1. Bayesian spatio-temporal modelling to assess the role of extreme weather, land use change and socio-economic trends on cryptosporidiosis in Australia, 2001–2018
- Author
-
Rose Hosking, Aparna Lal, Owen Forbes, and Karel Mokany
- Subjects
medicine.medical_specialty ,Environmental Engineering ,010504 meteorology & atmospheric sciences ,Population ,Cryptosporidiosis ,Distribution (economics) ,010501 environmental sciences ,01 natural sciences ,Extreme weather ,medicine ,Humans ,Environmental Chemistry ,Extreme Weather ,Land use, land-use change and forestry ,Socioeconomics ,education ,Waste Management and Disposal ,Socioeconomic status ,0105 earth and related environmental sciences ,education.field_of_study ,Land use ,business.industry ,Public health ,Australia ,Bayes Theorem ,Pollution ,Geography ,Socioeconomic Factors ,business ,Explanatory power - Abstract
Background Intensification of land use threatens to increase the emergence and prevalence of zoonotic diseases, with an adverse impact on human wellbeing. Understanding how the interaction between agriculture, natural systems, climate and socioeconomic drivers influence zoonotic disease distribution is crucial to inform policy planning and management to limit the emergence of new infections. Objectives Here we assess the relative contribution of environmental, climatic and socioeconomic factors influencing reported cryptosporidiosis across Australia from 2001 to 2018. Methods We apply a Bayesian spatio-temporal analysis using Integrated Nested Laplace Approximation (INLA). Results We find that area-level risk of reported disease are associated with the proportions of the population under 5 and over 65 years of age, socioeconomic disadvantage, annual rainfall anomaly, and the proportion of natural habitat remaining. This combination of multiple factors influencing cryptosporidiosis highlights the benefits of a sophisticated spatio-temporal statistical approach. Two key findings from our model include: an estimated 4.6% increase in the risk of reported cryptosporidiosis associated with 22.8% higher percentage of postal area covered with original habitat; and an estimated 1.8% increase in disease risk associated with a 77.99 mm increase in annual rainfall anomaly at the postal area level. Discussion These results provide novel insights regarding the predictive effects of extreme rainfall and the proportion of remaining natural habitat, which add unique explanatory power to the model alongside the variance associated with other predictive variables and spatiotemporal variation in reported disease. This demonstrates the importance of including perspectives from land and water management experts for policy making and public health responses to manage environmentally mediated diseases, including cryptosporidiosis.
- Published
- 2021
- Full Text
- View/download PDF