Air pollution and mortality in Madrid, Spain: a time-series analysis

To assess the relationship, if any, between air pollutant (sulfur dioxide and total suspended particulate) levels and mortality in the city of Madrid during the period 1986–1992, controlling for weather, season, and influenza epidemics. Methods: Daily death counts were obtained from the Regional Mortality Registry. Pollution data were supplied by the Municipal Monitoring Network. Time-series analysis methodology was used to assess the link between non-accidental as well as circulatory- and respiratory-disease mortality, on the one hand, and mean daily concentrations of SO₂ and total suspended particulate (TSP), on the other. Multivariate autoregressive integrated moving-average (ARIMA) models were used to adjust for season, temperature, relative humidity, and influenza. A␣sensitivity analysis was run to assess the robustness of the estimators. Results: Graphical analysis revealed a linear relationship between mortality and TSP. The relationship was logarithmic in the case of SO₂. TSP lagged 1 day and SO₂ lagged 3 days with an independent effect on mortality. This relationship was produced without the detection of a minimal threshold in emission values. Conclusions: These results support the hypothesis of an association between pollution levels and mortality between 1986–1992 in Madrid. Additional measures designed to reduce pollution levels without compromising thermal comfort should be implemented. Received: 28 January 1998 / Accepted: 6 July 1998