Health · 6 min read
What a Randomized Hearing Trial Found About Cognitive Decline
The ACHIEVE trial’s overall result, higher-risk subgroup finding, and the limits of dementia-prevention claims.

What a Randomized Hearing Trial Found About Cognitive Decline
Age-related hearing loss and cognitive decline often occur together. Observational studies have repeatedly found that older adults with greater hearing loss are more likely to experience faster cognitive decline or develop dementia. Those associations are important, but they leave the central causal question open. Hearing loss might contribute to cognitive decline, shared vascular or biological processes might cause both, or early cognitive changes might make hearing tests and communication harder. Social isolation, education, and access to care can also affect both measurements.
The ACHIEVE trial was designed to move from association toward intervention evidence. Its result is more informative—and less simple—than the claim that hearing aids either prevent or do not prevent dementia. Across the full randomized cohort, a three-year hearing intervention did not significantly slow cognitive decline compared with a health-education control. In a prespecified subgroup at higher baseline risk, the investigators found a benefit. That divergence supports further research, not a universal prevention claim.
What ACHIEVE compared
ACHIEVE stands for Aging and Cognitive Health Evaluation in Elders. Investigators led by Frank Lin enrolled 977 adults aged 70 to 84 at four United States sites. Participants had untreated mild-to-moderate hearing loss and did not have substantial cognitive impairment at enrollment. They were randomly assigned either to a hearing intervention or to a successful-aging health-education program.
The hearing intervention was more than handing over a device. It included bilateral hearing aids, fitting by audiologists, counseling, and instruction in communication strategies. The control group met with a health educator about chronic-disease prevention. That active control helped balance attention and contact between groups. The primary outcome was change over three years in a global cognition score assembled from a battery of neuropsychological tests.
Participants came from two populations. Some were already enrolled in the Atherosclerosis Risk in Communities, or ARIC, observational study. They were older and, on average, had more risk factors for cognitive decline. Others were newly recruited healthy community volunteers. Randomization occurred within these populations, and the analysis plan identified them before results were known.
The trial successfully changed hearing-related behavior. Device use and measures of communication improved in the intervention arm. This matters because a null cognitive result cannot readily be dismissed as a comparison between people who did and did not receive usable treatment.
The primary result was null
In the combined cohort, both groups showed some cognitive decline over three years. The difference in the rate of change between the hearing-intervention and health-education groups was not statistically significant. The investigators' published interpretation in The Lancet is direct: the intervention did not reduce three-year cognitive decline in the primary analysis of the total cohort.
That finding limits a common inference from observational data. An association between hearing loss and dementia risk does not by itself mean that treating hearing loss will reduce cognitive decline over a three-year interval. ACHIEVE supplied a randomized test and did not find an overall effect on its primary outcome.
It did not show that hearing care lacks value. Hearing aids can improve access to speech and communication, and clinical decisions need not wait for a dementia-prevention benefit. The United States National Academies' 2016 report on hearing health documented consequences of untreated hearing difficulty for communication, safety, participation, and quality of life. Those are direct treatment goals, separate from cognition.
The subgroup result requires both attention and restraint
The two recruitment populations produced different estimates. Among the healthier volunteers, cognitive change was slower overall and did not differ materially by assignment. Among ARIC participants, who had more baseline risk factors and a faster rate of decline, the hearing intervention was associated with a substantially smaller three-year decline than the control. The paper estimated a 48 percent reduction in the rate of cognitive change in that subgroup.
Because the subgroup was prespecified and reflected a visible difference in baseline risk, the finding is more credible than an exploratory split discovered after analysis. It is also consistent with a plausible threshold: an intervention may be easier to detect among people declining more quickly. But subgroup effects can be unstable even when prespecified. ACHIEVE was powered for the combined primary comparison, not for a definitive claim that one recruitment source identifies everyone who will benefit.
The percentage itself needs context. It describes a relative difference in the modeled rate of change on a composite score, not a 48 percent reduction in diagnosed dementia and not a restoration of lost cognition. The trial followed participants for three years. Dementia develops over longer periods, and the study was not designed primarily to count dementia diagnoses.
Several mechanisms remain plausible
One hypothesis is cognitive load. When speech is degraded, a listener may devote more attention to decoding it, leaving fewer resources for comprehension or memory. Better amplification can reduce that immediate burden. A second hypothesis is social engagement: easier conversation may help people maintain relationships and activities that are themselves associated with cognitive health. A third proposes longer-term changes in brain structure or function after reduced auditory input.
None of these paths has been established as the single mediator of dementia risk. Hearing aids do not correct speech perfectly, especially in noisy settings. Social participation depends on mobility, health, income, and available partners as well as audibility. Vascular disease can damage both auditory and cognitive systems. The 2020 Lancet Commission on dementia prevention classified hearing loss as a potentially modifiable risk factor, but “potentially modifiable” is a population-risk judgment, not proof that an individual intervention prevents dementia.
ACHIEVE's design also cannot blind participants or audiologists to hearing-aid use. Cognitive assessors could be masked, but expectations and behavior outside the clinic may still differ. The active control reduced that concern without eliminating it. Three years may be too short to observe some effects, or long enough to show that any average short-term cognitive effect is small. Longer follow-up is necessary to distinguish those possibilities.
What is established, contested, and unknown
Three conclusions can be stated with different strengths. First, the intervention improved hearing-related communication measures; that is direct trial evidence. Second, it did not significantly slow cognitive decline across the full cohort over three years; that is the primary randomized result. Third, it may slow decline in older adults at elevated risk; that is a supported but provisional interpretation of a prespecified subgroup.
Researchers still disagree about how much hearing loss causally contributes to dementia, which mechanisms dominate, and how to identify people most likely to gain cognitive benefit from treatment. Longer randomized follow-up, dementia outcomes, mediation analyses, and replication in other health systems could narrow those uncertainties.
For patients, the evidentiary boundary is straightforward. Treating hearing loss can be justified by hearing and communication benefits. It should not be promised as a general method of preventing dementia. The ACHIEVE trial made the question more precise: any cognitive benefit may depend on baseline risk, observation time, and the outcome measured.
Sources consulted
4 sources, every one with a link you can follow.
- Frank Lin and colleagues, “Hearing Intervention Versus Health Education Control to Reduce Cognitive Decline,” The Lancet (2023).
- National Academies of Sciences, Engineering, and Medicine, “Hearing Health Care for Adults,” consensus report (2016).
- Gill Livingston and colleagues, “Dementia Prevention, Intervention, and Care: 2020 Report30367-6/fulltext),” The Lancet (2020).
- National Institute on Aging, “Hearing Aids Slow Cognitive Decline in People at High Risk,” trial summary (2023).
