Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Oct 1.
Published in final edited form as: J Aging Health. 2022 Jan 31;34(6-8):765–774. doi: 10.1177/08982643211063338

The Longitudinal Association of Late-Life Visual and Hearing Difficulty and Cognitive Function: The Role of Social Isolation

Jeremy B Yorgason 1, Corinna Trujillo Tanner 2, Stephanie Richardson 3, Melanie S Hill 4, Brian Stagg 5, Markus Wettstein 6, Joshua R Ehrlich 7
PMCID: PMC9801657  NIHMSID: NIHMS1854881  PMID: 35100881

Abstract

Objectives:

Sensory impairments are prevalent among older adults and have been associated with cognitive challenges in later life, yet mechanisms are less well understood. We examined the mediating role of social isolation in the longitudinal relationship between self-reported sensory difficulty and impaired cognitive functioning among older adults.

Methods:

Data were taken from the NHATS Study, an annual survey of Medicare beneficiaries age ≥ 65. Participants (N = 6,338) provided data at Rounds 5, 6, and 7 (2015, 2016, 2017). Structural equation models were estimated to test longitudinal direct and indirect associations.

Results:

All sensory difficulties were negatively associated with all cognitive functioning measures cross-sectionally through social isolation. Longitudinally, VD and DSD were indirectly associated with cognitive functioning across time. HD had no longitudinal indirect associations with cognitive functioning through social isolation.

Discussion:

Social isolation is an important pathway through which late-life vision difficulty is associated with decreased cognitive function.

Keywords: Hearing, Vision, Cognitive Function, Social Factors, Longitudinal Analysis

Introduction

Sensory impairments, including visual, hearing, and dual sensory impairment, are highly prevalent among older adults, and the number of individuals affected is rapidly increasing as the population ages (Swenor et al., 2013). Self-reported visual impairment (VI), which affects 9% of adults 65 and older in the U.S. (Patel et al., 2020), and self-reported hearing impairment (HI), which affects 31% of adults age 60 to 69 and 63.1% of those 70 and older (Goman & Lin, 2016), have a negative impact on many domains of health, including cognitive functioning (Bainbridge & Wallhagen, 2014; Swenor et al., 2020; Shah et al., 2020; Shukla et al., 2020; Whitson et al., 2018; Lin & Albert, 2014). These challenges are compounded in the case of dual sensory impairment, which is present in up to 11% of adults age 60 and older (Swenor et al., 2013).

All sensory impairments increase risk for social isolation (Shah et al., 2020; Shukla et al., 2020), which cross-sectional research has demonstrated to be an important mediator of the relationship between sensory impairment and cognitive impairment (Whitson et al., 2018). However, when considering whether the association of sensory impairment with cognitive decline occurs via social isolation, it is essential to consider longitudinal data to provide evidence of processes that develop over time and to establish directionality. Establishing these associations in a nationally representative sample may allow for greater generalizability of findings. This study examined longitudinal associations between self- reported sensory difficulty and cognitive impairments indirectly through social isolation in a nationally representative sample of older adults.

Direct Links Between Sensory Impairment and Cognitive Functioning

VI, HI, and DSI are independently associated with cognitive impairment (Lin & Albert, 2014; Whitson et al., 2018; Zheng et al., 2018). The relationships between sensory impairments and cognitive impairment are complex, and numerous mechanisms have been hypothesized. Sensory impairments and cognitive impairment may share common causes, such as changes of the central nervous system (Baltes & Lindenberger, 1997), vascular disease, and neurodegeneration (Dichgans & Leys, 2017). Sensory impairments may also increase the cognitive load required for sensory processing, which may lead to poor cognitive outcomes (Pigeon et al., 2019) or result in direct alteration of brain structure, including both regional and whole-brain atrophy, due to decreased afferent sensory input (McEwen, 2000). Conversely, preserving sensory function may be protective of cognitive function and brain structure in vision (Tamura et al., 2004) and hearing (Dawes et al., 2015).

Sensory Impairment and Cognitive Functioning: The Role of Social Isolation

Social isolation has been proposed as a potential mediator of the association between decreased sensory and cognitive functions (Livingston et al., 2020). Vision and hearing are basic forms of interchange between individuals and their environment. When typical forms of communication are interrupted, there is evidence that social isolation often results (National Academies of Sciences, Engineering, and Medicine, 2020; Shah et al., 2020; Shukla et al., 2020). Individuals with self-reported vision impairments are less likely to engage in out-of-home leisure and social activities (Heyl et al., 2005). Socially engaging activities provide cognitive stimulation, which may protect against cognitive decline (Lövdén et al., 2005). This is consistent with the cascade hypothesis, in which cognitive decline is driven by a lack of cognitive stimulation related to sensory loss (Varadaraj et al., 2020). Established literature suggests that a lack of social engagement, or social isolation, may be a key modifiable risk factor that links sensory impairment with later cognitive impairment (Livingston et al., 2020). The current study fills this gap by providing a longitudinal perspective of social isolation as a mechanism linking sensory difficulty and cognitive functioning among older adults.

Current Study

In this study, we examined the longitudinal relationship between self-reported vision difficulty (VD), hearing difficulty (HD) and dual sensory difficulty (DSD), and cognitive function using a nationally representative sample from the National Health and Aging Trends Study (NHATS). We hypothesized that having one or more self-reported sensory difficulties would be associated with poorer cognitive functioning cross-sectionally, and longitudinally, across 1 and 2 years and that these relationships would be associated indirectly through social isolation.

Methods

Study Sample

The National Health and Aging Trends Study (NHATS) is a nationally representative panel study of Medicare beneficiaries age ≥ 65 years. The study began in 2011 with a total of 8,245 persons. In 2015, the sample was replenished, including 8,334 persons. The current analysis utilized 3 years of data spanning Round 5 in 2015 to Round 7 in 2017. Because NHATS data are publicly available and de-identified, the institutional review board at the first author’s institution deemed this study exempt. To decrease potential confounding and endogeneity, we excluded those individuals who did not live in a community setting (n = 1264) and those who additionally had probable dementia at Round 5 (n = 732), thus allowing for the inference of directionality of the effect of social isolation on later cognitive functioning. The remaining participants made up the analytic sample (N = 6338). Among the analytic sample, there was some attrition over time, with approximately 1% missing on measures of cognitive functioning at Round 5, approximately 14% missing on those measures at Round 6, and approximately 24% missing at Round 7.

Measures

Cognitive Function Measures

We employed cognitive measures collected in NHATS corresponding to orientation, executive function, and learning/memory (Kasper et al., 2013). To measure orientation, respondents were asked to recite the date, president, and vice president of the U.S. Executive function was assessed using a clock-drawing test. Learning/memory was evaluated using a delayed word-recall test.

Measures of Sensory Difficulty

Three separate variables were used to assess self-reported VD, HD, and DSD. VD was measured using a total of three items. If the participant reported being blind, or that they were unable to see well enough (including when using corrective lenses – glasses or contacts) to recognize someone across the street or to read newspaper print, they were then coded as having a VD. This method was used in prior studies using NHATS data (Ehrlich et al., 2019; Frank et al., 2019; Xiang et al., 2020).

A dichotomous measure of HD was constructed using four items. If the respondent reported difficulty with any of the items, they were then coded as having hearing impairment. Hearing impairment questions related to whether respondents could “hear well enough to carry on a conversation in a quiet room,” “hear well enough to carry on a conversation in a room with a radio or TV playing,” and “hear well enough to use the telephone,” and an item assessing whether participants were deaf (“yes” coded as 1, “no” coded as 0). Individuals were characterized as having HD only if hearing problems were severe enough to impact their functioning (whether or not they wore a hearing aid). People with hearing aids but who did not report problems with these listed items were not coded as having a hearing difficulty for this study.

Self-reported DSD was indicated if the participant reported having both HD and VD. In the current sample, 10% (n = 578) reported HD, 6% (n = 343) reported VD, and 3% (n = 140) reported DSD.

Social Isolation

Social isolation was assessed using a 5-item scale validated in NHATS (Cudjoe et al., 2020). Social isolation scores represent a count of whether or not participants endorsed specific situations. Participants were coded as having some social isolation if they (a) lived alone, (b) if they talked to one person or fewer about “important matters” during the last year, (c) if they did not attend religious services in the past month, (d) if they did not attend clubs/classes/organized activities in the past month, and (e) if they did not participate in volunteer work during the past month. Points were summed so that higher scores indicated greater social isolation, with scores ranging from 0 to 5.

Covariates

We included conceptually relevant covariates, including age, gender, marital status, race, education, smoking status, and diagnoses of heart disease, hypertension, diabetes, and stroke. Education, rather than income, was included as an indicator of socioeconomic status (Darin-Mattsson et al., 2017) because of its association with cognitive function (Heyl & Wahl, 2003).

Statistical Analysis

Descriptive statistics, bivariate correlations, and mean difference tests were estimated to provide initial information about the sample. Structural equation modeling, using Mplus (Muthén & Muthén, 2017), was used to test study hypotheses. As seen in Figure 1, we examined associations of sensory difficulty in Round 5 with social isolation in Rounds 5 and 6 and cognitive function in Rounds 5, 6, and 7. We simultaneously modeled direct (not attributable to social isolation) and indirect (attributable to social isolation) associations cross-sectionally and across 1- and 2-year periods (see Hayes, 2018). Models were adjusted for all covariates and prior-wave measures of outcomes. VD and HD were estimated together in the same models, allowing for direct comparisons between the two. Models including DSD were estimated separately due to its correlation with VD (r = .52) and HD (r = .47). Estimates were calculated using full-information maximum likelihood, an estimation approach that uses all available data to address missing data. Bootstrapping with 5,000 draws was used to adjust the standard errors associated with indirect effects (Hayes & Preacher, 2010).

Figure 1.

Figure 1.

Open in a new tab

Diagram of the Guiding Conceptual and Analytic Model

Results

Characteristics of the Study Sample

Table 1 summarizes the baseline characteristics of the 6,338 participants in this study. Participants ranged in age from 65 to 89 years, with the largest proportion in the 70 to 74 age range (26%). The majority were female (57%) and married (50%). Those with DSD were older and reported higher social-isolation scores than those with no or only one sensory difficulty. Participants without sensory difficulties had higher cognitive performance scores than those with HD, VD, or DSD, and those with DSD had significantly lower cognitive functioning scores than those with HD alone.

Table 1.

Descriptive Statistics from Round 5 of NHATS and ANOVA Analyses of Main Study Variables (N (%) or M (SD)).

Variable NHATS Total
Study Sample		 No Sensory
Difficulty
(n = 4,511)		 Hearing
Difficulty
(n = 578)		 Visual
Difficulty
(n = 343)		 Dual Sensory
Difficulty
(n = 140)		 Mean
Difference
p-values
Age Groups a b c d p < .001
 65–69 1,013 (15.98%) 829 (18.38%) 68 (11.76%) 50 (14.58%) 18 (12.86%)
 70–74 1,651 (26.05%) 1,310 (29.04%) 109 (18.86%) 78 (22.74%) 17 (12.14%)
 75–79 1,428 (22.53%) 1,053 (23.34%) 111 (19.20%) 86 (25.07%) 22 (15.71%)
 80–84 1,126 (17.77%) 753 (16.69%) 125 (21.63%) 55 (16.03%) 23 (16.43%)
 85–89 728 (11.49%) 410 (9.09%) 88 (15.22%) 52 (15.16%) 29 (20.71%)
 ≥ 90 392 (6.18%) 156 (3.46%) 77 (13.32%) 22 (6.41%) 31 (22.14%)
Sex a b c a p < .001
 Male 2,726 (43.01%) 1,823 (40.41%) 299 (51.73%) 112 (32.65%) 52 (37.14%)
 Female 3,612 (56.99%) 2,688 (59.59%) 279 (48.27%) 231 (67.35%) 88 (62.86%)
Education Level a b c c p < .001
 No Degree 1,211 (19.11%) 803 (17.80%) 144 (24.91%) 112 (32.65%) 61 (43.57%)
 High School 3,119 (26.38%) 1,164 (25.80%) 170 (29.41%) 102 (29.74%) 38 (27.14%)
 Some college 2,355 (26.87%) 1,257 (27.87%) 139 (24.05%) 76 (22.16%) 19 (13.57%)
 College degree 2,986 (25.59%) 1,192 (36.42%) 115 (19.90%) 46 (13.41%) 19 (13.57%)
Marital Status a a b b p < .001
 Married 1,760 (50.42%) 1,305 (50.72%) 159 (51.79%) 69 (37.70%) 27 (34.62%)
 Not Married 1,731 (49.58%) 1,268 (49.28%) 148 (48.21%) 114 (62.30%) 51 (65.38%)
Smokes Now a a a a p = .073
 No 5,863 (92.51%) 4,140 (91.80%) 540 (93.43%) 305 (88.92%) 132 (94.29%)
 Yes 474 (7.15% 370 (8.20%) 38 (6.57%) 38 (11.02%) 8 (5.71%)
Race/Ethnicity a b c c p < .001
 White, non-Hispanic 4,391 (69.28%) 3,031 (68.89%) 434 (76.14%) 180 (53.89%) 79 (58.09%)
 Non-White 1,799 (28.38%) 1,369 (31.11%) 136 (23.86%) 154 (46.11%) 57 (41.91%)
Social Isolation 2.46 (1.27) 2.42a 2.72b 2.74b 3.12c p < .001
Clock-Drawing 3.77 (1.07) 3.83a 3.69b 3.39c 3.16c p < .001
Delayed Word Recall 3.61 (2.01) 3.71a 3.16b 3.20b 2.47c p < .001
Orientation 6.69 (1.44) 6.79a 6.32b 6.17b 5.58c p < .001
Open in a new tab

Notes. Differences in superscripts a, b, c, and d represent significant mean group differences.

Associations of Self-Reported Sensory Difficulty, Social Isolation, and Cognitive Function

As seen in Models 1-3 of Table 2, VD in Round 5 was significantly associated with learning/memory scores 1 year later (β = −.03, p < .01), concurrent orientation (β = −.06, p < .001), and all three waves of executive function (Round 5: β = −.09, p < .001; Round 6: β = −.07, p < .001; Round 7: β = −.05, p < .01). In all three models, VD in Round 5 was associated with greater social isolation concurrently (β’s = .03 to .04, p < .01) and 1 year later (β’s = .03, p < .01). HD in Round 5 was significantly associated with concurrent, but not future, learning/memory (β = −.04, p < .01), and orientation (β = −.06, p < .001), and with social isolation (β’s = .05 to .06, p < .001). As seen in Models 1-3 of Table 3, DSD in Round 5 was significantly associated with concurrent, but not future, learning/memory (β = −.03, p < .05), orientation (β = −.07, p < .001), and executive-function (β = −.05, p < .01). In Tables 2 and 3, greater social isolation at Round 5 was significantly associated with lower scores on nearly all cognitive indicators in all rounds. Predictors account for between 30% and 47% of the variance in cognitive-functioning scores at Round 7, suggesting that the models provide a robust view of cognitive functioning in the context of important related factors.

Table 2.

Standardized Regression Coefficients From Structural Equation Model of Self-Reported Vision and Hearing Difficulty Predicting Cognitive Functioning Indirectly Through Social Isolation.

Predictor Variable Model 1
Learning/memory ß
(SE)		 Model 2
Orientation
ß (SE)		 Model 3
Executive
Function
ß (SE)
VD →R5 CI −.012 (.01) −.061*** (.01) −.087*** (.01)
VD →R6 CI −.032** (.01) −.010 (.01) −.066*** (.01)
VD →R7 CI .012 (.01) −.018 (.01) −.048** (.02)
HD →R5 CI −.039** (.01) −.062*** (.01) −.003 (.01)
HD →R6 CI −.006 (.01) −.011 (.01) .008 (.01)
HD →R7 CI −.015 (.01) .004 (.01) .011 (.02)
Social Isolation →R5 CI −.093*** (.01) −.074*** (.01) −.056*** (.01)
Social Isolation →R6 CI −.081*** (.01) −.029* (.01) −.033* (.01)
Social Isolation →R7 CI −.030** (.01) −.036** (.01) −.024± (.01)
VD →R5 Social Isolation .043*** (.01) .034** (.01) .033** (.01)
VD →R6 Social Isolation .031** (.01) .029** (.01) .029** (.01)
HD →R5 Social Isolation .050*** (.01) .051*** (.01) .058*** (.01)
HD →R6 Social Isolation −.001 (.01) .000 (.01) .002 (.01)
VD →R5 Social Isolation →R5 CI −.004** (.00) −.003* (.00) −.002* (.00)
VD →R6 Social Isolation →R6 CI −.003** (.00) −.001± (.00) −.001± (.00)
VD →R6 Social Isolation →R7 CI −.001* (.00) −.001* (.00) −.001 (.00)
HD →R5 Social Isolation →R5 CI −.005** (.00) −.004** (.00) −.003** (.00)
HD →R6 Social Isolation →R6 CI .000 (.00) .000 (.00) .000 (.00)
HD →R6 Social Isolation →R7 CI .000 (.00) .000 (.00) .000 (.00)
Chi-Square Model Fit 12,247.93*** 12,572.22*** 8,740.76***
RMSEA .02 .02 .02
CFI 1.00 1.00 1.00
R2 of R7 CI .47*** .47*** .30***
Open in a new tab

Note. N = 6,338. VD = vision difficulty; HD = hearing difficulty; R = Round; CI = cognitive impairment; RMSEA = root mean squared error of approximation; CFI = Comparative Fit Index. R5 was in 2015, R6 was in 2016, and R7 was in 2017. Arrows denote a directional regression; multiple arrows denote an indirect relationship. Covariates were age, gender, marital status, race, education, smoking status, and chronic health conditions (heart disease, hypertension, diabetes, and stroke).

±

p < .10

*

p < .05

**

p < .01

***

p < .001.

Table 3.

Standardized Regression Coefficients from Structural Equation Models of Self-Reported Dual Sensory Difficulty Predicting Cognitive Functioning Indirectly Through Social Isolation.

Predictor Variable Model 1
Learning/memory ß
(SE)		 Model 2
Orientation
ß (SE)		 Model 3
Executive
Function
ß (SE)
DSD →R5 CF −.031** (.01) −.065*** (.02) −.051** (.02)
DSD →R6 CF −.018 (.01) −.015 (.01) −.011 (.02)
DSD →R7 CF .002 (.01) −.001 (.02) −.019 (.02)
Social Isolation →R5 CF −.084*** (.01) −.077***(.01) −.064*** (.01)
Social Isolation →R6 CF −.086*** (.01) −.029* (.01) −.041** (.01)
Social Isolation →R7 CF −.032** (.01) −.037** (.01) −.031* (.01)
DSD →R5 Social Isolation .048*** (.01) .049*** (.01) .049*** (.01)
DSD →R6 Social Isolation .022* (.01) .023* (.01) .023* (.01)
DSD →R5 Social Isolation →R5 CF −.004** (.00) −.004** (.00) −.003** (.00)
DSD →R6 Social Isolation →R6 CF −.002* (.00) −.001 (.00) −.001± (.00)
DSD →R6 Social Isolation →R7 CF −.001 (.00) −.001± (.00) −.001 (.00)
Chi-Square 12,468.68*** 12,516.34*** 9,053.17***
RMSEA .02 .02 .03
CFI 1.00 1.00 1.00
R2 of R7 CF .47*** .47*** .30***
Open in a new tab

Note. N = 6,338. DSD = dual sensory difficulty, CF = cognitive function; RMSEA = root mean squared error of approximation, CFI = Comparative Fit Index; R = Round. R5 was in 2015, R6 was in 2016, and R7 was in 2017. Arrows denote a directional regression; multiple arrows denote an indirect relationship. Covariates were age, gender, marital status, race, education, smoking status, and chronic health conditions (heart disease, hypertension, diabetes, and stroke).

±

p < .10

*

p < .05

**

p < .01

***

p < .001.

Mediation Analyses

Results of mediation analyses indicated significant associations between sensory difficulty and cognitive outcomes mediated by social isolation (shown as percentage of total effect mediated) and are presented in Supplemental Table 1 and in Figures 2 and 3. Baseline VD was significantly associated with all three concurrent (Round 5) cognitive-functioning measures through social isolation (learning/memory: 25% of total effect, p < .01; orientation: 5% of total effect, p < .05; executive function: 2% of total effect, p < .05). Longitudinally, baseline VD was associated with learning/memory (9% of total effect, p < .01) 1 year later through Round 6 social isolation, and with learning/memory (9% of total effect, p < .01) and orientation (5% of total effect, p < .01) 2 years later through Round 6 social isolation.

Figure 2. Results from Tests of Indirect Effects of Vision Difficulty on Cognitive Outcomes Through Social Isolation.

Figure 2.

Open in a new tab

Indirect Effect Coefficients

R5 Learning/Memory a1*b1 = −.004 z = −3.418 p = .001 95% CI = −.007, −.002

R5 Orientation a1*b1 = −.003 z = −2.557 p = .011 95% CI = −.005, −.001

R5 Executive Function a1*b1 = −.002 z = −2.271 p = .023 95% CI = −.004, −.001

R6 Learning/Memory a2*b2 = −.003 z = −3.001 p = .003 95% CI = −.004, −.001

R7 Learning/Memory a2*b3 = −.001 z = −2.079 p = .038 95% CI = −.002, .000

R7 Orientation a2*b3 = −.001 z = −2.054 p = .040 95% CI = −.002, .000

Notes. R = Round of NHATS study; CI = confidence interval.

Figure 3. Results from Tests of Indirect Effects of Hearing Difficulty and Dual Sensory Difficulty on Cognitive Outcomes Through Social Isolation.

Figure 3.

Open in a new tab

Indirect Effect Coefficients

Hearing Difficulty

R5 Learning/Memory a1*b1 = −.005 z = −3.380 p = .001 95% CI = −.008, −.002

R5 Orientation a1*b1 = −.004 z = −3.151 p = .002 95% CI = −.007, −.002

R5 Executive Function a1*b1 = −.003 z = −2.895 p = .004 95% CI = −.006, −.001

Dual Sensory Difficulty

R5 Learning/Memory a1*b1 = −.004 z = −3.332 p = .001 95% CI = −.007, −.002

R5 Orientation a1*b1 = −.004 z = −3.189 p = .001 95% CI = −.006, −.002

R5 Executive Function a1*b1 = −.003 z = −2.993 p = .003 95% CI = −.006, −.001

R6 Learning/Memory a2*b2 = −.002 z = −2.040 p = .041 95% CI = −.004, .000

Notes. R = Round of NHATS study; CI = confidence interval.

Baseline HD was associated with all three concurrent (Round 5) cognitive-functioning measures through social isolation (learning/memory: 11% of total effect, p < .01; orientation: 6% of total effect, p < .01; executive function: 50% of total effect, p < .01), though HD was not associated with cognitive function longitudinally. Baseline DSD was associated with all three concurrent cognitive measures through social isolation (learning/memory: 11% of total effect, p < .01; orientation: 6% of total effect, p < .01; executive function: 6% of total effect, p < .01) and was significantly associated longitudinally with learning/memory (10% of total effect, p < .05) 1 year later through Round 6 social isolation. Indirect association effect sizes were generally small (Kenny, 2018).

Discussion

Using data from the nationally representative NHATS, we tested longitudinal associations between self-reported sensory difficulty and multiple measures of cognitive function, as well as whether this association was mediated by social isolation. Findings indicated that all cross-sectional associations between VD, HD, and DSD and cognitive function operated through social isolation. However, longitudinally, only the associations of VD and DSD with cognitive function appeared to be significantly mediated by social isolation. The proportion of indirect to total effects suggests that social isolation accounts for around or less than 10% in most cases, although in one case it accounted for 50% of the total effect of sensory difficulty on cognitive functioning. Taken together, results indicated that social isolation likely plays a small yet consistent role in the sensory–cognitive association.

Visual Difficulty

The finding of a longitudinal association between VD and cognitive function is in agreement with prior studies based on both self-reported VD (Davies-Kershaw et al., 2018; Maharani et al., 2018) and objectively measured visual function (Lee et al., 2020; Naël et al., 2019; Zheng et al., 2018). In fact, a recently published meta-analysis of longitudinal studies found that the odds of impaired cognitive function were significantly higher among adults with VD than among adults with normal vision (OR = 1.7; Vu et al., 2020). The current study builds on this literature, providing evidence of the association of VD with domain-specific cognitive function in a U.S. population-based sample. Notwithstanding the robust and consistent association of poor vision with cognitive function, vision has not yet been widely recognized in summaries of population attributable risk factors for dementia (Livingston et al., 2020).

In our study, VD was longitudinally associated with cognitive functioning 2 years later through social isolation. Few other studies have sought to test hypothesized mediators that might account for the association between visual and cognitive function. A cross-sectional study in Canada showed only weak mediating effects by social isolation between objectively measured sensory and cognitive difficulties (Hämäläinen et al., 2019). The researchers in that study reported that social factors were most important for cognitive abilities for females and older individuals, the latter of which was not surprising, as the sample ranged in age from 45 to 85 years. The current study extends those findings and provides longitudinal evidence supporting these associations. Although we accounted for age, gender, and other covariates, we did not explore how these characteristics might moderate direct and indirect associations between sensory difficulty and cognitive functioning. Future research is needed to better understand the specific groups for whom these associations may be most important. For instance, associations between sensory difficulties and cognition are found to be stronger when individuals have higher neuroticism scores (Gaynes et al., 2013; Wettstein et al., 2016). Additionally, social isolation accounted for only a fraction of the total association between vision and cognition, suggesting a need to investigate additional contributory mechanisms, including other mediators, such as genetics, cognitive reserve, sensory deprivation, and physiological as well as lifestyle factors.

Hearing Difficulty

As hypothesized and supported by prior research (Shukla et al., 2020), HD was associated with higher concurrent levels of impaired cognitive functioning and a significant fraction of this association was attributable to social isolation. Like VD, HD may impede social interactions, and prior research has suggested that this can lead to decreased cognitive functioning (Zheng et al., 2018). However, in the current study, HD was not significantly associated with a longitudinal decline in cognitive function. One possibility is that HD may have been underreported in the current sample. Some prior investigations have found that older adults tend to overestimate their hearing ability (Bainbridge & Wallhagen, 2014).

Dual Sensory Difficulty

Like VD and HD, DSD was associated with impaired cognitive functioning through social isolation cross-sectionally. DSD was also associated with learning/memory across 1 year through social isolation. Because VD and HD were independently associated with impaired cognitive functioning, it was not surprising that DSD was also associated with impaired cognitive functioning. In support of prior research, DSD may be a risk factor for cognitive decline and dementia in later life (Brenowitz et al., 2019).

Based on our findings, it appears likely that sensory difficulties have a considerable impact on concurrent and sometimes longitudinal social interactions and cognitive performance. Prior research has suggested that individuals with DSD may be at greater risk for adverse cognitive outcomes (Brenowitz et al., 2019). At the same time, it is possible that some older adults with sensory difficulties make adaptations to overcome social isolation, which may explain why some effects of sensory difficulties relating to cognitive functioning did not persist over time. Further research examining the nuances of how social networks, social connectedness, and loneliness are impacted by different sensory difficulties over long periods of time is needed to better understand the potential mediating role of these psychosocial constructs on cognitive outcomes.

Strengths and Limitations

Although prior studies have reported an association between sensory difficulties and impaired cognitive functioning, the pathways that account for this association have not been rigorously tested in longitudinal studies. A key strength of the current study is that it investigated the associations between sensory difficulties and cognitive function, as well as the mediating role of social isolation in a single longitudinal model. Additionally, the NHATS data used in this study are generalizable to the Medicare-eligible U.S. population age 65 years and older, which represents a population at high risk for sensory, cognitive, and social dysfunction. There were also several limitations to this study. Data on sensory difficulty and social isolation were based on self-reports, which may be subject to recall and social-desirability biases. Self-reported measures of sensory difficulty may in fact represent distinct latent constructs from objective measures of sensory status (e.g., visual acuity, pure-tone audiometry). However, both types of measures may be important in assessing functional status and outcomes related to sensory health (Gaynes et al., 2013), and there is substantial overlap between self-reported and objectively measured sensory function (Ng & Loke, 2015). Future work could compare associations between cognitive function and these two types of sensory measures. Data may also have been subject to survival bias, wherein the least healthy participants were less likely to continue study participation and contribute complete data. A follow-up analysis suggested that cognitive scores predicted attrition in the current sample. This could have led to an underestimation of the true rates of cognitive decline, biasing results toward the null hypothesis, as both sensory and cognitive challenges are negatively associated with survival (Ehrlich et al., 2021; Smith & Ismail, 2021).

Conclusion

Sensory difficulties impact not only the sight and hearing of older adults, but may also have a profound effect on other aspects of their lives, including social, cognitive, economic, and physical well-being (Burton et al., 2021). Consequently, there has been increasing interest in the association of sensory difficulty with cognition and dementia, because vision and hearing may represent readily modifiable risk factors that could possibly be leveraged to decrease cognitive decline and prevent dementia. This study provides novel evidence on the mediating role of social isolation in the association between sensory difficulty and cognitive function in a nationally-representative sample of older U.S. adults. In fact, social isolation may represent another viable intervention target to promote cognitive health among older adults with sensory difficulty—a group that is at high risk for cognitive decline. Furthermore, interventions which focus on reducing social isolation would be protective of multiple other negative health outcomes (National Academies of Sciences, Engineering, and Medicine, 2020). Research is needed to test additional hypothesized mediating pathways between sensory difficulty and cognitive functioning, strategies to reduce social isolation in older adults with sensory difficulty, particularly vision loss, and, ultimately, to conduct trials to determine whether cognitive decline may be mitigated through interventions to optimize sensory and social function.

Supplementary Material

Suppl Table 1

Acknowledgements

Funding:

This work was supported by the National Institutes of Health [grant number K23EY027848to JRE]; and unrestricted grants from Research to Prevent Blindness, New York, NY, to the Department of Ophthalmology & Visual Sciences, University of Utah and University of Michigan.

Footnotes

Conflicts of Interest:

None.

Contributor Information

Jeremy B. Yorgason, School of Family Life, Brigham Young University, 2079 JFSB, Provo, Utah 84602.

Corinna Trujillo Tanner, School of Nursing, Brigham Young University.

Stephanie Richardson, School of Family Life, Brigham Young University.

Melanie S. Hill, School of Family Life, Brigham Young University.

Brian Stagg, Department of Ophthalmology and Visual Sciences, University of Utah.

Markus Wettstein, Network Aging Research, Heidelberg University, Heidelberg, Germany.

Joshua R. Ehrlich, Department of Ophthalmology and Visual Sciences, University of Michigan.

References

  1. Bainbridge KE, & Wallhagen MI (2014). Hearing loss in an aging American population: Extent, impact, and management. Annual Review of Public Health, 35(1), 139–152. 10.1146/annurev-publhealth-032013-182510 [DOI] [PubMed] [Google Scholar]
  2. Baltes PB, & Lindenberger U (1997). Emergence of a powerful connection between sensory and cognitive functions across the adult life span: A new window to the study of cognitive aging? Psychology and Aging, 12(1), 12–21. https://doi,org-erl.lib.byu.edu/10.1037/0882-7974.12.1.12 [DOI] [PubMed] [Google Scholar]
  3. Brenowitz WD, Kaup AR, Lin FR, & Yaffe K (2019). Multiple sensory impairment is associated with increased risk of dementia among Black and White older adults. The Journals of Gerontology: Series A, 74(6), 890–896. 10.1093/gerona/gly264 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Burton MJ, Ramke J, Marques AP, Bourne RRA, Congdon N, Jones I, Ah Tong BAM, Arunga S, Bachani D, Bascaran C, Bastawrous A, Blanchet K, Braithwaite T, Buchan JC, Cairns J, Cama A, Chagunda M, Chuluunkhuu C, Cooper A, … Faal HB (2020). The Lancet Global Health commission on global eye health: Vision beyond 2020. Lancet Global Health, 9(4), E489–E551. 10.1016/S2214-109X(20)30488-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cudjoe TK, Roth DL, Szanton SL, Wolff JL, Boyd CM, & Thorpe RJ Jr. (2020). The epidemiology of social isolation: National health and aging trends study. The Journals of Gerontology: Series B, 75(1), 107–113. 10.1093/geronb/gby037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Darin-Mattsson A, Fors S, & Karcholt I (2017). Different indicators of socioeconomic status and their relative importance as determinants of health in older age. International Journal for Equity in Health, 16, 173–184. 10.1186/s12939-017-0670-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Davies-Kershaw HR, Hackett RA, Cadar D, Herbert A, Orrell M, & Steptoe A (2018). Vision impairment and risk of dementia: Findings from the English longitudinal study of ageing. Journal of the American Geriatrics Society, 66, 1823–1829. 10.1111/jgs.15456 [DOI] [PubMed] [Google Scholar]
  8. Dawes P, Emsley R, Cruickshanks KJ, Moore DR, Fortnum H, Edmondson-Jones M, McCormack A, & Munro KJ (2015). Hearing loss and cognition: The role of hearing aids, social isolation and depression. PLoS One, 10(3), 1–10. 10.1371/journal.pone.0119616 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dichgans M, & Leys D (2017). Vascular cognitive impairment. Circulation Research, 120, 573–591. 10.1161/circresaha.116.308426 [DOI] [PubMed] [Google Scholar]
  10. Ehrlich JR, Hassan SE, & Stagg BC (2019). Prevalence of falls and fall-related outcomes in older adults with self-reported vision impairment. Journal of the American Geriatrics Society, 67(2), 239–245. https://doi-org.erl.lib.byu.edu/10.1111/jgs.15628 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ehrlich JR, Ramke J, Macleod D, Burn H, Lee CN, Zhang JH, Waldock W, Swenor BK, Gordon I, Congdon N, Burton M, & Evans JR (2021). Association between vision impairment and mortality: A systematic review and meta-analysis. The Lancet Global Health, 9(4), e418–e430. 10.1016/S2214-109X(20)30549-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Frank CR, Xiang X, Stagg BC, & Ehrlich JR (2019). Longitudinal associations of self-reported vision impairment with symptoms of anxiety and depression among older adults in the United States. JAMA Ophthalmology, 137(7), 793–800. 10.1001/jamaophthalmol.2019.1085 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gaynes BI, Shah R, Leurgans S, & Bennett D (2013). Neuroticism modifies the association of vision impairment and cognition among community-dwelling older adults. Neuroepidemiology, 40(2), 142–146. 10.1159/000342762 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Goman AM, & Lin FR (2016). Prevalence of hearing loss by severity in the United States. American Journal of Public Health, 106(10), 1820–1822. 10.2105/AJPH.2016.303299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hämäläinen A, Phillips N, Wittich W, Pichora-Fuller MK, & Mick P (2019). Sensory-cognitive associations are only weakly mediated or moderated by social factors in the Canadian Longitudinal Study on Aging. Scientific Reports, 9(1), 1–8. 10.1038/s41598-019-55696-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hayes AF (2018). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-based Approach (2nd Ed). New York, NY: Guilford Press. [Google Scholar]
  17. Hayes AF, & Preacher KJ (2010). Quantifying and testing indirect effects in simple mediation models when the constituent paths are nonlinear. Multivariate Behavioral Research, 45, 627–660. 10.1080/00273171.2010.498290 [DOI] [PubMed] [Google Scholar]
  18. Heyl V, Wahl H-W, & Mollenkopf H (2005). Visual capacity, out-of-home activities and emotional well-being in old age: Basic relations and contextual variation. Social Indicators Research, 74(1), 159–189. 10.1007/s11205-005-6522-3 [DOI] [Google Scholar]
  19. Heyl V & Wahl H-W (2014). Experiencing age-related vision and hearing impairment: The psychosocial dimension. Journal of Clinical Outcomes Management, 21(7), 323–335. [Google Scholar]
  20. Kasper JD, Freedman VA, & Spillman BC (2013). Classification of persons by dementia status in the National Health and Aging Trends Study. Technical Paper #5. [Google Scholar]
  21. Kenny DA (2012). Mediation. Retrieved from http://davidakenny.net/cm/mediate.htm [Google Scholar]
  22. Kuo PL, Huang AR, Ehrlich JR, Kasper J, Lin FR, McKee MM, Reed NS, Swenor BK, & Deal JA (2021). Prevalence of concurrent functional vision and hearing impairment and association with dementia in community-dwelling Medicare beneficiaries. JAMA Network Open, 4(3), e211558. 10.1001/jamanetworkopen.2021.1558 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lee AT, Richards M, Chan WC, Chiu HF, Lee RS, & Lam LC (2020). Higher dementia incidence in older adults with poor visual acuity. The Journals of Gerontology: Series A, glaa036. 10.1093/gerona/glaa036 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lin FR, & Albert M (2014). Hearing loss and dementia—Who is listening? Aging and Mental Health, 18, 671–673. 10.1080/13607863.2014.915924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, Brayne C, Burns A, Cohen-Mansfield J, Cooper C, Costofreda SG, Dias A, Fox N, Gitlin LN, Howard R, Kales HC, Kivimäk M, Larsen EB, Ogunniyi A, … Mukadam N (2020). Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Advance online publication. Lancet. 10.1016/S0140-6736(20)30367-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lövdén M, Ghisletta P, & Lindenberger U (2005). Social participation attenuates decline in perceptual speed in old and very old age. Psychology and Aging, 20(3), 423–434. 10.1037/0882-7974.20.3.423 [DOI] [PubMed] [Google Scholar]
  27. Maharani A, Dawes P, Nazroo J, Tampubolon G, Pendleton N, & Sense-Cog WP1 Group. (2018). Visual and hearing impairments are associated with cognitive decline in older people. Age and Ageing, 47, 575–581. 10.1093/ageing/afy061 [DOI] [PubMed] [Google Scholar]
  28. McEwen BS (2000). Effects of adverse experiences for brain structure and function. Biological Psychiatry, 48, 721–731. 10.1016/s0006-3223(00)00964-1 [DOI] [PubMed] [Google Scholar]
  29. Muthén LK, & Muthén BO (2017). Mplus: Statistical analysis with latent variables: User’s guide (Version 8). Muthén & Muthén. [Google Scholar]
  30. Naël V, Peres K, Dartigues JF, Letenneur L, Amieva H, Arlco A, Scherlen AC, Tzourio C, Berr C, Carriere I, Delcourt C, Helmer C, Bertelsen G, Cosh S, Cougnard-Gregoire A, Dawes P, Delcourt C, Constantinidou F, Helmer C, … & von Hanno T (2019) Vision loss and 12-year risk of dementia in older adults: The 3C Cohort Study. European Journal of Epidemiology, 34(2), 141–152. 10.1007/s10654-018-00478-y [DOI] [PubMed] [Google Scholar]
  31. National Academies of Sciences, Engineering, and Medicine; Division of Behavioral and Social Sciences and Education; Health and Medicine Division; Board on Behavioral, Cognitive, and Sensory Sciences; Board on Health Sciences Policy; & Committee on the Health and Medical Dimensions of Social Isolation and Loneliness in Older Adults. (2020). Social isolation and loneliness in older adults: Opportunities for the health care system. National Academies Press. [PubMed] [Google Scholar]
  32. Ng JH-Y, & Loke AY (2015). Determinants of hearing-aid adoption and use among the elderly: A systematic review. International Journal of Audiology, 54(5), 291–300. 10.3109/14992027.2014.966922 [DOI] [PubMed] [Google Scholar]
  33. Patel N, Stagg BC, Swenor BK, Zhou Y, Talwar N, & Ehrlich JR, (2020). Association of co-occurring dementia and self-reported visual impairment with activity limitations in older adults. JAMA Ophthalmology, 138, 756–763. 10.1001/jamaophthalmol.2020.1562 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pigeon C, Li T, Moreau F, Pradel G, & Marin-Lamellet C (2019). Cognitive load of walking in people who are blind: Subjective and objective measures for assessment. Gait & Posture, 67, 43–49. 10.1016/j.gaitpost.2018.09.018 [DOI] [PubMed] [Google Scholar]
  35. Shah K, Frank CR, & Ehrlich JR (2020). The association between vision impairment and social participation in community-dwelling adults: A systematic review. Eye, 34(2), 290–298. https://doi-org.erl.lib.byu.edu/10.1038/s41433-019-0712-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Shukla A, Harper M, Pedersen E, Goman A, Suen JJ, Price C, Applebaum J, Hoyer M, Lin FR, & Reed NS (2020). Hearing loss, loneliness, and social isolation: A systematic review. Otolaryngology-Head & Neck Surgery, 162, 622–633. https://doi-org.erl.lib.byu.edu/10.1177/0194599820910377 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Smith EE, & Ismail Z (2021). Mortality risk models for persons with dementia: A systematic review. Journal of Alzheimer’s Disease, 80(1), 103–111. 10.3233/JAD-201364 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Swenor BK, Lee MJ, Varadaraj V, Whitson HE, & Ramulu PY (2020). Aging with vision loss: A framework for assessing the impact of visual impairment on older adults. The Gerontologist, 60, 989–995. 10.1093/geront/gnz117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Swenor BK, Ramulu PY, Willis JR, Friedman D, & Lin FR (2013). Research letters: The prevalence of concurrent hearing and vision impairment in the United States. JAMA Internal Medicine, 173(4), 312–313. 10.1001/jamainternmed.2013.1880 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tamura H, Tsukamoto H, Mukai S, Kato T, Minamoto A, Ohno Y, Yamashita H, & Mishima HK (2004). Improvement in cognitive impairment after cataract surgery in elderly patients. Journal of Cataract & Refractive Surgery, 30, 598–602. 10.1016/j.jcrs.2003.10.019 [DOI] [PubMed] [Google Scholar]
  41. Varadaraj V, Munoz B, Simonsick EM, & Swenor BK (2020). Vision impairment and participation in cognitively stimulating activities: The Health ABC Study. Journals of Gerontology: Series A, 76(5), 835–841. 10.1093/gerona/glaa184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Vu TA, Fenwick EK, Gan ATL, Man REK, Tan BKJ, Gupta P, Ho KC, Reyes-Ortiz CA, Trompet S, Gussekloo J, O’Brien JM, Mueller-Schotte S, Wong TY, Tham YC, Cheng C-Y, Lee ATC, Rait G, Swenor BK, Varadaraj V, …Lamoureux EL (2020). The bidirectional relationship between vision and cognition: A systematic review and meta-analysis. Ophthalmology, 128(7), 981–992. 10.1016/j.ophtha.2020.12.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wayne RV, & Johnsrude IS (2015). A review of causal mechanisms underlying the link between age-related hearing loss and cognitive decline. Ageing Research Reviews, 23(B), 154–166. 10.1016/j.arr.2015.06.002 [DOI] [PubMed] [Google Scholar]
  44. Wettstein M, Kuźma E, Wahl H-W, & Heyl V (2016). Cross-sectional and longitudinal relationship between neuroticism and cognitive ability in advanced old age: The moderating role of severe sensory impairment. Aging & Mental Health, 20(9), 918–929. 10.1080/13607863.2015.1049119 [DOI] [PubMed] [Google Scholar]
  45. Whitson HE, Cronin-Golomb A, Cruickshanks KJ, Gilmore GC, Owsley C, Peelle JE, Recanzone G, Sharma A, Swenor B, Yaffe K, & Lin FR (2018). American Geriatrics Society and National Institute on Aging Bench-to-Bedside Conference: Sensory impairment and cognitive decline in older adults. Journal of the American Geriatrics Society, 66, 2052–2058. 10.1111/jgs.15506 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Xiang X, Freedman VA, Shah K, Hu RX, Stagg BC, & Ehrlich JR (2020). Self-reported vision impairment and subjective well-being in older adults: A longitudinal mediation analysis. The Journals of Gerontology: Series A, 75, 589–595. https://doi-org.erl.lib.byu.edu/10.1093/gerona/glz148 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zheng DD, Swenor BK, Christ SL, West SK, Lam BL, & Lee DJ (2018). Longitudinal associations between visual impairment and cognitive functioning: The Salisbury Eye Evaluation Study. JAMA Ophthalmology, 136, 989–995. https://doi-org.erl.lib.byu.edu/10.1001/jamaophthalmol.2018.2493 [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.

Supplementary Materials

Suppl Table 1
NIHMS1854881-supplement-Suppl_Table_1.docx (14.9KB, docx)

RESOURCES