Neurofeedback and cognitive enhancement, a review
One of the common questions people ask me is whether neurofeedback can help cognitive decline associated with aging. Some very recent research articles and published reviews suggest cautious optimism. Yes…now we know that Neurofeedback can definitely increase attention and working memory performance in normal aging adults. But…while there have been reports of neurofeedback helping several types of dementia, there is no reason to expect EEG biofeedback to reverse progressive dementing conditions such as Alzheimer’s. Important questions are still unanswered regarding which particular protocols are the most effective…and at a deeper level, what types of working memory (cognitive vs. motor learning) are enhanced with neurofeedback?
I will focus on one paradigm-breaking article for this post, the first of my three reviews of neurofeedback and improved cognition.
JR Wang and S Hsieh published a definitive article in late 2013, boldly titled, “Neurofeedback training improves attention and working memory performance” (Clinical Neurophysiology, Volume 124, Issue 12, December 2013, Pages 2406–2420. The article provides a comprehensive review and then presents a brilliant experimental demonstration of neurofeedback efficacy. Following is my own review of that article.
In their Introductory section, Wang and Hsieh define the problem in the first sentence: “As individuals age, their physical, physiological, and psychological functions begin to deteriorate, which results in the progressive loss of competence.” They review the literature on the subject and then point out the key task: “it becomes increasingly important to understand how cognitive functioning can be preserved and promoted in old age.”
In subsequent subsections of the Introduction, they reviewed neurofeedback training historically, and then specifically in terms of various forms of feedback and the types of cognitive improvement that have been demonstrated and published in the professional literature (keep in mind that the EEG signal can be extracted in many ways—it can be alpha (8-10 Hz), theta (4-7 Hz), beta (12-15 or 15-18 Hz), gamma (36-44 Hz), various ratios of these (e.g., theta/beta ratio), or other pieces of the EEG signal—and then “fed-back” to the person receiving the information about their own extracted signal in many forms including direct and indirect visual and auditory analogues including games and distortion/restoration of commercial DVDs).
Wang and Hsieh then review neurofeedback and cognitive aging. They note that research has shown the effectiveness of neurofeedback for healthy young adults but the published evidence is sparse for older adults. One set of experiments has demonstrated that individuals from age 70-78 increased speed of information processing and resistance to interference (i.e. attention to task) with peak alpha frequency training. Another group demonstrated that feedback training to reduce the overall theta power in “normal elderly participants” improved performances on a number of measures compared to placebo, including the Verbal Comprehension Index and Verbal IQ scores of the WAIS-III, total scores on attention, executive function and memory functions. A third group could demonstrate changes in brainwave amplitudes (i.e., intensity of signal strength) but—in four training sessions (only)—could not demonstrate memory improvement.
So, in designing their own study, Wang and Hsieh were left with no comprehensive guideline on which aspect of the EEG signal in older adults would be most helpful to give feedback on. Two relatively recent studies have indicated that lower frontal midline theta seems to be a physiologic (neurophysiological) marker of aging.
Their study had three ingenious components. They decided to measure EEG before and after training to confirm that neurofeedback was changed by neurofeedback (30 channel data was collected, eyes open/eyes closed).
Next, they divided their 32 subjects (half college students, half elders age 61-72) into four groups. In a two-by-two design, they tested 8 older subjects with feedback and ANOTHER 8 older subjects receiving SHAM feedback, and they tested 16 college students who received either neurofeedback or SHAM feedback.
The final component used “targeted cognitive functions that might be associated with age-related EEG deviations.” They settled on two well-established cognitive tests, with well-documented ability to differentiate the cognitive decline in older compared to younger adults.
The first cognitive measure is known as the Attention Network Test (ANT) which probes three areas at once, as Wang and Hsieh write, “(1) the alerting component (i.e., the ability to prepare and sustain alertness for the processing of high-priority signals), which involves the thalamic, frontal, and parietal areas; (2) the orienting component (i.e., the component that allows one to attend to target items overtly or covertly and thereby improve processing efficiency), which involves the superior parietal lobe, temporo-parietal junctions and superior frontal cortex; and (3) the executive attention component (i.e., conflict resolution), which involves the anterior cingulate cortex and lateral prefrontal cortex.”
The ANT is a complex design involving flashing images on a computer screen with a task to pick out when selected designs appeared to replace a fixed + sign but not false cues.
The second instrument used is called the Sternberg memory test, a computer-based word-recognition test, involving a series of word lists in which one word from a prior list may be repeated, or not. (These are sophisticated tests!).
The neurofeedback (12 sessions, eyes-open training) was based on midline theta (4-7 Hz) with a roller coaster display in which the speed (and sound) increased with the intensity (power) of the theta signal. The SHAM feedback was another EEG band randomly selected and changed in each of the 12 SHAM sessions; subjects were blind to the conditions of the feedback. In Part III, I will return to the issue of SHAM neurofeedback.
As one might expect with such a complex experimental design, Wang and Hsieh developed a mass of data to report. Here are the highlights.
At baseline (pre-training) the older subjects had slower reaction times but higher accuracy scores on the ANT than the younger group; but the elders were slower and less accurate on the memory lest (the Strindberg word recognition test).
The interesting results, as expected, come with comparisons of before and after neurofeedback. There was a trend (p=0.6) toward neurofeedback resulting in quicker reaction times, young and old.
The ANT results were also broken down into effects testing three spheres of attention: 1) the alerting network, 2) the orienting network, 3) the conflict network. Here only statistically significant results will be noted.
1) “Neurofeedback has no effect on the alerting effect.”
2) “ Only the older neurofeedback group was able to improve orienting function…via neurofeedback training.”
3) “Both neurofeedback groups were able to improve executive function (i.e., reducing conflict score) through 12 sessions of neurofeedback training.”
Of the four test groups, the older subjects who received neurofeedback had a significant improvement in the memory-testing (Sternberg recognition test); the other groups did not show statistically significant training effects.
The other major area of the Wang and Hsieh study was the effect of neurofeedback on the EEG itself. What they demonstrated was that neurofeedback on theta activity led to measurable increase of eyes-open theta activity for both older and younger subjects. Sham feedback did not have that effect, and it was not attributable to inhibition feedback in other bands. Training effect was confirmed by analysis showing increased theta activity correlating with increased number of training sessions.
In conclusion, the new Wang & Hsieh study is not just a very good experimental design, it leads us to two important conclusions—one of which is surprising.
The first conclusion is that neurofeedback training can effectively help with cognitive problems associated with normal aging.
The second conclusion is that midline frontal theta up-training can mediate those benefits. That is surprising because technically and traditionally, neurofeedback researchers and practitioners have associated slow (theta) frontal activity with the activity that disrupts the components of attention. That is why many traditional protocols since the 1980s have either inhibited theta or, similarly encouraged a decrease in the theta/beta ratio.
This research illustrates how science changes the conventional wisdom. New findings shift boundaries of what we believe and bring a new perspective on knowledge that has been accumulating but been ignored (see Kuhn’s classic, The Structure of Scientific Revolutions).
Wang and Hsieh conclude like gentle revolutionaries:
Therefore, together with evidence showing improvement in cognitive performance and increased [frontal medial theta] activity, we suggest that the NFT protocol may function as a reconditioning process for older participants. Based on Finnigan and Robertson (2011), high resting [frontal medial theta] power, which is positively correlated with better cognitive performance, can serve as a sensitivity index of healthy cognitive aging that is free of severe cognitive deterioration or impairment. This idea implies that the current training protocol, i.e., uptraining the [frontal medial theta], may prevent or decrease cognitive deterioration. For the younger participants, we suggest that the NFT protocol may function as practice for executive function or inner focused enhance- ment. Furthermore, because some previous studies have shown that theta activity is associated with the integration of several brain regions and a deeply internalized state and quieting of the body, emotions, and thoughts (Green and Green, 1977), we suggest that uptraining theta activity may have a positive effect on mindfulness or concentration in healthy individuals.
In later blogs, I will review many other studies on cognition and neurofeedback and offer more commentary on these studies.
Thomas M Brod MD
July 14, 2014
 It would be interesting to know if the sham feedback—which wasn’t really “sham” at all, since it was training at various other EEG bands in different sessions–resulted in increases in those frequencies, even for the one session those “random” bands were used.