You are here:
ROADWAYS AND CHILDREN'S RESPIRATORY HEALTH: LAND-USE REGRESSION VERSUS PROXIMITY MEASURES OF EXPOSURE ASSESSMENT IN AN EPIDEMIOLOGIC STUDY
Citation:
NEAS, L. M., M. M. JOHNSON, S. MUKERJEE, L. SMITH, AND C. STALLINGS. ROADWAYS AND CHILDREN'S RESPIRATORY HEALTH: LAND-USE REGRESSION VERSUS PROXIMITY MEASURES OF EXPOSURE ASSESSMENT IN AN EPIDEMIOLOGIC STUDY. Presented at International Conference on Environmental Epidemiology and Exposure (ISEE/ISEA), Paris, FRANCE, September 02 - 06, 2006.
Impact/Purpose:
research results
Description:
Introduction: Previous studies of the respiratory health impact of mobile source air pollutants on children have relied heavily on simple exposure metrics such as proximity to roadways and traffic density near the home or school. Few studies have conducted area-wide monitoring of indicator pollutants coupled with land-use regression modeling for exposure assessment. Based largely on proximity studies, the State of California (U.S.) has restricted the location of new schools within 500ft (168m) from a roadway averaging more than 50k vehicles per day (vpd) in rural and 100k-vpd in urban areas. We examined adverse effects of mobile source emissions among children living outside California's buffer area using both simple exposure metrics and a land-use regression modeling approach. Methods: The EI Paso Children's Health Study was a cross-sectional study of children enrolled in the fourth and fifth. grades of public elementary schools in EI Paso, Texas. Measurements of nitrogen dioxide were made at 22 elementary schools. Exposures to mobile source emissions, as indicated by ambient N02, were estimated at each child's residence using a land-use regression model. For this analysis, we restricted the cohort to 62% of the children (n=2546) who lived outside the Rio Grand River valley (above 1700m). None of these children lived within 200m of a roadway with ;;eOk-vpd: only 3.2% lived within 200m of a roadway with ;;;50k-vpd; however, almost half lived within 200m of a roadway with ~ Ok¬vpd. Results: As expected from a GIS-based model, estimated N02 concentration increased by 4.8, 3.2, and 2.6ppb per km increase in proximity to 10k, 50k, and 90k-vpd roadways, respectively (95%CI's:[ 2.7,6.8], [2.7, 3.7], and [2.4, 2.9], respectively). N02 concentration estimated through land-use regression modeling was positively associated with current asthma adjusting for demographic, socioeconomic, parental, and random school effects (OR=1.7 for 10ppb N02, 95%CI:[1.1, 2.6]). Proximity to roadways with ;;;50k-vpd or more was positively associated with current asthma (OR =1.1 per 1.5km, 95%CI:[1.0, 1.2]). There were also marginally significant associations between current asthma and proximity to roadways with ~ Okand ;;eOk-vpd (OR =1.5 for 1.5km, 95%CI:[0.97, 2.4] and OR =1.04 per 1.5km, 95%CI:[1.0, 1.1], respectively). Discussion and Conclusions: These results provide evidence that adverse health affects associated with traffic-related air pollution occur well beyond a narrow (200m) buffer area surrounding major roadways, potentially impacting children living several kilometers away from these roadways. This is an abstract of a proposed presentation and does not necessarily reflect EPA policy.