Rapid Relief for Stubborn Depression: rTMS Plus Neurofeedback

In a brief but provocative clinical report, Bruno and colleagues (2021) describe a man with treatment-resistant depression (TRD) who received an intensive combination of repetitive transcranial magnetic stimulation (rTMS) and neurofeedback (NFB), showing substantial symptom improvement by day 5 and full remission by day 10. Because this paper was published more than a year ago, it belongs from the archives—yet the clinical question it raises feels extremely current: what happens when we stop treating depression as a purely “chemical” problem and start treating it as a brain-network regulation problem?

Neurofeedback is a form of biofeedback in which real-time brain signals (most commonly EEG) are translated into immediate visual or auditory feedback so the brain can learn—through reinforcement—more adaptive patterns of activity; biofeedback more broadly uses real-time physiological signals (like heart rate variability, muscle tension, or skin conductance) to train self-regulation.

In TRD, the appeal of these approaches is obvious. When medication trials begin to resemble a painful game of musical chairs, directly targeting brain function can become a coherent next step, especially when symptom relief is needed to restore daily functioning and hope. At the same time, “brain-based” does not mean “therapy-free.” For most non-urgent presentations, psychotherapy should remain the first order of business, because meaning-making, behaviour change, relational repair, and skills training are not optional add-ons—they are often the scaffolding that makes biological interventions stick.

The most interesting question, then, isn’t whether psychotherapy, medication, rTMS, or neurofeedback is the single “best” tool. The real question is sequencing and synergy: how do we combine methods so that symptom relief creates space for durable change, and durable change reduces the need for chronic crisis management?

Methods

The report describes a 37-year-old man admitted for a prolonged depressive episode that had not responded to extensive pharmacotherapy, consistent with treatment-resistant depression based on multiple prior antidepressant trials. At admission, his medication regimen was complex and included several antidepressants and augmenting agents, including fluoxetine, sertraline, mirtazapine, lamotrigine, lithium, clonazepam, and pregabalin.

The intervention combined sequential bilateral rTMS with EEG-neurofeedback in an intensive schedule: two treatment blocks per day across 10 days (a total of 20 rTMS sessions and 20 neurofeedback sessions).

The rTMS portion targeted the dorsolateral prefrontal cortex (DLPFC) bilaterally in sequence. Left DLPFC stimulation was delivered at 10 Hz with 3000 stimuli per session (40 pulses per train, 75 trains, 26-second inter-train interval) at 110% of the motor threshold (MT) for hand muscles. Right DLPFC stimulation followed at 1 Hz with 1200 stimuli per session (60 pulses per train, 20 trains) at 80% MT.

Immediately after each rTMS session, neurofeedback was administered twice daily in sessions described as 18/20 minutes. The training goal was to increase alpha and theta power at C3 and C4 (sensorimotor strip) under eyes-open conditions, using both visual and auditory feedback.

Clinical outcomes were tracked using standard symptom scales: the Hamilton Rating Scale for Depression, the Hamilton Anxiety Scale, and the Beck Depression Inventory at baseline (T0), after 10 combined sessions (day 5), and at the end of the protocol (day 10).

Results

Clinically, the report describes a rapid and substantial change. At baseline, the patient presented with severe depressive symptoms and moderate anxiety. After the first half of the protocol (10 rTMS sessions plus 10 neurofeedback sessions by day 5), the authors describe meaningful improvement in depressive symptoms. By day 10—after 20 combined sessions—the patient was reported as clinically remitted. Anxiety symptoms also decreased across the same time window.

From a tolerability standpoint, the combined protocol was described as safe and well tolerated. The only side effect reported was transient, mild headache at the beginning of treatment. That matters, because one of the most frustrating parts of TRD is not only the lack of response, but the accumulation of burdens: side effects, medication interactions, emotional exhaustion, and a growing sense that nothing will work.

One of the most clinically striking details is what happened to pharmacotherapy during improvement. As symptoms improved, medication was reduced: fluoxetine was suspended, and sertraline, mirtazapine, and clonazepam were reduced by half. While a single case cannot determine which component drove change—or whether the improvement would persist long-term—the pattern remains notable: symptom relief appeared quickly, was accompanied by decreased anxiety, and coincided with simplification of a complicated medication stack.

Discussion

This paper is a case report, so it can’t answer the big questions of efficacy, durability, or mechanisms with the certainty that randomized trials can. Still, it offers a clinically useful hypothesis: in TRD, there may be value in pairing a top-down neuromodulation tool (rTMS) with a bottom-up learning tool (neurofeedback) in close temporal proximity, potentially creating a synergy between state change and skill acquisition.

Why might sequencing matter? One plausible clinical interpretation is that rTMS shifts prefrontal network excitability and cognitive-emotional control capacity—helping the brain become more trainable—while neurofeedback then reinforces patterns associated with stability, regulation, and improved functional control. In this protocol, rTMS targeted bilateral DLPFC with excitatory-frequency stimulation on the left and inhibitory-frequency stimulation on the right, a pairing commonly intended to address imbalances in prefrontal activity implicated in depression. Neurofeedback then focused on increasing alpha and theta power at C3/C4, which may relate to calming arousal and supporting regulation through sensorimotor networks, though the report does not directly test mechanistic explanations.

In real-world clinical settings, TRD is rarely a purely biological puzzle. It is also a behavioural, interpersonal, and existential grind. That’s why—outside urgent safety situations—psychotherapy should remain first-line: it builds emotional literacy, cognitive flexibility, and behavioural traction, and it gives symptom relief somewhere to “land.” The interesting clinical move is not psychotherapy versus brain-based interventions, but psychotherapy plus targeted neurophysiological training when progress stalls, side effects accumulate, or functioning remains flat despite strong therapeutic work.

This case also touches a delicate but common reality: medication regimens can become extraordinarily complex over time, especially when clinicians are trying to help someone who is suffering and not improving. Even when such regimens are partially effective, they may create a fragile equilibrium that’s difficult to adjust. The reduction of medications reported here occurred in the context of improved symptoms and hospital monitoring, which is exactly the context in which careful simplification is most feasible. The takeaway is not “stop meds,” but rather: when symptoms improve, there may be an opportunity to rationalize treatment, reduce side-effect load, and regain clarity about what is actually helping.

Finally, the report raises an interpretive theme that deserves more attention across the field: measurement. If the brain is the organ we are treating, it is worth considering how often we proceed without measuring brain function in any meaningful way. While this case does not include qEEG outcomes, it naturally invites a more measurement-informed workflow: baseline symptom metrics, baseline physiology (sleep, arousal), and—where appropriate—brain-based assessment to guide selection and individualization of neuromodulation and neurofeedback targets.

Brendan’s perspective

In treatment-resistant depression, it’s easy for care to drift into a kind of clinical improvisation: adjust one medication, add another to patch the first, introduce an augmenting agent when the response stalls, then circle back because the side-effect burden becomes its own illness. I understand why it happens. Everyone is trying to help, and when someone is suffering, “doing something” can feel kinder than waiting. But over time, this can create pharmacological architecture that is intricate on paper and brittle in real life—hard to maintain, hard to interpret, and even harder to unwind.

This is where brain-directed interventions can become a more coherent next step—not as a rebellion against medication, but as a shift in logic. Depression is not only a chemistry story; it’s also a brain-network regulation story. Many people with persistent depression describe a nervous system that feels stuck in one gear: rumination that won’t release, motivation that won’t spark, sleep that won’t restore, and an emotional range that narrows until life feels flat. When that pattern persists across well-conducted psychotherapy and multiple medication trials, it is clinically reasonable to ask whether we should intervene more directly with brain function and self-regulation capacity.

I like to think of rTMS and neurofeedback as complementary tools that answer two different questions. rTMS is a state-shifter: it can nudge cortical excitability and network dynamics in ways that may reduce the “locked-in” quality of depressive circuitry. Neurofeedback is a learning tool: it gives the brain a way to practice, reinforce, and stabilize more adaptive patterns of regulation. The case report we’re discussing—rTMS followed immediately by EEG-neurofeedback twice daily over ten days—fits a simple clinical intuition: change the state, then teach the skill. The first step may make the nervous system more trainable; the second may help the gains become something the person can carry forward, rather than something that fades when the external input stops.

But there’s an important boundary condition here: psychotherapy first for most non-urgent cases. Not because brain-based approaches are “less real,” but because the brain learns in context. If someone’s life remains a chronic stress experiment—unrelenting threat cues, isolation, sleep disruption, unresolved grief, unstable relationships—the nervous system will keep adapting to survive. Psychotherapy helps change the experiment. It builds emotional literacy, cognitive flexibility, behavioural traction, and relational repair. Neurofeedback and rTMS often work best as accelerants and stabilizers for that deeper work, not replacements for it.

A second boundary condition is measurement. We routinely measure other organs before escalating treatment; brain-based care deserves the same respect. A qEEG assessment won’t diagnose a person or substitute for clinical judgment, but it can add useful constraints when treatment has become muddy. Is the presentation dominated by hyperarousal and threat physiology, or by underarousal and disengagement? Are there patterns consistent with cognitive rigidity or attentional instability? Is sleep physiology showing up as the hidden driver of mood symptoms? Even when qEEG findings are nonspecific, the act of measuring—paired with symptom tracking and sleep/autonomic metrics—often improves clinical decision-making simply by making the plan more testable.

In practical EEG-neurofeedback terms, this is where protocol individualisation matters. The study’s neurofeedback targeted increased alpha and theta at C3/C4 with eyes open. That might be a good fit for some people—especially those with anxious arousal, agitation, or difficulty settling. For others, “more calm” is not the immediate need; some depressed clients are already emotionally constricted, cognitively slowed, or fatigued, and overly sedating protocols can sometimes worsen flattening. This is why I treat protocols as hypotheses rather than prescriptions: adjust frequency targets, reward thresholds, electrode placements, and session dosing based on response.

In many clinics, sensorimotor strip training in the SMR range (often around 12–15 Hz, commonly at C3/C4) can be a useful stabilizing lever for sleep and arousal regulation. When it works well, it can support more stable sleep onset, reduce agitation, and improve the “braking system” that helps people pause instead of react. But it should be deployed thoughtfully and monitored closely, not treated as a universal fix. Depression is heterogeneous, and protocol choices should reflect the person in front of you—symptom profile, sleep quality, anxiety load, cognitive tempo, medication status, and how their nervous system responds session by session.

Integration is the difference between “a cool intervention” and a coherent care pathway. When brain-based treatment is embedded into a plan—psychotherapy with clear targets, sleep stabilization, autonomic regulation work (often HRV biofeedback when the stress system is clearly dysregulated), behavioural activation that is realistic rather than punitive, and coordination with the prescriber—patients often experience something deeper than symptom reduction: they experience agency. The nervous system stops being a mysterious adversary and becomes something that can be trained, negotiated with, and understood.

Finally, a word on medication: symptom relief can create an opportunity to simplify pharmacotherapy, but simplification should be approached with the same seriousness as initiation. The case report describes reduction of several medications as symptoms improved. That’s encouraging, yet it’s also a reminder that deprescribing is a clinical skill—slow, collaborative, and informed by the person’s stability rather than a calendar. The goal isn’t to “prove” anything about meds; the goal is to reduce suffering while improving long-term resilience and clarity. (And remember, unless you are the treating psychiatrist, medication choices just aren't your job.)

If there’s a unifying clinical theme here, it’s coherence. When psychotherapy provides meaning and behavioural leverage, and brain-based interventions provide state flexibility and self-regulation training, care begins to feel less like trial-and-error and more like guided learning. In the long arc of recovery, that shift—from managing symptoms to restoring regulatory capacity—may be one of the most hopeful directions we have.

Conclusion

This case report describes an intensive 10-day protocol combining sequential bilateral rTMS over the DLPFC with immediate follow-on EEG-neurofeedback training focused on increasing alpha and theta at C3/C4. In a single patient with treatment-resistant depression and moderate anxiety, the authors observed rapid improvement by day 5 and remission by day 10, alongside reduced anxiety, minimal side effects, and a notable simplification of an unusually complex medication regimen.

As a single case, the findings can’t prove efficacy or tell us whether the benefits persist. But the clinical idea is compelling: pair a neuromodulation tool that can shift brain state quickly with a self-regulation training approach that may help stabilize and maintain those shifts. In practice, the most powerful frame is integration—psychotherapy first for most non-urgent cases, then brain-based interventions when needed, guided by measurement wherever possible. When symptom relief and skill-building happen together, treatment can move from managing depression to restoring the brain’s capacity to regulate itself—and that’s a genuinely hopeful direction.

References

Bruno, A., Celebre, L., Tagliavia, G., Zoccali, R. A., & Muscatello, M. R. A. (2021). Combining intensive repetitive transcranial magnetic stimulation with neurofeedback in a case of treatment-resistant depression. Indian Journal of Psychiatry, 63(2), 199–200. https://doi.org/10.4103/psychiatry.indianjpsychiatry_351_20