Air pollution and hospital admissions for diseases of the circulatory system in three US metropolitan areas

Generalized additive models were used to analyze the time series of daily h ospital admissions for cardiovascular and cerebrovascular diseases over the period of 1987-1995 in three major metropolitan areas-Cook County, IL; Los Angeles County, CA; and Maricopa County, AZ-in the United States. In Cook and Maricopa Counties, admissions information was only available for the el derly (ages 65 and over), while in Los Angeles County, admissions informati on was available for all ages. In Cook County, daily monitoring information was available on PM10, CO, SO2, NO2, and O-3. In Los Angeles and Maricopa Counties, monitoring information was available daily on the gases, and info rmation on PM10 was available every sixth day. In Los Angeles County, infor mation on PM2.5 was also available every sixth day. In Cook and Los Angeles Counties, associations were found between each pollutant, with the excepti on of O-3, and admissions for cardiovascular disease, with the gases showin g the strongest associations. In two-pollutant models with PM and one of th e gases, the effect of the gases remained stable, while the effect of PM be came unstable and insignificant. In Maricopa County, the gases, with the ex ception of O-3, were weakly associated with hospital admissions for cardiov ascular disease, while PM was not. In two-pollutant models with two of CO, SO2, and NO2, the pattern of results is heterogeneous in the three counties . In all three counties, only weak evidence of any association between air pollution and cerebrovascular admissions was found.