Predicting personal exposure to air pollutants in children living in a high risk area. A Universal Bayesian Kriging approach

摘要

Background: A core challenge in epidemiological analysis of the impact of air pollution exposure on health is assigning exposures to individuals subjects at risk.GIS-based (Geographical Information Systems) pollution mapping has become one of the main tools for exposure assessment to ambient pollutants, which interpolation techniques, such as Kriging, have helped to utilize routine monitoring data to estimate levels of ambient air pollutants at unmeasured locations. Aim:to estimate the individual exposure to gaseous air pollutants of asthmatic children living in the area of Milazzo-Valle del Mela (Sicily, Italy) by Universal Bayesian Kriging approach. Methods:Weekly measurements for sulphur dioxide (SO2) and nitrogen dioxide (N02)were obtained from 21 passive dosimeters located at each school yards of the study area (November 2007-April 2008). The residences of 113 asthmatic children were geo-referenced. A Universal Bayesian Kriging approach was performed to predict individual exposure levels at each residential address, using as covariates land use information, altitude, distance to main roads and population density. Results: A large geographical heterogeneity in air quality was recorded suggesting complex exposure patterns. With an effective sample size of, we obtained a predicted mean level of 25.78 (± 10.61) μg/m3 NO2 and 4.10 (± 2.71) μg/m3 SO2 at 1,682 children's residential addresses, with a normalized root squared mean of 28% and 25%, respectively. The spatio-temporal distribution of SO2 concentrations showed a point source effect with a plume consistent with prevalent winds, whereas NO2 patterns were more stable and reflected mostly diffuse traffic emissions. Conclusion: Universal Bayesian kriging may be useful to predict residential concentrations from monitoring data. Heterogeneity of the spatio-temporal distribution of pollutants should be considered in order to estimate the real exposure to them.