Hearing aids and dementia risk: what the evidence shows

How to get your hearing checked in the UK

The NHS hearing-loss guidance advises seeing a GP for gradually worsening hearing or hearing that has not returned after treatment for an ear infection or wax build-up. The next step depends on the cause and may include referral to an audiologist.

Ask for an urgent GP appointment or NHS 111 advice for sudden hearing loss in either or both ears, deterioration over days or weeks, or hearing loss with earache or discharge. Do not put a sudden change on a routine shopping list for hearing aids or assume it is ordinary ageing.

For a routine assessment, describe the situations you want help with. Perhaps you hear a voice but miss words, follow one person in a quiet room but struggle at a meal, or find telephone conversations particularly difficult. Note when the change began and whether one ear seems worse. Those examples give the hearing professional more information than a general statement that everything is too quiet.

Tell the service if telephone booking or spoken instructions are difficult. Use an available accessible contact method or ask someone you trust to help arrange the appointment. You can also ask how information will be provided during and after the assessment so you are not relying on hearing every instruction first time.

What happens at a hearing assessment?

The NHS hearing-test page describes appointments lasting roughly 15 minutes to an hour, depending on the tests needed. In pure-tone audiometry, you respond to sounds through headphones. Speech audiometry uses speech material. Tympanometry checks eardrum movement and possible blockage behind it. You may have several tests, rather than necessarily all of them.

The specialist will discuss the results and whether further tests or treatment are needed. Ask them to relate the findings to your difficult situations: conversation in noise, a particular ear, television or work meetings. A test description becomes more useful when you understand which part of your experience it helps explain.

If aids are recommended, bring up practical handling needs before choosing. Small controls, batteries and fittings may be difficult with limited hand movement or vision. Ask to practise putting the aid in and taking it out, using the controls and cleaning it, rather than only watching a demonstration.

A companion can help recall examples or instructions if you want one there. Make clear which conversations matter to you and which tasks you need to manage yourself. The fitting should support your daily activities, including the unremarkable ones that are easy to forget to mention in a clinic.

Can you get hearing aids through the NHS?

NHS hearing-aid guidance states that a GP can refer you to an NHS provider. Aids are supplied free as a long-term loan, with free batteries, repairs, follow-up and aftercare. There can be a charge to replace an aid that is lost or damaged.

The aid is programmed for your hearing, with a follow-up appointment to check progress. Hearing aids can make speech and other sounds clearer and louder, but do not restore perfect hearing. Getting used to them may take weeks or months; contact the audiologist if there are problems rather than waiting indefinitely with an uncomfortable fit or unusable sound.

Before paying privately, ask for the full service and costs in writing: assessment, fitting, adjustments, batteries or charging equipment, repairs and replacement arrangements. Check what happens if the proposed aid does not meet the needs you discussed. A comparison based only on the small device in the display case can miss much of what you will use after purchase.

For follow-up, keep specific examples. 'The television is easier, but I still lose the conversation in our weekly meeting' gives the audiologist something to work on. Note whether the difficulty involves sound, comfort, handling or a particular environment, because those are different problems to discuss.

Why might hearing and brain health be connected?

The 2024 Lancet standing Commission includes hearing loss among potentially modifiable dementia risk factors. It reviews observational studies linking hearing loss with later dementia and finding lower risks among people using hearing aids.

Several explanations are possible. More mental effort may be spent decoding speech; difficulty communicating may reduce social participation; or vascular disease may affect both hearing and brain health. The report discusses these as possible pathways, rather than proof of one mechanism in every person.

Observational comparisons cannot settle whether hearing aids cause a difference. People who obtain and use aids may also differ in education, income, health behaviours and access to care. Those differences can influence later health. A randomised trial assigns the intervention rather than simply observing who already uses it, making the treatment question more directly testable.

That does not make an observational association worthless. It helps explain why the question deserves a trial, while keeping the association separate from what someone can expect after treatment. It also avoids treating hearing loss itself as a diagnosis of a future brain disease.

What did the main hearing-care trial find?

The ACHIEVE trial randomised 977 adults aged 70–84 with untreated hearing loss to hearing care or health education. Across three years, there was no significant overall difference in the combined cognitive-test score, which summarised performance across several tests.

A planned comparison examined recruitment groups. Among 238 participants from the ARIC cardiovascular-health study, hearing care was associated with 48% slower decline. Those participants were older, had more diabetes and hypertension and lower starting cognition; their control group declined 2.7 times faster than the newly recruited volunteers' controls. The 739 volunteers showed no significant treatment benefit.

'Planned' means specified before the results were known. This remained a subgroup analysis of cognitive-test changes, not a comparison of dementia diagnoses.

Care involved four one-hour audiologist visits, hearing aids in both ears fitted using real-ear measurements, device and communication instruction, and six-monthly booster visits. The comparator received health education. The trial tested ongoing care.

Why did a later headline say 62%?

A 2025 secondary analysis developed a cognitive-decline prediction model in a separate group of 2,692 people and applied it to the ACHIEVE participants. Among the quarter predicted to be at highest risk, decline over three years was estimated to be 61.6% slower with hearing care than with health education.

The 95% uncertainty interval was wide, from 33.7% to 94.1%. This concerned the rate of cognitive change in a selected group. It was not a 61.6% reduction in everyone's chance of developing dementia, and it was not an independent trial repeating ACHIEVE.

The 48% and 61.6% figures therefore answer related but different questions. The first concerns the existing cardiovascular-study recruitment group; the second uses a later risk model to identify a high-risk quarter. Seeing both percentages in headlines does not mean that two separate experiments established two levels of personal protection.

For someone deciding whether to seek hearing care, these results may add interest to the longer-term research. They cannot tell you that a particular model, price or fitting service will produce either percentage for you.

Did blood tests reveal a protective effect?

A September 2026 ACHIEVE analysis studied blood samples from 164 ARIC participants. It measured changes in GFAP, associated with neuroinflammatory processes, and NfL, associated with nerve-cell injury, over three years.

GFAP rose more slowly in the hearing-care group. The NfL difference was not statistically significant. These were non-specific biological markers in a subset of participants, rather than dementia diagnoses. The NIH-funded study's disclosures included industry relationships and donated hearing technologies.

This adds a possible biological clue for further investigation. It does not establish that a routine blood panel can show whether hearing aids are protecting someone's memory. A blood-marker change, a cognitive-test result and a clinical diagnosis answer different questions, so one should not silently replace another in the interpretation.

Make conversations easier while you wait

RNID's communication advice recommends getting attention before speaking, facing the person, keeping your face visible and reducing background noise. If repeating a sentence has not helped, try rephrasing. Written information can help with details such as names, times and numbers.

At a family meal, turn off the television and avoid several competing conversations. On a video call, use clear lighting and visible faces, with one speaker at a time; captions or chat can support details that are missed. Ask the person which adjustments help rather than assuming shouting or exaggerated lip movements will solve the problem.

These changes remain useful after fitting. The goal is participation, including situations where equipment alone cannot remove every difficulty. If conversation becomes easier but you still feel disconnected, our guide to feeling lonely around other people considers the quality of existing relationships and how to ask for the connection you need.

Difficulty hearing an instruction can look like forgetting it, but that does not explain every memory concern. If changes in memory affect daily life, raise them separately with a GP and mention hearing difficulties during assessment. Our everyday-memory guide covers practical support while keeping that clinical question distinct.