Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Epidemiology and Population Health

Degree Program

Public Health Sciences with a specialization in Epidemiology, PhD

Committee Chair

DuPre, Natalie

Committee Co-Chair (if applicable)

Baumgartner, Richard

Committee Member

Baumgartner, Richard

Committee Member

Taylor, Kira

Committee Member

Zhang, Charlie

Committee Member

Arnold, Forest

Author's Keywords

pneumonia; PM; mortality; particulate; pollution; Kentucky


The role of air pollution as an agent of elevated morbidity and mortality has come under increasingly intense scrutiny in recent years. As the body of literature grows, so too does our understanding of how extensive the harm of such pollutants can be. Particulate matter (PM) particularly harmful due to its small size, ability to travel great distance while airborne, and capacity to infiltrate numerous bodily systems, often to the detriment of those exposed. Besides being associated with increased mortality risk in more heavily exposed populations, PM exposure has also been associated with complications in the respiratory, cardiovascular, and nervous systems. Additionally, exposures to PM are often unequally distributed among populations, with the greatest burden placed upon already marginalized groups. The purpose of this dissertation study was to examine the impact of PM exposure at the time of hospitalization on pneumonia-related mortality risk among individuals hospitalized due to community-acquired pneumonia (CAP). The study sample for this dissertation work came from the Hospitalized Adults with Pneumococcal Pneumonia: Incidence (HAPPI) study conducted in Jefferson County, KY, USA from 2014 to 2016, containing adult residents of the county who were hospitalized due to CAP during that timeframe. PM10 and PM2.5 concentration data were derived from online sources published by the US Environmental Protection Agency (EPA) based on a limited number of air monitoring stations located within Jefferson County. A spatial interpolation method called spatial averaging was used to determine specific levels of PM exposure ascribed to each participant at the time of their hospitalization. Mortality within thirty days of hospitalization was the primary outcome examined as it is likely to reflect CAP-specific mortality. The relationships between PM exposures and mortality were examined using a variety of methods, including logistic and Cox proportional hazards regression and we additionally conducted effect modification analyses. Feasibility of alternative spatial interpolation methods was also examined. Age and race, as well as a number of hospitalization and medical history-oriented covariates, were found to be significantly associated with mortality in preliminary regression analysis. According to the preferred model containing all noted confounders and strong predictors, the primary PM exposures of interest were not found to be significantly associated with mortality; however, effect modification analysis did yield significant results. Black participants exposed to higher levels of PM10 had a greater odds of thirty-day mortality, based on both continuous PM10 (AOR = 1.19 per five-unit increase, 95% CI = 1.04-1.37) and PM10 quartiles (greatest versus lowest quartile; AOR = 2.70, 95% CI = 1.31-5.55). Female participants in the greatest PM10 exposure quartile had higher odds of death within thirty days versus those in the lowest quartile (AOR = 1.46, 95% CI = 1.01-2.14). The same elevated associations were not observed among white or male participants. Additionally, the more advanced spatial interpolation method kriging was not found to be feasible due to sparsity of monitor data at a given time point. The results of the study indicate that certain segments of the population may be disproportionately impacted by PM10 exposure. Additionally, the results provide insight into the risks and benefits associated with specific hospitalization / medical history-oriented covariates. Given the simplistic nature of spatial averaging and the inability to properly conduct kriging with the data that was available, future research on this subject would benefit greatly from alternative, more extensive sources of PM data within cities. For now, it appears that PM10 exposure is associated with elevated thirty-day mortality among Black and female populations within Jefferson County, KY.

Included in

Epidemiology Commons