Living Close to a Major Roadway, Particulate Matter Exposure, and Adiposity

In this issue of Obesity, Li et al. (1) present associations between adiposity measures and estimates of air pollution exposure in the Framingham Offspring cohort. The analysis is unique in two aspects: (1) most previous association studies for air pollutants have focused on short-term health outcomes such as acute cardiovascular events, whereas this study explores associations with adiposity measures (body mass index and subcutaneous and visceral adipose tissue) which would presumably develop over longer periods of exposure and (2) a comparison of two geo-spatial methods for estimating exposure to particulate matter with diameter less than 2.5 microns (PM_2.5) and other co-pollutants is presented (proximity to major roadways versus use of satellite-derived data).

Previously, research on the health effects of air pollution relied on sparsely placed ground-level monitors for exposure estimates. To improve spatial resolution of PM_2.5 exposure estimation, satellite-derived aerosol optical depth (AOD) has been employed (2,3). However because AOD measures light attenuation over a vertical column of the atmosphere, it does not necessarily reflect PM_2.5 near the surface of the earth, where people are exposed, and vertical distribution of aerosols varies substantially across landscapes and time (2,3). Li et al.’s method for estimating exposure to PM_2.5 using AOD improves upon previous attempts by incorporating additional predictors (such as elevation, population density, meteorological parameters, vegetation index) and increasing spatial resolution to 1 km² via use of a correction algorithm (4).

Interestingly, Li et al. found that the more spatially resolved satellite derived estimates of PM_2.5 exposure were not associated with adiposity measures, whereas proximity to major roadways was associated with adiposity measures. This is an important finding since it suggests other factors associated with living in close proximity to a major roadway are likely contributing to the association. Among several potential factors related to living in close proximity to a major roadway, exposure to noise and light may be a useful starting place since other studies have suggested effects on adiposity (5,6).

Teasing apart relationships between exposure to multiple cooccurring pollutants and socioeconomic position, and ultimately their independent and potentially interacting contributions to health outcomes, is a fundamental challenge in environmental epidemiology. As Li et al. note, residual confounding cannot be ruled out in this study. Collecting longitudinal data and more sophisticated methods for quantifying socioeconomic position, beyond level of education and census-tract-level income or home values, may be an important next step to determine the relationship between co-exposures, socioeconomic position, and health outcomes. For example, in addition to assignment to a satellite-derived grid cell for estimation of PM_2.5 exposure at the participants’ home, using the participants’ address for determining the home value or other characteristics may be one way to incorporate additional household-level factors.