Until the late 1990s, time series analysis was the main way we studied the impact of air pollution on human health. Typically, exposure data from a single, centrally located air monitoring station was used to examine the impact of pollutants on morbidity and mortality for all residents in a city. This was done to study acute events, such as spikes in air pollution, or to test long-term impacts of exposures on heart disease and lung disease. In the landmark Harvard Six Cities Study, every resident of a city studied was assigned the same exposure value. This allowed investigators to estimate pollution levels in the cities and compare differences in death rates (Figure 1).
In 1994, Glickman introduced spatial buffering as a proxy for air pollution exposure.12 Though simple by today’s standards, Glickman’s method assigned subjects to a particular exposure group if they lived within a buffer zone (often a circle) near a factory or road (Figure 2). Residents who lived outside the buffer zone were typically considered “unexposed.” This simple but insightful method allowed researchers to compare the impact of air pollution on different neighbourhoods. Within a decade the field rapidly diversified to include studies based on networks of monitoring stations, measurements of proximity to roadways, results of personal monitoring, and study methods incorporating spatial sampling and micrometeorology.3
Land Use Regression
Land use regression (LUR) is one of the most effective ways to estimate air pollution with greater geographic precision. LUR models are multivariate regression models that use the characteristics of a study area, such as traffic count, elevation and land cover, to enhance the quantification of pollutants. The result is a simple depiction of pollution concentrations for a geographic area. With land use regression, we can even assign pollutant exposures to specific dwellings in some neighbourhoods.
- 1. Ott WR. Human exposure assessment: the birth of a new science. J Expo Anal Environ Epidemiol. 1995;5(4):449-472.
- 2. Glickman T. Measuring environmental equity with geographical information systems. RRJ. 1994;116:17-20.
- 3. Hoek G, Beelen R, de Hoogh K, et al. A review of land-use regression models to assess spatial variation of outdoor air pollution. Atmos Environ. 2008;42(33):7561-7578.