Hearing aids showed no clear effect on dementia progression in 703 adults with mild cognitive impairment

Hearing loss appears near the top of most lists of dementia risk factors that a person might be able to change. Those who use hearing aids also tend to hold on to their thinking skills for longer than those whose hearing loss goes untreated. Put together, those two observations have produced a hopeful idea: fit the hearing aid and protect the brain.

Comparing users with non-users cannot test that idea, because buyers and wearers of hearing aids differ from everyone else in income, health and much else. A randomized trial can. A team led by Ying Chen at Shanghai Ninth People’s Hospital, part of Shanghai Jiao Tong University School of Medicine, has now run one where a benefit seemed most likely: in older adults with both hearing loss and early problems with memory and thinking. Writing in JAMA Neurology on September 28, they report that a hearing aid intervention did not significantly reduce the incidence of dementia-level impairment at 24 months. More of those given hearing aids did improve, a result the authors themselves treat with caution.

What is mild cognitive impairment?

Mild cognitive impairment, or MCI, describes memory or thinking that tests worse than expected for a person’s age while everyday life carries on much as before. The Shanghai team describe it as objective cognitive impairment with preserved daily functioning and a raised risk of conversion to dementia, and they add that someone with MCI can also drift back to normal scores.

It is common. A 2018 evidence review for the American Academy of Neurology put MCI prevalence at 6.7% for ages 60-64, rising to about a quarter of people in their early eighties. In the studies it gathered, roughly 15% of people over 65 with MCI went on to develop dementia within two years.

Those numbers make MCI an obvious place to test anything that might delay dementia. If treating hearing loss protects thinking, the effect ought to show soonest where risk is already raised.

Why hearing would affect thinking at all is still argued over. When the Shanghai team set out their plans for the trial, they listed three proposed explanations: the cognitive load hypothesis, common cause hypothesis, and cascade hypothesis. The first holds that straining to follow speech uses up mental effort that would otherwise go to remembering. The second says that aging damages the ear and the brain together, in which case treating the ear would not be expected to change the brain. The third pictures a chain reaction, in which poor hearing leads to less conversation and less stimulation, and thinking suffers in turn.

How the CHOICE trial tested hearing aids

The trial, known as CHOICE, screened 21,908 older adults at three hospital ear clinics and three community health stations in Shanghai. It enrolled 703 participants with coexisting MCI and moderate to severe hearing loss. Their average age was about 75, and none had used a hearing aid in the previous year.

Half were randomly assigned to receive free hearing aids for both ears, fitted by certified audiologists and adjusted at every visit. The other half received hearing care education: sessions of approximately 45 minutes on hearing loss, chronic disease, healthy habits and ways to protect thinking, timed to match the contact that hearing aid wearers had with staff. That comparison matters, because without it any gain among the wearers could simply reflect extra attention from a clinic.

Everyone was assessed at the start and after 6, 12 and 24 months with the Clinical Dementia Rating, a structured interview lasting 30 to 45 minutes that scores memory, orientation, judgment and daily activities. A score of 0 means unimpaired, 0.5 is the questionable range the trial used to define MCI, and 1 or more means dementia-level impairment.

Nobody can be kept unaware of whether they are wearing hearing aids, so the trial was open-label, meaning that participants and clinic staff knew who had been assigned to what. The interviewers were the exception. Hearing aids were taken out before each assessment so that whoever was scoring could not tell the two sets of participants apart. In all, 610 of the 703 completed the two-year visit.

Did hearing aids slow progression to dementia?

On the trial’s main question, the answer was no clear effect. After two years, 3.09% of the hearing aid group had progressed to dementia-level impairment, against 4.74% of the education group. The gap of about a percentage point and a half leans in the hopeful direction, but it is small enough to be chance.

Part of the reason is that so few participants progressed at all. The researchers had planned for far more. Their sample size calculation rested on earlier figures in which dementia-level impairment developed in 10% of hearing aid users with MCI compared with 18% of nonusers. With only a few dozen cases in total, the trial could not tell a modest benefit from none. The authors say so plainly: the rate was lower than they had planned for, “resulting in underpowered analyses for the primary outcome”. An underpowered trial is one too small to detect the effect it was looking for.

Why so few? The often-quoted figures for MCI turning into dementia come from specialist memory clinics, where patients arrive because something is clearly wrong. These participants were recruited from ear clinics and community health stations, and most were getting on with daily life.

The more striking number came from a secondary outcome, meaning a result the trial measured but was not built around. Among those given hearing aids, 15.34% moved from the MCI range back to an unimpaired score, compared with 2.36% of those given education. That is roughly one person in six against one in forty. When the researchers looked at the separate parts of the rating, memory showed the clearest difference.

The hearing aids themselves did their ordinary job. Wearers kept them in for an average of 5.9 hours a day, their measured hearing improved substantially with the devices in, and no study-related adverse events were reported.

Why the cognitive improvement with hearing aids needs caution

The authors call the improvement an exploratory result, writing that “this exploratory finding warrants cautious interpretation and further investigation”. Their own list of limitations explains the reluctance.

The rating is a judgment formed in an interview, and the trial relied on that subjective scale as the sole instrument for both screening and outcome assessment. Participants and their families knew who had hearing aids, and the devices were offered to the intervention group at no cost, which the authors say may have colored the answers given. Someone who has heard well for two years may also cope better in any long conversation, an interview included. The team took the hearing aids out for testing and offered written instructions, yet they concede that they cannot fully exclude residual effects of better hearing on the scores.

Two further details weaken the result. The entry rule of a 0.5 rating was not prespecified in the protocol or trial registration, which means it was not set down publicly before the trial began. And more participants dropped out on the education side than on the hearing aid side: of those who missed the final visit, there were 33 in the intervention group and 60 in the education group.

Who took part also limits how far the result travels. This was one city, more than 90% of participants lived with family, and Chinese is a tonal language, which the authors suggest could change how hearing treatment affects speech perception and thinking.

How CHOICE fits with ACHIEVE and earlier hearing aid studies

The belief that hearing aids protect the brain grew out of observational studies. In 2023 Brian Sheng Yep Yeo and colleagues at the National University of Singapore pooled them, also in JAMA Neurology. Across eight long-term studies, hearing aid users compared with participants with uncorrected hearing loss were about a fifth less likely to show cognitive decline. Research of that kind cannot separate the device from the kind of person who gets one, and Yeo’s team concluded that a benefit needed to be further investigated in randomized trials.

The first large trial reported in The Lancet the same year. The American study known as ACHIEVE, led by Frank Lin of Johns Hopkins, enrolled 977 adults aged 70-84 years with untreated hearing loss and without substantial cognitive impairment. After three years, the hearing intervention did not reduce 3-year cognitive decline in the primary analysis of the total cohort. But the effect differed between the two study populations that made up the trial. Volunteers drawn from a long-running heart study were older and had more risk factors for cognitive decline, and in that half hearing treatment appeared to help. Lin’s group suggested that it might reduce cognitive change over 3 years in populations of older adults at increased risk.

CHOICE was built to test that suggestion in adults at higher risk still, and its main result neither confirms nor rules it out. The direction favored hearing aids and the trial was too small to settle the matter. The Shanghai authors note that three other trials in adults with hearing loss and MCI are under way.

What the CHOICE result changes for hearing aids and dementia

It leaves the case for treating hearing loss where it was. The authors put it this way: “The hearing aid is an important intervention for hearing loss, and thus, for communication improvement, even if its effect on dementia prevention among individuals with MCI remains uncertain”.

They also point to how few of those who could use a hearing aid have one. Adoption runs at 25% to 40% in high-income countries, they write, whereas the rate remains below 10% in China.

For now, hearing aids are a treatment for hearing loss whose effect on dementia has not been shown in a trial.