Time series studies estimating health risks associated with short-term exposure to air pollution

Abstract

Epidemiological studies investigating the health effects of short-term exposure to air pollution most often assess the impact of variations in concentrations of air pollutants on the number of adverse health events from day to day, using regression models for analyzing time series data. Heterogeneity in the results of these studies is caused by differences in mixtures of pollutants, development level of health services, meteorological conditions, geographical location and characteristics of the population, but differences between the way regression models are formed and associated factor effect control are still present. The aim of this paper is to present the results of studies conducted in the world where the effects of air pollution on human health was investigated using time series models, and highlight the need for standardization of the regression models used in the analysis.

Epidemiološke studije koje istražuju zdravstvene efekte kratkotrajnog izlaganja aerozagađenju najčešće procenjuju uticaj varijacija koncentracija zagađujućih materija u vazduhu na broj negativnih zdravstvenih događaja iz dana u dan, korišćenjem regresionih modela za analizu vremenskih serija podataka. Do heterogenosti u rezultatima ovih studija dovode razlike u sastavu smeše polutanata, razvijenosti zdravstvene službe, meteoroloških prilika, geografskog položaja i karakteristika populacije, ali i još uvek su prisutne razlike u načinu formiranja regresionih modela i kontrole delovanja pridruženih faktora. Cilj ovog rada je prikaz rezultata studija sprovedenih u svetu u kojima je uticaj aerozagađenja na zdravlje ispitivan primenom modela vremenskih serija, kao i ukazivanje na potrebu standardizacije regresionih modela koji se koriste u analizi.