AIR-POLLUTION AND MORTALITY IN VALENCIA, SPAIN - A STUDY USING THE APHEA METHODOLOGY

Study objective - To assess the short term relationship between daily air pollution indicators (black smoke and sulphur dioxide (SO2)) and m ortality in Valencia. Design - This was an ecological study using time series data with application of Poisson regression. Daily variations in four selected outcome variables (total mortality, mortality in thos e over 70 years of age, and cardiovascular and respiratory mortality) were considered in relation to daily variations in air pollution level s for the period 1991-93. Setting - The city of Valencia, Spain. Main results - The mean daily mortality was 17.5, and the average daily lev els of air pollutants from the three monitoring stations included in t he study were, 67.7 mu g/m(3) for black smoke, and 39.9 mu g/m(3) for SO2. A significant positive association between black smoke and three of the four outcomes in the study was found. The estimated relative ri sk (RR) of dying corresponding to a 10 mu g/m(3) increase in mean dail y black smoke over the whole period was 1.009 (95% confidence interval (95% CI): 1.003, 1.015). For mortality in the group aged more than 70 years and for cardiovascular mortality, the RRs were 1.008 (95% CI: 1 .001, 1.016) and 1.012 (95% CI: 1.003, 1.022) respectively. The associ ation with SO, was less clear: it was only evident during the warm sea son. The estimated RRs in this case were 1.007 (95% CI: 0.999, 1.015) for total mortality, 1.009 (95% CI: 1.00, 1.21) for total mortality in those older than 70, and 1.012 (95% CI: 0.995, 1.026) for cardiovascu lar deaths. No significant association was found between mortality fro m respiratory diseases and either of the two pollutants. Conclusions - A positive relationship between air pollution and mortality was found in the short term, as has been shown in an important number of studie s carried out elsewhere. Although the current levels of particulate ai r pollution in Valencia are not very high, they could have an effect o n the number of premature deaths. Despite the fact that the associatio n is weak, it is important at the public health level both because of the numbers of population exposed and the possibility of establishing control measures.