Tinnitus (T) and hyperacusis (H) are common and often co-occurring problems that can severely impact one's quality of life. These conditions may be triggered by cochlear damage, often caused by loud noise or ototoxic drugs. Tinnitus is also a frequent symptom of conductive hearing loss and head or neck injuries, and both T and H are more prevalent in autism spectrum and other developmental disorders than in the general population (Int J Pediatr Otorhinolaryngol. 2015 Oct; 79(10):1683).
Many studies over the past several decades have shown that hearing loss can lead to stronger and more synchronous spontaneous firing in central auditory neurons that are deprived of peripheral input. The spontaneous hyperactivity may result from compensatory increases of central synaptic gains involving the potentiation of excitatory synaptic transmission and/or the depression of inhibitory transmission. This fits with the observation that most people perceive tinnitus with a pitch that lies in the frequency range of their hearing loss.
Hearing loss can also lead to increased brain activity in response to sound. In contrast to spontaneous activity, stronger sound-evoked responses are most pronounced at frequencies below the edge of a sloping, high-frequency loss, potentially accounting for the broad spectrum of sounds that hyperacusis sufferers find uncomfortably loud.
Successful treatment of hearing loss with corrective surgery, hearing aids, or cochlear implants often leads to a partial or even complete remission of T and H, but not in all cases. For instance, hearing aids may only reduce tinnitus loudness if the tinnitus pitch is within the aidable frequency range, typically up to about 6 kHz. Moreover, not all people with T and H are good candidates for surgery, amplification, or implantation.
Sound therapies have long been a popular alternative treatment for T and H and are also used in combination with hearing aids or implants. The original clinical sound stimulus was broadband noise, which was presented at levels just low enough to make the tinnitus less bothersome or just high enough to mask it completely (while the noise was on). Many other stimuli have since been tried, including narrow bands of noise or tones, music, and nature sounds. These have been adjusted for hearing loss, tailored to the tinnitus pitch, and recently paired with electrical stimulation of the vagus or trigeminal nerves. The long-standing hope for sound therapies was that they would at least partly reverse the central neural changes underlying T and H by increasing auditory nerve activity toward pre-hearing loss norms or by driving brain plasticity more directly. In addition to passive sound exposure, auditory perceptual training has also been utilized with these goals in mind.
DO SOUND THERAPIES WORK?
In a study that highlighted the potential of sound treatments for hyperacusis, Formby, et al., had normal-hearing young adults wear low-level broadband noise generators for two weeks, and found that this resulted in a reversible hypoacusis averaging at about 6 dB (J Acoust Soc Am. 2003 Jul; 114(1):55). Noreña and Chery-Croze followed this up by investigating a small sample of hyperacusis patients who had good low-frequency hearing, average high-frequency losses of 30 dB, and loudness discomfort levels (LDLs) that were 30 to 40 dB below normal at all measured frequencies (Neuroreport. 2007 Aug 6; 18(12):1251). They showed that 15 weeks of therapy with broadband tone ensembles, which were individually adjusted for hearing loss and presented at low sensation levels, resulted in an average 10 to 15 dB LDL increase (improvement) across frequency. While this study was not placebo-controlled, only limited benefits (LDL increases of <5 dB) were seen after two weeks, and measurements made one month after the 15-week treatment showed that LDLs had significantly decreased again. These findings suggest that sound therapy was responsible for the improved LDLs, and that the therapy would need to be maintained over time (though perhaps at a reduced dose) for its full benefits to be retained. While there have been several other promising small-sample studies, no larger clinical trials have been conducted, so the potential of sound therapies for hyperacusis remains underexplored.
In contrast, many studies have investigated the efficacy of various sound-based treatments for tinnitus, several of which have some grounding in animal models (although these are plagued with the difficulty of ascertaining whether the animals are experiencing tinnitus). A detailed account is provided in my recent review (Neuroscience. 2018 Sep 20; doi: 10.1016/j.neuroscience.2018.09.012). Animal studies do not appear to support the use of broadband noise or tone ensembles to treat tinnitus (especially tonal tinnitus), even if they could be helpful for hyperacusis, as previously noted. However, it is less clear that treatment with broadband noise could exacerbate tinnitus, especially at the low noise levels that are not intended to completely mask tinnitus (JAMA Otolaryngol Head Neck Surg. 2018 Oct 1; 144(10):938). While many studies have shown low-level broadband noise generators to benefit some tinnitus patients, no clinical trial has demonstrated significant reductions in tinnitus loudness when assessed psychophysically, and any improvement in tinnitus-related distress (measured using various questionnaires) has been modest in relation to that provided by counseling or hearing aid use.
Unfortunately, modest benefits have also been the rule for other sound therapies for tinnitus, including those better motivated by animal work. For example, stimulation with narrowband noise or tone ensembles at frequencies around the tinnitus pitch is intended to scale back the spontaneous hyperactivity and hypersynchrony that are assumed to underlie tinnitus. One possible reason for the typically modest outcomes seen with targeted narrowband sounds is that tinnitus pitch matches can vary widely (by an octave or more) between test sessions, so care is required in their measurement and monitoring. If narrowband sounds delivered around the tinnitus pitch can potentially reverse the central changes that drive tinnitus, then sounds delivered at off-tinnitus frequencies could potentially exacerbate tinnitus, which has sometimes been reported (Hear Res. 2013 Feb; 296:141).
Another approach is to notch-filter music around the tinnitus pitch, which at first glance seems at odds with the observation that presenting a notched broadband noise to normal listeners leads to the transient sensation of a phantom Zwicker tone, with a pitch in the notched frequency band. While notched music is not the same as notched broadband noise (music is more enjoyable to listen to and notched music may not give rise to Zwicker tones), the initial promising finding of a significant reduction in tinnitus loudness after seven to 12 months of notched music treatment (Proc Natl Acad Sci USA. 2010 Jan 19; 107(3):1207) has not been replicated by several larger studies.
Finally, it is not yet clear whether pairing tones with electrical stimulation of the vagus nerve via implanted electrodes (Sci Rep. 2017 Sep 20; 7(1):11960) or the trigeminal nerve via surface electrodes (Sci Transl Med. 2018 Jan 3; 10:422) will yield much better results. However, some of these patients reported a meaningful or even complete remission of tinnitus, and there is potential to further improve the acoustic and electrical stimulus pairing.
It is fair to say that some tinnitus patients respond quite well to sound therapy while others don't benefit at all. Much more work is needed to understand this and to better predict individual treatment outcomes. Some of the variability in treatment efficacy is undoubtedly due to the different etiologies and central manifestations of tinnitus and to differences in its perceptual qualities, especially its pitch (if tonal) and loudness. For example, while tinnitus pitch typically lies in the frequency range of the hearing loss, in some cases it is matched to the audiogram edge-frequency or even below. The relationship between the audiogram and tinnitus pitch (or more precisely, between the type and extent of peripheral damage and tinnitus pitch) should have important implications for which sound therapy could be most useful for a given patient. Sound therapy trials should include audiograms up to 16 or 20 kHz (DPOAEs and ABRs would also be helpful), as well as tinnitus pitch and loudness matches at regular intervals during the trial. Preferably, the sound stimulus should be verified with real-ear measurements.
Finally, it should be noted that abnormal brain activity related to T and H extends beyond the central auditory system, encompassing many other brain regions. In recent U.K. patient surveys, educational counseling and cognitive behavioral therapy were rated as the two most effective T and H treatment components, far ahead of in-ear or bedside broadband noise generators (Int J Audiol. 2016 Sep; 55(9):514). Nevertheless, counseling and psychotherapy only improve patients’ abilities to cope with T and H; they do not make these bothersome and potentially debilitating conditions disappear. Mindfulness and other stress management techniques are also useful, but again do not represent a cure. A cure would likely need to reverse the central auditory hyperactivity that underlies T and H. The extent to which this can be achieved with sound therapies remains to be seen.