Vagus Nerve Stimulation for Long COVID: What the Evidence Actually Shows
Introduction: A Plausible Idea That Needs Honest Scrutiny
Long COVID — the persistence of symptoms such as fatigue, "brain fog", breathlessness, and autonomic disturbance for months after a SARS-CoV-2 infection — has left millions of people searching for something that helps. Vagus nerve stimulation (VNS), and particularly the non-invasive ear-based form known as transcutaneous auricular VNS (taVNS), has been promoted in some quarters as a candidate. The biological story behind that idea is genuinely interesting. The clinical evidence, however, is early, small, and — where it is most rigorous — largely negative.
This article sets out what the research actually shows, rather than what device marketing or hopeful headlines might imply. The short version: the best-designed trials so far have not found that vagus nerve stimulation relieves long COVID symptoms, and in the most rigorous study the placebo group did at least as well as the active one. That conclusion may change as larger trials report, but it is where the evidence stands today.
Why Researchers Proposed VNS for Long COVID
The rationale rests on two overlapping observations about long COVID, both of which intersect with the known biology of the vagus nerve.
Persistent inflammation. The vagus nerve is the main pathway of the "inflammatory reflex" — a circuit through which the nervous system senses and restrains the production of inflammatory cytokines. In this model, first characterised by Tracey (2002), afferent (incoming) vagal fibres detect inflammatory signals from the body, while efferent (outgoing) fibres act through a downstream route known as the cholinergic anti-inflammatory pathway. That pathway operates in part via the spleen, where acetylcholine signalling on immune cells suppresses the release of tumour necrosis factor and other pro-inflammatory cytokines. This circuitry underpins much of the interest in VNS as an anti-inflammatory intervention (see our overview of VNS and inflammation). Because some long COVID symptoms have been linked to low-grade, ongoing immune activation, the hope was that engaging it might calm that activation down.
Autonomic dysfunction. Autonomic symptoms are among the most consistently reported features of long COVID. In a global survey of 2,314 affected adults, roughly two-thirds had questionnaire scores indicating moderate-to-severe autonomic dysfunction, and postural orthostatic tachycardia syndrome (POTS) was the most common associated diagnosis (Larsen et al., 2022). These problems — palpitations, light-headedness on standing, a racing heart rate — overlap with reduced parasympathetic (vagal) tone, which is often indexed by heart rate variability (HRV). Since taVNS is intended to engage vagal pathways, it was proposed as a way to "rebalance" autonomic function.
There is, however, a weak link in this chain of reasoning that is easy to overlook: it assumes taVNS reliably increases vagal tone in the first place — and even that step is not well established. A 2026 systematic review of neuromodulation and heart rate variability found that transcutaneous VNS produced no consistent, significant effect on HRV across randomised trials (Souza et al., 2026). So the intermediate premise — "taVNS raises vagal tone" — is itself unproven. A rationale is only as strong as its weakest link, and here more than one link remains unverified.
Khan et al. (2024), in a narrative review, articulate the overall reasoning while being explicit that the supporting data are "limited in scope and robustness". That is the key point to hold onto: biological plausibility is a reason to run a trial, not evidence that a treatment works.
What the Controlled Trials Show
The most informative studies are randomised, sham-controlled trials, because long COVID symptoms — especially fatigue — tend to improve over time on their own, and respond strongly to placebo. Without a sham comparison, it is impossible to tell whether any improvement was caused by the stimulation. Two such trials have now reported, and neither supports an effect.
Percin et al. (2025) randomised 42 people with post-COVID syndrome to active taVNS or sham stimulation, delivered for 30 minutes a day over 20 days. Fatigue improved in both groups — but the authors' own conclusion was that sham stimulation was superior to active taVNS for reducing fatigue severity. In other words, the placebo arm did better than the treatment arm. A telling detail sits inside the data: active taVNS did produce measurable increases in HRV, consistent with genuine engagement of cardiac vagal pathways — yet that physiological effect did not translate into greater symptom relief. Stimulating the nerve, in this trial, was not the same as helping the patient. It was a single-blind study with a small sample, so it should not be over-interpreted, but it certainly does not show benefit.
The PAuSing-pCF trial (Shahmandi et al., 2026) is larger and more rigorous: 114 adults with post-COVID fatigue were randomised across three arms — active stimulation, sham stimulation, and a separate placebo condition. That three-arm design matters, because it can distinguish a genuine treatment effect both from an inert placebo and from the non-specific effects of wearing and using the sham device itself. The primary fatigue outcome at eight weeks showed no significant difference between groups (active versus sham mean difference 2.76, 95% CI −5 to 11; p = 0.50). Fatigue fell modestly in every arm — the pattern you would expect from natural recovery and placebo response, not from an active treatment effect.
Taken together, the two best sham-controlled trials of taVNS for long COVID fatigue are a negative result and a null result. This is the heart of the matter, and it deserves a plain statement:
On the strength of the controlled evidence available today, vagus nerve stimulation has not been shown to treat long COVID. The trials designed specifically to detect a real effect did not find one.
The Pilot Studies — and Why They Read More Positively
Set against those controlled trials are smaller, uncontrolled reports that sound more encouraging — and it is worth understanding exactly why they differ.
The most cited is Zheng et al. (2024), a single-arm pilot in 24 women who used home-based taVNS for 10 days. The authors reported improvements in cognition, anxiety, depression, and sleep, with fatigue improving only at one-month follow-up and no change in sense of smell. These are real, carefully measured observations — but the study had no control or sham group. With participants on average around 20 months past their infection, there is no way to separate any benefit of stimulation from continued natural recovery, expectation effects, or the simple act of paying structured attention to one's health.
This is not a criticism of pilot studies as a category — they are how research begins, and they help justify the larger trials that follow. The problem arises only when uncontrolled results are presented as if they were proof. When the controlled trials are run, as they now have been, the apparent benefit does not survive.
Why Long COVID Is Especially Prone to False Positives
Three features of long COVID make rigorous controls unusually important here:
- Natural recovery. Many people improve gradually over months regardless of treatment, so "before and after" comparisons routinely overstate benefit.
- Strong placebo response. Fatigue, brain fog, and malaise are subjective and fluctuate, and they respond markedly to placebo — which is precisely why the sham arm matched or beat the active arm in the trials above.
- Outcome variability. Symptoms wax and wane day to day, so a short study can easily catch an upswing that has nothing to do with the intervention.
These are the very conditions under which uncontrolled studies mislead, and under which only sham-controlled trials can be trusted.
How Strong Is the Evidence Base Overall?
A 2026 systematic review by Balan et al. appraised the clinical evidence for taVNS in post-COVID-19 condition. It identified only a small number of studies (five, totalling around 154 participants), found that most carried a high risk of bias, and graded the certainty of the efficacy evidence as "very low" (GRADE) — the lowest rating on the scale. The single adequately double-blinded trial it included found no significant between-group differences on any outcome.
"Very low certainty" means future, better-designed studies could easily overturn current estimates in either direction. It is not a foundation on which to recommend a treatment.
What the Trials Tested — and What They Didn't
There is an important limitation built into the structure of this evidence: nearly every controlled trial to date has used fatigue as its primary outcome. That is a defensible choice — fatigue is long COVID's most common and disabling symptom — but it means the negative results speak most directly to fatigue, and far less to the condition's other forms.
Long COVID is not a single entity. Beyond fatigue, large cohorts describe a prominent autonomic phenotype — POTS and orthostatic intolerance affecting a substantial share of patients (Larsen et al., 2022) — alongside a cognitive phenotype dominated by brain fog and memory complaints. These are precisely the domains where a vagal intervention might, in principle, behave differently from how it behaves against fatigue. Yet they have barely been examined in adequately controlled trials. The uncontrolled Zheng pilot reported cognitive and mood changes, but, as discussed, could not separate them from placebo and natural recovery.
The honest position is therefore narrower and more specific than a blanket "VNS does not work for long COVID". It is this: for post-COVID fatigue, the controlled evidence is negative; for the autonomic and cognitive presentations, rigorous evidence barely exists in either direction. And absence of good evidence is not evidence of benefit — it is a reason for properly designed, phenotype-specific trials before anyone draws firm conclusions.
A Note on Safety Versus Effectiveness
None of this means taVNS is dangerous. Non-invasive vagus nerve stimulation is generally well tolerated, with side effects that are usually mild and transient — as discussed in our review of the safety profile of VNS. The Percin and PAuSing-pCF trials reported it was acceptable to participants.
But tolerability is not effectiveness. A treatment can be perfectly safe and still do nothing for the condition in question. The fact that taVNS is low-risk is a reason it remains worth studying — not a reason to assume it works.
The Bottom Line
For long COVID, vagus nerve stimulation is a reasonable scientific hypothesis with a coherent mechanism behind it — but it is not, on current evidence, an established or proven treatment. The most rigorous trials to date are null or negative, the overall evidence is of very low certainty, and the encouraging-sounding results come from studies that could not rule out placebo and natural recovery.
If you are living with long COVID and considering taVNS, the honest guidance is:
- Treat marketing claims of benefit with scepticism. They are not supported by the controlled evidence.
- Discuss any device use with a qualified clinician, particularly if you have cardiac or autonomic conditions.
- Watch this space rather than bank on it. Larger and better-designed trials are underway, and the picture may change. We will update this article as stronger evidence emerges.
You can explore the underlying studies in our Evidence Database, where each trial is listed with its design, sample size, and findings.
---
References
Balan, A. et al. (2026). Transcutaneous auricular vagus nerve stimulation for post-COVID-19 condition: a systematic review and critical appraisal of clinical evidence. Journal of Clinical Medicine, 15(11), 4247.
Khan, M.W.Z. et al. (2024). Vagal nerve stimulation for the management of long COVID symptoms. Infection Medicine, 3(4), 100149.
Larsen, N.W. et al. (2022). Characterization of autonomic symptom burden in long COVID: a global survey of 2,314 adults. Frontiers in Neurology, 13, 1012668.
Percin, A. et al. (2025). Effects of transcutaneous auricular vagus nerve stimulation on fatigue in post-COVID syndrome: a randomized, single-blind, sham-controlled study. International Journal of Clinical Practice, 2025, 5641307.
Shahmandi, M. et al. (2026). Effect of percutaneous auricular nerve stimulation on fatigue in adults with post-COVID fatigue (PAuSing-pCF): results of a randomised, sham-controlled trial. NIHR Open Research, 6, 35.
Souza, R. et al. (2026). Neuromodulation of heart rate variability: a systematic review. Autonomic Neuroscience, 263, 103379.
Tracey, K.J. (2002). The inflammatory reflex. Nature, 420(6917), 853–859.
Zheng, Z.S. et al. (2024). Transcutaneous vagus nerve stimulation improves Long COVID symptoms in a female cohort: a pilot study. Frontiers in Neurology, 15, 1393371.