|
Article Excerpt
Introduction
As students of health and health care, geographers have infused health research with a range of perspectives, approaches and techniques. An important contribution to the wider health community has been to shed light on the geographical patterning of disease, including the influence of place on health. The purpose of this paper is to lay the foundation for a long-term project concerned with the nature of places as evolving contexts of individual and population health.
A number of as yet disparate trends in social science health research lead us to consider the influence of neighbourhoods over health through time. First, the health inequalities literature features the role played by space and place in patterning and influencing health outcomes. Much of this work is at the neighbourhood scale, yet neighbourhoods have been used mainly to model health outcomes cross-sectionally, at single points in time. How do neighbourhoods themselves change in relative and absolute terms and how does their tracking present new research opportunities? Second, and related, is the emergent marriage of population and environmental health, two hitherto disparate literatures increasingly concerned with health and socio-economic inequalities. How do ambient health hazards and social stratification, separately and in combination, influence health? More generally, how do neighbourhood context and characteristics of residents interact to generate individual and community health outcomes? Focused on Vancouver, Canada, from 1976 to 2001, the objective is to develop a 'cohort' of neighbourhoods as the basis for a range of spatial epidemiologic study designs that incorporate explicitly the evolving spatial contexts of health.
Unlike other large cities, Vancouver's air quality record is very favourable: despite a 60 percent increase in total population and associated economic and transportation activities since 1976, all ambient criterion pollutants have steadily declined (GVRD 2003). It would seem Vancouver continues to earn its reputation as Canada's urban Eden (Wynn 1992). However, data from the regional air quality monitoring network--among the best the world over in terms of number and distribution of monitoring stations (Figure 1) and local studies underscore an emergent insight from air pollution epidemiology (Brauer et al. 2002; Henderson et al. 2004): that intra-urban variability in air pollution ranges as widely if not more so than between-city contrasts (Briggs et al. 1997; Brunekreef et al. 1997; Hoek et al. 2001; Brauer et al. 2003). Such variability within the region creates the potential for unequal exposures along neighbourhood socio-economic lines. How do these processes play out among Vancouver's neighbourhoods of affluence, despair and ethnocultural clustering?
[FIGURE 1 OMITTED]
This project is built retrospectively from 1976 to 2001 by developing a geodatabase of neighbourhoods in the Vancouver CMA. These neighbourhoods will be followed prospectively with each census period (2006 onwards) in the future (Figure 1). A commensurate set of 195 census tracts, or neighbourhoods, are constructed from varying numbers of tracts in each of the six censuses over the study period. Our approach is to explore the relationships between ambient health hazards and socio-economic status (SES) as defining features of neighbourhoods. We focus on air pollution and a range of markers of socioeconomic status to highlight the neighbourhood-level processes that can generate health outcomes among their residents, as individuals and communities, and point the way forward for health studies.
Population and Environmental HealthGaps and Opportunities
Health inequalities
The literature on health inequalities grows out of foundational studies that demonstrated a correlation between health status and social structure. Among the most important, the 'Whitehall studies' of the British civil service (Marmot et al. 1978, 1991) and the Black Report (DHSS 1980) popularized the correlation between occupational class and health, showing, for example, that mortality rates drop significantly with each step up the occupational classification (unskilled, semi-skilled, skilled manual and non-manual, managerial, professional) of the British civil service. Since this early work the health inequalities literature has mushroomed.
At an international scale health inequalities follow a health-and-development pattern such that more developed market economies exhibit better health indicators (e.g., longer life expectancy, lower infant mortality rates). But beyond a certain level of development, little is added to a population's health: life expectancy rises with wealth gains in poor countries but in developed economies wealth gains bring diminishing--almost flat--health benefits (Preston 1975; Rodgers 1979). Rather, it is social inequalities within developed countries that correlate with health. For example, Whitehall-1 (Marmot et al. 1978) found that the relative risk of mortality by coronary heart disease among clerical workers was 3.2 times that of administrative workers in the British civil service (for which occupational classes and health data constitute a robust cohort) over several decades up to the 1970s. Using aggregated data, rather than individuals, Kaplan et al. (1996) and Kennedy et al. (1996) were among the first to demonstrate the relationship between all-cause and cause-specific mortality and income inequality at the state level in the United States. Canadian research has also played a fundamental role in formalizing the population health perspective, particularly through the Canadian Institute for Advanced Research (Evans et al. 1994).
Growing out of this corpus of research is the central tenet of the health inequalities literature, namely that developed societies exhibit a SES-health gradient whereby population health status diminishes with rises in socio-economic status (Feinstein 1993). In the Whitehall studies, only 28 percent of the raised incidence of coronary heart disease (CHD) among clerical workers was explained by the traditional risk factors of cholesterol, smoking and blood pressure; the remainder was 'unexplained'. Further, life expectancy is not the only measure of a population's health: a range of diseases, including cardiovascular disease and cancer--the two leading causes of death in the developed world--underscore the SES-health gradient (Wilkinson 1996). (1)
Seemingly plausible explanations of the SES-health gradient have been offered, but refuted, in a number of powerful statements on population health inequalities. Individual health risk behaviours, health care, (2) genetic predisposition, level of economic development and material deprivation are not the principal reason why we see the SES-health gradient within developed countries (Rose 1992; Evans et al. 1994; Wilkinson 1996; Davey Smith 2003). Instead, some have argued that social structure is itself a 'fundamental cause' of a population's health (Link and Phelan 1995). This line of argumentation has spawned a number of responses and debate (Kaplan and Lynch 1997; Evans and Kantrowitz 2002) but agreement has settled around the influences of psychosocial stress and social cohesion (Evans et al. 1994; Wilkinson 1996). Seen as mediators in the SES-health relationship, one's social comparison with others, sense of social cohesion and social capital--all feeding into socially stratified psychosocial stress--produce measurable biological markers in human development and endocrine and immune system response to day-to-day life (Brunner 2000). What is important for our purpose is that the SES-health gradient is necessarily social, not individual.
Within this context, the geographical contribution to this literature has been to reveal the health inequalities between places within societies (Gatrell 2002; Curtis 2004 provides good overviews of the geographical patterning of disease and health care). Thus the Kaplan et al. (1996) and Kennedy et al. (1996) studies noted above demonstrated health and income inequality correlations across U.S. states. The Canadian east-to-west gradient in health status across provinces and metropolitan areas is well established (Gilmore 2004). At the local scale Ross et al. (2000) compared the relationship between age-adjusted ail-cause mortality and income inequality (percent income of poorest 50 percent of households) between Canadian and U.S. jurisdictions in 1990. In general, mortality and income inequality were both lower in Canada than in the United States and a significant relationship was found only for U.S. metropolitan areas. Wilkins et al. (2002) examined changes in mortality by income quintiles among Canadian metropolitan neighbourhoods (census tracts) from 1971 to 1996. Based on age-standardized mortality rates of major causes of death as well as life expectancy at birth, potential years of life lost (PYLL), and income-related excess PYLL before the age of 75 years, the disparity in disease between the richest and poorest neighbourhoods diminished for...
|
