Hearing difficulties and feelings of social isolation among Canadians aged 45 or older
by Pamela L. Ramage-Morin
The people who tend to thrive as they age are those who remain socially engaged.Note 1 They have a network of family and friends that allows them to participate in social life and achieve a sense of belonging and purpose.Note 2Note 3 Alternatively, people may be “socially isolated,” lacking social contact, support, and a sense of belonging.Note 4Note 5Note 6 “Social isolation” does not pertain to those who have voluntarily disconnected; rather, socially isolated people have an unmet need for meaningful social interactions, which is often identified as loneliness.Note 4Note 7 Those who are socially isolated are more likely to experience a poor quality of life, morbidity, and mortality.Note 2Note 8Note 9Note 10Note 11Note 12Note 13Note 14 Social isolation among the elderly is a particular concern,Note 13Note 15 as it is estimated that more than 30% of Canadian seniors are at high risk.Note 7
Hearing loss and associated communication difficulties can interfere with social activities and integration.Note 16 In 2012/2013, an estimated 4.5 million adults (19%) had some hearing loss in the range associated with normal speech; 8.4 million (35%) had high-frequency hearing loss, which is often related to aging.Note 17Note 18 At ages 70 to 79, 65% experienced loss in the speech frequency range, and almost everyone (94%) had some high-frequency hearing loss.Note 17 Because of a tendency to deny or minimize hearing loss and the insidious progress of the condition, only a fraction of people with loss actually reported hearing or communication difficulties—4% of adults, or fewer than a million.Note 16Note 17Note 19 As Canada’s senior population increases from around 6 million in 2015 to a projected 9 million in 2030, the number of people with hearing difficulties is expected to rise.Note 20
This study examines associations between hearing difficulties and social isolation. The presence of difficulties is based on self-reported ability to understand conversations, and therefore, reflects a functional limitation rather than a biological loss in hearing acuity. Hearing difficulties are further classified as corrected (able to hear with a hearing aid) or not (cannot hear in some situations, even with a hearing aid).Note 21
Social isolation has been conceptualized and measured9 in a number of ways.Note 6Note 7Note 9Note 22Note 23 For the present analysis, rather than objective measures such as network size or frequency of participation, the focus is on perceived or subjective social isolation, measured as a combination of loneliness and a weak sense of community belonging. Perceived social isolation reflects how people feel about their relationships and participation, and whether their desired circumstances differ from their actual situation.Note 4Note 22Note 24
Associations between hearing impairment and social isolation or loneliness have been explored in a variety of populations, although not in Canada, and often with smaller sample sizes.Note 5Note 16Note 25Note 26Note 27 This study presents information on the issue among Canadians aged 45 or older, based on a large sample representative of the household population across the 10 provinces.
The data are from the 2008/2009 Canadian Community Health Survey–Health Aging (CCHS–HA), a cross-sectional survey that targeted people aged 45 or older living in private dwellings in the 10 provinces. The survey excluded residents of the three territories, First Nations reserves, certain remote regions, and institutions, and full-time members of the Canadian Forces. Data were collected from December 1, 2008 through November 30, 2009, primarily using computer-assisted personal interviews. The combined household- and person-level response rate was 74.4%. Proxy respondents numbered 689 and comprised 2.2% of the sample (Appendix Table A). Because social isolation was determined from subjective questions that were not administered to these respondents, they were excluded, leaving a study sample of 30,176. Details about the CCHS–HA are available on the Statistics Canada website (www.statcan.gc.ca).
Levels of functional impairment for hearing, vision, speech, mobility, dexterity, pain, emotion, and cognition were based on the Health Utilities Index-Mark 3 (HUI3).Note 28Note 29 Each HUI3 attribute has five or six levels with corresponding utility-based scores ranging from 0.00 (most impaired) to 1.00 (no impairment). Continuous variables for each attribute were created using their respective utility scores for the logistic regression models. Prevalence estimates for hearing difficulties were based on an ordinal variable that distinguished between no (level 1), mild (level 2) and moderate/severe (levels 3 to 6) impairment.Note 30 The other types of functional impairment were dichotomized as no/mild (levels 1, 2) versus moderate/severe (levels greater than 2), except for cognition where levels 1 to 3 corresponded to no/mild impairment, and levels greater than 3, moderate/severe impairment.Note 30
Replicating the approach of Corna et al.,Note 21 hearing ability was classified as no difficulty, corrected, not corrected, or cannot hear at all, using the response patterns (1 = yes, 2 = no, 6 = not applicable) to the HUI3 hearing questions: Are you usually able to hear what is said in a group conversation with at least three other people without a hearing aid? (Q1); with a hearing aid? (Q2); Are you able to hear at all? (Q3); Are you usually able to hear what is said in a conversation with one other person in a quiet room without a hearing aid? (Q4); with a hearing aid? (Q5) (Text table 1).
Incontinence was established by asking: “Do you suffer from urinary incontinence?” Respondents were instructed to respond “yes” if their condition had been diagnosed by a health professional and had lasted, or was expected to last, at least six months.
Fear of falling was based on a positive response to the question, “Are you worried or concerned that in the future you might fall?”, and was asked of respondents aged 65 or older. A negative response was assumed for younger respondents.
Daily stress was classified as high (most days quite a bit or extremely stressful) or low (most days not at all, not very or a bit stressful).
In addition to age and sex, the sociodemographic variables in the analysis were household education—highest level obtained by any household member (less than postsecondary, postsecondary graduation or more); marital status; living arrangements (alone or with others; “others” could be spouse, children, friend, or other); and regular driver defined as having a valid driver’s licence and driving at least once in the past month. Labour force participation was based on the previous week: worked at job/absent from job/not working but looking for job. Questions were limited to respondents younger than 75; people aged 75 or older were classified as non-labour force participants.
Social isolation was derived from two variables that measured loneliness and community belonging. From the Three-Item Loneliness Scale, which was based on the Revised UCLA Loneliness Scale,Note 31 respondents were asked: “How often do you feel: that you lack companionship? left out? isolated from others?” Response category values (1 = hardly ever; 2 = some of the time; 3 = often) were summed. Respondents who scored 3 were categorized as not lonely versus scores 4 to 9.
Sense of community belonging was determined with one question: “How would you describe your sense of belonging to your local community? Would you say it is very strong? somewhat strong? somewhat weak? very weak?” Respondents were categorized as being socially isolated if they were lonely and had a somewhat weak or weak sense of community belonging (Text table 2).
The tetrachoric correlation between loneliness and sense of community belonging was 0.2, supporting the notion that one can simultaneously feel lonely but connected to the community, and vice versa. Additional mutually exclusive groups (lonely, solitary, and connected) were derived from the two variables.
Men and women were analyzed separately. Weighted cross-tabulations and logistic regression models were used to examine associations between independent variables and social isolation. A preliminary logistic model determined if an association existed between hearing difficulty and social isolation that was conditional on respondents’ age. The interaction term was not significant (data not shown); subsequent models pooled respondents and controlled for age. Multiple logistic regression models controlled for age (continuous) and ageNote 2 because of the non-linear relationship between social isolation and age (Figure 1), in addition to potential confounders known to be associated with social isolation. A minimum change of 0.05 on the HUI3 single-attribute utility scores was considered meaningful.Note 32Note 33 Marital status and stress were excluded from the logistic regression because of their respective correlations with living arrangements and emotional impairment. Data were weighted on age group, sex and province and adjusted for non-response. To account for survey design effects of the CCHS–HA, coefficients of variation and p-values were estimated, and significance tests were performed using the bootstrap technique.Note 34Note 35
Characteristics of study population
The study sample of 30,176 respondents was weighted to represent 13.3 million people aged 45 or older, with a mean age of 60.4 (Appendix Table A). Almost half (48%) were men, and most (69%) lived in households where at least one person was a postsecondary graduate.
In 2008/2009, an estimated 864,000 people (7%) aged 45 or older reported some hearing difficulty; for half of these people, the difficulty was mild rather than moderate or severe (Table 1). Around 5% had corrected hearing, and 2% could not hear what was said in some circumstances, even with a hearing aid. Hearing impairment was more common at older ages—21% among those aged 75 or older. Men were generally more likely than women to have a hearing impairment―8% versus 5%—although significant differences between the sexes were evident only in the older age groups.
Perceived social isolation
An estimated 1.9 million people—12% of men and 16% of women—experienced social isolation, in that they reported feelings of loneliness and a weak or somewhat weak sense of community belonging (Table 2).
Social isolation was more common at ages 45 to 59 than among most older age groups (Table 2), although the continuous data suggest that social isolation may rise again at older ages (Figure 1).
Overall, labour force participation was not associated with social isolation, although 45- to 59-year-olds who were not in the labour force were significantly more likely than individuals of the same age who were working or actively looking for work to feel isolated.
High levels of daily stress were associated with social isolation, and 28% (95% CI, 26 to 29) of 45- to 59-year-olds reported that most days were quite a bit or extremely stressful, significantly more than estimates for people aged 60 to 74 (14%; 95% CI, 13 to 15) or 75 or older (10%; 95% CI, 9 to 11) (data not shown).
Women were consistently more likely than men to be socially isolated (except for women who were widowed, separated or divorced, or living alone). Social isolation was associated with lower education (men), living alone, not having a spouse, incontinence (women), fear of falling, high daily stress, and some functional impairments―vision (women), mobility, pain, emotion, and cognition. The latter two are noteworthy; men and women with these difficulties were two to four times more likely to be socially isolated than were people without these impairments. Regular drivers were less likely than non-drivers to be socially isolated.
Hearing difficulties and perceived social isolation
Among men, hearing difficulties were not associated with social isolation (Table 2), even when sociodemographic characteristics and other conditions were taken into account (Table 3). By contrast, 23% of women with hearing difficulties reported feeling socially isolated, compared with 16% of women without hearing difficulties. This association remained when controlling for other factors―as hearing difficulties increased, so did the odds of being socially isolated (1.04).
Although point estimates for women suggest a gradient in the prevalence of social isolation from those with no hearing impairment to corrected and then uncorrected hearing difficulties, the difference between the latter groups was not significant (Table 2, Figure 2). Men with corrected hearing difficulties were more likely than those with no impairment to be socially isolated.
According to the present study, hearing difficulties were associated with social isolation for women but not men. These findings persisted when sociodemographic factors, other functional limitations, incontinence and fear of falling were taken into account. Associations between hearing difficulties and social isolation have been observed in other research, although the results were not stratified by sex.Note 5Note 25Note 36Note 37 For example, after reporting that the prevalence of social isolation did not differ between men and women aged 60 or older, Hawthorne et al.Note 5 analyzed both sexes together. By contrast, in this study, when both sexes were considered together, a significant relationship between hearing difficulties and social isolation was evident. The stratified analysis revealed that this was driven by the association among women. In every age group, women were significantly more likely than men to be socially isolated.
Hearing difficulties can lead to social isolation if people withdraw to avoid the challenge of following conversations or embarrassment over their hearing loss or use of a hearing aid.Note 37Note 38 Hearing loss has also been associated with poor mobility and falls, lower health status and cognitive decline, all of which can contribute to social isolation.Note 39Note 40Note 41Note 42 Although such associations were evident in the current analysis, hearing difficulties were independently associated with social isolation for women.
Despite substantial advances in assistive technology,Note 43 most Canadian adults (88%) with hearing loss do not use hearing aids.Note 17 The reasons include the cost of the devices and the belief that they are not needed.Note 38 This study compared social isolation among those whose hearing aids enabled them to understand what was said in conversations (corrected hearing) and those who could not hear in some situations despite hearing aid use (uncorrected). Although estimates of social isolation among those with corrected and uncorrected hearing did not differ significantly, the point estimates for women revealed the expected gradient. The opposite was true for men; those with corrected hearing appeared more likely than those with uncorrected hearing difficulties to be socially isolated. Schneider et al.Note 44 reported that hearing loss was associated with greater dependency on family and community members. It is possible that men with uncorrected hearing difficulties are more dependent, which results in social interactions that help protect against isolation. Dawes et al.Note 36 also found a positive association between hearing aid use and social isolation.
This study established the level of hearing difficulties using the HUI3, which assesses self-reported ability to understand conversations in different circumstances. In other studies that relied on self-reported hearing status, the measures and situations differed―for example, if respondents can hear well when the speech is loud or whispered, over the telephone, or in a noisy room.Note 5Note 25 Alternatively, some studies measured biological hearing loss using audiometric or other testing.Note 16Note 36 Monzani et al.Note 45 distinguished between the two when they described self-reported hearing difficulty as the subjective experience of disability that arises from actual hearing loss. Self-reported hearing difficulty underestimates actual hearing loss.Note 16Note 17 The people who self-report may be those with the most severe loss or participate in activities that depend heavily on hearing acuity (for example, musicians, birdwatchers).Note 25 Despite the differences, hearing loss established using audiometric testing and self-reported hearing difficulties have both been associated with poor social outcomesNote 16 and mortality.Note 11Note 39
Social isolation has been defined and measured in a number of ways.Note 4Note 6Note 13Note 22Note 23Note 46 This study used perceived or subjective social isolation, which reflects how respondents feel about their circumstances and the quality of their relationships rather than objectively assessing attributes such as the number of contacts, frequency of participation or living arrangements.Note 5Note 6Note 8Note 12Note 22Note 47Note 48 The subjective measure captures intimate, relational and collective feelings of loneliness―the three underlying constructs in the Revised UCLA Loneliness ScaleNote 31―and incorporates sense of community belonging. Subjective social isolation does not depend on an arbitrary decision about what size of network or frequency of participation is sufficient. Satisfaction with networks and their size are separate concepts that are not necessarily correlated.Note 8 The subjective measure best reflects the key question― whether people who have difficulty communicating because of hearing problems feel isolated from those around them.
Numerous age-related transitions can disrupt social networks and engagement: retirement and the concomitant change in role and social contacts, becoming a caregiver, changing health status, lack of transportation including the loss of the ability or desire to drive, death of significant others, and a move to alternate living arrangements.Note 13Note 22Note 49 The baby boom generation may be at a greater risk of social isolation than earlier generations because they are more likely to live alone, to have never married, and to have fewer children.Note 50
Results from this study indicate that social isolation was more common among 45- to 59-year-olds than among people aged 60 or older, which is consistent with the work of Cacioppo et al.Note 51 Older adults who retire may have more time for volunteering and other activities that foster social connections, whereas many in the younger cohort have responsibilities that create stress from the competing demands of work, childcare and eldercare and limit time and energy to connect with others.Note 52Note 53 This study found that high daily stress was associated with social isolation and that 45- to 59-year-olds were more likely than older people to report that their daily lives were quite or extremely stressful. Working–age people who were not in the labour force were more likely to be isolated. HawthorneNote 54 found that the unemployed, those with work injuries, students and homemakers were more likely to be isolated than the fully employed. It could be that employment-related activities protect people from isolation, or that middle-aged adults who are not in the labour force have health issues or family responsibilities that contribute to social isolation.
Interventions often aim to increase social interactions in group settings.Note 55Note 56 However, people with hearing limitations may avoid socially challenging situations and feel deprived by their reduced ability to participate.Note 45 Interventions designed for one-on-one interaction and those promoting the use of technology for non-verbal communication (for example, Internet) may be more effective for combating social isolation among those with hearing impairments.Note 55
This study demonstrates the importance evaluating men and women separately. Future research could focus on sex differences in the hearing loss/social isolation relationship. Research could further examine the role of assistive technology, including teletypewriter (TTY) services, improved telephone features, and personal computing devices for non-verbal communication such as e-mail and text messaging.Note 5Note 16Note 25 Co-pathologies could be explored―multiple health and disability problems increase with age and could contribute to social isolation and interfere with the adoption of technologies that aid social connections.Note 57 Finally, other mitigating factors could be examined such as the impact of family and friends, and level of social support.
The cross-sectional data are a limitation of this study. The temporal order of hearing loss and social isolation could not be established. People were classified according to their current hearing status, regardless of their hearing loss history. However, hearing impairment may be present at birth or occur over the life course; onset may be sudden or gradual. The timing and speed of hearing loss may affect individuals’ ability to adapt and their feelings of social isolation.
Members of the Deaf community who communicate with Sign language cannot be identified in the CCHS–HA―there is no reason to expect that they would be at any greater risk of social isolation than the hearing-abled.
Information about cochlear implants or the quality and use of hearing aids was not available. Transitions associated with aging that may contribute to social isolation could not be captured―changing roles, loss of significant others, changing living arrangements that disrupt social networks, and developing health issues that limit connections.
The exclusion of proxy respondents, who comprised 2.2% of the CCHS–HA sample, could weaken estimates of the association between functional impairment and social isolation. People with proxy respondents were more likely to be men, to have hearing difficulties, and to be older and more functionally impaired than were non-proxy respondents (Appendix Table A).
The results indicate that social isolation diminishes somewhat with age. This could reflect a healthy survivor effect whereby those who are socially connected and more likely to remain healthy and/or have support, are represented in this study, whereas those who are more isolated may experience earlier mortality or a move to an institution. Residents of long-term care facilities, who may be more likely to have hearing loss and experience social isolation, were excluded from the CCHS–HA.
Hearing impairment was found to be associated with social isolation among Canadian women aged 45 or older, but not among men, a difference that could be further examined in future research. Social isolation could become more prevalent as the number of seniors in Canada grows and the percentage of the population experiencing hearing impairment increases. The low rate of hearing aid use among people with hearing difficulties suggests that future research could examine whether assistive technology has mitigated the isolation experienced by people with hearing limitations.
- Date modified: