Impact of environmental factors on neglected emerging arboviral diseases

Abstract

Background: Brazil is a tropical country that is largely covered by rainforests and other natural ecosystems, which provide ideal conditions for the existence of many arboviruses. However, few analyses have examined the associations between environmental factors and arboviral diseases. Thus, based on the hypothesis of correlation between environment and epidemiology, the proposals of this study were (1) to obtain the probability of occurrence of Oropouche, Mayaro, Saint Louis and Rocio fevers in Brazil based on environmental conditions corresponding to the periods of occurrence of the outbreaks; (2) to describe the macroclimatic scenario in Brazil in the last 50 years, evaluating if there was any detectable tendency to increase temperatures and (3) to model future expansion of those arboviruses in Brazil based on future temperature projections.

Methodology/principal findings: Our model assessed seven environmental factors (annual rainfall, annual temperature, elevation, seasonality of temperature, seasonality of precipitation, thermal amplitude, and daytime temperature variation) for their association with the occurrence of outbreaks in the last 50 years. Our results suggest that various environmental factors distinctly influence the distribution of each arbovirus, with temperature being the central determinant of disease distribution in all high-risk areas. These areas are subject to change, since the average temperature of some areas has increased significantly over the time.

Conclusions/significance: This is the first spatio-temporal study of the Oropouche, Mayaro, Saint Louis, and Rocio arboviruses, and our results indicate that they may become increasingly important public health problems in Brazil. Thus, next studies and control programs should include these diseases and also take into consideration key environmental elements.

Fig 1. PCA of arboviruses.

This PCA showing the distribution of ORO (brown), MAY (blue), SLE (yellow) and ROC (red) cases according to environmental variables. The green ellipses show the main clusters: ROC, SLE and ORO+MAY.

Fig 2. The most influential environmental variables.

PCA of each disease showing which environmental variables are the most influential.

Fig 3. Occurrence of arboviruses.

Map of the Brazilian territory showing probability of areas with ORO (AUC: 0.79), MAY (AUC: 0.76), SLE (AUC: 0.85) and ROC (AUC: 099) during outbreaks in the last five decades. The maps were built using QGis software 2.10.1.

Fig 4. Temperature in Brazil during the last 50 years.

Map showing the increase in temperature in the last five decades throughout the Brazilian territory. The maps were created using QGis software (version 2.10.1).

Fig 5. Predicted ORO, MAY, SLE and ROC range expansion in Brazil based on GCM HadGEM2-ES.

The maps show the distribution under two climate change scenarios: RCP 2.6 (lower increase in greenhouse gas emissions) and RCP 8.5 (higher increase in greenhouse gas emissions). The maps were built using QGis software 2.10.1. *”Present” is the scenario in which disease outbreaks have been described, based on 1950–2010 climate data.