Yes, rheumatoid arthritis can cause hearing loss, though the link is underrecognized in routine care. The connection runs through two main routes: chronic systemic inflammation that damages the cochlea's blood supply, and direct joint damage in the temporomandibular joint or ossicular chain that blocks sound transmission. Published cohorts put sensorineural hearing loss prevalence in RA at roughly 20 to 60 percent, versus about 10 to 15 percent in age-matched adults without the disease. Most of that loss is mild and unnoticed at first, which is why screening remains inconsistent. The sections below separate mechanism from association and identify what monitoring actually catches.
How Rheumatoid Arthritis Reaches the Inner Ear
The cochlea depends on a fragile blood supply. The stria vascularis, which generates the endocochlear potential that drives hair cell function, is fed by a capillary network with fenestrated endothelial cells. That fenestration is useful for ion transport and also a weak point: circulating immune complexes, complement fragments, and activated platelets deposit there more easily than in vessels with tight junctions.
In RA, those complexes are plentiful. Rheumatoid factor and anti-citrullinated protein antibodies form circulating immune complexes that activate complement. The resulting microvasculitis narrows the cochlear microcirculation, and the stria vascularis is among the first structures to suffer. Ischemia lowers the endocochlear potential, outer hair cells lose their electromotility, and high-frequency thresholds rise first because the basal turn of the cochlea carries the highest metabolic demand.
Cytokines add a second hit. TNF-alpha, IL-6, and IL-1 beta cross the blood-labyrinth barrier, which is more permeable than the blood-brain barrier, and amplify local inflammation. In animal models, direct TNF-alpha application to the cochlea produces dose-dependent threshold shifts and hair cell loss. Human evidence is correlational rather than causal, but it fits the clinical observation that hearing worsens during disease flares.
In plain terms: the same inflammation that attacks joints can quietly starve and inflame the sound-sensing organ.
A third route is mechanical. The incudostapedial and incudomalleolar joints are true synovial joints, and the temporomandibular joint sits centimeters from the middle ear. Pannus formation in these joints is uncommon but documented, and it can produce a conductive deficit by stiffening the ossicular chain or by disrupting eustachian tube function. TMJ involvement on imaging is common in RA, reported in up to 40 to 80 percent of patients in imaging series, while clinically meaningful conductive hearing loss is far rarer.
Autoantibody cross-reactivity is the most speculative layer. Antibodies against type II collagen and the 60 kDa heat shock protein appear in some patients with autoimmune inner ear disease, and collagen II is a component of the tectorial membrane. Whether the same antibodies drive RA-related hearing loss is a hypothesis, not an established mechanism. It is worth knowing because it explains the minority of patients with rapidly progressive, bilateral sensorineural loss that behaves like autoimmune inner ear disease.
Can Rheumatoid Arthritis Cause Hearing Loss? Quantifying the Risk
The direction of the association is consistent across studies. The magnitude is not. Prevalence estimates for sensorineural hearing loss in RA range from roughly 20 to 60 percent, depending on the audiometric definition, the age of the cohort, and whether extended high frequencies are tested. Age-matched controls in the same studies typically sit at 10 to 15 percent.
Pooled analyses land on roughly a doubling of risk. A commonly cited range is 1.7 to 2.5 times the odds of hearing loss in RA versus controls after adjustment for age and sex. Adjustment for noise exposure, smoking, and diabetes moves the estimate by perhaps 10 to 20 percent, which is not enough to close the gap.
The most informative data are subclinical. Standard audiometry tests 0.25 to 8 kHz. Extended high-frequency audiometry tests 9 to 20 kHz, and that is where RA shows up first. Several cohorts describe RA patients with normal standard audiograms who nonetheless have thresholds 5 to 15 dB worse than controls at 9 to 16 kHz. Otoacoustic emissions, which measure outer hair cell function directly, are reduced in the same group.
Conductive hearing loss is less common and less well characterized. Reported prevalence ranges from about 5 to 15 percent, and the mechanisms are heterogeneous: ossicular joint involvement, TMJ-related eustachian tube dysfunction, and middle ear effusion. Tympanometry abnormalities are frequent but nonspecific, and many studies do not separate conductive from mixed patterns.
One reconciliation matters for reading your own audiogram. The WHO 2021 World Report on Hearing grades mild loss at 26 to 40 dB HL, moderate at 41 to 60 dB HL, and severe at 61 to 80 dB HL. A 5 to 10 dB shift is real but stays inside the normal range until the threshold crosses 26 dB HL. In plain terms, a measurable RA effect can coexist with a report that reads "normal hearing."
Subtypes, Flares, and Edge Cases
Not every RA patient carries the same hearing risk. Seropositive disease, meaning positive rheumatoid factor or anti-CCP, is associated with higher extra-articular burden, and extra-articular manifestations track with hearing loss in most cohorts. Nodular disease, vasculitis, and long disease duration point the same direction. Seronegative patients are not exempt, but their average risk is lower and their hearing loss is more often explained by age or noise.
Secondary Sjögren's syndrome is a distinct amplifier. Roughly 10 to 30 percent of RA patients meet criteria for secondary Sjögren's, and Sjögren's is independently associated with sensorineural hearing loss and autoimmune inner ear disease. When both diagnoses are present, attributing hearing loss to RA alone is a mistake and changes the treatment conversation.
Flares produce reversible threshold shifts. During active inflammation, thresholds can worsen by 5 to 10 dB and partially recover with disease control. This is why a single audiogram drawn during a flare can overstate permanent damage, and why repeat testing after treatment is more informative than a one-time snapshot.
Medication-related loss is a separate category with its own pattern. High-dose aspirin above roughly 2 g per day causes bilateral, symmetric, high-frequency sensorineural loss and tinnitus that usually resolves within a week of stopping. Hydroxychloroquine ototoxicity is rare and mostly tied to long cumulative doses. Methotrexate and TNF inhibitors are not classically ototoxic, though isolated case reports exist.
| Pattern | Underlying process | Typical audiogram | Reported share in RA cohorts | Clinical clue |
|---|---|---|---|---|
| Sensorineural | Immune-mediated cochlear microangiopathy | Downsloping high-frequency loss | 20 to 60 percent | Gradual, bilateral, tinnitus common |
| Conductive | Ossicular joint pannus, eustachian tube dysfunction | Air-bone gap with normal bone conduction | 5 to 15 percent | Aural fullness, normal inner ear function |
| Mixed | Both processes in long-standing disease | Air-bone gap plus high-frequency loss | 5 to 10 percent | Disease duration above 10 years |
| Medication-related | Salicylate or antimalarial toxicity | Symmetric high-frequency loss | Rare, dose-dependent | Temporal link to drug start |
| Sudden sensorineural | Possible vasculitis or autoimmune inner ear disease | Unilateral sudden loss | Under 1 percent but urgent | Needs same-week evaluation |
A Timeline of Hearing Change in Rheumatoid Arthritis
The table below synthesizes cross-sectional cohorts into an approximate time course. It is a framework, not a prognosis for any individual patient, because no single longitudinal cohort has tracked audiograms from diagnosis for 20 years.
| Disease stage | Typical audiometric finding | Approximate share affected |
|---|---|---|
| Year 0 to 1 (diagnosis) | Standard audiogram often normal; high-frequency dips in a minority | 5 to 15 percent |
| Years 1 to 3 | Subtle extended high-frequency loss, usually unnoticed | 15 to 25 percent |
| Years 3 to 5 | Detectable sensorineural loss; tinnitus more common | 20 to 35 percent |
| Years 5 to 10 | Cumulative sensorineural loss; mixed patterns in TMJ disease | 30 to 45 percent |
| Years 10 to 15 | Compounded by age-related presbycusis after 50 | 40 to 55 percent |
| Years 15 and beyond | More than half affected; severity tracks cumulative inflammation | Above 50 percent |
| Flare windows | Transient 5 to 10 dB shifts, partly reversible | Variable |
The ranges overlap by design. A 55-year-old with 12 years of seropositive disease and a noisy job may sit at the top of every row, while a 35-year-old in sustained remission may sit at the bottom. Disease activity, not calendar time alone, drives the slope.
Four Misconceptions That Delay Diagnosis
The first misconception is that hearing loss in RA is just aging or noise exposure. Adjusted analyses still show elevated risk after age matching, and presbycusis alone does not produce a two-fold excess. Noise history matters, but it does not absorb the RA signal.
The second is that the medications must be responsible. Salicylates are the main ototoxic culprit in this population, and that loss is largely reversible. Most RA-related hearing loss tracks with disease activity rather than drug exposure, so stopping a DMARD on suspicion alone usually removes protection without fixing the ear.
The third is that a normal audiogram means normal hearing. Standard audiometry stops at 8 kHz, while RA changes appear at 9 to 16 kHz and on otoacoustic emissions years earlier. A normal standard test is reassuring about conversational hearing, not about cochlear health.
The fourth is that RA-related hearing loss is always gradual and bilateral. Sudden unilateral sensorineural loss is a different process, often vasculitic or autoimmune, and it is time-critical. It deserves the same urgency as sudden vision loss.
Monitoring Hearing When You Have Rheumatoid Arthritis
A baseline audiogram at diagnosis or within the first year is the single most useful step. It creates a reference point against which later shifts can be judged. If extended high-frequency testing is available, request it, because it detects change 2 to 5 years earlier than standard audiometry in most RA cohorts.
Repeat testing every 1 to 2 years makes sense for most patients. Go annual if you have known high-frequency dips, take chronic salicylates, or work in noisy environments. A 10-item self-report screen such as the Hearing Handicap Inventory for the Elderly–Screening version takes about 3 minutes and catches functional decline that pure-tone thresholds miss.
Disease control is hearing protection. Treat-to-target strategies per the 2022 EULAR recommendations aim at remission or low disease activity, and cohorts with sustained low activity show slower progression of high-frequency loss. That is association rather than proof, but it is the best lever currently available.
Review ototoxic exposures at every visit. Keep chronic aspirin below 2 g per day when possible, and ask about cumulative hydroxychloroquine dose if hearing is declining. Smoking cessation and noise avoidance both matter, because each multiplies the RA effect rather than adding to it.
Seek same-week evaluation for any of the following red flags:
- Sudden hearing loss in one ear developing over 72 hours or less
- New severe tinnitus accompanied by vertigo or imbalance
- Hearing loss with facial weakness or numbness
- Rapidly progressive loss over weeks, especially when bilateral
- Aural fullness with fever, drainage, or severe pain
- New hearing loss after starting a new medication
What the Current Evidence Cannot Tell You
Nearly all human data are cross-sectional. That design cannot establish that RA causes hearing loss in any individual; it establishes that the two travel together more often than chance predicts. Causality is inferred from mechanism, not demonstrated by cohort structure.
Confounding is real and incompletely handled. Age, noise, smoking, diabetes, and socioeconomic status all affect hearing, and most studies adjust for only some of them. Residual confounding could shrink the true effect by a meaningful margin, though probably not to zero.
Definitions vary enough to matter. Some studies define hearing loss as a threshold above 25 dB at any frequency; others use a pure-tone average above 25 dB in the better ear. The WHO 2021 scale sets the mild threshold at 26 dB HL, and that single-decibel difference shifts prevalence estimates by several percentage points across studies.
No randomized trial has tested whether DMARDs prevent hearing loss. The hypothesis that early, sustained inflammation control preserves cochlear function is plausible and unproven. Until such a trial exists, the recommendation to control disease activity rests on general RA benefits rather than hearing-specific evidence.
The autoimmune inner ear disease overlap remains a hypothesis as well. It is clinically useful because it justifies a trial of immunosuppression in rapidly progressive cases, but it is not established pathophysiology. Patients with sudden or rapidly progressive loss should be evaluated by both rheumatology and otolaryngology rather than managed on the assumption that RA explains everything.