UC Berkeley

To investigate the extent to which traffic exposure affects the lung function of children with asthma and how local neighborhood factors may modify this relation, a merge of epidemiologic, environmental health and geographic methods is necessary. People and places are linked; therefore, it is necessary to consider place-effects on health as well as environmental exposures. The Fresno Asthmatic Children's Environment Study (FACES) is a longitudinal cohort study of children with asthma in Fresno, California that followed participants from 2000-2008 to explore short-term and long-term effects of ambient air pollution on lung function (as measured by spirometry, wheeze, and asthma symptoms). With publicly available data on traffic counts in Fresno, CA from 2000-2008, I built a spatial model of traffic exposure that varies both temporally and spatially for the FACES cohort. To capture and quantify neighborhood characteristics, I constructed individual neighborhoods based on global positioning software (GPS) data and walking distances around participant homes. To evaluate neighborhood deprivation, I collected geographic information system (GIS) data on parks, grocery stores, bus stops, etc. from publicly available sources and created an index based on Item Response Theory. To assess the marginal risk difference of lung function among children with asthma exposed to high levels of traffic pollution and those exposed to lower levels of traffic pollution (as measured by traffic density), I apply semi-parametric causal inference methods and use Targeted Maximum Likelihood Estimation (TMLE). More FACES participants who live in high deprivation neighborhoods are also farther away from high traffic areas. Neighborhood deprivation, as defined by a combination of GIS variables in this study, does not track well with US Census poverty. The marginal change in lung function from exposure to high neighborhood traffic to lower neighborhood traffic, without stratification for neighborhood deprivation, is -0.233 (95% CI -0.338, -0.129). The results can be interpreted as--the average decrease of FEV1 is 0.233 L, or there is a 12% reduction in lung function. Either neighborhood deprivation does not modify the effect of traffic on lung function or there is not enough data to evaluate this type of effect modification. The findings indicate that neighborhood exposure to traffic adversely affects lung function among the FACES cohort of children with asthma.