Abstract and Introduction
Abstract
Background: Living near traffic has been associated with asthma and other respiratory symptoms. Most studies, however, have been conducted in areas with high background levels of ambient air pollution, making it challenging to isolate an independent effect of traffic. Additionally, most investigations have used surrogates of exposure, and few have measured traffic pollutants directly as part of the study.
Objective: We conducted a cross-sectional study of current asthma and other respiratory symptoms in children (n = 1,080) living at varying distances from high-traffic roads in the San Francisco Bay Area, California, a highly urbanized region characterized by good regional air quality due to coastal breezes.
Methods: We obtained health information and home environmental factors by parental questionnaire. We assessed exposure with several measures of residential proximity to traffic calculated using geographic information systems, including traffic within a given radius and distance to major roads. We also measured traffic-related pollutants (nitrogen oxides and nitrogen dioxide) for a subset of households to determine how well traffic metrics correlated with measured traffic pollutants.
Results: Using multivariate logistic regression analyses, we found associations between current asthma and residential proximity to traffic. For several traffic metrics, children whose residences were in the highest quintile of exposure had approximately twice the adjusted odds of current asthma (i.e., asthma episode in the preceeding 12 months) compared with children whose residences were within the lowest quintile. The highest risks were among those living within 75 m of a freeway/highway. Most traffic metrics correlated moderately well with actual pollutant measurements.
Conclusion: Our findings provide evidence that even in an area with good regional air quality, proximity to traffic is associated with adverse respiratory health effects in children.
Introduction
Epidemiologic studies have linked proximity to busy roads with adverse health outcomes, including respiratory symptoms, asthma, adverse birth outcomes, and cardiopulmonary mortality (Brunekreef et al. 1997; Hoek et al. 2002; Wilhelm and Ritz 2003). Methods for estimating exposures to traffic pollutants have included neighborhood- or school-based estimates of traffic (Brunekreef et al. 1997; Kim et al. 2004), distance to freeways or busy roads (Gauderman et al. 2005), presence of a busy road within a given buffer (Venn et al. 2001), and traffic density within a given radius (English et al. 1999; Wilhelm and Ritz 2003). More recent studies have used geographic information systems (GIS) to estimate traffic exposure metrics. However, few have evaluated these GIS-based traffic metrics against measured traffic-related pollutants (Brauer et al. 2007; Gauderman et al. 2005; Hoek et al. 2002; Nicolai et al. 2003). Additionally, many of these studies were conducted in areas with moderate or high levels of regional air pollution.
We conducted the East Bay Children's Respiratory Health Study (EBCRHS) in the San Francisco Bay Area, California, a highly urbanized region of the United States where traffic is the major source of air pollution. This region ranks among the top four metropolitan areas with the worst traffic congestion in the United States (Schrank and Lomax 2005). However, the area experiences relatively good regional air quality due to onshore breezes. Thus, in contrast to most major metropolitan areas in the United States, there are only occasional exceedances of the federal ozone or fine particulate matter [particles ≤2.5 µm in diameter (PM2.5)] 24-hr standard. This allowed us to examine the impacts of local variations in traffic in the absence of significant levels of background ambient pollution.
In the first phase of this study, we found modest but statistically significant associations between measured traffic pollutants and recent episodes of asthma and bronchitis. In that analysis, we measured traffic-related pollutants at schools as indicators of neighborhood air pollution levels, which we used to estimate children's exposure to traffic emissions (Kim et al. 2004).
In this analysis, we sought to refine exposure estimates using GIS-derived traffic measures at the children's residences and to evaluate associations between residential proximity to traffic and respiratory health outcomes for the study population. We then evaluated whether traffic pollutants measured at the schools were independently associated with the health outcomes. We also evaluated the correlation of GIS-derived traffic proximity metrics and vehicular emissions for a subset of households using measurements of traffic-related pollutants (total nitrogen oxides and nitrogen dioxide).