6
How Does Where People Live Affect Their Health?

Health has always been a fundamental social concern, but apprehension over health issues has escalated in recent years in the wake of extensive media coverage of disease outbreaks, the rapid spread of infectious diseases around the world, growing evidence of the health impacts of exposure to the by-products of industrialization, and anxieties about the availability and affordability of health care. Because environmental factors play a fundamental role in shaping human health, locational issues are of central importance to addressing health questions. A variety of place-based influences affect health, including physical circumstances (e.g. altitude, temperature regimes, and pollutants), social context (e.g., social networks, access to care, perception of risk behaviors), and economic conditions (e.g., quality of nutrition, access to health insurance). Because locational influences are myriad and constantly shifting, and because people themselves are moving around at unprecedented rates (Chapter 7), understanding the health impacts of where people live is one of the most challenging, yet important, contemporary geographical problems.

The influence of location on health is clear even at the global scale. The best way to reduce the worldwide burden of disease may be to provide individuals with ready access to clean water, adequate nutrition, and rudimentary sanitation, yet the availability of these “big three” basic needs differs greatly from place to place. People’s access to immunization is perhaps the next most important variable in the health picture, yet access to immunization often depends on social circumstances and the distribution of health care facilities.

Much has been learned in the past about geographical influences on health through mapping the spread of diseases, access to care, and the treatment and prevention of illness. Coming more fully to terms with the impacts of location on human health, however, requires documenting, modeling, and predicting human health outcomes at individual- to population-level scales, while accounting for

  • Human mobility (e.g., daily, weekly, seasonal, life course),

  • Socioeconomic circumstance (e.g., income status, age, education, gender),

  • Behavioral risk factors (e.g., smoking, drinking, drugs, diet),

  • Changing environment (e.g., climate change, industrial development, urban expansion),

  • Time course of disease (e.g., cancer latency, induction period),

  • Genetics (e.g., determinants of predisposition to disease).

Addressing some of the major health challenges of the 21st century requires developing increasingly sophisticated theories, methods, visualizations, and tools that can help account for the intersecting impacts of these six variables in different locations (Figure 6.1).

The global resurgence of vector-borne diseases such as malaria, dengue fever, yellow fever, and Lyme disease (Gubler, 1998) highlights the need to enhance understanding of the geographical dimensions of disease occurrence and spread. Malaria alone kills an



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 67
6 How Does Where People Live Affect Their Health? H ealth has always been a fundamental social Much has been learned in the past about geo- concern, but apprehension over health issues graphical influences on health through mapping the has escalated in recent years in the wake of spread of diseases, access to care, and the treatment and extensive media coverage of disease outbreaks, the prevention of illness. Coming more fully to terms with rapid spread of infectious diseases around the world, the impacts of location on human health, however, re- quires documenting, modeling, and predicting human growing evidence of the health impacts of exposure health outcomes at individual- to population-level to the by-products of industrialization, and anxieties scales, while accounting for about the availability and affordability of health care. Because environmental factors play a fundamental role • Human mobility (e.g., daily, weekly, seasonal, in shaping human health, locational issues are of central life course), importance to addressing health questions. A variety of • Socioeconomic circumstance (e.g., income status, place-based influences affect health, including physical age, education, gender), circumstances (e.g. altitude, temperature regimes, and • Behavioral risk factors (e.g., smoking, drinking, pollutants), social context (e.g., social networks, access drugs, diet), to care, perception of risk behaviors), and economic • Changing environment (e.g., climate change, conditions (e.g., quality of nutrition, access to health industrial development, urban expansion), insurance). Because locational influences are myriad • Time course of disease (e.g., cancer latency, and constantly shifting, and because people themselves induction period), are moving around at unprecedented rates (Chapter 7), • Genetics (e.g., determinants of predisposition to understanding the health impacts of where people live disease). is one of the most challenging, yet important, contem- porary geographical problems. Addressing some of the major health challenges of The influence of location on health is clear even at the 21st century requires developing increasingly sophis- the global scale. The best way to reduce the worldwide ticated theories, methods, visualizations, and tools that burden of disease may be to provide individuals with can help account for the intersecting impacts of these six ready access to clean water, adequate nutrition, and variables in different locations (Figure 6.1). rudimentary sanitation, yet the availability of these “big The global resurgence of vector-borne diseases three” basic needs differs greatly from place to place. such as malaria, dengue fever, yellow fever, and Lyme People’s access to immunization is perhaps the next disease (Gubler, 1998) highlights the need to enhance most important variable in the health picture, yet access understanding of the geographical dimensions of to immunization often depends on social circumstances disease occurrence and spread. Malaria alone kills an and the distribution of health care facilities. 6

OCR for page 67
6 UNDERSTANDING THE CHANGING PLANET FIGURE 6.1 A three­dimensional geographical visualization of the residential mobility of 1,000 cases of amyotrophic lateral sclerosis (ALS) in southeastern Finland, distinguishing movers (blue) from nonmovers (red). The vertical lines represent periods of stability and the blue lines link the origins and destinations of the moves. Sabel et al. (2003) identified a birthplace cluster of the disease in southeast Finland—the first example of a significant ALS cluster being identified worldwide. This part of Finland has suffered from significant industrial pollution, and heavy metals are present in the environment. To determine whether these heavy metals may be implicated in the etiology of ALS, Sabel et al. (2009) used the detailed migration histories of the cases and controls to explore differences between movers and nonmovers. The results demonstrated that moving away from the area seems to be protective, meaning that the environ­ ment has a role to play in the disease etiology. SOURCE: Unpublished figure courtesy of Paul Boyle. role oF The geograPhical scieNces estimated 700,000 to 2.7 million people each year (Patz and Olson, 2006). The increasing movement of people Traditionally, epidemiologists allocate risk for specific and products is raising the specter of the spread of one diseases such as cancer to specific causes, including of the most potent malaria vectors, Anopheles gambiae, willingness to participate in high-risk behaviors (e.g., from Africa to South America and Southeast Asia. smoking), nutritional status, age, genetic predisposition, Such an event occurred in the 1930s when A. gambiae and gene–environment interactions. Studies are usually was accidentally introduced to Brazil. A vigorous cam- based on experiments (e.g., case-control studies, cohort paign to identify breeding areas and eradicate larvae studies) aimed at evaluating whether a specific factor is was required during the 1930s and 1940s to avert a near associated with an increased risk of disease. Although disaster (Parmakelis et al., 2008). More sophisticated useful, this approach often fails to take into account analyses of locational influences on malaria are vital to the range of locational influences that affect diseases or several current international initiatives aimed at halting the temporal and spatial complexities that arise from the spread of malaria, including the Multilateral Initia- disease latency and individual mobility through the tive on Malaria, which is part of the United Nations life course. The geographical sciences have a role to Children’s Fund, United Nations Development Pro- play in addressing such matters. Building from an early gramme, World Bank, and World Health Organiza- focus on disease patterns, the geographical sciences are tion Special Programme for Research and Training in devoting attention to the development of models and Tropical Diseases.1 visualizations that provide insight into space-time in- fluences on health and disease ( Jacquez et al., 2005a,b; Meliker et al., 2005). These models and visualizations 1See www.mimalaria.org (accessed January 20, 2010).

OCR for page 67
6 HEALTH help geographical scientists account for the impacts of Geographical scientists have found dramatic impacts of exposure to potentially hazardous substances in differ- differentially located health care facilities, at scales that ent places (Meliker and Jacquez, 2007). Geographical range from the global to the local (e.g., O’Meara et al., scientists are also exploring ways of using remote sens- 2009), and they have developed a range of techniques ing (satellite data and hyperspectral imagery) in the for optimizing the location of facilities to achieve desir- study of disease location and spread (Goovaerts et al., able goals (e.g., Rushton, 2003). 2005; Avruskin et al., 2008). Much health research makes use of statistical tech- The geographical sciences add at least four dimen- niques to make general conclusions from samples of sions to the investigation of disease incidence and laboratory animals or patients. When these experiments response, and of conditions such as drug addiction are controlled, by choosing random samples and giv- (Thomas et al., 2008). First, they provide a series of ing them identical treatments, the conditions conform techniques suited to the detection of patterns and precisely to the assumptions made in standard statistical anomalies in the incidence of disease, and to their inter- tests. In such cases significance levels can be determined, pretation in terms of disease-spreading and disease- and results generalized to the populations from which causing mechanisms. From the earliest successes the samples were randomly drawn. However, in research achieved by John Snow in unraveling the etiology of focused on actual fine-scale geographical patterns rather cholera ( Johnson, 2006) to modern techniques for than controlled conditions, these assumptions are rarely detecting statistical significance in apparent clusters valid (O’Sullivan and Unwin, 2003). Analysis at the (O’Sullivan and Unwin, 2003), a vast amount of pure national level often uses health statistics aggregated to and applied research has gone into the examination the county or even state level—units of analysis that of health and disease patterns, and their connection have their origins in previous centuries, vary enormously to other circumstances and processes in the human– in size, and average or smooth out much of the actual environment system (Cromley and McLafferty, 2002; variation that occurs at smaller scales. These statistical Waller and Gotway, 2004; Koch, 2005). Second, issues provide examples of the substantial problems con- geographical science tools allow researchers to ask fronting health researchers that the technical arsenal of questions about the factors present in an area that cor- the geographical sciences can help resolve. The follow- relate with disease outcomes, whether as causes or as ing questions illustrate the types of research that allow statistical correlates. Geographic information systems the geographical sciences to contribute to the develop- (GIS) have proved valuable in this regard because they ment of a multidisciplinary synthesis, which is needed support a range of multivariate analyses and can be to tackle major health issues in the years to come. used to analyze the impacts of factors at different scales. Third, geographical scientists have developed spatially research suBQuesTioNs explicit models of disease spread that incorporate con- cepts such as distance and connectivity directly into the how do diseases respond to changes in ecosystems models, often in the form of cellular automata2 (Schiff, and climate? 2008) or as agent-based models (Maguire et al., 2005). Finally, the concern of the geographical sciences with Spatial distributions of diseases are seldom uniform, location promotes consideration of access to health and understanding the reasons for their heterogeneity care. When facing conditions such as heart attacks, the can lead to valuable insights. As our ability to conduct distance between the patient and the emergency room spatial and temporal analysis has improved, so has our and the ability to get to the emergency room quickly ability to characterize both sides of the disease equa- can literally be the difference between life and death. tion—the genetics of diseases themselves and human risk factors. Advances in spatial and temporal analysis are also facilitating efforts to respond to changes in the 2Cellular automata models attempt to simulate processes oper- human–environment system. ating across geographical space by dividing the space into a set of On the genetic side, new insights into how and why normally rectangular cells and applying a series of rules to those cells to approximate changes in each cell’s state through time, in response diseases emerge and sometimes lead to pandemics such to the state of neighboring cells as well as to external factors.

OCR for page 67
0 UNDERSTANDING THE CHANGING PLANET as HIV and influenza can be achieved by coupling the organism). A recent National Research Council report study of molecular and spatial analysis. One example of states that “nearly half of the world’s population is this type of inquiry is a study on H5N1 avian influenza infected with at least one type of vector-borne patho- that analyzes genetic change using a spatial approach gen” (NRC, 2008c). With climate change potentially (Figure 6.2). As part of the study, a spatiotemporal expanding the spatial reach of vector-borne diseases influenza genotype database was built and integrated by increasing flooding, altering precipitation patterns, with ecosystems factors using GIS. The diffusion of and raising temperatures, there is a need for research different H5N1 genotypes and mutations that are re- that explores the link between climate change and such lated to virulence is being tracked, and the impacts of diseases. human–environment ecosystem factors (e.g., human A robust earth observing system that monitors key population distributions, farming systems, land use, climate variables is critical to predicting future disease climate) on H5N1 viral evolution are being measured outbreaks. Anyamba et al. (2009) propose using satel- using a spatial analytical approach called ecological lite measurements to examine the linkage between niche modeling. Similar efforts for other viruses can climate variability and the outbreak of Rift Valley help identify the human–environment ecosystems in fever in the Horn of Africa. Their work examines the which viruses change. relationship between outbreaks of the disease, which Climate change can have substantial impacts in the affect animals and humans, and the El Niño/Southern area of health, affecting, for example, the emergence Oscillation. They conclude that satellite observations and resurgence of vector-borne diseases (those diseases can be used to map climate changes, which in turn such as malaria or West Nile virus that are transmitted can be used to predict the location of outbreaks. Like to humans, animals, or plants via an insect or other Rift Valley fever, outbreaks of malaria, yellow fever, or B A FIGURE 6.2 (A) Geographical distribution of the genotypes identified for 142 H5N1 viruses isolated in Vietnam from 2003 to 2007, with the viral genotype of each H5N1 isolate mapped chronologically to show the time of genotype isolation in different regions of Vietnam. (B) Emerging H5N1 genotypes from introduction and reassortment in Vietnam from 2001 to 2007, showing the genetic change of each H5N1 virus segment using a different color corresponding to its precursor virus—a sort of family tree for the virus. SOURCE: Michael Emch, University of North Carolina, Chapel Hill Geography. Used with permission.

OCR for page 67
 HEALTH dengue fever are not just public health events; they are Future research is needed on the impact of different environmental, economic, and ecological as well. The climate change scenarios on melanoma incidence using geographical approach adopted in the Rift Valley fever the time-dependent locations of individuals coupled study could advance understanding of the relationship with models of climate change and of melanoma risk. between climate change and vector-borne disease out- breaks more broadly. how do disparities in geographical access to health Climate change could also affect the risk to humans care affect health and well-being? of noninfectious diseases such as melanoma. Melanoma The locations, number, and quality of health care is known to be caused by exposure to ultraviolet (UV) providers differ from place to place, and services may radiation. Ethnicity is a risk factor (epidermal melanin not be available in places where they are most needed. is protective), as is location (altitude, latitude), oc- Access to health care is not simply a matter of measur- cupation (indoor, outdoor), and individual behaviors ing distance to health facilities; access is affected by (e.g., outdoor lifestyle). Climate models predict highly socioeconomic status, cultural and social norms, and spatially heterogeneous changes in UV radiation as a function of atmospheric CO2 doubling (Figure 6.3). transportation networks. In a study of access to gov- FIGURE 6.3 Simulated changes in July incident solar radiation over the Pacific Northwest expressed as the difference between doubled and present atmospheric CO2 concentrations (2×­1×). SWI: incoming solar radiation (W/m2). The model was configured with a horizontal grid spacing of 15 km and 23 layers in the vertical. Solid black lines indicate U.S. state borders and elevation. SOURCE: Courtesy of Steve Hostetler and Justin Holman.

OCR for page 67
 UNDERSTANDING THE CHANGING PLANET ernment health services in Kenya, for example, Noor et al. (2006) show that a transportation model adjusted for actual use patterns and competition between health facilities provided a much better indicator of access to health care than did a model focused solely on distance. Their example illustrates that understanding the loca- tional circumstances that create inequalities in access to health care is a critical piece of the health picture. Studies analyzing geographical inequalities in health care access should build on increasingly detailed work on this topic. For example, Luo and Wang (2003) have developed an index of spatial accessibility as a way of analyzing the local supply of primary health services in relation to local demand. Their method was used to analyze spatial access to primary care in Illinois and its relationship to late-stage cancer (Wang et al., 2008). Not surprisingly, in Illinois, as in much of the United States, rural areas are characterized by lower levels of spatial access to primary care than urban areas (Figure 6.4). But why are some rural communities much more disadvantaged than others? Geographical scientists such as Cutchin (1997) have taken up this question, looking at the specific attributes of rural places that tend to attract primary-care physicians and encourage them to stay. Additional work in this vein can provide policy makers and community planners FIGURE 6.4 Spatial access to primary care in Illinois by ZIP with much needed insight into ways of addressing the code based on the 2­Stage Floating Catchment Area method. problem of rural health care access. Light tones indicate poor spatial access to primary care. The Another example of the promise of spatially explicit map was created as part of a study examining the combined effects of spatial access and socioeconomic factors on breast studies of health care access comes from ongoing cancer diagnosis. The results showed that poor geographical research on whether cancer patients living in commu- access to primary health care had significantly increased the nities with poor access to primary health care have a risk of late diagnosis. The study also showed that spatial access to primary health care was more important than similar access higher-than-average risk of late-stage diagnosis (cancer to mammography. SOURCE: Wang et al. (2008). diagnosed after it has spread to distant tissues or organs; Figure 6.5). The researchers conducting the study under- took multilevel analyses of the associations between late-stage risk, individual demographic variables, and contextual variables, describing socioeconomic charac- contexts (Rushton et al., 2004). McLaffery and Wang teristics of places and their spatial access to primary care also identify poor health outcomes in a dense urban set- (McLafferty and Wang, 2009). Their research showed ting where spatial access to care is quite high, according that, in Chicago, the high risk of late diagnosis among to GIS-based measures. cancer patients largely reflects the high concentration of These findings highlight the need for research vulnerable people living in economically disadvantaged investigating the connections between population vul- neighborhoods. Outside Chicago, poor spatial access nerability, place vulnerability (Chapter 3), and access to to primary care significantly heightens the risk of late health care. Among the most important topics to inves- diagnosis among breast and colorectal patients, sup- tigate are effects of time-space constraints on the use porting the findings of geographical scientists in other of primary health care by low-income urban residents

OCR for page 67
 HEALTH It is well established that spatial variability is a key factor in understanding HIV/AIDS; geographi- cal concentrations of HIV/AIDS are found even in many countries with very low HIV prevalence. In the Democratic Republic of the Congo, for example, Messina et al. (In review; Figure 6.6) documented a high degree of spatial variability in HIV/AIDS cases and found that the distribution of the disease was in- fluenced by such social factors as sexual practices, socio- economic status, and access to transportation. Future work in this vein could describe the spatial distribution of genetic subtypes of the HIV/AIDS virus, which in turn could facilitate the effort to understand not only FIGURE 6.5 Variation in risk of late­stage cancer diagnosis where imported cases originate, but where and how for the four major cancers in Illinois. The figure shows that risk is highest for cancer patients living in Chicago. The odds of late those cases evolve from earlier strains. Work on the diagnosis decrease away from Chicago, reaching their lowest impacts of socioeconomic status on the distribution of levels among patients living in large towns (defined here as towns with populations ranging from 10,000 to 50,000 and not located in metropolitan areas). This figure does not control for age or race, but the study looked at the impacts of both and found that the rural­urban gradient remained consistent for breast, colorectal, and lung cancers, but disappeared for prostate cancer. SOURCE: McLafferty and Wang (2009). and geographical variations in the quality of health care. There is also much to be gained from comparative analyses of the spatial dimensions of health care access in different parts of the world and for different types of health challenges (e.g., heart disease, HIV/AIDS, vector-borne diseases). how are social factors affecting the spread of diseases, such as hiV/aids in sub-saharan africa? HIV/AIDS is the leading cause of death in Sub- Saharan Africa, which remains the epicenter of the epidemic, accounting for 67 percent of all people living with HIV worldwide and 72 percent of all AIDS- FIGURE 6.6 HIV distribution in the Democratic Republic of related deaths (UNAIDS, 2008). In southern Africa, the Congo, showing differential patterns for men and women. A geographically based, population­representative sampling the worst-affected region, national adult prevalence scheme allowed Michael Emch and his colleagues to construct exceeds 15 percent in seven countries. Women account this visualization, which was then used to examine the impacts for nearly 60 percent of HIV infections in Sub-Saharan of social factors on the distribution of HIV/AIDS. Not surpris­ ingly, the study found that social factors are the most important Africa, and young women represent 67 percent of all drivers of the distribution of the disease. SOURCE: Michael new cases of HIV among people ages 15-24 living in Emch, University of North Carolina, Chapel Hill Geography. developing countries (UNAIDS, 2008). Used with permission.

OCR for page 67
 UNDERSTANDING THE CHANGING PLANET cholera in Bangladesh confirms the value of this type and living as internally displaced persons or refugees, of approach (Emch et al., 2010). survival compels many individuals to engage in trans- There is a long tradition of work on this subject in actional sex with other refugees or the host popula- the geographical sciences (Gould, 1993), but the range tion (Spiegel and Nankoe, 2004). The volatile mix of of contextual influences needs to be expanded. It is refugees, soldiers, desperate women with children, and becoming clear that more attention needs to be paid to the chaos associated with war facilitates HIV/AIDS the influence of attitudes toward sex, female empower- spread, but how this plays out in different places is not ment, and economic inequities on the spread of diseases well understood. (Kalipeni et al., 2007). Additional work focused on de- The efforts of multilateral organizations, such as veloping and analyzing data on the spatial distributions UNAIDS, have focused on assisting governments to of HIV/AIDS and relevant social circumstances can deal with the challenge of HIV/AIDS rather than improve understanding of the range of influences on evaluating how the apparatus of government—whether the vulnerability of different people in different places democratic, autocratic, or chaotic—affects HIV/AIDS (Kalipeni et al., 2007, 2008). spread. Yet government leadership, particularly good Globalization and migration are also changing governance, expressed by effective administration of sexual practices in Africa (Ampofo, 2001; Spronk, international aid for HIV/AIDS control, and public 2005). As a result of globalization, the media, including health programs including antiretroviral access, in- cinema, television, and the Internet, provide Africans fluence the success or failure of HIV/AIDS control with new sexual imagery, and undermine traditional efforts, as do policies that address poverty and social boundaries of sexuality and sexual expression. Further inequalities. Including such matters in a broadened effort to look at contextual influences on the geographi- research is needed on where these changes are having cal spread of HIV/AIDS can enhance understanding the greatest impacts, and how they are affecting the of how and why disease diffusion patterns are different spread of HIV/AIDS. Work in this vein could also from place to place. provide a more detailed understanding of the role of migrant labor on the diffusion of HIV/AIDS (Bassett a nd Mhloyi, 1991; Jochelson et al., 1991; Hunt, summarY 1996). Governance is another social factor affecting the The geographical sciences have a role to play in ad- spread of HIV/AIDS in Africa and elsewhere. Rela- vancing understanding of spatial variations in the tively little attention has been paid to the effects of spread of disease, access to care, and the treatment and civil wars and political instability on HIV/AIDS. As prevention of illness. The tools and approaches of the victims of armed conflict and as refugees (or internally geographical sciences can provide insights into loca- displaced persons), women and children are particularly tional factors affecting health, and they can illuminate vulnerable to sexual violence and abuse. Warring fac- important contextual influences on the diffusion of tions often use rape as an expression of violence and diseases. A major initiative to build upon and expand revenge against their enemies (Kalipeni and Oppong, work in the geographical sciences on health matters 1998). In refugee camps, women and girls remain sub- would benefit efforts to combat disease and promote jects of sexual coercion; separated from their families human well-being.