Neurofeedback for Sensory Over-Responsiveness

Hamed and colleagues’ 2022 feasibility study asks an interesting and clinically relevant question: could neurofeedback help adults with sensory over-responsiveness, or SOR? SOR is a pattern of heightened and often prolonged reactivity to ordinary sensory input. Sounds, textures, light, movement, or touch that others register as tolerable may feel aversive, overwhelming, or even painful. For many people, this is not just a matter of preference. It can affect participation, routines, relationships, and overall quality of life.

The question is clinically relevant because adult SOR remains clinically under-served. The condition has been associated with elevated anxiety, amplified pain sensitivity, and reduced well-being, yet intervention research in adults is still sparse. The authors build on prior resting-state EEG work suggesting reduced cortical activity in SOR, especially in the alpha band, and reasonably ask whether neuromodulation might help shift this physiology in a more adaptive direction.

In general terms, biofeedback refers to training that gives people real-time information about physiological signals such as heart rate, breathing, muscle activity, or skin conductance so they can learn to regulate them. Neurofeedback is a subtype that uses real-time measures of brain activity, most commonly EEG, to support self-regulation of neural states.

In this study, the authors targeted alpha enhancement, guided by the broader literature linking reduced alpha activity to pain, stress, and heightened arousal. The rationale is attractive. If adults with SOR show a profile consistent with reduced inhibitory rhythm activity at rest, then training toward stronger alpha expression might plausibly reduce sensory burden, dampen pain sensitivity, and improve day-to-day functioning. Still, feasibility studies are early steps, not proof-of-efficacy trials, and this paper is best read with that distinction firmly in mind.

Methods

This was a single-arm experimental feasibility study with four assessment points: baseline (T1), pre-intervention (T2), post-intervention (T3), and one-month follow-up (T4). The inclusion of both T1 and T2 is worth noting because it gave the investigators a short no-treatment comparison window before neurofeedback began. That design feature cannot substitute for a sham or active control, but it does help show whether observed changes were already drifting before treatment.

The final sample was very small: ten healthy women with SOR completed the study. Participants were between 21 and 50 years old, screened using the aversive scale of the Sensory Responsiveness Questionnaire, and excluded if they had neurological, psychiatric, neurodevelopmental, pain, medication, or other factors that could confound interpretation. Restricting the sample to women was intentional, based on reported sex differences in brain activity and neuroplasticity, but it also narrows generalizability.

The primary outcome was 5-minute resting-state EEG with eyes closed, recorded using a 64-channel cap. The authors focused on spectral power in the standard frequency bands: delta, theta, alpha, beta, and gamma. EEG preprocessing included re-referencing, filtering, visual inspection for noisy intervals, and artifact reduction using independent component analysis.

The intervention was relatively brief and technically interesting. Participants completed 10 neurofeedback sessions, twice weekly, each lasting about 45 minutes. Sessions began with a 1-minute resting adaptation period used to establish a session-specific alpha baseline. The software then set a target approximately 5% above that baseline. During the 19-minute training phase, participants listened through earphones to babbling brook sounds that became louder as alpha power increased. They kept their eyes closed, were asked to relax deeply without falling asleep, and were encouraged to maximize the volume, but were not taught explicit cognitive strategies. In that sense, this protocol leaned toward an implicit learning model, though the instructions still made the training goal behaviorally transparent.

Secondary outcomes were clinically well chosen for this population: life satisfaction, goal attainment, pain sensitivity, and state-trait anxiety. Taken together, they allowed the study to examine not just electrophysiology, but whether any physiological shifts might matter functionally.

Results

The central finding is a bit paradoxical, and that is precisely what makes the paper interesting. The intervention targeted alpha upregulation, and most participants appeared able to increase alpha during training sessions according to the system’s automatic scoring. Yet resting-state alpha power, the primary neurophysiological target, did not significantly change across the four time points.

Because the protocol was designed to increase alpha, the absence of alpha change in the primary outcome keeps the clinical interpretation provisional. This is not a trivial miss. It is the main mechanistic limitation of the study.

At the same time, the authors did observe changes elsewhere. Between pre-treatment and one-month follow-up, frontal-region EEG showed enhancement in delta, theta, beta, and gamma power, with the strongest effect size in frontal theta. Post hoc comparisons were most convincing for delta and theta at follow-up, suggesting that whatever learning occurred may have consolidated over time rather than appearing immediately at post-treatment. The lack of change between T1 and T2 also supports the idea that these shifts were not simply baseline instability.

Behaviorally, the pattern was more encouraging than the alpha result. Life satisfaction improved meaningfully from pre-treatment to follow-up, with a very large effect size. Goal attainment also improved, again with a very large effect size, and participants showed statistically significant gains already by post-treatment. Pain sensitivity moved in a favorable direction, with the total score improving significantly from pre- to post-treatment and moderate-to-large effect sizes persisting at follow-up. Trait anxiety, though not state anxiety, also showed a sub-large effect size by follow-up.

A final detail deserves attention: change in frontal delta power correlated with change in goal attainment, and peak alpha frequency correlated with pain sensitivity. These correlations are exploratory and should not be overinterpreted in a sample of ten, but they are exactly the kind of signal one hopes to detect in a feasibility paper. They suggest there may be a trainable relationship between neural state change and clinically relevant function, even if the original alpha-power hypothesis was not borne out.

Discussion

This study does not show that neurofeedback is an effective treatment for sensory over-responsiveness. It does, however, suggest that studying neurofeedback in SOR is reasonable, tolerable, and clinically worth pursuing.

The paper’s most important scientific message is not that alpha training “worked.” It is that the protocol may have engaged neurophysiological and behavioral processes in a more complex way than the original target model predicted. In other words, the intervention may have done something meaningful, just not the exact thing the authors first set out to measure.

Clinicians and researchers should keep that distinction in view. In neurofeedback, we often inherit elegant hypotheses about a target rhythm, train that rhythm, and then discover that the relevant therapeutic signal may lie elsewhere: in another band, in network dynamics, in trait-like changes that emerge only at follow-up, or in interactions among oscillations rather than isolated power values. This paper fits that broader pattern. Alpha was the rationale, but frontal delta and theta may have captured the more behaviorally relevant changes.

For clients and referring professionals, the clinically appealing part of the study is the symptom pattern. Participants reported better life satisfaction, lower pain sensitivity, lower trait anxiety, and better progress toward individualized goals. Those are not trivial endpoints, especially in a condition that is often experienced as diffuse, exhausting, and poorly understood. At the same time, the study design leaves open a range of alternative explanations: expectancy effects, repeated measurement effects, nonspecific therapeutic engagement, regression to the mean, or simple relief associated with being studied and supported consistently.

For neurofeedback professionals, the protocol design raises several practical questions. First, should alpha power really be the primary target in adult SOR, or is SOR better conceptualized as a broader arousal-regulation problem involving frontal control systems and multi-band dynamics? Second, is resting alpha mean power sensitive enough to capture clinically important change in this population? The authors themselves point toward a more nuanced interpretation, noting that alpha dynamics may matter more than average alpha power. Third, what would happen with a longer training course, more explicit strategy coaching, or state-dependent training that includes controlled sensory challenge rather than eyes-closed resting only?

There is also a broader translational point here. SOR is not merely “sensory sensitivity” in a casual sense. It appears entangled with pain processing, arousal regulation, and emotional distress. That makes it a particularly interesting candidate for EEG-based intervention, but it also means single-band protocols may be too blunt. If the underlying phenotype reflects dysregulated salience, threat appraisal, and sensory gating, then individualized protocols may ultimately outperform a fixed alpha-uptraining model.

The limitations are substantial and should stay front and center. The sample was tiny and restricted to women. There was no sham control, active comparison, or blinded expectancy control. The primary target was negative. The behavioral measures were self-report-heavy. Multiple comparisons were explored in a small dataset. And because the participants were otherwise healthy adults, these findings do not automatically extend to more complex clinical populations where SOR co-occurs with ADHD, autism, chronic pain, trauma, or anxiety disorders.

Still, feasibility studies are supposed to answer a modest question: is this line of work worth doing properly? Here, the answer appears to be yes. The intervention was implementable, retention among completers was acceptable, the outcome battery was clinically relevant, and the signal pattern, while mixed, was interesting enough to justify a better-powered randomized design.

Brendan’s perspective

This paper lands in a space I find especially interesting clinically: cases where the person is not simply anxious, inattentive, or in pain, but persistently overwhelmed by ordinary sensory life. Those clients often look deceptively “complex” on paper, when in practice the picture is often more coherent. Their nervous system seems to stay too available to incoming stimulation, too easily recruited into vigilance, and too slow to settle once activated. That does not mean every case of sensory over-responsiveness should be treated with EEG neurofeedback. It does mean this population deserves more precise attention than it usually receives.

From alpha targets to individualized protocols

One of the most useful aspects of this study is that it did not cleanly confirm its original target. The protocol aimed to uptrain alpha, yet the main resting alpha outcome did not shift. I do not see that as a failure so much as a reminder: the nervous system rarely reads our protocol manual.

In practice, sensory clients are often not well served by rigid, one-size-fits-all assumptions about which frequency “should” be trained. A person with sensory defensiveness may present with high baseline arousal, sleep fragility, somatic vigilance, pain amplification, and anxiety features, but the EEG pathway into those symptoms will not look identical across individuals. Some may benefit from classic alpha-oriented calming work. Others may respond better to SMR-oriented stabilization, especially when the clinical picture includes poor inhibition, restlessness, light sleep, or difficulty maintaining a settled but alert state. Others still may need work focused more frontally, especially where overwhelm quickly turns into emotional reactivity or loss of executive control.

This is where I think individualized protocol design becomes essential. I would be reluctant to treat sensory over-responsiveness as a simple “low alpha problem.” I would want to know what happens in eyes open and eyes closed states, how the client shifts under mild demand, whether posterior calming is weak, whether frontal overcontrol or inefficiency is present, and how symptoms cluster in real life. The study gives us permission to keep alpha in the conversation without pretending alpha is the whole story.

What SOR can teach neurofeedback clinicians about arousal

Sensory over-responsiveness is also a good reminder that arousal is not just “too much energy.” It is patterned. It has a sensory signature, a pain signature, an attentional signature, and often a relational one too.

Many sensory clients do not start treatment by saying, “I am dysregulated.” They say things like: tags drive me crazy, restaurants are exhausting, I dread fluorescent lighting, I cannot think when multiple sounds are happening, I snap after a long day, or I feel pain more intensely than other people seem to. Those are not separate complaints floating around randomly. Very often they reflect a system with poor sensory gating and an elevated cost of processing everyday input.

For neurofeedback clinicians, that means arousal work should not be reduced to making the client feel relaxed in the chair. The real question is whether the brain and body can become less reactive, less effortful, and less defensive in everyday sensory environments. In some cases, that may mean beginning more gently than we would with a straightforward attention or performance client. Shorter trials, lower cognitive demand, and careful monitoring of fatigue, irritability, headaches, or rebound sensitivity can matter a great deal.

It also means body-based support is not optional. I would be very inclined to combine EEG work with complementary regulation strategies: paced breathing, predictable session structure, sensory pacing, environmental modifications, and practical transfer exercises that help the client notice early signs of overload before they reach the point of no return. Sometimes the best neurofeedback decision is not a more sophisticated protocol. Sometimes it is recognizing that the client first needs enough physiological safety to learn.

Bridging SOR, pain sensitivity, and anxiety in EEG practice

What I especially appreciate in this paper is that it does not isolate sensory symptoms from pain and anxiety. Clinically, that overlap is common. The client who is hypersensitive to touch, noise, and visual clutter may also report migraines, fibromyalgia-like features, menstrual sensitivity, insomnia, digestive reactivity, or chronic tension. Another may look more psychological on referral, labeled as anxious or highly sensitive, when the sensory load is actually one of the main drivers of the anxiety.

This matters for EEG practice because protocol goals should follow the functional pattern, not just the diagnostic label. If the person’s day collapses under cumulative sensory load, then training should support better threshold management, recovery, and inhibitory control. If pain sensitivity is prominent, I would think carefully about protocols that promote calmer sensory processing without flattening alertness. If anxiety is secondary to a constantly over-recruited system, then “treating anxiety” alone may miss the more foundational issue.

In practical EEG terms, this is where I start thinking in layers rather than single symptoms. Posterior alpha support may still be useful in some presentations. SMR work may help with sensory stability and sleep-related restoration. Frontal work may be relevant when overwhelm rapidly becomes irritability, panic, or cognitive shutdown. The exact montage depends on the person, of course, but the broader point is that sensory over-responsiveness should push us toward formulation-based neurofeedback, not cookbook neurofeedback.

What this study offers, even with all its methodological limits, is a clinically encouraging message: sensory overwhelm may be trainable. Not always simply, and probably not through one universal mechanism, but trainable enough to deserve serious clinical curiosity. That feels like a worthwhile direction for the field.

Conclusion

This 2022 feasibility study offers an encouraging but appropriately complicated first look at neurofeedback for sensory over-responsiveness in adults. The encouraging part is that participants tolerated the intervention, most appeared able to engage with the training process, and several behavioral outcomes moved in a clinically favorable direction by follow-up, including life satisfaction, goal attainment, pain sensitivity, and trait anxiety. The complicated part is just as important: the study’s primary mechanistic target, resting alpha power, did not significantly change.

That combination makes this paper more useful than a simple positive pilot. It reminds us that early neurofeedback research often reveals a mismatch between the intended target and the observed outcome, and that such mismatches can still teach us something valuable. In this case, the findings hint that adult SOR may be trainable, but perhaps not through the exact physiological pathway the authors originally proposed.

For now, the study supports feasibility, not efficacy. It should not be used as stand-alone proof that alpha neurofeedback treats SOR. But it does provide a credible rationale for larger, controlled trials that use individualized protocols, stronger control conditions, and richer neurophysiological modeling. That is a worthwhile next step, and for a population with few adult-focused options, it is a genuinely promising one.

References

Hamed, R., Mizrachi, L., Granovsky, Y., Issachar, G., Yuval-Greenberg, S., & Bar-Shalita, T. (2022). Neurofeedback therapy for sensory over-responsiveness: A feasibility study. Sensors, 22(5), 1845. https://doi.org/10.3390/s22051845