Thesis – Loneliness and Social Connection Among Older Australians

Abstract

Loneliness is increasingly recognised as a determinant of health rather than a private misfortune, and Australia’s ageing population makes its distribution among older people a public health priority. This thesis examined the prevalence and predictors of loneliness among community-dwelling Australians aged 65 and over. A cross-sectional survey of 500 older adults across metropolitan, regional and remote areas measured loneliness with the six-item De Jong Gierveld Loneliness Scale and collected data on living arrangement, health, social participation and digital inclusion. Data were analysed using descriptive statistics and multiple linear regression. Moderate to severe loneliness affected 34.4 per cent of the sample and rose to 46.2 per cent among those aged 85 and over. Living alone, poorer self-rated health, regional or remote residence and financial strain independently predicted higher loneliness, while regular internet use and community group participation were protective. The findings support targeted, connection-focused responses within Australia’s aged care and primary health systems.

Introduction

Loneliness, the distressing feeling that arises when a person’s social relationships fall short of what they want, has moved from the margins of social policy to the centre of public health thinking. A landmark meta-analysis found that loneliness and social isolation raise the risk of premature death by a magnitude comparable to established behavioural risks such as smoking and physical inactivity (Holt-Lunstad et al., 2015). Mechanistic and longitudinal reviews link chronic loneliness to elevated blood pressure, disrupted sleep, cognitive decline and depression, establishing it as a modifiable determinant of health rather than an inevitable feature of growing older (Hawkley & Cacioppo, 2010; Courtin & Knapp, 2017).

These concerns are sharpened by demography. Australian Bureau of Statistics projections indicate that the share of the population aged 65 and over will continue to climb over coming decades, with the fastest growth among those aged 85 and over (ABS, 2023). The Australian Institute of Health and Welfare treats social isolation and loneliness as reportable population health issues and notes that older people are exposed to accumulating risk factors, including widowhood, retirement, reduced mobility and sensory decline (AIHW, 2024). The national organisation Ending Loneliness Together (2023) estimates that around one in three Australians experiences loneliness, and the Royal Commission into Aged Care Quality and Safety (2021) identified social isolation as a persistent harm within the aged care system that current arrangements do not adequately prevent.

Despite this policy attention, robust Australian estimates of loneliness that combine a validated measure with an analysis of its predictors in later life remain limited. This thesis addresses that gap. Three research questions guide the study:

1. RQ1: What is the prevalence of moderate to severe loneliness among community-dwelling Australians aged 65 and over?
2. RQ2: Which sociodemographic, health and social factors independently predict loneliness in this population?
3. RQ3: Are digital inclusion and community participation associated with lower loneliness once other factors are held constant?

Literature Review

The full review examined three bodies of evidence, summarised here: how loneliness is measured, what predicts it in later life, and what interventions reduce it.

Measuring loneliness

Loneliness is a subjective state and must be distinguished from objective social isolation, with which it correlates only moderately. Two validated instruments dominate the field. The University of California, Los Angeles scale treats loneliness as a single dimension, whereas the De Jong Gierveld scale separates emotional loneliness, the absence of an intimate attachment, from social loneliness, the absence of a wider network. The six-item short form has demonstrated sound reliability and validity across national survey samples and is well suited to older cohorts because of its brevity and low respondent burden (De Jong Gierveld & Van Tilburg, 2006). Direct comparison is complicated by measurement heterogeneity: a review of prevalence studies showed that estimates swing widely depending on whether a single-item or a multi-item scale is used and where the cut-point is set (Victor & Yang, 2012).

Predictors in later life

The predictors of loneliness among older adults are reasonably consistent across the international and Australian literature. Living alone, widowhood and the loss of a confidant are strong emotional-loneliness risks, while shrinking networks and withdrawal from paid work and community roles drive social loneliness (Courtin & Knapp, 2017). Poor self-rated health and functional limitation operate in both directions, restricting participation and being worsened by it. Australian analyses have added a geographic dimension, with earlier national work reporting that loneliness is unevenly distributed and sensitive to socioeconomic disadvantage (Baker, 2012), and more recent reporting highlighting the compounding disadvantage faced by older people in regional and remote areas where services and transport are thinner (AIHW, 2024).

Interventions

Intervention evidence is expanding but uneven. A frequently cited meta-analysis concluded that the most effective programs address maladaptive social cognition rather than simply increasing social contact, although group-based activities with an active ingredient of shared purpose also help (Masi et al., 2011). Community programs, volunteering and befriending schemes show promise, and digital inclusion has emerged as a plausible protective pathway by lowering the barrier to maintaining contact. Yet digital access is unequally distributed: older Australians record the lowest scores on the Australian Digital Inclusion Index, so an intervention that assumes connectivity may widen rather than narrow inequities (Thomas et al., 2023). This tension between community-based and technology-based responses motivates the present analysis.

Methodology

Design and sample

The study used a cross-sectional survey design. Community-dwelling Australians aged 65 and over were recruited between February and May 2024 through Council on the Ageing state networks, seniors’ community centres, University of the Third Age groups and public libraries across New South Wales, Victoria, Queensland and Tasmania. To avoid excluding less digitally connected people, the questionnaire was offered both online and in a large-print paper form with a reply-paid envelope. Of 536 questionnaires returned, 36 were removed for excessive missing data, yielding a final analytic sample of 500. Figure 1 sets out the conceptual model that organised the analysis.

Measures

Loneliness was measured with the six-item De Jong Gierveld Loneliness Scale, comprising three emotional and three social items that are dichotomised and summed to a total score ranging from 0, not lonely, to 6, intensely lonely (De Jong Gierveld & Van Tilburg, 2006). Consistent with ageing research, a total score of 3 or above was treated as indicating at least moderate loneliness. Predictor variables were age group, sex, living arrangement, marital status, remoteness classified against the Australian Statistical Geography Standard, self-rated health, financial strain, regular internet use and participation in a community, religious or interest group.

Analysis and ethics

Analysis proceeded in two stages. Descriptive statistics established prevalence overall and by subgroup, and multiple linear regression estimated the independent contribution of each predictor to the continuous loneliness score. Assumptions of linearity, normality of residuals and multicollinearity were checked, with all variance inflation factors below 2.0. The study was approved by the university Human Research Ethics Committee and conducted in accordance with the National Statement on Ethical Conduct in Human Research (NHMRC, 2023). Participation was voluntary and anonymous, and a plain-language list of support services was provided with every questionnaire given the sensitive nature of the topic.

Results

Sample characteristics and prevalence

The sample comprised 500 older adults, of whom 56 per cent were women. Just over half were aged 65 to 74, one third were aged 75 to 84, and 13 per cent were aged 85 and over. Table 1 reports the composition of the sample alongside mean loneliness scores and the prevalence of at least moderate loneliness within each subgroup. The mean total loneliness score was 2.4 out of 6 (SD = 1.8).

Table 1: Sample characteristics, mean loneliness score and prevalence of moderate to severe loneliness (n = 500)

Across the whole sample, 172 respondents recorded a total score of 3 or above. The prevalence of moderate to severe loneliness was therefore calculated as follows:

Prevalence = (respondents scoring 3 or above / total sample) x 100 = (172 / 500) x 100 = 34.4 per cent.

Prevalence rose steeply with age. Among the 65 respondents aged 85 and over, 30 were at least moderately lonely, giving a subgroup prevalence of (30 / 65) x 100 = 46.2 per cent, compared with 27.4 per cent among those aged 65 to 74. This gradient is consistent with national concern about the oldest old and with the accumulation of bereavement and functional loss in later life.

Predictors of loneliness

The multiple regression model was statistically significant and explained about one third of the variance in loneliness scores, F(8, 491) = 31.4, p < .001, R² = 0.34, adjusted R² = 0.33. Table 2 presents the coefficients. Living alone was the strongest predictor, followed by the two protective factors, community group participation and regular internet use. Poorer self-rated health, widowhood, financial strain, regional or remote residence and advanced age each made a smaller but statistically significant independent contribution.

Table 2: Multiple linear regression predicting the De Jong Gierveld loneliness score (n = 500)

The two protective coefficients are substantively important. Holding other factors constant, regular internet users scored on average 0.52 points lower on the loneliness scale, and members of a community group scored 0.61 points lower, than their counterparts. Because these estimates are adjusted for age, health, living arrangement and remoteness, they cannot be explained away as an artefact of healthier or younger people being more likely to be connected.

Discussion

The results provide an Australian answer to the three research questions. Moderate to severe loneliness affected roughly one in three community-dwelling older adults and nearly half of those aged 85 and over, a prevalence broadly in line with national estimates and with the concerns raised by the Royal Commission into Aged Care Quality and Safety (2021). Loneliness was not randomly distributed; it clustered among people who lived alone, were in poorer health, faced financial strain or lived outside the major cities, confirming that it behaves as a socially patterned determinant of health rather than a matter of individual temperament (Hawkley & Cacioppo, 2010; Courtin & Knapp, 2017).

Digital inclusion

Regular internet use was independently associated with lower loneliness, supporting the view that connectivity can help older people sustain relationships across distance. This finding must be read alongside the digital divide. Because older Australians record the lowest digital inclusion scores nationally (Thomas et al., 2023), programs that shift services and social contact online risk leaving behind the 39 per cent of this sample who did not use the internet regularly, the very group with the highest loneliness. Digital inclusion is therefore best understood as one instrument among several, valuable when paired with device access, affordable data and patient skills support delivered through libraries and seniors’ organisations, and counterproductive if treated as a substitute for human contact.

Community programs

Participation in a community, religious or interest group was the strongest protective factor after living arrangement. This aligns with intervention evidence that activities offering shared purpose, not merely proximity, are what reduce loneliness (Masi et al., 2011). The implication for the Australian aged care and primary health systems is practical. Social prescribing, in which general practitioners and My Aged Care assessors connect older people to local groups and volunteering, offers a low-cost pathway that fits within existing Medicare-supported primary care and the Commonwealth Home Support Programme. Sustained funding and referral infrastructure, rather than short-term pilots, would be required for such programs to reach the socially isolated.

Rurality

Older people in regional and remote areas carried a measurable excess of loneliness even after adjustment, echoing earlier Australian findings on the uneven geography of social connection (Baker, 2012; AIHW, 2024). Thinner transport, fewer services and greater distances plausibly explain the gap. Responses designed for metropolitan settings will not transfer automatically; outreach, transport assistance and support for local volunteer networks are likely to matter more outside the capital cities.

Limitations

Several limitations qualify these findings. The cross-sectional design cannot establish causal direction, and it is likely that loneliness and poor health, in particular, reinforce one another over time. Recruitment through community organisations may have under-represented the most isolated older people, meaning the true population prevalence is plausibly higher rather than lower. Self-report measures are subject to social desirability, and the study did not capture First Nations or residential aged care populations, whose experiences warrant dedicated investigation.

Conclusion

This thesis set out to quantify loneliness among older Australians and to identify what drives it. Using a validated scale and a sample of 500 community-dwelling adults aged 65 and over, it found that moderate to severe loneliness is common, rises sharply with age, and is concentrated among those who live alone, are in poorer health, experience financial strain or live in regional and remote areas. Crucially, two modifiable factors, regular internet use and participation in community groups, were independently protective, pointing to clear levers for policy and practice. The evidence supports a dual response: strengthening place-based community programs and social prescribing within Australia’s aged care and primary health systems, while expanding digital inclusion in a way that supplements rather than replaces human connection. Treating loneliness as a serious, measurable and modifiable determinant of health is a necessary step if an ageing Australia is to age well.

References

Australian Bureau of Statistics. (2023). General social survey: Summary results, Australia. ABS.

Australian Institute of Health and Welfare. (2024). Social isolation and loneliness. AIHW.

Baker, D. (2012). All the lonely people: Loneliness in Australia, 2001-2009. The Australia Institute.

Courtin, E., & Knapp, M. (2017). Social isolation, loneliness and health in old age: A scoping review. Health & Social Care in the Community, 25(3), 799-812.

De Jong Gierveld, J., & Van Tilburg, T. (2006). A 6-item scale for overall, emotional, and social loneliness: Confirmatory tests on survey data. Research on Aging, 28(5), 582-598.

Ending Loneliness Together. (2023). State of the nation report: Social connection in Australia 2023. Ending Loneliness Together.

Hawkley, L. C., & Cacioppo, J. T. (2010). Loneliness matters: A theoretical and empirical review of consequences and mechanisms. Annals of Behavioral Medicine, 40(2), 218-227.

Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10(2), 227-237.

Masi, C. M., Chen, H.-Y., Hawkley, L. C., & Cacioppo, J. T. (2011). A meta-analysis of interventions to reduce loneliness. Personality and Social Psychology Review, 15(3), 219-266.

National Health and Medical Research Council. (2023). National statement on ethical conduct in human research. NHMRC.

Royal Commission into Aged Care Quality and Safety. (2021). Final report: Care, dignity and respect. Commonwealth of Australia.

Thomas, J., McCosker, A., Parkinson, S., & Ganley, L. (2023). Measuring Australia’s digital divide: Australian digital inclusion index 2023. RMIT University and Swinburne University of Technology.

Victor, C. R., & Yang, K. (2012). The prevalence of loneliness among adults: A case study of the United Kingdom. The Journal of Psychology, 146(1-2), 85-104.