Article Text



Around 10% of people experience subjective tinnitus (the perception of sound, only audible to the patient, in the absence of an external auditory stimulus).1–3 It may be associated with hearing loss, anxiety, depression, sleep disturbance, concentration problems or reduced quality of life; for around 0.5% it is extremely disturbing.1–4 Risk factors include aging, significant noise exposure, drug therapy (e.g. aminoglycosides, NSAIDs, diuretics), or disorders of the outer, middle or inner ear or auditory nerve (e.g. ear wax, infections, vestibular schwannoma, otosclerosis).1,2,4 It may be due to excessive spontaneous activity in the auditory system and brain; if the signal (normally suppressed by the subconscious) becomes noticed it becomes more intrusive and annoying in a vicious cycle.5 Here, we discuss symptomatic drug and non-drug treatments for subjective tinnitus in adults. We do not cover treatment of underlying causes of tinnitus.

Statistics from

Assessment and referral

There are four different levels of service for tinnitus: primary care (GPs); local community-based tinnitus services (audiologists and hearing therapists); specialist hospital-based centres (multidisciplinary teams that include audiologists, hearing therapists, ENT specialists, audiovestibular physicians, and clinical psychologists); and supra-specialist assessment centres (multidisciplinary teams that can offer more complex audiological assessments and neurosurgical interventions).

Patients with non-troublesome tinnitus without hearing difficulties require only reassurance and education about the condition from their GP.1 Information and resources for patients, carers and healthcare professionals are available from the British Tinnitus Association ( Self-rated disease-specific scales may be used to assess severity (see Box).1,6 The history should include any impact on sleep, associated audio-vestibular symptoms, psychological/psychiatric conditions, history of noise exposure, drug intake and level of stress/anxiety. The ear should be examined, as should the cranial nerves for underlying neurological conditions.1

Box: Examples of disease-specific health-related quality of life scales6,7

Tinnitus Handicap Inventory (THI): 25 items, range 0–100 (also THI-12 version)*

Tinnitus Handicap Questionnaire (THQ): 27 items, range 0–2700*

Tinnitus Questionnaire (TQ): 52 items, range 0–84*

Tinnitus Reaction Questionnaire (TRQ): 26 items, range 0–104*

Tinnitus Severity Index (TSI): 12 items, range 0–48

*Validated scale.

If tinnitus persists and causes distress despite reassurance, or is associated with hearing loss, the patient could be referred to a community-based audiology service for assessment and support (including hearing aid fitting if necessary).1

Patients should be referred to a specialist service if they have:

  • distressing tinnitus and anxiety, depression or sleep disturbance;

  • unilateral or pulsatile tinnitus;

  • unilateral hearing impairment, sudden/fluctuating hearing loss or deterioration, dizziness, vertigo or imbalance or asymmetrical hearing loss;

  • other abnormalities of auditory perception (e.g. hyperacusis);

  • tinnitus and normal peripheral hearing, but difficulty hearing in noisy backgrounds, or with sound localisation or following complex auditory directions;

  • associated systemic or neurological disorders.1

A referral to a supra-specialist centre may be required for patients with profound hearing loss requiring cochlear implants or audiovestibular pathologies requiring specialist neuro-otological interventions.1

Non-drug treatment


A systematic review found nine randomised controlled trials (RCTs; mostly poor quality) of acupuncture for tinnitus.8 Five (n=167; data not pooled) compared acupuncture or electroacupuncture with sham-acupuncture and found no benefit (e.g. mean difference in THI 0.5, 95% CI –14.1 to +15.1).8 Two compared one-time scalp acupuncture with penetrating sham acupuncture at non-acupoints; in one (n=76), more patients having acupuncture achieved subjective symptom relief on a visual analogue scale (VAS; mean difference –1.34, p=0.02); the other (n=38) found no difference on otoacoustic emission amplitude. Two compared acupuncture with drugs (bandazol and flunarizine hydrochloride); acupuncture increased the response rate in one (risk ratio [RR] 2.20, 95% CI 1.27 to 3.89; n=90), but not the other (risk ratio 1.14, 95% CI 0.87 to 1.49; n=60). The review authors concluded that the trials are insufficient for definitive conclusions.8

Auditory perceptual training

Auditory perceptual training involves an active listening task (e.g. discrimination between tones of different frequencies, or concentrating to reduce tinnitus to match the loudness of a stimulus).9 A systematic review found ten studies (eight low quality, two moderate quality; only one RCT; n=1 to 63).9 None reported a significant change in an outcome measure (no data shown) and the review authors concluded that the evidence is not robust enough to guide treatment.9

Cognitive behavioural therapy

Cognitive behavioural therapy (CBT) uses relaxation, restructuring of thoughts, and exposure to exacerbating situations to promote habituation to a symptom.10 A Cochrane systematic review found eight RCTs (n=468).10 For tinnitus loudness, there was no difference between CBT and waiting list controls (six studies, standardised mean difference [SMD] 0.24, 95% CI –0.02 to +0.51) or other interventions (e.g. yoga, education; four studies, SMD 0.1, 95% CI –0.22 to +0.42). CBT had a small benefit on depression compared with waiting list controls (six studies, SMD 0.37, 95% CI 0.15 to 0.59) but not against other interventions (three studies, SMD 0.01, 95% CI –0.43 to +0.45). CBT improved quality of life scores more than waiting list controls (five studies, SMD 0.91, 95% CI 0.50 to 1.32) or other interventions (three studies, SMD 0.64, 95% CI 0.29 to 1.00).

A second systematic review included three RCTs (n not stated) of internet-based CBT, which gave a small improvement versus controls (waiting list, internet-based education without CBT, or therapist-delivered CBT; effect size [Cohen's di]=0.38, p=0.016).7

A later RCT (n=492), compared CBT plus multidisciplinary review with usual care (treatment for 8 months and follow up for 4 months).11 CBT gave small to medium effect sizes at 8 and 12 months (mean difference in health related quality of life: 8 months: d=0.18, p=0.0258; 12 months: d=0.24, p=0.0009. TQ: 8 months: d=0.41, p<0.0001; 12 months: d=0.43, p<0.0001. THI: 8 months: d=0.52, p<0.0001; 12 months: d=0.45, p<0.0001).

Hearing aids

Since hearing loss is associated with tinnitus, and straining to listen may increase central auditory gain, hearing aids have been suggested as a treatment even if hearing loss is relatively mild.4 However, a systematic review found no RCTs investigating hearing aids as a primary intervention for tinnitus using validated outcome measures.7 Decisions on when to start using a hearing aid are up to the individual patient and audiologist.4

Hyperbaric oxygen therapy

This involves breathing pure oxygen in a compression chamber in an attempt to increase the oxygen supply to the ear and brain, and reduce tinnitus severity.12 A Cochrane systematic review found two RCTs (n=53) in patients with acute tinnitus that reported greater mean improvement on a VAS with intervention than control (3.1 points in one trial and 0.4 points in the other; no standard deviations reported).12 A third RCT (n=25) found no difference between hyperbaric oxygen and medical therapy (RR 1.7, 95% CI 0.6 to 4.7).12 Two RCTs (n=83) in patients with chronic tinnitus found no difference between hyperbaric oxygen and controls (RR 0.68, 95% CI 0.232 to 1.42). The review authors concluded that there was no significant improvement with hyperbaric oxygen for chronic tinnitus, and the significance in acute tinnitus could not be assessed.12


Hypnosis is achieved by relaxing the body, then shifting attention from the external environment towards ideas suggested by the therapist; the goal is to gain self-control over behaviour, emotions or physiological processes.13 A systematic review found two RCTs of hypnotherapy.13 In the first study (crossover design, n=14), hypnotherapy focusing on tinnitus reduction was compared with hypnotherapy focusing on ego strengthening; the tinnitus itself was not altered, although five participants achieved relaxation. In the second study (parallel design, n=86) there was no difference in the tinnitus measures but 20 of the 44 subjects receiving hypnotherapy reported a general sense of well-being compared with 6 of 42 who received a single counselling session.13 The review authors concluded that hypnotherapy leads to relaxation and well-being that make tinnitus more bearable.13


Simple reassurance that tinnitus is common and the noise is harmless in itself may help.14 A systematic review found two RCTs assessing a self-help book compared with waiting-list controls.7 They provide moderate quality evidence; one trial (without therapist contact) reported a small effect (d=0.25) on the TRQ (50% reduction in TRQ score: 25% intervention vs. 14% controls, p value not stated) and the other (with therapist contact) a moderate effect (d=0.52) on the TRQ (50% reduction: 32% vs. 5%, p value not stated).

Low-power laser therapy

Low-power lasers are thought to work by reducing the compound action potential of nerves (e.g. the eighth nerve when the laser is directed into the cochlea).13 A systematic review found three RCTs (n=126); none showed any significant difference between laser and placebo on any outcome.13

We found two later RCTs of laser therapy (20 minutes daily for 3 months) versus sham laser.15,16 In the first trial (n=46), all patients received hypnotherapy and muscle relaxation.15 The change in the THI score was greater for the laser group (17.1 vs. 7.3 points, p<0.05).15 In the second trial (n=60), no differences were found between the groups for change in THI score (8.7 points vs. 8.8 points, p=0.97), a 0–100 VAS for tinnitus loudness (0.8 points vs. 3.1 points, p=0.69) or the minimum masking level (2.7 points vs. 1.4 points, p=0.42).16

Relaxation therapy

A systematic review found three RCTs investigating relaxation therapy (Qigong [a mindful exercise and active relaxation technique], progressive muscle relaxation training, or electromyography feedback), providing moderate or low quality evidence (d=1.63 on the TQ; d=0.43 on the THI-12; and d=0.05 on the Beck Depression Inventory).7 The review authors concluded that there is evidence for efficacy of relaxation for tinnitus intrusiveness, mixed evidence for depressive symptoms and no evidence for anxiety.7

Repetitive Transcranial Magnetic Stimulation

Repetitive Transcranial Magnetic Stimulation (rTMS) involves using an insulated coil to create a magnetic field over the left hemisphere, which produces a secondary electrical current in the hyperactive brain area.17 A systematic review found three sham-controlled RCTs (n=14, n=19 and n=42, respectively); each study was reported as showing significant benefits of rTMS over sham although no numerical data were presented.17 However, the review authors concluded that the results should be interpreted cautiously because there were flaws in all three studies and patients with psychological symptoms were excluded, so the results may not apply to the general tinnitus population.17

We found three further RCTs of rTMS versus sham.1820 The first (n=21), compared rTMS over the temporal cortex for 5 days with sham rTMS for 5 days in a crossover trial with a 1-week washout period between treatments.18 The VAS score for loudness decreased after active but not sham treatment, although the main effect comparing treatments was contaminated by carryover of active treatment into the sham week.18 Nine patients were ‘responders’ (i.e. experiencing ≥33% decrease in the loudness VAS after active treatment).18

The second trial (n=22) compared rTMS over the auditory cortex with sham rTMS.19 One week after treatment, TQ global scores had improved more in the active group (–8.58 points vs. +0.1 points with sham, p<0.01), as had THI scores (–8.33 points vs. 0 points, p<0.01); however, the difference was not significant at 1 month.19

The third trial (n=48) assessed 4 weeks of bilateral rTMS to the temporal or temporoparietal cortex or sham rTMS behind the mastoid.20 There was no significant difference on tinnitus severity between sham and temporal rTMS (mean difference 0.4, 95% CI –5.4 to + 6.7) or temporoparietal rTMS (mean difference 0.2, 95% CI –5.9 to +6.3).20

Sound enrichment or masking therapy

Enriching the sound environment (e.g. quiet uneventful music, a fan or a water feature) may be helpful.4 A systematic review found three RCTs investigating sound enrichment, providing low to moderate quality evidence (e.g. d=0.44 on the THI, not clinically significant).7

Masking devices introduce a sound, usually ‘white noise’ (like an out of tune radio signal), to completely mask the tinnitus, or at a low level to achieve habituation, or at a level where the patient can hear both their own tinnitus and the external noise at the same time.21 This may be combined with a hearing aid or be delivered via conventional equipment (e.g. headphones).21 A Cochrane systematic review found six trials (n=553) which varied in design and outcome measures, precluding meta-analysis.21 No significant change in tinnitus loudness or severity was found compared with other interventions (education, relaxation, coping strategies, counselling, tinnitus retraining therapy (TRT) or environmental sounds).21

Tinnitus retraining therapy

Tinnitus retraining therapy (TRT) is a combination of counselling and sound therapy/enrichment to counteract the pathological positive feedback process and promote habituation to the tinnitus.22 A Cochrane systematic review found one RCT (n=123) comparing TRT with masking.22 This study had a high risk of bias. Outcome data were presented for patients in three groups: participants' tinnitus being a ‘moderate problem’, ‘big problem’ or ‘very big problem’, using the THI, THQ and TSI. At 18 months, improvements were greater for TRT than for masking in each of the three groups (moderate problem: THI: 18.2 vs. 4.6, THQ: 489 vs. 178, TSI: 7.5 vs. 1.6; big problem: THI: 29.2 vs. 16.7, THQ: 799 vs. 256, TSI: 12.1 vs. 6.7; and very big problem: THI: 50.4 vs. 10.3, THQ; 1118 vs. 300, TSI: 19.7 vs. 4.8; no p values reported). The review authors concluded that a single, low-quality RCT suggests that TRT is much more effective than masking.


Relief of tinnitus has been reported as an unexpected side-effect of ultrasound.13 A systematic review found two RCTs (n=40 in each) comparing ultrasound with placebo.13 One found no significant difference in tinnitus measures; the other did not measure tinnitus severity, but a subjective feeling of improvement was more frequent in those on active treatment (no numerical data presented).13

Drug treatment

There are no medicines specifically approved for tinnitus treatment (apart from in the context of Meniere's disease, labyrinthitis or other underlying conditions) listed by the US Food and Drug Administration, the European Medicines Agency or the Medicines and Healthcare products Regulatory Agency; any drugs are therefore used ‘off-label’ for this symptom.5


Acamprosate (licensed for maintaining alcohol abstinence) acts as a glutamate receptor antagonist and GABA agonist, and has been tried as a means of reducing the glutaminergic (excitatory) transmission and increasing the GABA (inhibitory) activity in the auditory pathway.23

In an RCT (n=41), more patients on acamprosate reported any improvement in tinnitus disturbance than on placebo at day 90 (any non-zero difference [i.e. not necessarily clinically relevant]: 87% vs. 44%, p=0.004).24 The difference in the incidence of adverse events (all mild in severity) was not significant (12% vs. 20%). The authors concluded that acamprosate was a safe and successful treatment alternative.

In an RCT (n=40), participants with tinnitus received acamprosate or placebo for 45 days, followed by a washout period of 1 week.23 Acamprosate reduced the tinnitus score in more participants than placebo (any reduction [i.e. not necessarily clinically relevant]: 93% vs. 13%, p value not stated) and also reduced tinnitus matching loudness measures (8dB on active treatment vs. 0 on placebo, p<0.0001). The authors reported that the acamprosate was well tolerated without serious drug reactions.3

However, the clinical relevance of the results of both these studies is not clear.


Anticonvulsants have been suggested to reduce central auditory hyperactivity.25 A Cochrane systematic review found seven trials (n=453) comparing drugs used in epilepsy with placebo.25 The risk of bias of most studies was high or unclear.25 None of the three studies that used a validated questionnaire showed a significant positive effect of anticonvulsants.25 A meta-analysis of ‘any positive effect’ showed a small benefit of anticonvulsants (risk difference [RD] 14%, 95% CI 6% to 22%). A meta-analysis of ‘near or total eradication of tinnitus annoyance’ showed no effect (RD 4%, 95% CI –2% to +11%). Unwanted effects were reported by 18% of patients. The review authors concluded that anticonvulsants do not have a large positive effect in tinnitus.25

We found one later RCT (n=80), assessing gabapentin for 8 weeks versus placebo.26 This found no significant difference in decrease or resolution of tinnitus (50% vs. 55%), decrease in TSI score (3.53 vs. 2.28) or reduction in mean loudness score (1.50 vs. 1.43).26


Since tinnitus is frequently associated with depression, tricyclic antide­pressants (TCAs, e.g. amitriptyline), selective serotonin reuptake inhibitors (SSRIs, e.g. fluoxetine) and other antidepressants (e.g. trazodone) have been tried.5 Such drugs may work directly on the tinnitus by acting on receptors in the central auditory pathway, and/or indirectly by treating depression.5

A Cochrane systematic review found six trials (n=610) assessing antidepressants.5 Trial quality was generally low, and some of the trials had large losses to follow-up (reflecting significant unwanted effects of antidepressants).5 All the TCA trials suggested that there was a slight improvement in tinnitus but these effects may have been attributable to methodological bias.5 One high-quality SSRI trial found no improvement in any of the validated outcome measures used.5 The trial that used trazodone showed no benefit over placebo.5 The review authors concluded that there is insufficient evidence for antidepressants in tinnitus.5

We found one later RCT (n=75), assessing sertraline for 16 weeks versus placebo.27 This found that sertraline was more effective in improving the Psychological General Well-being Index (mean change 21 vs. 3, p=0.001; clinical relevance of a difference of this size not stated).27


Since tinnitus is frequently associated with anxiety, anxiolytics have been suggested, especially those which are also GABA agonists (e.g. benzodiazepines) as GABA is thought to be inhibitory in the auditory pathway.1,23 A systematic review found one high-quality RCT comparing alprazolam with chlorphenamine, which reported no significant difference in the THI (d=0.07 [not significant]).7

We found two further RCTs comparing an anxiolytic with placebo.28,29 In the first trial, compared with placebo, clonazepam significantly reduced tinnitus intensity (p=0.003) and annoyance (p=0.001), but four of the ten patients on clonazepam experienced drowsiness and nausea and two had sexual dysfunction.28 In the second (crossover) trial, 23 patients received vestipitant (a drug in development; an antagonist of neurokinin-1 receptors, which inhibit GABAergic neurones) or placebo for 14 days, with a 14-day washout in between.29 No statistically significant treatment effect was found (e.g. THI 29 vs. 28, p value not stated).29

Ginkgo biloba

It is thought that ginkgo biloba may promote increased blood flow and alter neurone metabolism.30 There is at least one product with a Traditional Herbal Registration (

A Cochrane systematic review included three RCTs (n=1,143) of ginkgo biloba versus placebo.30 None found a significant difference (e.g. improvement in tinnitus, 13.6% vs. 12.4%; loudness changed from 42.3dB to 39.0dB vs. 44.1dB to 45.1dB; mean difference in THI between groups –2.5, 95% CI –10.5 to +5.1).30 The most commonly reported unwanted effect of ginkgo biloba was mild gastrointestinal disturbance (e.g. stomach pains, change in bowel habit); serious unwanted effects are rare but include bleeding problems, interaction with anticoagulants and seizures.30


Melatonin has been proposed as a treatment for tinnitus, based on its favourable effects on sleep and its vasoactive and antioxidant properties.31 A systematic review found two RCTs (one high quality, one moderate, n not stated) comparing melatonin versus no treatment or placebo; neither found a significant effect on the THI (no data presented).7

We found two further RCTs.32,33 In the first trial (n=60), mean post-treatment scores on a 0–10 VAS were not significantly different (6.5 vs. 7.0, p=0.174) and most patients had no change in their tinnitus (60% vs. 78%, p value not stated); in the remainder, the intensity of the tinnitus was reduced and they had less distress.32 Melatonin did not cause more unwanted effects than placebo.32 In the second (crossover; n=61) trial, patients received melatonin or placebo for 1 month with a 1-month washout in between.33 More patients on melatonin than placebo experienced a decrease on at least two of tinnitus matching, TSI and self-rated tinnitus (57% vs. 25%, p value not stated), and sleep improvement (57% vs. 36%, p value not stated).33 No unwanted effects were reported.33

What do guidelines say?

The UK Department of Health Good Practice Guide for Provision of Services for Adults with Tinnitus notes that psychological elements of treatment appear to be particularly important and suggests that patients may be given one or more of the following interventions:

  • information/education;

  • hearing aids;

  • counselling and psychological support;

  • relaxation therapy;

  • CBT with the requisite professional supervision;

  • sleep management (including supervised CBT);

  • sound enrichment therapy;

  • tinnitus retraining therapy.1

None of the drug treatments that have been tried were recommended. However, in relation to supra-specialist services, the guidance included the comment: “consider off label prescriptions after assessment and discussion of risk/benefit and patient expectation”.

We found no guidelines on the management of tinnitus from the National Institute for Health and Clinical Excellence or the Scottish Intercollegiate Guidelines Network.


Patients with tinnitus require careful assessment including a clinical history and examination to identify any treatable causes of the condition and any associated hearing loss or psychological problems (e.g. anxiety, depression), supported by relevant audiovestibular tests. Few guidelines are available on treatment; a good practice guide suggests a number of non-drug treatment options, including several with supporting evidence from a number of studies (e.g. information/education, relaxation therapy, cognitive behavioural therapy and sound enrichment or masking therapy) and others for which there is little supporting evidence (hearing aids, counselling and psychological support, sleep management and tinnitus retraining therapy). Other non-drug treatments that have been tried include acupuncture, ‘auditory perceptual training’ (an active listening task such as frequency discrimination or concentrating to reduce tinnitus to match the loudness of a stimulus), hyperbaric oxygen therapy, repetitive transcranial magnetic stimulation (rTMS), low-power laser therapy, hypnotherapy and ultrasound; there is no strong evidence for any of these treatment options at present. Currently, no drugs are licensed for the treatment of tinnitus; antidepressants and anxiolytics may be useful in selected patients but many of the drugs that have been tried are of little or uncertain benefit and have unwanted effects.

View Abstract


  • ii Herbal medicine – all change (DTB 2011; 49: 37).

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.