No study has shown that Ozempic, Wegovy or Mounjaro cause tinnitus. What exists is a small number of tinnitus reports to the FDA, a tinnitus signal for liraglutide (an older drug in the same class), and a separate ear effect of rapid weight loss with a clear mechanism: a eustachian tube that stays open.
Ozempic and Wegovy are semaglutide; Mounjaro and Zepbound are tirzepatide. They belong to the class usually called GLP-1 receptor agonists (tirzepatide also activates a second hormone receptor, for GIP), and in the US they arrived in quick succession: Ozempic for type 2 diabetes in December 2017, Wegovy for weight management in June 2021, Mounjaro in May 2022 and Zepbound in November 2023.
What the prescribing information says
The US labels for all four drugs, as posted on DailyMed in 2026, do not mention tinnitus, hearing loss or the ear anywhere.
Dizziness is listed. In the weight-reduction trials summarised in the Wegovy label, 8% of people on the drug reported it against 4% on placebo, and the Zepbound label gives 4% to 5% across doses against 2% on placebo.
Each label also lists reactions reported after approval, and the ear is absent there too. The fuller record of those reports sits in the FDA's adverse-event database.
What the adverse-event reports show
Two 2025 studies searched the FDA Adverse Event Reporting System (FAERS), which holds reports of suspected side effects sent in by manufacturers, health professionals and the public, for ear problems with GLP-1 drugs.
- A count. A study in Otology & Neurotology found 97,237 adverse events reported with these drugs. Of those, 958 (0.99%) involved the ear, and 93 were tinnitus, fewer than one in a thousand. Hearing loss (515 reports) and vertigo (203) were both more common.
- A comparison. A study in The Laryngoscope asked whether each ear, nose and throat problem was reported more often with each drug than with other drugs. Liraglutide, approved in 2010 and one of the older drugs in the class, showed a significant signal for tinnitus, and exenatide one for hearing disability. Semaglutide's listed signals included loss of smell, dry mouth, altered taste and Bell's palsy; the abstract lists no tinnitus signal for semaglutide or tirzepatide. Tirzepatide contributed only months of data, because the analysis counted reports from a year after each drug's approval to the end of 2023.
A larger 2026 analysis in Obesity, covering 137,451 GLP-1 reports drawn from more than 18 million, found hearing loss reported more often when the drugs were used for diabetes than when they were used for weight control. The paper does not mention tinnitus.
What a "signal" means, and what it does not
A disproportionality signal compares reports with reports. It says that tinnitus makes up a larger share of the complaints filed about one drug than of those filed about others. It does not say how many people on the drug develop tinnitus, and it cannot.
The FDA spells this out on the public dashboard for the database, which it has folded into a wider Adverse Event Monitoring System:
- the existence of a report does not establish that the drug caused the event, which may have come from the disease being treated, another medicine or something else
- reports are not medically verified
- duplicate and incomplete reports are common
- the reports cannot be used to estimate how often an event occurs
The researchers say the same of their own work. The Obesity paper names underreporting and differential reporting bias (some events, and some drugs, being reported more readily than others) as the database's two major limitations, and calls its findings signals rather than definitive evidence of causality.
One detail from the Otology & Neurotology count shows how coarse the tool is. Its 97,237 events included no reports of autophony, hearing your own voice or breathing loudly in the ear, and two of ear fullness, yet the two patients the same paper described had exactly those symptoms. A reporting system can only count what reporters know to name.
Larger studies measured hearing and balance, not tinnitus
These studies compare groups of patients rather than counting reports. None of them measured tinnitus.
- Balance, against people not on the drugs. A 2025 study of US health records in the TriNetX network followed 419,497 adults starting semaglutide and 77,259 starting tirzepatide, each matched with someone with diabetes or obesity who was not given a GLP-1 drug. New diagnoses of vestibular (balance-system) disorders were more common on both drugs, though rare in absolute terms: by three years, 0.41% of semaglutide users against 0.16% of controls. The reported hazard ratios for semaglutide ran from 4.02 to 4.95, though those cumulative figures imply a smaller gap, about 2.5 times.
- Hearing and balance, against other diabetes drugs. A 2026 study in the same network compared 342,046 adults with type 2 diabetes starting a GLP-1 drug with 342,046 starting a different diabetes medicine, a DPP-4 or SGLT2 inhibitor. Diagnosed hearing impairment was more common on GLP-1 drugs, 9,522 cases against 7,898, a hazard ratio of 1.17. Peripheral vestibular disorders were slightly more common, at 1.08. The records held no hearing test results, and the authors caution that the association may not be causal.
- Randomised trials. A 2025 network meta-analysis of 29 randomised trials and 145,895 participants counted hearing-loss events recorded in trial registries. Only lixisenatide and a high dose of efpeglenatide were linked to more events than control. Oral semaglutide was borderline, and low doses of injectable semaglutide and of tirzepatide were linked to fewer.
The first two studies differ in the way that matters most. Against people not given a GLP-1 drug, the balance-disorder risk was roughly 2.5 to 5 times higher; against people starting another diabetes drug, it was 8% higher. The outcome definitions were not identical, so this is not a clean comparison, but it is a reminder, familiar from reading treatment trials, that the answer to "compared with what?" can carry much of a result.
Diabetes is part of that problem. In a 2026 Italian study of patients attending an audiology clinic with hearing loss, those with type 2 diabetes had worse average hearing thresholds than those without it, 64.4 against 56.4 dB, after adjusting for age, sex, body mass index and smoking. The Obesity finding on diabetes use points the same way. When a drug is prescribed for a condition already linked to hearing problems, a database cannot easily tell the drug's effect from the disease's.
The ear effect with a mechanism: a tube that stays open
Rapid weight loss is a recognised risk factor for a patulous eustachian tube, one that stays open when it should be closed, because the fatty tissue that helps support it is lost. A closed tube protects the middle ear from the sound of your own voice and breathing; an open one lets that sound through. A 2026 review of GLP-1 drugs in ear, nose and throat practice names the patulous tube among the possible effects of weight loss on these drugs.
A run of reports now ties it to GLP-1 drugs specifically:
- The 2025 Otology & Neurotology paper described two patients with autophony and ear fullness after starting a GLP-1 drug. In both, nasal endoscopy showed significant loss of tissue bulk in the cushions at the tube's opening.
- A 2025 German case series described seven adults, aged 28 to 56, who developed a patulous tube 4 to 10 months after starting semaglutide or tirzepatide, having lost 8.2% to 18.7% of their body weight. All seven had autophony, six had intermittent ear fullness and five could hear their own breathing. One improved with conservative measures; the other six had an implant material injected into the tube, after which their symptoms resolved.
- A 2026 case report described a 79-year-old man who had lost about 200 lb on semaglutide and developed intermittent fullness and popping in both ears. He was managed with hydration, intranasal saline and avoiding decongestants.
Hearing your own breathing in one ear is easy to describe as tinnitus, but it is a real sound reaching a space that should be sealed off from it. The patulous eustachian tube article explains how it differs from the ordinary, blocked kind of eustachian tube dysfunction.
The tell is posture. In the Japan Otological Society criteria set out in a 2020 review in JMA Journal, clear improvement when lying flat, or when the tube's opening is blocked, is one of the three findings a definite diagnosis requires.
Two cautions belong with this. The evidence is case reports and small series: a 2026 review of the parallel literature on bariatric surgery concluded that surgery cannot yet be established as an independent cause of a patulous tube. And the management differs from blocked-ear advice. The JMA Journal review lists rehydration and avoiding further weight loss among its first measures, and the 79-year-old's care included avoiding decongestants. The Wegovy label also warns of dehydration after nausea, vomiting or diarrhoea, though the evidence tying dehydration to tinnitus is thin.
Could these drugs help some tinnitus instead?
There are hints in the other direction too, and they are weaker still.
- Raised pressure around the brain. Idiopathic intracranial hypertension lists pulsatile tinnitus among its three main symptoms, with headache and vision loss, and weight loss is part of its established treatment. In a 2023 randomised trial, 16 women with the condition received exenatide, a GLP-1 drug, or placebo. Exenatide lowered the pressure within hours, by 5.7 cmCSF at 2.5 hours; the fall at 12 weeks met the trial's pre-set threshold of p below 0.1 but not the usual 0.05. A 2026 meta-analysis of eight mostly observational studies and 13,243 patients linked GLP-1 drugs to less optic disc swelling, headache and visual disturbance, and called the evidence low-certainty. Tinnitus is not among the trial's reported outcomes or the meta-analysis's primary ones. A 2025 study of US health records tracked tinnitus diagnoses in 555 people who started a GLP-1 drug and 555 treated conventionally, but there were too few cases to compare the groups.
- Hearing protection in animals. In chinchillas exposed to repeated blasts, liraglutide reduced hearing damage and hair-cell loss, most of all when given before the blast. An animal result is a reason to run trials, not evidence of benefit in people.
- One clinic study. In the Italian audiology study, GLP-1 use was associated with better hearing thresholds, but only 27 patients were taking semaglutide and 12 dulaglutide, and the authors call the finding novel and hypothesis-generating.
Put together, the evidence points in both directions, all of it indirect, and none of it answers whether these drugs change the risk of tinnitus.
If you notice ringing after starting one
- Do not stop the drug on your own. The FDA advises patients to talk to their doctor before stopping or changing how they take a medicine. GLP-1 drugs are not among the medicines with established ear toxicity, but new symptoms after a new drug are worth raising with the prescriber.
- Describe the sound precisely. Your own voice booming, your breathing audible in the ear, and relief when you lie down point towards a patulous tube, which an ENT can check. Use the word autophony, and say how much weight you have lost and how fast.
- Mention dizziness too. It is a listed side effect of Wegovy and Zepbound, and vertigo with tinnitus is evaluated differently from tinnitus alone.
- Treat a sudden drop in hearing as urgent, whatever you are taking. NIDCD advises treating sudden deafness as a medical emergency and seeing a doctor immediately; sudden sensorineural hearing loss explains why the window is short.
- Get whooshing in time with your pulse assessed promptly, especially with headaches or changes in vision. The when to see a doctor checklist sorts symptoms by urgency.
- Consider reporting it. The FDA encourages consumers and health professionals to report suspected side effects.
- If the ringing persists once the ear has been checked, it is managed like any other tinnitus, and the treatment comparison sets out what has evidence behind it.