Patterns of Multidimensional Social Exclusion Among Older Home Care Services Recipients

Abstract

Objectives

The guiding principle of current aging policies has been to promote older adults to live in their private homes, but little attention has been paid to social exclusion of older adults receiving home-based care. The aim of this study is to increase understanding on different patterns of multidimensional social exclusion among older adults receiving formal home care services, and through this to shed light on the possible challenges of current aging-in-place policies.

Methods

The survey data were collected in 2022 among older adults aged 65 to 102 years receiving home care services in Finland and merged with administrative data (n = 733). A latent class analysis was used to identify different types of social exclusion. Multinomial logistic regression modeling examined factors associated with different social exclusion types.

Results

Four social exclusion types were identified: (1) not excluded (16.9%), (2) homebound economically excluded (40.1%), (3) excluded from social relations (28.6%), and (4) multidimensionally excluded (14.3%). Poor self-rated health and poor functional ability significantly increase the risk of being multidimensionally excluded or homebound economically excluded. The group using home care and medical services the most are the most multidimensionally excluded. The group living in urban areas are more likely to be excluded from social relations.

Discussion

Different types of social exclusion should be acknowledged when addressing social exclusion among home care clients. Enhanced measures should be developed to support older adults using home and healthcare services the most, as they are at high risk of severe exclusion.

Older adults with functional limitations and care needs are at risk of different forms of social exclusion (Nilsen et al., 2022; Villar et al., 2021; Walsh et al., 2020). In many countries such as Finland, aging policies have aimed to support older adults’ abilities to live at home instead of long-term care facilities (Act on Supporting the Functional Capacity of the Older Population and on Social and Health Services for Older Persons 980/2012, Section 14; Ministry of Social Affairs and Health, Finland, 2012; see also Vasunilashorn et al., 2012). Living at home with home care services has been seen as a way to promote the idea of “aging in place”; pursuing a meaningful life by independence and self-management in familiar surroundings (Pani-Harreman et al., 2020).

Despite good intentions, aging-in-place-driven policies and the constant drive to save public resources may lead to a situation where older adults must cope in their private homes without adequate care and support (Kröger, 2022; Rostgaard et al., 2022). This raises concerns about whether older adults, even when receiving home care, still have unmet needs, which may transform their home into an unsafe place with feelings of insecurity and isolation, for example (Sihto & Van Aerschot, 2021). In this study, we focus on older adults who live in their private homes and receive formal home care services from the perspective of multidimensional social exclusion (see also Grenier & Guberman, 2009). Our aim is to increase understanding of different patterns of old-age social exclusion among home care services recipients and through this shed light on the possible challenges of current aging policies.

The context of the study lies in Finnish formal home care services, which consist of care and attention, promoting and maintaining functional ability and interaction, and home nursing (Social Welfare Act 1301/2014, Section 19; Ministry of Social Affairs and Health, Finland, 2014). In this study, regular home care recipient refers to a person who receives publicly organized regular home care, that is, at least one home care visit per week, and if necessary one or more support services such as a safety phone, meal service, and hygiene service. Regular home care is offered to older adults who are not able to manage activities of daily living (ADL) independently, with informal care, or with support services. Support service recipient refers to a person receiving only support services but is not regularly visited by home care workers. Here, we use the term home care services recipients indicating the overall target group of our study consisting of both client groups: older adults receiving regular home care and/or support services.

Multidimensional Old-Age Social Exclusion—Home Care Perspective

Social exclusion is widely understood as a complex and multidimensional process related to the lack of resources, rights, goods and services, and inability to participate in relationships and activities (Levitas et al., 2007). The multidimensionality of old-age social exclusion has been recognized over the last decades in conceptualizations and frameworks developed by social gerontologists (e.g., Scharf et al., 2001; Walsh et al., 2021). In this study, we recognize social exclusion as a dynamic phenomenon consisting of economic, social, services, community and spatial, and civic dimensions (Van Regenmortel et al., 2016; Walsh et al., 2017, 2021; see also Ristolainen et al., 2024). Grenier and Guberman (2009) have highlighted the need to focus on different forms of exclusion especially when examining disadvantaged older people in the context of home care. However, in previous research, knowledge on social exclusion of older home care recipients has been fragmented and has mostly focused on exclusion from social relations and unmet care needs.

Several studies have shown that home care services mainly focus on safety and independence in personal care and supporting the ADL, such as dressing and eating (e.g., Levasseur et al., 2014; Turjamaa et al., 2014), whereas needs related to social activities are more likely to remain unmet (Ristolainen et al., 2024; Turcotte et al., 2015). An Australian study indicated that around one third of older adults receiving home and community-based care saw their friends or attended community activities never or rarely (Siette et al., 2021). A study in Sweden showed that older home care recipients were less satisfied with their social networks compared to older adults receiving no care (Jarling et al., 2022). Moreover, inadequate home care services have been found to negatively affect social inclusion of home care recipients (Seidlein et al., 2020).

In previous studies focusing on older home care clients, factors associated with lower levels of social support and networks include decreased physical functioning, declined mental well-being (Dale et al., 2010), and lower cognitive function (Siette et al., 2020). Moreover, getting inadequate help and support has been found to be associated with loneliness and social isolation among home care clients in Finland (Ristolainen et al., 2024). From the perspective of home care professionals, a qualitative German study indicated that workers are concerned about older home care clients living in poverty and having limited opportunities for participation in social contacts and access to medical care (Messer, 2019). Additionally, some studies have shown that care workers are ethically burdened because they encounter home care recipients whose needs cannot be met due to restricted resources, for example (Ring et al., 2023; Vik & Eide, 2012). These previous findings indicate that older home care clients are at risk of different and often overlapping dimensions of social exclusion.

Also, from the broader perspective of aging research, a large body of literature has focused on domains of old-age social exclusion separately. However, some studies have examined the different dimensions of exclusion together by counting the numbers of dimensions excluded (Barnes et al., 2006; Macleod et al., 2019; Prattley et al., 2020) or by constructing a scale ranging from low to high levels of exclusion (e.g., Panek & Zwierzchowski, 2022), which easily loses the multidimensionality of the phenomena (Van Regenmortel et al., 2018). Despite careful development of composite measures for assessing multidimensional social exclusion there is lack of consensus on the most suitable method (Keogh et al., 2021). Preserving the multidimensionality of exclusion is important, as disadvantages and domains of exclusion co-occur in old age (Aartsen et al., 2023; Heap & Fors, 2015; Van Regenmortel et al., 2018), and diverse types of social exclusion require different preventions, interventions, and policy actions.

Recently, some researchers have aimed to overcome the challenges of studying multidimensional social exclusion by examining different types or profiles of exclusion among community-dwelling older adults (e.g., Aartsen et al., 2023; Van Regenmortel et al., 2018). In a Balkan study, four types of old-age exclusion were found, namely “low social exclusion risk,” “material exclusion,” “material and social exclusion,” and “multidimensional exclusion” (Aartsen et al., 2023). A Belgian study also found four profiles of old-age exclusion, namely “low risk,” “nonparticipating financially excluded,” “environmentally excluded,” and “severely excluded” (Van Regenmortel et al., 2018). The methodological orientation of our study draws from these previous studies, which have provided understanding of how to study different patterns of old-age social exclusion.

Based on existing knowledge, older home care recipients have unmet needs and are at risk of social exclusion, but this has been poorly recognized when evaluating and developing care services for older adults. Due to this and the fragmented research on old-age exclusion in home care, there is a clear need to examine multidimensional social exclusion from the perspective of older people who receive home care services and are at the forefront of “aging in place” policies. Our study draws on the research project called “Old-age social exclusion in home care—prevalence, meanings, and intervention” (SOLDEX) and examines survey and administrative data collected among older home care services recipients in Finland.

Research Aims

The aim of this study is to increase understanding on different patterns of old-age social exclusion in the context of formal home care services. To meet this aim, in our analysis, we:

1. Identify the types of social exclusion among home care services recipients.

2. Examine which factors are associated with different social exclusion types.

Method

Study Design and Data Collection

The data are derived from “Old-age social exclusion in home care—prevalence, meanings, and intervention” (SOLDEX) project (2021–2025). The main aim of the project is to gain understanding of the everyday lives and social exclusion of older home care clients and ways of reducing old-age exclusion in the context of home care. As older adults with poor health and increased care needs are often excluded from large population surveys (de Souto Barreto, 2012) and no representative data for home care recipients in Finland were available, this study therefore used newly collected survey data from home care services recipients, merged with administrative data accessed via Finnish Social and Health Data Permit Authority (Findata).

The survey data were collected in 2022 at two study sites in Finland, the city of Kuopio and Kainuu Social Welfare and Health Care Joint Authority. We used a total sampling covering both study sites. A questionnaire, information letter, two informed consent forms, and prepaid envelope were sent to all 65+ aged regular home care recipients (n = 2,284) at both study sites and to all 65+ aged support service recipients (n = 858) at one study site in early May. A letter reminding recipients to respond to the survey was sent to nonrespondents at the end of June. In September, we sent one more reminder letter together with study information, two consent forms, and the questionnaire. A total of 925 home care services recipients responded to the survey. The response rate was 29.4%. Three hundred and sixty-six of the respondents were support service recipients and 559 were regular home care recipients.

The participants were encouraged to ask for help if in need of assistance in completing the survey. Approximately half of the respondents were assisted by a family member, friend, researcher, home care worker, or other person. This meant reading the questions and filling in the survey, but not responding on behalf of the participant. Eight respondents were excluded because they had not answered the questions themselves. Finally, 917 respondents were eligible for the study, but after the data screening and due to missingness in social exclusion variables, 733 participants were included in this study. We compared the basic sociodemographic information of the participants excluded from the analyses (n = 184) to participants included in this study (n = 733). A statistically significant difference was found in education indicating that participants with higher education were better represented in the analysis group. However, participants with higher education were a relatively small group compared to participants with basic or secondary education (see Table 4), and therefore this is not expected to cause a significant bias in the results.

The survey data were merged with the administrative data, which consists of information extracted from care registers in Finland (Hilmo, Avohilmo, and Sosiaalihilmo). Care registers are based on use of health and social care services in terms of basic outpatient social and healthcare, institutional social care, and specialist healthcare. The care registers are from the period May 1, 2021, to December 31, 2022, which is approximately 12 months before and 8 months after the survey was sent.

Measures

Building on the framework of multidimensional social exclusion (Walsh et al., 2021), we utilized multiple scales and questions in the survey (see also Ristolainen et al., 2024). In this study, each dimension of social exclusion (economic, social relations, services, community, and spatial, civic) includes two subdimensions leading to the use of 10 different indicators (Table 1). Subdimensions were dichotomized into 0 (not excluded) and 1 (excluded).

Economic exclusion includes (1) inability to make ends meet (0 = easily or fairly easily; 1 = with some or great difficulty), and (2) material deprivation measured by asking whether the respondent is able to afford the following items: groceries, holidays, clothing, heating, glasses (Adena et al., 2015), and medicine (0 = two or fewer items cannot be afforded; 1 = three or more items cannot be afforded).

Exclusion from social relations includes (1) social isolation measured by the Lubben Social Network Scale, which ranges from 0 to 30 with lower scores indicating higher social isolation (Lubben et al., 2006) (0 = not isolated if scale score 12 or higher; 1 = isolated if scale score lower than 12), and (2) loneliness (0 = feeling lonely never, rarely, or sometimes; 1 = feeling lonely fairly often or all the time).

Exclusion from services includes (1) inadequacy and inaccessibility of social and healthcare services such as home care and support, residential services, transportation, personal assistant, social worker or case manager, general practitioner (GP), public health nurse, and dentist or dental hygienist (0 = none or one service inadequate or not accessed; 1 = two or more services inadequate or not accessed), and (2) inadequacy of help related to ADL, which include bathing, eating, getting out of and into bed, toileting, and dressing (Kröger et al., 2019) (0 = getting enough help for all activities if needed; 1 = not getting enough help for one or more activities).

Community and spatial exclusion includes (1) feeling insecure measured by asking whether the respondent feels insecure in their apartment, when home alone at night, when moving about in the local area, and due to remote location of the apartment (0 = feeling insecure sometimes or often in at most one item; 1 = feeling insecure sometimes or often in two or more items) and (2) not getting to go outside by asking for their thoughts on the following statement: I get to go outside and visit the places I want enough (0 = fully or somewhat agree; 1 = fully or somewhat disagree).

Civic exclusion includes (1) exclusion from civic participation measured by asking to what extent the following issues are true for you: following politics, discussing politics, membership of political society or organization, voluntary work, residential activities (see Taló & Mannarini, 2015) (0 = (a) having fewer activities that cannot be done compared to activities done often or sometimes, or (b) all activities are never done and there is no interest to do; 1 = more or equal number of activities that cannot be done compared to activities done often or sometimes), and (2) age discrimination based on two indicators: perceived discrimination of pensioners in society and self-experienced age discrimination (0 = not agree/somewhat agree and no; 1 = one or both items fully agree or yes; Table 1).

Other variables from the survey used in the analyses were age, gender, education, living alone or with someone, residential area, informal care, self-rated health (SRH), and functional ability. From the administrative data we used variables which indicated the amount of home care visits and the amount of emergency and GP visits.

Age was described in years. Gender had two categories “male” and “female.” Education was recoded into basic (primary school), secondary (vocational school/high school/community college), and higher (university of applied sciences degree/master’s degree/doctoral degree) education. Living status was based on the question, “How many people live in the same household with you?” and was dichotomized into “living alone” and “living with someone.” Residential area was recoded into urban area (city center/suburb) and rural area (village/countryside). Informal care was asked about using the question, “Is your spouse, child or other person your caregiver?,” with response options “yes” or “no.” Self-rated health was based on the question, “How do you feel about your current health status?,” with five response categories, which were recoded into good (very good/rather good), average (average), and poor (rather poor/very poor) health. Functional ability was based on five items of ADL described above and out-of-home mobility as an item of instrumental ADL (0 = copes without difficulties; 1 = does not cope independently). The values were summed to form a scale from 0 to 6 and then transformed to three categories: 0 = good, 1–3 = moderate, and 4–6 = poor.

Nonvirtual home care visits were counted over a 3-month period after the survey was sent. The participants were categorized according to the average number of home visits they received: 1 = 0–3 visits per month (support service recipients), 2 = 1–13 visits per week, and 3 = 2 or more visits per day. Emergency and GP visits to healthcare units or at home over a 20-month period (from May 1, 2021, to December 31, 2022) were summed to the total number of visits.

Analysis

To identify different types of social exclusion among older home care service recipients, we used latent class analysis (LCA), which has been used for analyzing associations in multivariate categorical data and increasingly to identify latent classes or typologies in research (McCutcheon, 1987; Weller et al., 2020). LCA clusters the categorical data and examines if distinct types of exclusion exist, and which dimensions of exclusion are more likely to co-occur (e.g., Aartsen et al., 2023; Van Regenmortel et al., 2018). A primary assumption of LCA is local independence, assuming that indicators are uncorrelated within each class (Visser & Depaoli, 2022), whereas the associations of the observed variables are attributable to the unobserved (i.e., latent) variable (McCutcheon, 1987).

Before running the LCA, respondents who were not socially excluded in any of the domains were excluded from the analysis, and “not excluded” became our first category. The sample size in LCA was 609 cases, which can be considered large enough because 500 cases is a sufficient sample size to support the validity of the LCA analysis (Nylund et al., 2007). Next, stepwise, different models were run by adding one class each time. To assess the model quality, we applied a combination of indicator variables, meaning that various fit indices as well as the conceptual understanding of the researched concept (social exclusion) were considered, as outlined by Nylund-Gibson and Choi (2018) and Weller et al. (2020, p. 292). Assessment of model fit employed several statistical criteria, including the Bayesian Information Criterion (BIC), the adjusted BIC, and the Akaike Information Criterion (AIC), where a lower BIC, adjusted BIC, and AIC value indicated superior model fit (Nylund et al., 2007; Nylund-Gibson & Choi, 2018; Weller et al., 2020); the Vuong–Lo–Mendell–Rubin adjusted likelihood ratio test, with a p value > .05 indicating statistical inferiority to a model with fewer classes (Lo et al., 2001; Nylund et al., 2007; Vuong, 1989). Frequently, the compilation of fit indices does not point toward a singular model. Rather, it is more prevalent for the fit indices to endorse one or two potential models. Therefore, we also considered latent class size (each more than 5% of the sample) and aimed for an entropy of 0.8 as it indicates a “good” classification of individuals into classes (Nylund-Gibson & Choi, 2018).

In order to examine which factors are associated with different types of social exclusion, we firstly performed bivariate analyses for relevant independent variables using the chi-square test for categorical variables, ANOVA for age, and Kruskall–Wallis test for emergency and GP visits due to non-normal distribution. Secondly, we performed multinomial logistic regression modeling (Kwak & Clayton-Matthews, 2002) to analyze whether there are multivariate associations between the selected independent variables and social exclusion types. We included age, gender, education, and living status as control variables and other variables if they were significantly associated with social exclusion types based on bivariate analyses. The multidimensionally excluded group was set as a reference category. We report relative risk ratios (RRR), which indicate the risk of being in the comparison category compared to the risk of being in the reference category in terms of the variable in question. Supplementary analyses were conducted to ensure a comprehensive presentation of the results by setting not excluded as a reference category.

LCA was analyzed using Latent GOLD 6.0 and other analyses were performed with StataIC 16.

Ethics

The research has been pre-assessed by the research ethical committee of the University of Eastern Finland. Informed consents were received from all participants.

Results

Leaving out the predefined subgroup of those not at risk of exclusion (n = 124), the LCA revealed several types of social exclusion. The fit statistics, decision criteria, and relative size of social exclusion types based on the LCA are presented in Table 2. We explored models with 2 to 7 latent classes. Different fit indices point toward different choices of models, which is common in LCA models (Nylund-Gibson & Choi, 2018). The lowest BIC was found in the two-class model, and the lowest adjusted BIC in the four-class model. The lowest AIC points toward the six-class model. Entropy was best in the three-class and six-class models. Based on interpretability, we did not include models with classes with a relative size below 0.05. This left us with the two-, three-, or four-class models. Considering the combination of the second-best BIC (BIC = 6,467.97), a better AIC than the two-class model (AIC = 6,326.79) and the higher entropy R² (Entropy R² = 0.78), the three-class model was selected.

The LCA identified three exclusion subgroups, which we defined as types of “homebound economically excluded” (cluster 1), “excluded from social relations” (cluster 2), and “multidimensionally excluded” (cluster 3). Home care clients in cluster 1 had a moderate probability of being excluded from getting to go outside and a moderate probability of being materially deprived. Cluster 2 consisted of home care clients having a high probability of being socially isolated and a moderate probability of feeling lonely. Finally, cluster 3 entailed older home care recipients combining a high probability of exclusion in almost all indicators (Table 3).

Combining the predefined subgroup of those not at risk of exclusion and assigning the other respondents to their best-fitting cluster by applying modal assignment (i.e., respondents were assigned to the cluster for which posterior class membership probabilities were the highest), we arrived at the following prevalences: 16.9% of the older home care clients are not excluded, 40.1% can be considered as homebound economically excluded, 28.6% are excluded from social relations, and 14.3% are multidimensionally excluded.

Table 4 presents descriptive statistics for all sociodemographic and other independent variables by social exclusion types. Mean age of the total sample was 84.4 (SD 7.3), most of the participants were women (74.1%), most of them lived alone (82.5%), and most lived in urban residential areas (67.4%). More than half (55%) were regular home care recipients as they were visited by a home care worker once a week or more, and the rest (45%) were support service recipients who had home care visits less than once a week.

According to bivariate analyses, there were significant differences in residential area (p = .031), SRH (p < .001), functional ability (p < .001), amount of home care visits (p < .001), and amount of GP and emergency visits (p < .001) across the four types of social exclusion. Those home care services recipients excluded from social relations were more likely to live in urban areas than those in the other social exclusion groups. Conversely, the multidimensionally excluded are overrepresented among those living in rural areas. SRH and functional ability were better among those who are not excluded in any social exclusion domain compared to those belonging to any other types of social exclusion. Home care services recipients with poor SRH are overrepresented among those being multidimensionally excluded or homebound and economically excluded. Additionally, those having poor functional ability are overrepresented in multidimensionally excluded home care services recipients. No significant differences in informal care were observed across the four social exclusion types. The multidimensionally excluded group had the highest number of home care visits, as well as GP and emergency visits, whereas those not excluded had the lowest number of visits.

The results from multinomial logistic regression (Pseudo R² = 0.12; Chi-square = 219.64, p < .001) are presented in Table 5. Of background variables, older age increases the likelihood of being not excluded rather than being multidimensionally excluded (RRR = 1.08, p = .001), but age is not significantly associated with other social exclusion types. There were also no differences in gender or education across the social exclusion types. In addition, home care services recipients living alone are less likely to be not excluded (RRR = 0.24, p = .002) and less likely to be homebound economically excluded (RRR = 0.34, p = .006) rather than multidimensionally excluded. Living in rural area is associated with a lower risk of being excluded from social relations (RRR = 0.48, p = .013) rather than being multidimensionally excluded. Supplementary analysis (Supplementary Table 1), which use not excluded as the reference group, indicated that home care services recipients living alone (RRR = 2.19, p = .028) or living in urban areas (RRR = 0.49, p = .012) have increased risk of being excluded from social relations rather than being not excluded.

Poor SRH and poor functional ability significantly increase the odds of being multidimensionally excluded rather than being not excluded (SRH, average, poor: RRR = 0.30, p = .018 and RRR = 0.05, p < .001; functional ability, moderate, poor: RRR = 0.27, p = .002 and RRR = 0.04, p < .001), being excluded from social relations (SRH, poor: RRR = 0.12, p < .001; functional ability, poor: RRR = 0.15, p < .001), or even in some respects being homebound economically excluded (SRH, average, poor: RRR = 0.37, p = .039 and RRR = 0.22, p = .001; functional ability, poor: RRR = 0.29, p = .007). However, home care services recipients with poor SRH (compared to good; RRR = 4.29, p < .001) or poor/moderate functional ability (compared to good; RRR = 2.99, p < .001 and RRR = 6.65, p = .001) are significantly more likely to be homebound economically excluded than to be not excluded (see Supplementary Table 1).

Having more home care visits is associated with a lower likelihood of being not excluded (RRR = 0.37, p = .016 and RRR = 0.28, p = .003) and being homebound economically excluded (RRR = 0.30, p = .007) rather than being multidimensionally excluded. In general, those with the lowest number of home care visits were more likely to be not excluded rather than to be excluded, irrespective of the social exclusion type (Supplementary Table 1). Similarly, older home care recipients with more GP and emergency visits have a slightly lower likelihood of being not excluded (RRR = 0.92, p = .025) and being homebound economically excluded (RRR = 0.94, p = .013) rather than being multidimensionally excluded.

Discussion

In this study, we have examined patterns of old-age social exclusion in the context of formal home care by identifying different types of social exclusion and their associated factors among older adults receiving home care services. We found that the majority of study participants (83.1%) were socially excluded on one or more subdimensions and only 16.9% were not excluded at all. Based on LCA, the socially excluded participants were clustered into three types of social exclusion: homebound economically excluded (40.1%), excluded from social relations (28.6%), and multidimensionally excluded (14.3%).

When examining the different types of social exclusion in terms of associated factors, we found that the multidimensionally excluded were more likely to have the poorest health and functional ability and have the most home care and GP/emergency visits. Home care clients living in rural areas were also more likely to be multidimensionally excluded than excluded from social relations. These findings indicate an accumulation of disadvantages within a small but recognizable group of older people receiving home care services. Similar results have also been found in previous studies among older adults in general (Heap & Fors, 2015; Van Regenmortel et al., 2018).

Based on the findings, home care services recipients considered as homebound economically excluded had poorer SRH than participants who were not excluded. This group was not visited by home care workers as often as the multidimensionally excluded. The homebound economically excluded also appeared to be somewhat excluded from social and healthcare services based on LCA. This may be due to the older home care clients’ inability to make ends meet or having difficulties due to physical accessibility challenges (see also Qiu et al., 2010).

The participants who were considered as excluded from social relations had better health and functional abilities compared to the multidimensionally excluded and the homebound economically excluded. In this group, the main unmet need appeared to be that the participants do not have enough close relationships. Despite better functional abilities, the excluded from social relations had more home care visits compared to the homebound economically excluded. Additionally, being excluded from social relations was more likely for home care services recipients living in urban areas. The urban–rural distinction has been found to be limited (Victor & Pikhartova, 2020), but in the context of our study it is likely that older adults living in rural areas have lived there for a long time and therefore have more social contacts close by (see also Hennessy & Innes, 2021).

Our findings show that home care services recipients are more likely to be not excluded if they are older and have better health and functional ability. The result regarding age contrasted with our expectation and were somewhat unexpected as poor health and increased care needs are generally known to be associated with older age (Gobbens et al., 2010). However, similar findings were found in a study on social exclusion of older population living in Balkan states (Aartsen et al., 2023) and in studies not solely focusing on older adults (e.g., Dykxhoorn et al., 2024). Additionally, in contrast to earlier findings (e.g., Aartsen et al., 2023; Ogg, 2005), we found no differences in gender and education across the social exclusion types of older adults. The nonsignificance of gender could point to a less gender-based sociocultural context in Finland (Al-Rashid et al., 2023), and therefore this finding might not be consistent across other countries and cultures (however, see, e.g., Miranti & Yu, 2015; Scharf et al., 2005). Overall, our findings indicate that in this specific target group of home care recipients, health, functional ability, and access to adequate care and support play a more vital role than gender and education in social exclusion.

Limitations

Our study has some limitations. We compared the study sample to all 65+ regular home care recipients in Finland in terms of age and gender. The distribution of age was almost similar, but women are overrepresented in our sample, which may have caused bias to the results. During data collection, we found that the older home care services recipients with very poor functional abilities or severe cognitive decline were not able to participate in the study. Therefore, the sample is not fully representative, and the results may show an overly positive picture of social exclusion among older home care services recipients. Additionally, the data is cross-sectional, which only allowed us to examine the association of some factors with the types of social exclusion. The data for example did not include retrospective information about life course factors, which have been recognized as important when studying multidimensional exclusion (Walsh et al., 2021).

The questionnaire was constructed based on indicators for which Finnish versions were available. For this reason, the measures of community and spatial exclusion are rather limited. Moreover, we have used thresholds to determine the exclusion for each subdimension, which eliminates variation in the degree of exclusion, but may still be reasonable for constructing variables for further analyses assessing the multidimensionality (see Keogh et al., 2021). For performing the LCA, we had to decide between two quite equal models. As the criteria are not strict, it would have been possible to choose a different model.

Conclusions and Implications

Our study has shown that most older adults receiving formal home care services may be regarded as socially excluded and that different types of social exclusion appear in this population. The different types of social exclusion and their associated factors should be acknowledged when aiming to prevent and reduce social exclusion of older adults receiving home-based care services.

Professional carers working daily with older home care clients play a key role in identifying and addressing old-age social exclusion. Care workers form an important interpersonal network for older clients by providing social support and enabling social inclusion (see also Norvoll et al., 2022). Moreover, special attention should be paid to the home care clients who already receive and use home care and healthcare services the most. This group appears to be at high risk of severe multidimensional exclusion and would benefit from even more intensive care and support in their daily lives. This care and support not only need to accommodate healthcare needs but also focus on reducing social exclusion, with attention, for example, to financial resources, civic participation, spatial inclusion, and access to community services.

Our main findings raise questions and concerns about the adequacy of current home care services in Finland and whether the private home in fact offers the possibility to receive the support and care needed to live a meaningful life. Considering the costs, it could be argued that aging policies that have aimed to reduce use of care services (e.g., Rostgaard et al., 2022) may in fact create a situation where exclusion of older adults leads to more costs both on the individual and societal levels. In summary, the study indicates that current aging-in-place-driven policies may in fact create and maintain social exclusion among older people living at home with multiple care needs (see also Grenier & Guberman, 2009). Are aging-in-place policies really designed to accommodate people with complex care needs? While aging in place often is idealized as “desired,” “needed,” or “what older people want themselves,” it has the danger to lead to older people being “stuck-in-place” (Russel, 2024). Our study underlines the prerequisite of adequate home care and support services, and the importance of recognizing the formal support structures as essential when aiming to promote meaningful life in the community (see also De Donder et al., 2024). Finally, there is a need for a more diversified approach to housing and care arrangements for older adults instead of the current polarized system of living at home or in an institution, while considering the continuity of care and older people’s own views (Barken, 2019).

We recognize that our study is country-specific, and the results are most relevant to the Finnish context, where aging policies have already focused on home-based care for several decades. It may well be that in other countries the home care services situation is better and older adults’ needs and rights are met in and outside their private homes. However, we believe that the study provides better understanding of old-age social exclusion also beyond the Finnish context and that the findings can be utilized when organizing and developing care and services for older populations across countries.

Funding

This work was supported by “The Research Council of Finland” (grant number #342267).

Conflict of Interest

None.

Author Contributions

H. Ristolainen designed the study, collected and processed data, carried out explanatory analyses, and drafted and revised the manuscript. S. Van Regenmortel designed the study, and drafted and revised the manuscript. L. De Donder designed the study, and commented and revised the manuscript. T. Vercauteren carried out the LCA analysis, and commented and revised the manuscript. J. Lehtiö collected data and revised the manuscript. E. Tiilikainen designed the study, supervised data collection, and drafted and revised the manuscript. All authors have read and approved the manuscript as submitted.

References