Hearing loss travels with faster cognitive decline and with incident dementia in older adults, which raises an obvious question nobody could answer from observation alone. Lin and colleagues ran the trial: nine hundred and seventy-seven people aged seventy to eighty-four with untreated hearing loss, to hearing aids or to health education, followed for three years.
Not in the trial as a whole. Three-year change in came out effectively identical in the two groups, about a fifth of a standard deviation of decline in each. The difference was two thousandths, at a p value of point nine six. A prespecified sensitivity analysis did find that the effect differed between the two populations the cohort was built from.
The association was already established: hearing loss goes with increased cognitive decline and incident dementia in older adults. What that cannot settle is direction or remedy, so the trial asked whether a hearing intervention reduces cognitive decline in cognitively healthy older adults who have hearing loss.
The design is unusual in a way that turns out to matter. The trial recruited adults aged seventy to eighty-four with untreated hearing loss and without substantial cognitive impairment, at four community sites, from two different populations. One was the long-running study of cardiovascular health. The other was healthy community members volunteering for this trial. From November two thousand and seventeen to October two thousand and nineteen the team screened three thousand and four people and randomly assigned nine hundred and seventy-seven. Half went to audiological counselling with hearing aids, half to health education on chronic disease prevention.
The cohort had a mean age of seventy-six point eight years, slightly more women than men, and was eighty-eight percent White. Two hundred and thirty-eight came from the observational study and seven hundred and thirty-nine were . The two groups were not alike. The first were older, carried more risk factors for cognitive decline, and started with lower cognitive scores.
In the primary analysis combining both, three-year cognitive change was not significantly different between groups. Each declined about a fifth of a standard deviation, a difference of zero point zero zero two at a p value of point nine six. A prespecified sensitivity analysis, however, showed the effect of the intervention differed between the two populations, with an interaction p value of point zero one. Other prespecified analyses varying the parameters did not change the overall picture, and no significant adverse events were attributed to either arm.
wat. Hearing aids did not slow three-year cognitive decline across the trial. Whether they do for people already at risk is the question it leaves open.
wat wat
The split between the two recruited populations is the finding worth carrying. One group was older, had more risk factors and lower baseline scores, which is where an intervention like this has room to show anything.
How do we know?
A multicentre, parallel-group, unmasked randomized controlled trial at four community sites in the United States. Participants were assigned one to one to audiological counselling with hearing aids or to individual health education sessions, followed every six months, and analyzed by intention to treat.
What is still unknown
The subgroup difference comes from a sensitivity analysis rather than the primary endpoint, so it is a signal about who to study next rather than a result to act on. The cohort was eighty-eight percent White, and neither arm was masked.
Learn something?
Where to next
What gets through while you are ignoring everyone else at a party?hearing
What happened to the memory of mice when their gut bacteria were swapped?aging
What do you learn from measuring the same fingernail for 35 years?aging