What ACHIEVE Really Tested

ACHIEVE is the concrete test behind the broader question about hearing aids and cognition. It did not ask whether people who happen to use hearing aids have better outcomes than people who do not. It randomly assigned a selected group of older adults with hearing loss to a supported hearing intervention or to an active health-education control, then followed global cognition for three years. Understanding that comparison makes the headline result more useful: the trial tested a defined care package in a defined population, not every possible hearing aid or every form of dementia prevention.

A supported hearing intervention, not a device-versus-nothing experiment

ACHIEVE enrolled 977 community-dwelling adults between 70 and 84 years old. They had adult-onset bilateral hearing loss in a range expected to benefit from conventional hearing aids and related audiological services, but they did not have substantial cognitive impairment at enrollment. Participants also had to be able to take part in the study's assessments, including being fluent English speakers. The study excluded recent hearing-aid users, people with permanent conductive hearing loss, people with major functional or visual limitations and people who could not or would not use hearing aids regularly.

Those criteria are not incidental details. They identify the population to which the direct randomized evidence applies most closely: older adults living in the community, with hearing loss that the investigators judged suitable for conventional amplification, and without substantial cognitive impairment. The trial does not directly answer what happens when hearing aids are started in midlife, when a person already has established dementia, when hearing loss is severe or conductive, or when the appropriate intervention is a cochlear implant or another form of rehabilitation. A carefully selected sample can make a trial interpretable, but it also makes generalization narrower.

The 977 participants were randomized in roughly equal numbers. The hearing-intervention group contained 490 people; the health-education control contained 487. Randomization is the feature that changes the kind of question the study can answer. Instead of comparing people who selected hearing aids with people who did not, the trial created groups whose assignment to the study conditions was determined by the research design. Differences that emerged afterward can therefore be interpreted more directly as differences associated with the assigned intervention, subject to the usual limits of the trial and the outcome being measured.

The hearing condition was more than a one-time prescription. Participants were fitted with hearing aids, instructed to wear them daily for the duration of the study and given four one-hour intervention sessions spaced across the two to three months after randomization. Re-instruction in device use and hearing-rehabilitative strategies was provided every six months. This matters because the exposure being tested was supported use of hearing care. It was not simply the biological fact of owning an aid, and it was not a test of whether a person could be handed a device without follow-up and expected to obtain the same result.

The control condition also received structured attention.