Elsevier

Social Science & Medicine

Volume 65, Issue 8, October 2007, Pages 1779-1791
Social Science & Medicine

Non-residential neighborhood exposures suppress neighborhood effects on self-rated health

https://doi.org/10.1016/j.socscimed.2007.05.051Get rights and content

Abstract

In prior research, neighborhood effects have often been weak or inconsistent in predicting specific causes of mortality and morbidity. To determine whether residential neighborhood effects are suppressed by exposure to other environments, we examined the effect on adult self-rated health of non-residential environments that figure in individuals’ daily routines. We linked the 2000 US Census data with the Los Angeles Family and Neighborhood Study (L.A.FANS) database, which consists of 3323 adults sampled from neighborhoods in LA County. Characteristics of census tracts where respondents lived, worked, shopped, sought medical care, worshipped and spent “other” time were obtained from the 2000 US Census. Weighted multilevel linear and clustered generalized ordered logistic regressions were used to estimate associations between self-rated health and non-residential neighborhood exposures after adjustment for individual-level factors and exposure to residential neighborhoods.

We found that residence in disadvantaged neighborhoods was associated with worse self-rated health. In a dose–response fashion, the greater the exposure to less disadvantaged non-residential neighborhoods in the course of routine activities, the greater the magnitude in improved self-rated health. Models including non-residential neighborhood exposure increase the magnitude and significance of the association between residential neighborhoods and health. In conclusion, individuals’ exposure to non-residential neighborhoods confounds and suppresses the association of residential neighborhoods with health and could explain why previous studies may not have found robust associations between residential neighborhood predictors and health.

Introduction

In the past decade, researchers, using multilevel models, have begun to examine residential neighborhood-level predictors (social and structural mechanisms) that might explain the geographic distribution of disease (Cohen et al., 2000; Robert, 1999; Yen & Syme, 1999). The most common area characteristics investigated have been aggregate measures of the socioeconomic characteristics of individuals who reside in these neighborhoods (Diez-Roux, 2001; Morris & Carstairs, 1991). Residential neighborhood socioeconomic characteristics (SES) is felt to be either a progenitor or a proxy of neighborhood environmental and psycho-social factors that may be associated with the development of various health outcomes.

Contextual studies have noted consistency in these studies documenting “independent” effects of neighborhood socioeconomic environment after controlling for individual-level factors on various health outcomes (Diez-Roux, 2001; Pickett & Pearl, 2001). But if we restrict these studies only to multilevel models that have adjusted for more than one individual level measure of socioeconomic status, though the overall effect suggests independent effects of neighborhood SES on various health outcomes (depression, Ross, 2000; heart disease, Diez-Roux et al., 2003; drug use, Boardman, Finch, Ellison, Williams, & Jackson, 2001; insulin resistance, Diez-Roux, Jacobs, & Kiefe, 2002; cardiovascular risk factors, Davey Smith, Hart, Watt, Hole, & Hawthorne, 1998; Duncan, Jones, & Moon, 1999; Lee & Cubbin, 2002; child mental health, Xue, Leventhal, Brooks-Gunn, & Earls, 2005; self-rated health, Patel, Eschbach, Rudkin, Peek, & Markides, 2003; mortality, Davey Smith et al., 1998), neighborhood socioeconomic predictors are frequently weak (Aneshensel & Sucoff, 1996; Borrell, Diez-Roux, Rose, & Clark, 2004; Cagney & Browning, 2004; Diez-Roux et al., 1999; Reijneveld, 1998; Robert, 1998; Robert & Reither, 2004; Van Lenthe & Mackenbach, 2002), and sometimes inconsistent across gender (Diez-Roux, Jacobs, & Kiefe (2002), Diez-Roux et al. (1997); Robert & Reither, 2004; Van Lenthe & Mackenbach, 2002) and race/ethnicity (Borrell et al., 2004; Diez-Roux et al., 1997; Lee & Cubbin, 2002) and across different study samples in the US (Aneshensel & Sucoff, 1996; Boardman et al., 2001; Borrell et al., 2004; Cagney & Browning, 2004; Diez-Roux, Jacobs, & Kiefe (2002), Diez-Roux et al. (1999), Diez-Roux et al. (1997); Kleinschmidt, Hills & Elliott, 1995; Lee & Cubbin, 2002; Patel et al., 2003; Reagan & Salsberry, 2005; Robert, 1998; Robert & Reither, 2004; Xue et al., 2005) and Europe (Davey Smith et al., 1998; Duncan et al., 1999; Ecob & Jones, 1998; Reijneveld, 1998; Van Lenthe & Mackenbach, 2002). It is easy to view these mixed results and query the role of residential neighborhoods in health (Diez-Roux (2003), Diez-Roux (2001); Kawachi & Berkman, 2003; Pickett & Pearl, 2001). But these results may also reflect an underlying variability within populations that are differentially affected by the contextual qualities of the residential neighborhood.

An individual's neighborhood of residence may be especially salient for those who are spatially segregated and socially isolated, as they tend to rely largely on the local environment for many aspects of their daily living (e.g., shopping, medical care). For others, the influence of the residential neighborhood may be less salient (Rankin & Quane, 2000) as modern telecommunications and transportation options allow an increasing number of social interactions to take place outside the residential neighborhood (Taub, Surgeon, Lindholm, Otti, & Bridges, 1977) and as a growing number of residents do not know their neighbors (Sampson, Raudenbush, & Earls, 1997; Wellman 1999).

Exposure to neighborhoods outside the residence may have both direct and indirect pathways in its association with health. Just as residential neighborhood SES has been associated with health, other neighborhoods where people spend their days may also have direct effects on health. Those same factors associated with SES (which are found in residential neighborhoods) in other neighborhoods outside of the residence may directly influence an individual's health outcome. But exposure to other neighborhoods may also affect self-rated health indirectly. Exposure to other neighborhoods implies that exposure to the residential neighborhood is reduced. Adjusting for the time spent in the residential neighborhood may adjust for individual variability of residential neighborhood exposure and “clarify” the true association of residential neighborhood on health. Exposure to multiple environments may also modify the influence of the residential neighborhood health benefits or risks on the individual.

Beginning the search for neighborhood contextual effects in residential areas where health outcomes have been shown to cluster makes inherent sense, but continued focus on only residential “context” may be limited. Geographic research on human activity-travel patterns, by incorporating both time and space in its models (Kwan, 2002), understands that the “action space/activity space” of individuals, or the geography of the individual's daily activity over time, is not limited to the residential neighborhood (Buliung & Kanaroglou, 2006; Friedrichs, Dangschat, Droth, & Kiehl, 1982; Gliebe & Koppelman, 2005; Law, 1999; Naess, 2006). Compared to administrative residential boundaries, this “activity space” may better represent an individual's “interaction space” (Guagliardo, 2004; Nemet & Bailey, 2000) or the environmental and social exposures that are associated in influencing health outcomes. But neither epidemiology nor geography studies have examined the contextual influence of an individual's “activity space” on health outcomes.

Using the Los Angeles Family and Neighborhood Study (L.A.FANS) database linked to US Census data, we use multilevel and clustered models to analyze the impact of non-residential neighborhood exposures on self-rated health as well as their effects on the association between residential neighborhood context upon health.

Section snippets

Sample

This study used data from the 2000–2001 L.A.FANS and the 2000 decennial US Census file. L.A.FANS is a panel study based on a stratified random sample of 65 neighborhoods (census tracts) that are representative of all neighborhoods and all households in Los Angeles County. Poor neighborhoods were over-sampled. In the 2000–2001 sample, an average of 41 households in each neighborhood were randomly selected and interviewed. A household survey asked adults and primary care givers about household

Results

The 3323 L.A.FANS respondents were predominantly young (mean age 39), female (69%), and Latino (57%). Nearly 70% of the total sample analyzed came from the two lowest SES neighborhood quartiles. Self-rated health followed a Gaussian distribution, but few rated their health as poor (4%). Table 1 presents respondent characteristics for the full sample.

Multilevel linear regression models, shown in Table 2, were used to estimate the direction of association among the individual socio-demographic

Discussion

Our study looked at the SES of neighborhoods where respondents worked, shopped, received medical care, went to worship and spent “other” time and found a salutary effect from being in a higher SES environment and detrimental health effects from being in a lower SES environment, whether the environment was the place of residence or the aggregated exposure to locations of their daily activities. There are many possibilities to explain these findings. The salutary effect of being in a higher SES

Limitations

This study has several limitations regarding the cross-sectional design, sampling strategy, the operationalization of non-residential neighborhood exposure and the treatment of missing values. The cross-sectional design of this study cannot prove social causation. Healthier people may choose to expose themselves to more advantaged neighborhoods, or healthier people may be more able to travel and to expose themselves to more advantaged neighborhoods.

Another limitation is the absence of data

Conclusion

Our study confirms and strengthens the notion that disadvantaged neighborhoods are associated with worse self-rated health and that those isolated there may be at the greatest risk. To understand contextual factors in health inequalities, consideration may need to be given to contextual factors in areas beyond the residential neighborhood.

Acknowledgments

We are grateful to Aimée Bower for her extremely capable work in programming, and Matthew Schonlau, Mitchell Wong, Phoenix Do, Susan Ettner, Marc Elliott, Richard A. Williams, Arleen Brown and Academic Technology Services at UCLA for statistical, theoretical and modeling advice. S. Inagami extends particular thanks to Steven Asch for his mentorship and support during her fellowship.

References (61)

  • J.D. Boardman et al.

    Neighborhood disadvantage, stress and drug use among adults

    Journal of Health and Social Behavior

    (2001)
  • L.N. Borrell et al.

    Neighborhood characteristics and mortality in the Atherosclerosis Risk in Communities Study

    International Journal of Epidemiology

    (2004)
  • R.N. Buliung et al.

    Urban form and household activity-travel behavior

    Growth and Change

    (2006)
  • K.A. Cagney et al.

    Exploring neighborhood-level variation in asthma and other respiratory diseases

    Journal of General Internal Medicine

    (2004)
  • D. Cohen et al.

    “Broken Windows” and the risk of gonorrhea

    American Journal of Public Health

    (2000)
  • D.A. Cohen et al.

    Neighborhood physical conditions and health

    American Journal of Public Health

    (2003)
  • G. Davey Smith et al.

    Individual social class, area-based deprivation, cardiovascular disease risk factors, and mortality: The Renfrew and Paisley study

    Journal of Epidemiology and Community Health

    (1998)
  • A. Diez-Roux

    Residential environments and cardiovascular risk

    Journal of Urban Health

    (2003)
  • A.V. Diez-Roux

    Investigating neighborhood and area effects on health

    American Journal of Public Health

    (2001)
  • A.V. Diez-Roux et al.

    Neighborhood characteristics and components of the insulin resistance syndrome in young adults

    Diabetes Care

    (2002)
  • A.V. Diez-Roux et al.

    Neighborhood of residence and incidence of coronary heart disease

    New England Journal of Medicine

    (2003)
  • A.V. Diez-Roux et al.

    Neighborhood differences in diet: The ARIC study

    Journal of Epidemiology and Community Health

    (1999)
  • A.V. Diez-Roux et al.

    Neighborhood environments and coronary heart disease: A multilevel analysis

    American Journal of Epidemiology

    (1997)
  • G.W. Evans

    Environmental stress and health

  • G.W. Evans et al.

    Socioeconomic status and health: The potential role of environmental risk exposures

    Annual Review of Public Health

    (2002)
  • Friedrichs, J., Dangschat, J., Droth, W., & Kiehl, K. (1982). Action spaces of urban residents. Presented at...
  • J. Gliebe et al.

    Modeling household activity-travel interactions as parallel constrained choices

    Transportation

    (2005)
  • M.F. Guagliardo

    Spatial accessibility of primary care: Concepts, methods and challenges

    International Journal of Health Geographics

    (2004)
  • E.L. Idler et al.

    Self-rated health and mortality: A review of twenty-seven community studies

    Journal of Health and Social Behavior

    (1997)
  • C. Jencks et al.

    The social consequences of growing up in a poor neighborhood

  • Cited by (158)

    View all citing articles on Scopus
    View full text