Social and environmental ‘determinants’ of health

Since joining the collaborative research environment of the IDCE Department at Clark University in 2006, I have broadened my research horizons to include GIS applications for public health and environmental justice.  I have pursued my research in a series of funded and unfunded research projects, collaborative projects with local government and community organizations, and published papers.  I apply cutting-edge techniques of spatial analysis/statistics – such as hot spots analysis and geographically weighted regression – to areas where they have not been used before, such as infectious diseases, heat-related mortality, and noise pollution. Two main strands of inquiry – methods development and application – are the core of my most recent research activities in the area of human health.  The National Institutes of Health (NIH)-funded National Children’s Study, in which Tim Downs and I have been involved since 2007, is the largest and longest-term study ever conducted to gauge how environmental, social and genetic factors impact the health, wellbeing, development and quality of life of a sample of 100,000 children in the U.S. For this study we developed indices of hazard and adaptability for each block-group and each town, then used them to stratify Worcester County, resulting in a spatial sampling frame that captured a full spectrum of vulnerability conditions (Downs et al., 2010).   By using health-relevant indicators to stratify, the approach may increase the ability to detect associations among health outcomes and social/environmental ‘determinants’ down the line. Currently we are adapting the same sampling approach to stratifying the patient catchment areas of pre-natal care providers, part of testing a provider-based recruitment model.

One of the most pressing issues in public health is to accurately characterize the spatial patterns and extents of various health outcomes, as well as their links to socio-economic and environmental ‘determinants’. This is a challenging undertaking because of the scarcity of fine-scale health outcomes data and the inherent difficulty in inferring causality when multiple factors are involved. One of the themes of my current research is to address the manner in which spatial analytical methods and GIS can be used in this context.  In one project we used areal interpolation and hot spots analysis to study the relationship between diesel emissions along major transportation corridors and the incidence of asthma and lung cancer in Massachusetts (McEntee and Ogneva-Himmelberger, 2008). Our results demonstrated that the towns with high levels of diesel emissions and high incidence of asthma corresponded to environmental justice areas. The work helped drive and support policy advocacy to curb emissions.  Our methodology can be readily adopted by other states because we used nationwide, publicly available data.

GIS technology and spatial statistics are indispensable tools for bringing a geographical perspective to studying the health impacts of global climate change. Because the anticipated effects of global warming include increases in the duration and intensity of “heat waves” it is important to analyze spatial variation in recent heat events and their impact on human health using spatial and time-series analysis. In a recent study we used geostatistics and linear regression to study the reasons for the spatial variability of heat-related mortality in Massachusetts, the first study of its kind, analyzing the spatial distribution of heat-related mortality at the municipal scale (Hattis, Ogneva-Himmelberger and Ratick, 2011).  We found that the percent African-American and percent elderly are both positively associated with increased mortality on hot days, but the degree of urbanization was not significant. With more temperature increases expected and the need to determine factors that affect heat-related mortality, we provided a new approach and preliminary results to facilitate similar assessments in other parts of the world.

Another major consequence of climate change is that increased ambient temperature will affect the range of vector-borne diseases, and in many ways exacerbate their health impacts. Therefore analyzing baseline spatial patterns of these diseases is essential for developing effective, evidence-based interventions. In a recent study we used hot spots analysis to explore spatio-temporal patterns of West Nile Virus in the U.S. (Carnes and Ogneva-Himmelberger, 2011).This paper is the first to analyze spatio-temporal dynamics of the WNV nationwide at the county scale. Using publicly available data, we identified counties with persistently high incidence of the disease and highlighted the need for fine-scale studies. The ability to accurately pinpoint high prevalence areas within these counties is essential to the development of an effective and efficient mitigation campaign.  Our research provided this critical information by identifying the counties where high disease incidence persisted over time.

The applications of key spatial analytical methods lead to improved understanding of social and environmental phenomena by professionals and researchers from various fields, including health practitioners. Since 2008, I have been involved in a research project with UMASS Medical School (in collaboration with Dr. Warren Ferguson) to improve the integration of family medicine and community health.  One of the goals of the project is to introduce GIS as a valuable tool into graduate medical education. For this project, I created a series of maps depicting the spatial distribution of patients for four community health centers in Central Massachusetts. These maps showed spatial patterns of all pediatric and adult patients, along with locations of health and human services agencies, mobile clinics, grocery stores, gyms, schools, parks, and other health-relevant resources. Separate maps were created for patients with asthma, hypertension, diabetes, and coronary heart disease. Using these maps, medical residents (who are practicing family medicine in these community health centers) can see where their patients live, and what access they have to resources. These maps will allow doctors to make specific, targeted recommendations to individual patients about where to go shopping for fresh vegetables, and where to go for a walk or to exercise, etc. Comparing density of patients with the location of health resources also enables practitioners to analyze the accessibility of care. A subset of these maps is available online at http://www.umassmed.edu/hahnemanndata.aspx .