Keywords: Bayesian model, urban pollution, air pollution, respiratory symptoms, children, asthma, carbon dioxide, CO2, coughs, phlegm, dispersion modelling, hierarchical logistic regression models, SIDRIA–2, traffic emissions, carbon emissions, respiratory health, Florence, Italy, health risks
A Bayesian model for studying urban air pollution and respiratory symptoms in children
The association between traffic–related air pollution and long–term respiratory health problems has been extensively studied. In this work we evaluated the effect of traffic–related air pollution on respiratory symptoms in children living in Florence, Italy. Children were selected from different schools part of the Italian Studies on Respiratory Disorders in Children and the Environment 2. Exposure to traffic air pollution was assessed through a dispersion model and weighted by distance using four different criteria. A Bayesian hierarchical logistic regression model was specified to assess the impact of traffic air pollution on cough or phlegm and asthma. Potential confounders were included in the analysis. Familiarity of asthma and exposure to second–hand smoking showed the strongest positive association with respiratory symptoms. No evidence of increasing risk of asthma with urban air pollution was found, while some evidence of an association was observed for carbon dioxide and cough or phlegm.