Artículo original

 

Self-rated health in Brazilian adults and elderly: Data from the National Household Sample Survey 2008

 

Percepción de la salud en adultos y ancianos brasileños: datos de la Encuesta Nacional de Hogares 2008

 

Doroteia Aparecida Höfelmann, PhD PH,⁽¹⁾ Leila Posenato Garcia, PhD Epidem,⁽²⁾ Lúcia Rolim Santana de Freitas, MSc Stat.⁽³⁾

 

(1) Post-Graduate Program in Nutrition, Federal University of Santa Catarina. Florianópolis, Brazil.

(2) Department of Nutrition, Federal University of Paraná. Curitiba, Brasil.

(3) Post-Graduate Program in Tropical Medicine, University of Brasília. Brasília, Brazil.

 

Corresponding author

 

---

Abstract

Objective. To investigate prevalence of poor self-rated health and its association with individual and household-level characteristics among adults and elderly in Brazil.

Materials and methods. Cross-sectional study with Brazilian National Household Sample Survey 2008 (n=257 816). Crude and multilevel-adjusted Poisson regression models were fitted.

Results. After adjusted analysis, poor self-rated health was significantly associated with higher household income, living alone, not having piped water nor garbage collection, lower education, not having health insurance, female sex, higher age, being a current or previous smoker, physical inactivity, having chronic diseases, having physical impairment. Subjects living in rural areas also had higher prevalence of poor self-rated health. The factors most strongly associated with the outcome were physical impairment and reporting three or more chronic diseases.

Conclusions. Socioeconomic, health related behaviors, and physical health were associated with poor self-rated health.

Key words: cross-sectional studies; environment and public health; health inequalities; housing; income; multilevel analysis; socioeconomic factors; Brazil.

---

Resumen

Objetivo. Investigar la prevalencia de la percepción negativa de salud y su asociación con características individuales a nivel de los hogares en adultos y adultos mayores de Brasil.

Material y métodos. Estudio transversal con datos de la Encuesta Nacional de Hogares de 2008 (n=257 816). Se estimaron modelos de regresión de Poisson multinivel crudos y ajustados.

Resultados. Después del análisis ajustado, la autopercepción negativa de salud se asoció significativamente con mayor ingreso, vivir solo, no tener agua corriente ni recolección de basura, baja educación, carecer de seguro de salud, sexo femenino, mayor edad, tabaquismo, inactividad física, enfermedades crónicas y deterioro físico. Los habitantes de zonas rurales también tuvieron mayor prevalencia de percepción negativa. Los factores más fuertemente asociados fueron impedimento físico y presentación de tres o más enfermedades crónicas.

Conclusiones. Factores socioeconómicos, comportamientos relacionados con la salud y salud física se asociaron con la percepción negativa.

Palabras clave: estudios transversales; medio ambiente y salud pública; desigualdades en la salud; vivienda; renta; análisis multinivel; factores socioeconómicos; Brasil.

---

 

Self-rated health represents a multidimensional concept that encompass multiple aspects, including socioeconomic, psychosocial and health status background.¹ The measure is used in several surveys and researches worldwide,^2-4 and strongly predicts morbidity and mortality.^1,5 Women, individuals with diseases, disabilities, unprivileged socioeconomic conditions, usually report poorer self-rated health.^2,6

Pathways by which socioeconomic status influences health should be those that affect health more generally, as biological determinants, health care, environmental exposure, and health related behaviors.^7,8 People living in more deprived socioeconomic conditions are exposed to overcrowding, physical stressors, violence, lack of safe places for social interactions and other factors,^2,7 which can lead to a harmful lifestyle, including use and abuse of smoking, drugs, alcohol, poor diet, among others.⁹

Lately, a renewed interest in impact of collective factors in health has taken place in epidemiological studies, beyond individual variables. The role of structural factors on health has been highlighted.¹⁰ Factors as contextual income, ethnic homogeneity, physical disorder, social capital,¹¹ violence, and others have been associated with several health outcomes.¹⁰

Brazilian population experienced changes in several health determinants, characterized with a society with high social inequalities.^8,12 In the last decades, the population living in urban areas has increased, and living conditions changed in both favorable and unfavorable ways. The impact of urbanization on health is not homogenous, it depends on social and historic aspects, in human relationship with the environment, and public policies.^13,14

According to the Brazilian Census 2010, 84% of the population resided in urban areas.¹⁵ The urbanization has traditionally been linked to development, which in turn is connected to health. Still, in several countries including Brazil, the urban population growth was reflected in the growth of slums, which are related to poor health.¹⁵ However, few studies have been conducted to investigate associations between self-rated health, at individual and household levels, including urbanicity, in middle and low income countries, including Brazil.^2,4,16

The aim of this study is to investigate the association between individual and household-level characteristics and self-rated health in the Brazilian population, using multilevel modeling techniques.

 

Materials and methods

Cross-sectional study using microdata from the National Household Sample Survey (PNAD, in Portuguese) carried out in 2008 by the Brazilian Institute for Geography and Statistics (IBGE, in Portuguese).

The PNAD uses a three-stage complex probabilistic sample: cities, census tracts and households and is representative of the national regional and state levels. The variables that define the structure of the sampling plan, called stratum and primary sampling unit (PSU), and the sample weights need to be considered in the statistical analysis. In 2008, PNAD included in its questionnaire a health supplement. Information on all 391 868 people from 150 591 sampled households was obtained by means of interviews; it relies on self-reporting. The IBGE discusses these sampling processing in more detail elsewhere.¹⁷

Persons under 20 years old were excluded of the analysis. The outcome variable, self-rated health, was collected by the question: "In general, do you consider your own health as very good, good, fair, poor or very poor?". For the analysis, the variable was dichotomized considering poor self-rated health to be grouping of the categories poor and very poor. This categorization used for the results of this study is consistent with previous research that examined self-rated health from PNAD data.^18,19

The independent variables investigated included socioeconomic behavioral, morbidity and demographic variables at the individual level, socioeconomic and environmental factors at the household level, and household location at the region level. The variables studied, according to their hierarchical levels of analysis, are presented in figure 1.

At the individual level the following variables were used: gender, age (measured in categories of 20-29, and 30-39, 40-49, 50-59, and 60 years and over), education level (0-3 completed years, 4-7 years, 8-10, and 11 years or more), health insurance (yes or no), skin color (white or yellow, and black, brown or indigenous), smoker (never, former and current), physical inactivity (yes or no), chronic diseases (none, 1 or 2, and 3 or more), and physical mobility (without limitation, a little restraint, and physical impairment).

In the PNAD, the variable current smoking was obtained through the questions: "Do you currently smoke any tobacco products?" and "Adding all cigarettes smoked in your lifetime, the total comes to at least five packs or a hundred cigarettes?". The variable was grouped into categories: smokers (individuals who had smoked at least five packs of cigarettes in their lives and who smoked at the time of conducting the research) former smokers (individuals who had smoked five packs but did not smoke at the time of conducting the research) and nonsmokers (individuals who did not smoke at least five packs in their lives).²⁰

Physical inactivity was defined as when a person reported not to practice any physical activity in all areas studied (commuting, work activity, cleaning the home, environment and physical leisure activities).

The information on chronic diseases included the self-report of 12 specific conditions or diseases, namely: column or back disease, arthritis or rheumatism, cancer, diabetes, bronchitis or asthma, hypertension, heart disease, chronic renal failure, depression, tuberculosis, tendonitis or tenosynovitis and cirrhosis. For each of these conditions it was asked if a doctor or health professional had diagnosed the person.²⁰

At the household level the following variables were used: quartile of household income per capita (poorest: 1st, 2nd, 3rd, and richest: 4th), type of household (tenant or owner), garbage collection (yes or no), and having piped water (yes or no).

At the region level, the following variables were used: household location (urban or rural), region (Southeastern, Midwestern, Southern, Northern, and Northeastern).

The classification of the status of the household in urban or rural, depends on its location, and is based on current legislation at the completion of the previous census. Urban households are those located in city areas, villages or isolated urban areas. Rural households are those located in the entire area outside these limits.²⁰

All analyses were adjusted for the sample design of the PNAD, including all the characteristics of the complex sampling design (weight, strata and primary sampling unit). The software STATA, version 10.0, was used to perform these analyses.

After description, bivariate analysis was conducted to estimate differences among proportions, through Fisher's exact test for categorical exposure variables and linear trend test for ordinal ones. The multivariate analysis, through multilevel Poisson regression with robust variance, was oriented by the hierarchic model shown in figure 1, using backwards regression, level by level. Variables that presented association with p-value p≤0.5 were kept in the model, aiming to control for possible confounding. Confidence intervals (95%) were calculated. Associations with p≤0.05 were considered statistically significant. In this paper, we have used standardized residuals to investigate if model assumptions such as homoscedasticity and normally distributed errors were violated.²¹

The database didn't identify the individuals who provided information, which preserves data anonymity. The study was conducted according to the ethical principles embodied in the Declaration of Helsinki and observing conditions of the Resolution n. 196 from the Brazilian National Health Council.

 

Results

Sample was comprised of 257 816 individuals, 52.6% women and 47.4% men. Smoking prevalence in the investigated sample was nearly 16 and 41% for physical inactivity. About 40% of individuals presented at least one chronic disease and 85% lived in urban areas. Half of the individuals reported black skin, light skin or indigenous color, while others reported white or Asian skin color. Table I presents complete variables description.

Table II presents crude and adjusted analysis, through multilevel robust Poisson's regression, of the association between self-rated health and the independent variables. In the crude analysis, poor self-rated health was positively associated with residing in rural area, residing in the Southern, Northern and Northeastern regions, living alone, lower household income, having household property, not having garbage collection, not having water supply, lower education, not having private health insurance, female gender, higher age, being black, brown or indigenous, being current or former smoker, being physically inactive, having chronic diseases, having limitation in physical mobility. After the adjusted analysis, poor self-rated health was associated with all these variables, except for skin color.

In the adjusted analysis, the variables most strongly associated with self-rated health were physical mobility and chronic diseases. For those variables, the prevalence ratios in the adjusted analysis were higher than those observed in the crude analysis. Individuals presenting limited physical mobility had a 688% higher probability of evaluating poorly their health than those who did not report limited physical mobility. Individuals who reported having three or more chronic diseases had a 457% higher probability of having poor self-rated health than those who did not report those diseases. Having 60 years of age and belonging to the first quartile of household income per capita were also strong risk factors for having poor self-rated health.

 

Discussion

This paper aimed to evaluate households and individual factors associated with poor self-rated health in the Brazilian population. Most of the socioeconomic and demographic variables located in all analysis levels (individual household and region) were associated with the outcome.

Among the household variables investigated, the presence of basic assets as garbage collection, piped water, as well as the condition of owning other own house, were positively associated with better self-rated health, and those associations remained even after adjustment for other individual level variables. The association among a variety of socioeconomic measures and self-rated health has been demonstrated in several studies.^2,22,23 That factor indicates a strong and persistent socioeconomic disadvantage in health for people in the bottom of the social scale.

In Brazil, garbage collection is a particular issue, considering that besides being a marker of living conditions in the area of residence, and health problems related to the accumulation of garbage, there is an additional concern for dengue transmission, since the uncollected garbage serves as breeding site for the mosquito vectors.²⁴

A constellation of pathways connecting socioeconomic disadvantage and poor health outcomes has been identified in epidemiological literature. Socioeconomic status underlies three major determinants of health: health care, environmental exposure, and health behavior. In addition, chronic stress associated with lower socioeconomic status may also increase morbidity and mortality.⁷

People living in rural areas presented higher prevalence of poor self-rated health. In other study using the same database, De Moraes and colleagues²⁵ found that after controlling for individual and environmental factors, the association between household location area and self-rated health was modified and lost their statistical significance. It is important to point out that in that study they work with self-rated health in an ordinal instead of dichotomized form of categorization. However, authors observed significant interaction among household location area and gender, skin color, self-reported morbidity, ownership of basic assets and percentage of households with adequate dwelling quality.²⁵

The percentage of rural Brazilian population experienced important declines in last decades. In 1960, most of the country population was living in rural areas, except in the Southeastern region (57% in urban area). In the 2010 Brazilian census, 15.6% of the population was residing in rural areas, varying in range from 7 to 26%, in Southeastern and North and Northeastern regions, respectively.^17,26

There is some evidence that health indicators are not only better in urban than rural areas (especially in less wealthy nations) but that the urban poor fare better than the nonurban poor.²⁷ Vlahov and colleagues²⁷ reviewed the hypothesis about the urban "advantage" in health, and point out some explanations, including: "the proximity of wealth and poverty within cities brings benefits to those less well-off, the availability of higher levels of social support and greater social cohesion in urban than nonurban areas, the offer of more access to the necessities of life, a physical environment that is conducive to health, and finally, cities through their size and density offer the potential for political mobilization and social movements, enabling urban populations to win more resources for health, another possible route to a health advantage".²⁷

In Brazil, the urban population is heterogeneous. Urban areas shelter not only the people with the best socioeconomic conditions, but also those with very poor living conditions, particularly the population that lives in slums.²⁸ However population living in rural areas experience lower socioeconomic conditions, including lower monthly per capita income (less than half part of urban), the illiteracy in people with 15 years or more is 7.5% in urban against 23.5% in rural area. Furthermore, one third of inhabitants in rural areas do not have piped water, while in cities this proportion doesn't reach 3%, and the health insurance coverage is lower in rural areas.²⁹ Those are some possibilities that can partially explain the higher prevalence of poor self-rated health among poor residents in rural areas.

In this study, those that adopted healthy behaviors, as never smokers and people who practice physical activity presented better health. The association between engagement in health related behaviors and self-rated health has been demonstrated in other studies.³⁰ Health related behaviors represent one of the mechanisms behind higher socioeconomic advantage in health outcomes.⁷ Kim,³¹ studying a Korean population, observed that engaging in regular exercise significantly mediated the relationship between education and self-rated health as well as between poverty and self-rated health. Finally, poverty and regular exercise had a greater impact on self-rated health in old age than in middle age.³¹

The variables most strongly associated with self-rated health in this study were physical impairment, and reporting of three or more chronic diseases. Health status variables and self-rated health were consistently associated in most studies. In fact, the self-rated health variable is sensitive to changes like physical health decay.^5,32 The effect of disease in health is associated with the complexity of the therapeutic process and the psychological and financial resources available to the individual to deal with the illness. Disease jeopardizes people's quality of life, altering the reproduction of social conditions for existence by limiting the performance of their everyday and occupational activities.³³ In a study conducted among a workers' population, Höfelmann and Blank³⁴ found that psychosocial (-25.59%), socioeconomic (-9.29%), and occupational variables (10.54%) were important confounders in the association between self-rated health and chronic diseases and/or symptoms.

The limitations of the present study are related, mainly, to methodological aspects from the PNAD database. The prevalence of chronic disease and other self reported measures could be underestimated.³⁵ The cross-sectional design of the study does not allow drawing of causal inferences, since the information about exposure and outcome was obtained at the same time.

A potential source of bias is related to the use of proxy respondents for some information. Several authors have studied that limitation in PNAD data,^36,37 including its implication for self-rated health prevalence.^19,38

Furthermore, it is known that individuals with worse socioeconomic position experience lower survival rates. Thus, their participation in the research may be underrepresented in the sample due to survival bias. As a result, socioeconomic lags may be even stronger than those observed in the studied population.^39,40

Some variables that would allow for a more detailed analysis, such as size of the municipality, were not available in the data set. It is recommended for future studies to explore such variables.

Despite of all those facts, we found important socioeconomic gradients between self-rated health, and both individual and household level variables were associated with the outcome. Furthermore, health related behaviors and physical health status were associated with poor self-rated health.

Our findings reinforce the importance that measures to reduce inequalities in health should be multidisciplinary, involving different civil society sectors, and focused in providing better living conditions to socioeconomic unprivileged groups, which can be more vulnerable to physical problems in health.

 

References

1. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38(1):21-37.         [ Links ]

2. Cremonese C, Backes V, Olinto MT, Dias-da Costa JS, Pattussi MP. Neighborhood sociodemographic and environmental contexts and self-rated health among Brazilian adults: a multilevel study. Cad Saude Publica 2010;26(12):2368-2378.         [ Links ]

3. Franzini L, Giannoni M. Determinants of health disparities between Italian regions. BMC Public Health 2011;10:296.         [ Links ]

4. Hurtado D, Kawachi I, Sudarsky J. Social capital and self-rated health in Colombia: the good, the bad and the ugly. Soc Sci Med 2011;72(4):584-590.         [ Links ]

5. Jylha M. What is self-rated health and why does it predict mortality? Towards a unified conceptual model. Soc Sci Med 2009;69(3):307-316.         [ Links ]

6. Garcia LP, Höfelmann DA, Facchini LA. Self-rated health and working conditions among workers from primary health care centers in Brazil. Cad Saude Publica 2010;26(5):971-980.         [ Links ]

7. Adler NE, Newman K. Socioeconomic disparities in health: pathways and policies. Health Aff (Millwood) 2002;21(2):60-76.         [ Links ]

8. Pavao AL, Coeli CM, Lopes Cde S, Faerstein E, Werneck GL, Chor D. Social determinants of the use of health services among a public university workers. Rev Saude Publica 2012;46(1):98-103.         [ Links ]

9. Weyers S, Dragano N, Richter M, Bosma H. How does socio economic position link to health behaviour? Sociological pathways and perspectives for health promotion. Glob Health Promot 2010;17(2):25-33.         [ Links ]

10. Diez-Roux AV, Mair C. Neighborhoods and health. Ann N Y Acad Sci 2010;1186:125-145.         [ Links ]

11. Kawachi I, Kennedy BP, Glass R. Social capital and self-rated health: a contextual analysis. Am J Public Health 1999;89(8):1187-1193.         [ Links ]

12. Belon AP, Barros MB, Marin-Leon L. Mortality among adults: gender and socioeconomic differences in a Brazilian city. BMC Public Health 2012;12(1):39.         [ Links ]

13. Caiaffa WT, Ferreira FR, Ferreira AD, Oliveira CD, Camargos VP, Proietti FA. [Urban health: "the city is a strange lady, smiling today, devouring you tomorrow"]. Cien Saude Colet 2008;13(6):1785-1796.         [ Links ]

14. Vlahov D, Freudenberg N, Proietti F, Ompad D, Quinn A, Nandi V, et al. Urban as a determinant of health. J Urban Health 2007;84 suppl 3:i16-26.         [ Links ]

15. Instituto Brasileiro de Geografia e Estatística. Censo demográfico 2010: características da população e dos domicílios. Resultados do universo. Rio de Janeiro: IBGE, 2012.         [ Links ]

16. Giatti L, Barreto SM, Cesar CC. Unemployment and self-rated health: neighborhood influence. Soc Sci Med 2011;71(4):815-823.         [ Links ]

17. Brazilian Institute of Geography and Statistics. Pesquisa Nacional por Amostra de Domicílios: Um panorama da saúde no Brasil, acesso e utilização dos serviços, condições de saúde e fatores de risco e proteção à saúde, 2008. Rio de Janeiro: IBGE, 2010.         [ Links ]

18. Dachs J. Determinantes das desigualdades na auto-avaliação do estado de saúde no Brasil: análise dos dados da PNAD/1998. Ciênc Saúde Coletiva 2002;7(4):641-657.         [ Links ]

19. Dachs J. Determinantes das desigualdades na autoavaliação do estado de saúde no Brasil: análise dos dados da PNAD/1998. Cien Saude Colet 2002;7(4):641-657.         [ Links ]

20. Instituto Brasileiro de Geografia e Estatística. PNAD - Nota Técnica. Rio de Janeiro: IBGE, 2009.         [ Links ]

21. Rabe-Hesketh S, Skrondal A. Multilevel and longitudinal modeling using Stata. 2 ed. Texas: Stata Press, 2008.         [ Links ]

22. Quesnel-Vallee A. Self-rated health: caught in the crossfire of the quest for 'true' health? Int J Epidemiol 2007;36(6):1161-1164.         [ Links ]

23. Theme-Filha MM, Szwarcwald CL, Souza-Junior PR. Socio-demographic characteristics, treatment coverage, and self-rated health of individuals who reported six chronic diseases in Brazil, 2003. Cad Saude Publica 2005;21 suppl :43-53.         [ Links ]

24. Penna MLF. Um desafio para a saúde pública brasileira: o controle do dengue. Cad Saude Publica 2003;19(1):305-309.         [ Links ]

25. De Moraes JR, Moreira JP, Luiz RR. [Association between self-reported state of health among adults and the location of the home: ordinal logistic regression analysis using PNAD 2008]. Cien Saude Colet 16(9):3769-3780.         [ Links ]

26. Instituto Brasileiro de Geografia e Estatística. Cidades. Rio de Janeiro: IBGE, 2011; vol. 2011, 2010.         [ Links ]

27. Vlahov D, Galea S, Freudenberg N. The urban health "advantage". J Urban Health. 2005;82(1):1-4.         [ Links ]

28. Kassouf AL. Acesso aos serviços de saúde nas áreas urbana e rural do Brasil. Rev Econ Sociol Rural 2005;43(1):29-44.         [ Links ]

29. Instituto de Pesquisa Econômica Aplicada. PNAD 2008: Primeiras análises - O setor rural. Brasília: Ipea, 2010.         [ Links ]

30. Conry MC, Morgan K, Curry P, McGee H, Harrington J, Ward M, Shelley E. The clustering of health behaviours in Ireland and their relationship with mental health, self-rated health and quality of life. BMC Public Health 2011;11:692.         [ Links ]

31. Kim J. The mediating effects of lifestyle factors on the relationship between socioeconomic status and self-rated health among middle-aged and older adults in Korea. Int J Aging Hum Dev 2011;73(2):153-173.         [ Links ]

32. Lima-Costa MF, Cesar CC, Chor D, Proietti FA. Self-rated health compared with objectively measured health status as a tool for mortality risk screening in older adults: 10-year follow-up of the Bambui Cohort Study of Aging. Am J Epidemiol 2012;175(3):228-235.         [ Links ]

33. Minayo-Gomez C. O desafio do conhecimento: pesquisa qualitativa em saúde. 8 ed. São Pualo: Hucitec, 2004.         [ Links ]

34. Hofelmann DA, Blank N. Identification of confounders in the association between self-reported diseases and symptoms and self-rated health in a group of factory workers. Cad Saude Publica 2008;24(5):983-992.         [ Links ]

35. Alonso A, Beunza JJ, Delgado-Rodríguez M, Martínez-González MA. Validation of self reported diagnosis of hypertension in a cohort of university graduates in Spain. BMC Public Health 2005;5:94.         [ Links ]

36. Jardim R, Barreto S, Gonçalves L. Confiabilidade do informante secundário em inquéritos de saúde. Revista Brasileira de Estudos de População 2009;26:141-144        [ Links ]

37. Lima-Costa M, Barreto S, Giatti L. A situação socioeconómica afeta igualmente a saúde de idosos e adultos mais jovens no Brasil? Um estudo utilizando dados da Pesquisa Nacional por Amostras de Domicílios-PNAD/98. Cien Saude Colet 2002;7(4):813-824.         [ Links ]

38. Dachs J, Santos A. Auto-avaliação do estado de saúde no Brasil: análise dos dados da PNAD/2003. Cien Saude Colet 2006;11(4):887-894.         [ Links ]

39. Barros MB, Francisco PM, Zanchetta LM, Cesar CL. [Trends in social and demographic inequalities in the prevalence of chronic diseases in Brazil. PNAD: 2003- 2008]. Cien Saude Colet 2011;16(9):3755-3768.         [ Links ]

40. Mackenbach JP, Looman CW, van der Meer JB. Differences in the misreporting of chronic conditions, by level of education: the effect on inequalities in prevalence rates. Am J Public Health 1996;86(5):706-711.         [ Links ]

 

Recieved on: November 27, 2014
Accepted on: July 31, 2014

 

Corresponding author:
Dr. Doroteia Aparecida Höfelmann.
Departament of Nutrition, Botanical Campus.
Federal University of Paraná. Av. Lothário Meissner 632. 80210-170. Jardin Botánico. Curitiba, PR, Brasil.
E-mail: doroaph@gmail.com

 

Declaration of conflict of interests. The authors declare that they have no conflict of interests.