The Association Between Hearing Loss and Multiple Negative Emotional States in the U.S. Hispanic Population

Abstract

Objective:

Hearing loss (HL) has been linked to commonly-studied detrimental mood states such as loneliness and depression. However, its relationship with other negative emotions remained largely unstudied. We explore the association between HL and anxiety, anger, hostility, poor self-esteem, and pessimism in a national cohort of U.S. Hispanic adults.

Study Design:

Cross-sectional study

Settings:

Multi-centered U.S. national epidemiologic study (Hispanic Community Health Study; HCHS)

Methods:

Subjects were ages 18–75 with completed audiometric and emotional survey data. Multivariable regressions controlling for age, gender, and education were conducted to analyze the association between HL, measured by 4-frequency pure tone average (PTA), and emotional states. States included anxiety (Speilberger Trait Anxiety Scale-10), anger (Speilberger Trait Anger Scale), hostility (Cook Medley Cynicism Scale-13); poor self-esteem (Self-Esteem Scale-10); and pessimism (Revised Life Orientation Test).

Results:

4,120–4,341 participants met inclusion criteria, depending on the specific survey; average age was 46.7 years (SD=13.7) and average PTA was 13.8 dB (SD=10.1). Controlling for age, gender, and education, HL was associated with all outcomes. Specifically, for every 10 dB worsening in HL, the anxiety score worsened by 0.41 (0.23–0.60), the anger score worsened by 0.40 (0.22–0.58), the hostility score worsened by 0.16 (0.04–0.27), the self-esteem score worsened by 0.25 (0.12–0.38), and the pessimism score worsened by 0.17 (0.04–0.30) (all P<0.01).

Conclusions:

HL is related to numerous negative mood states beyond loneliness and depression. This includes worse anxiety, anger, hostility, self-esteem and pessimism. Future studies should investigate whether treating HL improves negative emotional states.

Introduction

Hearing loss (HL) is the third most common chronic physical condition in U.S. adults¹, and yet only 11% who have at least unilateral hearing loss and subjective moderate difficulty hearing use hearing aids^2,3. This represents a significant proportion of the population at risk for the negative sequalae associated with HL⁴. Many studies have recently shown strong associations between HL and numerous deleterious medical comorbidities including depression, cognitive decline, dementia, and most recently, serious mental illness^5–13. HL has also been linked with poorer interpersonal support smaller social networks, and more familial conflict^14–16. The far-reaching effects of HL in other areas of life is still being explored.

The relationship between HL and negative emotional states is a largely unprobed area of research that might yield insight into how hearing loss contributes to depression, cognitive decline, and poorer socialization. These harmful states do not occur in a hearing-challenged individual spontaneously; rather they are likely the result of years of sustained negative emotional states that ultimately result in withdrawing from their social life, pulling back from their systems of support, and possibly lashing out against loved ones. These poorer socialization states may then result in a reduction of stimulating cognitive input, which is a risk factor for depression and dementia.

This study aimed to quantify the association between adult HL and negative emotional states using five independent surveys: the Speilberger Trait Anxiety Scale-10 queried anxiety¹⁷; the Speilberger Trait Anger Scale probed anger¹⁸; the Cook Medley Cynicism Scale-13 assessed hostility¹⁹; the Self-Esteem Scale-10 measured self-esteem²⁰; and the Revised Life Orientation Test evaluated optimism²¹. To our knowledge, anger, hostility, self-esteem, and optimism have never been studied in relation to HL before. Higher levels of anxiety have previously been linked to HL, but HL and anxiety have not been studied in a Hispanic population specifically. Hispanics are the second fastest growing racial or ethnic subpopulation in the United States²² and may be disproportionally affected by hearing loss compared to non-White counterparts²³ due to aging; from 2010 to 2020, the percentage of U.S. adults over the age of 18 who were of Hispanic or Latino origin jumped from 14.2% to 16.8%, while the proportion of other racial or ethnic groups comprising the U.S. adult population fell. A rapidly growing and aging population is likely to experience more hearing loss comparatively than other populations.

Determining the relationship between HL and these negative emotional states is not only critical in elucidating the causal pathway between HL and medical comorbidities like dementia and depression, but also vital in understanding the myriad ways HL may affect people.

Methods

This was a cross-sectional epidemiological study conducted in the Hispanic Community Health Study/Study of Latinos (HCHS-SOL). HCHS-SOL includes a comprehensive physical exam, including an audiometric hearing test, and a variety of surveys. Participants are U.S. Hispanics ages 18–76 residing in Chicago, Miami, San Diego, and the Bronx area of New York City. HCHS-SOL is a longitudinal multi-centered study; however, only the first wave of data (2008–2011) was available, so cross-sectional analysis was performed. The inclusion criteria for this specific analysis was HCHS participants with both audiometric and emotional survey data, explained in detail below.

Multivariable linear regressions controlling for age, gender, and education were conducted to analyze the association between hearing loss and different measures of emotional states. Hearing loss was measured by the 4-frequency air conduction pure tone average (PTA) in the better ear at 500, 1000, 2000 and 4000 Hz and reported continuously in decibels (dB). This hearing test was conducted using supra-aural headphones in a sound booth per HCHS-SOL protocols. Education was a continuous variable representing number of years of schooling completed. The emotional state surveys included the Speilberger Trait Anxiety Scale-10 measuring anxiety; the Speilberger Trait Anger Scale measuring anger; the Cook Medley Cynicism Scale-13 measuring hostility; the Self-Esteem Scale-10 measuring self-esteem; and the Revised Life Orientation Test measuring optimism.

The Speilberger Trait Anxiety Scale-10 consisted of ten self-descriptive statements probing anxiety concerns versus self-contentment; participants were asked to rate the veracity of each statement to their own experience right now in that moment. Options were “not at all”, “somewhat”, “moderately so”, and “very much so.” The composite score was continuous on a scale of 0–30, with higher values indicating more feelings of anxiety¹⁷.

The Speilberger Trait Anger Scale consisted of ten self-descripting statements probing the frequency and degree to which each participant had anger; participants were asked to rate the frequency with which they felt the particular statement was true on a scale of “never” to “almost always”. The composite score was continuous on a scale of 0–30, with higher values indicating more frequent and intense anger¹⁸.

The Cook Medley Cynicism Scale-13 consisted of 13 true-false statements probing participants’ cynicism in different areas of life. The composite score was continuous on a scale of 0–13, with higher scores indicating higher cynical hostility¹⁹.

The Self-Esteem Scale-10 consisted of ten self-descriptive statements probing an individual’s feelings of self-worth. Participants rated the accuracy of each statement on a scale from “strongly agree” to “strongly disagree”. The composite score was continuous on a scale of 0–30, with higher scores indicated higher self-esteem²⁰.

The Revised Life Orientation Test consisted of nine self-descriptive statements probing participants’ outlooks on life. Participants chose to agree or disagree with the statements on a scale from “I disagree a lot” to “I agree a lot”. The composite score was continuous on a scale of 0–24, with higher scores indicated higher levels of optimism²¹.

Statistical analysis, as described above, was performed in R version 4.0.5 (R Foundation for Statistical Computing, Vienna, Austria) in RStudio version 1.4.1106 (Boston, MA). The Columbia University IRB deemed this study of anonymous data to be not human subjects research and thus exempt from IRB approval.

Results

13,551 participants age 18 years and older had audiometric data. Of those, 4,404 had completed at least 1 socialization instrument, comprising the primary analytics sample. 60.6% (n=2,670) were female, and average age was 46.7 years (SD=13.7 years, range: 18 to 75 years). Average PTA was 13.8 dB (SD=10.1 dB, range: −10 to 105 dB) (Table 1).

Table 1. Sample demographics.

The anxiety score was measured by the Speilberger Anxiety Scale-10 where higher scores correspond to more anxiety. The anger score was measured by the Speilberger Anger Scale where higher scores correspond to more frequent and intense anger. The hostility score was measured by the Cook Medley Cynicism Scale-13 where higher scores correspond to more hostility. The self-esteem score was measured by the Self-Esteem-Scale where higher scores correspond to more self-esteem. The optimism score was measured by the Revised Life Orientation Test where higher scores correspond to more optimism.

Total Sample, N	4,404
Gender
Female (%)	2,670 (60.6)
Male (%)	1,734 (39.4)
Age years
Mean (SD, range)	46.7 (13.7, 18 to 75)
Hearing Level dB
Mean (SD, range)	13.8 (10.1, −10 to 105)
Anxiety Score pts
N (%)	4,303 (97.7)
Mean (SD, range)	7.7 (5.3, 0 to 29)
Anger Score pts
N (%)	4,340 (98.5)
Mean (SD, range)	6.8 (5.3, 0 to 30)
Hostility Score, pts
N (%)	4,120 (93.6)
Mean (SD, range)	8.3 (3.3, 0 to 13)
Self-Esteem Score, pts
N (%)	4,318 (98.0)
Mean (SD, range)	21.5 (4.1, 7 to 30)
Optimism Score, pts
N (%)	4,341 (98.6)
Mean (SD, range)	16.7 (3.8, 1 to 24)

The Speilberger Trait Anxiety Scale-10 was completed by 4,303 participants. Average composite score was 7.7 (SD=5.3, range: 0–29). Controlling for confounders (age, gender, and education), for every 10 dB worsening in hearing, the score increased by 0.41 (95% CI: 0.23–0.60; P<0.01). This relationship is depicted visually in Figure 1A. In summary, as hearing loss worsens, individuals are more likely to feel more anxiety (Table 2).

Figure 1.

Increased hearing loss is associated with more anxiety (A), anger (B), hostility (C), poorer self-esteem (D), and less optimism (E), adjusted for confounders. Hearing and emotional-states are plotted on the Y-axes respectively. Every dot represents an individual person with the adjusted line-of-best-fit in blue (all p < .01).

Table 2. Associations between hearing loss and emotional states for every 10 dB worsening in hearing loss, controlling for age, gender, and education.

As hearing loss worsens, the anxiety, anger and hostility scores increase significantly (positive coefficients); as hearing loss worsens, the self-esteem and optimism scores decrease significantly (negative coefficients).

Test	Coefficient*	95% CI	P Value
Anxiety	0.41	0.23 to 0.60	<0.01
Anger	0.40	0.22 to 0.58	<0.01
Hostility	0.16	0.04 to 0.27	<0.01
Self-Esteem	−0.25	−0.28 to −0.12	<0.01
Optimism	−0.17	−0.30 to −0.04	<0.01

^*All regressions are multivariable and controlled for covariates.

The Speilberger Trait Anger Scale was completed by 4,340 participants. Average composite score was 6.8 (SD=5.3, range 0–30). Controlling for confounders, for every 10 dB worsening in hearing, the score increased by 0.40. (95% CI: 0.22–0.58; P<0.01). This relationship is depicted visually in Figure 1B. In summary, as hearing loss worsens, individuals are more likely to feel anger more intensely and frequently (Table 2).

The Cook Medley Cynicism Scale-13 (measuring hostility) was completed by 4,120 participants. Average composite score was 8.3 (SD=3.3, range 0–13). Controlling for confounders, for every 10 dB worsening in hearing, the score increased by 0.16 (95% CI: 0.04–0.27; P<0.01). This relationship is depicted visually in Figure 1C. In summary, as hearing loss worsens, individuals are more likely to feel more cynical hostility (Table 2).

The Self-Esteem Scale-10 was completed by 4,318 participants. Average composite score was 21.5 (SD=4.1, range 7–30). Controlling for confounders, for every 10 dB worsening in hearing, the score decreased by 0.25 (95% CI: 0.12–0.38; P<0.01). This relationship is depicted visually in Figure 1D. In summary, as hearing loss worsens, individuals are more likely to have poorer self-esteem (Table 2).

The Revised Life Orientation Test (measuring optimism) was completed by 4,341 participants. Average composite score was 16.7 (SD=3.8, range 1–24). Controlling for confounders, for every 10 dB worsening in hearing, the score decreased by 0.17 (95% CI: 0.04–0.30; P<0.01). This relationship is depicted visually in Figure 1E. In summary, as hearing loss worsens, individuals are more likely to feel less optimistic and more pessimistic (Table 2).

In a sensitivity analysis, study location (Chicago, Miami, San Diego, and the Bronx) was added as a covariate to all models. This resulted in a minimal attenuation of the estimates, except for hostility, which decreased to from 0.16 to 0.12. All estimates all remained significant. We similarly included a sensitivity analysis for marital status as covariate and saw no change in the estimates. Under 1% of participants wore hearing aids. However, to ensure that these participants didn’t attenuate the effects (as they would be treating their exposure), we ran a sensitivity analysis with hearing aid use as a covariate to all models. This also resulted in a minimal attenuation (under 10%) of the estimates with all remaining highly significant.

Discussion

HL was significantly associated with all measured negative emotional states, controlling for potential confounders. Specifically, HL was associated with more anxiety, more intense and more frequent anger, more cynical hostility, poorer self-esteem, and less optimism. This follows other papers that have shown that HL is associated with other negative socioemotional states. This includes worse socialization (less interpersonal support, weaker social networks, more family conflict)¹⁶ and loneliness.²⁴ There is also evidence of a relationship between HL and clinical disorders such depression^5,6,25–27, even when levels of HL are minimal.²⁸ The relationship between HL and negative emotional states observed in the present study may lie on a causal pathway between HL and depression.

This paper adds to the greater literature that HL may be linked to deleterious conditions rather than exist as an isolated communication or quality-of-life issue. Beyond socioemotional states, HL has also been linked to dementia, cognitive decline, serious mental illness, hospitalizations, and falls^{13,25,29–39}. The cognitive relationship is present even for minimal level of HL.^30,40–42 Disturbingly, even though older adults are at risk for many of these disorders, their HL may be less likely to be treated than younger individuals.⁴ These data contribute to the growing field of the potential impact of HL beyond audition and the importance of equitable treatment in often overlooked groups.

The relationships between HL and emotional states were not only statistically significant, but also clinically meaningful. The reported associations in Table 2 were for a 10 dB worsening in hearing. However, this represents a modest drop of less than a single category of HL (e.g., from mild to moderate HL). Across a greater spectrum of human hearing, for example, from normal to severe HL, one would predict a more meaningful worsening in emotional state. This can be visualized in Figure 1, which is also controlled for confounders. An individual with severe HL (80 dB), on average, would have an anger score of about 10 points compared to an individual with normal HL (10 dB) who would have an anger score of about 7 points, representing a 3-point difference in score. This 3-point difference corresponds to 10% of the entire 0 to 30 anxiety scale range. While this may seem modest at first, the standard deviation of the anger scores was only 5.3 points, since most individuals do not have extremely high scores. Thus, the 3-point difference in anger from normal to severe hearing loss represents more than half a standard deviation difference in the anger scale. This suggests a true real-world effect. The other emotional state outcomes had a similarly meaningful relationship.

These emotional outcomes that were related to increased HL can be further categorized into two broad types: increased detrimental mood states and decreased positive mood states. The association between HL and anxiety, anger, and hostility can be characterized as the former, while the association between HL and poorer self-esteem and less optimism fall into the latter. Increased levels of anxiety, anger, and hostility have been linked with numerous deleterious health conditions including a predisposition towards addiction, increased days of work missed, and increased psychiatric healthcare costs^43,44.

The decrease in positive mood states associated with increased HL is significant for an implied paucity of personal emotional fortitude. Those who do not have reserves of optimism or self-esteem are likely to react more poorly to stress which HL can also exacerbate creating a vicious cycle. Such individuals may demonstrate emotional lability and outbursts that can put a strain on social relationships¹⁵. This may be a way that HL predisposes such people to reclusion, isolation, and then depression¹⁴. The exact pathway between HL, depression, and social isolation has yet to be elucidated, but it is likely bidirectional and mediated in by biopsychosocial feedback loops⁴⁵. It’s conceivable that these loops are disinhibited by a lack of positive emotional states, namely optimism and self-esteem, and further heightened by a stress response to HL that includes increased negative emotional states, like anger, anxiety, and hostility, in the place of positive emotional responses.

This is the first study exploring HL and other mood states beyond depression and anxiety. To our knowledge, anger, hostility, self-esteem, and optimism have not been analyzed in relation to HL. This is not the first study to examine HL and anxiety⁴⁶, but it is the first study to do so in a Hispanic only population, which is important because Hispanics have been a historically neglected ethnic group in research. A prior study examined anxiety, measured by the General Anxiety Disorder-7 (GAD-7) in adolescents ages 12–18 at a metropolitan otology clinic in the Southeastern United States. They found that adolescents with severe to profound hearing loss had higher rates of anxiety⁴⁷. In the Health, Aging, and Body Composition (Health ABC) study, the odds of anxiety were 1.32 times higher in those with mild HL (25–40 dB) and 1.59 times higher in those with moderate or greater HL (>40 dB), compared to no HL.⁴⁸ Finally, a longitudinal population study of 10,566 Taiwanese adults explored the reverse causal relationship and found an increased risk of HL among those with a diagnosed anxiety disorder⁴⁹.

This study has several limitations. First, it is an observational study that can only identify association; no causation can be inferred. Furthermore, it is cross-sectional; the observations bear no temporality, which also makes it difficult to infer causation (i.e., it is not possible to show that HL occurred before the negative emotional states occurred). While we cannot rule out the possibility of reverse causation, whereby negative emotional states cause HL, this seems less probable. While highly speculative, it is theoretically possible, for example, that an angry individual might listen to louder music and thus have higher risk of noise-induced hearing loss. However, given the possibility that HL might lead to negative emotional states as well as the low risk of treating hearing loss with hearing aids, it seems reasonable to recommend hearing testing and treatment in clinical practice to maintain emotional well-being in later life.

In addition, the emotional surveys queried how an individual felt about anxiety, anger, hostility, self-esteem, and optimism on the specific day of intake, and do not attempt to capture emotional well-being beyond that point in time. While we controlled for potential confounders in our analysis, it is not possible to control for all possible confounders. Covariates not included in this exploration due to lack of data, but could be useful in further analyses, include use of health services and status of living alone or with others.

Additionally, the proportion of this particular population that uses hearing rehabilitation devices such as hearing aids is unfortunately low, less than 1%, obfuscating an exploration of whether such interventions might mitigate the associations seen here between hearing loss and negative emotional states. Finally, that this study was performed in HCHS-SOL in a Hispanic population limits the generalizability of these findings to the rest of the population. Importantly, these findings also cannot be applied to hearing impaired individuals living in Deaf communities because such communities have entirely different communications strategies, namely American Sign Language. In contrast, non-culturally Deaf individuals who lose hearing over time nearly always remain in hearing communities and thus struggle with their original societal form of communication.

However, these limitations also provide opportunities for further research, such as confirming findings in additional ethnic/racial groups. Additionally, HCHS-SOL will ultimately contain a second wave of data that would allow for longitudinal analysis of HL and emotional states. Future studies should also examine treating HL and evaluate subsequent emotional and socialization measures to look for improvement.

In conclusion, HL is associated with many harmful emotional states largely not previously described including anxiety, anger, hostility, poorer self-esteem, and pessimism. HL is a diagnosis with potential impact far beyond audition. Understanding the myriad ways HL may burden our patients is important not only in elucidating the mechanisms behind HL and serious medical comorbidities such as depression, but also in promoting appropriate awareness, prevention, and timely treatment for vulnerable individuals.