Local-Level Risk Factors for Diarrheal Disease

This study investigated local-level risk of diarrheal diseases in rural Bangladesh using a comprehensive geographic information system (GIS) database of health and population events. This work is unique to cholera research because it involves measuring neighborhood-level socio-environmental risk factors with spatial analytical methods.

Using an ecological approach, the project has identified environmental, socioeconomic, and biological risk factors for several different diarrheal diseases including cholera and shigellosis. The project included a study that examined relationships between cholera and large flood control embankments and provided supporting evidence concerning the existence and importance of an aquatic cholera reservoir.

Geography of Avian Influenza Evolution

There exist regional ecosystems that encourage or facilitate faster or more radical AIV evolution. To explore this hypothesis, this project used computational genetics and geographic approaches to build a public, spatially referenced AIV genotype database and then investigate relationships between human-environment factors and AIV evolution.

The specific objectives were to: (1) Classify influenza viral genotypes using their genomic sequence data; (2) Construct an Influenza Genotype-Geographic Database (IGGD) of viruses and hypothesized population-environment drivers; and (3) Analyze the impacts of human-environment ecosystem factors on influenza viral evolution. This project was used to generate a systematic description of the spatio-temporal patterns of influenza viral genotypes and enhance our understanding of ecosystem drivers of influenza viral evolution.

Geographical Analysis in Vaccine Trials

This project utilized ecological and spatial analytical methods to evaluate vaccine efficacy. Conventional vaccine trials often assume that vaccine efficacy is geographically homogeneous. However, spatial variation of ecological factors and disease burden may influence a vaccines protective effect. Ignoring location and ecological differences among vaccine trial participants can affect the vaccine efficacy estimate and ignore possible herd. This project addressed these factors in two vaccine trial settings: 1) cholera vaccine trial in Bangladesh, and 2) malaria vaccine trial in Malawi.

Cholera Prediction Using Environmental Information

Cholera is a major problem in many parts of the world, and it is endemic in Bangladesh and Vietnam. Research into the indirect causes of cholera indicates that environmental factors such as sea surface temperature and ocean chlorophyll concentration play a role in outbreaks. This investigation used 25 years of data to examine the environmental drivers of cholera in Bangladesh and Vietnam using population-derived and spatially derived variables. Our research suggests that the effects of both sea surface temperature and ocean chlorophyll concentration have a lag effect and that more rainfall and river discharge usually reduces cholera outbreaks unless localized flooding that inundates water and sanitation systems has the opposite effect. These within and between site comparisons and prediction models are applicable for development of an early warning system for cholera. The project is presently being extended to simultaneously investigate the local-level environmental drivers described above and regional climate drivers such as El Nino.

Bacterial and Viral Pathogens in Groundwater in Bangladesh

Over the past three decades, Bangladesh has experienced a nearly universal switch of human consumption from surface water to groundwater in order to improve health. This switch has not proven successful for two reasons: at least a third of the population pumping groundwater from existing tube wells in the country is exposed to toxic levels of arsenic, and diarrheal morbidity remains a severe problem for all age groups in rural Bangladesh. This study determines whether or not households that switched from shallow high-arsenic wells to shallow low-arsenic wells may have increased their exposure to certain microbial pathogens (cholera, E. coli, rotavirus, and shigella).

This study answers two research questions:

1) Has poor sanitation in densely populated villages resulted in widespread contamination of shallow aquifers?

2) Has the transition from surface water to tube wells exposed the population to toxic levels of arsenic resulting in spatial and temporal patterns in the distribution of certain types of diarrheal diseases?

Integration of Spatial and Social Network Analysis

Our objectives were to develop new techniques for spatial-social network analysis within the realm of medical geography and spatial epidemiology in the case of infectious disease, particularly cholera and other diarrheal illness. Considering space in addition to social interaction allows an accurate perspective on disease transmission and may thus contribute to public health initiatives.

Data used in these analyses comes primarily from the Matlab, Bangladesh study area and the Health and Demographic Surveillance System administered there by the International Center for Diarrheal Disease Research, Bangladesh (ICDDR,B). This specific project was originally initiated with the introduction of social networks to a previously conducted study of cholera vaccine efficacy in Bangladesh. Using data from the vaccine trial, which took place between 1984-1988, positive cholera incidence in both placebo recipients and vaccines was measured in relation to vaccinated individuals within household-level social and kinship networks. Our results overall suggest that environmental factors play a stronger role in cholera occurrence, which is consistent with literature supporting the hypothesis of primary transmission in humans (i.e. through environmental contact with the cholera pathogen).

Our next steps include re-examining the existing data using networks on a smaller scale and refining the spatial methodologies used. These methods will also be applied to diseases and health outcomes beyond cholera, including shigellosis, which we expect to show stronger association with social networks than cholera. Furthermore, a comprehensive review of social-spatial methods used in geography and other fields is in progress.