If tinnitus is a problem created by the brain rewiring itself, can we fix it by rewiring the brain back? That's the core question behind neuroplasticity-based sound therapy -- and after two decades of research, we're starting to get answers.
Why Your Brain Creates a Phantom Sound
Tinnitus isn't a problem with your ears. It's a problem with your brain's response to reduced auditory input.
When hearing damage occurs -- from noise exposure, aging, or other causes -- the auditory cortex loses input at certain frequencies. But the affected neurons don't go silent. Instead, three things happen:
- Central gain increase. The brain compensates for reduced input by turning up its internal amplifier. Schaette and Kempter (2006) built a computational model demonstrating that this homeostatic plasticity -- the brain's attempt to maintain stable neural activity levels -- inadvertently creates spontaneous hyperactivity. It's like cranking the volume on a noisy amplifier when the signal weakens: you hear more static.
- Tonotopic map reorganization. The auditory cortex is organized by frequency -- low tones at one end, high tones at the other. When neurons lose their normal input, neighboring neurons expand into the vacated territory. Pantev et al. (1998) measured this with magnetoencephalography (MEG) and found tinnitus patients had cortical map shifts of 5.3 mm vs. 2.5 mm in controls. Critically, the degree of reorganization correlated strongly with tinnitus severity (r = 0.82, P < 0.01).
- Reduced lateral inhibition. In the healthy auditory cortex, neurons suppress their neighbors to sharpen frequency tuning. When hearing loss creates a gap in input, this inhibitory network breaks down in the affected region, allowing synchronized spontaneous neural firing to emerge -- perceived as a phantom sound.
All three mechanisms are forms of neuroplasticity -- the brain's ability to reorganize itself. The important implication: if tinnitus is created by maladaptive plasticity, it may be reversible through targeted beneficial plasticity.
The Difference Between Masking and Retraining
This distinction matters. Sound masking -- using white noise or music to drown out tinnitus -- provides temporary relief while the sound is playing. Once you turn it off, the tinnitus returns unchanged. The underlying neural circuitry hasn't changed.
Neuroplastic sound retraining is fundamentally different. It uses structured sound exposure to induce lasting changes in brain circuitry. Multiple approaches exist, each targeting different mechanisms:
Approach 1: Notched Sound Therapy
How it works: Music or white noise is modified to remove a frequency band centered on the patient's individual tinnitus frequency. The spectral "edges" created by this notch are thought to attract lateral inhibition into the notched region, actively suppressing the hyperactive neurons generating the tinnitus signal.
Key evidence: Okamoto et al. (2010) demonstrated in 39 chronic tinnitus patients that 12 months of tailor-made notched music training (TMNMT) significantly reduced both subjective tinnitus loudness and auditory cortex activity at the tinnitus frequency (measured by MEG). This was the first study to verify both behavioral and neurophysiological changes from notched sound therapy.
More recent evidence has been encouraging. Tong et al. (2023) conducted the largest head-to-head randomized controlled trial comparing TMNMT against Tinnitus Retraining Therapy in 120 patients. After 3 months, both treatments were effective, but TMNMT appeared superior -- with simpler processes and higher patient compliance.
A 2025 meta-analysis pooling 14 randomized controlled trials and 793 patients found notched music therapy produced significant improvements over conventional music, with the effect growing stronger over time: THI score improvements of -8.6 points at 3 months and -24.6 points at 6 months.
Approach 2: Tinnitus Retraining Therapy (TRT)
How it works: Developed by Pawel Jastreboff in 1990, TRT targets a different part of the problem. Tinnitus becomes clinically distressing not just because of the auditory signal, but because the limbic system (emotion) and autonomic nervous system (stress response) react to it as a threat. TRT combines directive counseling -- reclassifying tinnitus from a threatening to a neutral signal -- with low-level broadband sound therapy to reduce the tinnitus signal-to-noise ratio.
Jastreboff explicitly warns against masking in TRT: patients cannot habituate to a sound they're trying to cover up. The goal is to hear the tinnitus and the therapy sound simultaneously, training the brain to filter out the tinnitus as irrelevant -- like how you eventually stop hearing a ticking clock.
Clinical evidence: Multiple case series report approximately 80% of patients achieving significant improvement. However, most of this evidence is Level 4 (case series), and a 2020 Cochrane Review noted that TRT shows no significant advantage over standard sound therapy or hearing aids alone -- suggesting the sound component, rather than the specific TRT protocol, may be the active ingredient.
Approach 3: Enriched Acoustic Environments
How it works: Rather than removing frequencies (notching), this approach adds them back. Exposure to a spectrally enriched sound environment containing the missing frequencies can reverse tonotopic map reorganization.
Key evidence: Norena and Eggermont (2005) showed in animal models that cortical maps altered by noise trauma returned to normal when exposed to enriched acoustic environments containing the affected frequencies -- despite remaining hearing loss. The missing input was effectively restored, preventing the central gain increase that drives tinnitus.
The Frontier: Bimodal Neuromodulation
The newest approach targets spike-timing-dependent plasticity (STDP) in the dorsal cochlear nucleus -- a brainstem structure implicated in tinnitus generation. By precisely timing auditory stimulation with electrical tongue stimulation, devices like Lenire aim to recalibrate the neural timing rules that have gone awry.
A 2024 multi-site controlled trial published in Nature Communications reported that 58.6% of participants achieved clinically significant improvement with bimodal treatment vs. 43.2% with sound only. In real-world follow-up data from 220 patients, the responder rate was 91.5% with a mean THI improvement of 27.8 points.
What the Science Actually Shows
The evidence for neuroplasticity-based sound therapy is real but nuanced:
- It works for many people, but not everyone. Effect sizes are moderate. Not every patient responds, and we don't yet fully understand why.
- Longer treatment produces better results. The 2025 meta-analysis found improvements roughly tripled between 3 and 6 months of notched therapy. This is consistent with the gradual nature of neuroplastic change.
- The neurophysiology is verified. These aren't just subjective reports. MEG, fNIRS, and EEG studies show measurable cortical reorganization in response to treatment (Huang et al. 2022).
- The field is still maturing. Many trials have small sample sizes, and results across studies aren't perfectly consistent. A 2024 meta-analysis found no significant difference between TMNMT and music controls on THI scores, while the 2025 meta-analysis of 14 RCTs did find significant improvement. The evidence base is growing but not yet definitive.
The trajectory of research over the past 25 years -- from Pantev's initial cortical mapping work, through Okamoto's proof-of-concept, to recent large RCTs and meta-analyses -- is one of steady, incremental validation. Sound can retrain the tinnitus brain. The question we're still answering is exactly how much, for whom, and with which specific approach.