Diabetes is linked to tinnitus, but modestly: a 2025 meta-analysis of 2,277,719 people found about 18% higher odds, and UK Biobank data associate type 2 diabetes with constant and upsetting tinnitus. Both are observational, so they show that the two travel together, not that one causes the other — and the far larger percentages quoted elsewhere come from much more selected groups of people.
The number, drawn from 2.3 million people
A 2025 meta-analysis in Biomolecules & Biomedicine pooled 12 observational studies covering 2,277,719 participants and found diabetes associated with tinnitus at an odds ratio of 1.18 (95% CI 1.06–1.31, P = .002), with moderate heterogeneity (I² = 51%).
By the standards of observational evidence this one is unusually well behaved. Sensitivity analyses held. Egger's test found no sign of publication bias (P = .29). Subgroup analyses found no significant differences by geographical region, mean age, sex distribution, how tinnitus was diagnosed, or the statistical model used — meaning the result does not hinge on any one of those choices.
And the number it produces is small. An odds ratio of 1.18 is roughly 18% higher odds — not a doubling, not a transformation of risk. That figure is worth carrying into the rest of this article, because the individual studies below report percentages that sound far more dramatic, and were measured in far more selected groups of people.
What the UK Biobank data shows
A 2026 analysis in Acta Diabetologica used UK Biobank data from 136,181 people to look beyond whether someone has tinnitus, at how often it is there and how much it bothers them. The paper describes itself as a population-based cohort study, but the analysis was cross-sectional: it compared diabetes and tinnitus at the same point in time rather than following people forward.
- Type 2 diabetes was associated with higher odds of having tinnitus now, of constant tinnitus (odds ratio 2.49, 95% CI 1.44–4.31), and of upsetting tinnitus (6.36, 95% CI 2.3–17.54).
- Insulin use was associated with lower odds of current and of upsetting tinnitus than taking no glucose-lowering medication (0.61 and 0.42).
- Diabetes status and HbA1c were associated with tinnitus mainly in younger women.
Read those carefully. The interval around 6.36 runs from 2.3 to 17.54, which means the estimate is imprecise. The insulin finding is an association in a snapshot, not evidence that insulin protects the ear. And a published letter on the study raised two further cautions: diabetes type was assigned by an algorithm without confirmatory tests such as autoantibodies or C-peptide, and age was treated as a simple adjustment across the 40 to 70 age range, as if the association were the same at every age. Its authors call the data hypothesis-generating, to be interpreted cautiously until confirmed.
What the UK Biobank data adds, then, is a direction worth following — an association with how constant and how upsetting tinnitus is, not only with whether it is present — rather than a settled answer.
The pattern smaller studies had already suggested
Before these larger analyses, smaller studies had pointed the same way, with much bigger numbers. A Hungarian study compared 103 people with type 2 diabetes seen at a university clinic against 589 people with type 2 diabetes and 15,622 without diabetes in a national health screening programme. It found hearing impairment and/or tinnitus in 80% of the clinic patients, 34% of the diabetics in the screening programme, and 14% of the non-diabetics. A study of 250 patients at a diabetes centre in Iran found tinnitus in 66 of them (26.4%).
Those percentages are exactly why the pooled odds ratio matters. A diabetes clinic sees the people who came in with a problem; a national screening programme sees everyone. The 80% figure describes the first group, and the OR of 1.18 describes something much closer to the second. When a statistic about your own risk sounds alarming, the first question to ask is which population it was measured in — and the largest, least selected measurement available here returns a modest number.
Why diabetes might affect the ear
The proposed mechanism centers on the inner ear's blood supply and neural tissue, both of which are sensitive to the kind of small-vessel damage (microvascular disease) diabetes is well known for causing elsewhere in the body — the same underlying process implicated in diabetic retinopathy (eye damage) and neuropathy (nerve damage). Diabetes-related hearing research has also documented structural changes specific to the inner ear, including thickening of small blood vessel walls in the cochlea and changes to the auditory nerve, which would plausibly affect both hearing and the auditory signal processing implicated in tinnitus. The Iranian study's authors suggest tinnitus in diabetes may be one indicator of neuropathy or of small-vessel disease in the inner ear.
Does blood sugar control matter?
The evidence here is mixed rather than settled. In UK Biobank, HbA1c — the standard measure of long-term blood sugar control — was associated with tinnitus, mainly in younger women. The Hungarian study found no significant correlation between HbA1c and hearing impairment or tinnitus. The Iranian study found no link between tinnitus and either fasting blood sugar or HbA1c, though it did link longer diabetes duration with tinnitus, and found hearing problems more severe in people who had had diabetes for more than ten years.
Good blood sugar control is worth pursuing for its established effects on diabetes's other complications. Whether it changes tinnitus specifically has not been shown consistently, and it would be overstating the evidence to present tinnitus as a gauge of how well your diabetes is controlled.
What this means practically
If you have diabetes — particularly if you've had the condition for many years — a baseline hearing test is a reasonable thing to add to your regular diabetes care, even without noticing an obvious hearing problem yet, given how easy gradual hearing change is to miss subjectively (covered in more depth in the article on tinnitus and hearing loss). If tinnitus is already present, mention it to whoever manages your diabetes care as well as to an audiologist, so it is part of the same picture. If diabetes has also affected your kidneys, kidney disease and tinnitus sets out what the evidence shows there. If it began after starting a new diabetes drug, GLP-1 drugs and tinnitus covers what is known about one widely used class.