Modeling Precipitation, Acute Gastrointestinal Illness, and Environmental Factors in North Carolina, USA

Public Deposited
Analytics

Downloadable Content

Download PDF
Request Version for Screen Reader
Last Modified
  • October 17, 2023
Creator
  • Downs, Kristen D.
    • Affiliation: Gillings School of Global Public Health, Department of Environmental Sciences and Engineering
Abstract
  • Increasing intensity and frequency of extreme weather events due to climate change underscores the importance of understanding the influence of hydroclimatic variability on health. Meteorological drivers affect rates of acute gastrointestinal illness (AGI), but the association between precipitation and AGI, the sensitivity to modeling decisions, and the effects of sociodemographic and environmental risk factors are not well characterized. Furthermore, methodological differences may reduce inter-study comparability and can affect model estimates.In this dissertation, we reviewed the methodologies of recent time series AGI-weather studies, including outcome and exposure variables, data sources, spatiotemporal aggregation, and model specification. To investigate the sensitivity of the association between AGI and precipitation to exposure definitions and effect measure modification (EMM), we used AGI emergency department (ED) visit and weather data (2008-2015) from North Carolina (NC) to develop daily, ZIP code-level quasi-Poisson generalized linear models and distributed lag models. We compared multiple precipitation metrics: absolute (total precipitation), extreme (90th, 95th, and 99th percentiles with and without zero-precipitation days), and antecedent (cumulative wet-dry days; 8-week wet-dry periods). We assessed for potential EMM by physiographic region, the density of hogs in concentrated animal feeding operations (CAFOs), and percent of population on private drinking water wells.Depending on exposure definition, we observed an overall cumulative decrease of 1-18% in AGI ED rates following extreme precipitation events (over 0-7 days), with stronger effects associated with heavier rainfall, and a 2% (95% CI: 1.02, 1.03) increase after antecedent (8-week) wet periods. Inverse statewide results following extreme precipitation—dominated by the demographic weight of urban centers in the Piedmont region—were consistent with dilution effects posited by the concentration-dilution hypothesis but obscured dramatic sub-state variation. While EMM by private wells was inconclusive, region and hog density strongly modified the associations observed, with increased AGI ED rates following 95th percentile precipitation in the mountains (18%), coastal plains (19%), and areas exposed to hog CAFOs (7-15%). Our results reveal the vulnerability of mountainous, coastal, and CAFO-impacted areas in NC to rainfall-exacerbated AGI risk. This dissertation highlights the hazards of data aggregation and importance of precipitation exposure definitions and effect measure modification when modeling climate-health relationships.
Date of publication
Keyword
DOI
Resource type
Advisor
  • Whittington, Dale
  • Characklis, Gregory
  • Konrad, Charles
  • Levy, Karen
  • Richardson, David
  • Waller, Anna
  • West, Jason
Degree
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2022
Language

Relations

Parents:

This work has no parents.

In Collection:

Items

Thumbnail Title Date Uploaded Visibility Actions
file details Downs_unc_0153D_21309.pdf 2023-02-02 Public Download