Sensory impairments and loneliness among older Europeans: Insights from a cross-national longitudinal study

doi:10.21203/rs.3.rs-5036552/v1

Download PDF

Research Article

Sensory impairments and loneliness among older Europeans: Insights from a cross-national longitudinal study

https://doi.org/10.21203/rs.3.rs-5036552/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Sensory impairments are linked to loneliness, but the reasons are not well understood. This study aims to investigate the associations between sensory impairments and loneliness over time, considering sex, age, and European regions, and to examine the mediating roles of activity participation, activities of daily living (ADL), depressive symptoms, quality of life (QoL), cognitive function, and social network. We conducted a cohort study of 83,329 Europeans aged 50 + years, participating in at least two waves of the Survey of Health, Ageing and Retirement in Europe (SHARE) from 2011–2022. Loneliness was measured on a scale from 3–9. Mixed-effects regression models with sex-interactions, adjusting for socio-demographic factors and medical history, were used. Compared to those with good vision and hearing, individuals with vision impairment (VI) (coef. = 0.12, 95% CI: 0.10; 0.14), hearing impairment (HI) (coef. = 0.11, 95% CI: 0.08; 0.14), and dual sensory impairment (DSI) (coef. = 0.28, 95% CI: 0.25; 0.31) had a higher risk of loneliness. Dose-response relationships were observed in all associations. Most associations were consistent across sex and age groups, but the association with HI was strongest for females up to age 79 years but only observed in males thereafter. Associations varied across Europe, with the strongest associations in Northern Europe for females with VI and males with HI. For DSI, the associations were mainly mediated by activity participation (17.4%), depressive symptoms (17.3%), and QoL (16.6%). To mitigate the effect of sensory impairments on loneliness, particularly limitations in activity participation should be targeted.

Vision

hearing

dual sensory impairment

loneliness

mediation

Loneliness is often conceptualized as the subjective experience of a lack of satisfying human relationships (Andersson 1998). Loneliness affects a considerable share of older individuals globally and is increasingly addressed as a priority public health concern (World Health Organization 2021). Studies suggest that 20–34% of older individuals across Europe, China, Latin America, and USA experience feelings of loneliness (World Health Organization 2021), and loneliness is associated with severe health-related consequences such as cardiovascular disease, anxiety, depression (The Lancet 2023) and dementia (Lara, et al. 2019).

Sensory impairments are frequent among older adults and visual- (VI) and hearing impairment (HI) contribute significantly to the global disease burden (Ehrlich, et al. 2021; Sharma, et al. 2021). In Europe, one-third of individuals aged 50 years and above report either VI or HI (Hansen, et al. 2023). The consequences of sensory impairments are numerous and imply, among others, depressive symptoms (Ahrenfeldt, et al. 2024), cognitive decline (Möller, et al. 2024), decreased communication performance (Heine and Browning 2002), increased risk of falls (Ogliari, et al. 2021) and even an increased mortality risk (Ehrlich, et al. 2021; Feng, et al. 2022).

The association between sensory impairments and loneliness remains an area of uncertainty within current research. While several studies have examined HI and its association with loneliness, findings are inconsistent. A systematic review of 14 primarily cross-sectional studies indicated a link between HI and loneliness among individuals above 60 years, with a stronger association observed in females than in males (Shukla, et al. 2020). Similarly, a scoping review of 57 articles covering the entire lifespan suggested a comparable trend (Bott and Saunders 2021). However, contradicting these reviews, recent longitudinal research found no significant association between HI and loneliness (Hajek, et al. 2024).

Fewer studies have explored the effects of VI and dual sensory impairment (DSI)—the combination of VI and HI—on loneliness. A Norwegian cross-sectional study revealed that loneliness was more prevalent among 736 visually impaired adults compared to the general population, with the strongest association found among those aged 36 to 50 years (Brunes, et al. 2019). Complementing this, a longitudinal study from the German Ageing Survey, involving approximately 3,500 individuals aged 40 and older, showed that the onset of severe VI was linked to increased loneliness over time (Hajek, et al. 2024). Additionally, a large cross-sectional study of middle-aged and older Canadians demonstrated consistent associations between VI, HI, DSI, and loneliness across different ages and sexes (Mick, et al. 2018).

Further complexity is added by findings from a cross-sectional study in China involving 3,069 older adults, which reported that DSI had a more pronounced effect on loneliness than VI or HI alone, but this effect was only significant in rural areas (Wang, et al. 2022). Meanwhile, a nationally representative German study found that individuals with DSI were lonelier than those without sensory impairments (Hajek and König 2020a). Contradicting these findings, a Malaysian study with 229 older participants found no significant association between DSI and loneliness (Harithasan, et al. 2020).

Overall, the evidence remains inconclusive and is derived mainly from cross-sectional studies, which limits the ability to draw firm conclusions about causality. Additionally, little attention has been paid to understanding the underlying mechanisms that might explain the associations between sensory impairments and loneliness. This gap in knowledge underscores the need for further research to clarify these relationships and to explore potential explanatory factors.

Here, we aim to investigate the associations between sensory impairments and loneliness across sex, age, and European regions. Moreover, we aim to determine to which extent these associations are mediated by activity participation, limitations in activities of daily living (ADL), depressive symptoms, cognitive decline, quality of life (QoL) and social network. We hypothesize that sensory impairments will be associated with loneliness over time, particularly among individuals with DSI and among those who report the most severe vision and hearing difficulties. Based on previous evidence (Ahrenfeldt, et al. 2024; Möller, et al. 2024; Shukla, et al. 2020), we hypothesize that the association will be stronger in females than in males. Additionally, we hypothesize that the associations will be mediated by the aforementioned variables; however, perhaps most pronounced for activity participation and depressive symptoms.

Data source and study population

This study utilizes data from the Survey of Health, Ageing and Retirement in Europe (SHARE), which covers 28 European countries and focuses on the health, well-being, and life circumstances of individuals aged 50 years and above (Borsch-Supan, et al. 2013). Launched in 2004–2005, SHARE conducts biennial household-level interviews. SHARE uses computer-assisted personal interviewing for data consistency, and employs standardized questionnaires translated into national languages to ensure uniformity across countries (Borsch-Supan, et al. 2013). In SHARE the same participants are interviewed in each wave, if possible, but to maintain representation of younger cohorts and to counteract attrition, the SHARE survey is supplemented by refreshing samples. Besides self-reported questions, the SHARE data collection includes cognitive and physical tests (Borsch-Supan, et al. 2013).

In this study, we included Europeans aged 50 years and older who participated in at least two waves of SHARE between 2011 (wave 4) and 2022 (wave 9). Waves 1–3 were not included due to lack of information on loneliness. Moreover, we excluded individuals with missing age (n = 8), and those lacking data on vision and hearing impairments at baseline (i.e., the first time a participant took part in SHARE) (n = 596). To analyze the development of loneliness over time, information on sensory impairments, covariates and mediators was gathered from the baseline wave, and information on loneliness was obtained from subsequent waves (i.e., waves 5–9).

Loneliness

Information about loneliness was based on three items from the SHARE interview: 1) “How much of the time do you feel a lack of companionship?” 2) “How much of the time do you feel left out?” and 3) “How much of the time do you feel isolated from others?” Responses to the questions were registered on a three-point Likert scale: “Often”, “Some of the time”, and “Hardly ever or never” and subsequently aggregated into a continuous loneliness scale consistent with previous SHARE research (Hajek and König 2020b; Prell, et al. 2023). The scale was computed based on a short version of the Revised University of California at Los Angeles Loneliness Scale (R-UCLA) (Hughes, et al. 2004). The scale ranged from 3 to 9, where 3 corresponded to “Not lonely” and 9 corresponded to “Very lonely”. The R-UCLA has been harmonized for use in SHARE research (Malter and Börsch-Supan 2013).

Sensory impairments

The assessment of vision and hearing impairments was conducted using specific questions posed to the participants during the interviews. VI was assessed by the following two questions: “How good is your eyesight for seeing things up close, like reading ordinary newspaper print (using glasses or contact lenses as usual)?” to detect close vision and “How good is your eyesight for seeing things at a distance, like recognizing a friend across the street (using glasses or contact lenses as usual)?” to detect distant vision. Hearing was assessed by asking participants to rate their ability to hear using a hearing aid if needed. The response options for all three questions were "Excellent", “Very good”, “Good”, “Fair” and "Poor." Those reporting "Fair" or "Poor" in vision or hearing were categorized as impaired. Based on prior methodologies (Ahrenfeldt, et al. 2024; Möller, et al. 2024; Ogliari, et al. 2021), participants were grouped into "Good vision and hearing," "VI (i.e., VI but not HI),” "HI (i.e., HI but not VI)," and "DSI (i.e., both VI and HI)". Moreover, vision and hearing assessments were combined into a 5-point Likert scale to facilitate dose-response analysis.

Socio-demographic variables

Socio-demographic variables were collected during the first participant enrollment and included sex (males/females), age (categorized in 5-year intervals from 50 to 85, with 85 + as an open-ended category), and regional grouping of European countries into Northern (Denmark, Sweden), Western (Austria, Germany, the Netherlands, France, Switzerland, Belgium, Luxembourg), Southern (Spain, Italy, Greece, Portugal), and Eastern Europe (Czech Republic, Poland, Hungary, Slovenia, Estonia, Croatia) in line with prior studies (Ahrenfeldt, et al. 2024; Ahrenfeldt, et al. 2021; Hansen, et al. 2023; Möller, et al. 2024; Tolstrup Wester, et al. 2022). Additional variables included cohabitation status (partner presence as yes or no), educational level (low: ISCED 0–2, medium: ISCED 3–4, high: ISCED 5–6), household income (deciles, using SHARE's imputed data) (SHARE 2022), and medical history based on self-reported doctor diagnoses including heart attack, stroke, hypertension, diabetes, Parkinson’s Disease, lung disease, cancer, and arthritis.

Possible intermediate variables

Activity participation was quantified based on the types of activities respondents engaged in during the past year. The total activity score ranged from 0 to 9 (Litwin and Stoeckel 2016). These included voluntary work, educational courses, sports or social clubs, political or community events, reading, word or number games, and card games. Additionally, two questions assessed physical activities categorized by energy level: Moderate and vigorous. To be counted, activities had to be participated in at least once per month.

The ADL scale included six tasks: Dressing, walking, bathing, eating, getting in or out of bed, and using the toilet, with a total score ranging from 0–6 (Katz, et al. 1963).

Depressive symptoms were assessed by 12 questions measuring different aspects of depression such as mood, sleep, and concentration. Participants could answer yes or no to each question resulting in a score from 0–12, with higher scores indicating more depressive symptoms (Mehrbrodt, et al. 2019; Prince, et al. 1999).

QoL was assessed using the CASP-12 scale. This scale comprises 12 items rated on a four-point Likert scale (often, sometimes, rarely, and never) and covers four domains of need (Hyde, et al. 2003). The overall CASP-12 score range from 12 to 48 points, where higher scores denote better QoL (Mehrbrodt, et al. 2019).

Cognitive function was assessed using three cognitive tests performed during the personal interview (fluency, immediate and delayed recall). Similar to prior SHARE studies (Ahrenfeldt, et al. 2018; Ahrenfeldt, et al. 2019; Ahrenfeldt, et al. 2023; Ahrenfeldt and Möller 2021; Nielsen, et al. 2022), the individual scores of the three tests were standardized using the means and standard deviation (SD) obtained in the youngest age group (50–54-year-olds), and then summed and converted into a t-score with a mean of 50 and an SD of 10 in the youngest age group.

The social network variable was synthesized into a composite measure based on five attributes: Network size, proximity, frequency of contact, emotional support, and relationship diversity. Each attribute was scored on a 0–4 scale, with the total raw score ranging from 0 to 20. This score was then adjusted to a 0–4 scale, consistent with previous research (Litwin and Stoeckel 2016).

Statistical analyses

To investigate the associations between sensory impairments and development of loneliness, mixed effects linear regression models were used. The models produced coefficients (coef.) with 95% confidence intervals (CIs) using good vision and hearing as the reference group. The analysis encompassed three models: Model 1: Adjusted for sex, age, survey wave, and geographic region. Model 2: Further adjusted for household partner presence, educational level, and household income. Model 3: Included adjustments for medical history. We also examined sex differences in these associations by integrating an interaction term between sensory impairments and sex. The fully adjusted model was used to address the dose-response relationships between sensory impairments and loneliness using "Excellent" as the reference category. Also, it was used to conduct stratified analyses by age groups (50–64, 65–79, 80 + years), and European regions. Differences across age groups and European regions were evaluated using the Wald test.

We conducted two sensitivity analyses: First, we repeated the main analysis among only the sub-sample of the study population with no feelings of loneliness at baseline (i.e., those who scored 3 on the loneliness scale and thus stated “Hardly ever or never” to all three questions on loneliness, corresponding to 57.1% of the study population). Second, to evaluate how much hearing aids influenced the results, we divided the sub-population with good vision into four groups: 1) Good hearing and no use of hearing aid, good hearing but use of hearing aid, impaired hearing and no use of hearing aid, and impaired hearing but use of hearing aid. We investigated the associations with loneliness for the four groups, with good hearing and no use of hearing aid as the reference category.

Mediation analyses with robust clustered standard errors assessed how the potential mediators influenced the associations with loneliness. This approach involved separating the total effect of the association into a controlled direct effect and a mediated effect, and to estimate the proportion eliminated (PE), which accounts for the mediation effect of the potential mediator. To establish the 95% CIs for the PE, we employed bootstrapping with 100 replicates. This method was proposed by Yu et al. (Yu, et al. 2022), and originally suggested by VanderWeele (VanderWeele 2014).

All statistical analyses were conducted using Stata version 18.0, and multiple testing adjustments were made using the Holm-Bonferroni method in R version 4.3.1.

Table 1 presents the study population including 36,423 males and 46,906 females, corresponding to 266,994 interviews. At baseline, the mean age was 64.8 years (SD = 9.6 years). A total of 61.5% of the males and 63.1% of the females reported good vision and hearing, while 17.8% males and 21.8% females reported VI, 12.3% males and 7.7% females reported HI, and 8.3% males and 7.4% females reported DSI. Overall, 69.5% males and 74.3% females used glasses or contact lenses, and 5.8% males and 4.0% females used a hearing aid. The average loneliness score at follow-up was 3.8 (SD = 1.3) for males and 4.0 (SD = 1.5) for females (Table 1).

Table 1

Characteristics of the study population including Europeans who participated in at least two waves of SHARE between 2011 and 2022 stratified by sex and European regions.
	Total population		Northern Europe		Western Europe		Southern Europe		Eastern Europe
	Males	Females	Males	Females	Males	Females	Males	Females	Males	Females
Individuals	36,423 (43.7)	46,906 (56.3)	4,184 (46.4)	4,831 (53.6)	12,861 (44.7)	15,922 (55.3)	7,557 (44.5)	9,419 (55.5)	11,821 (41.4)	16,734 (58.6)
Age 50–64 years	19,002 (52.2)	25,211 (53.8)	2,020 (48.3)	2,407 (49.8)	7,092 (55.1)	9,086 (57.1)	3,531 (46.7)	4,829 (51.3)	6,359 (53.8)	8,889 (53.1)
Age 65–79 years	14,643 (40.2)	17,568 (37.5)	1,809 (43.2)	1,968 (40.7)	4,827 (37.5)	5,386 (33.8)	3,327 (44.0)	3,700 (39.3)	4,680 (39.6)	6,514 (38.9)
Age 80 years and above	2,778 (7.6)	4,127 (8.8)	355 (8.5)	456 (9.4)	942 (7.3)	1,450 (9.1)	699 (9.3)	890 (9.5)	782 (6.6)	1,331 (8.0)
Age, mean (SD)	64.9 (9.3)	64.6 (9.8)	65.4 (9.5)	65.2 (9.8)	64.3 (9.4)	64.1 (10.0)	66.2 (9.3)	65.2 (9.8)	64.6 (8.9)	64.8 (9.5)
Good vision and hearing	22,408 (61.5)	29,595 (63.1)	3,000 (71.7)	3,691 (76.4)	8,157 (63.4)	10,717 (67.3)	4,490 (59.4)	5,553 (59.0)	6,761 (57.2)	9,634 (57.6)
Vision impairment	6,499 (17.8)	10,246 (21.8)	404 (9.7)	578 (12.0)	1,997 (15.5)	2,924 (18.4)	1,562 (20.7)	2,402 (25.5)	2,536 (21.5)	4,342 (26.0)
Hearing impairment	4,447 (12.3)	3,593 (7.7)	602 (14.4)	401 (8.3)	1,906 (14.8)	1,495 (9.4)	745 (9.9)	596 (6.3)	1,224 (10.4)	1,101 (6.6)
Vision and hearing impairment	3,039 (8.3)	3,472 (7.4)	178 (4.3)	161 (3.3)	801 (6.2)	786 (4.9)	760 (10.1)	868 (9.2)	1,300 (11.0)	1,657 (9.9)
Use of glasses/contact lenses	25,306 (69.5)	34,823 (74.3)	3,587 (85.7)	4,375 (90.6)	9,267 (72.1)	12,319 (77.4)	5,367 (71.0)	7,119 (75.6)	7,085 (59.9)	11,010 (65.8)
Use of hearing aid	2,123 (5.8)	1,874 (4.0)	514 (12.3)	394 (8.2)	1,008 (7.8)	847 (5.3)	224 (3.0)	225 (2.4)	377 (3.2)	408 (2.4)
Partner in household	30,373 (83.4)	31,604 (67.4)	3,429 (82.0)	3,433 (71.1)	10,400 (80.9)	10,650 (66.9)	6,570 (86.9)	6,879 (73.0)	9,974 (84.4)	10,642 (63.6)
Low education	11,959 (33.1)	19,001 (40.8)	1,048 (25.3)	1,308 (27.3)	3,010 (23.7)	5,615 (35.7)	4,872 (65.0)	6,556 (70.3)	3,029 (25.7)	5,522 (33.1)
Medium education	15,417 (42.7)	18,017 (38.7)	1,636 (39.5)	1,540 (32.1)	5,564 (43.8)	6,383 (40.6)	1,559 (20.8)	1,811 (19.4)	6,658 (56.5)	8,283 (49.7)
High education	8,754 (24.2)	9,528 (20.5)	1,459 (35.2)	1,952 (40.7)	4,133 (32.5)	3,739 (23.8)	1,069 (14.3)	966 (10.4)	2,093 (17.8)	2,871 (17.2)
Income, median (IQR)	21,201 (9,793 − 44,703)	16,950 (7,714 − 36,799)	48,200 (31,218 − 68,657)	41,560 (26,226 − 62,846)	40,570 (24,500 − 69,309)	34,043 (19,858 − 59,189)	16,148 (9,647 − 27,534)	14,236 (8,186 − 24,374)	8,808 (5,068 − 14,153)	7,245 (4,015 − 12,212)
Heart attack	5,207 (14.3)	4,638 (9.9)	536 (12.8)	385 (8.0)	1,706 (13.3)	1,212 (7.6)	965 (12.8)	701 (7.5)	2,000 (16.9)	2,340 (14.0)
Hypertension	14,063 (38.7)	18,905 (40.4)	1,536 (36.7)	1,697 (35.2)	4,441 (34.6)	5,379 (33.8)	2,962 (39.3)	3,723 (39.7)	5,124 (43.4)	8,106 (48.5)
Stroke	1,522 (4.2)	1,520 (3.3)	198 (4.7)	168 (3.5)	469 (3.7)	480 (3.0)	229 (3.0)	183 (2.0)	626 (5.3)	689 (4.1)
Diabetes	4,703 (12.9)	5,032 (10.8)	429 (10.3)	371 (7.7)	1,570 (12.2)	1,361 (8.6)	1,126 (14.9)	1,193 (12.7)	1,578 (13.4)	2,107 (12.6)
Parkinson’s Disease	233 (0.6)	221 (0.5)	27 (0.7)	16 (0.3)	78 (0.6)	76 (0.5)	58 (0.8)	41 (0.4)	70 (0.6)	88 (0.5)
Lung disease	2,194 (6.0)	2,661 (5.7)	217 (5.2)	289 (6.0)	819 (6.4)	951 (6.0)	471 (6.3)	419 (4.5)	687 (5.8)	1,002 (6.0)
Cancer	1,840 (5.1)	2,762 (5.9)	243 (5.8)	358 (7.4)	792 (6.2)	1,072 (6.7)	251 (3.3)	387 (4.1)	554 (4.7)	945 (5.7)
Arthritis	5,988 (16.5)	13,379 (28.6)	662 (15.8)	1,326 (27.5)	2,254 (17.5)	4,521 (28.4)	1,134 (15.1)	3,078 (32.8)	1,938 (16.4)	4,454 (26.7)
Loneliness at follow-up, mean (SD)	3.8 (1.3)	4.0 (1.5)	3.5 (1.0)	3.7 (1.2)	3.6 (1.1)	3.9 (1.3)	3.9 (1.4)	4.4 (1.7)	4.0 (1.3)	4.1 (1.5)
Number of observations at follow-up	78,638 (42.8)	105,027 (57.2)	10,097 (45.6)	12,043 (54.4)	30,534 (44.0)	38,895 (56.0)	15,682 (43.8)	20,097 (56.2)	22,325 (39.6)	33,992 (60.4)
All estimates are numbers (%) unless stated otherwise.
SD = standard deviation. IQR = interquartile range
Missing observations are excluded from percentage calculations. The following variables had missing data: education 0.8% and comorbidities 0.2%. For all other variables there were 0.0% missing values.

[Table 1 about here]

Figure 1A-C shows the results of the associations between sensory impairments and development of loneliness for the three models. Although the associations became slightly weaker when further adjusting for socioeconomic status and medical history, significant associations were found for both VI (coef. = 0.12, 95% CI: 0.10; 0.14), HI (coef. = 0.11, 95% CI: 0.08; 0.14) and DSI (coef. = 0.28, 95% CI: 0.25; 0.31) in model 3. A significant interaction with sex was found for HI (p = 0.010) with stronger associations for females than for males (Supplementary Tables 1 and 2). When investigating the dose-response relationships, we found no difference in loneliness between males reporting “Very good” and “Excellent”, otherwise the risk of loneliness increased gradually with worsening levels of VI, HI and DSI, respectively showing significant dose-response relationships in all models (p < 0.001). The highest estimates were observed for those who reported “Poor”. This was most pronounced among participants who reported DSI (coef. = 0.56, 95% CI: 0.49; 0.62) (Fig. 1D-F, Supplementary Table 3).

Some fluctuations were found across age groups. The association between VI and loneliness among females increased from ages 50–79 years. Similarly, the association between DSI and loneliness among males increased from ages 50–80 + years. On the contrary, the association between HI and loneliness among females decreased from ages 50–80 + years with no association in the oldest age group (Fig. 1G-I, Supplementary Tables 4 and 5). A significant interaction with sex was found for HI in all age groups, with the strongest association at ages 50–79 years among females, and the strongest association among males at ages 80 + years (Supplementary Tables 2 and 4).

Fluctuations across the four European regions were also observed. The associations were stronger in Northern than in Western, Southern and Eastern Europe for females with VI and for males with HI. Moreover, females with HI and DSI had stronger associations in Northern than in Southern Europe. Moreover, for females with HI and DSI the associations were more pronounced in Western Europe than in Southern and Eastern Europe. Also, a stronger association was found for males with HI in Western than in Southern Europe (Supplementary Tables 6 and 7). In all European regions significant associations between sensory impairments and loneliness were found except from all associations for HI in Southern Europe and for HI in Eastern Europe among males (Fig. 1J-L and Supplementary Table 6).

When keeping only participants with no reporting of loneliness at baseline, increased risk of developing loneliness was found in all models; however, compared with the main analysis, the estimates were slightly less pronounced for VI (coef. = 0.05, 95% CI: 0.03; 0.07), HI (coef. = 0.03, 95% CI: 0.00; 0.06) and DSI (coef. = 0.14, 95% CI: 0.11; 0.18) (Supplementary Table 8). When investigating the influence of the use of hearing aid, we found a gradual increase in loneliness with the use of hearing aid and impaired hearing. Compared to people with good hearing who did not use a hearing aid, people with good hearing who used a hearing aid (coef. = 0.06, 95% CI: 0.01; 0.11), people with impaired hearing who did not use a hearing aid (coef. = 0.11, 95% CI: 0.09; 0.14) and those with impaired hearing who used a hearing aid (coef. = 0.16, 95% CI: 0.08; 0.24) had higher risk of loneliness (Supplementary Table 9).

After adjusting for multiple testing most associations remained significant (Supplementary Tables 1, 3, 4, 6, 8 and 9).

[Figure 1 about here]

When we investigated the mediating effect of the six different variables, we found that activity participation mediated 17.4% (95% CI 16.3%-18.5%) of the association between DSI on loneliness; however, with no significant mediation effect was observed for HI and only a minor effect for found for VI (2.7%; 95% CI 2.4%-3.0%). Depressive symptoms significantly mediated the associations of both VI (7.6%, 95% CI 6.9%-8.4%) and HI (7.6%, 95% CI 6.7%-8.5%), but most notably for DSI (17.3%, 95% CI 16.1%-18.6%). In addition, also QoL mediated the associations of VI (8.2%, 95% CI 7.3%-9.0%), HI (7.0%, 95% CI 5.8%-8.2%) and DSI (16.6%, 95% CI 15.4%-17.7%) on loneliness. Cognitive function had a small mediating effect on loneliness for both VI (1.1%, 95% CI 0.8%-1.3%), HI (1.0%, 95% CI 0.7%-1.3%) and DSI (3.0%, 95% CI 2.7%-3.3%). A similar pattern was found regarding limitations in ADL, which mediated 1.6% (95% CI 1.3%-1.8%) of the association with VI, 1.0% (95% CI 0.7%-1.4%) for HI and 4.0% (95% CI 3.5%-4.4%) for DSI. Social networks did not mediate loneliness, except for a minimal effect in cases of VI (0.2%, 95% CI 0.1%-0.3%) and DSI (0.3%, 95% CI 0.1%-0.5%) (Table 2).

Table 2

The mediating role of activity participation, activities of daily living (ADL), depressive symptom, quality of life, cognitive function, and social network in the associations between sensory impairments and loneliness.
	Activity participation	Activities of daily living (ADL)	Depressive symptoms	Quality of life	Cognitive function	Social network
Vision impairment	2.7 (2.4-3.0)	1.6 (1.3–1.8)	7.6 (6.9–8.4)	8.2 (7.3-9.0)	1.1 (0.8–1.3)	0.2 (0.1–0.3)
Hearing impairment	0.0 (0-0.4)	1.0 (0.7–1.4)	7.6 (6.7–8.5)	7.0 (5.8–8.2)	1.0 (0.7–1.3)	0.1 (0-0.3)
Dual sensory impairment	17.4 (16.3–18.5)	4.0 (3.5–4.4)	17.3 (16.1–18.6)	16.6 (15.4–17.7)	3.0 (2.7–3.3)	0.3 (0.1–0.5)
The proportions are presented with 95% confidence intervals.
Adjusted for sex, age, wave, region, partner in household, education, household income, and history of diseases including heart attack, hypertension, stroke, diabetes, Parkinson’s disease, lung disease, cancer, and arthritis.

[Table 2 about here]

This large, longitudinal study demonstrated persistent dose-response relationships, showing that sensory impairments are associated with loneliness among middle-aged and older Europeans. This was particularly the case among people with the most impaired vision and hearing and among people with DSI. The associations were generally consistent across sex, age and European regions, though fluctuations were found. The association between HI and loneliness was stronger for females than for males from ages 50–79 years. The strongest associations were found in Northern Europe particularly for females with VI and males with HI. The mediation analysis demonstrated that particularly preventing limitations in activity participation, depressive symptoms and QoL could reduce a significant portion of the total effect of sensory impairments on loneliness, particularly among people with DSI.

Previous research from England has shown that older males are less affected by social isolation and the absence of close relationships compared to older females, who are more likely to report feelings of loneliness (Ratcliffe, et al. 2024), and research has indicated that the association between HI and loneliness is stronger in females than in males (Shukla, et al. 2020). In line, this study found a significant sex-interaction for HI in the fully adjusted model. However, this pattern varied across age: Females showed stronger associations up to 79 years, whereas males in the oldest age group exhibited a stronger association. Despite some sex-specific trends regarding VI and DSI, the overall associations with loneliness were similar across sexes.

Consistent associations were also found across age groups with few exceptions. The associations between VI and loneliness in females and between DSI and loneliness in males, strengthened with age, likely due to the progression of sensory impairments with advancing age. Conversely, our findings indicate that the association between HI and loneliness weakens with advancing age in females, a trend that mirrors recent research on DSI and depressive symptoms (Ahrenfeldt, et al. 2024). This suggests that as women age, the impact of HI on feelings of loneliness may diminish, potentially due to adaptive coping mechanisms over time or changes in social expectations. Furthermore, recent studies have highlighted that in middle age (50–64 years), the link between HI and negative health outcomes is stronger in females than in males (Ahrenfeldt, et al. 2024; Möller, et al. 2024). This pattern aligns with our findings, indicating that midlife may be a particularly vulnerable period for women experiencing HI. However, these insights highlight the complexity of the associations between sensory impairments and loneliness, and further research is needed to explore the underlying mechanisms driving these sex differences.

In this study, we observed the strongest associations between sensory impairments and loneliness in Northern and Western Europe, while these associations were weaker in Southern and Eastern Europe. In fact, no significant association was found between HI and loneliness in Southern Europe. One potential explanation for these differences lies in the cultural practices related to living arrangements. In Southern and Eastern Europe, it is common for older individuals to cohabit with their adult children (d'Hombres, et al. 2018). In contrast, solitary living is more prevalent among older adults in Northern Europe, where independent living is more culturally ingrained (Sundström, et al. 2009). This cultural context could contribute to the stronger association between sensory impairments and loneliness observed particularly in Northern Europe. Additionally, household size has been shown to influence loneliness, particularly in Southern and Eastern Europe, where larger households appear to offer a protective effect against loneliness (Nyqvist, et al. 2019). This protective factor might explain the weaker association between sensory impairments and loneliness in these regions. Differences in healthcare access, especially in the availability and use of sensory aids, may further contribute to the regional disparities. For instance, while the underuse of hearing aids is a common issue across Europe, it is most pronounced in Southern and Eastern Europe (Hansen, et al. 2023). As a result, individuals in Northern Europe who still experience hearing difficulties, might be in worse health than those in Southern and Eastern Europe. Moreover, the sensitivity analysis provided additional insights, revealing that individuals with HI who used a hearing aid still experienced significant loneliness, suggesting that the current use of hearing aids does little to alleviate this feeling. This finding raises the need for further research into more comprehensive interventions.

As societies continues to age, addressing loneliness becomes increasingly crucial from a public health perspective. For example, research has shown that interventions designed to reduce loneliness can have significant benefits, such as lowering mortality rates and reducing healthcare usage among older individuals (Pitkala, et al. 2009). Recent evidence have demonstrated that sensory impairments are associated with the development of depressive symptoms (Ahrenfeldt, et al. 2024) and cognitive decline over time (Möller, et al. 2024), and that sensory impairments, particularly DSI, are associated with reduced QoL (Tseng, et al. 2018). These conditions can isolate individuals, increasing their risk of experiencing loneliness. In our study, we observed that limitations in activity participation, depressive symptoms, and QoL mediated the most significant portions of the total effect of sensory impairments on loneliness. However, contrary to expectations, we did not find that social network mediated the associations between sensory impairments and loneliness to any substantial degree. However, this could be due to the way social network have been measured and further studies should be conducted to confirm these findings.

From a public health perspective, the present findings suggest that interventions aimed at enhancing activity participation, such as social environments that actively include and support individuals with sensory impairments could be particularly effective in reducing loneliness among sensory-impaired individuals, especially those with DSI. By fostering inclusive communities and providing opportunities for meaningful social interaction, it may be possible to reduce the burden of social withdrawal and, consequently, the experience of loneliness. Specific recommendations include designing social programs that are accessible to those with sensory impairments, perhaps by promoting the use of assistive technologies to facilitate communication. These strategies may play a role in improving QoL for older adults with sensory impairments, ultimately reducing the incidence of loneliness.

This study has several strengths. The key strength was the large sample size comprising 83,329 individuals who were interviewed at least twice, which gave the analysis sufficient statistical power to analyze the associations across sex, age groups and European regions in up to 11 years. Moreover, SHARE follows standardized procedures across all countries, including quality control measures (Börsch-Supan et al., 2013). This supports the validity of the cross-regional comparisons.

The study has also limitations. First, VI, HI and DSI were only measured at baseline. Since sensory impairments become more pronounced with advancing age (Hansen, et al. 2023), some participants categorized with no impairments at baseline might have developed sensory impairments during follow-up, which might have underestimated the associations. All data were obtained through self-reports. Self-reporting comes with some uncertainty (Pannucci and Wilkins 2010). Still, previous research has found self-reports of sensory impairments sufficiently valid (Ferrite, et al. 2011; Whillans and Nazroo 2014). However, prior research suggested that loneliness might be subject to some degree of negative social stigma (d'Hombres, et al. 2018). Thus, the associations regarding loneliness might be subject to responder bias which might have underestimated the results.

Our research underscores persistent associations between sensory impairments and loneliness in a dose-response manner. Most associations were consistent across sex and age groups, though the association with HI was stronger for females than for males up to age 79 years. After age 79, the association was observed only in males. Associations varied across European regions with the strongest associations in Northern Europe for females with VI and males with HI. Our mediation analyses suggested that mitigating limitations in activity participation, ADL, depressive symptoms, cognitive decline and QoL could potentially reduce a significant portion of the total effect of sensory impairments on development of loneliness. Particularly addressing limitations in activity participation could be of public health relevance.

Declaration of competing interest:

None

Funding:

There was no specific funding for this study.

Author Contribution

Author contributions: MSSH: Conceptualisation, Validation, Writing - original draft, Writing - review & editing. JL, SM, LS, NKK: Supervision, Validation LJA: Conceptualisation, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualisation, Writing - original draft, Writing - review & editing. All authors reviewed and approved the final manuscript before submission.

Acknowledgement

This paper uses data from SHARE Waves 4, 5, 6, 7, 8 and 9 (DOIs: 10.6103/SHARE.w4.900, 10.6103/SHARE.w5.900, 10.6103/SHARE.w6.900,10.6103/SHARE.w7.900, 10.6103/SHARE.w8.900), see Börsch-Supan et al. (2013) for methodological details. The SHARE data collection has been funded by the European Commission, DG RTD through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: GA N°211909, SHARE-LEAP: GA N°227822, SHARE M4: GA N°261982, DASISH: GA N°283646) and Horizon 2020 (SHARE-DEV3: GA N°676536, SHARE-COHESION: GA N°870628, SERISS: GA N°654221, SSHOC: GA N°823782, SHARE-COVID19: GA N°101015924) and by DG Employment, Social Affairs & Inclusion through VS 2015/0195, VS 2016/0135, VS 2018/0285, VS 2019/0332, and VS 2020/0313. Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C, RAG052527A) and from various national funding sources is gratefully acknowledged (see www.share-project.org).

Ahrenfeldt, L. J., et al. 2018 Comparison of cognitive and physical functioning of Europeans in 2004-05 and 2013. Int J Epidemiol.
Ahrenfeldt, L. J., et al. 2024 Sensory impairments and depressive symptoms in Europe: a cross-national cohort study. Aging Ment Health:1-9.
Ahrenfeldt, L. J., et al. 2021 Burden and prevalence of risk factors for severe COVID-19 in the ageing European population - a SHARE-based analysis. Z Gesundh Wiss:1-10.
Ahrenfeldt, L. J., et al. 2019 Gender differences in cognitive function and grip strength: a cross-national comparison of four European regions. Eur J Public Health 29(4):667-674.
Ahrenfeldt, L. J., et al. 2023 Cognitive function among religious and non-religious Europeans: a cross-national cohort study. Aging Ment Health:1-9.
Ahrenfeldt, Linda Juel, and Sören Möller 2021 The Reciprocal Relationship between Socioeconomic Status and Health and the Influence of Sex: A European SHARE-Analysis Based on Structural Equation Modeling. International Journal of Environmental Research and Public Health 18(9):5045.
Andersson, Lars 1998 Loneliness research and interventions: A review of the literature. Aging & Mental Health 2(4):264-274.
Borsch-Supan, A., et al. 2013 Data Resource Profile: the Survey of Health, Ageing and Retirement in Europe (SHARE). Int J Epidemiol 42(4):992-1001.
Bott, A., and G. Saunders 2021 A scoping review of studies investigating hearing loss, social isolation and/or loneliness in adults. Int J Audiol 60(sup2):30-46.
Brunes, A., B. Hansen M, and T. Heir 2019 Loneliness among adults with visual impairment: prevalence, associated factors, and relationship to life satisfaction. Health Qual Life Outcomes 17(1):24.
d'Hombres, Béatrice, et al. 2018 Loneliness - an unequally shared burden in Europe.
Ehrlich, J. R., et al. 2021 Association between vision impairment and mortality: a systematic review and meta-analysis. Lancet Glob Health 9(4):e418-e430.
Feng, X., et al. 2022 Association of hearing loss with total and cause-specific mortality in US adults. Environ Sci Pollut Res Int 29(4):5032-5042.
Ferrite, S., V. S. Santana, and S. W. Marshall 2011 Validity of self-reported hearing loss in adults: performance of three single questions. Rev Saude Publica 45(5):824-30.
Hajek, A., et al. 2024 Vision and hearing problems and psychosocial outcomes: longitudinal evidence from the German Ageing Survey. Soc Psychiatry Psychiatr Epidemiol 59(5):879-886.
Hajek, A., and H. H. König 2020a Dual sensory impairment and healthcare use: Findings from a nationally representative sample. Geriatr Gerontol Int 20(6):602-606.
— 2020b Which factors contribute to loneliness among older Europeans? Findings from the Survey of Health, Ageing and Retirement in Europe: Determinants of loneliness. Arch Gerontol Geriatr 89:104080.
Hansen, R. S., et al. 2023 Sex differences in vision and hearing impairments across age and European regions : Findings from SHARE. Wien Klin Wochenschr.
Harithasan, D., et al. 2020 The impact of sensory impairment on cognitive performance, quality of life, depression, and loneliness in older adults. Int J Geriatr Psychiatry 35(4):358-364.
Heine, C., and C. J. Browning 2002 Communication and psychosocial consequences of sensory loss in older adults: overview and rehabilitation directions. Disabil Rehabil 24(15):763-73.
Hughes, M. E., et al. 2004 A Short Scale for Measuring Loneliness in Large Surveys: Results From Two Population-Based Studies. Res Aging 26(6):655-672.
Hyde, M., et al. 2003 A measure of quality of life in early old age: the theory, development and properties of a needs satisfaction model (CASP-19). Aging Ment Health 7(3):186-94.
Katz, S., et al. 1963 STUDIES OF ILLNESS IN THE AGED. THE INDEX OF ADL: A STANDARDIZED MEASURE OF BIOLOGICAL AND PSYCHOSOCIAL FUNCTION. Jama 185:914-9.
Lara, E., et al. 2019 Does loneliness contribute to mild cognitive impairment and dementia? A systematic review and meta-analysis of longitudinal studies. Ageing Res Rev 52:7-16.
Litwin, H., and K. J. Stoeckel 2016 Social Network, Activity Participation, and Cognition: A Complex Relationship. Res Aging 38(1):76-97.
Malter, Frederic, and Axel Börsch-Supan 2013 SHARE Wave 4: Innovations & Methodology. https://share.cerge-ei.cz/documentation/Wave2010/share_wave4_innovations_and_methodology.pdf
Mehrbrodt, Tabea, Stefan Gruber, and Melanie Wagner 2019 Scales and Multi-Item Indicators in the Survey of Health, Ageing and Retirement in Europe. SHARE Working Paper Series 45-2019.
Mick, P., et al. 2018 Associations between sensory loss and social networks, participation, support, and loneliness: Analysis of the Canadian Longitudinal Study on Aging. Can Fam Physician 64(1):e33-e41.
Möller, S., et al. 2024 Sensory impairments and the risk of cognitive decline and dementia across sex, age, and regions: Longitudinal insights from Europe. Arch Gerontol Geriatr 127:105584.
Nielsen, C. R., et al. 2022 Development in life expectancy with good and poor cognitive function in the elderly European Population from 2004-05 to 2015. Eur J Epidemiol 37(5):495-502.
Nyqvist, F., M. Nygård, and T. Scharf 2019 Loneliness amongst older people in Europe: a comparative study of welfare regimes. Eur J Ageing 16(2):133-143.
Ogliari, G., et al. 2021 Subjective vision and hearing impairment and falls among community-dwelling adults: a prospective study in the Survey of Health, Ageing and Retirement in Europe (SHARE). Eur Geriatr Med 12(5):1031-1043.
Pannucci, C. J., and E. G. Wilkins 2010 Identifying and avoiding bias in research. Plast Reconstr Surg 126(2):619-625.
Pitkala, K. H., et al. 2009 Effects of psychosocial group rehabilitation on health, use of health care services, and mortality of older persons suffering from loneliness: a randomized, controlled trial. J Gerontol A Biol Sci Med Sci 64(7):792-800.
Prell, T., A. Schönenberg, and K. G. Heimrich 2023 The impact of loneliness on quality of life in people with Parkinson's disease: results from the Survey of Health, Ageing and Retirement in Europe. Front Med (Lausanne) 10:1183289.
Prince, M. J., et al. 1999 Development of the EURO-D scale--a European, Union initiative to compare symptoms of depression in 14 European centres. Br J Psychiatry 174:330-8.
Ratcliffe, J., P. Galdas, and M. Kanaan 2024 Older men and loneliness: a cross-sectional study of sex differences in the English Longitudinal Study of Ageing. BMC Public Health 24(1):354.
SHARE 2022 Survey of Health, Ageing and Retirement in Europe. Release guide 8.0.0.
Sharma, R. K., A. Chern, and J. S. Golub 2021 Age-Related Hearing Loss and the Development of Cognitive Impairment and Late-Life Depression: A Scoping Overview. Semin Hear 42(1):10-25.
Shukla, A., et al. 2020 Hearing Loss, Loneliness, and Social Isolation: A Systematic Review. Otolaryngol Head Neck Surg 162(5):622-633.
Sundström, G., et al. 2009 Loneliness among older Europeans. Eur J Ageing 6(4):267.
The Lancet 2023 Loneliness as a health issue. The Lancet 402(10396):79.
Tolstrup Wester, C., et al. 2022 Prayer frequency and COVID-19 vaccine hesitancy among older adults in Europe. Vaccine 40(44):6383-6390.
Tseng, Y. C., et al. 2018 Quality of life in older adults with sensory impairments: a systematic review. Qual Life Res 27(8):1957-1971.
VanderWeele, T. J. 2014 A unification of mediation and interaction: a 4-way decomposition. Epidemiology 25(5):749-61.
Wang, Q., et al. 2022 Dual Sensory Impairment as a Predictor of Loneliness and Isolation in Older Adults: National Cohort Study. JMIR Public Health Surveill 8(11):e39314.
Whillans, Jennifer, and James Nazroo 2014 Assessment of visual impairment: The relationship between self-reported vision and ‘gold-standard’ measured visual acuity. British Journal of Visual Impairment 32(3):236-248.
World Health Organization 2021 Social isolation and loneliness among older people: advocacy brief. Geneva: World Health Organization.
Yu, Y., et al. 2022 Gestational Diabetes Mellitus and the Risks of Overall and Type-Specific Cardiovascular Diseases: A Population- and Sibling-Matched Cohort Study. Diabetes Care 45(1):151-159.

No competing interests reported.

SupplementaryTables.docx

Download PDF

Editor assigned by journal
20 Sep, 2024
Submission checks completed at journal
05 Sep, 2024
First submitted to journal
05 Sep, 2024

You are reading this latest preprint version

Sensory impairments and loneliness among older Europeans: Insights from a cross-national longitudinal study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Data source and study population

Loneliness

Sensory impairments

Socio-demographic variables

Possible intermediate variables

Statistical analyses

Results

Discussion

Conclusion

Declarations

Declaration of competing interest:

Funding:

Author Contribution

Acknowledgement

References

Additional Declarations

Supplementary Files

Status:

Version 1