Counting on a Pipe Dream: The Effect of Inadequate Tap Water on Dengue Fever in Karnataka, India

Abstract

Dengue fever, a vector-borne neglected tropical disease, is a major public health issue in Karnataka, a state on the coast of southwestern India. While weather, environment, socioeconomic factors, and population demography are all known to influence dengue fever incidence, literature findings regarding mechanisms for transmission are mixed and often surprising. One such surprising finding is how wider access to piped water can actually increase the rates of dengue fever in an area. Intuitively, piped water access should lower the risk of dengue fever, because we would expect that people with tap water have no need for water storage tanks which can breed the Aedes Aegypti mosquito that transmits dengue fever. This thesis delves into the relationship between access to piped water and dengue fever incidence by challenging the assumption that houses with tap water don’t store any water. To do this, I break the people who have tap water access into groups – those with safe and consistent tap water, those with tap water that is prone to interruptions, and those with tap water that is contaminated by chemicals that make it unsafe to drink. I compare these groups to each other, and to the population who don’t have any tap water at all, in terms of their relative risk of dengue fever. First, I model dengue fever cases as a linear function of access to piped water, climate variables, socioeconomic status, and population demographics. This linear regression finds that piped water access is associated with higher dengue fever incidence. However, the results of this model indicate that some of the predictor variables have nonlinear relationships to dengue fever, limiting the conclusions one can draw from a linear model, so I also used a Generalized Additive Model (GAM) to predict the effects of water access groups on dengue fever incidence. The GAM allows each variable to have an independent nonlinear relationship with the predictor, making it ideal for modelling nonlinear effects of seasonal and demographic variables on disease. Three GAM fits serve to illuminate the complex dynamics between water access groups in terms of dengue risk. The main finding of this paper is that untrustworthy piped water sources are associated with a higher risk of dengue fever. In particular, piped water that shuts off frequently has the strongest association with dengue, and may be driving the positive association between dengue fever and general piped water access. This indicates that people who cannot trust their tap to provide enough water for their everyday needs are prone to store water as a backup source. The findings suggest that there might even be an underlying psychological effect of the threat of water scarcity encouraging long-term water hoarding in this group. Long-term undisturbed water storage provides an ideal habitat for the Aedes dengue vector to breed, a mechanistic link to disease which is supported by the literature. Furthermore, the findings suggest that if all piped water was safe and reliable, there would be no positive association between dengue and tap water. The international community today strives to extend tap water access to all people, yet anticipates that piped water schemes will initially suffer from continuity and quality issues as a result of global warming related water shortages. In light of these findings, it is crucial that policymakers in low-middle income countries understand that areas with poor quality piped water systems are particularly vulnerable to dengue fever.