Short-Term Effects of Air Pollution on Childhood Respiratory Illness in Piraeus, Greece, 1987–1992: Nonparametric Stochastic Dynamic Analysis

The short-term effects of air pollution on childhood respiratory illness are analyzed for the city of Piraeus, Greece during the 1987–1992 period using a distinct, nonparametric, stochastic dynamical system approach. This approach accounts for the time continuity of the phenomenon, including contemporaneous, lagged, and unstructured cumulative effects. It also accounts for the presence of autocorrelation and nonstationary structures in the series and aims at revealing the potentially subtle relationships among the examined pollutants (smoke, CO, SO2, NO2, O3), meteorological factors (temperature, humidity), and hospital admissions. The approach is also capable of assessing the individual effects of any isolated pollutant, as well as the total effects of all measured pollutants. The study is based upon 4042 hospital admissions, and its results indicate statistically significant effects of each individual pollutant, as well as meteorological factor, on hospital admissions. It is interesting that these effects appear stronger for the second half of the examined period, despite declines in the observed mean concentrations for most pollutants. The total effects of the four pollutants (excluding O3) are characterized by multiple coherence measures of 18.37 and 21.66% in the first and second halves of the examined period, respectively, and are confirmed as statistically significant.