Landscape Immunity, Climate, and Yellow Fever Risk

About This Grant

ABSTRACT In 2018, there were over 100,000 severe infections and 50,000 deaths from yellow fever globally. Case fatality rates are estimated to be as high as 20-60%, though mild cases and diagnostic challenges make estimating disease burden difficult. Though a vaccine is available, vaccination rates vary significantly both between and within countries. The 2016-2019 yellow fever epidemic in Brazil was the country’s most significant in 70 years, with the outbreak shifting towards population centers in the east. Despite an estimated national vaccination coverage of 70%, yellow fever has continued to be a public health threat, due partly to heterogeneous vaccine coverage at the municipality level. Yellow fever in Brazil is primarily sylvatic, with transmission between non- human primates and humans. Higher levels of forest fragmentation and low levels of native vegetation are associated with increased risk of outbreaks in non-human primates. Studies of land use and human risk have often used proxies such as vegetation indices and areas impacted by fire. More research is needed to clarify the relationship between land use, especially land use heterogeneity, and the risk of human yellow fever cases. Furthermore, as climate change intensifies, the risk of vector-borne disease is shifting along with a shift in areas most suitable for vector survival and proliferation. These changes need to be incorporated into efforts to model future yellow fever risk. This study will evaluate the association between land use, climate, and the risk of human yellow fever cases in Brazil. Aim 1 will use machine learning to characterize fine scale land use change from 2016-2022 in 6 Brazilian states with varying histories of yellow fever infection, including changes in land use types and landscape heterogeneity. Aim 2 will assess associations between land use, environmental degradation, and incident yellow fever cases via the development of spatial models of risk. Aim 3 will combine data on land use with predications of climatic variables under various climate change scenarios to predict areas at future risk for yellow fever in Brazil. Findings from these analyses will allow the identification of landscape features associated with yellow fever cases in Brazil at an improved level of granularity and improve understanding of yellow fever risk to humans under current and future conditions. This will result in comprehensive risk estimates relevant to yellow fever control programs and the efficient allocation of vaccines. Development of this novel approach will also allow for future expansion of this predictive work to other countries and diseases which are sensitive to climatic and land use changes.