The short version
- In a double-blind French trial of 145 adults with mild hearing loss, a hearing aid running a sound generator cut tinnitus distress by about 21 points on a 100-point scale, versus 6 for placebo.
- Adding amplification to the sound generator did not help more; the combined arm scored slightly lower than the sound generator alone.
- The trial lasted 12 weeks, used a self-report questionnaire, and only covered people whose tinnitus came with mild hearing loss in both ears.
For adults whose tinnitus comes with mild hearing loss, a hearing aid programmed to play a steady background sound reduced tinnitus distress about three times as much as a placebo aid in a 12-week French trial.
The design
The study, published in Scientific Reports by P. Fournier, H. Thai-Van, P. Reynard and colleagues and reported this week by Hearing Health & Technology Matters, is the kind of trial tinnitus research rarely gets: double-blind, placebo-controlled, and run across six specialist tinnitus clinics in France.
Researchers enrolled 145 adults with chronic bothersome tinnitus and mild hearing loss in both ears; 128 finished. Mean age was 54. Every participant was fitted with the same hearing aids, which were programmed three ways:
- Placebo amplification: the aid was worn and programmed as a placebo, with no generated sound.
- Sound generator only: the aid played a generated background sound, with no amplification.
- Sound generator plus amplification: both.
Every arm also received standardized counselling, so the comparison isolates what the device did on top of talking to a clinician. Participants wore the aids for 12 weeks.
The results
The main outcome was the Tinnitus Functional Index, a self-report questionnaire scored from 0 to 100 that covers how intrusive the tinnitus is and how much it interferes with sleep, concentration, mood and daily life.
- Sound generator only: down about 21 points.
- Sound generator plus amplification: down about 18 points.
- Placebo: down about 6 points.
Asked whether they had improved meaningfully, 88% of the sound-generator group said yes, 77% of the combined group, and 33% of the placebo group.
Two things in those numbers are worth pausing on. The placebo arm improved, which is what counselling plus the expectation of a device does; the 6-point drop is the floor the real treatments had to clear. And adding amplification to the sound generator did not add benefit, which is mildly surprising, since amplifying everyday sound is itself a common tinnitus strategy.
The limits
Twelve weeks is short. Tinnitus is a condition people live with for decades, and the question that matters to them is whether the relief lasts at month 12, not month 3. This trial cannot answer it.
The Tinnitus Functional Index is a self-report measure. That is the standard in the field, because there is no objective way to measure how loud someone’s tinnitus is, but it means the result is a change in how people feel about their tinnitus rather than a change in the sound itself.
The population is narrow. Everyone had mild hearing loss in both ears. The trial says nothing about tinnitus in people with normal hearing, which is a large share of sufferers, or about people with more severe loss.
There was no amplification-only arm in this phase, so the trial cannot say whether a standard hearing aid, without the sound generator, would have done as well.
Why it matters
Sound therapy for tinnitus has a long history and thin evidence, mostly because the trials have been small, unblinded, or both. Six clinics, a placebo device, double-blinding, and a 128-person completion count make this one of the stronger tests in the literature. The effect size is large by the field’s standards.
The practical reading for the specific group studied: a hearing aid with a sound generator, prescribed alongside counselling, is a reasonable thing to ask an audiologist about. Whether it is the aid, the sound, or the counselling-plus-ritual doing the work is partly answered here, since all arms got counselling and only the sound-generator arms improved.
What to watch
Whether the group publishes a longer follow-up of these patients, and whether a later phase adds the missing amplification-only arm. Until then, the result holds for 12 weeks in people with mild hearing loss, and that is the honest size of it.
