Relationship between social cohesion and the care burden of primary family caregivers in central Tokyo, Japan

Abstract Objectives To clarify how the care burden of primary family caregivers is associated with social cohesion in an urban area of Tokyo, Japan. Study design Cross‐sectional study. Methods A questionnaire survey of primary family caregivers was conducted in Tokyo in 2015. Social cohesion was examined using the social capital indicators of Kondo et al, and the care burden of primary family caregivers was assessed by the Zarit Care Burden Interview Scale in Japanese short version (J‐ZBI_8). Data were analyzed by multiple regression models. Ethics approval was obtained to carry out this research. Results Seventy‐nine caregivers responded to the questionnaires. After excluding 6 caregivers who did not respond appropriately to the questionnaire, 73 caregivers were included in the analyses. The average age of the primary family caregivers was 68.9 ± 12.7 years old, and that of the patients receiving care was 83.1 ± 10.0 years old. “Receipt of instrumental support” was significantly associated with reduced burden of care for family caregivers as assessed by the J‐ZBI_8 score (P = .027). Conclusion This study suggested that social cohesion was significantly associated with reduced burden of care for primary family caregivers. Especially, the results suggested that “social support: receipt of instrumental support” was associated with a lower burden of care after adjustment for confounding factors. It is important to understand family structure and social community differences such as informal social support for future policy making.


| INTRODUCTION
Serious concerns have been raised about family caregiving of older adults in many countries. There is not only a growing demand for families to provide elderly care but also growing evidence that caregiving itself poses risks related to mental, physical, and economic health for some people. In line with the increasing elderly population in Japan, 1 the Ministry of Health, Labour and Welfare aims to maintain the dignity of the elderly and support independent living so that they can live their own lives as much as possible in the community. 2 It is essential to provide comprehensive and continuous home care in cooperation with local medical and nursing care agencies. However, as life expectancy lengthens and age and family structure change, concerns are emerging about both the lack and burden of elderly care. 3 Zarit et al 4 define care burden as "the degree of damage the caregiver suffers in terms of emotional, physical health, social life and financial condition as a result of caring for a relative." Long-term care insurance systems provided in Japan to support elderly care currently classify patients based on their activities of daily life or level of dementia. At present, the type of family structure in which a patient lives is not considered a criterion for registration.
Family composition is greatly influenced by the promotion of home care. For elderly people who are admitted to hospital, most factors making it difficult to discharge them from hospital are related to nursing care issues. 5,6 The availability of care tends to be affected by the situation of the caregivers. Regional long-term care systems never work without the support of the local community.
Some studies show that support from the community depends on its social cohesion. Nevertheless, there is little evidence on whether social cohesion reduces the burden of caregivers who provide home nursing care. Social support is believed to promote health by reducing mental stress. 7 Chan et al 8 define social cohesion as a state of affairs concerning both the vertical and the horizontal interactions among members of society as characterized by a set of attitudes and norms that include trust, a sense of belonging, and the willingness to participate and help, as well as their behavioral manifestations. Cramm et al 9 found that neighborhood social capital and social cohesion are accessed via individual, group, or community membership, and these resources are norms of reciprocity, civic participation, trust in others, and benefits of membership. Putnam 10 stated that social capital refers to physical objects, and human capital refers to the properties of connections among individuals, social networks, and the norms of reciprocity and trustworthiness that arise from them. Social support is broadly defined as both emotional and instrumental support. Noguchi 11 said that emotional support is support that works on emotions, such as "listening to worries and worries" and "enhancing," and instrumental support is "nursing and care" and "lending money." Social support is a means of substantive support.
In a study in the United States on the association of social capital with depression, the risk of depression was reduced by 57% in a group of local residents with high confidence compared to the group with low confidence. 12 In a survey conducted in Japan that examined regional differences in the burden of care for the family caring for their elderly showed that there was better nursing care and a higher care burden in urban areas than in sparsely populated areas. 13 However, after correcting for the factors related to care burden, the tendency for higher care burden in urban areas was no longer remarkable. The relationship between social capital and depression has been well discussed, 7,12-14 and numerous papers have reviewed the care burden of primary caregivers. 15 There is also research on regional differences in care burden. 16 However, the relationship between the care burden of home caregivers and regional social cohesion has been little explored. The present study aimed to clarify the relationship between social cohesion and family care burden to support the development of an enhanced caregiver support system by use of a questionnaire survey. We also discuss the difference between two areas, namely the center of Tokyo (this study) and a rural island (a previous study conducted using the same questionnaire) 17 in Japan.

| Subjects
The survey was conducted as a placement method survey using a self-administered questionnaire. The study subjects were primary family caregivers who lived with patients over the age of 65 who contracted with visiting nurse services in Tokyo. The home-visit nursing system in Japan is based on medical insurance and long-term care insurance. Funding is provided through health insurance premiums paid by the insured, contributions from working-age individuals, and public sources. When medical treatment is received, the co-payment is 20% (30% for individuals with incomes over a certain level). When visiting nursing services are used, the co-payment rate is 10% (or 20% or 30%). In 2015, 796 home-visiting nurse offices were operating in Tokyo. With cooperation from five visiting nursing offices, each located in an adjacent ward in Tokyo, a questionnaire survey was conducted during home visits to service recipients. The average number of patients requiring long-term care during the survey at the five visiting nursing establishments was 261 patients for 3 months. As this study was intended for the primary caregivers, elderly long-term care insurance recipients living alone were excluded from the survey. The 261 patients were the average number of patients over the implementation period of 3 months and include only patients classified as longterm care level 1 to 5. Hospitalized patients were excluded. Those who did not sign the consent form and those who did not respond to the Japanese short version of the Zarit Care Burden Interview Scale (J-ZBI_8) were also excluded from the analysis. The questionnaire was distributed to the family caregiver by the home-visiting nurse, and the family caregiver filled out the questionnaire, which was then collected by the home-visiting nurses. The survey was conducted from October 9 to December 9, 2015.

| Questionnaires
The questionnaire contents are explained in detail in a previous study conducted in an island area of Japan. 17 The questionnaire assessed the following categories.  f. Caregiver's social network: using the frequency of meeting with friends and acquaintances as a question item.
Questions e and f asked for frequency, but we classified them into two (Yes/No) categories. The J-ZBI_8 19 was also used. The rating scale of the J-ZBI_8 (with scores ranging from 0 to 32) is an objective measure that can be used as an international comparison to capture the care burden of caregivers. 4,20 The short version is widely used. 21,22 Confounding factors included caregiver's sex, age, and occupation, length of caregiving, educational background, length of residence time, household, sex of the patient under care, and patient's age.

| Statistical analyses
To confirm internal consistency, Cronbach's α 19 was obtained for the J-ZBI_8. Characteristics of the participants are summarized as the mean with SD or median (Q1, Q3) for continuous variables and number (%) for categorical variables. The score of J-ZBI_8 was used as an outcome value after log-transformation, and the effects of the exploratory variables were analyzed using simple and multiple regression models. Social capital variables of support variables a, b, c, and d and participation in organized activities (e) and social networks (f) were used along with the other confounding factors.
Arai and Zarit 23 used receiver operating characteristic analysis of the J-ZBI_8 to determine a reference cutoff value as an indicator of depressive symptoms. The total score of the J-ZBI_8 was classified into a dichotomous variable of ≥13 vs ≤12, and a logistic regression model was used with the odds ratio and associated 95% confidence interval (CI) calculated.
In the multiple regression analyses, a stepwise method (with inclusion and exclusion criteria selected at the 20% significance level) was used as a method to select variables. A two-sided significance level of 5% was considered to indicate statistical significance. All statistical analyses were performed using SAS Ver. 9.4.

| Ethical considerations
This study was carried out in accordance with the Declaration of Helsinki, a set of ethical principles regarding human experimentation for the medical community. To protect personal information, all information remained anonymous. Informed consent of each subject to participate in this study was obtained by their signing of the questionnaire document. This study was aimed at the primary caregiver who is a member of the patient's family, and the only patient information collected was sex and age. Only the primary caregivers signed the questionnaire. Ethics approval was granted by the ethics committee of Teikyo University (approval no.  and that of the Tokyo-Hokuto Health Co-operative Association (approval no. 76).

| RESULTS
Among the primary caregivers surveyed, 79 responded (the number of patients serviced by the five home-visit nursing stations, including the elderly living alone, was 261. Responses were obtained from 79 primary caregivers of these patients, giving a response rate of 30.3%). After we excluded six caregivers who did not respond to the questionnaire, 73 primary family caregivers were included in the analysis. Because this survey was intended for the primary caregivers, the number of patients included those living alone. Therefore, the exact response rate was expected to be higher than 30%. The average age of the primary family caregivers was 68.9 ± 12.7 years old, and they included 23 men (31.5%) and 50 women (68.5%). The average age of the patients receiving care was 83.1 ± 10.0 years old, and they included 31 men (43.7%) and 40 women (56.3%) ( Table 1). Cronbach's α for the J-ZBI_8 was .89.
The results of the single and multiple regression analyses are shown in Table 2. After adjustment for confounding factors, the results of the regression model with stepwise variable selection showed a significant difference for "social support: receipt of instrumental support" (P = .027). The results of the logistic regression analysis are shown in Table 3. Similar to the results of the multiple regression analysis, the odds ratio of having or not having "social support: receipt of instrumental support" was 0.24 (95% CI = 0.07-0.81, P = .024). The results of both analyses indicated that the receipt of instrumental support was associated with less care burden after adjustment for confounding factors.

| DISCUSSION
The relationship between a caregiver's social cohesion and primary family care burden was examined via a placement method survey in the Tokyo metropolitan area. The results suggested that "social support: receipt of instrumental support" was associated with less care burden after adjustment for confounding factors.
Among the items in this survey referring to social support, "instrumental support provided care for the resident given by others when the caregiver cannot provide care" tended to increase the care burden of the home caregiver. The primary caregiver may have anxiety about his/her health condition, and when it deteriorates, there may be no one the caregiver can rely on, and thus their burden of care may potentially increase. Similar results were observed in the island survey, that is, there was a significant difference in the results related to "social support." 17 In addition, there were significant differences in social support related to "reception of emotional support" and "participation in organized activities" in the island survey (Appendix Table   A1, A2).  The primary family caregivers who live on islands tended to seek emotional support and participation in organizational activities (social and positive self-role). In contrast, among those living in Tokyo, the burden of care for the primary caregivers was not associated with "emotional support" or "participation in organizational activities," which indicates the frequency of participation in the activities of a group organization, but rather was associated with "having receipt of instrumental support". Surveys in urban and rural areas in South  27 The number of single households and nuclear family households tends to increase as people locate to the city centers in Japan.
In 2015, the percentages of employees who changed their jobs due to their requirement to provide home care were 0.6% for men and 1.1% for women. 28  Not included in the logistic regression analysis. A stepwise method was used for variable selection in the multiple logistic regression analysis with the inclusion and exclusion criteria of 20%. Variables not selected in the stepwise method are indicated by "-".
hours. It may also be difficult for caregivers to keep working, even if they have reduced their working hours, and there may be situations in which work style must be changed to provide care.
In the long-term care insurance system, the degree of care required is determined without taking either family background or family structure into consideration, so it is essential to consider the care burden of the primary caregiver. It has been suggested that caregivers receive partial support, especially in areas with weak social cohesion. For primary caregivers in central Tokyo, "participation in organizational activities" did not affect the level of care burden, possibly because no family member other than the caregiver is available to provide elderly care and also to go to work to earn money. According to the results of this survey, the less social support available as "instrumental support", the higher is the burden of long-term care. There is no one to care for the patient other than the main caregiver, who must take care of the patient every day and thinks that he/she must provide this care alone. Considering the Japanese family structure, it can be said that there are many hurdles to promoting home care.
Home caregivers and primary family caregivers who are unable to "participate in organizational activities" are required to further strengthen their social ties when seeking support from local community members and social support from medical social workers when their care recipient returns from the hospital to the community. In medical institutions, medical social workers and nurses mainly support the patient's discharge. However, as the length of hospital stay is being reduced, the problems faced by patients are becoming more complex, and increasing social support is becoming a challenge. 31 Recently, United Kingdom introduced Social Prescribing, which allows third parties (called link workers) to connect to local resources.
One report found that Social Prescribing saved £647 000 (about 100 million Japanese yen). 32 Kumakawa et al 33 pointed out that various organizational activities and regional resource information of organizations are not systematically organized. In addition, they suggested that informal regional resources need to be identified and developed.
Informal social resources will become increasingly necessary not only for caregivers but also for the patients and local residents.

| Limitations
The present study showed the importance of community cohesion in possibly reducing the burden on home caregivers. Because the number of variables was large and the number of survey subjects was small, it was difficult to use certain items as adjustment factors, and thus the degree of care required could not be adjusted as a variable in the analysis. Furthermore, it was not possible to identify all of the elderly living alone. In addition, some patients receiving care were admitted to or discharged from the hospital during the survey period, and thus an exact response rate could not be calculated.
A recent study found that social support for caregivers is associated with physical effects and psychological stress. 38 This suggests the possibility that, to reduce care burden, interventions should focus more on "feeling connected" rather than on "building connections."

| CONCLUSION
Despite the above limitations, this study suggested that social cohesion was significantly associated with reduced burden of care for primary family caregivers who live in Tokyo. Especially, the results suggested that "social support: receipt of instrumental support" was associated with less care burden after adjustment for confounding factors. However, the type of support differed somewhat from that received by caregivers living in the island community we surveyed before. It is important to understand family structure and social community differences such as informal social support for future policy making.

ACKNOWLEDGMENTS
We thank the primary caregivers and the staff at the visiting nursing facility corporation. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CONFLICT OF INTEREST
The authors declare no conflict of interest. Yuki Naganuma had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.

TRANSPARENCY STATEMENT
The corresponding author (Y.N.) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; that any discrepancies from the study as planned have been explained.

DATA AVAILABILITY STATEMENT
The datasets generated and/or analyzed during the current study are not publicly available due to ethical restrictions. Not included in the logistic regression analysis. A stepwise method was used for variable selection in the multiple logistic regression analysis with the inclusion and exclusion criteria of 20%. A variable that was not selected in the stepwise method is indicated by "-".