Research Article: The Role of Hydrogeography and Climate in the Landscape Epidemiology of West Nile Virus in New York State from 2000 to 2010

Date Published: February 6, 2012

Publisher: Public Library of Science

Author(s): Michael G. Walsh, Abdisalan Mohamed Noor.


The epidemiology and ecology of West Nile virus (WNV) have not yet been completely described. In particular, the specific roles of climate and water in the landscape in the occurrence of human WNV cases remain unknown. This study used Poisson regression to describe the relationships between WNV cases and temperature, precipitation, and the hydrogeography of the landscape in New York State from 2000 to 2010. Fully adjusted models showed that hydrogeographic area was significantly inversely associated with WNV cases (incidence rate ratio (IRR) = 0.99; 95% C.I. = 0.98–0.997, p = 0.04), such that each one square kilometer increase in hydrogeographic area was associated with a 1% decrease in WNV incidence. This association was independent of both temperature, which was also associated with WNV incidence (IRR = 2.06; 95% C.I. = 1.84–2.31, p<0.001), and precipitation, which was not (IRR = 1.0; 95% C.I. = 0.99–1.01, p = 0.16). While the results are only suggestive due to the county-level aggregated data, these findings do identify a potentially important surveillance signal in the landscape epidemiology of WNV infection.

Partial Text

West Nile virus (WNV) was introduced into the United States during the summer season of 1999 in New York City. The virus was able to establish itself in each of the 48 contiguous states by 2005. Indeed, all the contiguous states other than Oregon and Washington were reporting human cases by the 2003 season [1]. This virus has disseminated throughout North America with extraordinary swiftness [2], and has contributed to multiple and ongoing local and regional epidemics, which is in contrast to the stable endemicity in Africa and southern Asia, or sporadic occurrence in Europe [2]. While Culex mosquitoes and passerine birds have been fairly well established as the primary vectors and reservoir hosts, respectively [3], important aspects of the landscape epidemiology of WNV remain unknown. Moreover, in New York State WNV epidemics had given way to WNV endemicity within only a few years after its introduction. While some evidence is building regarding the key mechanisms of such rapid distribution, open questions remain, particularly with respect to virus transmission in the context of certain geographic phenomena. For example, there is now some evidence that the primary Culex vectors are different east and west of the Mississippi River, with Culex pipiens being the primary vector in the east, and C. tarsalis being the primary vector in the west [4]–[7]. In addition, both precipitation and temperature seem to play pivotal roles in WNV ecology, which is to be expected given that 1) mosquitoes require water for ovipositioning and 2) higher temperatures shorten the time required to complete their life cycle. Higher precipitation and temperature have both been associated with increased WNV transmission in the United States [8]. Interestingly, however, low precipitation in concert with high temperature has also been associated with greater incidence of WNV infection in humans [9], [10]. Reasons for the latter inverse relationship between low precipitation and increased human WNV cases are not clear, but may be due to more widespread dispersal of the mosquitoes’ preferred hosts, passerine birds, during periods of local drought [11]. Land cover has also been used to delineate geographic differences in WNV transmission. It has been suggested that land use differences between the eastern and western United States correspond to differences in WNV ecology. These differences are purported to reflect differences in habitat preferences between the two primary WNV vectors, respectively, C. pipiens in the east and C. tarsalis in the west [12]. Nevertheless, we still understand little about the extraordinarily complex landscape epidemiology of WNV. In particular, the distribution of water across the landscape, i.e. the hydrogeography, may be of particular importance in the distribution of WNV [13]–[15]. Moreover, the role of hydrogeography in concert with other climate factors, such as precipitation and temperature, have not been adequately described as a component of WNV epidemiology and disease transmission. Nor have these aspects of the physical landscape been considered in combination with WNV surveillance in birds, which could be quite useful given that passerine birds are both the reservoir for WNV and the preferred hosts of C. pipiens. As such, these birds are fundamental to the disease ecology of WNV in the northeastern part of North America [3], [16], and may be important in other areas of North America as well [17]. This paper seeks to take a hydrogeographic approach to human WNV occurrence by combing data on surface water with climate data during each of the years of WNV transmission, from 2000 to 2010, in the 62 counties of New York State. This investigation hypothesized that the hydrogeography, specifically the total area represented by large bodies water, would exhibit a negative association with WNV infection in humans independent of the weather. Secondarily, it was hypothesized that WNV infections in birds in these same counties may have the potential to mediate the association between water in the landscape and human infection with WNV.

All incident cases of WNV in humans were obtained from the New York State Department of Health by county for each of the years from 2000 through 2010 [18]. In addition, surveillance of WNV in dead birds was also recorded in each county from 2000 through 2008. NYS DOH identifies human WNV cases by an ongoing surveillance system, which was initially adopted during and after the 1999 epidemic in New York City [19]. Briefly, New York State’s Department of Health relied on the Health Information Network during the initial outbreak to identify cases of WNV disease. Subsequently, in preparation for sustained transmission in the following season, the Department developed a surveillance program that combined the monitoring of mosquitoes, bird populations, and human cases. The coordinated actions of the Health Information Network and the newly developed species monitoring served as the foundation for the subsequent decade of WNV surveillance in New York State [19].

During the 11 year period from 2000 to 2010, 457 human cases of WNV occurred across the state of New York. These cases were distributed across just 29 of 62 counties (47%), ranging from 1 to 128 cumulative incident cases per county. New York State has a cumulative surface water area of 2619 square kilometers. The median hydrogeographic area across all counties is 40 square kilometers, but these counties also demonstrated considerable hydrogeographic variability, ranging from less than two square kilometers to 421 square kilometers.

This report found strong independent associations between WNV cases and both climatic and hydrogeographic factors across 11 years of observance in New York State. Specifically, increasing precipitation was significantly associated with decreasing WNV cases, but not after considering the amount of surface water in the landscape and temperature. Conversely, increasing hydrogeographic area was significantly associated with decreasing WNV cases independently of both precipitation and temperature. Temperature was consistently positively associated with WNV cases having adjusted for both precipitation and hydrogeography. Moreover, the association between water in the environment and human WNV infection may be mediated by birds during the same years. It should be stressed that these findings cannot be interpreted as causal, but they do suggest the possibility for an important relationship between hydrogeography and climate in the complex epidemiology of WNV infection.