EcoHealth DOI: 10.1007/s10393-015-1034-3
Ó 2015 International Association for Ecology and Health
Original Contribution
An Examination of the Demographic and Environmental Variables Correlated with Lyme Disease Emergence in Virginia Sara E. Seukep,1 Korine N. Kolivras,1 Yili Hong,2 Jie Li,2 Stephen P. Prisley,3 James B. Campbell,1 David N. Gaines,4 and Randel L. Dymond5 1
Department of Geography, Virginia Tech, 115 Major Williams Hall, Blacksburg, VA 24061 Department of Statistics, Virginia Tech, Hutcheson Hall, Blacksburg, VA 24061 3 Department of Forest Resources and Environmental Conservation, Virginia Tech, 313 Cheatham Hall, Blacksburg, VA 24061 4 Division of Environmental Epidemiology, Virginia Department of Health, 109 Governor St., Richmond, VA 23219 5 Department of Civil and Environmental Engineering, Virginia Tech, 200 Patton Hall, Blacksburg, VA 24061 2
Abstract: Lyme disease is the United States’ most significant vector-borne illness. Virginia, on the southern edge of the disease’s currently expanding range, has experienced an increase in Lyme disease both spatially and temporally, with steadily increasing rates over the past decade and disease spread from the northern to the southwestern part of the state. This study used a Geographic Information System and a spatial Poisson regression model to examine correlations between demographic and land cover variables, and human Lyme disease from 2006 to 2010 in Virginia. Analysis indicated that herbaceous land cover is positively correlated with Lyme disease incidence rates. Areas with greater interspersion between herbaceous and forested land were also positively correlated with incidence rates. In addition, income and age were positively correlated with incidence rates. Levels of development, interspersion of herbaceous and developed land, and population density were negatively correlated with incidence rates. Abundance of forest fragments less than 2 hectares in area was not significantly correlated. Our results support some findings of previous studies on ecological variables and Lyme disease in endemic areas, but other results have not been found in previous studies, highlighting the potential contribution of new variables as Lyme disease continues to emerge southward. Keywords: Lyme disease, GIS, medical geography, Virginia, spatial poisson regression
INTRODUCTION Lyme disease, the most common vector-borne disease in the U.S. (Bacon et al. 2008), is caused by the bacterium Borrelia burgdorferi, and is primarily transmitted by the Ixodes scapularis or Ixodes pacificus tick. The geographic range (Centers for Disease Control and Prevention 2012a)
Correspondence to: Sara E. Seukep, e-mail: [email protected]
and prevalence (Figure 1) of Lyme disease has increased across the U.S. since its original identification in 1975 in Lyme, Connecticut (Steere et al. 1978). This study examines the continued emergence of Lyme disease in the southeastern U.S. by focusing on Virginia, the study area for this research. Virginia has recently experienced a rapid expansion of Lyme disease with a quintupling of cases from 2004 to 2011 (Figure 1), and a spread in case distribution from the densely populated northern part of the state toward the more rural southwestern corner, as illustrated in Figure 2
S. E. Seukep et al.
U.S. Lyme Disease Cases
35,000 30,000 25,000 20,000 15,000 10,000 5,000 0
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year Virginia Lyme Disease Cases 1600 1400 1200 1000 800 600 400 200 0
2000
2001
2002
2003
2004
2005
2006
2007
Year
(Brinkerhoff et al. 2014; Kelly et al. 2014; Li et al. 2014). Using Virginia as a model for examining the continued emergence of Lyme disease, we seek to identify potential contributing factors for its emergence in underlying human and ecological variables. One approach for examining Lyme disease’s emergence is to determine where amplifying organisms and humans intersect. White-tailed deer, primary hosts for adult ticks, especially thrive in ‘‘edge’’ environments, and most specifically, the intermediate zones between forest fragments and open, vegetated spaces (DeNicola 2000). This ecotone, or transitional area between biomes, allows deer to forage in the open space and allows them to retreat when necessary into forested areas. White-footed mice, primary reservoirs for B. burgdorferi, thrive in most forest environments. However, as an opportunistic species, whitefooted mice outcompete other small mammals in disrupted areas of small, fragmented forests, which are created when larger forests are subdivided as a result of land use changes (Yahner 1992). Therefore, the ecotone between forested and open, vegetated areas supports populations of both white-tailed deer and white-footed mice (Despommier et al. 2007). Horobik et al. (2007) specifically examined disease transmission risk at the herbaceous-forest ecotone
2008
2009
2010
2011
Figure 1. Comparison of U.S. Lyme Disease Cases & Virginia Lyme Disease Cases. Sources (Virginia Department of Health 2011; Centers for Disease Control and Prevention 2012b).
and concluded that, while the proportion of infected ticks and therefore entomological risk was higher in the forest interior, human activity at edges appears to be the driving factor that leads to increased incidence rates in areas with greater levels of herbaceous-forest edges. This research quantifies relationships between Lyme disease incidence, and different land cover types and their interspersion, degree of forest fragmentation, and demographic features in an area newly experiencing Lyme disease emergence. Though many studies have looked at land cover or ecotones at the individual residential level (Glass et al. 1995; Dister et al. 1997; Cromley et al. 1998), few have tried to capture broad-scale land cover patterns, reflecting how land use decisions can affect incidence rates over a relatively large area. This study is informed by broad-scale land cover and Lyme disease studies by Jackson et al. (2006a) and Brownstein et al. (2005) that sought to determine the land cover patterns associated with incidence rates. The research presented here examines landscape-level factors associated with Lyme disease emergence since little is known about the role of ecological factors in facilitating the disease’s spread into the southeastern U.S.; this study addresses that gap by characterizing the landscapes associated with Lyme disease cases in Virginia.
An Examination of the Demographic and Environmental Variables
Figure 2. Time series map of Virginia Lyme disease emergence.
The research objective was to examine ecological and demographic factors that may have influenced the emergence of Lyme disease in Virginia. Specifically, we hypothesized that the percentage of land classified as forest and herbaceous cover would be positively correlated with Lyme disease incidence rates, and the percentage classified as developed land would be negatively correlated with Lyme disease incidence rates. Additionally, we hypothesized that the interspersion of these land cover types would be significantly correlated with Lyme disease incidence rates. Finally, we hypothesized that small forest fragments would be positively correlated with incidence rates.
BACKGROUND Determining areas where disease vectors, reservoirs, and humans are most likely to interact can help identify areas of disease risk. Much research has examined the residential landscape as the point of contact between ticks and hosts. Several studies [e.g., Dister et al. (1997) and Frank et al. (1998)] have examined environmental data at residences to determine which variables may affect tick densities. Tick
density on a property could affect the likelihood of human– tick interaction and thus, Lyme disease. Both studies concluded that lawns are negatively correlated and wooded land is positively correlated with tick density. In addition to wooded landscapes, population density was also considered. Cromley et al. (1998) examined residential settings and found that people living in high and medium density zones (1–2 acre lots) were less likely to contract Lyme disease than those living in low-density developments (2–5 acre lots or more). Therefore, it appears that low-density residential zones with at least partially forested lots present a greater Lyme disease risk. Researchers have also considered how forest fragments around homes impact disease risk. The fragmentation of habitats of the species involved in Lyme disease transmission was found to be ‘‘of pivotal importance’’ by EstradaPena (2009, p. 153), when examining how rates of infected ticks varied in an endemic area. The significance of forest fragments more specifically was studied by Yahner (1992) who examined the relative abundance of white-footed mice, B. burgdorferi’s most competent reservoir, in clearcut environments. He found that the relative abundance of white-footed mice as compared to other small mammals
S. E. Seukep et al.
increased as the size of forest fragments decreased. Allan et al. (2003) found that smaller forest fragments (2 ha), potentially due to the higher relative abundance of white-footed mice in smaller fragments. Finally, Brownstein et al. (2005) yielded similar results, finding that small forest fragments were positively correlated with nymphal tick infection rates. However, in Brownstein et al.’s study, small forest fragments were negatively correlated with human disease rates, indicating that other variables appear to influence human Lyme disease rate. Jackson et al. (2006a) examined forest fragmentation trends and land cover ecotones, linking these variables to Lyme disease in Maryland. Lyme disease infections were positively correlated with percent forested-herbaceous edge contrast, percentage forest cover, and income. The study also found a weakly positive correlation between the number of small forest fragments (
Ó 2015 International Association for Ecology and Health
Original Contribution
An Examination of the Demographic and Environmental Variables Correlated with Lyme Disease Emergence in Virginia Sara E. Seukep,1 Korine N. Kolivras,1 Yili Hong,2 Jie Li,2 Stephen P. Prisley,3 James B. Campbell,1 David N. Gaines,4 and Randel L. Dymond5 1
Department of Geography, Virginia Tech, 115 Major Williams Hall, Blacksburg, VA 24061 Department of Statistics, Virginia Tech, Hutcheson Hall, Blacksburg, VA 24061 3 Department of Forest Resources and Environmental Conservation, Virginia Tech, 313 Cheatham Hall, Blacksburg, VA 24061 4 Division of Environmental Epidemiology, Virginia Department of Health, 109 Governor St., Richmond, VA 23219 5 Department of Civil and Environmental Engineering, Virginia Tech, 200 Patton Hall, Blacksburg, VA 24061 2
Abstract: Lyme disease is the United States’ most significant vector-borne illness. Virginia, on the southern edge of the disease’s currently expanding range, has experienced an increase in Lyme disease both spatially and temporally, with steadily increasing rates over the past decade and disease spread from the northern to the southwestern part of the state. This study used a Geographic Information System and a spatial Poisson regression model to examine correlations between demographic and land cover variables, and human Lyme disease from 2006 to 2010 in Virginia. Analysis indicated that herbaceous land cover is positively correlated with Lyme disease incidence rates. Areas with greater interspersion between herbaceous and forested land were also positively correlated with incidence rates. In addition, income and age were positively correlated with incidence rates. Levels of development, interspersion of herbaceous and developed land, and population density were negatively correlated with incidence rates. Abundance of forest fragments less than 2 hectares in area was not significantly correlated. Our results support some findings of previous studies on ecological variables and Lyme disease in endemic areas, but other results have not been found in previous studies, highlighting the potential contribution of new variables as Lyme disease continues to emerge southward. Keywords: Lyme disease, GIS, medical geography, Virginia, spatial poisson regression
INTRODUCTION Lyme disease, the most common vector-borne disease in the U.S. (Bacon et al. 2008), is caused by the bacterium Borrelia burgdorferi, and is primarily transmitted by the Ixodes scapularis or Ixodes pacificus tick. The geographic range (Centers for Disease Control and Prevention 2012a)
Correspondence to: Sara E. Seukep, e-mail: [email protected]
and prevalence (Figure 1) of Lyme disease has increased across the U.S. since its original identification in 1975 in Lyme, Connecticut (Steere et al. 1978). This study examines the continued emergence of Lyme disease in the southeastern U.S. by focusing on Virginia, the study area for this research. Virginia has recently experienced a rapid expansion of Lyme disease with a quintupling of cases from 2004 to 2011 (Figure 1), and a spread in case distribution from the densely populated northern part of the state toward the more rural southwestern corner, as illustrated in Figure 2
S. E. Seukep et al.
U.S. Lyme Disease Cases
35,000 30,000 25,000 20,000 15,000 10,000 5,000 0
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year Virginia Lyme Disease Cases 1600 1400 1200 1000 800 600 400 200 0
2000
2001
2002
2003
2004
2005
2006
2007
Year
(Brinkerhoff et al. 2014; Kelly et al. 2014; Li et al. 2014). Using Virginia as a model for examining the continued emergence of Lyme disease, we seek to identify potential contributing factors for its emergence in underlying human and ecological variables. One approach for examining Lyme disease’s emergence is to determine where amplifying organisms and humans intersect. White-tailed deer, primary hosts for adult ticks, especially thrive in ‘‘edge’’ environments, and most specifically, the intermediate zones between forest fragments and open, vegetated spaces (DeNicola 2000). This ecotone, or transitional area between biomes, allows deer to forage in the open space and allows them to retreat when necessary into forested areas. White-footed mice, primary reservoirs for B. burgdorferi, thrive in most forest environments. However, as an opportunistic species, whitefooted mice outcompete other small mammals in disrupted areas of small, fragmented forests, which are created when larger forests are subdivided as a result of land use changes (Yahner 1992). Therefore, the ecotone between forested and open, vegetated areas supports populations of both white-tailed deer and white-footed mice (Despommier et al. 2007). Horobik et al. (2007) specifically examined disease transmission risk at the herbaceous-forest ecotone
2008
2009
2010
2011
Figure 1. Comparison of U.S. Lyme Disease Cases & Virginia Lyme Disease Cases. Sources (Virginia Department of Health 2011; Centers for Disease Control and Prevention 2012b).
and concluded that, while the proportion of infected ticks and therefore entomological risk was higher in the forest interior, human activity at edges appears to be the driving factor that leads to increased incidence rates in areas with greater levels of herbaceous-forest edges. This research quantifies relationships between Lyme disease incidence, and different land cover types and their interspersion, degree of forest fragmentation, and demographic features in an area newly experiencing Lyme disease emergence. Though many studies have looked at land cover or ecotones at the individual residential level (Glass et al. 1995; Dister et al. 1997; Cromley et al. 1998), few have tried to capture broad-scale land cover patterns, reflecting how land use decisions can affect incidence rates over a relatively large area. This study is informed by broad-scale land cover and Lyme disease studies by Jackson et al. (2006a) and Brownstein et al. (2005) that sought to determine the land cover patterns associated with incidence rates. The research presented here examines landscape-level factors associated with Lyme disease emergence since little is known about the role of ecological factors in facilitating the disease’s spread into the southeastern U.S.; this study addresses that gap by characterizing the landscapes associated with Lyme disease cases in Virginia.
An Examination of the Demographic and Environmental Variables
Figure 2. Time series map of Virginia Lyme disease emergence.
The research objective was to examine ecological and demographic factors that may have influenced the emergence of Lyme disease in Virginia. Specifically, we hypothesized that the percentage of land classified as forest and herbaceous cover would be positively correlated with Lyme disease incidence rates, and the percentage classified as developed land would be negatively correlated with Lyme disease incidence rates. Additionally, we hypothesized that the interspersion of these land cover types would be significantly correlated with Lyme disease incidence rates. Finally, we hypothesized that small forest fragments would be positively correlated with incidence rates.
BACKGROUND Determining areas where disease vectors, reservoirs, and humans are most likely to interact can help identify areas of disease risk. Much research has examined the residential landscape as the point of contact between ticks and hosts. Several studies [e.g., Dister et al. (1997) and Frank et al. (1998)] have examined environmental data at residences to determine which variables may affect tick densities. Tick
density on a property could affect the likelihood of human– tick interaction and thus, Lyme disease. Both studies concluded that lawns are negatively correlated and wooded land is positively correlated with tick density. In addition to wooded landscapes, population density was also considered. Cromley et al. (1998) examined residential settings and found that people living in high and medium density zones (1–2 acre lots) were less likely to contract Lyme disease than those living in low-density developments (2–5 acre lots or more). Therefore, it appears that low-density residential zones with at least partially forested lots present a greater Lyme disease risk. Researchers have also considered how forest fragments around homes impact disease risk. The fragmentation of habitats of the species involved in Lyme disease transmission was found to be ‘‘of pivotal importance’’ by EstradaPena (2009, p. 153), when examining how rates of infected ticks varied in an endemic area. The significance of forest fragments more specifically was studied by Yahner (1992) who examined the relative abundance of white-footed mice, B. burgdorferi’s most competent reservoir, in clearcut environments. He found that the relative abundance of white-footed mice as compared to other small mammals
S. E. Seukep et al.
increased as the size of forest fragments decreased. Allan et al. (2003) found that smaller forest fragments (2 ha), potentially due to the higher relative abundance of white-footed mice in smaller fragments. Finally, Brownstein et al. (2005) yielded similar results, finding that small forest fragments were positively correlated with nymphal tick infection rates. However, in Brownstein et al.’s study, small forest fragments were negatively correlated with human disease rates, indicating that other variables appear to influence human Lyme disease rate. Jackson et al. (2006a) examined forest fragmentation trends and land cover ecotones, linking these variables to Lyme disease in Maryland. Lyme disease infections were positively correlated with percent forested-herbaceous edge contrast, percentage forest cover, and income. The study also found a weakly positive correlation between the number of small forest fragments (
