Bimodal neuromodulation pairs sound with mild electrical stimulation of the tongue; its best-known device, Lenire, has US marketing authorisation through the FDA De Novo pathway, but none of its trials has compared it with a sham device. It is a new approach to tinnitus, distinct from sound therapy, TRT, or CBT, and Lenire was the first of its kind to be granted that authorisation, in March 2023, on the strength of clinical trial data rather than being marketed as a general wellness or masking product.
What it does
The best-known device, Lenire (made by Neuromod Devices), pairs two forms of stimulation delivered at the same time: audio tones through headphones, and mild electrical pulses delivered to the surface of the tongue through a small device. The underlying idea, supported by earlier animal research, is that combining sound with a second sensory input suppresses tinnitus-related neural activity in the brain's auditory pathways more effectively than sound alone — targeting the same central auditory hyperactivity thought to drive tinnitus in the first place, but through two simultaneous channels rather than one.
A typical course involves two 30-minute sessions a day, self-administered at home, over roughly 12 weeks, under the guidance of an audiologist trained in the device.
The specific brain mechanism behind the pairing
The theory has a fairly precise anatomical target: the dorsal cochlear nucleus (DCN), a brainstem structure where auditory signals from the ear and somatosensory signals (touch, in this case from the tongue) physically converge on the same neurons. Foundational animal research found that pairing sound with somatosensory stimulation at the DCN produces spike-timing-dependent plasticity — a lasting change in how strongly those neurons respond to sound afterward, with the direction of the change (strengthened or weakened response) depending on the precise timing between the two signals. This is the specific mechanism bimodal devices are designed to exploit: using carefully timed tongue stimulation alongside sound to drive the DCN's overactive tinnitus-related firing back down, rather than simply distracting from or masking the tinnitus the way conventional sound therapy does.
The evidence behind it
This is one of the more heavily studied tinnitus devices, and the design of each study matters more than its headline number:
- TENT-A1 (2020) enrolled 326 adults in a randomised, double-blinded study that allocated them between three parallel arms with different stimulation settings. All three arms received active bimodal treatment; there was no sham arm and no sound-only arm. Symptom scores fell significantly by the end of the 12-week treatment, and for specific settings the improvement continued for 12 months afterwards.
- TENT-A2 (2022) randomised 194 participants into four arms. Its two main arms compared different bimodal settings; an exploratory arm had sound only for the first six weeks and bimodal treatment for the second. Again there was no sham arm.
- TENT-A3 (2024), the pivotal trial behind the FDA authorisation, was a single-arm, within-subject pivotal trial (all 112 participants had sound-only first, then bimodal), run across three sites. Its paper states that because of this design no randomisation or masking was applicable. Among the 44 participants who still had moderate or worse tinnitus — a THI score of 38 or more — at the end of the sound-only weeks, 58.6% cleared the 7-point responder threshold during the bimodal weeks, against 43.2% during the sound-only weeks. In the full cohort that advantage was not observed: 63.3% responded during the sound-only weeks and 43.3% during the bimodal weeks, the second figure counting improvement on top of what sound had already achieved.
- On that trial, Lenire was granted marketing authorisation via the FDA De Novo pathway (DEN210033, March 2023), the route for a new kind of device with no legally marketed equivalent. It is not a premarket approval and not a 510(k) clearance.
- A retrospective chart review from a clinic in Alaska reported a 91.5% responder rate among 212 patients with moderate or worse tinnitus after about 12 weeks of treatment. It is an uncontrolled, retrospective chart review, with no comparison group and no sound-only stage, so it measures improvement from the start of treatment rather than what the tongue stimulation added. The closest matching figure from TENT-A3 is its responder rate over the full 12 weeks, also counted from the start of treatment: 79.4% of the full cohort, and 76.0% of the 44 still moderately bothered at the six-week visit. That is a before-and-after figure with no control too, and trial participants spent their first six weeks on sound alone. The chart review's authors write that longer use of the device than prescribed cannot be ruled out as a contributor to its higher rate, and every patient in it received tinnitus counselling and education before choosing a treatment plan. It does not confirm the trial's comparison, which it could not do, because it had no sound-only stage to compare against.
- A chart review from a clinic in New York, of 140 patients, reported 81.8% responders among those with moderate or worse tinnitus, also with no comparison group.
- No device-related serious adverse events were reported in TENT-A3 or in the Alaska chart review.
What these trials cannot show
TENT-A3's design is a reasonable one for a device that cannot be blinded, and it is the design the FDA recommended. Its structural limit is that bimodal treatment always came second. The fixed order cannot separate the tongue stimulation from time, expectation or order effects, so a better bimodal stage is not, on its own, evidence against a placebo response. The trial's authors write that a sham control was not possible, because both the sound and the tongue stimulation are noticeable. No sham-controlled trial of Lenire exists, and the two randomised trials compared bimodal settings with each other, not with an inactive device.
Who ran the studies
Neuromod Devices, which makes Lenire, is behind most of this evidence, and the papers say so. TENT-A2 was sponsored by the company, and its authors declare that seven of them are employees, consultants or shareholders of Neuromod Devices and three sit on its paid clinical advisory board. Eight of TENT-A1's thirteen authors list the company as their affiliation. TENT-A3 was funded by Neuromod Devices, which the paper names as the study sponsor, and declares one author, H. H. Lim, as a consultant with financial interests in the company. The Alaska chart review declares that its first author is a commercial provider of Lenire treatment and that Lim is a consultant with financial interests in Neuromod Devices; Lim is also an author of the New York review, which lists the company among Lim's affiliations. The 2026 meta-analysis below declares no conflicts of interest.
None of this makes the results wrong. It does mean every Lenire study on this page was either funded by the manufacturer or has an author with a declared tie to it.
How it compares against the other neuromodulation approaches
Most tinnitus evidence arrives one treatment at a time — this one against sham, that one against a waitlist. A 2026 systematic review and meta-analysis in The Laryngoscope did something more useful: it put every neuromodulation approach for chronic tinnitus into one analysis. 26 randomised trials, 1,576 participants, published between 2015 and 2025.
What it found, modality by modality:
- Transcranial electrical stimulation (11 trials, 372 people) — pooled SMD −0.36 (95% CI −0.75 to 0.02). Crosses zero. See tDCS.
- Repetitive TMS (8 trials, 432 people) — −0.15 (−0.37 to 0.07), with I² = 0%. See TMS.
- Acoustic coordinated reset (1 trial, 100 people) — no advantage over broadband noise.
- Vagus nerve stimulation (2 trials, 90 people) — modest benefit, with safety concerns confined to implanted approaches. See vagus nerve stimulation.
- Bimodal stimulation (4 trials, 582 people) — consistent, clinically meaningful reductions, often 10 to 20 points on the THI or TFI, with durability up to 12 months.
The authors' conclusion names it directly: among the modalities reviewed, bimodal stimulation shows the most consistent and durable clinical improvements. The review admitted only randomised trials.
Two things are worth holding alongside that. The bimodal arm also had the largest participant base of any modality in the review — 582 of the 1,576 — which is partly why its estimate is steadier than the others. And the same conclusion asks for standardised, adequately powered trials with harmonised protocols and long-term follow-up, which is not the language of a settled question.
Adverse events across the non-invasive modalities were mild and transient.
Who it's for, and how to access it
Bimodal neuromodulation is generally positioned for people with moderate-to-severe, bothersome tinnitus who haven't found sufficient relief from more conservative approaches. It requires an audiologist trained in the device — it isn't a consumer product you buy and self-administer without guidance — and is available through specialized tinnitus clinics in the US, with the US Department of Veterans Affairs specifically noted as one access pathway.
Cost and practical considerations
As a newer device, coverage varies by insurance plan, and it isn't yet a guaranteed covered benefit everywhere — worth confirming directly with a clinic offering it before committing, along with the total cost of the 12-week protocol including audiologist supervision.
Where this fits among tinnitus treatments
Given a completed pivotal trial and a US marketing authorisation, bimodal neuromodulation is one of the better-studied newer options for tinnitus that hasn't responded well to sound therapy or hearing aids alone — worth discussing directly with an audiologist, particularly for tinnitus significant enough to be scored as moderate or worse on standard clinical measures. Two caveats belong in that conversation: none of its trials has a sham arm, and every Lenire study so far involves the manufacturer. Tinnitus devices compared sets Lenire beside Neosensory Duo, Levo, Neuromonics and Desyncra, with what each was tested against.