METHODS: This case-control study analyzed fixed-location and in-vehicle TRAP exposure during pregnancy of 175 subjects (56 cases, 119 controls), aged 0-18 years. In-vehicle TRAP was estimated using GIS based on commuting routes and measured and modeled on-road concentrations. Spatially interpolated monthly ambient pollutant concentrations (PM10, PM2.5, NOx, NO2, O3 and CO) were estimated for residential home locations. A modified Gaussian line source dispersion model (CALINE4) was employed to estimate residential TRAP (NO2, NOx, CO). Unconditional logistic regression was used to model odds of asthma as a function of exposure.
RESULTS: In-vehicle concentrations were around twice as high compared with residential exposures. We found no associations of asthma risk with either GIS-estimated or questionnaire-reported commute travel time, or modeled in-vehicle exposures. In univariate models, there were positive associations of increased asthma risk from exposure to both residential exposures to ambient air pollutants (except O3) and residential TRAP exposures during pregnancy. However, these associations were markedly confounded by demographic covariates, especially age.
CONCLUSIONS: A univariate finding was that in-utero exposure to residential TRAP was associated with risk for asthma development but not in-vehicle exposure. Nevertheless, in adjusted models we found no significant associations, possibly due to small sample size and limitations of the case-control design.