TYPHOID, PAST AND PRESENT: Rain-Driven Seasonality and the Influence of Municipal Water Systems on Disease Dynamics

Abstract

Despite successful control of the disease in the United States and the developed world almost a hundred years ago, typhoid fever causes over 21.7 million cases of illness and 217,000 deaths each year (Crump et al. 2004). It is important to continue research into the disease dynamics and factors which led to the success of typhoid control in the developed world if we hope to lessen and eliminate this burden of disease. This thesis seeks to better quantify the influence of environmental drivers, particularly rainfall, on typhoid dynamics as well as explore the impact of the development of various municipal water systems on typhoid prevalence and patterns. To do this, typhoid data is analyzed from six American cities, including New York, NY, Pittsburgh, PA, Philadelphia, PA, Washington, DC, Chicago, IL and Baltimore, MD, between 1888 and 1932 as well as Kathmandu, Nepal, between 1993 and 2011. Seasonal and multi-year cycles as well as correlation with rainfall are analyzed using power spectral density analysis, magnitude squared coherence estimates, and wavelet analysis. Typhoid dynamics are also explored through the adaptation of a doubly stochastic TSIR (time series susceptible-infected-recovered) model. Through these tests, it was discovered that the strength of annual cycles of typhoid is correlated with water source, as cities with remote, pristine water sources brought to the city by way of aqueduct experienced strong, clear annual cycles while cities with water sources more vulnerable to pollution had more variability and longer trends exhibited by the typhoid cycles. Rainfall and typhoid cycles are also correlated. This correlation did not seem to be related to water source, but did seem to be associated with strength of rainfall cycle. Finally, using a TSIR model to predict disease dynamics of historical typhoid data in New York and Philadelphia revealed that the model was better equipped to predict the regular and strong seasonality of New York rather than the variability of typhoid in Philadelphia. These results show the impact of water sources and water treatment on typhoid dynamics, as well as the influence of environmental drivers on typhoid patterns, and the importance of considering these factors when designing typhoid models or public health interventions.