Satisfaction with the neighborhood environment and health in older elderly : cross-sectional evidence from the Bambuí Cohort Study of Aging

In order to investigate the association between satisfaction with the neighborhood environment and self-rated health among older elderly, data from 814 participants of the eleventh wave of the Bambuí Cohort Study of Aging were analyzed using robust Poisson regression analyses. Those elderly with higher satisfaction with their neighborhoods (PR = 0.75; 95%CI: 0.63-0.87) were less likely to report worse self-rated health. The number of chronic diseases (two, PR = 1.69; 95%CI: 1.05-2.70, three or more, PR = 1.99, 95%CI: 1.273.13), difficulty in performing daily activities (PR = 1.51; 95%CI: 1.28-1.78), presence of depressive symptoms (PR = 1.68; 95%CI: 1.44-1.95) and frequency of leisure-time exercise in previous 90 days (less than once a week, PR =1.24; 95%CI: 1.03-1.50) were all positively and significantly associated with poor self-rated health. This study provided empirical evidence that satisfaction with the neighborhood environment was associated with the health of the older elderly. The findings further suggest the potential importance of including this indicator in analyses of place and health among the elderly. Aged; Urban Health; Health Status; Cohort Studies Introduction Self-rated health is one of the most widely used health indicators in health research 1. The measure, that reflects an individual’s global evaluation of their overall health, has shown robust validity as well as good test-retest reliability 2. Because of these properties, some authors argue that, for research purposes, self-rated health may serve as a reasonable substitute for multi-item measures of health status 3,4. For the aged population, previous studies have documented that self-rated health is a consistent predictor of functional disability and mortality rates even after adjustment for sociodemographic characteristics and other clinically relevant factors or after controlling for more objective measures of health and health behaviours 5,6,7. Moreover, self-rated health is also associated with increased rates of outpatient service utilizations and hospitalizations 5. Several worldwide studies have demonstrated that poor self-rated health among the elderly is associated with female gender 6,8; low income 9; presence of comorbidities 10, depression 10,11,12,13, disabilities 6,12; lack of physical exercise 12,14, inability to go out alone 11; access to health services, including health care coverage and high outpatient use 5,15. Less understood, however, is the neighborhood’s influence on the health of older adults 16. S390 ARTIGO ARTICLE SATISFACTION WITH THE NEIGHBORHOOD ENVIRONMENT AND HEALTH S391 Cad. Saúde Pública, Rio de Janeiro, 27 Sup 3:S390-S398, 2011 According to Stronegger & Titze 17, neighborhoods are the most important place to establish connections with other individuals, daily routine activities and consumption habits, therefore their physical and social environments affect the health and health behavior of residents. This can be particularly relevant for the elderly, given the combination of declines in physical and cognitive functioning that tends to accompany ageing, which leads to a greater dependence on the immediate residential neighborhood for their health and well-being 16,18. Despite the existence, in recent years, of considerable research examining how health outcomes vary according to where one lives 19,20,21, there are relatively few available publications about the impact that neighborhood environment might have on the health of the elderly. Most of this evidence has been collected in developed countries and among elderly living in large cities 16. Studies have revealed, for instance, associations between poor individual health and neighborhoods with low socioeconomic levels 22,23,24. Similar associations were observed between perceived resources and/or problems (e.g., traffic, trash or litter, safety/crime) and health 25,26,27. Since the most basic environmental unit in which people live and conduct their daily activities is the neighborhood, its effects on health can be accessed through measures based on the perceptions of the residents therein. The purpose of this investigation is to ascertain whether there is an association between satisfaction with the neighborhood environment and self rated health among older elderly living in a small Brazilian city, after adjusting for sociodemographic characteristics, health conditions and lifestyle risk factors.


Introduction
Self-rated health is one of the most widely used health indicators in health research 1 .The measure, that reflects an individual's global evaluation of their overall health, has shown robust validity as well as good test-retest reliability 2 .Because of these properties, some authors argue that, for research purposes, self-rated health may serve as a reasonable substitute for multi-item measures of health status 3,4 .
For the aged population, previous studies have documented that self-rated health is a consistent predictor of functional disability and mortality rates even after adjustment for sociodemographic characteristics and other clinically relevant factors or after controlling for more objective measures of health and health behaviours 5,6,7 .Moreover, self-rated health is also associated with increased rates of outpatient service utilizations and hospitalizations 5 .
Several worldwide studies have demonstrated that poor self-rated health among the elderly is associated with female gender 6,8 ; low income 9 ; presence of comorbidities 10 , depression 10,11,12,13 , disabilities 6,12 ; lack of physical exercise 12,14 , inability to go out alone 11 ; access to health services, including health care coverage and high outpatient use 5,15 .Less understood, however, is the neighborhood's influence on the health of older adults 16 .
According to Stronegger & Titze 17 , neighborhoods are the most important place to establish connections with other individuals, daily routine activities and consumption habits, therefore their physical and social environments affect the health and health behavior of residents.This can be particularly relevant for the elderly, given the combination of declines in physical and cognitive functioning that tends to accompany ageing, which leads to a greater dependence on the immediate residential neighborhood for their health and well-being 16,18 .
Despite the existence, in recent years, of considerable research examining how health outcomes vary according to where one lives 19,20,21 , there are relatively few available publications about the impact that neighborhood environment might have on the health of the elderly.Most of this evidence has been collected in developed countries and among elderly living in large cities 16 .Studies have revealed, for instance, associations between poor individual health and neighborhoods with low socioeconomic levels 22,23,24 .Similar associations were observed between perceived resources and/or problems (e.g., traffic, trash or litter, safety/crime) and health 25,26,27 .
Since the most basic environmental unit in which people live and conduct their daily activities is the neighborhood, its effects on health can be accessed through measures based on the perceptions of the residents therein.The purpose of this investigation is to ascertain whether there is an association between satisfaction with the neighborhood environment and self rated health among older elderly living in a small Brazilian city, after adjusting for sociodemographic characteristics, health conditions and lifestyle risk factors.

Methodology Study area and participants
The Bambuí Cohort Study of Aging is a population-based cohort study of older adults that has been conducted in Bambuí, a city of approximately 15,000 inhabitants located in the state of Minas Gerais in southeastern Brazil.The eligible population for the cohort consisted of all residents aged 60 years or older on 1 January 1997 (1,742 inhabitants), who were identified by a complete census of the city.A response rate of 92 per cent was achieved and 1,606 individuals participated in the baseline interview.Since the first contact, the whole cohort is invited to participate yearly in the research.The annual follow-up visits consist of standardized interviews at home and verification of death certificates, conducted by trained interviewers.From 1997 to 2007, 641 participants died and 96 (6%) were lost to followup 28 .Specific questions on satisfaction with the neighborhood were added in the household survey in 2008 (eleventh wave of the study).
The Bambuí Cohort Study of Aging was approved by the Ethics Research Committee of the Oswaldo Cruz Foundation (Fundação Oswaldo Cruz), Brazil.More details on the cohort can be seen elsewhere 28,29 .
This study consists of a cross-sectional analysis of the eleventh wave of Bambuí Cohort Study of Aging.All the 865 cohort members who were alive in 2008 were selected to participate in the present analysis.

Outcome
The dependent variable was self-rated health status measured by a single question 30  that had an eigenvalue of 5.16 and explained 64.5% of the variation between the items.Subsequently, this one extracted factor was dichotomized about the median, forming the categories "low" and "high", representing the indicator Neighborhood Satisfaction for analyses.In addition, an internal consistency test was performed to examine whether the eight items reflected one single dimension, i.e. satisfaction with the neighborhood.A Cronbach's alpha of 0.79 was found, indicating a high internal consistency of the calculated coefficient 33 .

Potential confounders
The following groups of variables were included in the analysis as potential confounding factors: (1) sociodemographic characteristics: gender, age, education (0-3; 4-7 and > 8 complete years of schooling), marital status, monthly personal income (< 2; 2-3 and > 4 times the Brazilian minimum wage), head of the household (yes, no); (2) health conditions: number of selected chronic conditions, difficulty or inability to perform selected activities of daily living (ADLs), presence of depressive symptoms; (3) lifestyle risk factors: leisure-time physical activity in previous 90 days and current smoker (yes, no).
For chronic conditions we used a checklist of 10 conditions based on the report of previous medical diagnosis for hypertension, diabetes mellitus, stroke, arthritis/rheumatism, thyroid problems, myocardial infarction, angina pectoris, heart disease, chagas disease and depression.Difficulty or inability to perform selected activities of daily living was defined as having any difficulty to perform at least one of the following activities: bathing, dressing, getting out of bed, walking from room to room, using the toilet or eating.Presence of depressive symptoms was measured by the 12-item version of the General Health Questionnaire (GHQ-12); cut-off point ≥ 5 was used based on previous study in the elderly population of Bambuí 34 .

Statistical analyses
All statistical analyses were carried out using the Stata software, version 10.0 (Stata Corp., College Station, USA).
Univariate and multivariate relationships between self-rated health and neighborhood satisfaction and also with other characteristics of the study population were analysed.All regression analyses were based on Poisson regression with a robust error variance 35 .The initial multivariate model included all independent variables with p-value less than 0.20 in the univariate analy-sis.A backward elimination process eliminated non significant variables at p-value < 0.05.The prevalence ratio (PR), 95% confidence interval (95%CI), and p-value are presented.

Results
Among 865 individuals who were alive in 2008, 814 (94.1%) participated in the present analysis.Non-participation was due to lost to follow-up.Participants and non-participants were similar in age [mean (standard deviation -SD) age of 77.7 (5.7) and 78.2 (5.6) years, respectively, p = 0.550], but women were more likely to participate than men (95.5% vs. 91.5%,p = 0.015).
A description of the participants' characteristics is shown in Table 1.The participants were predominantly composed of individuals with less than seven years of schooling (91.1%), widows (52.2%), and with monthly income below twice the Brazilian minimum wage (79.2%).Almost half self-rated their health as fair or poor and reported difficulty in performing ADLs.The majority of the participants suffered from chronic illness (71.4% had two or more conditions) and were non-exercisers (67.3%).About one in four were not satisfied with how the neighborhood was being taken care of (23.1%), or were not proud when telling others where they live (18.4%) or even would like to move out of the neighborhood (18.7%).
The results of the univariate analyses of factors associated with self-rated health are displayed in Tables 2 and 3. Socio-demographic characteristics most strongly associated with worse self-rated health were: female gender (PR = 1.29; 95%CI: 1.07-1.54),schooling (4-7 years, PR = 0.77; 95%CI: 0.65-0.92),and with not being the head of the household (PR = 1.22; 95%CI: 1.03-1.46).With respect to the neighborhood environment, higher levels of satisfaction with the neighborhood showed protective effects on self-rated health (PR = 0.63; 95%CI: 0.54-0.75).Other socio-demographic characteristics such as age, marital status and monthly personal income were unrelated to self-rated health.
In Table 4, we display the significant results of the multivariate analysis of factors associated with self-rated health.In the adjusted analysis, elders with higher satisfaction with the neighborhood (PR = 0.75; 95%CI: 0.63-0.87)were less likely to report worse self-rated health.In addition, the number of chronic diseases (two, PR = 1.69; 95%CI: 1.05-2.70,three or more, PR = 1.99; 95%CI: 1.27-3.13),difficulty to perform activities of daily living (PR = 1.51; 95%CI: 1.28-1.78),presence of depressive symptoms (PR = 1.68; 95%CI: 1.44-1.95)and frequency of leisure-time exercise in previous 90 days (less than once a week, PR = 1.24; 95%CI: 1.03-1.50)were all positively and significantly associated with poor self-rated health.

Discussion
The place of residence is endowed with both physical and social attributes that can affect the health of individuals 36 .This is corroborated by the results of our study, which showed a significant association between satisfaction with the neighborhood environment and self-rated health among the older elderly residing in a small city.Additionally, despite the lack of other similar studies with which to make a direct comparison with this finding, our results are also in accordance with previous literature demonstrating association between different aspects of neighborhood environment and self-rated health among elderly living in large cities 22,24,25,27,37 .Although there are conceptual models suggesting the importance of environmental determinants of health among older adults, the mechanisms through which the neighborhood influences their health have yet to be elucidated 16 .
The uniqueness of this study was the simultaneous analyses of both individual and perceived neighborhood environment predictors on selfrated health in a very old population (mean age of 77.7 years).In previous studies, neighborhood attributes have been characterized in terms of The neighborhood is a good place to live.Likes the neighbors and the house 6.6 Proud when tell others where he/she lives 18.4 The neighbors are willing to help each other 10.1 Children and young people treat adults with respect 6.1 The neighbourhood is a good place for children to play; and a good place to raise teenagers the quality of the local environment, neighborhood satisfaction, social organization, local community and socioeconomic status of the neighborhood 21,31,38,39 .We used a composite measure representing Satisfaction with the Neighborhood Environment.This approach allowed us to assess more social aspects of the neighborhood, such as feelings about neighborhood qualities that could not be assessed through objective questions.In fact, studies suggest that subjective indicators may be more relevant for individual health than objective measures 37,40 .Such variables can be more proximal determinants of health, above and beyond socioeconomic influences 40,41 .According to Roosa et al. 42 , individuals develop their own "filters" through which they perceive their environments.Therefore, individuals living in the same context may have different experiences about their neighborhoods.
Our definition of neighborhood refers to a person's immediate residential environment.There are multiple ways in which researchers define neighborhood environments 19 .Criteria can be historical, based on administrative boundaries, based on people's perceptions 20 .In this regard, however, it is important to remark that older adults tend to spend a greater proportion of their lives closer to home, therefore, their proximal environment could be more relevant to their health and well-being 16 .Number of chronic diseases, difficulty to perform activities of daily living, presence of depressive symptoms and frequency of leisure-time exercise in previous 90 days remained significantly associated with self-rated health among the older elderly.These results confirm previous observations of the existence of association between selfrated health and medical conditions 10,11,12,13,14 and also between self-rated health and health behaviors 12,14 .
Our study's strengths include its population basis, the very low rate of non-response and the fact that it represents, to the best of our knowledge, the first effort at examining the association between self-rated health and satisfaction with the neighborhood environment among the older elderly.On the other hand, although our findings suggest that neighborhood satisfaction may have a role to play in determining self-rated health, some caveats must be considered when interpreting this result.Due to the cross-sectional design of the study we cannot deduce causal relationships between self-rated health and related factors, but merely describe probable associations, even though the main purpose of our analyses of such risk factors was to estimate the magnitude of these associations.The use of self-reported data for both the outcome and the neighborhood characteristics could generate same-source bias, that is, the possibility of a spurious association between the two because the measurement errors in both reports are correlated or because the outcome affects the perception or report of the neighborhood attribute 36 .
In conclusion, this population-based study provided empirical evidence that satisfaction with the neighborhood environment was directly associated with the health of the older elderly in a small Brazilian city.Thess results support the potential importance of including this indicator in analyses of place and health among the elderly.Our results also provide evidence supporting the need to develop area-based programs and strategies related to the built environment.Therefore, the ways in which the elderly perceive their neighborhoods may be beneficial to their health and well-being.

9 . 5
Would like to move out (yes) 18.7 ADL: activities of daily living; GHQ-12: 12-item version of the General Health Questionnaire.* In multiples of the monthly Brazilian minimum wage (total US$ 250 during the study period).
Idoso; Saúde da População Urbana; Nível de Saúde; Estudos de Coortes Cad.Saúde Pública, Rio de Janeiro, 27 Sup 3:S390-S398, 2011 Contributors T. C. B. Luz and F. A. Proietti participated in the design, data analysis, interpretation of the results of the study and drafting the manuscript.C. C. César contributed with the statistical modeling and interpretation of the statistical findings.M. F. Lima-Costa contributed in the analysis, interpretation of data and the writing of the manuscript.All authors critically revised and approved the final manuscript.

Table 1
Selected characteristics of the study participants.The Bambuí Cohort Study of Aging, 2008.

Table 2
Univariate association between socio-demographic characteristics, satisfaction with the neighborhood and fair/poor self-rated health.The Bambuí Cohort Study of Aging, 2008.
* Pearson chi-square test; ** In multiples of the monthly Brazilian minimum wage (total = US$ 250 during the study period).

Table 3
Univariate association between health characteristics and health behaviors and fair/poor self-rated health.The Bambuí Cohort Study of Aging, 2008.

Table 4
Statistically signifi cant results for the multivariate analysis of factors associated with fair/poor self-rated health.The Bambuí Cohort Study of Aging, 2008.