Date of Award

January 2025

Document Type

Open Access Thesis

Degree Name

Master of Public Health (MPH)

Department

School of Public Health

First Advisor

Kai Chen

Second Advisor

Nicole Deziel

Abstract

Daily ambient PM2.5 (particulate matter ≤ 2.5 µm) concentration poses substantial health risks. PM2.5 comes from various sources, leading to regional differences in concentrations. Urban areas generally experience higher PM2.5 concentrations compared to rural areas. However, urban-rural differences in the relationship between daily PM2.5 concentration and adverse health effects remain unclear. The mortality effects of daily PM2.5 concentration across all counties in North Carolina were analyzed both temporally and spatially over a five-year period (2015-2019). A two-stage time series analysis was conducted using a quasi-Poisson regression model with various lag structures to estimate the relative mortality risk for each county. A meta-analysis was then performed to pool these estimates. All analyses were stratified by urbanicity. Further analysis using the selected lag model included sensitivity and spatial analyses. For spatial analysis, both Global and Local Moran’s I statistics were calculated to determine spatial autocorrelation. For single-day lags, lag 0 and lag 1 showed statistically significant associations. A 10 μg/m³ increase in daily ambient PM2.5 concentration was associated with a relative risk (RR) of 1.02 (95% CI:1.00, 1.03) at lag 0 and 1.01 (95% CI: 1.00, 1.02) at lag 1. For cumulative-day lags, a 10 μg/m³ increase in PM2.5 was associated with significantly elevated relative risks of 1.02 (95% CI: 1.00, 1.03) at lag 0-1 (lag 0 to lag 1), 1.02 (95% CI: 1.00, 1.03) at lag 0-2 and lag 0-3, 1.02 (95% CI: 1.00, 1.04) at lag 0-4 and lag 0-5. Cumulative-day lags were significant for urban counties, while no significant associations were observed in rural counties. However, there was no statistically significant difference in the effect estimates between urban and rural counties. Spatial analysis showed no significant clustering of total mortality risk (Global Moran’s I = -0.06, p = 0.76); Local Moran’s I identified a few significant clusters including two hot and cold spots in southeastern urban counties. Sensitivity analysis confirmed model robustness, with consistent PM2.5 -mortality associations. Overall, this study found that daily ambient PM2.5 concentration is associated with an increased risk of mortality, particularly in urban areas with cumulative exposure. However, the differences between urban and rural areas were not statistically significant, and no clear spatial clustering of mortality risk was detected across North Carolina.

Comments

This is an Open Access Thesis.

Download
Open Access

This Article is Open Access

Share

COinS