Air Pollution Effects on Mortality - Sinergyc and Isolated Effects Using Time-Series and Case-Crossover Analysis

Abstract

Exposure to air pollution and weather conditions occurs simultaneously, therefore, in addition to the isolated effect, factors interactions and modification effects are key issue. The present study characterized the effect of air pollution, isolated and synergistic effects, on mortality, from secondary data. Three methods were applied to evaluate the association of average daily concentration of pollutants (PM10, NO2,O3) on mortality caused by cardiovascular and respiratory diseases in the city of Sao Paulo - Brazil, from 1998 to 2008 . We compared the estimates of relative risk produced in several case- crossover approaches, time-stratified bidirectional and matched by the confounding factor, i.e. average temperature, to the results of a traditional time-series analysis. The risk factors synergistic effect was evaluated by the graphical interpretation of response surfaces generated by bivariate models. No statistical differences were observed between the results from the case crossover and time-series analyses . The relative risk percent changes for cardiovascular mortality associated with an increase of 10µg/m3 in the 24-h average concentrations of PM10 and NO2 are, respectively, 0.85% (0.45-1.25) and 0.26% (0.04-0.48). The percentage increase in the risk of respiratory mortality was 1.60 (0.74-2.46) and 1.29% (0.46-2.12), respectively, for PM10 and O3. The case-crossover analyzes confirmed that the positive association parameterized for pollutants is unlikely to be caused by confounding by temperature in time-series analysis. The simultaneous exposure to several levels of environmental factors, such as temperature and air pollution, may represent conditions as harmful as the predicted to extreme concentrations, due to the combined effect.