Estimates of adverse health effect due to outdoor air pollution from epidemiological studies can be used in setting the regulatory standards and help improve public health. The objective of this paper to is to use time-series analysis to examine the association between counts of deaths and ambient PM2.5 concentration accounting for confounders including meteorology and long-term and seasonal trends in mortality. Multiple models including Poisson generalized linear models, Bayesian Poisson models, and Bayesian negative binomial models were used to examine the health effects associated with PM2.5 concentrations.
We found positive associations between mortality and ambient PM2.5 concentrations but none of the estimates from the three models are statistically significant. We also found that the negative binomial model fits the data better compared to a Poisson regression model, suggesting the importance of accounting for over-dispersion in mortality count data.
Table of Contents
About this Master's Thesis
|Subfield / Discipline|
|Committee Chair / Thesis Advisor|
|Time-series analyses of the association between mortality and ambient PM2.5 concentration ()||2018-08-28||
|Sample R code for analysis.docx ()||2018-08-28||