Long-Term Exposure to Ambient Air Pollution and Cognitive Function in Older U.S. Adults: The Multi-Ethnic Study of Atherosclerosis and Air Pollution

摘要

Background: On-road vehicles are a major source of air pollutant emissions. As a result, spatial patterns of air pollution near roadways have been studied exhaustively, and many spatial models such as land use regression (LUR) identify traffic as a main source of pollutant heterogeneity. However, relatively little is known about spatial patterns resulting from other common urban sources. Methods: We deployed an aerosol mass spectrometer (AMS), a state-of-the-art instrument for quantifying particulate matter (PM) composition, in a mobile laboratory. Measurements were conducted in Pittsburgh, PA and Oakland, CA. AMS provides real-time measurement of PM composition between organic and inorganic (sulfate, nitrate, ammonium) fractions, as well as full aerosol mass spectra of organic PM. Factor analysis of organic aerosol mass spectra allows for quantification of the contributions of both fresh emissions (e.g., traffic and cooking emissions) and aged secondary PM. Results: Urban spatial variations in PM exposure are dominated by fresh emissions, primarily from traffic and food cooking. Concentrations vary temporally, with higher traffic contributions in the morning and cooking dominating at midday and evening. In Pittsburgh, locations with both high traffic and high restaurant density have up to 2 μg m-3 higher PM than the urban background. In Oakland, local traffic and cooking sources are more important than emissions from the large commercial port in driving pollutant spatial variations. Traffic also increases concentrations of secondary PM in downtown Oakland. The relationships between traffic- and cooking-emitted PM and typical land use variables is non-linear, which complicates the building of LUR models for source-specific PM. Conclusions: Source-resolved measurements of PM2.5 reveal spatial patterns that cannot be described by traffic sources alone. Cooking is also an important determinant of PM2.5 variations in US urban areas.