Impact of Sociodemographic Factors on Use of Formal Social Services in an Older Swedish Population

Background: In Sweden, societal support for older people is the responsibility of the municipalities. However, due to Sweden’s current aging-in-place policy for older people, there is a need to assess how the use of such services varies based on sociodemographic factors. The aim of this study was to describe the use of different forms of social services and institutional long-term care (ILTC) in an older population and to evaluate the impact of sociodemographic factors. Methods: This was a cross-sectional register-based study, including all individuals aged 65 years and older in two Swedish municipalities in 2010, 2015, and 2019. The study analyzed the use of social services and ILTC in relation to sex, place of birth, cohabitation status, and type of housing. Results: Women, those born in Sweden, and those living in an apartment were more likely to receive assistance than men, those born abroad, or living in single family houses, respectively. People living alone were consistently more likely to have assistance, as well as ILTC. Conclusions: There may be a discrepancy between the individual’s need and the assistance provided from the municipality in certain sociodemographic groups in the older population in Sweden.


Introduction
In Sweden, as well as worldwide, the population increasingly tends to live longer. Between 2002 and 2019, the life expectancy from birth increased in Europe (from 80.9 to 84.0 years for women, and from 74.3 to 78.5 years for men) [1]. In Sweden, people aged 65 years and older had similar increase in life expectancy, i.e., about four years, during the past decades.
In Sweden, care for the old is a part of the welfare system. The Swedish aging-in-place policy implies that older people should live in their own housing for as long as possible and be provided with the necessary social services to do so. This policy relies on the Social Services Act [2], which regulates the services and support aimed at those needing it to be able to live an independent life in their own home. The aim of social services for older people is to enable a dignified and good life. It is the responsibility of the National Board of Health and Welfare, a government agency under the Ministry of Health and Social Affairs, to ensure that older people have the possibility to live independently, under safe conditions, and have an active and meaningful existence in community with others. In contrast to many other countries, the responsibility for societal support and service for older people in Sweden lies on the municipalities, not relatives or private operators [3]. Thus, the municipality is responsible for ensuring that older people receive good housing and the support and services that they need [2]. This could also include the health care This was done in in two sociodemographically different municipalities in southern Sweden, to establish if there were general differences in terms of the provided social services that might serve as fundament for future efforts and distribution of social care among older people.

Study Design
This was a cross-sectional register-based study on people aged 65 years or older living in Malmö and Kristianstad, two municipalities in the most southern part of Sweden (Scania), during 2010, 2015, and 2019. As described in more detail below, register data were used to identify the study population, as well as to collect information on sociodemographic factors and social services, ILTC, and mobility devices.

Study Context
Measured by number of inhabitants, Malmö is the third largest municipality in Sweden, with a population ranging from around 299,000 to 344,000 during the study years [15]. However, geographically, Malmö is one of the smaller municipalities in Sweden, covering an area of around 330 square kilometers. Thus, it is one of the most densely populated municipalities in Sweden (2193 inhabitants/km 2 ). In contrast, Kristianstad is the 25th largest municipality in Sweden with a population, ranging from around 79,500 to 85,700 during the study years, and covers an area of 1 800 square kilometers (69 inhabitants/km 2 ) [15]. The vast difference in population size and density between the two municipalities contributes to disparities in sociodemographic conditions, which may impact need and provision of care and social service, as well as the ability of the municipalities to provide social services.

Data Sources
This study was based on three national registers. The register of the total population (RTB) is maintained by Statistics Sweden, the agency responsible for official statistics and other government statistics. It comprises demographic data on all people residing in Sweden. Through this, we identified all people aged 65 years and more and living in Malmö or Kristianstad on 31 December, 2010, 2015, or 2019. Furthermore, we collected information on cohabitation status and type of housing for the 31 December, the year prior to inclusion (i.e., 2009, 2014, and 2018, respectively). The social services register, maintained by the Swedish National Board of Health and Welfare, comprises data on all social services, including ILTC, provided according to the Social Services Act [2]. From this register, we obtained data on social services for each study year (i.e., 2010, 2015, and 2019). Technical aids supplied by the municipality are registered in the Sesam register using ISO (International Organization for Standardization) codes, together with the date when the person received the aid and, when relevant, the date when the aid was returned. For this study, we retrieved information on all mobility aids (ISO codes 120303, 120306, 120309, 120312, 120316, 120603, 120606, 120609, 120612, 121806, 122203, 122218, 122306, and 122409) provided to all people in Malmö and Kristianstad between 1 January 2010 and 31 December 2019.

Sociodemographic Factors
Social services, ILTC, and mobility devices were assessed for four sociodemographic factors: 1.
Sex (woman or man).

2.
Place of birth (born in Sweden or abroad).

3.
Cohabitant status (living alone or cohabiting) on the 31st of December the year before the study year. This term focuses on the living condition and not on the civil status of an individual.

4.
Type of housing (living in an apartment or single-family house (house)) on the 31st of December the year before the study year.

Social Services, ILTC, and Mobility Devices
Use of social services and mobility devices were aggregated into a four-level-outcome, based on the level of assistance provided at each level: 1.
ILTC: People who were registered as living in or moving to an institutional long-term care home during the year. 2.
Major assistance: People who had home care service, in the form of personal care, respite service, or companion service; hence, individual's dependent on personal assistance to be able to continue to live in their own housing. 3.
Minor assistance: People registered as having a safety alarm, meals-on-wheels, or using mobility devices, such as wheelchairs, walkers, walking frames, crutches, or canes; thus, those who only need "non-personal" aid to live an independent life in their own housing.

4.
No assistance: People with no registered use of social service or mobility devices.
People were classified based on the highest level of assistance needed. Thus, individuals who were granted accommodation in an ILTC home during a study year belonged to this group, and this group only, although they might also be registered for minor assistance in terms of e.g., mobility aid or safety alarm.
In 2010, only yearly data (i.e., one data point for the whole year) were available, and thus only one registered mobility device or service record during the year was required to be included in any of the four groups. In 2015 and 2019, monthly data were available. For these years, to be considered as having major assistance, at least three months of the relevant social services during the year, or the relevant social services during both November and December of each respective year, were required. This restriction was made to avoid including people that might only have had a temporary need of major services, e.g., post-surgery or after being hospitalized.

Study Populations
Six cohorts were established, comprising people living in Malmö or Kristianstad, aged 65 years or more, and alive 31 December, 2010, 2015, and 2019.
As we were unable to determine the cohabitation status and type of housing for people who moved from other municipalities to Malmö or Kristianstad in 2010, 2015, and 2019, or who had missing data on residence on those dates, these were excluded (n = 604, n = 790, and n = 782 for 2010, 2015, and 2019, respectively). We further excluded people with a need for supported living during any time during the study period (n = 95 in 2010, n = 128 in 2015, and n = 123 in 2019), or in housing (taxation unit) other than a house or apartment, or with missing data on type of housing (n = 1101 in 2010, n = 1389 in 2015, and n = 1432 in 2019). Using information from the RTB, we categorized people as cohabiting if their family type was married/partnership (with or without children) or cohabiting (with or without children). People listed as living in a single parent or single household were categorized as living alone. However, we excluded those with a role in the family listed as "child", as the living situation and need for these might go beyond the scope of this study (n = 26 in 2010, n = 40 in 2015, and n = 53 in 2019).
Thus, the final cohorts, i.e., those used in the analyses, for Malmö comprised 45,195, 47,892, and 52,552 people for 2010, 2015, and 2019, respectively. The corresponding numbers for Kristianstad were 15,237, 16,971, and 18,046. The age distribution for the sociodemographic groups in the six cohorts were similar across years and municipalities (Table 1).
There were distinct differences in sociodemographic factors between the two municipalities cohorts (Table 2). This was especially apparent when it came to inhabitants born abroad, where the number of immigrants in Malmö was 22-28%, while the corresponding numbers for Kristianstad were 7-9% (Table 2). In addition, there was a clear difference in cohabitant status, where the majority of people in Malmö lived alone (≈54%), while in Kristianstad, most people were cohabiting (≈57%). The same pattern was found for housing situation, where a noticeable number in Malmö lived in apartments (≈74%), while most in Kristianstad lived in houses (≈67%).

Statistics
To address the study objectives, inferential statistics was applied. The associations between sociodemographic factors and level of support (none, minor, major, and ILTC) were assessed using the relative risk (RR) of receiving such support for men vs. women, born in Sweden vs. born abroad, living alone vs. cohabiting, and living in apartment vs. living in single-family house. RRs with 95% confidence intervals (CIs) were estimated using generalized linear models, with Poisson distribution and identity link. Separate analyses were performed for minor assistance vs. no assistance, major assistance vs. no assistance, and ILTC vs. no assistance. For each sociodemographic factor, RRs were estimated in models including only the factor itself (crude), models including the factor itself and year of birth as a continuous variable (adjusted), and in models including all sociodemographic factors and year of birth (multivariate). For each pair of sociodemographic factors, the interaction was evaluated in models including both factors and their cross-term, as well as year of birth.

Results
The results are presented below by sociodemographic factor.

Level of Assistance Provided According to Sex
Women were consistently more likely than men to receive minor or major assistance and ILTC, although statistical significance was not always reached (Table 3). In Kristianstad in 2019, the increase in major assistance for women was higher among those living alone than among those cohabiting; and in Malmö in 2010, the increase in ILTC was higher among those born in Sweden (Table A1). No other interaction was found between sex and the other sociodemographic factors. Table 3. Number of men and women, stratified by year and municipality, in the different outcome categories (n (%)), and relative risks with 95% confidence intervals for comparisons of women vs. men, adjusted for year of birth and the other sociodemographic variables (place of birth, cohabitant status, type of housing). Bold indicates statistically significant differences.

Level of Assistance Provided According to Place of Birth
People born in Sweden were consistently more likely than those born abroad to receive major assistance. There were indications of the same pattern for minor assistance and ILTC (Table 4). Stratified analyses revealed that an increased likelihood of major assistance and ILTC was primarily found among those living alone (Table A2). No other consistent interactions were found between place of birth and the other sociodemographic variables. Table 4. Number of people born in Sweden and abroad, stratified by year and municipality, in the different outcome categories (n (%)), and relative risks with 95% confidence intervals for comparisons of being born in Sweden vs. being born abroad, adjusted for year of birth and the other sociodemographic variables (sex, cohabitant status, type of housing). Bold indicates statistically significant differences.

Level of Assistance Provided According to Cohabitant Status
Living alone was consistently a risk factor for all three outcomes (minor and major assistance, and ILTC) across years and municipalities (Table 5). Stratified analyses showed that the increase in major assistance and ILTC associated with living alone was higher among those living in apartments than those living in houses in 2010 and 2015 (Table A3). Moreover, in Malmö, those being born in Sweden and living alone were more likely to be provided with major assistance and ILTC than those being born abroad and living alone. No other consistent interactions were found between cohabitation and the other sociodemographic variables.

Level of Assistance Provided According to Type of Housing
Those living in an apartment rather than a single-family house were consistently more likely to have all three outcomes (minor and major assistance, and ILTC) across years and municipalities, although this was not always statically significant (Table 6). Stratified analyses revealed that the increase in major assistance and ILTC associated with living in an apartment was higher among those living alone than those cohabiting (Table A4). Table 6. Number of people living in a house or apartment, stratified by year and municipality, in the different outcome categories (n (%)), and relative risks with 95% confidence intervals for comparisons of living in an apartment vs. living in a house, adjusted for year of birth and the other sociodemographic variables (sex, place of birth, cohabitant status). Bold indicates statistically significant differences.

Discussion
Being a woman and living alone were associated with a higher use of both minor and major assistance and living in an apartment (as opposed to a house) was associated with a higher use of major assistance. Although being born in Sweden did not affect the use of social services in itself, living alone carried a higher risk among those born in Sweden than among those born abroad.
The focus of this study was older people in different sociodemographic groups and their use of social services. In the interpretation of the results, it is important to remember that we have measured use of social services, not need of them. Thus, discrepancies between groups (e.g., those between men and women) are not necessarily due to actual differences in need, but may also come from people with need finding assistance elsewhere (e.g., from an informal caregiver, such as a relative, partner or friend), health-seeking behavior or experiencing barriers to applying for social services (e.g., due to lack of knowledge about the possibility of obtaining support or how to apply for it).
In the present study, women were more likely than men to receive both minor and major assistance. This is in line with a previous study by Brändström et al. [10], in which Swedish women received twice as much home care service as men. One possible explanation could be differences in health-seeking behavior between men and women, as women tend to seek more health care in response to both physical and mental health concerns [12]. Women in Sweden, as well as in the rest of the EU, tend to live longer than men [1,16]. One consequence of this is that more women live alone at an advanced age (75% at the end of their lives compared to 30% among men [16]). However, in our study the differences between men and women in use of social services remained, even after adjusting for age, as well as place of birth, cohabitant status. This indicates that there may be other aspects involved. This is also supported by Mackian [11], who states that health-seeking behavior is not an isolated event but part of a major life puzzle involving an individual's social, cultural, and socioeconomic situation.
Among those living alone, people born in Sweden were more likely to receive major assistance than those born abroad. This may be due to differences in health care seeking behavior and reliance on family bonds for support and help among people born in Sweden and those born abroad [13]. This possible explanation is supported by a study that estimated that, compared to Spain [17], older persons in Sweden provided almost double the amount of caregiving hours. These differences are mostly due to differences in demography and household structures, as the large majority of older people in Sweden tend to live with one partner only or alone, whereas a more complex family structure is predominant in Spain [17]. Another possible explanation for the discrepancy between those born in Sweden and those born abroad is that services for older people are not as accessible to older migrants, because of an inadequate provision of information and a lack of policy integration [14]. Even so, a study by Songur [18] revealed that in Sweden, immigrants from the middle-East and Africa used as much social services as Swedes, but refrained from using ILTC by compensating for these increased needs with family or informal caregivers. The older immigrants' use of social services was also enabled by younger family members taking care of applications and communication between the care recipient and the social administration [18]. That this growing group of older migrants might be discriminated against in obtaining formal care and must rely upon informal caregivers is of considerable societal concern.
In the present study, people living in apartments were more likely than those living in single-family houses to receive major assistance. One explanation could be that people in need of increased help and support might have difficulties taking care of a house and garden and thus might have already moved into an apartment at an earlier stage [19]. This hypothesis is strengthened by a previous study, which showed that among older Swedish house owners it was more common to change from a house to an apartment with an increase in age [19]. This study also revealed that a change of marital status, such as becoming a widow/widower or divorced, increased the probability of moving. Such changes in life situation would also mean that an informal caregiver, in terms of a spouse, would disappear.
Our findings support this by showing an interaction between living alone and living in an apartment, in terms of a higher relative risk for major assistance for those living alone in an apartment than those living alone in a single-family house (and, correspondingly, for those living alone in an apartment and those cohabiting in an apartment).
The results from the present study indicate that older people receiving social services are primarily those living alone, which is in line with other studies [6,16]. This implies that for those living together with a partner or other cohabitant, the partner/cohabitant is the one who provides, e.g., personal help and other household chores such as washing and cleaning [16]. These informal caregivers are most probably older partners fulfilling the needs that otherwise would have been provided for by formal care. This is supported by a study by Jegermalm and Torgé [20], in which the cohabitant family carer was most likely married and 50% provided care for a spouse (often a person older than 75 years). This is worth careful consideration, as informal caregivers to older people, in case of a partner or cohabitant, will most likely be of an older age themselves and might thus have limited health and physical strength. However, even when the informal caregiver is someone other than a partner, e.g., a child, their age, health, and physical strength may be an issue. Since an older population increases the age of the parents, many adult children have to take responsibility for their parents, even when they themselves are of advanced age and might need support and health care. This strain, also in the sense of time and possible economic cutbacks, that the informal caregiver burden might imply upon individuals should not be neglected. A study by Berglund et al. [5] revealed that informal caregivers were more likely than non-caregivers to report non-adherence to treatment, and to report refraining from seeking medical care for their own care needs.
In most cases, when there is a close non-cohabiting relative acting as an informal caregiver, this is often a woman [16,21]. As women in the past decades have begun to work full-time outside the home, their roles as informal caregivers is of concern, as they often, still, have the major responsibility for the home and care of children. Thus, the addition of being an informal caregiver puts a major strain on them. These high informal workloads often result in a trade-off, where the women need to choose between cutting back on workhours or accessing supplemental help. A trade-off that is not so much based on the care recipient's (i.e., the older relative's) well-being, as that of the caregiver's [9]. A study by Stanfors et al. [21] on caregiving times and trade-offs showed that in Sweden, in contrast to the United Kingdom and Canada, caregivers did not trade off time in paid work for caregiving, but instead they had less leisure time. They argue that trade-offs with leisure should not be ignored but demanded further investigation because they likely have health implications, especially for women.
Despite the notable difference between the two municipalities in cohabitant status and types of housing, the patterns of obtained social service for different sociodemographic groups were the same in the two municipalities, where women and individuals living alone seem to receive more assistance. These findings also remained consistent across all three study years. This consistency strengthens our findings and stresses the fact that there are sociodemographic differences in the use of social services and ILTC in the older Swedish population.
A strength of the present study is that municipalities are required to report all care and services provided according to the Swedish Social Services Act to the National Board of Health and Welfare. Thus, the coverage of data is high. However, the level of detail and classification of services and their delivery differed between 2010 and the following years (2015 and 2019). However, as our major findings remained stable throughout all years and both municipalities, we believe this would have had negligible effects on our results.
Another study limitation is that we were unable to identify couples in this study. Although this does not affect the results in the present study, it prevents validation and quantification of the extent and burden the role of informal caregiver might impose on spouses.
There seems to be a high risk of discrimination against certain groups of older people potentially in need of social services, as well as against informal caregivers. Women and immigrants were mentioned earlier, but previous studies have also shown that individuals with higher income and education are more likely to pay for extra hours of services, both for themselves and for their relatives [10]. These groups are also more likely to be able to communicate with the authorities and thus increase the chances of claiming their legal rights to formal care [6,10,16,22]. The use and need of informal caregivers in Swedish society to obtain support and care is problematic. A study by Alwin et al. [21] concluded that the care provided by informal caregivers for frail older people is substantial and represents a considerable economic value. Although their calculations were associated with some uncertainty, the population size indicated that supporting informal caregivers should be a prime concern for society [21].
From our results, it is obvious that discrepancies either in need of or provision of social services for different socio-economic groups exist. Furthermore, these discrepancies are independent of municipality differences in sociodemographic structure and organization of social services. Moreover, they remained stable during the investigated decade. As the study was performed in a Swedish setting, the results are not necessarily generalizable to other countries, with other forms of care of older people. However, it could be argued that if discrepancies regarding provision of such care are found in Sweden, where the welfare system is designed to provide equal support for all, regardless of socio-economic ability, in countries with policies that rely more heavily on the individual's ability to provide for themselves and their families, sociodemographically vulnerable groups may face major difficulties during ageing.
There is a need to avoid discriminating against older people, in terms of possibilities to live a healthy and independent life or becoming a major burden for potential informal caregivers (in terms of an increased workload and thus cutbacks in both time, health, and economic factors). Our study highlights the need for further research on the need and provision of social services, and efforts of informal caregivers, within different sociodemographic groups.

Conclusions
Older people's use of social services seems to be subject to their sociodemographic characteristics. It is therefore likely that there are varying discrepancies between individual needs and the provision of social services (demanding personal assistance) among sociodemographic groups in the older population in Sweden. When society does not meet the needs of the older individual, this might be compensated for by informal caregivers. There is a major risk that supporting older relatives and partners/cohabitants poses an extra burden, and thus affects certain groups negatively. Older partners or cohabitants, children to individuals of advanced age, women, and people of another ethnicity might be overrepresented as informal caregivers and thus could have a substantial extra workload, limiting or decreasing their leisure time, physical and psychological health, and income.
Future research should focus on targeted studies of how needs and services are met, or compensated for, among older individuals in different socioeconomic groups and their cohabitants, relatives, and other possible informal caregivers. It is particularly important to assess these differences in countries with varying policies for the care of older people.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Swedish Ethical Review Authority (protocol code 2020-01958).

Informed Consent Statement:
In line with requirements from the Swedish Ethical Review Authority, no informed consent was obtained from the participants. However, information about the study, including who to contact to have personal data removed from the data base, was published online.

Data Availability Statement:
Since the data comprise information about single individuals and are detailed enough to enable identification of, at least, some of the people included, they cannot be made publicly available. However, as the database was compiled with national register data, other researchers may contact the register holders (the Swedish National Board of Health and Welfare, Statistics Sweden, and Malmö and Kristianstad municipalities) to obtain access to the registries used in this study, and thereby generate an identical database.
Acknowledgments: Carolina Hansen, Development secretary in Social Work at Social administration, Malmö municipality for clarifying the municipalities registers and routines for granting different forms of social support.

Conflicts of Interest:
The authors declare no conflict of interest.