Exercise probably helps tinnitus modestly over time — more physical activity is linked to lower tinnitus severity — but some people find workouts make it worse, more often the harder they train, and the two findings fit together.
"Does exercise help tinnitus" doesn't have a single clean answer — and the reason why is useful to understand, rather than something to smooth over into a simple yes.
The case that exercise generally helps
An online survey that 1,030 adults with tinnitus started (about 40% of responses were excluded as incomplete) found higher levels of physical activity were significantly associated with improved quality of life and lower tinnitus severity. Physical activity accounted for a small but statistically significant independent portion of the variation in tinnitus severity scores — 0.8% — separate from its broader quality-of-life benefits. The authors' own conclusion was careful: physical activity may be a management strategy, but further testing is needed.
A 2025 scoping review examining physical activity's effect on tinnitus — covering aerobic exercise, resistance training, yoga, and other mind-body approaches — found preliminary evidence that regular physical activity may offer real benefit as an adjunct to tinnitus management, particularly for people with co-occurring anxiety, depression or sleep disturbance. It suggests several overlapping mechanisms: reduced stress, improved cardiovascular and musculoskeletal health, and effects on neuroplasticity (the brain's capacity to change how it processes signals, relevant to how tinnitus perception is thought to develop and persist). It also says plainly that the current evidence is limited and heterogeneous, and that well-designed trials are needed to define the best type, intensity and duration of activity.
The honest complication: intensity seems to matter
A separate international web-based survey complicates the simple "exercise helps" framing. It gathered responses from 6,115 people with tinnitus through an online tinnitus forum, and found that the higher the intensity of the workout, the higher the share of participants who said it had a negative effect on their tinnitus, and the lower the share reporting a positive one. For moderate exercise, 18.3% reported a negative effect and 11.5% a positive one.
The survey also compared people whose tinnitus had been diagnosed as somatic with those diagnosed with other types. The two groups did not differ on moderate or high-intensity exercise; the difference was at the other end — people with somatic tinnitus more often reported light exercise making their tinnitus worse.
The researchers proposed one explanation for why their results ran against the survey above: the temporary rise in blood pressure that accompanies more intense exercise, combined with tinnitus's association with high blood pressure (covered in the article on tinnitus and blood pressure). Two cautions apply. It was a convenience sample of forum users, who may not represent everyone with tinnitus, and everything in it is self-report.
How to hold both findings without contradiction
These aren't incompatible once you separate the timeframes involved: regular, moderate physical activity appears to have a real, if modest, positive relationship with tinnitus severity and quality of life over time — but a single bout of high-intensity exercise may produce a temporary, short-term spike for some people, similar in principle to the flare-up pattern covered in the article on tinnitus flare-ups. Both effects being real at the same time is a good illustration of why "does X help tinnitus" often needs a more specific answer than yes or no — the intensity, duration, and timeframe all matter.
What this means practically
- Regular, moderate exercise — the kind broadly recommended for general health — has real, if modest, supportive evidence for tinnitus specifically, beyond its well-established general health benefits
- If you notice tinnitus spikes specifically during or right after intense exercise, that's a recognized, plausible pattern, not something unusual — worth tracking using the same approach suggested in the flare-ups article, and worth mentioning to a doctor if it's a consistent, significant pattern, particularly to rule out a blood pressure component
- Light-to-moderate intensity exercise is a reasonable starting point if you're unsure how your own tinnitus responds, since negative reports rose with intensity in the survey — though if your tinnitus changes with jaw or neck movement, keep an eye on light sessions as well
- A whooshing or thumping that arrives during exertion is a different phenomenon from tinnitus that is generally worse afterwards, with a different workup — exertional tinnitus covers why it is assessed rather than trained through
The one risk factor that survived a causal test
Almost everything known about tinnitus risk is association: two things occur together, and nobody can say which drives which. Mendelian randomisation is the design built to get past that. It uses genetic variants people are born with as a stand-in for lifelong exposure — and because genes are allocated before any of the outcomes happen, a genetic association is far harder to explain away as reverse causation.
A 2026 Mendelian randomisation study in the Journal of International Medical Research applied it to 13 common life factors, using UK Biobank as a discovery cohort and FinnGen for validation.
In UK Biobank, four looked significant:
- Waist circumference — OR 1.37 (95% CI 1.09–1.71), p = 0.006
- Waist-to-hip ratio — OR 1.71 (1.17–2.50), p = 0.005
- Fasting insulin — OR 2.07 (1.07–4.01), p = 0.03
- Fasting glucose — OR 0.70 (0.54–0.90), p = 0.007
In FinnGen, a different one did: smoking, OR 1.20 (1.03–1.40), p = 0.018.
Combining the two cohorts, exactly one survived: waist circumference, OR 1.27 (1.08–1.49), p = 0.003.
That attrition is the most instructive thing in the study, and its authors say so — describing it as underscoring the critical role of meta-analysis in Mendelian randomisation. Four findings in one cohort, one in the other, one left standing at the end. It is a compact demonstration of why a single significant result, even from a well-designed study, is not yet a fact.
What remains is worth having precisely because it is a causal claim rather than an association: central obesity, measured at the waist, appears to be a modifiable target for tinnitus prevention. Note what that specifies — not weight in general, not BMI, but where it sits.
Be careful about what this does and does not license. It is not evidence that exercise prevents tinnitus, which no one has demonstrated. It is that the single lifestyle factor with causal evidence behind it happens to be one that physical activity acts on directly — which is a reason to take the rest of this article seriously, not a promise attached to it.
And a small trial that changed the thing the genetics pointed at
The Mendelian randomisation result says central obesity is causally linked to tinnitus. The obvious next question is whether reducing it helps — and a 2025 randomised controlled trial in the European Archives of Oto-Rhino-Laryngology tested something close to that.
It enrolled obese elderly people with metabolic syndrome and chronic subjective tinnitus, and ran a 12-week programme of diet restriction plus treadmill walking, with a study group of 30.
The study group improved on a long list: BMI, rated tinnitus severity, the components of metabolic syndrome including waist circumference, fasting glucose and insulin, blood pressure, triglycerides and HDL — and on both the Tinnitus Handicap Inventory and rated tinnitus-related discomfort. The control group did not improve on any of them.
What makes this worth including is the convergence, not the trial. Two methods that share none of the same weaknesses point at the same variable: a genetic design that cannot easily be confounded by lifestyle, and an intervention that directly reduced waist circumference and saw tinnitus move with it. That is a better reason to take the finding seriously than either on its own.
What makes it worth discounting is everything else. Thirty people in the study group. A narrow population — obese, elderly, with metabolic syndrome — which is not most people reading this. And the result is reported as significant change within the treated group against none in the control, rather than as a quantified difference between them, which is a pattern that overstates effects.
So: a promising, under-powered trial pointing the same way as a much stronger causal design. Not a prescription, and not nothing. If metabolic health is part of your own picture, diabetes and tinnitus covers the blood-sugar side of it.