Hearing and Mobility in Aging—The Moderating Role of Neuropsychological Function

Abstract

Background

In older persons, neuropsychological function, especially sensorimotor function, is strongly associated with mobility. Hearing impairment is related to poor mobility, and we hypothesize that such relationships would be stronger in persons with compromised sensorimotor function than those with preserved sensorimotor function.

Methods

We analyzed 601 cognitively normal (Mini-Mental State Examination ≥24 or free of cognitive impairment/dementia) Baltimore Longitudinal Study of Aging participants aged 50 or older with concurrent data on hearing (better ear pure-tone average [PTA]), mobility (6 m usual gait, 400 m endurance walk), and neuropsychological function including attention via Trail Making Test (TMT)-A, executive function via the difference between TMT-A and B, and Digit Symbol Substitution Test, and manual dexterity via Pegboard performance. We examined the association of PTA, each cognitive measure, and their interaction with mobility using multivariable linear regression, adjusted for covariates.

Results

After covariate adjustment, there were significant interactions between PTA and all cognitive measures in the association with 400 m endurance walking speed as the outcome. There was also a significant interaction between PTA and pegboard nondominant hand performance in the association with 6 m gait as the outcome. In all these cases, the PTA–mobility relationship was stronger among participants with lower cognition.

Conclusion

Impaired hearing more strongly relates to poor mobility among those with compromised neuropsychological performance, especially manual dexterity reflecting the motor–cognitive interface and sensorimotor function, than those with preserved neuropsychological performance. Future longitudinal studies are needed to test whether hearing impairment more strongly predicts mobility decline over time in older persons with compromised neuropsychological function.

Slow walking is common in older adults and predicts multiple adverse health outcomes, such as falls and cognitive impairment (1). Poor mobility has multiple etiologies, including deficits in the central nervous, musculoskeletal, cardiopulmonary, and sensory systems. The role of the auditory system in mobility has received increased attention as it provides temporal and spatial sound cues necessary to maintain orientation during walking (2). By using acoustic information in the environment, an individual can perceive motion through space. Impaired hearing can adversely affect mobility through reduced auditory environmental cues and/or imposing a cognitive load on existing neural resources (3). Impaired hearing may also indicate subclinical damage in the central nervous systems, which contributes to poor mobility.

Previous studies have shown that hearing impairment (HI) is associated with poor mobility, including slow gait speed and poor walking endurance, but whether poor cognition exacerbates this association is not well understood (4). With age, HI imposes a cognitive load, and mobility tasks, especially challenging dual-tasking and long-distance endurance walks, also increase cognitive demands, thus older adults with poor cognition may be more vulnerable than those with preserved cognition during walking (4). Also, when hearing loss is an early manifestation of neurological impairment as indicated by concurrent cognitive performance, it may be more strongly correlated with poor mobility. Overall, we hypothesize that HI may more adversely affect mobility among those with poor cognition than those with preserved cognition.

In this study, we examined whether selected neuropsychological function modifies the relationship between hearing and mobility performance in a sample of cognitively normal community-dwelling older adults. Based on known relations of neuropsychological function, hearing, and mobility, we hypothesize that attention, executive function, and manual dexterity would modify the relationship between hearing and mobility (5–7).

Method

Study Population

We used data from the Baltimore Longitudinal Study of Aging (BLSA), which began in 1958 (8). The Institutional Review Board of the National Institutes of Health approved this study protocol. All participants provided written informed consent at each BLSA visit. This study was also preregistered with the Center for Open Science (www.osf.io/z39nr/) before the study was conducted in support of transparency in research.

We first identified 681 participants aged 50 or older who had concurrent data at the most recent visit on hearing (better ear pure-tone average [PTA]), mobility performance (usual gait speed, 400 m endurance walk), and cognitive function (attention, executive function, and manual dexterity) between 2012 and 2019. Because we aimed to examine the role of neuropsychological function in the relationship between hearing and mobility during usual aging, we focused on participants without overt neurological conditions and dismobility. Participants were excluded if they had evidence of cognitive impairment as indicated by a Mini-Mental State Examination score <24 (n = 12) and/or dismobility as indicated by usual gait speed ≤0.6 m/s (n = 7). Participants were also excluded if they were diagnosed with Parkinson’s disease (PD), stroke, cognitive impairment, or dementia (n = 61). PD and stroke were based on self-reported information. Diagnoses of cognitive impairment and dementia followed the standard BLSA procedures described previously (7). Diagnosis of dementia follows the criteria in the Diagnostic and Statistical Manual, third edition, revised (9). Mild cognitive impairment was determined using the Petersen criteria (10). The final analytic sample in this study consisted of 601 cognitively normal participants.

Hearing Assessment

Hearing was measured using pure-tone air conduction audiometry in a sound-attenuating booth with insert earphones. A speech frequency (0.5–4 kHz) PTA was calculated for each ear. PTA in the better ear was used for analysis.

Mobility

Measures of mobility included usual gait speed over 6 m and 400 m endurance walk. Usual gait speed was measured using 2 trials on a 6 m course in an uncarpeted corridor with the faster trial used in the analysis. The 400 m walk is a self-paced endurance walking test (11). During the 400 m walk, participants were instructed to complete ten 40 m laps as quickly as possible, and the time to complete the full 10 laps was recorded. Standardized verbal encouragement was given at each lap along with the number of laps remaining.

Neuropsychological Function

This study examined 3 neuropsychological domains, including attention, executive function, and manual dexterity (12). Attention was measured using the Trail Making Test (TMT) Part A. Executive function measures included the difference between TMT Part B and Part A (Delta TMT) and Digit Symbol Substitution Test. Manual dexterity was assessed using Purdue Pegboard Test, including the dominant and nondominant hand performance.

Covariates

Covariates included age, sex, race, years of education, body weight, and height. Body weight was related to both cognition and mobility performance. Body height was related to walking speed.

Statistical Analysis

We assessed the relationship between hearing and mobility using linear regression. We then tested whether performance on each cognitive measure of interest modified the relationship between hearing and mobility by adding an interaction between PTA and each cognitive measure of interest. To decompose statistically significant interactions, the association between hearing and mobility was examined for participants with low (−1 SD), average (at sample mean), or high (+1 SD) neuropsychological performance.

We performed a secondary analysis using hearing as a categorical measure based on PTA cutoffs by the World Health Organization (13). Normal hearing was defined as a PTA of ≤25 decibels of hearing level (dB HL), mild HI was defined as a PTA from 26 to 40 dB HL, and moderate or greater HI was defined as a PTA of >40 dB HL.

Significance was set at 2-tailed p < .05. All analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC).

Results

Participants’ characteristics are presented in Table 1. This sample had a mean age of 71.5 years with a mean gait speed of 1.2 m/s (57% women, 27% Black). Based on PTA cutoffs, 57.4% of the sample were considered to have normal hearing, 25.3% mild HI, and 17.3% moderate or greater HI (Table 1).

Higher PTA in the better ear was associated with slower 6 m gait speed and slower 400 m endurance walking speed (standardized β = −0.237, 95% CI: −0.325, −0.149, p < .001; standardized β = −0.299, 95% CI: −0.366, 0.231, p < .001, respectively). These associations were not significant after adjustment for age, sex, body weight, and height (standardized β = 0.042, 95% CI: −0.062, 0.146, p = .427; standardized β = −0.013, 95% CI: −0.079, 0.054, p = .704, respectively).

In the association with 6 m gait speed as the outcome, there was a significant interaction between PTA and pegboard nondominant hand performance (Table 2), indicating that the relationship between PTA and 6 m gait speed differed by pegboard nondominant hand performance. In participants with lower pegboard nondominant hand performance (1 SD below the sample mean), those with HI had slower gait speed than those with normal hearing (Figure 1A). In participants with higher pegboard nondominant hand performance (1 SD above the mean), the relationship between hearing and gait speed was substantially weaker (Figure 1A). The interactions between PTA and other neuropsychological measures of interest were not significant (Table 2).

In the association with 400 m endurance walking speed as the outcome, there were significant interactions between PTA and all neuropsychological measures considered (Table 2), indicating that the relationship between PTA and 400 m endurance walking speed differed by neuropsychological performance. In participants with lower neuropsychological performance (1 SD below the sample mean), those with HI had lower 400 m endurance performance than those with normal hearing (Figure 1B). In participants with higher neuropsychological performance (1 SD above the sample mean), the relationship between hearing and 400 m endurance performance was substantially weaker (Figure 1B).

Results remained similar when PTA was used as a categorical measure (Supplementary Table 1).

Discussion

In this sample of cognitively normal and well-functioning older adults, neuropsychological function modifies the relationship between hearing and mobility performance. Impaired hearing more adversely affects mobility among those with compromised neuropsychological function than those with preserved function, perhaps due to a cognitive load imposed by HI. Among cognitive measures examined, manual dexterity modifies relationships with both short-distance 6 m gait speed and long-distance 400 m endurance performance. These may suggest that subtle deficits in manual dexterity, reflecting the motor–cognitive interface and sensorimotor function, play a key role in the relationship between hearing and mobility.

We observed several neuropsychological measures modified the relationship between hearing and 400 m walk performance, and only manual dexterity assessed using pegboard nondominant hand performance modified the relationship with 6 m gait speed. These findings may be explained by differences between the 2 walking tests and their relationships with cognition. The 400 m walk is a challenging endurance walk and requires the fastest pace possible, while the 6 m walk is a short distance walk at a usual pace. Compared to the 6 m walk, the challenging 400 m walk assesses an individual’s endurance capacity, may require an increased demand for attentional and executive resources, and is more strongly associated with cognition (7). Age-related cognitive resource limitations become observable during the 400 m walk due to the increased cognitive demand. Compromised attention and executive function may play a more important role in the relationship of hearing with 400 m walk than 6 m walk. Because the maximal or submaximal speed is not assessed during the 6 m walk at a usual pace, it is unclear what 6 m usual gait measures relative to an individual’s capacity. This may also explain why few neuropsychological measures modified the relationship with 6 m usual gait speed.

Manual dexterity, specifically the pegboard nondominant hand performance, modifies relationships with both 6 m gait speed and 400 m endurance walk performance. One unique aspect of the pegboard test relative to other cognitive measures is that it is considered a motor–cognitive interface measure and represents a sensorimotor function. The influence of manual dexterity on the association between hearing and short-distance 6 m walk may be more due to motor than cognitive processes. The role of manual dexterity in the association between hearing and mobility may also be explained by shared brain structures. The sensorimotor cortex plays a key role in encoding temporal predictions used in auditory perception (14) as well as in mobility performance (15). Other brain areas vital for sensorimotor integration, including the parietal lobe and the thalamus, are associated with both hearing and mobility performance (15,16). Future studies are warranted to understand brain mechanisms underlying the interrelationships of hearing, neuropsychological function, and mobility.

This study has limitations. First, the cross-sectional design does not imply temporality or causation. Second, the BLSA sample is relatively healthier, less diverse, and more educated than the general older population. This study has several strengths. First, we included 2 mobility measures: one was relatively easy at a usual pace and the other was challenging at the fastest pace possible. Second, this study included multiple cognitive function measures, which allowed us to test which cognitive domains modify the relationship. Third, this study sample is well-characterized with rigorous adjudication of cognitive impairment and dementia, which allowed us to focus on cognitively normal participants. Finally, secondary analyses were performed to check the strength of these relationships. Our findings remained robust and consistent between continuous and categorical measures of hearing.

Findings from this study may provide new insight into prevention and interventions to preserve or improve mobility that involves a hearing component. Initial intervention studies examining the use of hearing aids in improving mobility have yielded mixed findings; some studies failed to observe improved mobility (17–19). These studies have not considered neuropsychological components. Whether a subgroup with compromised neuropsychological function benefits more from these audiological interventions than those with preserved function requires further investigation. Recent data have shown that combined cognitive and aerobic training improved mobility performance in older adults with age-related hearing loss, suggesting the benefit of incorporating cognitive components (20).

Conclusion

Impaired hearing more adversely affects mobility among those with compromised neuropsychological function than those with preserved function, perhaps due to a cognitive load imposed by HI. Future studies are needed to expand the current analysis to the longitudinal setting and test the hypothesis that HI more strongly predicts mobility decline over time in face of cognitive impairment.

Funding

This research was supported in part by the Intramural Research Program of the National Institute on Aging.

Conflict of Interest

None declared.

Author Contributions

D.A.P. and Q.T. developed the concept and study design. D.A.P. and Q.T. performed the statistical analysis. B.J.S. provided statistical guidance and contributed to the data interpretation. D.A.P., B.J.S., and Q.T. drafted the manuscript. J.A.D., N.M.A., E.M.S., S.M.R., F.R.L., and L.F. critically evaluated the manuscript and contributed to the data interpretation. All authors edited and approved the manuscript.

References