The implications of climate change on dengue virus transmission in Bangladesh

Abstract

The transmission of dengue fever is already being impacted by the changing climate. This phenomenon poses a considerable public health challenge for countries like Bangladesh, where regular seasonal outbreaks of dengue fever are already prevalent. This thesis aims to investigate how changing climate will impact long-term dengue epidemiology in Bangladesh as a whole and more specifically in Dhaka, the capital city of the country over the 21st Century.

Several statistical models have been developed to estimate the short-term risk of dengue outbreaks as a function of climate variables but the underlying causal relationships that contribute to dengue transmission and the observed patterns of dengue epidemiology are not accounted for in these models. Initially, we determined the suitability of using climate projections for 21st Century from Global Climate Models (GCM) to assess the impact of changing climate on future dengue risk in Bangladesh setting. We then used the GCM output to assess the impact of changing climate on one aspect of dengue transmission by calculating the change in vectorial capacity (VC) of Aedes aegypti mosquitoes at a seasonal level for all regions in Bangladesh under two future climate change scenarios. The analysis indicates that the annual VC in all divisions of Bangladesh is expected to consistently exceed the threshold for dengue transmission throughout the 21st Century, regardless of the climate change scenarios considered. However, during the latter half of the century, there is a projected decline in the annual VC compared to the period between 1986 and 2005. Despite this, monthly VC variations reveal that the winter/dry season could see an increase in VC, potentially leading to a longer dengue season with outbreaks occurring year-round.

The application of the VC calculation is limited by the fact that it only accounts for temperature and does not consider the impact of other climate variables such as rainfall and humidity, as well as the role of host immunity. To incorporate these factors, we then developed a mechanistic dengue transmission model that considers the influence of temperature, rainfall, and humidity on the transmission of two different dengue serotypes among human hosts and mosquito vectors. We calibrated and validated the model against observed dengue epidemiology data from Dhaka for 1995-2014 using observed climate data as input. We then used GCM output for two future climate change scenarios to simulate the model for two future periods (2030-2049 and 2080-2099) to assess the potential changes in dengue epidemiology in Dhaka. When utilizing observed climate data and climate projections from GCMs specific to Dhaka, our mechanistic model reasonably reproduced the observed dengue epidemiology in Dhaka between 1995 and 2014 in terms of the recurring annual dengue outbreaks, the seasonal pattern of transmission, and the increase in seroprevalence. Simulations for 2030-2049 indicate that dengue transmission is likely to increase regardless of the combination of initial seroprevalence, GCM, and climate change scenario, when compared to the baseline period of 1995-2014. However, for the period 2080-2099, the projected changes in dengue transmission vary, with some combinations of initial seroprevalence, GCM, and climate change scenario predicting a slight increase and others indicating a decrease. The simulations also suggest the seasonal pattern of dengue infections is likely to change in future, with more pronounced change projected for the 2080-2099 period, resulting in a lengthening of the dengue season.

The primary contribution of this thesis is to present a modelling framework that considers the anticipated changes in the future climate and immunological factors to project the long-term risk of dengue epidemics. The model is flexible enough to be adapted to other settings and other pathogens transmitted by the same mosquito vector.