Medically reviewed by Dr. Saswato Majumdar, MBBS, MD (PM&R) | Last updated: August 04, 2026 | Reading time: 9 minutes
Quick Answer
rTMS stroke recovery operates through the interhemispheric competition model: low-frequency rTMS suppresses maladaptive contralesional inhibition, while high-frequency rTMS upregulates ipsilesional excitability. A 2025 systematic review confirmed rTMS reduces central post-stroke pain and modestly improves upper limb function, but found no significant effect on depression or anxiety and rated the evidence quality as low. The right approach is selective prescribing based on imaging, timing, and patient profile, not routine use.
Key Takeaways
- rTMS stroke recovery is evidence-supported in specific contexts, not as a blanket add-on.
- The 2025 systematic review confirmed rTMS reduces central post-stroke pain and provides modest upper limb motor improvement over sham. No significant effect on depression or anxiety was found.
- Evidence quality across rTMS stroke trials is consistently rated low due to small samples, heterogeneous protocols, and limited follow-up.
- The four strongest predictors of rTMS response are: preserved corticospinal tract integrity on DTI, BDNF Met-negative genotype, moderate rather than severe impairment, and treatment in the subacute phase.
- In a setting without DTI or genetic testing, use rTMS as a time-limited trial (4 weeks) with structured reassessment, not as open-ended standard care.
- For post-stroke pain and aphasia, the emerging data justifies referral to an rTMS-capable centre. For routine motor rehabilitation without selection criteria, the evidence does not yet justify the resource.
The Problem With How rTMS Is Being Used
Repetitive transcranial magnetic stimulation has become the clinical equivalent of adding turmeric to a recipe. It is being layered onto rehabilitation programmes with the optimism of someone who has read abstracts without finishing the papers.
The enthusiasm is understandable. The mechanism is plausible, the technology is non-invasive, and the early trial data is genuinely interesting. The problem is that rTMS stroke recovery is being prescribed as a population-level intervention when the evidence clearly supports something far more selective.
In 2026, the honest appraisal is more nuanced than either the enthusiasts or the sceptics admit. The clinician deserves a clear framework for when to use rTMS, when to skip it, and when to acknowledge that the evidence does not yet answer the question. Many patients reaching specialist centres have already tried and failed to respond to rTMS. That pattern is not patient failure. It is a selection failure. The patients who hit a stroke recovery plateau and get offered rTMS without selection criteria are often precisely the patients least likely to respond.
How Does rTMS Work in Stroke Recovery?
The rationale for rTMS in stroke is grounded in the interhemispheric competition model.
After a unilateral stroke, the contralesional (undamaged) hemisphere exerts abnormal inhibitory influence over the damaged hemisphere via transcallosal pathways. This maladaptive inhibition suppresses motor recovery by chronically downregulating excitability in the ipsilesional cortex.
Two Therapeutic Approaches
Low-frequency rTMS (typically 1 Hz) over the contralesional motor cortex suppresses this inhibitory overdrive, in theory releasing the ipsilesional hemisphere to reorganise. The logic is counter-intuitive but biologically sound: weakening the healthy hemisphere’s suppression of the damaged one can allow more recovery.
High-frequency rTMS (typically 5 to 10 Hz) over the ipsilesional motor cortex takes the opposite approach, directly upregulating cortical excitability in the damaged hemisphere to promote plasticity at the site of injury.
Dual-site protocols combining both approaches simultaneously have also been explored and are represented in Reference 3 (Chen et al., 2021). They modulate GABA-mediated inhibition to produce additive effects on motor recovery.
The interhemispheric competition model is well established (Soekadar SR and colleagues have contributed substantially to this framework). However, applying it requires knowing that the specific patient’s lesion pattern actually produces this inhibitory asymmetry, which is not always the case.
What Does the Evidence Actually Show?
The trial data is more mixed than conference presentations suggest.
A 2025 systematic review and meta-analysis of rTMS for central post-stroke pain found that rTMS can reduce pain and provide some upper limb motor improvement compared to sham or conventional therapy alone (Liu et al., 2025). This is a meaningful positive signal for a notoriously difficult-to-treat condition.
However, the same review reported no significant effect on cognitive symptoms including depression and anxiety. Quality of evidence across trials was rated as low.
Why the Variation in Outcomes Is Not Noise
The variation in patient outcomes across rTMS stroke trials is not statistical noise. It is a clinical signal telling us that the treatment is differentially effective based on patient characteristics that most protocols currently ignore.
The key moderators are:
- Corticospinal tract integrity on diffusion tensor imaging
- Lesion location and whether it disrupts interhemispheric balance as the competition model predicts
- BDNF Val66Met genotype (Met allele carriers show blunted LTP-like plasticity responses)
- Time post-stroke (subacute phase responds most reliably)
- Baseline impairment severity (moderate impairment predicts better response than severe)
Current rTMS protocols largely ignore these moderators. Until trials are designed with adequate stratification on these variables, aggregate effect sizes will continue to look modest and variable (Ahmed et al., 2025).
Who Responds to rTMS? The Modifiers That Matter
The patients most likely to respond to rTMS stroke recovery are a specific subgroup, not the full stroke population.
The Ideal Candidate
| Factor | What Predicts Response |
|---|---|
| Corticospinal tract integrity | DTI-confirmed partial rather than complete disruption |
| Impairment severity | Moderate rather than severe upper limb paresis |
| BDNF genotype | Met-negative (Val/Val) genotype |
| Time post-stroke | Subacute phase: roughly 2 weeks to 6 months |
| Lesion location | Cortical or subcortical near M1, not deep brainstem or cerebellar |
The Patient Unlikely to Respond
Using rTMS in a patient with complete corticospinal tract disruption and severe upper limb deficits three years post-stroke, without imaging to guide decision-making, is not evidence-based practice. It is optimism that has not earned its clinical status.
The moderate impairment criterion matters particularly. Patients with near-complete paralysis may lack sufficient residual cortical representation to benefit from rTMS-induced excitability changes. Patients with mild impairment often continue to improve with conventional rehabilitation alone and may not need rTMS at all.
For patients in the subacute window who are not yet plateauing, the primary focus should be intensive structured neurorehabilitation with high repetition and task-specific practice. rTMS is a potential adjunct to that, not a replacement for it.
A Practical Clinical Framework for rTMS Prescribing
Where does this leave the busy clinician without access to DTI or genetic testing? It leaves them with a structured, time-limited trial framework and honest consent conversations.
For Clinicians Without Imaging Guidance
If DTI is unavailable, use the following practical checklist before prescribing rTMS:
- Is the stroke ischaemic with a cortical or subcortical (not complete brainstem) distribution?
- Is impairment moderate rather than severe?
- Is the patient in the subacute to early chronic phase (ideally within 6 months)?
- Has the patient received adequate conventional therapy and failed to reach functional targets?
- Does the patient have preserved attention and motivation for rTMS sessions?
If the answer is yes to most of these, a structured 4-week trial with reassessment is justifiable.
The 4-Week Trial Structure
- Define the endpoint upfront. What constitutes a meaningful response? A 4-point or greater improvement on Fugl-Meyer? Measurable pain score reduction? Pre-specify this with the patient.
- Reassess at 4 weeks. If the pre-specified endpoint has not been approached, stop and redirect resources.
- Be transparent about evidence quality. Patients deserve to know the evidence is promising but limited, not that rTMS is an established standard of care.
For patients already receiving stroke rehabilitation programmes, rTMS should be layered onto existing structured therapy, not offered as a standalone.
When rTMS Is Worth Referring For
Despite the cautions above, there are specific clinical scenarios where rTMS stroke recovery referral is justified by the evidence.
Central Post-Stroke Pain
The 2025 systematic review found rTMS reduces central post-stroke pain compared to sham (Liu et al., 2025). Central post-stroke pain is notoriously refractory to conventional analgesia. For patients with documented central pain who have failed pharmacological management, referral to an rTMS-capable centre is justifiable and consistent with the current evidence base.
Aphasia in Selected Patients
Emerging data on rTMS for post-stroke aphasia is increasingly promising. Low-frequency rTMS over Broca’s area homologue in the right hemisphere has shown benefit in several controlled trials. This is not yet at the level of clinical guideline recommendation, but for selected patients with non-fluent aphasia who have plateaued on speech therapy, referral to a specialist rTMS programme is reasonable.
Subacute Upper Limb Motor Recovery With DTI Guidance
For centres with DTI capability, patients in the subacute phase with preserved corticospinal tract integrity who have not yet responded to intensive conventional therapy are the strongest motor recovery candidates. At these centres, rTMS as an adjunct to structured task-specific rehabilitation is supported by the clinical trial data (Kim et al., 2020).
For these patients, rTMS sits alongside other neuromodulatory options that the HCAH clinical team considers, including vagus nerve stimulation for upper limb recovery in chronic stroke.
Frequently Asked Questions
What is rTMS for stroke and how does it work?
Repetitive transcranial magnetic stimulation delivers repeated magnetic pulses to specific brain regions to modulate cortical excitability. In stroke, low-frequency rTMS suppresses the overactive undamaged hemisphere, while high-frequency rTMS stimulates the damaged hemisphere directly. Both approaches aim to restore the interhemispheric balance that stroke disrupts.
Does rTMS help all stroke patients?
No. The evidence supports rTMS as beneficial for specific patient profiles, particularly those with moderate impairment, preserved corticospinal tract integrity, subacute stroke timing, and specific lesion patterns. Using rTMS without these selection criteria produces inconsistent results and wastes clinical resources.
What is central post-stroke pain and can rTMS help?
Central post-stroke pain is a neuropathic pain syndrome caused by stroke affecting the sensory pathways in the brain or thalamus. It is often refractory to standard analgesia. A 2025 systematic review found that rTMS reduces central post-stroke pain compared to sham stimulation, making it one of the better-supported specific indications for rTMS in stroke.
How many rTMS sessions does a stroke patient typically need?
Most clinical trials have used 10 to 20 sessions over 2 to 4 weeks. There is no universally agreed protocol. The practical approach is to define a response criterion at baseline, deliver a structured 4-week course, and reassess. If pre-specified targets have not been approached, the trial should stop.
What is the BDNF Val66Met genotype and why does it affect rTMS response?
BDNF (brain-derived neurotrophic factor) supports neuronal plasticity. The Val66Met polymorphism affects how much BDNF is released during activity-dependent plasticity. Met allele carriers have reduced BDNF secretion and show blunted long-term potentiation-like responses to rTMS. This makes them less likely to respond to rTMS-based rehabilitation. Genetic testing for this polymorphism is not routine, but it explains much of the outcome variation seen across rTMS trials.
Can rTMS be used for post-stroke depression?
Despite its established role in primary depression (rTMS is FDA-approved for treatment-resistant depression), the 2025 systematic review found no significant effect of rTMS on depression or anxiety in stroke patients. The mechanisms may differ. Post-stroke depression involves structural lesion effects, vascular factors, and adjustment reactions that may not respond to the same protocols used in primary depression. Antidepressants and psychological therapies remain first-line for post-stroke depression.
Is rTMS safe for stroke patients?
rTMS is generally well-tolerated in stroke patients. The main risks are headache, scalp discomfort, and a small risk of seizure (estimated at less than 1 in 1,000 sessions with standard protocols). Patients with metal implants in or near the skull, cochlear implants, or certain cardiac devices are contraindicated. Careful pre-screening and standard safety protocols manage these risks effectively.
How does rTMS compare to vagus nerve stimulation and BCI-FES for stroke?
All three are neuromodulatory approaches that enhance cortical plasticity through different mechanisms. VNS is surgically implanted and paired precisely with movement, producing strong localised plasticity signals. BCI-FES uses the patient’s own motor intention as the timing signal. rTMS modulates baseline excitability at the population level without requiring real-time pairing with movement. rTMS is the least invasive and most accessible but also the least targeted. The right tool depends on patient profile, available centre expertise, and what conventional therapy has already achieved.
Conclusion
rTMS stroke recovery is a legitimate clinical tool that has been rendered less useful than it could be by indiscriminate prescribing. The evidence supports it for specific patients, specific indications (central post-stroke pain, selected aphasia), and specific phases. Applied with selection criteria and structured reassessment, it is worth having in the clinic. Applied as a blanket add-on, it dilutes resources and erodes patient trust when results disappoint.
The honest 2026 position is not to abandon rTMS. It is to stop adding it to everything, and to start asking whether this specific patient, at this specific timepoint, with this specific lesion pattern, is actually the patient the evidence was about. Usually, the answer to that question will sharpen both the decision and the outcome.
Medical Disclaimer
This article is for educational and clinical reference purposes and does not replace personalised medical advice. rTMS protocols and patient selection criteria should be determined by a qualified neurologist or rehabilitation physician with experience in non-invasive brain stimulation. Individual response depends on lesion location, corticospinal tract integrity, stroke timing, and other clinical factors.
References
- Liu L, Zhou J, Wang Y, Li J, Liu J. Repetitive transcranial magnetic stimulation for central post-stroke pain: a systematic review. Frontiers in Neuroscience. 2025;19:1487391. ⚠️ Doctor to verify exact citation before publish.
- Kim WS, Kwon BS, Seo HG, Oh BM, Han TR. Low-frequency repetitive transcranial magnetic stimulation over contralesional motor cortex for motor recovery in subacute ischemic stroke: a randomised sham-controlled trial. Neurorehabilitation and Neural Repair. 2020;34(10):856 to 867.
- Chen QM, Yao FR, Sun HW, Gao ZK, Qi J, Xu DS. Combining inhibitory and facilitatory repetitive transcranial magnetic stimulation treatment improves motor function by modulating GABA in acute ischemic stroke patients. Restorative Neurology and Neuroscience. 2021;39(6):419 to 434.
- Mann SK, Malhi NK. Repetitive transcranial magnetic stimulation. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2025.
- Ahmed N, Wang B, Zhou Y, Khalid S. Clinical efficacy of NIBS in enhancing neuroplasticity for stroke recovery. Journal of Neuroscience Methods. 2025;379:110182. ⚠️ Doctor to verify exact citation before publish.
