.

Migraines are characterized by disregulation of central arousal function, which also impinges upon sympathetic and parasympathetic balance. The problem is fundamentally one of instability, for which typical pharmacological agents are not a good answer. The remedy is to increase fundamental stability in the brain, so that the excursion into migraines cannot be as readily triggered. In its role in aborting active migraines, the EEG training may be compellingly promoting a particular state of arousal that stabilizes against the ongoing excursions in arousal level. Even in the case of training at the higher EEG frequencies, reinforcement of an increased EEG amplitude is in effect to reward quiescence. We will return to this theme later.

In training to remediate migraines, sessions are of course preferably conducted during an asymptomatic period. The obvious signposts of whether the correct protocol has been selected may then be unavailable. A general pattern has emerged, however, in which migraines generally require both the higher and lower frequency training to improve stability, with a bias toward the lower-frequency (SMR) training unless the migraines are PMS-related, in which case a bias toward higher-frequency (beta) training prevails. A client may need to keep records of their migraine incidence in order to document the improvement as early as possible to confirm the choice of protocol. Also, migraines are not usually the only symptom affected by the training. The individual will respond favorably to the correct training in other ways, such as improved sleep and mood regulation. In general, if there are adverse consequences in any of a number of areas, an adjustment in protocol is called for.

Panic Attacks

Panic attacks may be considered another paroxysmal brain state in which inappropriate collective activity is subjectively perceived as a panic reaction. It arises out of a matrix of vulnerability to anxiety. Stabilizing the brain against excursions such as panic attacks is quite readily achievable with EEG biofeedback training, and protocol selection is generally straightforward. As in the case of migraines, such stability is difficult to achieve with pharmacological means.

One striking and illustrative case must be mentioned. A woman who had been in treatment for panic anxiety and agoraphobia for ten years, with repeated hospitalizations, long-term psychotherapy, and extensive pharmacological intervention, was eventually given EEG training by the same psychologist who had worked with her for ten years. After only eight sessions, she was able to vacation with her husband in Las Vegas, mixing easily in crowds, and declaring later that she felt anxious only once. On the basis of cases such as these, panic attacks are seen as fundamentally issues of brain instability rather than of psychological state. There may have been psychological underpinnings, but panic susceptibility takes on a life of its own. Onset of panic excursions appear to be chaotic in character, and in its mature form need not have a behavioral antecedent.

Bruxism

Bruxism often responds quite readily to EEG biofeedback training in the beta and SMR domains. Intuitively, bruxism would appear to be a stress reaction, one for which relaxation training might be the appropriate remedy. However, the fact that bruxism is also commonly observed during general anesthesia makes it more reasonable to regard it simply as a consequence of disregulation of arousal, or even of underarousal.

Sterman has proved the direct connection of SMR-training at sensorimotor cortex with motor inactivity in cats, and the identification holds true in primates as well (Sterman, 1978) Hence, SMR-training would appear to be the appropriate remedy. This is generally true, but sometimes an instability in arousal requires beta training also. In the present instance, it is still preferred to regard the process as a normalization of arousal, with whatever frequency training is required to accomplish that objective in a particular instance, and that in consequence of such normalization motor system activation will normalize as well.

In clinical experience, it has been found possible to normalize nocturnal bruxism behavior as it is commonly observed in children with attentional disorders, as well as long-standing conditions in which major restorative dental work has been mandated by the persistent bruxism. This fairly general clinical success supports the hypothesis of bruxism as having a central nervous system origin as opposed to being primarily a disorder attributable to such factors as malocclusion (Parker, 1990, McNeill, 1990).

Hypoglycemia

Hypoglycemia can be regarded for present purposes as disregulation of blood glucose level, which is presumptively also under the management of the central nervous system. One of the common failure modes of a feedback regulatory system is that it can go into damped oscillation. A small stimulus can send the system in to near oscillation, from which it recovers only slowly. In this case, the small stimulus may be a sugar challenge, or even a challenge with a sugar substitute.

It has been found quite generally that conditions of hypoglycemia can be normalized with EEG training in the higher frequency domains of SMR and beta. The measure in this case is simply behavioral. No studies of glucose level after EEG training have been done, to our knowledge. However, it is observed that the cognitive, behavioral, and mood aspects of hypoglycemia remediate with the training, and dietary restrictions can often be abandoned after the training reaches completion. Thus, either the glucose levels have been stabilized through improved regulatory function, or the brain has been made more tolerant to the fluctuations in ambient glucose level.

A case can be made that glucose regulation is directly affected by the training on the basis of comparison with Type I diabetics undertaking the training. A reduction in insulin requirement may be observed in these cases (in which of course the glucose level is being actively monitored). A more stable blood sugar level is implied. It has been observed in some Type II diabetics, for whom dietary management was becoming insufficient, that EEG training could delay indefinitely the onset of insulin replacement therapy.

Sleep Disorders

The field of SMR-beta EEG biofeedback training got its start in the context of sleep studies (Sterman, 1970). Since the early years, it has become clear that one of the first observable data consequent to EEG training relates to the quality of sleep. The EEG training has emerged as a powerful tool in the management of ordinary sleep disorders such as sleep onset difficulties, frequent waking, nightmares and night terrors. (More challenging sleep disorders such as sleep apnea, narcolepsy, and nocturnal myoclonus are not included here.) A relationship has been observed between pattern of sleep disturbance and affective disorders. Thus, sleep onset difficulties correlate with anxiety, and difficulty in staying asleep are correlated with depression. The inability to find one's way to bed, and sleeping only a few hours each night, is associated with mania. The protocols used in these cases are identical to the approaches used for anxiety, depression, and mania, respectively.

Nightmares can be seen as an anxiety phenomenon for purposes of protocol selection. Night terrors, on the other hand, are presumably a paroxysmal event. They generally respond to a combination of higher- and lower frequency training. Nocturnal elimination disorders (enuresis, encopresis) generally respond readily to the training in the young. However, enuresis which survives into adulthood may require a greater variety of protocols and a greater number of training sessions, and may even be entirely refractory to training with any protocol we have devised. Enuresis can be considered a concomitant of disregulation of arousal during sleep. Encopresis, on the other hand, could be a paroxysmal phenomenon, possibly requiring more extended training.

Efficacy for common sleep disorders can be invoked in support of the model that improved regulation of arousal is the primary mode of action of EEG biofeedback training. It can also be used to exclude the hypothesis that overt behavior may be trained as opposed to control mechanisms. After all, rehearsed behavioral strategies are not likely to be relevant as the brain manages its own sleep state transitions. Sometimes aborting a pattern of enuresis or nocturnal bruxism can be accomplished within one to four training sessions. The person most surprised may be the child himself, unaware of having made any behavioral adjustment whatsoever.

Anxiety

In principle, anxiety responds exceedingly well to EEG biofeedback training. In this respect, it is similar to peripheral biofeedback. In practice, however, just as with peripheral biofeedback a considerable problem with patient compliance may be observed. Whereas the competence of the technique in remediating anxiety is now beyond question, there are other factors that can affect compliance adversely. The dynamics of the training can elicit performance anxiety; the anxious person may have difficulty abandoning the perceived'but ambiguous''comfort zone' of the anxiety state (to wit, my vigilance is keeping me alive). By the same token, the anxious person may have difficulty perceiving a more relaxed and controlled state as being desirable. In some individuals, compliance can be increased if the lower- frequency alpha training is also employed early on in training. However, the latter is not the focus of this survey.

It is primarily for reasons of compliance that we have ranked anxiety as more problematic, on the whole, than panic disorder. This probably contrasts significantly with what is found with other therapies. We consider panic disorder as a paroxysmal condition. The EEG training is manifestly quite competent in stabilizing the brain against the minor paroxysmal events such as panic disorder and night terrors previously discussed. Obtaining such stabilization may, curiously, be quicker and easier than comprehensive management of an anxiety susceptibility.

Click for Next Page