Hearing Aids May Slow Cognitive Decline in Higher-Risk Older Adults
A major trial found no overall cognitive benefit from hearing care, but subgroup results suggest some older adults may benefit. New biomarker findings add context.
Table of Contents
DISCLAIMER
This article is for informational purposes only and does not constitute medical advice. The statements in this article have not been evaluated by the FDA. The information presented is based on published research and should not be used as a substitute for professional medical guidance. Consult your physician before starting any supplement or health protocol.
The overall result deserves the first sentence
Hearing treatment may help preserve thinking abilities in some older adults, but the ACHIEVE trial did not establish a cognitive benefit across its full study population. That distinction should frame any claim that hearing aids prevent dementia. The trial’s main result and its more encouraging subgroup findings address different questions; neither should disappear when the findings become a headline.
According to the 2023 ACHIEVE report in The Lancet, investigators randomly assigned 977 adults aged 70–84 with untreated hearing loss and without substantial cognitive impairment to hearing intervention or health education. Over three years, the difference in change on the main cognitive measure was approximately zero: 0.002 standard-deviation units, with a 95% confidence interval from −0.077 to 0.081.
The same report suggested a more favorable result among 238 participants recruited from a longstanding cardiovascular study. In that subgroup, hearing intervention was associated with 48% less cognitive decline over three years. The remaining 739 participants, newly recruited community volunteers, showed no statistically significant cognitive benefit. The analysis examining differences between recruitment groups was prespecified, which strengthens its credibility without making it the trial’s primary finding.
A reasonable interpretation may therefore be narrower than either enthusiastic promotion or outright dismissal: hearing care could have cognitive benefits that depend on a person’s underlying risk and the period of observation. The trial does not justify promising that a hearing aid will preserve an individual’s memory. It also leaves room for a meaningful benefit that an average across dissimilar participants may obscure.
What the experiment actually compared
The intervention may be better understood as a hearing-care program than as a device test. The ACHIEVE study description reports audiologist sessions, hearing aids, and instruction in using technologies and communication strategies. The comparison group received an organized health-education intervention. This design asked whether offering supported hearing care could improve outcomes relative to another structured program.
That distinction matters when translating the findings into a purchase decision. A consumer may buy a device without receiving the fitting, troubleshooting, communication support, or continued encouragement available in a trial. Conversely, someone receiving ordinary clinical care may obtain support that differs from the research program. The results cannot identify a particular retail product as the component responsible for the possible cognitive benefit.
The active comparison may also make interpretation more realistic. An experiment against no contact could mix the effects of hearing treatment with the effects of attention and participation. Comparing two organized programs may help reduce that concern, although it cannot make the experiences identical. The study description supports interpreting the intervention as a package; it does not establish which part of that package might matter most.
For a reader evaluating a commercial claim, the practical question is whether the advertised service resembles the tested intervention. A claim about a small electronic device may sound more specific than the experiment allows. The finding could reflect amplification, learning to communicate more effectively, ongoing professional support, or a combination. Separating those contributions would require a study designed to compare them.
Random assignment may strengthen causal interpretation because treatment allocation does not depend on participants choosing hearing aids for themselves. It does not, however, remove every uncertainty. Participants may differ in how they use an intervention, some outcomes may be difficult to measure, and trial volunteers may differ from people seen in routine practice. A credible experiment can reduce uncertainty without settling every clinical question.
Why underlying risk may change the result
The recruitment groups may have differed in ways relevant to detecting cognitive change. As the study team explains in its overview of the trial, ACHIEVE combined participants from the Atherosclerosis Risk in Communities study, known as ARIC, with newly recruited volunteers. This offered an opportunity to examine hearing treatment across populations rather than assuming all older adults would respond alike.
A useful statistical possibility is that a protective intervention may be easier to detect where decline is occurring more quickly. Imagine two groups followed for the same interval: one changes substantially, while the other changes little. Even if an intervention could eventually help both, the second group may provide less opportunity to observe separation during that interval. This illustration explains a possibility; it does not prove why ACHIEVE’s groups differed.
A 2025 secondary analysis examined risk more directly. Researchers developed a prediction model using 2,692 dementia-free ARIC participants outside the trial and applied it to ACHIEVE participants’ baseline characteristics. Among participants in the highest quarter of predicted risk, the hearing intervention was associated with 61.6% slower cognitive decline over three years. The reported confidence interval was wide, from 33.7% to 94.1%.
This analysis may strengthen the hypothesis that several baseline vulnerabilities together help identify people more likely to benefit. It does not establish a ready-made consumer screening tool. A model used to investigate treatment effects in a study may need further validation before it can support an individual clinical decision, especially in populations different from those used to develop it.
The 2025 result also needs to be understood as another analysis of ACHIEVE rather than an independent replication. Reexamining one experiment may reveal an important pattern, but it does not create a second experiment. A new trial enrolling participants selected for higher predicted risk could provide a more direct test of the hypothesis.
Neither percentage should be read as a personal forecast. The estimates describe relative differences in cognitive decline on study measures, not the proportion of dementia cases prevented. A 48% or 61.6% reduction in a measured rate of change cannot be translated into the same reduction in an individual’s chance of receiving a dementia diagnosis. That conversion would require evidence about diagnoses themselves.
The 2026 blood findings offer a mechanistic clue
A secondary analysis published in 2026 investigated blood-based markers that might help explain the cognitive findings. In 164 ARIC participants with the required measurements, researchers compared three-year changes in glial fibrillary acidic protein, or GFAP, and neurofilament light chain, or NfL. The paper treated these as nonspecific indicators related to neuroinflammation and neurodegeneration, respectively.
Hearing intervention was associated with a slower rise in GFAP, with a statistically significant difference between groups. The NfL estimate pointed in a similar direction, but its confidence interval included no difference and the result was not statistically significant. Research therefore suggests a possible biological signal, while leaving uncertainty about whether the intervention affected the processes represented by both markers.
The authors also reported that the statistically significant effects did not exceed their reference change values, thresholds intended to account for measurement error and biological variation. That detail may limit the clinical interpretation of the average group difference. A result detectable in a statistical model need not be large enough to interpret confidently as a meaningful change in one person’s blood test.
These findings may help motivate mechanistic experiments, but they do not establish that hearing treatment reverses neurodegeneration or treats Alzheimer’s disease. Nonspecific biomarkers cannot identify a complete disease pathway on their own. Nor does a change in one marker show that the marker caused a change in cognition. The two outcomes could move together for reasons that require further investigation.
For readers, the distinction between a clue and a clinical endpoint is especially useful. A blood result may make a proposed explanation more plausible without demonstrating fewer dementia diagnoses, greater independence, or a lasting treatment effect. Those outcomes would need their own measurements. The biomarker analysis adds another layer to ACHIEVE; it does not replace the main trial result.
Plausible mechanisms need separate tests
Hearing treatment may influence everyday communication in ways relevant to cognitive activity. The National Institute on Deafness and Other Communication Disorders notes that hearing loss can make conversation difficult and may contribute to isolation. It also describes research into changes in sound processing. These observations offer context for possible pathways, rather than proof that amplification prevents cognitive disease.
One proposed pathway may involve listening effort. If spoken information becomes easier to distinguish, a person might devote less mental effort to reconstructing the words. That could leave more capacity for understanding or remembering the conversation. The hypothesis is testable, but an improvement in listening effort would not by itself demonstrate a lasting change in the underlying course of cognitive decline.
Another possible pathway may involve participation. Easier conversations could make social activities more accessible, creating opportunities for engagement that a person previously found tiring or frustrating. Yet the causal chain requires several separate links: hearing care would need to change communication, communication would need to change participation, and that change would need to affect later cognition. Evidence for one link cannot stand in for all three.
Measurement also deserves attention. A cognitive task delivered through spoken instructions may be harder for someone who does not hear those instructions clearly. Improving access to a test could improve performance without necessarily altering the underlying condition the test aims to assess. This is a methodological concern to examine, rather than a reason to assume every favorable cognitive result is an artifact.
Future studies may help distinguish these explanations by measuring communication, listening effort, participation, cognition, and biological markers at multiple points. The order of change could be informative: a proposed mediator should plausibly change before the outcome it is thought to influence. Even then, a mechanism would require careful analysis rather than a persuasive story attached after the result.
What readers can reasonably do with the findings
For someone experiencing difficulty hearing, the possible cognitive benefit may be an additional reason to consider assessment, rather than the sole justification for care. NIDCD describes hearing aids and other communication options as approaches that may help depending on the type and severity of hearing loss. Better access to conversation can be a relevant goal even when future cognitive outcomes remain uncertain.
In the United States, the FDA explains that adults with perceived mild to moderate hearing loss may purchase over-the-counter hearing aids without first seeing a hearing professional. That access category does not establish that every device or every purchaser will reproduce ACHIEVE’s findings. Regulatory availability and evidence for a cognitive outcome answer different questions.
A practical discussion with a hearing professional may focus first on the difficulty someone wants to address: following conversation in noise, understanding speech at home, or using a telephone. These goals may make an assessment more concrete than an abstract promise of dementia prevention. They also provide outcomes a person can discuss without waiting years to learn whether cognitive change differs.
The FDA’s information may help readers understand routes to obtaining hearing care, while the trial may help frame expectations. Neither supports choosing a device on the basis of a claimed dementia-prevention percentage. Price, fit, usability, and available support may be relevant to a purchase, but ACHIEVE did not rank commercial products by their ability to preserve cognition.
The remaining research questions are specific. Would an independent trial reproduce the higher-risk finding? Would longer follow-up show a difference in clinical diagnoses or everyday function? Could a simpler care program provide comparable results? Would risk prediction work across more varied populations? Answers to these questions may determine how far the promising subgroup findings can inform clinical practice.
For now, the most defensible interpretation may be that supported hearing care has a plausible, unevenly distributed cognitive benefit that requires further confirmation. The overall null result sets a boundary around broad promises. The subgroup and biomarker results offer reasons to continue testing a more targeted hypothesis, while communication needs provide an immediate, individual context for considering hearing care.
Frequently asked questions
Do hearing aids prevent dementia?
Research suggests possible cognitive benefits in some higher-risk older adults, but ACHIEVE did not establish dementia prevention or an overall cognitive benefit over three years.
Does the 48% finding apply to everyone?
No. That estimate concerned cognitive decline in the ARIC subgroup and may not apply to other populations or predict an individual’s outcome.
Do the blood findings show disease reversal?
No. The 2026 analysis suggests a possible GFAP signal, but biomarker changes do not establish reversal of neurodegeneration or a cure.
This article summarises published research and is not medical advice. Consult your healthcare provider before changing your routine.
Frequently Asked Questions
Do hearing aids prevent dementia?
Does the 48% finding apply to everyone?
Do the blood findings show disease reversal?
Sources
- The Lancet: ACHIEVE randomized controlled trial, 2023(2026)
- ACHIEVE Study: About the trial(2026)
- Alzheimer's & Dementia: Cognitive benefits vary by predicted risk, 2025(2026)
- eClinicalMedicine: ACHIEVE blood-based biomarker analysis, 2026(2026)
- NIDCD: Age-Related Hearing Loss(2026)
- FDA: How to Get Hearing Aids(2026)
Stay Updated on Longevity Science
Weekly research digests. No spam, unsubscribe anytime.
Related Articles

AI Drug Discovery for Longevity: What Research Is Coming
How artificial intelligence is accelerating longevity drug discovery, from target identification to clinical trial design, and what it means for aging research.
13 min readBlood Biomarkers for Aging: The Essential Panel
Learn which blood biomarkers are most important for tracking biological aging, what optimal ranges look like for longevity, and how to interpret your results.
10 min readContinuous Glucose Monitors and Aging Optimization
Continuous glucose monitors may help optimize metabolic health for longevity. Learn how CGM data can inform anti-aging dietary and lifestyle strategies.
10 min read