Social isolation and loneliness are descriptive terms that are often used interchangeably, although they refer to two distinct concepts. Social isolation is an objective and quantifiable measure, while loneliness is a subjective emotional experience (De Jong-Gierveld and Kamphuls, 1985; Holt-Lunstad et al., 2015). Social isolation is characterized by a limited relational network and social contacts (Holt-Lunstad et al., 2015), while loneliness describes the unpleasant feeling resulting from the discrepancy between desired and existing social relationships (Pinquart and Sörensen, 2003; Boldy and Grenade, 2011). Loneliness can be further subdivided into “social loneliness” (the perceived lack of a good-quality social network) and “emotional loneliness” (the perceived lack of intimate/emotional support from a significant other) (Weiss, 1973; Gierveld et al., 2006). Thus, an individual’s perception of the size of their social network and the emotional connections they have will influence whether they feel lonely or not (Weiss, 1973; Gierveld et al., 2006; Cacioppo et al., 2015).

Even though the terms’ “social isolation” (negative) and “social support” (positive) are used interchangeably, they do not refer to diametrical concepts (Berkman et al., 2000; Holt-Lunstad et al., 2015). While social isolation is an objective measure of the number of social connections (Holt-Lunstad et al., 2015), social support is a subjective measure of the actual or perceived availability of psychological and material resources provided by the social network (Cohen, 2004). Social support further can be divided into three categories based on the resources: tangible (provision of materials), informational (guidance or support), and emotional (empathy, care) (House and Kahn, 1985).

Loneliness is associated with depressive symptoms (Cacioppo et al., 2006, 2010; Zebhauser et al., 2014), reduced physical activity (Hawkley et al., 2009), decreased satisfaction with life (Zebhauser et al., 2014; Bai and Knapp, 2016) and poor subjective wellbeing (Pinquart and Sörensen, 2003). Loneliness has also been associated with cognitive impairment (Donovan et al., 2017; Kim et al., 2020), increased risk of dementia (Zhou et al., 2018) and cardiovascular disease, diabetes and migraine (Christiansen et al., 2016). Social isolation has been associated with falls, re-hospitalization, cardiac heart disease, cancer, and nutritional risk (Nicholson, 2012). Current literature highlights that loneliness and social isolation resulting from poor social relationships increase the mortality rate among older adults (Luo et al., 2012; Holt-Lunstad et al., 2015). On a positive note, social support increases the resilience to stressful events (Ozbay et al., 2007) and suicidal ideation (Zhang et al., 2018), plays a protective role against depression (Gariépy et al., 2016) and reduces the Diabetes burden (Kaya and Caydam, 2019). Identifying factors that contribute to developing strong positive social relationships in older adults could help reduce this population’s high morbidity and mortality.

Age-related hearing loss (ARHL) is prevalent in later life and has an increasing trend across ages; 10.9–17.6% in 60–69 years, 41.9–51.2% in 80–89 years, and 52.9–64.9% in 90 years and above (World Health Organization [WHO], 2021b). ARHL is associated with an increased risk of cognitive impairment (Jayakody et al., 2017), dementia (Dalton et al., 2003; Lin et al., 2011), and Alzheimer’s disease (Lin et al., 2011), poor quality of life (Dalton et al., 2003), physical inactivity (Gispen et al., 2014), social isolation (Strawbridge et al., 2000), as well as depression, anxiety, and stress (Jayakody et al., 2018). ARHL is characterized by loss of peripheral hearing sensitivity and decreased ability to understand speech, mainly when there is background noise (Davis et al., 2016). Thus, hearing impairment poses a major challenge as it impairs the ability to hear, listen and understand the intended message. Hence, the consequences of hearing impairment affect not only the hearing-impaired individual but their communication partners/significant others as well (Kamil and Lin, 2015). Communication partners experience poorer quality of life and relationship satisfaction, a restricted social life and increased communication burden, and (Kamil and Lin, 2015).

Communication refers to the “bidirectional transfer of information, meaning, and intent between two or more individuals” (Kiessling et al., 2003). Communication forms part of a social act which involves expressing oneself and relating to others and is moderated by the emotions, attitudes and beliefs of the communication partners and the rules of the society (Lemke and Scherpiet, 2015). Verbal communication requires both participants to hear, listen and comprehend (Kiessling et al., 2003). According to the World Health Organization’s International Classification of Functioning, Disability, and Health (World Health Organization [WHO], 2001), communication disability due to hearing impairment is an outcome of interactions between sensory impairment and participation in life. Hence, communication impairment resulting from hearing loss restricts active involvement in social and cultural activities leading to withdrawal and feelings of loneliness (Strawbridge et al., 2000) and social isolation (Pronk et al., 2011; Zebhauser et al., 2014). Several studies have found that hearing loss is associated with a higher risk of social isolation and loneliness (Pronk et al., 2014; Mick and Pichora-Fuller, 2016; Contrera et al., 2017). It is possible that people with hearing loss withdraw from social situations due to difficulties in communicating and following conversations, which leads to social isolation. This association has been found more frequently in women compared to men (Mick et al., 2014), possibly due to several reasons: (i) Women feel more comfortable in reporting social isolation/loneliness compared to men (Shukla et al., 2020); (ii) women rely more on verbal communication for social networking (Maltz and Borker, 1982) and (iii) become socially and emotionally vulnerable if they are unable to effectively connect with their social network due to their hearing impairment (Shukla et al., 2020).

Hence, it is important to investigate the impact of untreated hearing loss on social and emotional loneliness, social isolation, perceived social support, and psychological discomfort (symptoms of depression, anxiety, stress).

This was a cross-sectional study of a community-derived sample of volunteers and adults in contact with the clinical services of the Ear Science Institute of Australia in Perth, Western Australia. Participants were either native or fluent English speakers older than 40 years, did not use any hearing aids or hearing implants, had bilateral symmetrical pure tone audiometric thresholds of hearing sensitivity and did not have any morbidities or disabilities that prevented them from completing the assessments.

The assessment materials consisted of measures of hearing, mental health, social support and interaction, and social and emotional loneliness. The Human Research Ethics Committee of the University of Western Australia approved the study protocol (RA/4/1/7368), and all participants provided written informed consent.

The statistical analyses were carried out in SPSS Statistics for Windows version 25.0 (Armonk, NY: IBM Corp.) and R version 3.5.1 (R Core Team, 2018). Univariate linear regression was used to assess the association between Social Support, Social Interaction, and BE4PTA. The statically significant outcomes observed in the univariate analysis were then used in multiple linear regression. A P-value of less than 0.05 was considered to indicate statistical significance.

All authors accessed the study data and reviewed and approved the final manuscript. We used the STROBE statement to check the completeness of our study report (Von Elm et al., 2014).

After univariate analysis, hearing was not statistically significant associated with SS. (P = 0.07). Poor satisfaction with social support scores was observed in individuals who live alone (P = 0.003), and experience more stress (P < 0.001), anxiety (P = 0.001), and depression (P < 0.001) or vice versa (see Tables 2, 3).

Two variables have a significant effect on SI in the univariate analyses: depression (P = 0.01) and age (P < 0.001). After multivariate analysis for confounders, aging (P = 0.01) and depression (P = 0.01) continued to be statistically significant. These results suggest that higher depression scores and older age contributes to increased social isolation. A statistically significant interaction between age and hearing was observed (P = 0.04). Poorer SI scores are seen for people who are not only older but with poorer hearing. Results are summarized in Table 4. No significant effect of years of formal education, premorbid IQ (NART-R), smoking, alcohol ≥ 14 drinks per week or exercise were observed on SI and SS.

The total number and percentage of participants for each EL and SL category are reported in Table 5. Logistic multinominal regression analysis did not show any effect of gender, premorbid IQ (NART-R), alcohol ≥ 14 drinks per week, smoking and exercises on EL. Living alone (P = 0.01) significantly contributed to the intense levels of emotional loneliness. Age (P = 0.02) and (P = 0.02), depression (P < 0.001) and (P < 0.001), anxiety (P < 0.001) and (P < 0.001), stress (P < 0.001) and (P < 0.001), BE 4PTA (P < 0.001) and (P < 0.001) significantly contributed to both moderate and intense levels of EL, respectively (see Tables 5, 6).

An adjacent category ordinal regression analysis was conducted to investigate this association between EL and BE 4PTA, living alone and age. The adjacent category logistic regression model was implemented in the VGAM framework (Yee and Wild, 1996; Yee, 2010, 2015) using the R statistical language (R Core Team, 2018). This adjacent ordinal model considers the impact of the predictors on the odds ratio on the adjacent ordinal response levels with the following mathematical model,

In our case, the p_i denote the probabilities of being at an emotional loneliness level i. This is modeling the log-odds as a linear function of the predictors in design matrix X and the linear predictor β. Therefore, the full model would involve the three logistic equations,

Only three equations are modeled since there are only three transitions to cover the four EL ordinal levels. The likelihood ratio test was used to investigate the proportional odds assumption, i.e., if there are different predictor effects at different ordinal levels, but there was no statistical evidence for this with a p-value of 0.86, the same effect of the predictors is present across all adjacent transitions; only an intercept adjustment is required. Furthermore, there was no statistical evidence to suggest an interaction between the predictor terms. The final model is,

where β_0,i is an intercept term for each transition level that gives the estimated log odds for the adjacent levels at the baseline predictor level. Here, the baseline predictors are for participants that live alone (Living alone 0). The living alone variables in the equation above are indicators that take the value 1 for participants that live alone. The exponentiated effect estimates are shown above in the equation. The corresponding p-values and 95% CI are BE4PTA [0.048, (95% CI: 1.00 to −1.01)], Age [0.002, (95% CI: 0.95 to −0.99)], Living alone [0.044, (95% CI: 1.01 to −2.36)], and Depression [2.5×10⁻⁹, (95% CI: 1.06–1.16)]. Interpreting this model suggests that a participant living alone increases their odds of transitioning to the next higher EL level by 54.6%. For each unit of BE4PTA, a participant’s odds of transitioning to the next higher EL level increased by 0.6%. For example, if a person whose hearing deteriorates from no hearing loss to mild to moderate hearing loss, where the BE4PTA increases from 25 to 50 dB, they will have 1.006²⁵ = 1.16 or a 16% increase in their odds of transitioning to the next higher EL level.

Multinominal analysis revealed that age, gender, premorbid IQ (NART-R), anxiety, BE4PTA, smoking, and alcohol ≥ 14 drinks per week had no effect on SL. Three of the variables exhibited a statistically significant association with the SL response. Living alone (P = 0.01), depression (P = 0.01), and stress (P = 0.01) had a significant impact on intense levels of SL. Results are summarized in Table 7.

This is a cross-sectional study; therefore, results must be interpreted cautiously. A study with more power and a longitudinal design could provide more insight into the causal relationship between untreated hearing loss, loneliness, social isolation, and social support.

A follow-up manuscript will investigate the effect of hearing loss treatment using hearing aids/hearing implants that could help alleviate loneliness, social isolation, and depression. Future research is required to investigate whether social support could buffer the association between hearing loss and emotional loneliness.