Lenire's pivotal trial found an advantage over sound alone only in people still moderately bothered after six weeks of sound therapy, not in its full cohort of 112, and no sham-controlled trial of the device exists.
Lenire, made by Neuromod Devices, is the first device of its kind to be granted marketing authorisation in the US, through the FDA De Novo pathway (DEN210033, March 2023). Bimodal neuromodulation devices covers the wider category. This page looks at what the pivotal trial found, how it was designed, who it included, what it did not measure, and what getting the treatment involves, separately from the marketing built on top of it.
What the device does
Lenire pairs sound delivered through headphones with mild electrical stimulation delivered to the tongue via a device held in the mouth, timed together — the theory being that this combined, bimodal stimulation drives more meaningful neuroplastic change in the brain's auditory and attention pathways than sound alone. It is a prescription device: a clinician experienced in tinnitus prescribes it, and the patient then uses it at home. It is not an over-the-counter purchase.
What the FDA authorisation is, and is not
Lenire was granted marketing authorisation via the FDA De Novo pathway (DEN210033, March 2023). De Novo is the route for a new kind of device that has no legally marketed equivalent to be compared against. The FDA classified Lenire as a class II prescription device, subject to special controls, to temporarily relieve the symptoms of tinnitus in adults with at least moderate tinnitus as measured by the Tinnitus Handicap Inventory.
That is neither a premarket approval, the route for class III devices, nor a 510(k) clearance, which rests on equivalence to a device already on the market. So in the FDA's own sense of the words, Lenire is neither approved nor cleared — and this page used both words until it was corrected.
How the pivotal trial was designed
The TENT-A3 trial, which supported the De Novo authorisation, enrolled 112 participants across three sites in three countries. It was a single-arm, within-subject pivotal trial: all 112 participants had sound-only stimulation first, then bimodal — six weeks of each, always in that order, with each person's second six weeks compared against their first. There were no separate groups. The paper is direct about it: because of the single-arm, repeated-measures design, no randomisation or masking was applicable. The design was the one the FDA recommended.
A responder was defined as someone improving at least 7 points on the Tinnitus Handicap Inventory within a stage. So a responder in the bimodal stage had to gain another 7 points on top of whatever the sound-only weeks had already achieved.
Participants were asked to use the device for up to an hour a day, in two 30-minute sessions, for the 12 weeks. The trial counted at least 18 hours of use in each six-week stage as compliant, and 92.0% met that in the sound-only stage and 82.4% in the bimodal stage.
What it found
- In the full cohort of 112, the paper reports that the advantage for bimodal treatment was not observed. 63.3% responded during the sound-only weeks and 43.3% during the bimodal weeks that followed.
- In the 44 participants still scoring 38 or more on the THI at the six-week interim visit — the moderate-or-worse group — 58.6% responded during the bimodal weeks against 43.2% during the sound-only weeks (p=0.022). The Tinnitus Functional Index, a secondary measure with a 13-point threshold, pointed the same way: 45.5% against 29.6% (p=0.010).
The favourable result is in a subgroup defined at the end of the sound-only stage. Everyone enrolled scored 38 or more at screening, so the 44 are the people still at that level after six weeks of sound alone; participants who had dropped below it were not part of that comparison. Even within the subgroup, over 40% did not reach the responder threshold on bimodal treatment.
What this design cannot tell you
Because everyone had the same order, the bimodal weeks were always the later weeks. Anything that shifts tinnitus scores over time — natural fluctuation, the expectation that the second, "full" stage is the real treatment, simply having been in a study for longer — falls on the bimodal side of the comparison. The fixed order cannot separate the tongue stimulation from time, expectation or order effects. The authors' answer is that the bar was set high, since bimodal had to add a further 7 points beyond what sound had already done. That is a real hurdle, but it is not a control.
No sham-controlled trial of Lenire exists. The TENT-A3 authors write that a sham control was not possible, because both the sound and the tongue stimulation are noticeable, so participants would know if they had been given a sham. The two earlier trials did not have one either. TENT-A1 randomised 326 adults between three parallel arms with different stimulation settings, all of them active bimodal treatment. TENT-A2 randomised 194 participants into four arms: two main arms comparing bimodal settings, plus exploratory arms that included one given sound alone for six weeks before bimodal. Both were double-blinded, and neither compared the device against an inactive version of itself.
Who the trial included, and who it left out
A trial's result applies most directly to people who resemble the people in it, and TENT-A3's entry criteria were narrow.
Who was in it. Adults aged 18 or over with subjective tinnitus lasting between three months and ten years, scoring at least 38 on the Tinnitus Handicap Inventory, and with no more than a set amount of hearing loss — at most 40 dB HL between 250 Hz and 1 kHz, and at most 80 dB HL between 2 and 8 kHz. They were recruited partly through an online forum and radio advertisements. Of 2,877 people who filled in the online pre-screening form, 222 were screened in clinic and 112 enrolled.
Who was excluded. People with:
- objective or pulsatile tinnitus, or an ear abnormality on examination that might be causing the tinnitus
- Ménière's disease, a temporomandibular joint disorder, or burning mouth syndrome
- a previous diagnosis of psychosis or schizophrenia, a neurological condition that may lead to seizures or loss of consciousness, severe cognitive impairment, or very high anxiety scores
- any type of electro-active implanted device
- a head or neck injury needing hospital or a doctor's visit in the previous 12 months
- a hearing aid started in the previous 90 days, or prescription medications or treatments started or stopped in the previous three months that might affect the results
- pregnancy, or oral piercings that could not be removed
- previous use of Lenire, a previous tinnitus trial, or an experimental or surgical tinnitus treatment
None of that means Lenire is known to be unsafe for those groups. It means the trial did not test it on them, so its figures do not tell you what to expect if you are in one. Several of the exclusions — pulsatile tinnitus above all — are also reasons to have the tinnitus medically assessed before trying any device.
What the trials did not measure
What happens after treatment stops. The pivotal trial ended at the 12-week visit. Its authors list as a limitation that long-term benefits were not assessed after treatment ended, because of time constraints the FDA set for the data behind the De Novo submission. They point to TENT-A1 and TENT-A2, which followed participants for 12 months after treatment — trials that, as above, had no inactive comparison either.
Loudness. The outcomes the paper reports are the Tinnitus Handicap Inventory, the Tinnitus Functional Index, a general quality-of-life index, two satisfaction questions and safety. It reports no measure of how loud the tinnitus was. The results describe how much the tinnitus interfered with people's lives, not whether the sound changed.
Quality of life, in the end. The quality-of-life index, the Health Utilities Index Mark 3, barely moved between screening and the final visit. The authors explain why: participants were generally healthy apart from their tinnitus, and 78 of the 80 who completed it already had the best possible score on its hearing item at the start, so it could not detect much. They had included it partly because such data matter when seeking reimbursement for a treatment. The honest reading is that the trial produced no usable evidence on quality of life either way.
Other treatments. None of the three trials compared Lenire with CBT, hearing aids or any other established tinnitus treatment. The only comparison with anything but another bimodal setting was with sound stimulation delivered by the same device.
The safety and satisfaction data
No device-related serious adverse events were reported in the trial. Of the 62 device-related adverse events, 96.8% were classified mild, and every adverse event resolved during the study except one mild case that was lost to follow-up. At the final visit, 62.9% of the 105 participants who attended said they had benefited from the bimodal stage, and 88.6% said they would recommend the treatment to someone with tinnitus. Those are self-reports from people who had just finished twelve weeks of treatment, with no comparison group asked the same questions.
Outside the trial
A retrospective chart review from one US clinic reported a 91.5% responder rate — an improvement of more than 7 THI points from the start of treatment — among 212 patients with moderate or worse tinnitus after about 12 weeks. It is an uncontrolled, retrospective chart review: no comparison group and no sound-only stage, so it measures how much people improved from the start of treatment, not what the tongue stimulation added.
The pivotal trial reports the closest matching figure. Everyone it enrolled had scored 38 or more at screening, the same threshold the chart review used. Over the full 12 weeks, counted from the start of treatment, 79.4% of the full cohort cleared the same 7-point bar (95% confidence interval 71.6% to 87.2%, against 86.9% to 94.5% for the chart review), as did 76.0% of the 44 still moderately bothered at the six-week visit. That, too, is a before-and-after figure with no control. The two were not reached the same way: trial participants spent their first six weeks on sound alone, while every patient in the chart review used the same bimodal setting.
The chart review's authors write that its patients may have used the device for longer than prescribed, and that this cannot be ruled out as a contributor to a responder rate higher than the trials'. Every patient in it also received tinnitus counselling and education before choosing a treatment plan. What it cannot do is confirm the trial's comparison: with no sound-only stage, it has nothing to set the tongue stimulation against.
A second US chart review, published in 2026, looked back at 140 patients treated at a New York clinic between May 2023 and January 2024. Among those with moderate or worse tinnitus, 81.8% improved by at least 7 THI points after 12 weeks (95% CI 70.9% to 89.3%), with an average fall of 23.8 points; 71.2% cleared a stricter 11-point bar. Patients with slight or mild tinnitus showed nearly zero change on average. That split fits the authorisation being limited to at least moderate tinnitus, but it has a second explanation the design cannot rule out: people who start higher have more room to fall, and in an uncontrolled study some of that fall happens without any treatment at all, a pattern set out in what happens if you do nothing.
Cost and access
Price. Neuromod Devices does not publish one. Its FAQ, read in September 2026, says the company gives clinicians guidance about a recommended retail cost, but each clinician sets the price of the treatment programme, taking into account their own time and expertise as well as the device, and it directs readers to their local clinic. The manufacturer's pages read for this article say nothing about insurance coverage. No independent source reviewed here reports a price, so the figure you are quoted will come from the clinic.
Where it is offered. The manufacturer says Lenire is available at specialised clinics across the US, with a clinic-finder map and a waiting-list form where there is none nearby.
Veterans. On 17 June 2024 the manufacturer announced that its US company had been awarded a Federal Supply Schedule contract, which it says makes Lenire a treatment option for veterans through the Department of Veterans Affairs. Its veterans page, read in September 2026, says many VA clinics do not yet offer it and suggests asking at a local VA clinic. Whether it is prescribed still depends on a clinical assessment.
Time. In the trial, the commitment was up to an hour a day for 12 weeks, plus clinic visits for fitting and review.
Who paid for the research
The pivotal trial was funded by Neuromod Devices, the company that makes Lenire, which the paper names as the study sponsor; three of its co-authors are from the contract research organisation the sponsor delegated the study to. The paper declares no competing interests except for one author, H. H. Lim, a consultant with financial interests in Neuromod Devices. TENT-A2's authors declare that seven of them are employees, consultants or shareholders of the company and three sit on its paid clinical advisory board, and eight of TENT-A1's thirteen authors list Neuromod Devices as their affiliation. The first US chart review declares that its first author is a commercial provider of Lenire treatment, and that Lim is a consultant with financial interests in the company. The 2026 chart review lists Lim among its four authors, with Neuromod Devices among his affiliations.
None of that makes the results wrong. It does mean that every study on this page was either funded by the manufacturer or has an author with a declared tie to it.
How it compares with other neuromodulation approaches
Lenire's evidence base is more extensive than that of most other neuromodulation devices for tinnitus, including newer bimodal entrants and single-mode approaches like TMS or vagus nerve stimulation: three trials of more than 100 people each, and a completed pivotal trial behind a US marketing authorisation rather than early pilot data alone. That doesn't make it a cure, and it doesn't make it sham-controlled: the comparison that would show whether the tongue stimulation itself does the work has not been run.
Before you commit
- Check you match the people it was tested on. The authorisation is for adults with at least moderate tinnitus on the THI; in the trial, the advantage over sound alone appeared only in people still at that level after six weeks of sound. Ask the clinic for your score before and after.
- Have the tinnitus medically assessed first if you fall into one of the excluded groups, above all if it is pulsatile or followed a head or neck injury.
- Ask for the full price in writing, including the device, the fitting and any follow-up visits, since the manufacturer leaves pricing to each clinic.
- Weigh it against the alternatives. CBT and, where there is hearing loss, hearing aids have guideline support.
- Expect some chance it will not help. Over 40% of the favourable subgroup did not reach the responder threshold on bimodal treatment.