DAILY VARIATION IN SIZE-FRACTIONATED PARTICULATE AIR POLLUTION AND CARDIOVASCULAR AND RESPIRATORY MORTALITY:

J Pekkanen,P Tiittanen,P Aalto,M Kulmala

DOI: https://doi.org/10.1097/00001648-200509000-00186

2005-01-01

Epidemiology

Abstract:ISEE-191 Introduction: Daily variations in the levels of particulate air pollution have been associated with cardiorespiratory mortality, but there are few studies on ultrafine particles. In Helsinki, previous analyses have used mainly PM10 (mass of particles <10 μm in diameter), but PM10 levels are influenced by resuspended dust, in addition to combustion sources, which produce smaller particles. Methods: Number and size distribution of particulate air pollution was monitored with DMPS in 1998–2002 in Helsinki at a central monitoring site (1775 days). In these preliminary analyses, we analysed three size fractions as indicators of the whole size distribution of particles, i.e. number of particles between 13–18 nm (NC13–18) and 61–84 nm (NC61–84), and PM 2.5 (particles <2.5 μm). The daily number of cardiovascular and respiratory deaths was obtained from Statistics Finland. Data were analysed using Poisson regression by first building a basic model, to which the pollutants were added one by one. Only data lagged up to 3 three days were analysed. The basic models were built mainly based on the lowest AIC criteria using trend for time, temperature, relative humidity, barometric pressure, and dummies for weekdays and holiday. Results: Daily average levels were 2045 (range 140–7160) particles/cm3 for NC13–18, 1010 (range 940–6860) particles/cm3 for NC61–84, and 9.9 (range 1.1–52) μg/m3 for PM 2.5, 8.0 (range1–20) for cardiovascular and 1.8 (range 0–8) for respiratory deaths. The correlation of PM 2.5 was 0.05 with NC13–18 and 0.57 with NC61–84. Associations of particles with mortality differed by season, so all analyses were conducted separately for winter (Nov–Feb), summer (May–Aug), and ‘mid-season’ (rest of the year). NC13–18was positively associated with respiratory mortality in the summer, but negatively in the winter. Other associations were weak. Conclusions: Current preliminary analyses show few associations between daily variation in levels of size-fractionated particulate air pollution and mortality in Helsinki. The observed effect modification by season on the association between the smallest size fraction (13–18 nm) and respiratory mortality is intriguing and may be partly related to the worse penetration of ultrafine particles during winter into indoor air.