Neurofeedback May Protect Older Adults' Cognitive Abilities

Graciela Alatorre-Cruz is a research scientist near Mexico City at the National Autonomous University of Mexico. She led researchers in Mexico, Norway, and the United States in a recent study that examined how EEG neurofeedback (NFB) may alter brain structure and function to maintain brain health in ageing (Alatorre-Cruz et al., 2022).

This is important because there is a normal decline in some cognitive abilities as well as an increasing risk for dementia as we age, and the ageing population is increasing as various healthcare approaches lead to longer life expectancy. Efforts to bolster cognitive abilities during normal ageing may be protective against neurocognitive disorders in older adults.

What is the Science?

The science underlying this study rests on the observation that certain electroencephalographic (EEG) patterns in older adults carry meaningful information about cognitive ageing. As outlined by Alatorre-Cruz, resting-state EEG in older adults shows reduced absolute alpha power and a slowing of the peak alpha posterior dominant rhythm. Alpha may also be topographically reorganized with increased frontal activity. Occasional delta increases in temporal regions may also be observed.

In particular, excessive theta activity in the resting EEG is a well-established predictor of later neurocognitive decline.

Long before memory loss or executive dysfunction becomes behaviorally apparent, the brain often signals its vulnerability through subtle shifts in frequency rhythms. Neurofeedback has been used previously with older adults, both with and without neurocognitive decline.

For older patients, protocols have aimed to normalize aberrant EEG activity and employed various approaches that upregulated SMR/theta, alpha, beta, or beta/alpha. Protocols with healthy older adults have included training the EEG pattern of an expert meditator, uptraining frontal midline theta, downtraining theta, and uptraining alpha, SMR, beta, and gamma.

From earlier work by the same research group, it was known that neurofeedback that trains older adults to reduce elevated theta activity could induce measurable reorganization in the brain’s electrical patterns and even improve cognitive performance. The rationale for choosing theta as the target for downtraining by Alatorre-Cruz’s group was based on resting-state EEG patterns in healthy older adults and the risk of cognitive decline associated with elevated theta.

What remained unknown was whether such changes would last. This is important because clinical usefulness depends not only on immediate effects but also on their durability over time. The few previous studies with follow-up by other investigators included subjects with mild neurocognitive dysfunction and a 1-month interval. They found that 20, but not 10, sessions of NFB were necessary to maintain gains.

What Did They Study?

Alatorre-Cruz’s (2022) study examined NFB as a potential preventive strategy for healthy older adults who exhibit elevated theta EEG risk markers but have not yet shown cognitive decline. The question was whether modifying these theta rhythms could yield enduring benefits, as evidenced by EEG and cognitive measures, one year after NFB.

They studied twenty healthy adults over the age of sixty who showed abnormally high theta power in at least one of 19 standard 10-20 electrode sites, a pattern linked in earlier longitudinal research to elevated risk for cognitive decline years later. Participants were women and men aged 60 or older, with at least average IQ, at least 9 years of education, and cognitively normal on baseline neuropsychological testing, who were free of neurological and psychiatric disease. 19-channel eyes-closed EEG and cognitive measures were administered at baseline, 2 months after NFB, and 1 year after NFB.

How Did They Do It?

Participants were randomly assigned to a true neurofeedback group or to a sham control group that received identical auditory tones unrelated to their brain activity.

All participants underwent a 19-channel quantitative EEG and a comprehensive set of cognitive tests before treatment, 2 months after completing the program, and again at a 1-year follow-up. Cognitive tests included the 6 summary scales and the Full Scale IQ from the Wechsler IQ test, as well as 10 indices from an abbreviated standardized battery of neuropsychological measures (NEUROPSI), all administered in Spanish.

This design enabled tracking of both short- and long-term changes across multiple cognitive domains and EEG frequency bands, including delta, theta, alpha, and beta. This design also allowed the investigators to distinguish genuine learning of EEG self-regulation from expectancy or placebo responses.

Participants completed 30 30-minute z-score eyes-closed neurofeedback sessions over 10 to 12 weeks. During each session, the program monitored absolute theta power at the EEG electrode showing the participant’s most abnormal baseline value. Participants were told that more feedback was better, but were not provided with strategies for producing it.

For the 10 subjects in the experimental group, whenever theta power fell below a starting threshold of 70%, feedback was delivered in the form of an auditory tone. However, thresholds were adjusted every 3 minutes as needed to maintain a success rate between 60% and 80%. Two subjects in the control group dropped out, leaving 8 who underwent the same procedure; however, the tones were delivered randomly rather than according to brain activity.

What Did They Find?

EEG Results

Two months after training, both groups showed significantly reduced theta power at the trained electrode (i.e., the electrode at which theta was most extreme at baseline). Still, the magnitude of the reduction was significantly larger in the neurofeedback group, suggesting that EEG-specific learning had occurred.

Another distinguishing feature of the experimental group was the widespread reorganization of EEG activity observed 2 months after training. They showed global decreases in delta and theta activity and increases in beta power, a shift toward faster and more efficient neural rhythms commonly associated with healthier cognitive ageing.

The control group showed only minor local theta reductions and no meaningful changes in higher frequency bands.

One year after training, the neurofeedback group maintained and, in some cases, strengthened the earlier improvements, with more pronounced decreases in theta and increases in beta relative to their pretreatment EEG. The control group showed a reduction in theta power and a limited increase in beta power relative to baseline.

In a supplementary table, the authors report theta z-scores for each subject at each time point, showing that, one year after training, a marginally larger percentage of NFB subjects had absolute theta power within the normal range than control participants.

Cognitive Results

Results from the Wechsler indices showed no change, except that the experimental group showed an increase in Full Scale IQ from baseline to post-training, followed by a decrease from post-training to the 1-year follow-up that was larger than that observed in the control group.

Results on the executive functions subscale of the NEUROPSI in the NFB group showed a significant improvement at 1 year. In contrast, the control group showed declines over the same periods. Although effects were not uniform across all cognitive measures, the sustained enhancement in executive functioning aligned with the electrophysiological normalization produced by training.

There were no significant correlations between NFB learning and any cognitive measure.

What Are the Strengths and Limitations?

A major strength of the study was its rigorous design, which included random assignment, a sham control condition, blinding of participants, and a long follow-up period, rarely seen in neurofeedback research with older adults.

The training protocol was intensive, with 30 sessions rather than the fewer sessions typical of earlier work, and it replicated prior findings confirming EEG plasticity in this population.

The study also combined electrophysiological analysis with standardized cognitive assessment at multiple time points, providing converging evidence of neural and behavioral change.

Although baseline theta appeared greater in the experimental group than in the control group, the authors did not compare pretraining theta power between groups. Regression toward the mean is a statistical phenomenon where extreme values of a variable (e.g., theta) tend to be closer to the mean when measured a second time. It is possible that regression toward the mean, at least in part, accounted for the reduction in theta after training, and more so for the experimental group, whose baseline theta was even more extreme than that of the control group.

It is also possible that eye movement artifact was not well controlled for before training and was reduced following the experience of regularly sitting quietly during the training sessions. Eye movement artifact is often a contributor to false elevations in low-frequency activity during EEG recordings.

Practice effects may be a factor in interpreting the Wechsler and NEUROPSI results. Both measures do have alternate forms, but the authors do not explicitly report using them to prevent the possible appearance of improvement at post-training and follow-up intervals. However, the pattern of results suggests that practice effects were not an issue.

Another limitation is the absence of measures of meaningful day-to-day activity (e.g., the Executive Function Performance Test; Baum et al., 2016), which might have added ecological validity to the study.

The principal limitation was the small sample size, which restricts generalizability and statistical power. Experimenters were not blinded to group assignment, and neurofeedback sessions did not include online artifact correction, a consensus guideline in neurofeedback methodology.

Because participants knew they had EEG risk markers, lifestyle changes during the study may also have contributed nonspecifically to the observed improvements. Nevertheless, these limitations do not diminish the clear group differences across EEG and cognitive measures, which point to neurofeedback-specific effects.

What Is the Impact?

The study provides evidence that neurofeedback training designed to reduce excessive theta activity may help preserve, or even improve, cognitive health in older adults at risk of decline, as evidenced by significant improvements in executive function in the NFB group but not in the control group.

The finding that EEG patterns in both theta and beta bands not only improved but also remained better after 1 year suggests a durable reorganization of neural dynamics rather than a limited, temporary training effect. The documented improvement in executive functions reinforces the possibility that modifying early electrophysiological risk markers may translate into meaningful functional benefits.

By treating individuals who are still clinically healthy, the research moves neurofeedback into a preventive framework and offers a promising, noninvasive method for maintaining brain resilience as people age.

It also contributes to a growing body of work advocating for high-quality, long-term follow-up in neurofeedback studies, which is essential for evaluating clinical relevance.

Key Takeaways

1. Neurofeedback effectively reduced excessive theta activity, a known predictor of cognitive decline, in healthy older adults.

   
   

2. EEG changes in the neurofeedback group were durable and generalized to an untrained EEG band (beta), persisting and, in some cases, strengthening after 1 year.

   
   

3. Executive functions improved significantly only in the neurofeedback group, suggesting likely causal effects of NFB training.

   
   

4. The study supports neurofeedback as a potential preventive health intervention for older adults with EEG risk markers, warranting larger-scale research.

Glossary

absolute power: a quantitative measure of the energy contained in an EEG frequency band, typically expressed across defined spectral ranges such as delta, theta, alpha, beta, and gamma.

alpha activity: an intermediate speed EEG frequency whose appearance in the occipital lobes when eyes are closed suggests restful wakefulness.

beta activity: a pattern of faster EEG frequencies reflecting an alert and engaged cortical state often associated with efficient cognitive processing.

delta activity: a slow EEG rhythm that is typically prominent during deep sleep and may indicate dysfunction when elevated during wakefulness.

ecological validity: the extent to which a study’s methods, settings, and tasks approximate real-world conditions closely enough that its findings can be generalized to everyday behavior

electroencephalogram: a noninvasive recording of the brain’s electrical activity obtained through scalp electrodes.

Executive Function Performance Test (EFPT): a standardized, performance-based assessment that evaluates an individual’s executive functioning in real-world tasks by measuring initiation, organization, sequencing, judgment, and safety during activities such as cooking, phone use, bill payment, and medication management

executive functions: a set of higher-order cognitive processes that include planning, inhibition, working memory, and cognitive flexibility.

gamma activity: a very fast EEG frequency that may be involved in active information processing and perceptual binding.

neurofeedback: a form of operant conditioning in which individuals learn to modify their brain activity using real time feedback.

NEUROPSI: a standardized neuropsychological battery developed in Mexico that assesses attention, memory, language, visuospatial skills, and executive functions in Spanish-speaking adults across a wide age and education range

regression toward the mean: the statistical tendency for extreme scores on a first measurement to move closer to the average on a subsequent measurement because part of their extremity reflects random variation rather than stable characteristics

theta activity: a low-frequency EEG rhythm that, when elevated at rest in older adults, serves as an early risk marker for later cognitive impairment.

Wechsler test: a standardized intelligence scales developed by David Wechsler that assess general cognitive ability through subtests measuring verbal comprehension, perceptual reasoning, working memory, and processing speed, most commonly represented by the WAIS for adults.

References

Alatorre-Cruz, G. C., et al. (2022). One-year follow-up of healthy older adults with electroencephalographic risk for neurocognitive disorder after neurofeedback training. Journal of Alzheimer’s Disease, 85, 1767–1781. https://doi.org/10.3233/JAD-215538

Baum, C. M., Wolf, T. J., Wong, A. W. K., Chen, C. H., Walker, K., Young, A. C., Carlozzi, N. E., Tulsky, D. S., Heaton, R. K., & Heinemann, A. W. (2017). Validation and clinical utility of the executive function performance test in persons with traumatic brain injury. Neuropsychological rehabilitation, 27(5), 603–617. https://doi.org/10.1080/09602011.2016.1176934

About the Author

Dr. John Raymond Davis is an adjunct lecturer in the Department of Psychiatry and Behavioural Neurosciences at McMaster University's Faculty of Health Sciences. His scholarly contributions include research on EEG changes in major depression and case studies on neurological conditions. ​

Support Our Friends

Neurofeedback May Protect Older Adults' Cognitive Abilities.pdf

Download PDF • 7.77MB