Page 25 - IJPS-11-2

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

older adults’ loneliness is the interplay of their needs “Living with household members” was coded as “not
and resources, such that if their resources cannot fulfill living alone.”
their needs, especially in terms of emotional and social Loneliness was measured by the question, “Do you
needs, they will feel lonely (Bedard-Thomas et al., often feel lonely and isolated?” This item, measured on a
2019; Teater et al., 2021). Certain demographic and five-point Likert scale, has been used widely in previous
personality variables, such as older age, being female, studies and has good predictive validity in health outcomes
and lower socioeconomic status, predispose people (Wei et al., 2022; Yang & Gu, 2021). Following previous
to loneliness (Cohen-Mansfield et al., 2016; Pinquart studies (Wei et al., 2022; Yang, 2021) and considering the
& Sorensen, 2001). As such, informed by Andersen’s highly skewed nature of this loneliness measure in the
behavior model (Andersen, 1995), this study attempts to CLHLS, the item was dichotomized to combine ratings
examine the typology and correlates of living alone and of “sometimes,” “often,” and “always” into the “lonely”
loneliness. Specifically, the correlates were categorized as category (coded as 1) and ratings “seldom” and “never”
predisposing, enabling, and need factors. into the “not lonely” category (coded as 0).
1.2. The current study Combining living arrangements, four categories were
generated, namely, living alone and feeling lonely, living
The correlates of living arrangements and loneliness have alone but not lonely, not living alone but lonely, and neither
been well documented in the literature; however, such living alone nor feeling lonely.
research might overlook the complexity of older adults’
circumstances. Thus, as per Andersen’s behavioral Model, We categorized the variables into predisposing,
this study attempted to examine the typology of living enabling, and need factors. Predisposing factors included
alone and loneliness—living alone and feeling lonely, living the following: age (65 – 79, 80 – 89, 90 – 99, and 100+);
alone and not lonely, not living alone but lonely, and neither sex (men vs. women); residence (rural vs. urban); marital
living alone nor feeling lonely—and the correlates as well status (currently married vs. not married); educational
as urban–rural differences in a nationally representative attainment in terms of the number of years of education
sample of Chinese older adults. received (0, 1 – 6, 7+); and occupation before retirement
(white-collar occupation vs. others).
2. Methods Enabling factors included financial independence (yes
2.1. Data vs. no), family economic status (rich vs. not rich), family
emotional support, participation in individual and group
This study used data from the 2018 CLHLS. More details activities, and community old-age care services. Family
about the sampling can be found in previous studies (Gu, emotional support included the following questions:
2008; Gu et al., 2021). Because we focused on adults aged “Whom do you ask first for help when you have problems/
65 and older and given that some variables had missing difficulties?,” “To whom do you usually talk most frequently
data, including loneliness (1,766 missing cases), living in daily life?,” and “Whom do you talk to first when you
alone (259 missing cases), self-rated health (1,238 missing need to tell something about your thoughts?” When
cases), and chronic disease (2,260 missing cases), we family members were selected, the response was coded as
conducted 10 imputations using the multiple imputation 1; otherwise, it was coded as 0. The Cronbach’s alpha for
method for variables with missing values. The final sample this item was 0.87. Participation in individual activities
for the study was 14,469. Ethical approval for this study was measured by four items, namely, gardening, reading
was obtained from the Institutional Review Board of newspapers and/or books, raising domestic animals, and
Peking University (Approval number: 2022-147). watching television and/or listening to the radio. When
older adults participated in at least one of the above
2.2. Measurements
activities, it was considered participation in individual
The dependent variable was typology based on living activities; otherwise, it was considered non-participation.
alone and loneliness. Living arrangements were measured Meanwhile, group activity participation was dichotomized
by the question, “Whom do you live with now?” The and coded based on playing cards and/or mahjong and
answers included “with household member(s),” “alone,” organized social activities. It was coded as participation
and “in a nursing home.” As community home-based care for those who participated in at least one activity, and
is the main way of providing old-age care and older adults otherwise as non-participation. Community services
in China primarily prefer to live in their communities were measured by eight old-age care services, including
or at home, this study focused on community-dwelling recreational activities and psychological support. If at
older adults and excluded institutionalized older adults. least one service was available, community services were

Volume 11 Issue 2 (2025) 19 https://doi.org/10.36922/ijps.4184