Do Hearing Aids Help Prevent Cognitive Decline?

Abstract

Dementia is a devastating disease and public health concern that affects ~50 million individuals globally, a number projected to triple by 2050. A growing body of research has identified an independent association between age-related hearing loss (ARHL) and both dementia and impaired cognition, as well as other conditions of the elderly (e.g., depression, loneliness, increased risk of falls, and frailty). Given limited efficacy of existing pharmacologic treatments for dementia, treating risk factors is a crucial strategy. A severely undertreated and highly prevalent disease in the elderly, ARHL is a compelling potential target for preventative strategies for cognitive decline and dementia. In this Best Practice, we ask if hearing aid (HA) use has a positive impact on cognitive decline and dementia. Studies investigating cognitive outcomes with HA usage have largely comprised of observational studies. A cross-sectional study by Qian et al. examined the association between HA usage and cognition in 100 older adults with HL who did or did not use a HA. Cognitive outcomes included the Mini-Mental State Examination (MMSE; measures global cognitive function) and Trail Making Test, Part B (TMT-B; measures executive function).1 The HA group performed better on the MMSE but not the TMT-B compared to the non-HA group, despite having worse hearing at both high and low frequencies. However, the study did not adjust for potential confounders (e.g., age and socioeconomic status). Prospective, observational, cohort studies with larger sample sizes have demonstrated the longitudinal association of HA usage and improved cognitive performance or slower cognitive decline. A study by Amieva et al. examined the association between HL, HA usage, and cognitive decline over time in 1,276 adults >65 years from the Personnes Agées QUID study.2 HL and HA usage were determined with an initial questionnaire assessing self-perceived HL and HA usage. Participants underwent interval cognitive testing using the MMSE over 25 years. HL subjects without HAs had a faster rate of cognitive decline compared to normal hearing controls. In contrast, HL subjects with HAs had no significant difference in cognitive decline compared to controls. The authors conclude that though self-reported HL is associated with accelerated cognitive decline in older adults, HA usage attenuated this decline. Interestingly, there was no direct comparison between HA and non-HA groups among those with HL. Maharani et al. investigated the longitudinal association between HA use and cognitive function in an uncontrolled prospective cohort study of 2,040 adults ≥50 years from the Health and Retirement Study who used HAs for the first time partway through the study.3 Cognitive outcomes were measured every 2 years over a period of 18 years and based on episodic memory scores (sum of immediate and delayed recall of 10 words); scores pre- and post-HA use were compared. There was a slower rate of decline in episodic memory scores after using HAs compared to before using HAs. Some studies did not find demonstrate significant findings. An 8-year longitudinal study of men aged ≥62 years examined the relationship between self-reported HL, self-reported HA use, and risk of subjective cognitive function (SCF) decline.4 Longitudinal analysis demonstrated that the magnitude of the independent association between incident SCF decline and HL was not significantly attenuated (P = .4) by HA use in individuals with severe HL (non-HA group n = 65, adjusted risk ratio 1.37; HA group n = 179, adjusted risk ratio 1.18). In the only randomized controlled trial (RCT) conducted in individuals who are not specifically cognitively impaired, Mulrow et al. conducted a study of 188 elderly veterans with HL who were randomly assigned to receive a HA (n = 95) or join a waiting list (n = 99). HL was defined as a better-ear hearing threshold of ≥40 dB at 2,000 Hz.5 Participants were randomly assigned to an immediate HA group or waiting list group. Cognitive function was assessed using the Short Portable Mental Status Questionnaire (SPMSQ). Subjects from both groups were similar in baseline demographic, clinical, and HL characteristics. Subjects who received HAs demonstrated a small but significantly improved mean SPMSQ scores over 4 months of 0.28 points compared to the control group (test is scored from 0 to 10 by counting total number of errors; 0 to 2 indicates intact intellectual function). Although results are preliminary, a growing number of studies demonstrate a cross-sectional and longitudinal association between HA usage and improved cognition or decreased rate of cognitive decline (Table I). Nonetheless, these studies also exhibit significant limitations. The RCT by Mulrow et al. had limited follow-up (4 months) and a small sample size.5 Several studies employed self-reported measures of HL, HA usage, and HA possession.2-4 This may introduce bias and skew the effect size in either direction. Subjective perception of cognitive function (i.e., SCF) may not accurately reflect true cognitive function, particularly in a disorder, which can include poor insight. Moreover, most of the studies assessed global cognitive function, rather than specific cognitive domains. Future studies should investigate specific domains, such as perceptual-motor function, language, learning and memory, executive function, complex attention, and social cognition (per the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition [DSM-5]). Several studies also had limited external validity; the sample population of Curhan et al. (2019) consisted of predominantly white male health-care professionals with high socioeconomic status), while that of Mulrow et al. (1990) consisted of male veterans.4, 5 None of the studies included subjects with dementia or major cognitive impairment. From these studies alone, it is unclear if HA usage would benefit individuals who have baseline cognitive impairment (secondary and tertiary prevention) versus individuals with normal cognition (primary prevention). Modest effect sizes can also be attributed to several factors, including inability to address inconsistent HA usage (i.e., from subjective report of HA usage) and suboptimal fittings (i.e., monaurally fittings in Mulrow et al. [1990]). Maharani et al. did not employ a control group in their study.3 Although all studies but one (Qian et al., [2016])1 accounted for demographic and psychosocial characteristics, participants using HAs may have differed from those not using HAs in unmeasured covariates. Though the evidence is encouraging, clear causal linkage between HL and cognition has not been established. Furthermore, given the slow nature of cognitive decline, establishing benefit of HA on cognition will require well-controlled long-term studies. In the interval, HA usage for prevention of cognitive decline is warranted given the low risk and strong theoretical benefit of HA. Well-designed RCTs with accurate measures of HL, HA usage, and cognitive performance (of a variety of domains over time) are required to demonstrate conclusive evidence that HAs help attenuate or prevent cognitive decline in individuals with HL. Three prospective cohort studies (level 2), one cross-sectional study (level 3), and one RCT (level 1) were evaluated.