Abstract

This study investigates the associations between daily asthma hospital admissions and concentrations of ambient pollutants in Montreal, Quebec from 1992 to 1999. Health data was collected from the hospital discharge database (Med-Echo) provided by the Quebec Ministry of Health and Social Services. Selected stationary monitoring sites in and around Montreal were used to obtain daily mean air pollutant concentrations of particulate matter having aerodynamic diameters of 10 om or less, coefficient of haze, sulfur dioxide, ozone, nitrogen dioxide, carbon monoxide and hydrogen sulfide. Weather data was procured from Environment Canada's weather station at Dorval International Airport (45 28'N, 73 45'W). A time series analysis in the format of a general additive model (GAM) was used to model the logarithm of daily asthma admissions as a function of the daily averaged air pollutants. Non-parametric smoothing functions, specifically locally weighted regression smoothers (LOESS), were used to remove confounding factors such as seasonal fluctuations, calendar year, and weather. The data set was analyzed according to the whole year and seasons (cool and warm periods). Asthma admissions were studied by age. Carbon monoxide and coefficient of haze were positively associated with asthma hospital admissions at lags of 0, 1, and 3-day mean.