One in Three Cases of Dementia May be Attributable to Hearing Loss: Population-Based Study

Nearly one in three new cases of dementia in older adults may be attributable to hearing loss, according to new data published in JAMA Otolaryngology—Head & Neck Surgery. The prospective cohort study revealed that among participants with clinically significant hearing loss measured by audiometry, 9.9% developed dementia during the 8-year follow-up period, compared to just 4.7% of those with normal hearing. Statistical modeling estimated that approximately one-third (32.0%) of incident dementia may be associated with audiometric hearing loss.

The authors, led by Jason Smith, PhD, from the Johns Hopkins Bloomberg School of Public Health, in Baltimore, reported that population attributable fractions (PAF) of cognitive decline related to hearing loss were greater in those aged 75 years and older (30.5%), among women (30.8%), and among White individuals (27.8%). They also found that self-reported hearing loss was not associated with an increased risk of dementia, with a hazard ratio of less than 1, making PAF unquantifiable, they wrote.

“Interventions for sensory health in late life might be associated with a broad benefit for cognitive health,” the authors wrote. With dementia prevalence expected to triple globally in coming decades,2 addressing modifiable risk factors like hearing loss takes on increased urgency, they added.

Smith and colleagues found clinically significant hearing loss measured by audiometry in 66.1% of the study population of 2946 community-dwelling adults (mean age 75 years) and self-reported hearing loss among 37.2%. Analyses revealed similar PAF by hearing loss severity, ie for mild hearing loss, 16.2% and for moderate or greater hearing loss, 16.6%, suggesting that even mild hearing impairment may contribute significantly to dementia risk.

The team cites previous estimates of dementia associated with hearing loss in the US that range from 2% to 19%,3,4 making the current finding of 32% notable. One key explanation for the variance is the type of measure used to quantify hearing loss. Many past studies have relied on self-report, which may result in poor hearing being misclassified as normal the researchers said. The error "results in substantially lower prevalence estimates than what would be expected for older adults in the US." Their data confirm the magnitude of the effect, as self-reported hearing loss showed no dementia risk association.

For the population-based study, Smith and colleagues analyzed data from 2,946 community-dwelling adults aged 66 to 90 years without dementia at baseline. Participants were enrolled in the Atherosclerosis Risk in Communities Neurocognitive Study (ARIC-NCS) across four state field centers.

Data were collected between 2011 and 2019, with a single hearing assessment conducted at visit 6 (2016-2017). Researchers measured hearing loss both objectively using pure-tone audiometry and subjectively via self-reporting. Investigators determined a dementia diagnosis using a standardized algorithm that incorporated neuropsychological testing, informant interviews, and medical records.

Among the limitations of the analysis the investigators note sourcing data from a community-based cohort of self-identified Black and white adults and the generalizability of prevalence estimates to the wider US population could be limited. By design, ARIC oversampled adults who self-identified as Black for greater representation.

Smith's team compares the preventive potential from addressing hearing loss in late life with other potentially modifiable risk factors, citing PAFS for midlife obesity, 18%; physical inactivity, 12%; low education, 12%, compared with the 32% for hearing loss found in their study.5 They point out that the ability to control "upstream factors" such as noise and ototoxic medications, and the various interventions available for treatment of impaired hearing offer a powerful public health message about cost effectively maintaining auditory health across the lifespan.

References