Short-term association of air pollution and health in Poland. Results from multi-city study

Background Although time-series studies of the impact of air pollution on mortality are numerous in the literature there is noticeable lack of the results from multi-city studies in Poland. The aim of our analysis is to fill this gap and give estimates based on the most actual data. Methods A multi-city ecological time series analysis was carried out. A database contained a daily number of deaths from all causes (excluding external), by sex and age groups (below 65, 65 and more), PM10 and PM2.5 levels as well as meteorological conditions in 3 agglomerations and 19 big cities in Poland. The analysis was performed using generalized additive models. The distributed lag model method was used to investigate the mortality displacement phenomenon. Results Two versions of time-series analysis considering the level of air pollution from the current and previous day (acute effect) as well as from the past 30 days (allowing for mortality displacement) in the total population as well as in sub-populations determined by sex and age-group were done in each localization. The pooled effect estimate was calculated based on estimates obtained in each city. The value of the relative risk of death from all causes associated with an increase of air pollution with PM10 and PM2.5 was equal to (respectively) 1.005 (1.003-1.006) and 1.004 (1.002-1.005) for acute effect and 1.016 (1.011-1.021) and 1.021 (1.015-1.026) for delayed pollution effect. Higher relative risks were obtained for females and in the older age group. Conclusions A statistically significant short-term effect of air pollution with both PM10 and PM2.5 on mortality was found. The relative risks for PM2.5 were higher than for PM10. Population of females and people aged 65 years and above were more at risk. Key messages The existence of the short-term effect of dust air pollution on mortality in Poland has been confirmed. The air pollution effect modification by sex and age was found.