December 2, 2003

Biofeedback Society of California Conference, 2003

Author: Siegfried Othmer

The BSC had its largest meeting in seven years November 5-8 in Irvine. This is particularly welcome news because normally it is the Northern California venues---in alternate years---that draw the largest participation. A good spirit prevailed.

I gave a short course on developing a general theory of self-regulation. This theme was also reflected in my talk at a panel discussion, in which I addressed the spectrum theory of ADHD and Autism. In the short course I talked about the increasing overlap between peripheral biofeedback and neurofeedback. For purposes of this discussion, I lumped HEG in the peripheral category rather than the neurofeedback bin. Both Hershel Toomim and Jeff Carmen like to have HEG to be taken as neurofeedback modalities. This sentiment is probably not shared by most people coming from the biofeedback side, and it doesn’t fit my world view either. To my mind, neurofeedback has to engage with the timing and frequency basis of brain organization, i.e. directly with synaptic events.

If we think only in terms of the traditional peripheral modalities, then the overlap between biofeedback and neurofeedback lies largely in the realm of conditions with a primarily autonomic involvement---migraines, asthma, panic anxiety, stress responses, anxiety, Reflex Sympathetic Dystrophy, etc. The moment we consider HEG as well, however, the overlap of biofeedback with neurofeedback becomes much greater. Now we have to consider ADHD, the autistic spectrum, fibromyalgia, multiple chemical sensitivities, and many other conditions as being accessible to both types of modalities. And with respect to conditions like migraine, HEG appears to be entirely competitive with our latest techniques in EEG biofeedback.

If Hershel Toomim or Jeff Carmen were writing this, it would read very differently. Each of us is still viewing the world very much from the perspective of our own trenches. This came home to me recently when Hershel expressed surprise that we could address the emotional and affective disconnects in autistic children with neurofeedback. He had thought that to be the unique province of his HEG. Of course our collective findings for affect regulation have been common knowledge among our practitioners for years, and this capability forms the basis of our approach to Reactive Attachment Disorder as well. I could hardly believe that this would be news to Hershel.

And Jeff would no doubt consider neurofeedback for migraines to be the new kid on the block, when in fact we have been wildly successful with migraines over many years, and have been watching him catch up as he developed his HEG to greater maturity and got greater experience with it. The fact that the world at large chose not to pay attention does not alter this reality.

Even as I was making the case during this conference for a unified perspective that brings together the insights from both biofeedback and neurofeedback, the extreme fissures that still prevail in our field surfaced again and again. Sue and I attended Dick Gevirtz’s workshop on chronic pain on the last day of the conference. Gevirtz talked about the mechanism for the development of myofascial pain. Here we have highly localized pain originating in specific muscle spindles that may originally have been triggered in the course of the ordinary business of protecting muscles from over-exertion. The muscle spindles protect the muscles by signaling pain, but then they did not return to baseline. This represents a fundamental shift in our traditional understanding. The problem here was not to be found in “tense muscles.” The EMG levels in the neighborhood of the trigger points were normal. It was the muscle spindles, which are embedded among the muscle fibers and communicate their state of tension, that had altered their setpoint of function. Among their functions is calling a halt to muscle over-exertion through pain, such as we observe in muscle spasms, cramps, and muscle pulls. In myofascial pain we are seeing a perfectly normal pain response selectively and locally gone wrong.

The problem is highly localizable, and does not appear to be central in origin, although the pain is a function of sympathetic drive. Since the pain is peripheral and local, there does not seem to be a case here for addressing central mechanisms of pain through neurofeedback, according to Gevirtz. On the other hand, conventional EMG training doesn’t help that much either because we are not really dealing with “muscle tension” in the colloquial sense. As it happens, Gevirtz’s recent interest in Heart Rate Variability training paid off. This turns out to be the best biofeedback option for myofascial pain. The mechanism by which HRV training is effective remains obscure. However, the hypothesis is that the “challenge” of HRV drives the sympathetic and parasympathetic control loops involving cardiac regulation into better balance. This sounds remarkably similar to the way we like to understand neurofeedback as well. But to continue with the story.

On another part of the program, past AAPB President Paul Lehrer talked about asthma. There were not enough handouts to go around, so I am having to go entirely on memory here. In a small study of asthma, Lehrer had compared Heart Rate Variability training with a wait list control, and with a “placebo” EEG protocol, namely frontal alpha training. It had already been established in earlier work that conventional breath training didn’t do much to improve asthma susceptibility. The surprising finding was that HRV did remarkably well in reducing the incidence and severity of asthmatic events, and in reducing medication utilization. The other surprise was that the “placebo” EEG protocol did just about as well.

After the formal talk, I challenged Lehrer on his choice of a placebo. I thought this was mischievous at best, and diabolical at worst. And it served him right that it worked entirely too well. Lehrer said that the EEG training had not improved the ancillary variables related to the breath that they had been tracking---such as breathing resistance, breath exhalation volume, etc. But then neither had the HRV training! And when I reminded him that the medication reduction was nearly as great with EEG as with HRV on his very own charts, he tried to sweep that under the rug. Yes, there was a response, he said, “but we believe the EEG response to have been a classical placebo response.” This even though it nearly matches his HRV data in nearly all respects. There is of course no scientific way to discriminate at this point whether the improvement shown by the EEG cohort was a placebo response or was “real.” That would require yet another independent test. But what if we accept Lehrer’s hypothesis, and assume that the EEG data reflect only a classical placebo. One would then have to conclude that the HRV data are only marginally better than placebo! So there would be really very little to talk about. That would hardly be agreeable to Paul, either. Without a doubt, we were seeing slight-of-hand on stage to try to make perfectly good EEG-induced improvement disappear. This, despite the fact that nobody in their right minds would propose frontal alpha training for asthma if they were actually intending to be helpful. If one had to choose a placebo EEG feedback protocol, they did not do badly. What if a decent protocol had been used? They might have found what we have known for years from our own experience, namely that we can be routinely helpful with asthma susceptibility.

So we have here two instances in which HRV paradoxically gave good results, suggesting a mechanism through improved sympathetic/parasympathetic balance.
I suggested to Lehrer that he might like to use a proper EEG protocol as part of his design in the future, and he answered, “I’d be interested if you have the data to justify it.” Mind-boggling. I have in fact been telling him about our asthma data for years. And now his own data tend, if anything, to be supportive. What this discussion illustrates is that Lehrer and Gevirtz seem to deal with EEG as some sort of abstraction that is separate from their world. They seem to handle it with tongs, and talk about it in standard categories such as “placebo.” There seems to be no realization that EEG training can similarly effect a re-balancing of sympathetic and parasympathetic activation, and certainly no awareness that with EEG training one can achieve improved stability in nervous system functioning in considerable generality. This all rests, I believe, on the stark distinction that has been historically made between the autonomic nervous system and the central nervous system, even though they are obviously involved ultimately in a network of mutual regulation.

But there is more. In the case of Gevirtz and myofascial pain, his model is dealing directly with the muscle spindle activity that we think we are training in the traditional Sterman model. It would seem to be only too obvious to reach into the grab-bag of tricks and come up with SMR-training. But by now, in reflection on his new findings around HRV, Gevirtz postulated that the limbic system must be involved, and of course training at the sensorimotor strip wouldn’t have anything to do with that now, would it?

So my hopes of bringing some synthesis into our broken field came thudding down to earth on the last day of the conference with this display of radical parochialism. Somehow these people intend to go on living their lives without caring one whit about what happens to the brain, or to the EEG. How is such compartmentalization of thinking even possible in the biofeedback field, the very field that should be open to integrated models of function? The pathology in the field is in some ways greater than that in the patients. Sue points out that when it comes to the EEG, Lehrer and Gevirtz seem simply to have absorbed what Sterman and Lubar have been saying. Without other evidence out of their own experience, why should they do otherwise? Indeed. Those of us active in the field have no qualms about moving beyond Sterman and Lubar, but if one doesn’t have that corrective on one’s belief system, what then?

By the time these two programs took place on the last day, I had already enthusiastically proclaimed the more optimistic message about an impending unification of the field around the new findings of HEG and EEG feedback. Such a general understanding must be based on the more abstract Disregulation Model, one that rises above the particulars of specific control systems. This Disregulation Model concerns itself with all of the feedback control loops managed by our central nervous system. Any physiological variable that is subject to measurement anywhere on the body can be used in a feedback paradigm to effect improved regulation. If there is overt disregulation, then the feedback can be used to restore proper regulation. If no overt disregulation is present, then the feedback works simply in an exercise fashion: The system is disturbed out of the prevailing state, unleashing restoring forces that move to reestablish the intended state. That exercise likewise strengthens the functioning of the control apparatus.

In proposing the Disregulation Model as the basis for unification of the field, I am also saying that this model has huge explanatory power, once it is understood. The problem is that it will have its own language, one that is not taught to mental health professionals. So it may not resonate with people immediately. But then it is a long way also from serotonin receptors to depression, and mental health professionals didn’t care to know about that either until they thought they had to.

Below the level of the over-arching Disregulation Model, our models have to be more specific. Thus, with regard to EEG biofeedback we are establishing improved functioning of neuronal networks, of their local activation/de-activation, and of their linkages to other brain regions. In this context, we see brain organization in terms of the “small-world” network model, one in which certain “hubs” play crucial roles. One of the primary hubs is the thalamus; another is the brainstem; others are the basal ganglia. The thalamus in particular plays a central role in the generation of the EEG rhythms we observe. The failure modes of the thalamocortical networks are subsumed in the Rodolfo Llinas model of Thalamocortical Dysrhythmias, although one must allow for the possibility that the thalamus may not be the ultimate cause, but rather merely the mediator of the disregulation of interest. In any event, resolution lies in the remediation of the observed dysrhythmias.

Any manifestation of the disregulated activity---reduced perfusion; reduced oxygen demand; elevated muscle tone; etc. can be used equally well in the task of re-regulation. The outstanding question is one of efficiency in this process rather than of raw efficacy. One does not in any sense have to zero in on the “beating heart” of the disregulated activity. One simply has to observe it somewhere.

There was a talk at the conference about the latest findings with regard to the cerebellum, that obscure organ to which we had assigned nothing more substantial than the organization of smooth muscular movements. What a waste of good resources, in that the cerebellum has more neurons than cortex by a factor of ten. Turns out that much more may go on there with respect to sensory integration and even emotional regulation. So the thought that immediately commands our attention is where and how one might train the cerebellum. It does not contribute much to the EEG; we know that. So the EEG does not provide a handle. But wait. If the cerebellum is in the regulatory loop with respect to emotional processing, sensory integration, and other functions, then we have been training it already! After all, we have no direct way of training the basal ganglia either, but we believe that we are impacting on basal ganglia function with our training. The same holds for the brainstem.

If we take our network model seriously, then we realize that there is not a problem to be solved here. If something is part of the regulatory network, then it will get trained. If not, then it is probably irrelevant. So with the cerebellum moving closer to center stage, nothing has really changed other than our thinking, and with it our “circuit diagram”! Our existing approaches to neurofeedback training have not suddenly become inadequate.

Come to think of it, just how has thinking about getting to the source of the problem taken hold in the biofeedback field in the first place? Aren’t we the ones who put electrodes on finger tips to train sympathetic nervous system response? Any measurement anywhere that detects the disregulation is good enough. With that information, from wherever derived, we close the loop back to the brain through an information pathway.

In the talk on the cerebellum, it was pointed out that when a severe problem of dysfunction exists there, the person may actually function better without any cerebellum at all. This may at first seem surprising, but we see the same thing in cortex as well. Taking out part of the brain to improve overall function has been a part of medicine now for some time, and we are even taking out whole hemispheres just to stop seizures. This can also be regarded from the network perspective: A dysfunctional region of the brain can perturb brain networks broadly. But if the source of disruption is removed, brain networks can learn to work around the holes that are left behind.

These considerations also elevate disregulations to the top rung of our concerns, displacing the famous and largely fictional lesion. Small lesions are probably commonplace, but it is disregulations that limit functionality, and these may have very little to do with the observed lesions.

In summary, unification of biofeedback and neurofeedback is to be found only at the highest, most inclusive level, that of the Disregulation Model. Below that, our models must particularize to the language appropriate to our respective tools. When it comes to our principal modalities, we find that EEG, HEG, and now HRV are striking in the generality of their effects, and in the similarity. This only makes sense if we see each of these techniques impinging upon an integrated network of regulation, and that it is a secondary issue as to how that network is sampled, or accessed.

As a final epitaph to where the leaders of the biofeedback field have brought us, I report that Dick Gevirtz said at the end of his workshop that one would not be inclined to use EEG to train migraines, since they are a vascular phenomenon. I kid you not. This in 2003. What a down-note on which to end the conference. Let’s hope that at the AAPB meeting there will be a chance to update everyone on the new findings regarding migraines. Dick Gevirtz is program chairman this year.

Finally, I would like to mention the renewed attention that Peter Litchfield is bringing to the breath, and to its measurement using end-tidal CO2 concentration. Here we have another technologically very straight-forward means of assessing the quality of regulation. It contrasts to EEG work particularly in the fact that we can exert some conscious control over the breath. For anxiety and other conditions, we really need to augment our EEG training with breathwork as an element of mastery. The combination of this with breath-training of heart function makes this an obligatory field of study. Naras Bhat—a lecturer who is simply a joy to listen to—calls defensive breathing a “heartless matter.” He suggests that simply proper breathing constitutes a remedy for many of the electrical storms of the heart.

Peter Litchfield proposes a chemical dependency model of over-breathing. Things reach a point where over-breathing feels natural. Then when a person is catapulted into crisis, the tendency is to over-breathe more, thus worsening the condition. Such a crisis could occur simply around breathing itself. Peter also points out that over-breathing is involved in a lot of the bad actors we are concerned with---the autistic spectrum and other conditions we have lumped into the over-arousal category.

When Peter talks about the chemical cascade of disregulation that results from over-breathing, the explanatory power of this perspective on disregulation is remarkable. But what interests me even more is the fact that we have another key tie-in here to emotional regulation. It will be recalled that the HeartMath people really weren’t that interested in Heart Rate Variability training for the sake of heart health per se, but rather as a means of reawakening the heart of which the poet writes. It was a pathway to emotional deepening and to emotional regulation. We have here a way of reconnecting to our “heart space.” Many will not be able to do heart rate variability training successfully until they manage their more fundamental problem of over-breathing. And that may not happen until we can show it to them. The first step in meditation with hand-rails.