Environmental risk factors in infectious diseases: studies in waterborne disease outbreaks, Ebola, and Lyme disease

• Main Author: Mekaru, Sumiko Rachel
• Language: en_US
• Published: Boston University 2015
• Online Access: https://hdl.handle.net/2144/11144

Thesis (Ph.D.)--Boston University === The resurgence of infectious diseases and global climate change's potential impact on them has refocused public health's attention on the environment's role in infectious disease. The studies in this dissertation utilize the increased availability of satellite image-derived data sets with fine temporal and geographic granularity and the expansion of epidemiologic methods to explore the relationship between the environment and infectious disease in three settings. The first study employed a novel study design and analytic methods to investigate the hypothesis that heavy rainfall is an independent risk factor for waterborne disease outbreaks (WBDOs). We found that a location experiencing a heavy rainfall event had about half the odds of a WBDO two or four weeks later than did a location without a heavy rainfall event. The location-based case-crossover study design utilized in this study may help to expand the research methods available to epidemiologists working in this developing field. The second study employed a location-based case-crossover study design to evaluate standardized differences from historic average of weekly rainfall in locations with a recorded introduction of Ebola into a human. For each 1.0 unit z-score decrease in total rainfall, the odds of an Ebola introduction three weeks later increased by 75%. Given the severity of Ebola outbreaks and the dearth of knowledge about indicators of increased risk, this finding is an important step in advancing our understanding of Ebola ecology. The third study used GIS methods on remote sensing data to estimate the association between peridomestic forest/non-forest interface within 100, 150, 250 meters and Lyme-associated peripheral facial palsy (LAPFP) among pediatric facial palsy patients. After adjustment for sex, age, and socio-economic status, children with the highest level of forest edge in the three radii of analysis had 2.74 (95% CI 1.15, 6.53), 4.58 (1.84, 11.41), and 5.88 (2.11, 16.4) times the odds of LAPFP compared to children with zero forest edge in those radii. This study is the first to examine environmental risk factors for LAPFP. Each of these studies advances the techniques used to investigate environmental risk factors for infectious disease through study design, case definition, data used, or exposure definitions.