2022 Theses Doctoral
Quantifying the effect of extreme and seasonal floods on waterborne infectious disease in the United States
The severity of flood events is predicted to increase as a consequence of climate change and may lead to a higher burden of waterborne infectious diseases in the United States. Contaminated floodwater transports bacterial, protozoal, and viral pathogens that typically cause moderate intestinal or respiratory disease, but can also lead to more serious disseminated infections among immunocompromised, young, and older people. Hydroclimatology and drinking water infrastructure influence the transmission of disease, but their roles are not well-understood and may vary by pathogen-type or geographic region. Specific outbreaks of waterborne disease have been attributed to major floods and cases have been positively associated with some meteorological variables, but the association between infections and flooding has not been systematically examined. In this dissertation, we examine the association between seasonal and extreme floods and parasitic and bacterial infections using multiple flood-indicator variables and exposure definitions.
In Chapter 2, we use multimodel inference and generalized linear mixed models to determine the effect of seasonal meteorology on hospitalizations across the US. We found that hospitalization rates were generally higher in rural areas and in places that relied on groundwater for drinking water sources. Soil moisture, precipitation, and runoff were associated with significant increases in hospitalizations for Legionnaires' disease, Cryptosporidiosis, and Campylobacteriosis, respectively. In Chapter 3, we use 23 years of weekly case data to examine the effect of cyclonic storms on six waterborne infections in a conditional quasi-Poisson statistical model. Storm exposure was defined separately for distinct storm hazards, namely wind speed and cumulative rainfall, and effects were examined over 3 weeks post-storm. We found that exposure to storm-related rainfall was associated with immediate and lagged increases in cases.
In Chapter 4, we use a nonparametric bootstrap to examine the effect of anomalous meteorological conditions, i.e. extremes unrelated to cyclonic storms, on Legionnaires' disease hospitalizations. We also assess the effect of exposure to specific cyclonic storms in a GLMM framework and compare these approaches. Extreme precipitation and months with cyclonic storms were positively associated with Legionnaires' disease hospitalizations. Determining the effect of flooding on Legionnaires' disease is particularly important as it causes severe illness and has steadily increased in incidence for 20 years.
An objective of this dissertation was to develop a framework for examining flood-disease dynamics in the context of hydrometeorological and infrastructure-related factors that may influence transmission. We demonstrated that drinking water source, rurality, and geography may play an important role in these dynamics; the analyses also underscored, however, the urgent need for more extensive epidemiological surveillance and water quality data. Climate change will likely place a considerable strain on aging water infrastructure in the US. A nuanced understanding of flood-disease dynamics is central to mitigating these effects.
- Lynch_columbia_0054D_17489.pdf application/pdf 3.07 MB Download File
More About This Work
- Academic Units
- Environmental Health Sciences
- Thesis Advisors
- Shaman, Jeffrey L.
- Ph.D., Columbia University
- Published Here
- September 21, 2022