Investigating the Short-term Effects of Air Pollution on Daily Mortality Among the Elderly Using Generalized Additive Distributed Lag Modelling

Abstract

PP-29-210 Background/Aims: Melbourne, Australia, with a population of nearly 4 million, is a city that typically experiences low levels of air pollution. Despite having low annual levels of air pollution, air quality events such as bushfires, dust storms, and smog do occur. Moreover, these events typically occur in combination with periods of extreme heat, compounding the environmental stress on the local population. We aim to quantify the short-term associations between air pollution and mortality in Melbourne across 2 dimensions—-the range of the pollutant and its distributed lag effect. Methods: The concentration–response relationship between daily ambient air pollution and daily all-cause mortality in the elderly in Melbourne over the period of 2002–2006 has been assessed using generalized additive distributed lag modelling. After controlling for long-term trends, seasonality, day-of-the-week, and weather single-pollutant models were fit, using a distributed lag of 5 days for ozone (O3), particulate matter ≤10 μm (PM10), and nitrogen dioxide (NO2). Results: We found the overall increase of 1.13% in DACMORT65+ over 5 days of lag for a 10 μg/m3 increase in PM10 above the mean during the period of 2002–2006. A 10 ppb increase in O3 above the mean resulted in an increase of 1.25% and a 10 ppb increase in NO2 above the mean resulted in an increase of 1.54%. The maximum positive effect for each pollutant was experienced within a window of 2 days following exposure. Conclusion: The effect of PM10 on mortality had the longest duration followed by O3 and NO2, respectively. These findings indicate that the effects of air pollution in Melbourne are quite modest and of relatively short duration.