Page 26 - IJPS-11-2

International Journal of
Population Studies Living alone and loneliness in older adults

considered to be available; otherwise, they were considered disabled, and poor self-rated health. Men had a higher
to be unavailable. likelihood (b = 0.257, p < 0.001) of living alone and feeling
Finally, need factors comprised health-related lonely, whereas currently married (b = −3.926, p < 0.001)
variables, including the activities of daily living (ADL), the and rich older adults (b = −0.736, p < 0.001) had a lower
instrumental activity of daily living (IADL), Mini-Mental likelihood of living alone and feeling lonely. Regarding
State Examination (MMSE) scores, self-rated health, and enabling factors, family emotional support (b = −0.536,
chronic disease. Participants were classified as disabled p < 0.001) and individual activity participation (b = −0.452,
in terms of ADL or IADL if they reported difficulty in p < 0.001) were related negatively to the typology of living
one of the six basic ADL activities (e.g., bathing and alone and feeling lonely. Regarding need factors, older
dressing; α = 0.90) or one of the eight IADL activities (e.g., adults who were in poorer health were less likely to live
cooking and shopping; α = 0.94). Cognitive functioning alone and be lonely (ADL disabled: b = −0.950, p < 0.001;
was measured using the Chinese version of the MMSE, SRH: b = −0.938, p < 0.001, respectively).
which has a maximum score of 30 (α = 0.89; Lagona & In addition, compared with the reference category,
Zhang, 2010). It was coded as a binary variable: cognitive the profile for older adults living alone but not lonely was
impairment (0 – 23) and cognitively normal (24 and above; younger age, currently not married, poor family economic
Yang, 2021). Self-rated health (SRH) was measured by the status, non-ADL disabled, and good self-rated health.
question, “How do you rate your health at present?” It was Currently married (b = −2.894, p < 0.001) and rich older
dichotomized and recorded as “good” and “bad” based adults (b = −0.357, p < 0.001) were less likely to live alone
on a five-point Likert scale. Finally, chronic disease was but not be lonely. Regarding need factors, older adults in
measured by whether the participants had chronic diseases poorer health were less likely to live alone but not be lonely
diagnosed by physicians. compared with those with good SRH (b = 0.149, p = 0.025),
who were more likely to live alone but not be lonely (ADL-
2.3. Statistical analysis disabled: b = −0.976, p < 0.001).
First, a descriptive analysis of the variables was performed Compared to the reference category, the characteristics
using percentages and frequencies. The urban–rural of older adults who were not living alone but were
difference in the variables was analyzed using the chi- lonely were younger age, urban residence, currently
square test. Second, considering that the dependent not married, being illiterate, not having a white-collar
variable was a four-category variable, we used multinomial occupation before retirement, poor family economic
logistic models. Not living alone and not feeling lonely status, no individual activity participation, unavailability
was regarded as a reference category. In addition, the of community services, ADL disabled, IADL disabled,
multinomial logistic models were also separately run cognitive impairment, and poor self-rated health. Urban
among urban and rural older adults to examine urban– older adults had a higher likelihood (b = 0.094, p = 0.045),
rural differences. All analyses were performed using Stata whereas currently married (b = −0.848, p < 0.001),
v17.0.
educated for 1 – 6 years (b = −0.137, p = 0.020), white-collar
3. Results occupation before retirement (b = −0.278, p = 0.003), and
rich older adults (b = −0.426, p < 0.001) were less likely to
3.1. Descriptive results not live alone but be lonely. Regarding enabling factors,
As Table 1 shows, approximately 7.84% of people surveyed individual activity participation (b = −0.192, p = 0.001)
lived alone and felt lonely, 8.69% lived alone but did not feel and the availability of community services (b = −0.105,
lonely, 20.13% did not live alone but felt lonely, and 63.34% p = 0.022) were related negatively to the typology of not
neither lived alone nor felt lonely. Except for chronic living alone but feeling lonely. In terms of need factors,
diseases, differences were observed in the four types older adults in poorer health were more likely to not
of living alone and loneliness in terms of predisposing, live alone but be lonely, whereas those with good SRH
enabling, and need factors. (b = −0.825, p < 0.001) were less likely (ADL disabled:
b = 0.143, p = 0.013; IADL disabled: b = 0.302, p < 0.001;
3.2. Multinomial logistic regression results cognitively impaired: b = 0.130, p = 0.025, respectively).
As Table 2 shows, compared with the reference category, The results for those living alone and feeling lonely
the profile for older adults who lived alone and were and living alone but not feeling lonely were largely similar
lonely was younger age, older men, currently not married, among urban and rural older adults. However, for urban
poor family economic status, no emotional support from older adults, those with a non-white-collar occupation
the family, no individual activity participation, not ADL before retirement and those with cognitive impairment

Volume 11 Issue 2 (2025) 20 https://doi.org/10.36922/ijps.4184