Abstract and Introduction
Abstract
Lyme disease, caused by the bacterium Borrelia burgdorferi and transmitted in the eastern United States by the black-legged tick (Ixodes scapularis), is increasing in incidence and expanding geographically. Recent environmental modeling based on extensive field collections of host-seeking I. scapularis ticks predicted a coastal distribution of ticks in mid-Atlantic states and an elevational limit of 510 m. However, human Lyme disease cases are increasing most dramatically at higher elevations in Virginia, a state where Lyme disease is rapidly emerging. Our goal was to explore the apparent incongruity, during 2000–2011, between human Lyme disease data and predicted and observed I. scapularis distribution. We found significantly higher densities of infected ticks at our highest elevation site than at lower elevation sites. We also found that I. scapularis ticks in Virginia are more closely related to northern than to southern tick populations. Clinicians and epidemiologists should be vigilant in light of the changing spatial distributions of risk.
Introduction
Lyme disease (LD), caused by the bacterium Borrelia burgdorferi and transmitted in the eastern United States by the black-legged tick (Ixodes scapularis), is the most common vector-transmitted disease in North America. Maintained in an enzootic cycle comprising competent vertebrate reservoir host species, B. burgdorferi is transmitted to humans by the bite of an I. scapularis nymph or adult that acquired infection during a blood feeding as a nymph or larva. Although the principal reservoir host for this pathogen, the white-footed deer mouse, Peromyscus leucopus, is wildly distributed throughout North America, LD is generally confined to 2 geographic foci in the eastern United States: 1 in the upper Midwest and 1 in the Northeast. Densities of host-seeking I. scapularis nymphs correlate significantly with cases of human LD, but this species has been reported throughout much of eastern North America. Nationally, LD incidence increased during 1992–2002, but overall numbers of confirmed cases have since remained relatively stable.
In some locations, LD incidence recently has increased dramatically; in Virginia, the number of confirmed cases nearly tripled from 2006 to 2007 (http://www.vdh.virginia.gov/epidemiology/surveillance/surveillancedata/index.htm) to ≈12.4 cases per 100,000 residents, well above the 1998–2006 average of 2.2 per 100,000. A 1990 report of LD cases in Virginia noted that the disease was rare in the early 1980s but apparently increased in incidence and geographic distribution through the late 1980s, leading the authors to conclude that the disease was expanding southward. Before 2006, most studies of I. scapularis ticks in Virginia focused on the eastern and southeastern parts of the state and found that densities of I. scapularis ticks declined, as did their rate of infection with B. burgdorferi, with distance from the coast. Several early surveys for I. scapularis ticks in Virginia's neighboring states of North Carolina and Maryland also found them to be most abundant on the Coastal Plain but absent or less common in the Piedmont and Appalachian Mountains. During 1983–1987, Apperson et al. surveyed 1,629 hunter-killed deer from the Coastal Plain, Piedmont, and Appalachian Mountain regions of North Carolina and found I. scapularis ticks only on deer from the Coastal Plain. Amerasinghe et al. surveyed 1,281, and 922 hunter-killed deer in 1989 and 1991, respectively, at sites from the Coastal Plain to the Appalachian Mountains of Maryland and found I. scapularis ticks on 59%–70% of deer on the Coastal Plain, fewer on deer in the Piedmont Region, and on only 1%–5% of deer in the Appalachian Mountains.
Although I. scapularis ticks exist in the southeastern United States, they are most easily detected by drag sampling, a method used as a proxy for risk to tick exposure, in areas associated with highest LD incidence, i.e., the Northeast (New Jersey through Massachusetts) and upper Midwest (Wisconsin and Minnesota). The difference in apparent abundance of I. scapularis ticks and risk for LD between the northern and southeastern United States has been the subject of much discussion and debate and might be related, either through behavioral or physiologic mechanisms, to genetic differences between I. scapularis populations in these regions. Population genetic structure of I. scapularis ticks has shown that dynamic range shifts are likely to have occurred in recent evolutionary history and that 2 distinct lineages within this species can be identified; a relatively genetically uniform "American clade" exists in the northern United States (although this lineage has also been detected in the South), and a genetically diverse "southern clade," members of which have been found only in the South. Although other nomenclatures have been proposed for these 2 lineages (e.g., clades A and B for northern and southern lineages, respectively, we follow the terminology established by Norris et al.: "American" describes the widely distributed yet less diverse clade and "southern" describes the geographically restricted yet more diverse mtDNA clade of I. scapularis ticks.
Range expansion of I. scapularis ticks over relatively short periods has been observed. Moreover, recent environmental modeling, based on extensive field collections of host-seeking I. scapularis ticks, suggests that this species suggests that the range of this species is expanding widely and its occurrence in a given area depends on the lack of abiotic drivers, vapor pressure deficit and elevation. In Virginia, studies found that I. scapularis ticks were concentrated in in northern sites; very few ticks were reported in other parts of the state. In contrast, human LD cases at inland, higher-elevation locations have increased in recent years in Virginia (http://www.vdh.virginia.gov/epidemiology/surveillance/surveillancedata/index.htm). The incongruity between human case and vector abundance datasets might be explained by recent (i.e., since 2007) spatial and/or numerical expansion of I. scapularis populations. We hypothesized that density of B. burgdorferi–infected ticks would be highest in counties associated with high incidence of human disease if epidemiologic data represent cases in tick-endemic areas. In contrast, low numbers of infected ticks in areas of high human disease might indicate either misdiagnosis or allochthonous exposure.