What a major hearing aid trial found about memory and thinking

Your father has started answering questions nobody asked. At dinner he laughs a beat late, or not at all, and he has stopped calling his brother because, he says, everyone on the phone mumbles. Then a headline crosses your feed saying hearing aids may protect the brain, and an over-the-counter pair is a click away.

It is tempting to buy them and call it dementia prevention. The evidence supports something narrower and, in its way, more useful.

Hearing loss and cognitive decline travel together in long-running studies. When researchers finally tested hearing treatment in a large randomized trial, though, the main result was a tie. A benefit appeared in one higher-risk subgroup, and nowhere in the trial did treatment reduce diagnoses of cognitive impairment. Meanwhile, since October 2022, adults have been able to buy hearing aids without seeing anyone at all. Here is what the research shows, where it stops, and when a store shelf is the wrong first stop.

Why researchers suspected hearing in the first place

In 2011, Frank R. Lin of Johns Hopkins and colleagues published an analysis from the Baltimore Longitudinal Study of Aging. They followed 639 people whose hearing had been measured in the early 1990s, all free of dementia at the start. Over a median of almost 12 years, 58 developed dementia.

Risk climbed with the degree of loss. Compared with normal hearing, mild loss carried about 1.9 times the risk, a borderline result whose confidence range just touched no effect. Moderate loss carried 3 times the risk and severe loss nearly 5 times, though the severe group held only six people. Lin’s team ended with a question rather than a conclusion: “Whether hearing loss is a marker for early-stage dementia or is a modifiable risk factor for dementia deserves further study.”

That question still frames the field. Researchers have proposed three routes from the ears to the mind: the extra mental effort of decoding muffled speech, effects of reduced sound input on brain structure, and a slow retreat from conversation and activity. All three are proposed mechanisms, not demonstrated outcomes.

The Lancet’s standing commission on dementia, in its 2024 report, counts hearing loss as a midlife risk factor with a relative risk of about 1.4. It attributes roughly 7 percent of dementia worldwide to hearing loss, a modeling figure that assumes the link is causal. Writing in 2023, ACHIEVE’s investigators noted that earlier Commission reports had called hearing loss the single largest modifiable factor. In the 2024 table, high LDL cholesterol in midlife rounds to the same 7 percent. The Commission’s other 12 factors are taken apart in our explainer on dementia risks most people miss.

The exposure is common. The National Institute on Deafness and Other Communication Disorders (NIDCD), part of NIH, reports disabling hearing loss in 55 percent of Americans 75 and older, based on 2015 to 2020 national survey data. Among Americans 71 and older living with dementia, a 2024 Johns Hopkins analysis of 2021 national data found hearing loss on testing in about 79 percent, and only about 22 percent of those reported using hearing aids.

The observational promise, and its catch

A 2023 meta-analysis in JAMA Neurology, from the National University of Singapore, pooled eight long-term studies with 126,903 participants. Hearing aid users had a 19 percent lower hazard of cognitive decline than people whose loss went uncorrected. In shorter studies, cognitive test scores ran about 3 percent higher after people began using hearing aids.

The catch is who wears hearing aids. People who get fitted tend to differ from people who don’t in ways that also protect thinking. ACHIEVE’s authors spelled this out: measured factors such as education and income, and unmeasured ones such as health habits, can produce the appearance of a benefit. Random assignment is the standard way to strip that out.

Inside the ACHIEVE trial

The Aging and Cognitive Health Evaluation in Elders study, known as ACHIEVE, was funded by the National Institutes of Health and published in The Lancet in July 2023. It enrolled 977 people aged 70 to 84 in four US communities: Forsyth County, North Carolina; Jackson, Mississippi; the Minneapolis suburbs; and Washington County, Maryland. Everyone had untreated hearing loss, no substantial cognitive impairment, and a willingness to wear hearing aids.

Half received the hearing intervention: four hour-long sessions with an audiologist, two hearing aids fitted to their loss, accessories for phones and television, coaching on communication strategies, and refresher visits twice a year. The other half received a health education program on preventing chronic disease and disability, built to match the time and attention.

Participants came from two pools. Some 238 were longtime members of the Atherosclerosis Risk in Communities (ARIC) study, a heart-health cohort running since the late 1980s. The other 739 were new community volunteers. Compared with the volunteers, the ARIC group was older and more often Black, had less education and income, had more diabetes and high blood pressure, more often lived alone, and started with lower cognitive scores.

The main result: a tie

After three years, both groups had slipped by the same amount, 0.200 and 0.202 standard deviations on a combined thinking score. The statistical test returned a p value of 0.96, about as close to no difference as a trial gets. The trial’s published abstract in The Lancet puts it plainly: “The hearing intervention did not reduce 3-year cognitive decline in the primary analysis of the total cohort.”

As a hearing treatment, the intervention worked. By year three, participants wore their aids an average of 7.2 hours a day, and their scores on a questionnaire about everyday communication difficulty fell by roughly half. The control group’s scores edged up.

The subgroup that drew the headlines

A sensitivity analysis, planned before the results were known, split participants by where they were recruited. The two pools responded differently. The 238 people drawn from ARIC lost ground 48 percent more slowly if they got hearing treatment. The newly recruited volunteers showed no such difference.

The authors’ leading explanation is that the volunteers were declining slowly to begin with. ARIC controls lost ground 2.7 times faster than volunteer controls, and a treatment can only slow a decline that is happening. The authors also raised a healthy-volunteer effect, since people who answer an ad for a new study tend to be healthier. They raised practice effects too: the volunteers were taking these tests for the first time and may have improved with repetition, while ARIC members had taken such tests many times before.

Keep the subgroup in proportion. It held 238 people, and the confidence interval for the difference ran from 0.022 to 0.360 standard deviations, a range from almost nothing to substantial. A 2025 secondary analysis in Alzheimer’s & Dementia ranked all 977 participants with a statistical risk model and found the steepest benefit among those predicted to decline fastest. That strengthens the hypothesis that hearing care matters most for people already on a faster path. As a model-based look back at the same trial, it cannot confirm it.

What did not change

The trial also tracked a harder outcome: new cognitive impairment, defined as adjudicated mild cognitive impairment, dementia, or a meaningful drop on a screening test. Hearing treatment made no significant difference overall, with a hazard ratio of 0.90 and a range from 0.61 to 1.33, or in either pool. The authors note that three years may be too short to see it, and a follow-up study of the same participants is underway.

Limits that cut both ways

• No blinding. Participants and testers could see who wore hearing aids.
• Spoken-only tests. Two of the 10 cognitive tests used only spoken material, so untreated controls might score lower simply from mishearing. The authors point out that the strongest ARIC effect was in language tests that did not rely on spoken-only items.
• Controls who bought their own aids. About 16.5 percent of the control group got hearing aids on their own during the trial, which would blur any true difference.

Secondary findings, labeled as such

Three later analyses of ACHIEVE drew attention. Each is secondary, so each is a lead rather than a verdict.

• Falls. In a prespecified exploratory analysis in Lancet Public Health in 2025, the hearing group averaged 1.45 falls over three years against 1.98 for controls, about 27 percent fewer, consistent in both pools.
• Social networks. A 2025 analysis in JAMA Internal Medicine found the hearing group kept about one more person in their social circle over three years. The investigators said they did not know whether that difference was clinically meaningful, and that the loneliness changes were not. The wider evidence on social ties and thinking is weighed in our article on isolation and dementia.
• Blood markers. A 2026 analysis in eClinicalMedicine of 164 ARIC participants found a slower rise in GFAP, a blood protein associated with brain inflammation, with hearing treatment. A second marker, of nerve-cell damage, did not differ significantly. A marker in the blood is not a memory you get back.

A low test score can be a hearing problem

Most cognitive screening happens by voice. A 2015 study of 301 older adults by University of Toronto researchers found that people with hearing or vision loss were less likely to pass the Montreal Cognitive Assessment, a common 30-point screen, even after scoring adjustments meant to allow for sensory problems. The authors warned that cognitive ability can be underestimated when sensory loss is overlooked. ACHIEVE’s designers took steps so hearing loss would not skew its testing, and kept spoken-only tests to two of ten.

So whether a disappointing score reflects hearing rather than memory is a fair question for whoever gave the test, including whether hearing was checked first. How cognitive screens are built and scored belongs to a different piece in this series.

Over-the-counter hearing aids: what changed in 2022

The Food and Drug Administration published its over-the-counter rule on August 17, 2022, and the category took effect on October 17, 2022. The regulation, at 21 CFR 800.30, covers air-conduction hearing aids that need no surgery, for people 18 and older with perceived mild to moderate hearing impairment, sold without the supervision, prescription or involvement of a licensed professional. It also caps how loud the devices can get.

Perceived means you judge your own loss. No test is required to buy one. According to the FDA, the devices are not meant for severe or profound loss, and anyone under 18 needs a prescription. FDA rules do not require a warranty, and the agency urges buyers to understand the return or warranty window so there is time to try the device.

What does the research say about fitting them yourself? A 2026 meta-analysis in Otolaryngology-Head and Neck Surgery pooled 15 studies with 774 participants, mostly short trials of 10 days to 8 weeks. Self-fitted devices produced results comparable to professional fitting, measured largely by participants’ own reports. The authors rated the certainty of that evidence low.

Real users describe friction. In a 2026 study in the Journal of the American Academy of Audiology, 78 adults 55 and older recruited by San Diego State University researchers tried over-the-counter devices. Many reported trouble with setup and handling, and they wanted more support. In a 2026 survey of 1,013 US adults 50 and older with no hearing aid experience, 39 percent said they were comfortable buying over the counter, yet 84 percent preferred care led by a provider, and insurance coverage predicted which way people leaned. In October 2026, two University of California, San Francisco researchers titled a JAMA commentary with the argument that regulatory reform by itself has not been enough to guarantee benefits for consumers.

Here is the trade-off stated plainly. Over-the-counter devices remove appointments and gatekeepers. ACHIEVE, however, tested a full package of audiologist care, counseling and follow-up, so even its subgroup result says nothing direct about a device fitted alone at the kitchen table.

When a clinician comes before the store

On its over-the-counter hearing aid page, the FDA lists problems that should send a person to a doctor, preferably an ear, nose and throat specialist. Grouped by what they signal:

• Something changed fast or unevenly. Hearing dropped suddenly within the past six months; one ear is clearly worse; ringing or buzzing is in only one ear; or hearing worsens and then improves again.
• Something is wrong with the ear itself. Pain or discomfort; blood, pus or fluid draining within the past six months; a lot of wax, or something possibly stuck inside; or an ear shaped unusually from birth or injury.
• Balance is involved. Strong dizziness or a spinning or swaying sensation, called vertigo.

The FDA separately points people toward a hearing professional if they cannot follow speech even in a quiet room, or if loud sounds such as engines or power tools do not come through well. Those signs may point to loss beyond the mild-to-moderate range the devices are built for.

Sudden loss is the urgent case. NIDCD’s page on sudden sensorineural hearing loss, last updated in 2018, calls it a medical emergency that warrants seeing a doctor immediately. It often strikes one ear, and treatment delayed more than two to four weeks is less likely to reverse it.

For gradual loss, NIDCD describes the main players. Otolaryngologists are physicians who diagnose and treat ear conditions. Audiologists are not physicians; they test hearing, identify the type and degree of loss, and are licensed to fit hearing aids. Hearing instrument specialists are state-licensed to perform basic hearing tests and fit and dispense aids.

Who pays for what

Original Medicare does not cover hearing aids or the exams to fit them, according to Medicare.gov. The person pays all of those costs. Some Medicare Advantage plans offer hearing benefits that Original Medicare lacks; the details vary by plan.

Part B does cover diagnostic hearing and balance exams when a clinician orders them. Once every 12 months, a Medicare beneficiary can see an audiologist without an order for a non-acute problem, such as hearing loss that developed over years. After the Part B deductible, the person pays 20 percent of the Medicare-approved amount, plus a copayment in a hospital outpatient setting.

The US Preventive Services Task Force, in its 2021 statement, concluded that evidence was insufficient to weigh the benefits and harms of screening adults 50 and older who have no symptoms. That statement does not apply to people who notice trouble hearing. Its evidence review also found no study that had tested whether hearing treatment changed rates of dementia. ACHIEVE came later and still did not show fewer diagnoses.

Where the evidence stands

Here is each claim, sorted into the library’s evidence verdicts:

• Hearing loss goes with faster decline and more dementia. Positive evidence, contested interpretation. Whether it is a cause or an early marker is still open.
• Hearing treatment slows cognitive decline in older adults generally. Tested and not supported over three years in ACHIEVE’s primary analysis.
• Hearing treatment slows decline in higher-risk older adults. Positive evidence, contested interpretation. It rests on a planned subgroup of 238 and a model-based reanalysis, and needs its own trial.
• Hearing aids prevent dementia. Insufficient. No trial has shown it, and ACHIEVE found no drop in new cognitive impairment.
• A guideline recommends hearing aids to prevent dementia. None of the US bodies cited here, including the Preventive Services Task Force, does. The Lancet Commission, which is not a clinical guideline, says hearing aids should be accessible to people with hearing loss.

The most useful takeaway

The strongest case for hearing care is hearing: following the dinner conversation, catching the punchline, keeping the phone calls going. ACHIEVE showed that a full audiology package delivers that, and its secondary findings on falls and social ties are encouraging leads. What it did not show is protection against dementia for older adults in general. If the change in your parent’s hearing is sudden, lopsided or painful, it belongs with a doctor first, and a sudden drop belongs there quickly. If it has crept up over years, Part B’s once-a-year audiologist visit, available without an order, is one route to learning what kind of loss it is before money goes to a device.

Sources

1. Archives of Neurology. “Lin FR, Metter EJ, O'Brien RJ, Resnick SM, Zonderman AB, Ferrucci L. "Hearing loss and incident dementia." Arch Neurol, 2011. PubMed. DOI: 10.1001/archneurol.2010.362.” pubmed.ncbi.nlm.nih.gov
2. University College London (UCL Discovery), accepted manuscript of The Lancet, 2024;404(10452):572-628. DOI: 10.1016/S0140-6736(24)01296-0. “Livingston G, Huntley J, Liu KY, et al. "Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission" (accepted version, text).” discovery.ucl.ac.uk
3. University College London (UCL Discovery), accepted manuscript table file. “Livingston G, Huntley J, Liu KY, et al. "Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission," Table 1 (relative risks and population attributable fractions).” discovery.ucl.ac.uk
4. National Institute on Deafness and Other Communication Disorders, National Institutes of Health. “Quick Statistics About Hearing, Balance, & Dizziness (last updated September 20, 2024).” nidcd.nih.gov
5. JAMA Network Open. “Nieman CL, Garcia Morales EE, Huang AR, Reed NS, Yasar S, Oh ES. "Prevalence of Hearing Loss and Hearing Aid Use Among Persons Living With Dementia in the US." JAMA Netw Open, 2024. PubMed. DOI: 10.1001/jamanetworkopen.2024.40400.” pubmed.ncbi.nlm.nih.gov
6. JAMA Neurology. “Yeo BSY, Song HJJMD, Toh EMS, et al. "Association of Hearing Aids and Cochlear Implants With Cognitive Decline and Dementia: A Systematic Review and Meta-analysis." JAMA Neurol, 2023. PubMed. DOI: 10.1001/jamaneurol.2022.4427.” pubmed.ncbi.nlm.nih.gov
7. The Lancet. “Lin FR, Pike JR, Albert MS, et al. "Hearing intervention versus health education control to reduce cognitive decline in older adults with hearing loss in the USA (ACHIEVE): a multicentre, randomised controlled trial." Lancet, 2023. PubMed. DOI: 10.1016/S0140-6736(23)01406-X.” pubmed.ncbi.nlm.nih.gov
8. Alzheimer's & Dementia. “Pike JR, Huang AR, Reed NS, et al. "Cognitive benefits of hearing intervention vary by risk of cognitive decline: A secondary analysis of the ACHIEVE trial." Alzheimers Dement, 2025. PubMed. DOI: 10.1002/alz.70156.” pubmed.ncbi.nlm.nih.gov
9. The Lancet Public Health. “Goman AM, Tan N, Pike JR, et al. "Effects of hearing intervention on falls in older adults: findings from a secondary analysis of the ACHIEVE randomised controlled trial." Lancet Public Health, 2025. PubMed. DOI: 10.1016/S2468-2667(25)00088-X.” pubmed.ncbi.nlm.nih.gov
10. JAMA Internal Medicine. “Reed NS, Chen J, Huang AR, et al. "Hearing Intervention, Social Isolation, and Loneliness: A Secondary Analysis of the ACHIEVE Randomized Clinical Trial." JAMA Intern Med, 2025. PubMed. DOI: 10.1001/jamainternmed.2025.1140.” pubmed.ncbi.nlm.nih.gov
11. eClinicalMedicine. “Pike JR, Huang AR, Reed N, et al. "Effects of hearing intervention on blood-based biomarkers of neuroinflammation and neurodegeneration: a secondary analysis of the ACHIEVE randomised controlled trial." EClinicalMedicine, 2026. PubMed. DOI: 10.1016/j.eclinm.2026.104164.” pubmed.ncbi.nlm.nih.gov
12. Aging, Neuropsychology, and Cognition. “Dupuis K, Pichora-Fuller MK, Chasteen AL, Marchuk V, Singh G, Smith SL. "Effects of hearing and vision impairments on the Montreal Cognitive Assessment." Neuropsychol Dev Cogn B Aging Neuropsychol Cogn, 2015. PubMed. DOI: 10.1080/13825585.2014.968084.” pubmed.ncbi.nlm.nih.gov
13. US Food and Drug Administration. “OTC Hearing Aids: What You Should Know (content current as of 05/03/2023).” fda.gov
14. Legal Information Institute, Cornell Law School (text of the Code of Federal Regulations). “21 CFR 800.30, Over-the-counter hearing aid controls (87 FR 50748, Aug. 17, 2022).” law.cornell.edu
15. Otolaryngology-Head and Neck Surgery. “De Sousa KC, Almufarrij I, Kruger M, Manchaiah V, Munro KJ, Swanepoel DW. "Effectiveness of Over-the-Counter Hearing Aids Versus Professionally Fitted Devices: A Systematic Review and Meta-analysis." Otolaryngol Head Neck Surg, 2026. PubMed. DOI: 10.1002/ohn.70306.” pubmed.ncbi.nlm.nih.gov
16. Journal of the American Academy of Audiology. “Coco L, Weeks T, Rodriguez C, et al. "Perceptions of Over-the-Counter Hearing Aids: A Qualitative Study Informed by the Technology Acceptance Model." J Am Acad Audiol, 2026. PubMed. DOI: 10.3766/jaaa.250056.” pubmed.ncbi.nlm.nih.gov
17. American Journal of Audiology. “Singh J, Dhar S. "Understanding Hearing Health Care Choices and Customer Archetypes in an Era of Over-the-Counter Hearing Aids." Am J Audiol, 2026. PubMed. DOI: 10.1044/2026_AJA-26-00032.” pubmed.ncbi.nlm.nih.gov
18. JAMA. “Phillips KA, Cassel CK. "Over-the-Counter Hearing Aids 4 Years Later: Regulatory Reform Alone Is Insufficient to Ensure Consumer Benefits." JAMA, 2026. PubMed. DOI: 10.1001/jama.2026.19796.” pubmed.ncbi.nlm.nih.gov
19. National Institute on Deafness and Other Communication Disorders, National Institutes of Health. “Sudden Sensorineural Hearing Loss (last updated September 14, 2018).” nidcd.nih.gov
20. National Institute on Deafness and Other Communication Disorders, National Institutes of Health. “Who Can I Turn to for Help with My Hearing Loss? (last updated October 24, 2024).” nidcd.nih.gov
21. Centers for Medicare & Medicaid Services (Medicare.gov). “Hearing aids.” medicare.gov
22. Centers for Medicare & Medicaid Services (Medicare.gov). “Hearing & balance exams.” medicare.gov
23. US Preventive Services Task Force. “Hearing Loss in Older Adults: Screening (final recommendation, March 23, 2021).” uspreventiveservicestaskforce.org