Elsevier

Health & Place

Volume 22, July 2013, Pages 82-89
Health & Place

Who does well where? Exploring how self-rated health differs across diverse people and neighborhoods

https://doi.org/10.1016/j.healthplace.2013.03.006Get rights and content

Highlights

  • Neighborhood disadvantage amplified the impact of individual-level SEP on SRH.

  • Using a graded measure of SRH detected stronger relationships for poor and excellent outcomes.

  • Individual- and area-level disadvantage measures should be considered simultaneously when examining SRH.

Abstract

This work establishes whether neighborhood disadvantage amplifies the impact of socioeconomic position (SEP) on a graded measure of self-rated health (SRH). SRH data were taken from 10,932 adults recruited across 200 Brisbane neighborhoods. After adjusting for demographics, those who lived in the most disadvantaged neighborhoods were more likely to report poor SRH than those living in the least disadvantaged neighborhoods (OR=2.67). Those with the lowest SEP and lived in the most advantaged neighborhoods had a similar probability of reporting excellent SRH as those with the highest SEP living in the most disadvantaged neighborhoods. This work highlights the importance of examining SEP and neighborhood-level disadvantage simultaneously when planning communities.

Introduction

The relationships between social inequalities and health outcomes is well established in developed countries; those who are socioeconomically disadvantaged or live in deprived neighborhoods experience higher levels of disease and die earlier compared with those more advantaged (Crombie et al., 2002, Strategic Review of Health Inequalities in England Post 2010, 2010). This phenomenon manifests itself in part through compositional (individual-level) and contextual (area-level) exposures (Kawachi, 2002, Macintyre, 2007). Health and place-based research has established potential causal pathways between these exposures on health outcomes, supported by quantitative and qualitative evidence showing who you are and where you live are associated with health behaviors and outcomes (Torsheim et al., 2004, Franzini et al., 2005, Cummins et al., 2007, Chandola et al., 2003, Sacker et al., 2000, Wilkinson and Pickett, 2006, Giles-Corti and Donovan, 2003, Turrell and Mathers, 2001).

Using the deprivation amplification argument (Macintyre, 2007), compositional and contextual effects compound; that is, having a low socioeconomic position (SEP) and living in deprived areas expose individuals to double disadvantage. Conversely, there is a ‘pulling-up’ effect for those of low SEP residing in more advantaged areas. When matched by SEP, those living in less deprived areas have better health profiles than those living in more deprived areas (Cummins et al., 2007, Strategic Review of Health Inequalities in England Post 2010, 2010). Earlier work suggests contextual factors have only a modest effect on explaining health outcomes when compared with compositional factors (Pickett and Pearl, 2001). More recent work using multilevel modeling, however, has further considered the complexities of health and place-based research, suggesting that the strength of associations between contextual factors and health outcomes is stronger than previously thought, but differs by the scale of administrative units applied (Giles-Corti et al., 2005), health outcomes examined, and measures of area-level exposure (Riva et al., 2007).

Self-rated health (SRH) is associated with overall mortality and morbidity across the socioeconomic spectrum (Burström and Fredlund, 2001, Kaplan and Camacho, 1983), and is influenced by both individual-level SEP and neighborhood disadvantage (Browning and Cagney, 2002, Kennedy et al., 1998, Torsheim et al., 2004, Franzini et al., 2005). Different SRH reference points are used differently by population groups; for example when asked about SRH, older adults are more likely to consider presence of chronic conditions, whereas those with higher education attainment are more likely to reflect on their general health status (Krause and Jay, 1994). However, a meta-analysis demonstrated the global measures of SRH accurately predict mortality after adjusting for functional status, depression, and co-morbidities (DeSalvo et al., 2006). Franzini et al. (2005) developed pathways for conceptualizing how neighborhood-level disadvantage (both physical and social aspects) contribute to SRH. Neighborhood-level disadvantage was associated with lower collective efficacy, trust, and social capital, and higher levels of social and physical disorder, fear of crime, and racism; demonstrating the relationship between neighborhood disadvantage and SRH is mediated by the social and physical characteristics of the area. Other work, using student data obtained from 22 countries, demonstrated additive, rather than interactive effects, of SEP and neighborhood-level disadvantage on SRH (Torsheim et al., 2004). When SEP was considered alone, the most deprived students were three times more likely to report poor SRH than the least deprived students. However, when SEP and neighborhood-level disadvantage were considered together, the most deprived students were eight times more likely to report poor SRH compared with the least deprived students (Torsheim et al., 2004). Together, this emphasizes the importance of studying the multilevel structure of disadvantage, and indeed whether there is a double disadvantage, when considering SRH.

The rationale for choosing specific indicators of SEP is often unjustified, and different measures are frequently treated as having the same conceptual underpinning and underlying constructs, despite empirical evidence suggesting otherwise (Dutton et al., 2005). Using multiple measures of SEP, such as education attainment and household income, has been suggested as a way of revealing different patterning of socioeconomic markers on health outcomes through diverse causal pathways (Geronimus and Bound, 2007). Furthermore, education attainment and household income have established unique and predictive contributions to SRH (Lantz et al., 2001). These two measures of SEP are appropriate for the ‘baby boomer’ population being investigated here (i.e., those born between 1946 and 1964), as most have attained their highest qualification and career advancement, therefore near maximum earning potential at the time of being studied.

As well as conceptual considerations, a further limitation of research using SRH as an outcome is that the measure is frequently collapsed into a dichotomy (poorer versus better SRH), despite being initially assessed on a three- or five-point scale (Power et al., 1998, Mackenbach et al., 1997, Rahkonen et al., 1995). Using this reductionist data treatment approach potentially obscures important patterns of SRH and results in a loss of information and statistical efficiency (Agresti, 1984). An approach that considers SRH in a more sensitive manner (i.e., using a multinomial rather than binomial outcome) contributes further to our understanding of the outcome as a continuum in relation to measures of disadvantage (Manor et al., 2000).

Taking into account this earlier work, we are interested in exploring how different exposures to disadvantage may result in inequalities for SRH in middle-aged adults, living in a relatively affluent country (Australia) (United Nations Development Program, 2011). We believe it reasonable to expect education attainment and household income to be independently, and differently associated with SRH, with the strength of these associations varying by neighborhood-level disadvantage. The first aim of this paper was to establish the multilevel relationship of disadvantage on categorical responses of SRH in middle-aged Australian adults. The second aim was to examine the association between two measures of SEP on categorical responses of SRH, and whether these differed across neighborhoods that varied by disadvantage. We explore whether double disadvantage exists for those with low SEP in the Australian context, and the protective effects, if any, that living in more affluent neighborhoods may provide. We unpack these relationships using multilevel approaches and finer-grained measures of SRH.

Section snippets

Methods

This research uses the 2007 survey data (wave I) of the HABITAT (How Areas in Brisbane Influence Health and Activity) study. HABITAT is a longitudinal (2007–2011) study of ‘baby boomer’ women and men living in Brisbane, Australia. Detailed information about the study can be found elsewhere (Burton et al., 2009). HABITAT was granted ethical approval by The University Human Research Ethics Committee at the Queensland University of Technology (ID3967H).

Results

Overall, 17.8% of participants reported poor SRH, 38.3% reported good SRH, and 43.8% reported excellent SRH. Those living in the most disadvantaged neighborhoods were less likely to have qualifications above school-level, were more likely to have a lower household income, and to not be in paid employment compared with those living in more advantaged neighborhoods (Table 1).

Those of lower SEP, and those living in more disadvantaged neighborhoods were more likely to report poor or good (rather

Discussion

Our findings support well-established associations between health and disadvantage, suggesting that even in a relatively affluent country like Australia that experiences a high quality of life (United Nations Development Program, 2011), those receiving lower household incomes are doubly disadvantaged if they live in less affluent neighborhoods. This inequality attenuates if they live in more advantaged neighborhoods. After accounting for common compositional variables, neighborhood-level

Acknowledgments

Funding: HABITAT is supported by two (Australian) National Health and Medical Research Council Project Grants (ID 339718, 497236). GT is supported by an (Australian) National Health and Medical Research Council Senior Research Fellowship (ID 1003710). BGC is supported by an (Australian) National Health and Medical Research Council Principal Research Fellowship (ID 1004900).

References (49)

  • A. Agresti

    Analysis of ordinal categorical data

    (1984)
  • R. Andersen et al.

    Access to medical care for low-income persons: How do communities make a difference?

    Medical Care Research and Review

    (2002)
  • W. Browne

    MCMC estimation in MLwiN (version 2.10)

    (2009)
  • C. Browning et al.

    Neighborhood structural disadvantage, collective efficacy, and self-rated physical health in an urban setting

    Journal of Health and Social Behavior

    (2002)
  • B. Burström et al.

    Self-rated health: Is it a good predictor of subsequent mortality among adults in lower as well as higher social classes?

    Journal of Epidemiology and Community Health

    (2001)
  • N. Burton et al.

    HABITAT: A longitudinal multilevel study of physical activity change in mid-aged adults

    BMC Public Health

    (2009)
  • T. Chandola et al.

    Social inequalities in health by individual and household measures of social position in a cohort of healthy people

    Journal of Epidemiology and Community Health

    (2003)
  • I. Crombie et al.

    Closing the health inequalities gap: An international perspective

    (2002)
  • G. Davey Smith et al.

    Education and occupational social class: which is the more important indicator of mortality risk?

    Journal of Epidemiology and Community Health

    (1998)
  • K. DeSalvo et al.

    Mortality prediction with a single general self-rated health question. A meta-analysis

    Journal of General Internal Medicine

    (2006)
  • D. Dillman

    Mail and internet surveys: The tailored design method

    (2000)
  • T. Dutton et al.

    Measuring socioeconomic position in population health monitoring and health research. Health Inequalities Monitoring Series

    (2005)
  • I. Ericksson et al.

    Self-rated health. Comparisons between three different measures. Results from a population study.

    International Journal of Epidemiology

    (2001)
  • A. Geronimus et al.

    Use of census-based aggregate variables to proxy for socioeconomic group: Evidence from national samples

    American Journal of Epidemiology

    (2007)
  • Cited by (0)

    View full text