Lyme Disease, Virginia, USA, 2000-2011
Lyme Disease, Virginia, USA, 2000-2011
During 1995–1998, the Virginia Department of Health counted 55–73 LD cases per year. The number increased to 122 cases in 1999, and cases continued to increase through the early 2000s. Although Virginia's LD activity during 2000–2005 was focused primarily on northern Virginia and the Eastern Shore of Virginia (a peninsula extending south from Maryland on the eastern side of the Chesapeake Bay), small numbers of LD cases were recorded in counties across Virginia, including counties in the most southern and southwestern parts (Figure 2). During 2006–2007 the incidence of LD increased substantially in counties throughout the Appalachian Mountains (Figure 2). After the change in the SCD in 2008, many of the most southern and southwestern counties that had recorded LD cases before 2008 ceased to report cases, and the geographic progression of LD appeared as a compact front that progressed from county to county from northeast to southwest. LD cases were not observed again in any of the far southwestern counties until 2011, by which time LD was considered endemic to many of the counties immediately to their northwest (Figure 2).
(Enlarge Image)
Figure 2.
Progressive geographic spread of human Lyme disease across Virginia, 2001–2011. Data were reported by the Virginia Department of Health http://www.vdh.virginia.gov/epidemiology/surveillance/surveillancedata/index.htm. Cases per 100,000 population were calculated by county or city census estimate data published for the year preceding the year of the report.
We collected 2,549 ticks from the field: 2,192 Amblyomma americanum (1 larva, 1,917 nymphs, 274 adults), 306 I. scapularis (304 nymphs, 2 adults), 50 Dermacentor variabilis (all adults), and 1 I. dentatus (nymph). Sampling site was a major determinant of I. scapularis density (F = 71.07, p<0.0001, degree of freedom [df] = 3), as was sampling date (F = 6.85, p=0.024, df = 1). Post hoc comparisons indicated that tick density at the highest elevation site (9.55 nymphs/200 m) was significantly greater than at any other site and that tick density at GR (1.66 nymphs/200 m) was significantly higher than at site AB (0.25 nymphs/200 m) (Table). We detected B. burgdorferi DNA in 48 I. scapularis nymphs, 45 of which produced unambiguous sequence reads for at least 1 locus (ospC or intergenic spacer region. Infection prevalence varied significantly among sites (likelihood ratio test, G = 16.3, p<0.0001, df = 3); the prevalence of infection was significantly higher at site LE (0.2) than at sites CR (0.00) and GR (0.04). Because of low sample size, site AB did not yield a reliable estimate of infection prevalence (Figure 3).
(Enlarge Image)
Figure 3.
Variation in estimated prevalence of Borrelia burgdorferi infection in Ixodes scapularis nymphs at 4 field sites in Virginia. Sites are arranged west to east from left to right. LE, Lesesne State Forest; AB, Appomattox-Buckingham State Forest; GR, University of Richmond–owned field site; CR, Crawfords State Forest. Error bars represent 95% CIs.
Analysis of I. scapularis 16S sequences yielded 17 haplotypes (GenBank accession nos. KF146631–47) from 85 individual nymphs (14 haplotypes from 44 ticks at LE, 1 from 2 ticks at AB, 6 from 21 ticks at GR, and 4 from 18 ticks at CR). Maximum-likelihood phylogenetic reconstruction using Tamura 3-parameter model indicated that all haplotypes detected fall within the American clade; none of the ticks we sampled were phylogenetically identified as southern clade I. scapularis (Figure 4). In addition to an overall increase in human LD cases (from 136 in 2000 to an average of >1,000 in 2010 and 2011), we observed a significant spatial shift of the geometric center of LD incidence in Virginia. The longitude value associated with the centroid of each year's LD incidence depended significantly on year from 2000 to 2011 (F = 12.48, p = 0.005, r = 0.56) (Figure 5). Latitude values did not change significantly over time (F = 0.14, p = 0.71, r = 0.01). We also calculated the average LD incidence per county for 2000–2006 (before the dramatic spike in cases in Virginia) and for 2007–2011 to identify counties in which the largest increases in cases occurred (Table).
(Enlarge Image)
Figure 4.
Maximum-likelihood phylogenetic reconstruction of Ixodes scapularis lineages based on 16S rRNA gene sequences using Tamura 3-parameter model (35). All samples beginning with IS were collected during this study; reference sequence GenBank accession numbers are indicated, as were sampling locations (2-letter state abbreviation). The clade containing samples collected in GA, FL, NC, OK, and SC is known as the Southern clade (sensu Norris et al. [20]); the clade containing all samples from this study, indicated by the prefix IS, represents the American clade (more complete explanation of these terms is provided in the text). Bootstrap values at nodes are based on 500 replicates.
(Enlarge Image)
Figure 5.
Centroids of annual incidence, by county, Virginia, 2000–2011. The size of each circle represents the annual number of cases reported by the Virginia Department of Health and is proportional to annual incidence (cases/100,000 population). Black arrow represents the mean linear direction of annual movement among centroids during 2006–2011 (these years indicate the recent dramatic increase in Lyme disease incidence in Virginia).
Results
During 1995–1998, the Virginia Department of Health counted 55–73 LD cases per year. The number increased to 122 cases in 1999, and cases continued to increase through the early 2000s. Although Virginia's LD activity during 2000–2005 was focused primarily on northern Virginia and the Eastern Shore of Virginia (a peninsula extending south from Maryland on the eastern side of the Chesapeake Bay), small numbers of LD cases were recorded in counties across Virginia, including counties in the most southern and southwestern parts (Figure 2). During 2006–2007 the incidence of LD increased substantially in counties throughout the Appalachian Mountains (Figure 2). After the change in the SCD in 2008, many of the most southern and southwestern counties that had recorded LD cases before 2008 ceased to report cases, and the geographic progression of LD appeared as a compact front that progressed from county to county from northeast to southwest. LD cases were not observed again in any of the far southwestern counties until 2011, by which time LD was considered endemic to many of the counties immediately to their northwest (Figure 2).
(Enlarge Image)
Figure 2.
Progressive geographic spread of human Lyme disease across Virginia, 2001–2011. Data were reported by the Virginia Department of Health http://www.vdh.virginia.gov/epidemiology/surveillance/surveillancedata/index.htm. Cases per 100,000 population were calculated by county or city census estimate data published for the year preceding the year of the report.
We collected 2,549 ticks from the field: 2,192 Amblyomma americanum (1 larva, 1,917 nymphs, 274 adults), 306 I. scapularis (304 nymphs, 2 adults), 50 Dermacentor variabilis (all adults), and 1 I. dentatus (nymph). Sampling site was a major determinant of I. scapularis density (F = 71.07, p<0.0001, degree of freedom [df] = 3), as was sampling date (F = 6.85, p=0.024, df = 1). Post hoc comparisons indicated that tick density at the highest elevation site (9.55 nymphs/200 m) was significantly greater than at any other site and that tick density at GR (1.66 nymphs/200 m) was significantly higher than at site AB (0.25 nymphs/200 m) (Table). We detected B. burgdorferi DNA in 48 I. scapularis nymphs, 45 of which produced unambiguous sequence reads for at least 1 locus (ospC or intergenic spacer region. Infection prevalence varied significantly among sites (likelihood ratio test, G = 16.3, p<0.0001, df = 3); the prevalence of infection was significantly higher at site LE (0.2) than at sites CR (0.00) and GR (0.04). Because of low sample size, site AB did not yield a reliable estimate of infection prevalence (Figure 3).
(Enlarge Image)
Figure 3.
Variation in estimated prevalence of Borrelia burgdorferi infection in Ixodes scapularis nymphs at 4 field sites in Virginia. Sites are arranged west to east from left to right. LE, Lesesne State Forest; AB, Appomattox-Buckingham State Forest; GR, University of Richmond–owned field site; CR, Crawfords State Forest. Error bars represent 95% CIs.
Analysis of I. scapularis 16S sequences yielded 17 haplotypes (GenBank accession nos. KF146631–47) from 85 individual nymphs (14 haplotypes from 44 ticks at LE, 1 from 2 ticks at AB, 6 from 21 ticks at GR, and 4 from 18 ticks at CR). Maximum-likelihood phylogenetic reconstruction using Tamura 3-parameter model indicated that all haplotypes detected fall within the American clade; none of the ticks we sampled were phylogenetically identified as southern clade I. scapularis (Figure 4). In addition to an overall increase in human LD cases (from 136 in 2000 to an average of >1,000 in 2010 and 2011), we observed a significant spatial shift of the geometric center of LD incidence in Virginia. The longitude value associated with the centroid of each year's LD incidence depended significantly on year from 2000 to 2011 (F = 12.48, p = 0.005, r = 0.56) (Figure 5). Latitude values did not change significantly over time (F = 0.14, p = 0.71, r = 0.01). We also calculated the average LD incidence per county for 2000–2006 (before the dramatic spike in cases in Virginia) and for 2007–2011 to identify counties in which the largest increases in cases occurred (Table).
(Enlarge Image)
Figure 4.
Maximum-likelihood phylogenetic reconstruction of Ixodes scapularis lineages based on 16S rRNA gene sequences using Tamura 3-parameter model (35). All samples beginning with IS were collected during this study; reference sequence GenBank accession numbers are indicated, as were sampling locations (2-letter state abbreviation). The clade containing samples collected in GA, FL, NC, OK, and SC is known as the Southern clade (sensu Norris et al. [20]); the clade containing all samples from this study, indicated by the prefix IS, represents the American clade (more complete explanation of these terms is provided in the text). Bootstrap values at nodes are based on 500 replicates.
(Enlarge Image)
Figure 5.
Centroids of annual incidence, by county, Virginia, 2000–2011. The size of each circle represents the annual number of cases reported by the Virginia Department of Health and is proportional to annual incidence (cases/100,000 population). Black arrow represents the mean linear direction of annual movement among centroids during 2006–2011 (these years indicate the recent dramatic increase in Lyme disease incidence in Virginia).
Source...