Abstract
Background
Hearing impairment (HI) could be a risk factor for cognitive decline, but cognition could plausibly also affect psychoacoustic assessment of hearing with audiometry. We examined the temporal sequence of hearing and cognitive function among nondemented, community-dwelling older adults.
Methods
Hearing and cognition were assessed between 2012 and 2015 and 2 years thereafter in 313 nondemented participants aged ≥60 years in the Baltimore Longitudinal Study of Aging. Poorer hearing was defined by pure-tone average of 0.5–4 kHz tones in the better-hearing ear. Cognitive measures with either visual or auditory inputs were Trail-making Test Part B; Digit Symbol Substitution Test; California Verbal Learning Test immediate recall, short delay, and long delay; Digit Span Forward/Backward; Benton Visual Retention Test; and Mini-Mental State Examination. We used linear regression models for cross-sectional associations at each timepoint and autoregressive, cross-lagged models to evaluate whether baseline hearing impairment (Time 1) predicted cognitive performance 2 years after baseline (Time 2) and vice versa.
Results
Cross-sectionally, there were no associations between poorer hearing and cognitive performance. Longitudinally, poorer hearing was associated with declines in California Verbal Learning Test immediate (β = −0.073, SE = 0.032, p = .024), short-delayed (β = −0.134, SE = 0.043, p = .002), long-delayed (β = −0.080, SE = 0.032, p = .012) recall, and Digit Span Forward (β = −0.074, SE = 0.029, p = .011).) from Time 1 to Time 2. Cognitive performance at Time 1 did not predict change in hearing status at Time 2.
Conclusions
Audiometric hearing impairment predicted short-term cognitive declines in both California Verbal Learning Test and auditory stimuli for attention.
Keywords: Hearing, Cognition, Executive function, Memory
Hearing impairment (HI) reflects peripheral auditory functional impairment (1). It is prevalent in two-thirds of adults ≥70 years (2). HI is cross-sectionally associated with poorer memory, executive function, global cognition, and mental status (3,4), and longitudinally with faster rates of decline in global cognition in older adults (4–10). Hypothesized mechanisms underlying the association of HI and cognitive function suggest that impaired hearing may contribute causally to cognitive decline and dementia through either information degradation or sensory deprivation (11–14). However, whether cognitive function could also contribute to changes in hearing, possibly through effects of subclinical cognitive impairment on psychoacoustic testing performance with audiometry, is unclear. Reverse causation, whereby cognitive impairment results in worse hearing test performance, is also plausible. The temporal sequence of HI and cognitive change has been identified as a gap in current knowledge of the association between HI and cognitive performance (15).
Previous work on examining the association of sensory loss with cognitive change cross-sectionally and longitudinally has found associations of sensory loss with cognitive change. Cross-sectionally, sensory, sensorimotor, and intellectual functioning are highly correlated at older ages (16,17). In addition, sensory processing, age, and global cognitive performance are moderately correlated with one another (18). Longitudinally, Lindenberger and Ghisletta (19) reported moderate correlations between cognitive and sensory declines (close and distant visual acuity and hearing) in 516 participants, ranging from ages 70 to 103, followed up to 13 years. Our study builds on these findings through examining the temporal sequence of HI and cognitive function using data from the Baltimore Longitudinal Study of Aging (BLSA), a prospective cohort study of aging.
Besides examining the associations between peripheral auditory function and cognition, we aimed to evaluate the relationships of peripheral auditory function by ear with cognition. There could be laterality differences in the way peripheral auditory function is processed, thus affecting performance on cognitive measures. Adults with left-hemispheric language lateralization can be more accurate in reporting verbal stimuli arriving at the right ear than left ear (20). Kimura (21) explained that this right ear advantage could be attributed to the right ear being connected to language centers located on the left hemisphere through contralateral pathways. These may affect the performance on auditory cognitive tests, as opposed to performance on visual cognitive tests.
To examine whether HI predicts cognition, as tested by either visual or auditory inputs, 2 years later and vice versa, we used cross-lagged models to examine whether HI was associated with change in cognition in three domains, that is attention/executive function, memory, and mental status, 2 years after baseline and or whether baseline z scores of neuropsychological measures were associated with declines in hearing 2 years after baseline. We hypothesized that poorer baseline hearing is associated with declines in cognition 2 years after baseline, but that lower baseline cognition is not associated with declines in hearing 2 years after baseline. In addition, we examined the laterality differences of the associations of HI and cognition to determine if laterality of HI affected the associations found in the main study.
Methods
Study Participants
Participants, ranging from ages 61 to 98 years, were community-dwelling BLSA volunteers (22). Participants aged 60–79 years have biennial study visits, whereas participants aged ≥80 years have annual study visits. Participants had two assessments of both audiometric and cognitive test data collected between 2012 and 2017 to be included in this analysis (N = 319). We further restricted the sample to those without cognitive impairment at the first timepoint (N = 313). The first assessment occurring from 2012 to 2015 was defined as baseline, and the second assessment occurred 2 years after baseline.
The local institutional review board approved the research protocol for this study, and written informed consent was obtained at each visit from all participants.
Peripheral Audiometry
Pure-tone audiometric testing, a measure of the peripheral auditory system sensitivity (21), was conducted by a trained technician using insert earphones and with an Interacoustics AD-629 audiometer in a sound-attenuating booth. A speech-frequency pure-tone average (PTA) of air-conduction thresholds at 0.5, 1, 2, and 4 kHz was calculated for each ear. All thresholds were measured in decibel of hearing level (dB HL). Hearing was defined by the PTA in the better-hearing ear.
Cognitive Measures
We examined tests of visual and auditory attention/executive function, verbal and visual memory, and mental status at both timepoints. Attention/executive function tests included the completion times of Trail-Making Test Part B (TMT-B) (23), number of symbol/digit pairs from Digit Symbol Substitution Test (DSST) (24), and length of digits for Digit Span Forward and Backward (25). We calculated the difference of the completion times between TMT-B and Trail-Making Test Part A (23) to account for the processing speed component and log-transformed the difference to account for the nonnormality of the distribution of the difference. Verbal memory tests included total number of words recalled from the immediate, short-delay, and long-delay measures recall from the California Verbal Learning Test (CVLT) (26,27). Visual memory was assessed by the total number of errors calculated from Benton Visual Retention Test (BVRT) (28). Mental status was assessed by the Mini-Mental State Examination (MMSE) (29).
All test scores were standardized to z scores based on the means and standard deviations (SD) of the baseline measures to facilitate comparisons across all tests. Higher SDs of the TMT-B and BVRT indicate lower performance (higher values indicate longer completion times of the TMT-B and higher number of errors on the BVRT). Higher values for the other tests reflect higher performance.
Adjustment Covariates
Age, sex (male vs female), years of education, race (white vs nonwhite), and vascular burden were adjustment covariates collected at each visit. Vascular burden was defined as the presence of ≥1 cardiovascular risk factors versus none. Cardiovascular risk factors included the presence of hypertension, elevated total cholesterol (≥200 mg/dL of total cholesterol), diabetes, obesity (body mass index of ≥30 kg/m2), and current smoking (self-report smoking status) (30). Hypertension was defined as diastolic blood pressure ≥90 mm Hg, systolic blood pressure ≥140 mm Hg, and or antihypertensive treatment. Diabetes was defined as glycated hemoglobin (HbA1c) ≥6.5% (31).
Statistical Analysis
First, we compared the differences in the covariates for the two timepoints, using chi-square tests for categorical variables and two-sample t tests for continuous variables. Then, we examined cross-sectional and temporal sequences of hearing and cognition. To examine cross-sectional associations between hearing and cognition at each assessment, we used multivariable linear regression models.
To evaluate the temporal sequences of hearing and cognition, we used bivariate autoregressive cross-lagged models from structural equation models (Figure 1). Both autoregressive and cross-lagged associations were examined simultaneously in these models. The autoregressive analyses examined the association between hearing at time t and hearing at time t +1 as well as cognition at time t and cognition at time t + 1. The cross-lagged analyses examined the association between hearing at time t and cognition at time t + 1 while adjusting for cognition at time t, and vice versa. Both autoregressive and cross-lagged associations were adjusted by age, sex, education, and race at Time 1. A significant association between hearing at time t and cognition at time t + 1 would be beyond the variance accounted for by the autoregressive association between cognition at time t and cognition at time t + 1 and covariates. The regression coefficient is equal to the effect size. Statistical analyses were performed using Stata 15.0 (32) and Mplus 8.0 (33).
Figure 1.
Significant cross-lagged and autoregressive associations between hearing and cognition. Solid line: significant associations. Dashed line: nonsignificant associations.
As a secondary analysis, we evaluated the associations of PTA of left and right ears with each cognitive assessment to determine whether patterns of results were similar for each ear.
Sensitivity Analyses
We repeated statistical analyses in three subsamples. The first was restricted to hearing aid nonusers at both timepoints, as this would represent a natural HI history. Although this may be a biased sample as those with hearing aids are typically of a higher socioeconomic status, we also repeated the analyses in a subset without moderate/severe HI. Lastly, as the associations could be driven by those with cognitive impairment, we re-ran the models to see if the inferences changed on the removal of three cognitively impaired participants (six observations).
Results
Characteristics of Study Sample
The characteristics of the study sample are shown in Table 1. The mean age at the first visit was 73.9 (SD = 8.0) years, and there were 130 (41.5%) men in the sample. The mean years of education were 17.2 (SD = 2.4) years, and 68.1% of the sample was white. At the second visit, three (1.0%) had cognitive impairment. The characteristics were similar at both timepoints, except for elevated total cholesterol (Table 1). Mean PTA in the better ear was 28.9 (SD = 13.3) dB HL at baseline and 29.5 (SD = 14.4) dB HL 2 years later.
Table 1.
Characteristics of Baltimore Longitudinal Study of Aging (N = 313)
| Time 1 | Time 2 | p Value for the Difference | |
|---|---|---|---|
| Sample Characteristics | N = 313 | N = 313 | |
| Age, mean (SD) | 73.9 (8.0) | 76.2 (8.1) | <.001 |
| Men, n (%) | 130 (41.5) | 130 (41.5) | 1.000 |
| Education, in years, mean (SD) | 17.2 (2.4) | 17.2 (2.4) | 1.000 |
| Cognitive impairment, n (%) | — | 3 (1.0) | — |
| White, n (%) | 213 (68.1) | 213 (68.1) | 1.000 |
| Cumulative vascular burden†, n (%) | .662 | ||
| 0 | 85 (27.2) | 59 (18.9) | |
| 1 | 130 (41.5) | 133 (42.5) | |
| 2 | 72 (23.0) | 92 (29.4) | |
| 3 | 25 (8.0) | 24 (7.7) | |
| 4 | 1 (0.3) | 5 (1.6) | |
| 5 | 0 (0.0) | 0 (0.0) | |
| Hypertension, n (%) | 155 (49.7) | 165 (52.7) | .448 |
| Elevated total cholesterol, n (%) | 75 (24.0) | 117 (37.4) | <.001 |
| Diabetes, n (%) | 39 (12.5) | 43 (13.7) | .636 |
| Obesity, n (%) | 79 (25.2) | 76 (24.3) | .718 |
| Current smoking, n (%) | 5 (1.6) | 8 (2.6) | .400 |
| Hearing aid use, n (%) | 50 (16.0) | 69 (22.0) | .053 |
| Pure-tone average in the better ear, mean (SD) | 28.9 (13.3) | 29.5 (14.4) | .563 |
| Mild hearing impairment‡, n (%) | 106 (34.2) | 96 (30.8) | .362 |
| Moderate hearing impairment§, n (%) | 59 (19.0) | 73 (23.4) | .191 |
| Severe hearing impairment||, n (%) | 3 (1.0) | 3 (1.0) | 1.000 |
Notes: †Sum of presence of hypertension, elevated total cholesterol, diabetes, obesity, and current smoking.
‡≥25–40 decibels of hearing level (dB HL) in better ear.
§≥41–70 dB HL in better ear.
||>70 dB HL in better ear.
Cognitive test scores for TMT-B, DSST, CVLT, and MMSE are shown for both visits in Supplementary Table 1, and ranges for all continuous variables are provided in Supplementary Table 2.
The Cross-sectional Association Between Hearing and Cognition
The results from cross-sectional associations between hearing and cognition at each visit are shown in Table 2. For both visits, a 10-dB HL PTA difference (poorer hearing) was not associated with cognitive performance.
Table 2.
Cross-sectional Associations† of Pure-Tone Average (per 10 Decibels of Hearing Level) With Cognition at Times 1 and 2 in the Baltimore Longitudinal Study of Aging (n = 313)
| Cognitive Tests‡ | Time 1 | Time 2 | ||||||
|---|---|---|---|---|---|---|---|---|
| β | Standard Error | T-statistic | p Value | β | Standard Error | T-statistic | p Value | |
| Trail-Making Test Part B§ | 0.061 | 0.062 | 0.981 | .327 | 0.053 | 0.055 | 0.954 | .340 |
| Digit Symbol Substitution Test | −0.056 | 0.052 | −1.078 | .281 | −0.007 | 0.045 | −0.164 | .870 |
| California Verbal Learning Test (CVLT) Immediate Recall | −0.021 | 0.055 | −0.383 | .702 | −0.027 | 0.047 | −0.569 | .570 |
| CVLT Short Delay Recall | 0.014 | 0.057 | 0.253 | .800 | −0.066 | 0.050 | −1.336 | .181 |
| CVLT Long Delay Recall | −0.002 | 0.058 | −0.035 | .972 | −0.062 | 0.051 | −1.220 | .222 |
| Digit Span Forward | −0.019 | 0.060 | −0.311 | .756 | −0.065 | 0.048 | −1.337 | .181 |
| Digit Span Backward | −0.040 | 0.058 | −0.701 | .483 | −0.067 | 0.045 | −1.483 | .138 |
| Benton Visual Retention Test§ | −0.038 | 0.053 | −0.719 | .472 | −0.024 | 0.051 | −0.476 | .634 |
| Mini-Mental State Examination | −0.110 | 0.058 | −1.908 | .056 | −0.036 | 0.058 | −0.618 | .537 |
Notes: †Adjusted by baseline age, sex, race, vascular burden, and education.
‡Standardized to baseline mean and standard deviation.
§Higher scores indicate worse performance.
The Temporal Sequences of Hearing and Cognition
After covariate adjustment, the autoregressive associations (eg, the association of a measure with itself at follow-up) of hearing and cognitive measures across all timepoints were significant at p < .001 (data not shown).
Table 3 shows the results of the autoregressive, cross-lagged model to evaluate whether hearing was associated with change in cognition or vice versa. None of the baseline cognitive tests were associated with change in hearing status (Table 3). Poorer hearing was associated with declines primarily on verbal memory tests, that is the CVLT immediate, short-delay, and long-delay recall, and Digit Span Forward (Table 3). Poorer hearing was associated with a decline of 0.073 SD of the CVLT immediate recall (SE = 0.032, p = .024), a decline of 0.134 SD on the CVLT short-delay recall (SE = 0.043, p = .002), a decline of 0.080 SD of CVLT delayed recall (SE = 0.032, p = .012), and a decline of 0.074 SD of Digit Span Forward from baseline (SE = 0.029, p = .011). As a secondary analysis, we examined the associations of poorer hearing of left and right ears separately with cognition. Results were substantively similar, except for poorer left-ear hearing being associated with increases in the 0.063 SD of total number of errors of the BVRT from baseline (SE = 0.031, p = .043; Table 3).
Table 3.
Cross-lagged Associations† Between Pure-Tone Average (PTA) per 10 Decibels of Hearing Level and Cognition‡ at Times 1 and 2 in the Baltimore Longitudinal Study of Aging (n = 313)
| β | SE | T-statistic | p Value | β | SE | T-statistic | p Value | ||
|---|---|---|---|---|---|---|---|---|---|
| PTA of better ear | |||||||||
| TMT-B§ → PTA | −0.030 | 0.023 | −1.309 | .190 | PTA → TMT-B§ | 0.066 | 0.041 | 1.594 | .111 |
| DSST → PTA | −0.009 | 0.024 | −0.375 | .707 | PTA → DSST | 0.000 | 0.021 | −0.019 | .985 |
| CVLT Immediate → PTA | −0.024 | 0.023 | −1.055 | .291 | PTA → CVLT Immediate | −0.064 | 0.029 | −2.179 | .029 |
| CVLT Short-Delay → PTA | −0.038 | 0.023 | −1.665 | .096 | PTA → CVLT Short-Delay | −0.089 | 0.030 | −2.990 | .003 |
| CVLT Long-Delay → PTA | −0.008 | 0.023 | −0.365 | .715 | PTA → CVLT Long-Delay | −0.078 | 0.031 | −2.509 | .012 |
| Digit Span Forward → PTA | −0.019 | 0.023 | −0.804 | .421 | PTA → Digit Span Forward | −0.074 | 0.029 | −2.544 | .011 |
| Digit Span Backward → PTA | −0.031 | 0.023 | −1.350 | .177 | PTA → Digit Span Backward | −0.048 | 0.032 | −1.517 | .129 |
| BVRT§ → PTA | 0.028 | 0.023 | 1.256 | .209 | PTA → BVRT§ | 0.056 | 0.034 | 1.628 | .104 |
| MMSE → PTA | 0.002 | 0.023 | 0.088 | .930 | PTA → MMSE | −0.059 | 0.046 | −1.278 | .201 |
| PTA of right ear | |||||||||
| TMT-B§ → PTA | −0.052 | 0.029 | −1.805 | .071 | PTA → TMT-B§ | 0.078 | 0.040 | 1.952 | .051 |
| DSST → PTA | 0.004 | 0.030 | 0.151 | .880 | PTA → DSST | 0.002 | 0.021 | 0.101 | .919 |
| CVLT Immediate → PTA | −0.032 | 0.029 | −1.112 | .266 | PTA → CVLT Immediate | −0.059 | 0.029 | −2.065 | .039 |
| CVLT Short-Delay → PTA | −0.027 | 0.029 | −0.921 | .357 | PTA → CVLT Short-Delay | −0.093 | 0.029 | −3.201 | .001 |
| CVLT Long-Delay → PTA | −0.005 | 0.029 | −0.182 | .856 | PTA → CVLT Long-Delay | −0.066 | 0.031 | −2.155 | .031 |
| Digit Span Forward → PTA | −0.013 | 0.029 | −0.432 | .666 | PTA → Digit Span Forward | −0.069 | 0.028 | −2.435 | .015 |
| Digit Span Backward → PTA | −0.041 | 0.029 | −1.432 | .152 | PTA → Digit Span Backward | −0.032 | 0.031 | −1.006 | .314 |
| BVRT§ → PTA | 0.034 | 0.029 | 1.181 | .238 | PTA → BVRT§ | 0.021 | 0.034 | 0.620 | .536 |
| MMSE → PTA | 0.016 | 0.029 | 0.554 | .579 | PTA →MMSE | −0.026 | 0.046 | −0.574 | .566 |
| PTA of left ear | |||||||||
| TMT-B§ → PTA | −0.013 | 0.024 | −0.556 | .578 | PTA → TMT-B§ | 0.051 | 0.037 | 1.369 | .171 |
| DSST → PTA | 0.011 | 0.024 | 0.457 | .647 | PTA → DSST | 0.000 | 0.019 | 0.012 | .990 |
| CVLT Immediate → PTA | −0.007 | 0.024 | −0.300 | .764 | PTA → CVLT Immediate | −0.049 | 0.027 | −1.841 | .066 |
| CVLT Short-Delay → PTA | −0.013 | 0.023 | −0.570 | .568 | PTA → CVLT Short-Delay | −0.068 | 0.027 | −2.480 | .013 |
| CVLT Long-Delay → PTA | −0.002 | 0.015 | −0.155 | .877 | PTA → CVLT Long-Delay | −0.098 | 0.039 | −2.508 | .012 |
| Digit Span Forward → PTA | −0.021 | 0.024 | −0.866 | .387 | PTA → Digit Span Forward | −0.063 | 0.026 | −2.376 | .018 |
| Digit Span Backward → PTA | −0.017 | 0.024 | −0.720 | .471 | PTA → Digit Span Backward | −0.039 | 0.029 | −1.367 | .172 |
| BVRT§ → PTA | −0.042 | 0.023 | −1.800 | .072 | PTA → BVRT§ | 0.063 | 0.031 | 2.026 | .043 |
| MMSE → PTA | 0.001 | 0.024 | 0.049 | .961 | PTA → MMSE | −0.048 | 0.042 | −1.143 | .253 |
Notes: BVRT = Benton Visual Retention Test, CVLT = California Verbal Learning Test, DSST = Digit Symbol Substitution Test, MMSE = Mini-Mental State Examination, SE = standard error, TMT-B = Trail-making Test Part B.
Bold values indicate p < .05.
†Adjusted by baseline age, sex, race, vascular burden, and education.
‡Standardized to baseline mean and standard deviation.
§Higher standard deviations indicate worse performance.
Sensitivity Analyses
We examined the cross-sectional associations of hearing with cognition in three subsamples: hearing aid nonusers (n = 244), those without moderate/severe hearing loss at either timepoint (n = 229), and those without cognitive impairment 2 years after baseline (n = 310). We did not find any cross-sectional associations between hearing and cognition in any of these subsamples (Supplementary Tables 3–5).
Then, we evaluated the cross-lagged associations of hearing with cognition in these three subsamples. Among hearing aid nonusers, baseline cognition was not associated with hearing changes. Poorer baseline hearing was associated with declines in z scores in CVLT short-delay recall (β = −0.099, SE = 0.042, p = .018), Digit Span Forward (β = −0.137, SE = 0.042, p = .001), and MMSE (β = −0.175, SE = 0.061, p = .004) from baseline (Supplementary Table 6). Among those without moderate/severe hearing loss at either timepoint, there was a bidirectional association between CVLT short-delay recall and PTA (Time 1: β = −0.047, SE = 0.023, p = .041; Time 2: β = −0.116, SE = 0.056, p = .038). Poorer baseline hearing was associated with declines in SD of Digit Span Forward (β = −0.174, SE = 0.057, p = .002), increases in SD in BVRT total errors score (β = 0.168, SE = 0.068, p = .014), and worse mental status (β = −0.289, SE = 0.077, p < .001) from baseline (Supplementary Table 7). Among those without cognitive impairment, the findings mirrored those of the main analyses (Supplementary Table 8).
Discussion
In a sample of nondemented community-dwelling older adults, we examined the association between hearing and cognition cross-sectionally and longitudinally to determine the association of one with change in the other. Cross-sectionally, we did not find any associations of poorer hearing with cognition at either visit. Longitudinally, we found that poorer baseline hearing was associated with declines in immediate, short-, and long-delayed recall and Digit Span Forward 2 years later. Baseline cognitive performance was not associated with change in hearing 2 years later.
Our cross-sectional findings were inconsistent with other previous studies examining the effects of any HI on cognitive performance. Deal and coworkers (8) found an association of HI with longitudinal decline in delayed word recall test and incidental learning. A previous cross-sectional BLSA study reported that greater HI was associated with lower scores on measures of mental status, memory, and executive function (4), yet we did not replicate these findings, possibly due to differences in baseline sample characteristics. Our sample was older with poorer hearing, higher hearing aid use, fewer whites and males, and lower prevalence of hypertension and diabetes. However, HI was marginally associated with baseline mental status, similar to results from the initial cross-sectional study.
Our cross-lagged findings suggest that poorer baseline hearing in the better ear may affect verbal memory performance (immediate, short and long delay) and attention to auditory stimuli (Digit Span Forward) 2 years later. This could be resultant of participants with HI unable to hear the spoken words during an auditory test. Although hearing could be a confound in cognitive performance, the neurocognitive testing in the BLSA is carried out under standardized psychometric testing conditions—for example a quiet room devoid of background noise and distractions, and with highly experienced psychometrists speaking face-to-face with participants.
Another possibility is that HI degrades auditory input, so that a participant can still hear what is being said, for example repeating a sentence the psychometrist said, but the additional processing needed to account for the degraded auditory signal greatly reduces the ability to encode in working memory. Poorer hearing may have prevented participants from optimally encoding the list of words on the CVLT and the length of the digit span. We also examined a test of visual memory (BVRT) and tests of visual executive function (TMT-B and DSST). Although both the estimates for the associations of poorer baseline hearing in the better ear with change in visual memory and executive function tests were in the same direction, the magnitude was not significant.
We examined these associations within two subsamples: those who do not use hearing aids and those without moderate/severe hearing loss. In both subsamples, poorer baseline hearing was associated with declines in z scores of short-delay verbal memory and Digit Span Forward from baseline. Among those who do not use hearing aids, poorer baseline hearing was associated with decline in mental status. Within a sample restricted to those without moderate/severe HI, poorer baseline hearing was associated with both declines in visual memory and mental status 2 years after baseline. This finding suggests that moderate/severe HI individuals could have been influential in the main study findings. In the main study, we did not find any associations of poorer hearing with visual memory and mental status.
We evaluated whether baseline hearing measures were associated with change in mental status 2 years after baseline or vice versa. Although we did not find any associations in our main analysis, poorer hearing predicted decline in mental status when the sample was restricted to those without moderate/severe HI and hearing aid nonusers, a finding consistent with other studies (4,5,8). Mechanisms underlying this association are unclear. One hypothesized mechanism could be that HI and cognitive decline share common factors, that is age, vascular burden, and social factors, that affect both conditions (11,13,14). Another hypothesized mechanism is HI and effortful listening increase cognitive load in which greater cognitive resources are needed for peripheral auditory processing, thus affecting working memory and other cognitive processes (8,34). However, we did not find any associations between peripheral auditory function and tests of executive function, that is TMT-B and DSST, consistent with findings from another study (8). In addition, HI may contribute to brain structural changes and atrophy, constituting a “second hit” on the brain in conjunction with other brain insults from cerebrovascular disease and Alzheimer’s disease pathology (11).
We did not find any associations of cognitive performance with change in hearing status. The lack of association may contribute to evidence that suggests that HI is a risk factor of cognitive decline, especially because we found robust associations of hearing impairment with cognitive decline in a 2-year period. On the other hand, our inability to find associations between cognition and subsequent HI could be due to the short time interval in which we were examining the longitudinal relationships between hearing and cognition. It is possible that longer follow-up intervals might demonstrate prospective associations between cognition and HI.
We examined whether greater peripheral HI in the right ear would be associated with steeper declines in cognition. We found that poorer baseline hearing in both right and left ears was associated with steeper declines in short- and long-delay verbal memory and attention to auditory stimuli 2 years after baseline, suggesting no laterality differences in these associations. In addition, poorer baseline hearing in the left ear was associated with declines in visual memory.
Strengths of the study included a sample of older adults aged 60 years and older with repeated assessments of objective peripheral auditory function and neuropsychological tests collected at multiple timepoints, resulting in 626 observations. The strengths of our study build on previous literature in this area, that is Berlin Aging Study, in which a moderate correlation was found between sensory loss and cognitive decline (19).
Our study has several limitations. First, there is limited heterogeneity of the sample, because BLSA participants are highly educated and have a high socioeconomic status. Second, a 2-year time lag might not have been long enough for cross-lagged associations, yet we saw that auditory measures were associated with declines in tests of verbal memory and attention to auditory stimuli after 2 years, not vice versa. Third, hearing aid use increased from 16.0% to 22.0%, possibly attenuating our findings. We examined hearing aid nonusers at both timepoints to examine if hearing aid use confounded our results. We found similar results, but poorer hearing was associated with worse mental status among hearing aid nonusers. In addition, we restricted the sample to those who did not have moderate/severe HI. Within this subsample, poorer hearing was associated with declines in verbal memory, attention to auditory stimuli, visual memory, and mental status. Fourth, an unmeasured variable could have affected the associations of HI with cognitive change.
In summary, we found that HI is associated with declines in verbal memory and auditory stimuli for attention after a 2-year period, but not executive function with auditory or visual inputs. In contrast, we did not find evidence of cognitive functioning being associated with changes in hearing over a 2-year follow-up period. These findings suggest that decreases in peripheral auditory function may affect verbal cognitive abilities in a short period of time. These findings warrant further investigation of the longitudinal associations of poorer hearing with cognitive change throughout adulthood.
Funding
This research was supported by the Intramural Research Program of the National Institutes of Health (NIH), National Institute on Aging (NIA). Dr. Deal was supported by NIH/NIA grant K01AG054693. The authors of this manuscript include employees of the Intramural Research Program of the NIA. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Supplementary Material
Acknowledgments
We would like to thank the participants and staff of the BLSA and the Laboratory of Behavioral Neuroscience.
Conflicts of interest statement
None reported.
References
- 1. Pickles J. An Introduction to the Physiology of Hearing. Bingley, UK: Emerald; 2008. [Google Scholar]
- 2. Lin FR, Niparko JK, Ferrucci L. Hearing loss prevalence in the United States. Arch Intern Med. 2011;171:1851–1852. doi:10.1001/archinternmed.2011.506 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Lin FR. Hearing loss and cognition among older adults in the United States. J Gerontol A Biol Sci Med Sci. 2011;66:1131–1136. doi:10.1093/gerona/glr115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Lin FR, Ferrucci L, Metter EJ, An Y, Zonderman AB, Resnick SM. Hearing loss and cognition in the Baltimore Longitudinal Study of Aging. Neuropsychology. 2011;25:763–770. doi:10.1037/a0024238 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Deal JA, Betz J, Yaffe K, et al. ; Health ABC Study Group. Hearing impairment and incident dementia and cognitive decline in older adults: the Health ABC Study. J Gerontol A Biol Sci Med Sci. 2017;72:703–709. doi:10.1093/gerona/glw069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Lin FR, Metter EJ, O’Brien RJ, Resnick SM, Zonderman AB, Ferrucci L. Hearing loss and incident dementia. Arch Neurol. 2011;68:214–220. doi:10.1001/archneurol.2010.362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Gallacher J, Ilubaera V, Ben-Shlomo Y, et al. Auditory threshold, phonologic demand, and incident dementia. Neurology. 2012;79:1583–1590. doi:10.1212/WNL.0b013e31826e263d [DOI] [PubMed] [Google Scholar]
- 8. Deal JA, Sharrett AR, Albert MS, et al. Hearing impairment and cognitive decline: a pilot study conducted within the Atherosclerosis Risk in Communities Neurocognitive Study. Am J Epidemiol. 2015;181:680–690. doi:10.1093/aje/kwu333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gurgel RK, Ward PD, Schwartz S, Norton MC, Foster NL, Tschanz JT. Relationship of hearing loss and dementia: a prospective, population-based study. Otol Neurotol. 2014;35:775–781. doi:10.1097/MAO.0000000000000313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lin FR, Yaffe K, Xia J, et al. ; Health ABC Study Group. Hearing loss and cognitive decline in older adults. JAMA Intern Med. 2013;173:293–299. doi:10.1001/jamainternmed.2013.1868 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Lin FR, Albert M. Hearing loss and dementia—who is listening?Aging Ment Health. 2014;18:671–673. doi:10.1080/13607863.2014.915924 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Loughrey DG, Kelly ME, Kelley GA, Brennan S, Lawlor BA. Association of age-related hearing loss with cognitive function, cognitive impairment, and dementia: a systematic review and meta-analysis. JAMA Otolaryngol Head Neck Surg. 2018;144:115–126. doi:10.1001/jamaoto.2017.2513 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Humes LE, Young LA. Sensory–cognitive interactions in older adults. Ear Hear. 2016;37 (suppl 1):52S–61S. doi:10.1097/AUD.0000000000000303 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Li KZ, Lindenberger U. Relations between aging sensory/sensorimotor and cognitive functions. Neurosci Biobehav Rev. 2002;26:777–783. doi:10.1016/S0149-7634(02)00073-8 [DOI] [PubMed] [Google Scholar]
- 15. Whitson HE, Cronin-Golomb A, Cruickshanks KJ, et al. American Geriatrics Society and National Institute on Aging bench-to-bedside conference: sensory impairment and cognitive decline in older adults. J Am Geriatr Soc. 2018; 66:2052–2058. doi:10.1111/jgs.15506 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lindenberger U, Baltes PB. Sensory functioning and intelligence in old age: a strong connection. Psychol Aging. 1994;9:339–355. doi:10.1037/0882-7974.9.3.339 [DOI] [PubMed] [Google Scholar]
- 17. Baltes PB, Lindenberger U. Emergence of a powerful connection between sensory and cognitive functions across the adult life span: a new window to the study of cognitive aging?Psychol Aging. 1997;12:12. doi:10.1037/0882-7974.12.1.12 [DOI] [PubMed] [Google Scholar]
- 18. Humes LE, Busey TA, Craig J, Kewley-Port D. Are age-related changes in cognitive function driven by age-related changes in sensory processing?Atten Percept Psychophys. 2012;75:508–524. doi:10.3758/s13414-012-0406-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Lindenberger U, Ghisletta P. Cognitive and sensory declines in old age: gauging the evidence for a common cause. Psychol Aging. 2009;24:1–16. doi:10.1037/a0014986 [DOI] [PubMed] [Google Scholar]
- 20. Hugdahl K. Symmetry and asymmetry in the human brain. Eur Rev. 2005;13:119–133. doi:10.1017/S1062798705000700 [Google Scholar]
- 21. Kimura D. Functional asymmetry of the brain in dichotic listening. Cortex. 1967;3:163–178. doi:10.1016/S0010-9452(67)80010–8 [Google Scholar]
- 22. Shock N, Greulich R, Andres R, et al. Normal Human Aging: the Baltimore Longitudinal Study of Aging. Washington, DC: US Government Printing Office; 1984. [Google Scholar]
- 23. Reitan RM. Trail Making Test: Manual for Administration and Scoring. Tucson, AZ: Reitan Neuropsychology Laboratory; 1986. [Google Scholar]
- 24.Smith A. The symbol-digit modalities test: a neuropsychologic test of learning and other cerebral disorders. In: Learning Disorders. Seattle, WA: Special Child Publications; 1968: 83–91. [Google Scholar]
- 25. Blackburn HL, Benton AL. Revised administration and scoring of the digit span test. J Consult Psychol. 1957;21:139–143. doi:10.1037/h0047235 [DOI] [PubMed] [Google Scholar]
- 26. Delis DC. CVLT-II: California Verbal Learning Test: Adult Version. San Antonio, TX: Psychological Corporation; 2000. [Google Scholar]
- 27. Delis DC, Kramer JH, Kaplan E, Thompkins BAO.. CVLT: California Verbal Learning Test–Adult Version: Manual. San Antonio, TX: Psychological Corporation; 1987. [Google Scholar]
- 28. Benton A. Revised Visual Retention Test: Clinical and Experimental Applications. New York, NY: Psychological Corporation; 1974. [Google Scholar]
- 29. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198. doi:10.1016/0022- 3956(75)90026-6 [DOI] [PubMed] [Google Scholar]
- 30. Gottesman RF, Schneider AL, Zhou Y, et al. Association between midlife vascular risk factors and estimated brain amyloid deposition. JAMA. 2017;317:1443–1450. doi:10.1001/jama.2017.3090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Chamberlain JJ, Rhinehart AS, Shaefer CF Jr, Neuman A. Diagnosis and management of diabetes: synopsis of the 2016 American Diabetes Association Standards of Medical Care in Diabetes. Ann Intern Med. 2016;164:542–552. doi:10.7326/M15-3016 [DOI] [PubMed] [Google Scholar]
- 32. StataCorp. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC; 2017. [Google Scholar]
- 33. Muthén L, Muthén B.. Mplus User’s Guide. 8th ed Los Angeles, CA: Muthén & Muthén; 1998–2017. [Google Scholar]
- 34. Tun PA, McCoy S, Wingfield A. Aging, hearing acuity, and the attentional costs of effortful listening. Psychol Aging. 2009;24:761–766. doi:10.1037/a0014802 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.