tVNS for migraine

This page is intended for UK healthcare professionals. Clinical information, including dosing parameters, effect sizes, and trial-level data, is included.

The clinical context

Migraine is the second leading cause of years lived with disability worldwide and the leading cause among women aged 15 to 49. In the UK, roughly 10 million adults live with migraine, and approximately 1 million have chronic migraine (15 or more headache days per month for at least 3 months, with at least 8 meeting migraine criteria).

Current first-line preventives provide only partial benefit. Topiramate achieves a 50% responder rate of 35 to 45% but carries significant cognitive, mood, and teratogenic side effects. CGRP monoclonal antibodies achieve 50% responder rates of 27 to 61% and are available through the NHS only after failure of at least three prior preventive treatments. OnabotulinumtoxinA achieves approximately 45% responder rate at 24 weeks but requires specialist clinic visits every 12 weeks.

Medication overuse headache (MOH) affects 50 to 80% of chronic migraine patients presenting to specialist headache clinics. Any treatment free from drug interactions that does not contribute to medication overuse has intrinsic value.

tVNS offers home-based, self-administered neuromodulation with no drug interactions, a predominantly local side-effect profile, no MOH risk, and substantially lower cost than CGRP monoclonal antibodies or onabotulinumtoxinA.

Neurobiological mechanisms

Trigeminovagal connection

The trigeminovascular system is the primary neural substrate of migraine. The vagus nerve and trigeminal nerve converge in the brainstem: the NTS has direct and indirect connections with the trigeminal nucleus caudalis and the broader trigeminocervical complex (TCC). Animal studies demonstrate dose-dependent inhibition of dural-vascular trigeminal neurons by VNS, with suppression of glutamate release and microglial activation in the TCC (Oshinsky et al., 2014; Akerman et al., 2017).

Cortical spreading depression (CSD) inhibition

CSD is the slowly propagating depolarisation wave underlying migraine aura and thought to trigger the trigeminovascular cascade. Chen et al. (2016) demonstrated that both invasive VNS and transcutaneous VNS inhibit CSD frequency and elevate electrical thresholds to the same degree. Chen et al. (2020) established that this inhibition occurs exclusively through central mechanisms via the NTS. Liu et al. (2024) identified a glutamate-dependent TrkB activation mechanism. CSD inhibition is intensity-dependent.

This CSD-inhibiting property is particularly relevant for migraine with aura, and may explain the clinical finding that these patients respond preferentially.

CGRP pathway

VNS reduces CGRP expression in trigeminal ganglia in animal models. Chen et al. (2022) demonstrated attenuation of CGRP expression triggered by CSD. This provides a mechanistic link to the same neuropeptide pathway targeted by erenumab, fremanezumab, galcanezumab, and the gepants.

Serotonergic and noradrenergic modulation

NTS projections to the dorsal raphe nuclei (serotonin) and locus coeruleus (norepinephrine) provide direct pathways for enhancing descending inhibition of trigeminal nociception. Zhang et al. (2019) demonstrated using fMRI that taVNS at 1 Hz modulates locus coeruleus activity and increases its connectivity with somatosensory cortex in migraine patients. This connectivity increase was significantly negatively correlated with migraine attack frequency.

Anti-inflammatory effects

VNS suppresses CSD-induced neuroinflammation, reducing cortical COX-2 expression and c-Fos expression in the trigeminal nucleus caudalis, directly relevant to the neuroinflammatory component of migraine pathophysiology.

A key parameter difference from epilepsy

An important finding: in migraine, 1 Hz stimulation may outperform 25 Hz (the opposite of epilepsy, where 25 Hz outperforms 1 Hz). The Straube 2015 trial is the basis for this observation. The frequency-dependent difference may reflect distinct pathophysiological targets: epilepsy benefits from higher-frequency modulation of cortical excitability via locus coeruleus norepinephrine pathways, while migraine may benefit from lower-frequency modulation of descending serotonergic inhibition and trigeminovascular processing.

This is based on a single trial. Dose-response replication is needed, but the finding has practical implications for protocol selection.

Clinical evidence

Auricular tVNS

Straube et al. (2015): key auricular trial

Randomised, double-blind, chronic migraine. Compared 25 Hz vs 1 Hz at the left cymba conchae, 4 hours daily, 3 months.

Per protocol (n=40):

1 Hz group: reduction of 7.0 +/- 4.6 headache days per 28 days
25 Hz group: significantly smaller reduction (p < 0.05)
Both groups showed significant HIT-6 and MIDAS improvements
Both reduced acute medication intake

The 7-day reduction exceeds what has been reported for other nerve stimulation procedures in chronic migraine. However, this trial compared two active frequencies rather than active versus sham, so the absolute treatment effect size cannot be determined. This is the most important caveat.

Migraine-specific fMRI studies

Zhang et al. (2019) and Chen et al. (2023) provided mechanistic validation demonstrating brainstem pathway engagement and functional connectivity changes correlating with clinical improvement.

Cervical tVNS (gammaCore, for context)

PRESTO trial (Tassorelli et al., 2018): acute treatment

248 patients with episodic migraine, 10 centres.

2-hour pain freedom: 30.4% vs 19.7% (p=0.067, not significant)
30-minute pain freedom: 12.7% vs 4.2% (p=0.012)
60-minute pain freedom: 21.0% vs 10.0% (p=0.023)

PREMIUM trial (Diener et al., 2019): episodic prevention

332 patients, double-blind, sham-controlled.

ITT primary: migraine days -2.26 vs -1.80 (p=0.15, not significant)
Modified ITT (>=67% adherence, n=278): migraine days -2.27 vs -1.53 (p=0.043); acute medication days -1.94 vs -1.14 (p=0.039)

Adherence-dependent result: patients who used the device consistently showed significantly better outcomes.

PREMIUM II trial (Najib et al., 2022): episodic prevention

Terminated early due to COVID-19. Of 336 enrolled, 113 formed the prespecified modified intention-to-treat population (active n=56, sham n=57) — roughly 60% below the planned sample, leaving the trial underpowered.

Monthly migraine days (primary endpoint): -3.12 (active) vs -2.29 (sham); difference -0.83; p=0.23 -- not statistically significant
50% responder rate: 44.87% vs 26.81% (p=0.05)
Migraine with aura subgroup: a prespecified subgroup analysis found that participants with aura responded preferentially (specific effect sizes are reported only in the full paper).

Meta-analyses

Zhuo et al. (2023, 6 RCTs, 845 patients), auricular and cervical analysed separately:

Low-frequency auricular nVNS: significant reduction in monthly headache days (MD -1.8, p=0.02) and headache intensity (SMD -0.7, p=0.009)
Cervical nVNS: not significant for headache day reduction (MD -0.46, p=0.23)
Overall 50% responder rate: OR 1.69 (p=0.009)

The meta-analytic finding that auricular tVNS significantly reduces headache days while cervical tVNS does not reach significance is noteworthy and suggests distinct advantages for the auricular approach.

IHS 2025 guidelines

Weak recommendation for gammaCore for preventive migraine treatment. No formal recommendation for auricular tVNS devices due to insufficient qualifying RCTs, though guidelines described non-invasive neuromodulation devices as "promising alternatives to drug treatment" that are "safe and generally well tolerated and devoid of drug interactions."

Stimulation parameters for migraine

Parameter	Straube chronic migraine	fMRI studies in migraine
Frequency	1 Hz (outperformed 25 Hz)	1 Hz
Pulse width	250 microseconds	200 microseconds
Intensity	Maximum tolerated, sub-pain	1.5 to 5 mA
Duty cycle	Continuous	Continuous
Daily duration	4 hours	30 min twice daily
Stimulation site	Left cymba conchae	Left cymba conchae

Predicting response

Migraine with aura: most consistently identified predictive factor. In PREMIUM II, a prespecified subgroup analysis found that migraine-with-aura patients responded preferentially. Mechanistically consistent with CSD inhibition.
Adherence: PREMIUM demonstrated significantly better outcomes in patients using the device at least 67% of the time. Consistent daily use is a key determinant.
fMRI-based prediction: early fALFF patterns can predict taVNS efficacy in migraine without aura (published 2022). Not yet clinically practical.

Safety

No serious device-related adverse events were reported in the published migraine trials cited here (auricular and cervical). Common mild and transient effects: tingling, itching, erythema at electrode site, headache, ear discomfort. The Instructions for Use list a fuller range. In the PREMIUM trial, dropout due to adverse events was actually lower in the active group (1.2%) than sham (5.2%).

A systematic review of the safety and tolerability of tVNS (Redgrave et al., 2018; 51 studies, 1,322 participants) concluded that tVNS is safe and well tolerated at the doses tested to date. The most common side effect was local skin irritation from the electrode (18.2%); 35 participants (2.6%) withdrew because of side effects; and of the serious adverse events recorded, only 3 (across 1,322 participants) were judged possibly related to tVNS, with symptomatic bradycardia in 1 (0.08%).

Practical guidance for UK clinicians

Candidate selection

Adults with chronic migraine (>=15 headache days/month) who have not responded adequately to at least one preventive medication
Patients with episodic migraine seeking non-pharmacological prevention, particularly those with migraine with aura
Patients at risk of or already experiencing medication overuse headache
Patients unable to access or tolerate CGRP-targeted therapies or onabotulinumtoxinA
Exclusions (per the IFU): pregnancy; active implants (cardiac pacemaker, implanted vagus nerve stimulator, cochlear implant); cerebral shunt; sore, broken, or diseased skin at the site. Cardiac arrhythmia warrants medical clearance before use (a caution, not an absolute exclusion).

Treatment initiation

Establish a baseline headache diary (minimum 4 weeks)
Based on the Straube protocol: 1 Hz, 250 microsecond pulse width, maximum tolerated intensity, 4 hours daily at the left cymba conchae
Alternative shorter protocol: 30 minutes twice daily at similar parameters
Assess response at 4, 8, and 12 weeks using headache diary (migraine days, headache days, acute medication days)
Allow at least 12 weeks before judging response

Patient counselling

Low-frequency auricular tVNS can significantly reduce headache days and intensity based on available evidence
If the patient has migraine with aura, there is a biologically coherent reason for a potentially stronger response
Benefit builds over time; allow 8 to 12 weeks of consistent use
The device does not contribute to medication overuse and can be used alongside any existing treatment
Building sessions into daily routine improves adherence, which is strongly associated with better outcomes

Regulatory status

The tVNS E device holds Class IIa certification under EU MDR 2017/745 with migraine as a specifically listed indication. There is no specific NICE guidance for tVNS (auricular or cervical) for migraine. NICE IPG652 covers transcutaneous stimulation of the cervical branch for cluster headache only. The IHS 2025 guidelines gave gammaCore a weak recommendation for preventive migraine but did not formally recommend auricular devices.

Key evidence gaps

No large sham-controlled RCT for auricular tVNS in chronic migraine.
Optimal auricular parameters for migraine not established by dose-response studies.
The 1 Hz vs 25 Hz finding needs replication.
No controlled data beyond 3 months for auricular tVNS in migraine.
No head-to-head trials comparing tVNS with CGRP monoclonal antibodies, onabotulinumtoxinA, or rTMS.
Chronic migraine evidence base is more limited than episodic for cervical nVNS.
Paediatric evidence pending (BMJ Open protocol published 2025).

This page is based on evidence available up to April 2026.

tVNS is a Class IIa medical device manufactured by tVNS Technologies GmbH, Germany. Distributed in the UK by Anatomical Concepts UK Ltd.