Association between living arrangement and quality of life for older adults in the community*

Objective: to compare the sociodemographic and economic characteristics of the older adults in the community according to the living arrangement and to verify the association between the type of living arrangement and the quality of life scores. Method: a cross-sectional epidemiological study conducted with 796 older adults in the community. To assess quality of life (dependent variable), network and social support (adjustment variable), validated and applied chi-square tests, descriptive statistical analysis, multiple comparison analysis (ANOVA) and multiple linear regression model (p<0.05) were used. Results: the older adults who lived only with their spouses had better quality of life scores in all domains and facets, except in the death and dying domain, which did not show any significant difference. The lowest scores for quality of life were identified in the groups with the presence of children and, exceptionally, in the domain of social relationships and, in the facets death and dying and intimacy, those who lived alone had worse assessments. In the adjusted model, there was an association between the type of living arrangement and the different domains and facets of quality of life. Conclusion: living arrangement was associated with quality of life scores for older adults in the community, even after adjusting for the gender, age, number of morbidities, and social support variables.

the individual's perception of their position in life, in the context of the culture and value system in which they live and in relation to their goals, expectations, standards and concerns" (11) .
In the gerontological literature, a number of studies have found that the older adults who live with adult children (5,12) , live alone (5,7) or with other family members (with no presence of the spouse) (7,13) have showed harms in QoL. However, these studies did not adjust for social support, which will be analyzed in this study.
Regardless of the type of living arrangement, social support promotes significant engagement and emotional support, as well as it avoids social isolation (14) , therefore, being able to play a mediating role between the living arrangement and QoL. Results of a research study conducted with Chinese older adults living alone suggest that social support can mitigate the negative effects on QoL (9) . In another study also carried out in China, urban older adults who lived alone were the most vulnerable to the worst health-related quality of life scores, and social interaction reduced this negative effect (14) . However, so far, no research study has been identified to verify the association between living arrangement and the QoL domains and facets of urban older adults in the community through the application of a specific instrument for this age group and adjustments for the potential confounding variables, such as social support.
Given the above, some questions emerge: Which type of living arrangement contributes the most to the QoL of urban older adults in the community? Does the relationship between the living arrangement and the QoL scores of the older adults remain the same after controlling for social support? It should be mentioned that the configuration of living arrangement in Brazil does not depend exclusively on the will of the older adults or of their families, but involves a variety of historical, sociocultural, political, economic and demographic aspects, which can favor or harm QoL (15) . Therefore, it is believed that investigating the relationship between different types of contemporary living arrangements and the QoL of the older adult population in a Brazilian municipality will contribute to elucidating the potential and specific challenges of each Bolina AF, Araújo MC, Hass VJ, Tavares DMS.
housing context, which, in turn, need be considered by researchers, educators, health and social care professionals and, above all, by public policy-makers.
As a member of the family health team, the nurse has a primary role in caring for the older adults in the community, with the housing context and the social support network aspects that should be analyzed during the nursing consultation.
This study has the following objectives: to compare the sociodemographic and economic characteristics of the community's older adults according to living arrangement; and to verify the association between the type of living arrangement and the quality of life scores.

Method
A quantitative approach study, of the home survey type, observational, analytical and cross-sectional, developed in the urban area of the city of Uberaba, located in the state of Minas Gerais.
Older adults (60 years old or more) living in the urban area of that municipality took part in the study.
The sample was defined based on procedures for multistage cluster sampling. The sample size calculation considered a determination coefficient in a multiple linear regression model with five predictors, with a level of significance or type I error of and type II error of β = 0,1 , therefore resulting in an a priori statistical power of 90%. The values described above were inserted in the PASS (Power Analysis and Sample Size) application, version 13, obtaining n=813.
The inclusion criteria were the following: being 60 years old or older and living in the urban area of the municipality of Uberaba-MG. Institutionalized older adults were excluded; as well as those with communication problems (deafness not corrected by hearing aids and severe speech disorders), those who were not found after three attempts by the interviewer, and those who presented cognitive decline associated with a final score ≥ 6 points in the PFEFFER Functional Activities Questionnaire. Based on such criteria, 17 older adults were excluded: 12 for having cognitive decline and a score > 6 points in the questionnaire, and five that did not have complete data in relation to the outcome variables. The final sample of the study comprised 796 older adults.
Data collection was carried out at the older adults' homes from May 2017 to June 2018, by means of direct interviews using the instruments described below.
Initially, the older adult's cognitive assessment was carried out through the application of the Mini Mental State Examination (MMSE), translated and validated in Brazil (16) and, in a positive screening for cognitive decline, the PFEFFER Functional Activities Questionnaire (17) was applied to the older adults. This scale has 11 questions with a maximum score of 33 points and allows verifying the presence and severity of cognitive decline through the assessment of functionality and the need for assistance from other people. In the present study, for situations in which the PFEFFER result was less than six points, it was defined that the interview would be  (19) . The scale consists of five dimensions (material support, affective support, emotional support, information, and social interaction), assessed by means of questions in which the older adults indicate how often they consider each type of support available, in case of need: 1 (never); 2 (rarely); 3 (sometimes); 4 (almost always), and 5 (always). Based on the answers, the scale generates a score, and the higher the score, the better the social support (19) .
To assess the living arrangement (preferred predictor variable), the older adult was asked "Do you live at your home?" and offered the following answer autonomy; past, present and future activities; social participation; death and dying; and intimacy) (21) . It is noteworthy that the instruments are complementary for assessing QoL, as the WHOQoL-Bref measures this variable in a generic way and the WHOQoL-Old is specific for older adults. The scores were measured using syntax and vary from 0 to 100, where the highest value represents the best QoL.
The collected data were processed in spreadsheets using the Excel ® program, in two databases for consistency assessment. Then, they were exported to

Results
The final sample of the study comprised 796 older adults, divided into six groups according to the type of As shown in Table 1 (Table 3).

Discussion
In this study, the highest percentage of older and couple with children (18.9%) (15) .
Regarding gender, other studies have also verified a predominance of older adult women living alone (3,6,15) , which can be explained by the higher life expectancy of women in Brazil (22) . In addition, men in a situation of widowhood or separation tend to remarry (15) ; converging with the result of this study, where the highest percentage of older adults living with their spouse and others pertained to the male gender.
Regarding the age group, there was a higher percentage of older adults who lived alone aged between 70 and 80 years old. National data corroborate the findings of the present study by showing a prevalence of older individuals (75 years old or more) living in single-person homes (6) . A research study carried out in Canada also identified a higher proportion of older adult people living alone compared to the others, in both genders (4) . Over time, older adults become more susceptible to living alone due to widowhood. In Brazil, the estimated life expectancy of the population in 2017 was 76 years old (23) .
Although the number of surviving generations has increased due to the increase in life expectancy, there has been a weakening of the ability of families to provide care for the older adults due to several factors: reduction in the fertility rates, greater insertion of women in the labor market, and presence of ageism in modern society (1) .
It should be mentioned that the increase in the older adults' single-person homes represents an indicator of a successful aging process, as they can experience this moment independently and autonomously (3) . However, over the years, the risk of functional impairment increases (14,24) , which can hinder access to the health services and the performance of daily tasks (3) .
With regard to schooling, a similar result was evidenced in a study carried out using data from the National Health Survey, in which the majority of the Brazilians older adults reported having elementary education or lower (76.7%) (25) . The schooling level eventually interferes in the health-disease process, in the search for health services, and in adherence to treatments (26) . Consequently, older adults who live alone and have a low level of schooling can be more vulnerable to worsening health conditions compared to those who live with a partner or who have more years of study (6) .
In this study, there was predominance of older adults who received a minimum wage in all types of living arrangements. As age advances, health expenditures can increase, especially for the purchase of medications (27) , with the risk of compromising the basic needs of the  older adults who receive this monthly amount. This result is of particular concern for the older adults who live alone, since a nationwide study with data from the Family Budget Survey identified that older adults in single-person homes commit 63.38% of their income to meet basic needs such as housing, health care and food at home (28) . According to the authors, the commitment of income affects the availability of resources for leisure activities and other essential expenses (28) , which can compromise quality of life for the older adults in this housing condition.

WHOQoL-Bref
With regard to the comparison of the QoL scores with living arrangements in the bivariate analysis, previous findings corroborate the results obtained in the present study, as they indicate that cohabitation with the spouse was positively related to the QoL of older adults (5,7) or because they show that living alone or with their children/others resulted in worse assessments (9,(12)(13)29) .
However, these surveys did not adjust for social support, which was analyzed in the current study.
It is known that the type of living arrangement can interfere in the social interactions of the older adults and in the provision of available resources in their daily lives (7) . A study carried out with older adults who lived alone in South Korea showed that social support was a determining variable for the QoL domains Partially ratifying these findings, in a study developed in Turkey, the older adults who lived with other family members and without a spouse also had lower QoL scores in the domains of autonomy and social participation (13) .

Conclusion
This study revealed that the type of living arrangement was associated with QoL scores for older adults in the community, even after adjusting for the gender, age, number of morbidities, and social support variables. In relation to the group that lived only with the spouse, older adults who lived alone had a lower QoL score in the intimacy facet, and those who lived only with their children, lower QoL scores in functioning of the senses, autonomy, and intimacy.
Also in the comparison with the group that lived only with the spouse, older adults who lived with the spouse and others had a worse assessment of QoL in autonomy and social participation; those who lived with children and others, a worse assessment of functioning of the senses, past, present and future activities, social participation, and intimacy; and, finally, in the group of other types of living arrangement, a worse assessment in the intimacy facet.
Therefore, the findings of this research converge to the understanding that each modality of family composition has its strengths and weaknesses that need to be considered by researchers, educators, health and social care professionals and, above all, by policy-makers.