Neurofeedback at Infra-low Frequencies of the EEG

It is time for an update on our collective experience with infra-low frequency neurofeedback training. One impetus is the continuing confusion in the rest of the field about the implications of this kind of training, both theoretically and practically. Some are still skeptical of the whole enterprise. Others are coming to terms with it, but would prefer to regard it as filling certain niches within the panoply of neurofeedback approaches. Such pigeon-holing would nicely leave unperturbed whatever has come before in terms of neurofeedback approaches and understandings. While it is probably true that no single neurofeedback technique will cover all the bases, the infra-low training gives every sign of being broadly applicable to the concerns that surface in neurofeedback offices. It's not just for PTSD and for the autism spectrum. It follows, then, that it must be folded into our thinking generally about neurofeedback and cannot be compartmentalized.

The most striking clinical reports do relate to PTSD and to the autism spectrum, but that is simply because expectations are so modest with regard to both of those conditions. A reporting bias has crept in because we tend to emphasize those cases that defy the unbelief among the unbelievers. It continues to surprise that results in such "intractable" cases are so quickly achieved. One report found "huge improvements" in only sixteen sessions with respect to startle response, hypervigilance, troubling memories, disturbed sleep, and paranoia. A middle-age woman with an abuse history "benefited greatly" in only five sessions, by which time she reported that her PTSD symptoms were gone (although some sleep issues remained to be dealt with).

The fact that some reports are so striking gives the impression that the infra-low training is a distinct entity that must be discussed in its own terms. In our own perspective, however, it is part of a continuum in development which utilizes the same basic approach we have now used for over a decade. There has simply been a progression toward the inclusion of ever lower EEG frequencies as experience was gained and as the software allowed a more extended range of operation. The same optimization procedure that we always use with trainees had us bumping up against the lower limit of our software with every iteration, from NeuroCybernetics to Brainmaster to BioExplorer to Cygnet.

As the range in frequencies was extended downward, the effectiveness of the training increased both in terms of results and in terms of the rate at which these results were achieved. This can perhaps best be illustrated with case vignettes from practitioners who were just beginning to explore the infra-low frequency region:

"The autistic child who had over 200 sessions of neurofeedback was still babbling during sessions until I started working at the 0.01 Hz frequency; she has quietened down during the last 4 sessions. She was able to focus on the game without babbling."

And another report on an intellectually challenged teenager (which also points up the frequency specificity of the training):

"I have been walking him down, frequency-wise, with excellent effect. Last session we dipped down from .01 to .005 for the first time. Mom said he was the happiest over the next couple of days she'd ever seen. He's also been much more self-reliantâ€"doing things on his own he normally seems not to know how to do or feels he can't. He's also become far more talkative and is additionally becoming far more intelligible. For the first time he started talking in full sentences. (All the training was done with the inter-hemispheric training of T3-T4)."

At the next session, a progression to even lower frequencies seemed to be in order. But at 0.003 Hz the mood turned sour in a child that had come in the door happy and smiling. He covered his eyes and didn't want to train any more. But when the electrodes came off he just sat balled up in the chair and didn't want to move. Upon invitation from the clinician, he was willing to try again and grabbed the dolphin (for tactile feedback). The obvious remedy was to go back to 0.005 Hz, but his expression remained sour. His brain had moved"At 0.004 Hz, however, a huge grin swept across his face and he looked like a brand new boy. The smile never left him for the duration of the session.

Such frequency sensitivity defies belief, but it is routinely being observed. This presents both a challenge and an opportunity. Sensitivity to training is typically such that the immediate report from the trainee can be used to fine-tune the training effectively. The frequency at which the person feels best is also the frequency at which training is most effective. Further, this is the frequency at which the person feels most in tune with the signal being presented. There is a natural quality to the training under these conditions, one in which the person feels in harmony with the process. The fact that such subtle--and not so subtle--state shifts are subject to our immediate influence takes some getting used to, however. This is where clinician skill comes in. A trainee may simply motor on even under duress because that's what they think they are paying for. They know that medicine does not always taste good. A clinician typically needs to cultivate state awareness on the part of the trainee over time to refine the training progressively.

It turns out that the addition of the infra-low frequency range to our repertoire has improved our outcomes across the board with all of the various conditions that we encounter in our work. For example, a person who came with the desire to master binge drinking had had only five sessions over a period of seven weeks when he reported at the last session that he hadn't had a drink since the previous session three weeks prior, despite the intervening year-end holidays. A severely anorexic teenage girl in a physically delicate state was turned around in short order with only a few neurofeedback session at infra-low frequencies. A young boy with medically intractable seizures became seizure-free over only a few sessions. He trained successfully at 4 Hz.

As stated above, the training effects are clearly stronger at the resonance frequency for each individual, and that fact could not be fully exploited until we extended our signal bandwidth down to 0.001 Hz. Even now, we see a pile-up of clients at the lowest frequency. In fact, about fifty percent of our clients optimize at the lowest frequency. This of course has something to do with the kinds of clients that now fill our office---autistic and bipolar children, PTSD, and migraine. On the other hand, even ADHD children often optimize at the lowest frequency. The distribution doesn't correlate with diagnosis particularly; it may instead depend more on severity. The more intractable conditions--of whatever stripe or label--may be more likely to involve the deeper disregulations that we target with the infra-low training.

The stronger effects we get with resonance frequency training manifest first in the experience of state shifts that can often be quite profound. In the early days of our neurofeedback work, when we still did our combination of SMR and beta training, state shifts were the driver in the optimization procedure also, but the shifts were not as noticeable for two principal reasons. Firstly, we were not at the optimum reward frequency for most people, and secondly we were largely working with a population of ADHD kids who are not good reporters on their own state in any event.

The ability to move people very quickly to more functional and more stable states is indeed welcome, but by itself it does not make the case that learning has occurred. The consolidation of learning still takes time. But learning clearly happens faster when it takes place while the person is an optimal state already. In the SMR/beta training paradigm, the movement toward more appropriate state regulation was necessarily more gradual, as the neurofeedback process was typically taking place under non-optimal conditions.

The latest development in our work is the finding that an optimum reward frequency in the infra-low frequency region does not rule out other optima in different parts of the EEG spectrum. The optimization procedure is local in the frequency domain rather than global. So a person who thrives with the infra-low training may also benefit from training in the neighborhood of the SMR band. Again, however, an optimization procedure can make a considerable difference. At the higher frequency we are engaging different brain subsystems, but each of these can apparently exhibit an incredible degree of frequency specificity in many individuals.

To date our focus on the higher frequencies has concentrated on the dominant cortical resting frequencies. By implementing synchrony protocols the principal role is once again played by the relative phase between the two sites at the target frequency, just as we believe to be the case for the bipolar training. Each of these resting frequencies evokes a different response, with the common characteristic that each leads to a greater sense of calmness, of being in control, and of generally being in a good place.

 1  |  2