Researchers have found that carefully synchronized electrical stimulation of two brain regions can improve aspects of emotional action control in people with high levels of social anxiety.
The 2026 study is particularly interesting because the stimulation was designed around a specific pattern of communication between brain regions rather than simply applying electrical current to one area of the brain. Participants performed better when researchers synchronized different brain rhythms in the prefrontal and sensorimotor cortices, and brain imaging suggested that the stimulation altered communication between those regions (Meijer et al., 2026).
The researchers describe their findings as opening a path toward mechanistically informed interventions for persistent avoidance in anxiety disorders.
That possibility deserves attention.
But there is a substantial distance between performing better on an emotional approach-avoidance task while receiving experimental brain stimulation and successfully treating social anxiety disorder.
The study did not test whether participants became less socially anxious. It did not measure whether they avoided fewer real-life social situations after the experiment. It did not compare the intervention with psychotherapy or medication. And the participants were selected for high social anxiety scores rather than recruited as a clinical treatment group with confirmed social anxiety disorder.
So what did the experiment actually show?
Key Takeaways
- Researchers studied 49 adults selected for high self-reported social anxiety.
- Participants received a form of noninvasive brain stimulation called transcranial alternating current stimulation, or tACS.
- Electrical rhythms were applied to the lateral prefrontal cortex and sensorimotor cortex while participants performed an emotional approach-avoidance task.
- In-phase stimulation improved behavioral performance and strengthened functional coupling between the targeted brain regions.
- Participants whose brains showed stronger physiological engagement with stimulation experienced greater improvement in emotional action control.
- The experiment builds directly on earlier work using a similar stimulation technique in nonanxious participants.
- Participants were not recruited on the basis of a confirmed clinical diagnosis of social anxiety disorder.
- The study did not demonstrate improvement in social anxiety symptoms, real-world avoidance, quality of life, or long-term functioning.
- The findings therefore support a possible brain mechanism and experimental intervention, not an established treatment for social anxiety disorder.
Why Avoidance Matters in Social Anxiety
Social anxiety can involve considerably more than ordinary nervousness before meeting unfamiliar people or speaking in public.
People with substantial social anxiety may fear being watched, judged, embarrassed, rejected, or negatively evaluated by others. Avoiding feared situations can provide immediate relief, but repeated avoidance can also help maintain anxiety by limiting opportunities to learn that feared outcomes may not occur.
This makes avoidance an important target in psychological research and established treatments for social anxiety disorder.
Meijer et al. (2026) focused on something more specific: the brain’s ability to override automatic tendencies to approach or avoid emotionally meaningful social signals.
Their question was whether manipulating communication between particular brain regions could improve that control.
How Can Electricity Influence Brain Activity?
The researchers used transcranial alternating current stimulation, commonly abbreviated tACS.
Unlike implanted brain stimulation, tACS does not require surgery. Electrodes placed on the scalp deliver weak alternating electrical currents.
The important feature of this experiment was not simply where the researchers stimulated the brain. It was also when the electrical rhythms occurred relative to one another.
The investigators targeted communication between the lateral prefrontal cortex, or lPFC, and the sensorimotor cortex, or SMC.
Previous research had linked communication between these areas to emotional action control (Bramson et al., 2018).
The researchers specifically targeted interactions between two types of neural rhythms:
Theta activity at 6 Hz was applied over the right lateral prefrontal cortex.
Gamma activity at 75 Hz, modulated by the 6 Hz theta rhythm, was applied over the left sensorimotor cortex.
The gamma activity could be synchronized with either the peaks or troughs of the theta rhythm.
That created two active conditions called in-phase and anti-phase stimulation.
There was also a sham condition intended to serve as a control (Meijer et al., 2026).
What Did the Participants Actually Have to Do?
This is one of the most important details for understanding the findings.
Participants were not sent into a party, workplace, restaurant, classroom, or other social environment to see whether they became less avoidant.
They were lying inside an MRI scanner performing a computerized approach-avoidance task.
They held a joystick and viewed happy and angry faces.
During relatively intuitive trials, they were instructed to pull the joystick toward themselves for happy faces and push it away for angry faces.
During the more difficult condition, those instructions were reversed.
Participants had to push away happy faces and pull angry faces toward themselves.
The second condition requires overriding an automatic emotional response and following a conflicting rule.
Each participant completed 576 trials across two testing days (Meijer et al., 2026).
During sham stimulation, participants made significantly more errors when they had to override the automatic emotional response. The odds of an incorrect response were approximately 42% higher during these incongruent trials than during the easier congruent trials.
The task therefore produced the emotional-control challenge the researchers expected.
But it remains a laboratory measure of emotional action control.
That distinction becomes important when considering claims about treating social anxiety.
What Did the Brain Stimulation Change?
In-phase stimulation produced several notable findings.
First, it strengthened functional coupling between the dorsolateral prefrontal cortex, or dlPFC, and sensorimotor cortex during the emotional-control challenge (Meijer et al., 2026).
The researchers also examined a brain region they had not targeted, the primary visual cortex. They did not observe the same connectivity effect there, providing some evidence that the effect was specific to the targeted network.
The behavioral results were also linked to what researchers called target engagement.
Participants whose dlPFC showed a stronger physiological response to stimulation tended to experience greater improvement in emotional action control during in-phase stimulation compared with anti-phase stimulation.
The preregistered analysis found a statistically significant engagement-dependent effect, and the association was present during in-phase stimulation but not during anti-phase stimulation (Meijer et al., 2026).
That combination of behavioral and imaging evidence is one of the study’s strengths.
Researchers did not simply observe that participants pressed a joystick more accurately. They also found evidence that the stimulation affected communication within the brain circuit they were attempting to manipulate.
The Effect Was Not the Same for Everyone
One of the more interesting findings may be the variation between participants.
The behavioral benefit of in-phase stimulation was related to how strongly an individual’s brain responded to the intervention.
Participants showing greater stimulation-related changes in the dlPFC experienced greater improvement in emotional action control.
The researchers also found that this measure of target engagement varied with trait anxiety. People with higher trait-anxiety scores showed stronger dlPFC responses to stimulation (Meijer et al., 2026).
This could eventually matter if brain stimulation becomes personalized.
A fixed stimulation protocol may not affect every person’s brain in exactly the same way. Differences in anatomy and ongoing neural activity can influence how an externally applied electrical field interacts with the brain.
However, the researchers caution that direct measurements of neural responses currently appear more informative about stimulation engagement than anxiety scores alone.
In other words, the study does not provide a simple rule such as “more severe anxiety means brain stimulation will work better.”
How Does This Compare With the Earlier Study?
The 2026 experiment did not emerge in isolation.
Previous work by the same research group had already examined this brain circuit and stimulation technique.
Bramson et al. (2018) reported evidence linking emotional action control to theta-gamma coupling between lateral prefrontal and sensorimotor regions.
Bramson et al. (2020) then used dual-site tACS in nonanxious participants and found that synchronizing these rhythms could facilitate control over automatic emotional action tendencies.
That earlier experiment provided an important foundation for the new study.
The 2026 research asks a logical next question: Would a similar manipulation work in people with high social anxiety?
The answer appears to be yes, at least for the laboratory measure tested.
But the comparison also revealed something interesting about the brain.
Social Anxiety May Change Which Prefrontal Region Does the Work
The new study did not simply reproduce the earlier neural findings.
In the previous nonanxious participants, emotional control had been associated with a more anterior prefrontal region called the lateral frontal pole, or FPl.
The highly anxious participants in the new study showed stronger involvement of the dorsolateral prefrontal cortex instead (Meijer et al., 2026).
The authors interpret this as consistent with earlier evidence that people with high anxiety may rely on a different prefrontal circuit when controlling emotional actions.
They also examined whether the difference might simply reflect the sex composition of the studies because the earlier experiment involved men, while the new study included both women and men.
The new sample included 38 women and 11 men in the relevant analysis. Researchers found no significant evidence that sex explained the difference in prefrontal recruitment.
That makes anxiety-related differences a plausible explanation, although the study cannot by itself establish why the neural shift occurs.
Synchronized brain stimulation improved emotional action control and altered communication between targeted brain regions in people with high social anxiety, but the study did not show that tACS treats social anxiety disorder.
Were These Patients With Social Anxiety Disorder?
This is where careful wording matters.
The study involved highly socially anxious individuals, but that does not mean every participant had clinically diagnosed social anxiety disorder.
Researchers initially recruited 52 students from Radboud University. After three exclusions, 49 participants remained.
Participants were selected using the Liebowitz Social Anxiety Scale, or LSAS.
To qualify, they needed a score of at least 30. The average LSAS score was 64.4, with scores ranging from 31 to 106 (Meijer et al., 2026).
The researchers chose that cutoff because previous work found that it provides useful sensitivity and specificity when screening for social anxiety disorder.
But a screening score is not the same as a clinical diagnosis established through a diagnostic evaluation.
That distinction does not weaken the experiment’s finding that stimulation affected emotional control in people with substantial social anxiety.
It does limit how confidently the results can be generalized to patients receiving treatment for diagnosed social anxiety disorder.
Did Brain Stimulation Treat Their Social Anxiety?
No. The study did not establish that.
This is probably the most important distinction for readers.
The researchers demonstrated an improvement in emotional action control during an experimental task.
They did not demonstrate that stimulation reduced participants’ social anxiety scores after treatment, reduced fear of social situations, made participants more willing to enter feared situations, reduced real-world avoidance, improved relationships or occupational functioning, prevented relapse, or produced lasting clinical improvement.
The intervention was administered while participants performed the laboratory task.
The experiment therefore answers a mechanistic question much more convincingly than a therapeutic one.
It suggests that researchers can manipulate a neural circuit associated with emotional action control and influence behavior while that circuit is being challenged.
Whether doing so can meaningfully improve social anxiety disorder remains an unanswered question (Meijer et al., 2026).
Why the Word “Control” Can Be Misleading
“Improved emotion control” can sound as though participants became better at regulating anxiety generally.
That is broader than what was measured.
The experiment examined emotional action control, meaning the ability to override an automatic approach or avoidance response and execute a competing action.
It did not measure all aspects of emotional regulation.
Nor did it show that people could consciously control feelings of anxiety after stimulation.
This distinction matters because social anxiety involves much more than approach-avoidance motor responses. It can involve anticipatory anxiety, negative self-evaluation, physical symptoms, attention biases, fear of scrutiny, avoidance, and post-event rumination.
A technique that improves one neural process involved in avoidance would not necessarily improve all of those symptoms.
The Study Has an Important Strength: It Was Preregistered
The experiment was preregistered.
That means important aspects of the researchers’ hypotheses and analysis plan were specified before the results were analyzed.
The study was also designed to replicate and extend an effect previously reported in nonanxious participants.
Researchers calculated the required sample size based on the effect observed in their earlier work and planned for approximately 50 participants (Meijer et al., 2026).
These features make the study stronger than a small exploratory experiment in which researchers test numerous possibilities and report whichever happen to produce significant results.
At the same time, preregistration does not eliminate the need for independent replication or clinical trials.
Does This Mean tACS Is a New Treatment for Social Anxiety Disorder?
Not yet.
Calling this a treatment for social anxiety disorder would move beyond the evidence.
The study provides a proof of mechanism showing that synchronized noninvasive electrical stimulation can alter a neural circuit associated with emotional action control in highly socially anxious individuals.
A clinical treatment study would require substantially different outcomes.
Researchers would need to recruit people with diagnosed social anxiety disorder, administer treatment across an appropriate period, measure validated clinical symptoms before and after treatment, evaluate real-world functioning and avoidance, determine how long any improvement lasts, and compare the intervention with an appropriate control.
Eventually, researchers would also want to know how it compares with established treatments.
The current experiment was not designed to answer those questions.
Could It Eventually Complement Exposure Therapy?
This is an interesting possibility, but it remains hypothetical.
Avoidance plays an important role in maintaining anxiety. Effective psychological treatment often involves learning to approach rather than continually avoid feared situations.
If researchers eventually establish that synchronized brain stimulation reliably improves the neural control processes needed to override avoidance, one conceivable application could be pairing stimulation with behavioral therapy.
But the 2026 study did not test tACS combined with exposure therapy.
It therefore cannot tell us whether stimulation would make psychotherapy more effective, accelerate treatment, help people who have not responded to therapy, or produce any additional clinical benefit.
Those would be valuable questions for future randomized trials.
What Are the Main Limitations?
The most important limitation is the difference between the experimental outcome and the clinical claim readers might infer.
The study measured performance on a joystick-based emotional action-control task, not improvement in social anxiety disorder.
The sample was also relatively small, with 49 participants completing the experiment, and participants were young adults recruited from a university population. Their mean age was 24 years.
Participants were selected by self-reported social anxiety rather than a confirmed clinical diagnosis for every participant.
The stimulation was tested during a highly controlled task inside an MRI scanner. Everyday social interactions are considerably more complicated. They involve unpredictable people, conversation, self-consciousness, competing cognitive demands, memories, expectations, and consequences.
The researchers themselves identify this as an important future question. They note that it remains unknown whether improved dlPFC-SMC coordination would remain behaviorally useful when emotional challenges occur alongside other cognitive demands, as they commonly do in real life (Meijer et al., 2026).
Finally, this study does not establish whether the observed effects persist after stimulation ends.
What Does This Research Actually Tell Us?
The most interesting result is narrower, but arguably more scientifically meaningful, than saying that brain stimulation can treat social anxiety.
Researchers identified a brain circuit involved in overriding automatic emotional actions. Earlier work showed that synchronizing activity between parts of this circuit could improve emotional action control in people without high anxiety (Bramson et al., 2020).
The 2026 study extended that experiment to people with high social anxiety. In-phase stimulation improved performance during an emotional approach-avoidance task, strengthened functional communication between targeted prefrontal and sensorimotor regions, and produced greater behavioral improvements in participants whose brains showed stronger physiological engagement with the stimulation (Meijer et al., 2026).
Together, the findings provide evidence that precisely timed stimulation of communication between brain regions can influence a neural process relevant to social anxiety.
The next question is considerably more difficult: can changing that mechanism produce durable improvements in social anxiety symptoms and real-world avoidance?
Answering that will require clinical trials involving people with diagnosed social anxiety disorder, validated symptom measures, longer-term follow-up, and outcomes that reflect everyday functioning rather than performance on a laboratory task.
Until such evidence exists, dual-site tACS should be viewed as a promising experimental neuroscience approach, not an established treatment or alternative to existing therapies for social anxiety disorder.
References
Bramson, B., Jensen, O., Toni, I., & Roelofs, K. (2018). Cortical oscillatory mechanisms supporting the control of human social-emotional actions. The Journal of Neuroscience, 38(25), 5739–5749. https://doi.org/10.1523/JNEUROSCI.3382-17.2018
Bramson, B., den Ouden, H. E. M., Toni, I., & Roelofs, K. (2020). Improving emotional-action control by targeting long-range phase-amplitude neuronal coupling. eLife, 9, e59600. https://doi.org/10.7554/eLife.59600
Meijer, S., Bramson, B., Toni, I., & Roelofs, K. (2026). Improving emotion control in social anxiety by targeting rhythmic brain circuits. The Journal of Neuroscience, 46(18), e0769252026. https://doi.org/10.1523/JNEUROSCI.0769-25.2026
