VECTOR-BORNE DISEASES, SURVEILLANCE, PREVENTION
Expansion of Zoonotic Babesiosis and Reported Human Cases,
Connecticut, 2001–2010
KIRBY C. STAFFORD III,1,2 SCOTT C. WILLIAMS,1 LOUIS A. MAGNARELLI,1
ANUJA BHARADWAJ,3 STARR-HOPE ERTEL,4 AND RANDALL S. NELSON4
J. Med. Entomol. 51(1): 245Ð252 (2014); DOI: http://dx.doi.org/10.1603/ME13154
ABSTRACT To document the expansion of human babesiosis in Connecticut, we analyzed reservoir
host sera for seroreactivity to Babesia microti Franca and reviewed Connecticut human surveillance
case data collected during 2001Ð2010. Sera from white-footed mice, Peromyscus leucopus RaÞnesque,
from 10 towns in 5 counties, collected at 4 Ð7-yr periods between 2001 and 2010, were tested for total
immunoglobulins. The prevalence of B. microti-positive mice was compared with conÞrmed and
probable human case reports tabulated by the Connecticut Department of Public Health. The highest
babesiosis and rodent seroprevalence rates were in New London County, where this protozoan disease
was Þrst documented in the state. However, human cases and reservoir host infection increased
signiÞcantly from 2001Ð2005 to 2005Ð2010 and in other parts of the state. Clinicians should be aware
that the disease is not conÞned to long-established endemic areas of the state.
KEY WORDS Human babesiosis, Babesia microti, Peromyscus leucopus, white-footed mouse,
epidemiology
Human babesiosis in the northeastern United States is
caused by the rodent protozoan parasite Babesia microti Franca, which is usually transmitted through the
bite of the blacklegged tick, Ixodes scapularis Say. The
disease is also transmissible congenitally and through
blood transfusion (Vannier and Krause 2012). Babesiosis has long been endemic in several coastal areas in
the northeastern United States, particularly parts of
Long Island, New York, Rhode Island, and islands off
of Cape Cod, Massachusetts (Spielman et al. 1981).
However, the number and distribution of reported
human cases of babesiosis in the region have noticeably increased over the past 2 decades since the Þrst
reported human cases in Stonington, CT, in 1988
([CDC] Centers for Disease Control 1989), inland
New Jersey in 1993 (Herwaldt et al. 2003), and the
Lower Hudson River Valley, New York, in 2001
(Kogut et al. 2005). The number of reported cases in
the Lower Hudson River Valley increased 20-fold
from 2001 through 2008, and cases have also been
reported from eastern Pennsylvania (Joseph et al.
2011, Perez Acosta et al. 2013). Those most susceptible
to babesiosis are the elderly, immunocompromised,
and people who lack a functioning spleen (Homer et
al. 2000, Krause 2002). A recent analysis of babesiosis
in Medicare beneÞciaries in the United States from
1 The Connecticut Agricultural Experiment Station, 123 Huntington St., P.O. Box 1106, New Haven, CT 06504.
2 Corresponding author, e-mail: [email protected].
3 Current address: Connecticut Department of Emergency Services
and Public Protection, 278 Colony St., Meriden, CT 06451.
4 Connecticut Department of Public Health, 410 Capitol Ave.,
MS#11EPI, P.O. Box 340308, Hartford, CT 06134.
2006 to 2008 found that 10 states and the District of
Columbia accounted for 83.3 to 84.7% of all cases, with
the highest rates among the elderly in the known
endemic states of Connecticut, Rhode Island, New
York, and Massachusetts (Menis et al. 2012).
The white-footed mouse, Peromyscus leucopus
RaÞnesque, is the principal rodent reservoir for B.
microti in the northeastern United States. Seropositivity rates to B. microti in this animal in endemic areas
can be high (Stafford et al. 1999, Magnarelli et al.
2013). The prevalence of this pathogen is correlated
with higher tick densities (Mather et al. 1996). In 1999,
human seroprevalence and parasitemia in blood donors in Connecticut were signiÞcantly higher in an
endemic county than in a county considered nonendemic (Leiby et al. 2005). In this study, we compared
the reported incidence of human babesiosis over the
10-yr period from 2001 to 2010 with serological evidence of antibodies to B. microti in P. leucopus in
endemic and emerging areas of Connecticut.
Materials and Methods
Surveillance for Babesiosis. Human babesiosis was
made a physician-reportable disease to the Connecticut Department of Public Health (CTÐDPH) in October 1989, and laboratory-reportable in January 1990.
During the study period, a conÞrmed case was deÞned
as identiÞcation of the parasite within red blood cells
on a peripheral blood smear or by polymerase chain
reaction (PCR) testing of whole blood, while a probable case was based on positive serological test results
with the identiÞcation of antibodies to B. microti with
0022-2585/14/0245Ð0252$04.00/0 䉷 2014 Entomological Society of America
246
Table 1.
County
FairÞeld
LitchÞeld
New Haven
New London
Tolland
JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 51, no. 1
Description of Connecticut study areas and mouse trapping effort, 2001–2010
Town or study
area
Trapping
period
No. sites or
plots/yr
Total no.
trap nights
Total
capturesa
Westport
Weston
Redding
Salisbury
Canaan
Cornwall
North Branford
Old Lyme
Groton
MansÞeld
2001Ð2006
2001Ð2006
2007Ð2010
2002Ð2007
2002Ð2007
2002Ð2007
2007Ð2010
2001Ð2003
2001Ð2003
2007Ð2010
5Ð12
5Ð10
3Ð6
4Ð13
4Ð9
4Ð13
3Ð6
1
7
3Ð4
1,511
1,315
1,500
1,547
1,504
1,474
1,500
330
1,005
780
317
202
302
188
140
180
1,028
82
189
159
a
Total captures include both P. leucopus and the deer mouse, P. maniculatus. The deer mouse was not included in the analyses as shown
in Table 3.
a titer of 1:256 or higher (Ertel et al. 2003). The
number of conÞrmed and probable case reports was
combined, and the incidence per 100,000 population
was determined by year, town, and county. The number of conÞrmed babesiosis cases for the 169 towns in
Connecticut is available on the CTÐDPH Web site
(www.ct.gov/dph). The 2000 population census data
for Connecticut were used to calculate the rate for
2001 through 2009, and the 2010 census Þgures were
used for 2010. To simplify tabulation, the human case
data were summarized for the periods 2001Ð2005,
2006 Ð2010, and the entire 10-yr period. Babesiosis was
made nationally reportable in January 2011 (Herwaldt
et al. 2012). National statistics are tabulated at the
county level.
Study Areas. Mice were live-trapped at varying periods and locations from April or June through August
2001Ð2010. Trap sites were located on both private and
public property in 10 towns located in Þve of ConnecticutÕs eight counties (Table 1). A total of 2,787
rodents were captured over 12,466 trap nights (one
trap night equals one trap placed overnight), which
represented 1,822 individual P. leucopus and 318 recaptures. From 2001 to 2003, trapping was conducted
in the Mumford Cove (MC) community of Groton
and at a residence in Old Lyme, New London County
in association with a white-tailed deer, Odocoileus
virginianus, reduction study (Kilpatrick and LaBonte
2003, Dolan et al. 2004). The MC community of 112
homes consisted of ⬇38.8 ha of residential development and ⬇41.3 ha of undeveloped woodland area in
MC proper. Trapping was conducted at both treated
and untreated home sites and adjacent woodland areas.
During 2001Ð2006, there were 5Ð10 and 5Ð12 mousetrapped properties in the towns of Weston and Westport, respectively, in FairÞeld County. Similarly, in
LitchÞeld County, mice were trapped during 2002Ð
2007 at 4⫺13 home properties each in the three towns
of Salisbury, Canaan, and Cornwall (Table 1). These
residential properties varied in size and landscape
characteristics, but all had lawn bordered by mixed
deciduous forest, and little understory vegetation due
to over-browsing by white-tailed deer. Stone walls
were also found on many of the properties. Only
untreated sites related to some tick management stud-
ies were used in FairÞeld County. From 2007Ð2010,
mice were captured in three geographically separate
study areas on South Central Connecticut Regional
Water Authority property in North Branford, New
Haven County; on Centennial Watershed State Forest
property in Redding, FairÞeld County; and at the
University of Connecticut Forest and on municipal
lands in MansÞeld, Tolland County. These forested
sites were characterized primarily by the presence of
sugar maple (Acer saccharum Marshall) with mixed
oak (Quercus spp.) in North Branford, or white ash
(Fraxinus americana L.) and red maple (Acer rubrum
Redding and MansÞeld) as previously described (Williams et al. 2009, Williams and Ward 2010). Many of
these areas had moderate to heavy stands of mature
Japanese barberry, Berberis thunbergii de Candolle,
which was shown to provide habitat conducive to
higher tick densities (Williams et al. 2009, Williams
and Ward 2010). Data for each habitat type for each
location were pooled for the purposes of this study.
Mouse Sampling. Peromyscus were captured using
rigid or folding Sherman box traps (H.B. Sherman
Traps Inc., Tallahassee, FL) baited with peanut butter
and/or apple. Mice were anesthetized with methoxyßurane or isoßurane, received a numbered ear tag
(National Band and Tag, Newport, KY), examined for
ticks, and a blood sample was collected by cardiac
puncture (0.1Ð 0.2 ml) under protocols for limited
biweekly (or longer period) sampling approved by
The Connecticut Agricultural Experiment StationÕs
Institutional Animal Care and Use Committee. Serum
was obtained after centrifugation of clotted blood
samples at 5,000 ⫻ g for 5 min and stored at ⫺60⬚C until
antibody analyses were conducted. Mice were released at the site of capture upon recovery from anesthesia. Total captures included P. leucopus and, in
LitchÞeld County, the deer mouse, Peromyscus
maniculatus LeConte, but the deer mouse data were
not included in the Þnal analysis.
Serological Assays. Indirect ßuorescent antibody
staining methods were used to detect total immunoglobulins to B. microti. Sera from laboratory-reared P.
leucopus were used as negative controls. Antigen of B.
microti consisted of infected erythrocytes obtained
from Golden Syrian hamsters (Mesocricetus auratus)
and whole blood from an infected person diagnosed
January 2014
STAFFORD ET AL.: ZOONOTIC BABESIOSIS AND HUMAN CASES IN CONNECTICUT
247
Results
Fig. 1. Number of reported conÞrmed and probable human cases of babesiosis, Connecticut, 2001Ð2010.
with human babesiosis (Anderson et al. 1991, Anderson and Magnarelli 2004). Details on the source of the
ßuorescein-conjugated antibodies and antigen preparation have been previously described (Anderson et
al. 1991, Magnarelli et al. 1997). Reactivity of infected
erythrocytes is considered positive for antibodies to B.
microti at a serum dilution of ⱖ1:80. Therefore, for the
purposes of this study, prevalence of infection with B.
microti means prevalence of B. microti antibody-positive mice. A total of 1,529 blood samples from P.
leucopus, tested for reactivity to B. microti, were used
for the analyses. Multiple positive results from an individual mouse captured more than once was only
counted as positive once.
Statistical Analysis. With the human incidence data
and proportion of infected mice meeting normality
and equal variance requirements, a two-way ANOVA
was used to examine both the incidence of babesiosis
and prevalence of the mice seropositive for B. microti
by year and county with means separation by the
HolmÐSidack method (Systat Software Inc. 2008, SigmaPlot version 11, Systat Software Inc., San Jose, CA).
Although there were different trapping intervals at
different locations, overall differences in disease incidence between the Þrst half (2001Ð2005) and second
half (2006 Ð2010) of the 10-yr period were compared
with a StudentÕs t-test. The z-test on proportions was
used to compare the prevalence of infection in P.
leucopus between the two 5-yr periods and speciÞc
counties.
Human Babesiosis. The number of human babesiosis cases reported in Connecticut generally increased
over the 10-yr period from 2001 to 2010 (Fig. 1),
especially in comparison with that reported from 1991
to 2000. A nearly sixfold increase in reported cases to
the CTÐDPH was seen during the study period compared with the previous 10 yr (1,699 cases [904 conÞrmed and 795 probable] vs. 296 cases, respectively;
Cartter and Ertel 2001). A nearly twofold increase was
seen during the study period when comparing the Þrst
half of study data with the latter half (605 cases reported during 2001Ð2005 vs. 1,094 cases reported during 2006 Ð2010; Table 2). The greatest number of babesiosis cases was reported in 2008 (n ⫽ 333; Fig. 1).
Case-patients ranged from 0 (2 ⬍ 1 yr of age) to 100
yr in age with a mean age of 56.6 (⫾0.46) yr (n ⫽
1,684). Of the 1,699 cases, 710 were female, 927 were
male, with 62 for which gender was unknown. A slight
majority were conÞrmed cases, diagnosed by positive
blood smear (51%) or PCR (2%), while 47% were
diagnosed by serology and classiÞed as a probable
case. Nearly half (49%) were positive by combined
PCR and serology.
Cases of babesiosis were reported in residents from
148 (87.6%) of ConnecticutÕs primary 169 towns, and
all eight counties (Fig. 2a and b). For 8.2% of the cases,
the county was unknown (Table 2). During the study
period, the incidence of disease differed signiÞcantly
by year (F ⫽ 90.201; df ⫽ 9; P ⱕ 0.001) and county (F ⫽
46.192; df ⫽ 7; P ⱕ 0.001). New London County had
a signiÞcantly (HolmÐSidaki multiple comparison; t ⫽
8.67Ð14.536; all P ⱕ 0.001) higher incidence of disease
than all other counties with a mean of 24.21 cases per
100,000 population, and accounted for 37.1% of the
total number of case reports (Table 2). Unlike the
period from 1991 through 2002 (Ertel et al. 2003),
there was no distinct increasing trend (annual county
data not shown), possibly reßecting the endemicity of
babesiosis in the area. Nevertheless, there was a signiÞcant (one-sample t-test, t ⫽ 6.550; df ⫽ 5; P ⫽ 0.001)
annual ßuctuation in disease incidence in the county
during the same period (Table 3). Similarly, there was
no distinct trend statewide, but the greatest number of
babesiosis cases occurred during 2005, 2007, and 2008,
Table 2. Number of reported cases and annual mean incidencea of human babesiosis for Connecticut and Connecticut counties,
2001–2005, 2006 –2010, and 2001–2010
County
FairÞeld
Hartford
LitchÞeld
Middlesex
New Haven
New London
Tolland
Windham
Unknown
Connecticut
a
2001Ð2005
2006 Ð2010
All years 2001Ð2010
No. cases (%)
Inc (SEM)a
No. cases (%)
Inc (SEM)a
Total cases (%)
Inc (SEM)a
96 (15.9)
28 (4.6)
9 (1.5)
65 (10.7)
42 (7.0)
279 (46.1)
15 (2.5)
23 (3.8)
48 (7.9)
605 (100)
2.18 (0.60)
0.65 (0.23)
0.99 (0.44)
8.38 (2.48)
1.02 (0.29)
21.54 (4.05)
2.20 (0.77)
4.22 (1.32)
Ð
5.15 (1.21)
205 (18.7)
73 (6.7)
53 (4.8)
99 (9.0)
76 (7.0)
351 (32.1)
64 (5.9)
81 (7.4)
92 (8.4)
1,094 (100)
4.62 (1.32)
1.69 (0.28)
5.78 (1.15)
12.71 (3.98)
1.84 (0.42)
26.88 (3.10)
9.20 (2.01)
14.63 (1.93)
Ð
9.67 (1.43)
301 (17.7)
101 (5.9)
62 (3.7)
164 (9.7)
118 (6.9)
630 (37.1)
79 (4.7)
104 (6.1)
140 (8.2)
1,699 (100.0)
3.40 (0.80)
1.17 (0.24)
3.39 (0.99)
10.55 (2.33)
1.43 (0.28)
24.21 (2.56)
5.70 (1.55)
9.43 (2.06)
Ð
7.41 (0.97)
Inc ⫽ annual mean incidence per 100,000 pop (SEM ⫽ standard error of the mean).
248
JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 51, no. 1
Fig. 2. Distribution of reported human babesiosis and mice seropositive for B. microti in Connecticut, 2001Ð2010. A)
Number of babesiosis cases by town. B) Mean incidence per 100,000 population and, in parentheses, total number of cases
of babesiosis by county (top row) and percentage of mice seropositive for B. microti with total number of mice tested (bottom
row; e.g., ND, no data).
apparently due to a high proportion of probable cases,
and decreased in 2009 and 2010 (Fig. 1).
Counties adjacent to New London County, Windham and Middlesex, had the second and third highest
incidence, respectively, during the study period. Although Windham County showed the greatest increase (3.5⫻) in incidence from the Þrst half to the
second half of the study period, LitchÞeld County
experienced a signiÞcant (t ⫽ ⫺3.911; df ⫽ 8; P ⫽
0.004) nearly sixfold increase in incidence between
the 5-yr periods (Table 2). Combined rates from New
London, Windham, and Middlesex counties were signiÞcantly higher than those in the two western counties (t ⫽ 3.804 Ð 4.520; P ⱕ 0.001). Central New Haven
and Hartford counties had the lowest incidence of
human babesiosis. There was no signiÞcant difference
in disease incidence between Windham, Tolland, LitchÞeld, and FairÞeld counties (P ⬎ 0.05). The inci-
Table 3. Number and proportion of white-footed mice with evidence of infection with B. microti by indirect fluorescent antibody
staining methods–Connecticut, by county, 2001–2010
County
FairÞeld
LitchÞeld
New Haven
New London
Tolland
a
b
Year
Total
captures
No. individuals
(no. tested)
No. (%)a mice
positive for B. microti
Incidence human
babesiosisb
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2002
2003
2004
2005
2006
2007
2007
2008
2009
2010
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
44
102
180
63
57
72
90
118
24
70
67
111
97
97
101
43
94
151
15
96
56
118
71
19
11
14
55
66
20
18
26 (25)
61 (44)
143 (39)
59 (41)
32 (18)
66 (13)
90 (90)
118 (118)
24 (24)
52 (52)
67 (60)
98 (98)
89 (81)
85 (74)
98 (77)
43 (42)
94 (94)
151 (151)
15 (15)
53 (53)
42 (42)
70 (65)
53 (25)
16 (11)
9 (9)
14 (14)
55 (55)
66 (66)
20 (20)
13 (13)
5 (20.0)
4 (9.1)
2 (5.1)
7 (17.0)
1 (5.6)
0 (0.0)
47 (52.2)
65 (55.1)
12 (50.0)
22 (42.3)
6 (10.0)
6 (6.1)
20 (24.7)
19 (25.7)
20 (26.0)
5 (11.0)
77 (81.9)
102 (67.5)
13 (86.7)
41 (77.4)
24 (57.1)
41 (63.1)
16 (64.0)
7 (63.6)
4 (44.4)
10 (71.4)
18 (32.7)
9 (13.6)
4 (20.0)
2 (15.4)
0.9
1.6
1.7
2.3
4.4
1.8
8.4
7.1
2.5
3.3
0.0
0.0
2.2
1.6
3.8
7.7
2.9
2.8
1.5
1.0
15.8
21.6
20.1
13.5
36.7
22.8
14.7
13.2
5.1
7.9
Percentage positive is based on individual mouse testing positive at least once if recaptured.
Incidence per 100,000 population for the county each year.
January 2014
STAFFORD ET AL.: ZOONOTIC BABESIOSIS AND HUMAN CASES IN CONNECTICUT
Fig. 3. Proportion of white-footed mice seropositive for
B. microti by county, Connecticut, 2001Ð2005 and 2006Ð2010.
dence of human babesiosis in LitchÞeld County increased during the period from 2002 to 2007 (0.0 Ð7.7
cases per 100,000 population), although the difference
was not signiÞcant (t ⫽ 2.152; df ⫽ 5; P ⫽ 0.0840; Table
3). However, there was a signiÞcant difference in the
incidence of human babesiosis across the entire study
period of 2001Ð2010 (t ⫽ 3.676, df ⫽ 8, P ⫽ 0.006).
There was a doubling of human disease incidence in
FairÞeld County from 2.18 to 4.62 cases per 100,000
population (Table 2).
Prevalence in White-Footed Mice. During the study
period, there was no signiÞcant annual difference (P ⬎
0.05) in the prevalence of Babesia infection in mice,
but the difference between counties was highly signiÞcant (F ⫽ 22.969; df ⫽ 4; P ⱕ 0.001). The prevalence
of antibodies to B. microti in P. leucopus trapped in
New London County from 2001 to 2006 was relatively
high, with 61.4% of 166 mice testing positive for B.
microti (Table 3), a proportion similar to that found in
1997 (76.9% of 26 mice; z ⫽ 1.301; P ⫽ 0.191; Stafford
et al. 1999). In addition, the difference in the prevalence of infection between the Þrst sample years (i.e.,
2002Ð2004) and the subsequent 2-yr interval was virtually the same (61.5 vs. 60.9%, respectively; z ⫽
⫺0.176; P ⫽ 0.860). In New Haven County, the prevalence of Babesia reactivity in the mice was relatively
high (74.4% of 313 mice; Table 3; Fig. 2b; Fig. 3), and
the proportion of seropositive P. leucopus in New Haven County was slightly greater than in New London
County (61.4%).
Babesia seroprevalence in FairÞeld County was also
relatively high, at least during the last 4 yr of trapping,
and showed a signiÞcant (z ⫽ 8.061; P ⱕ 0.001) increase in the proportion of infected mice during the
entire study period. FairÞeld County was the only
county for which mouse trap data were available from
2001 through 2010. The seroprevalence for B. microti
in white-footed mice in FairÞeld County was not signiÞcantly (z ⫽ 0.180; P ⫽ 0.857) different from that
detected in 1997. Nevertheless, there was a signiÞcant
difference in the proportion of infected animals in
FairÞeld County between 1997 and the more recent
2006 Ð2010 period (z ⫽ 2.871; P ⫽ 0.004). Examining
the annual data from Table 3 across the same 5-yr
intervals for which the disease incidence was tabulated, we Þnd that the proportion of antibody-positive
mice increased from 11.4 to 49.2%.
249
More notably, the increase in seroprevalence in P.
leucopus in LitchÞeld County was highly signiÞcant
(z ⫽ 3.938; P ⱕ 0.001) between the period 2002Ð2003
(7.6%), when there were no reported human cases,
and the subsequent 4-yr period (2004 Ð2007; 23.0%).
There was no signiÞcant difference between the 2002Ð
2005 and 2006 Ð2007 periods. There was no signiÞcant
difference in mouse seroprevalence between Tolland,
FairÞeld, or LitchÞeld counties (P ⱖ 0.05). While no
trap data were available for Tolland County for the
earlier 2001Ð2005 study period, 21.4% of P. leucopus
were seropositive for B. microti, a prevalence similar
to that observed in LitchÞeld County but signiÞcantly
less (z ⫽ 3.786; P ⱕ 0.001) than that in New London
County.
Discussion
The number of cases of human babesiosis reported
to the Connecticut Department of Public Health increased by over Þvefold between 2001 and 2008, especially in some areas of the state, although cases
subsequently dropped by nearly half in 2009 and 2010.
This may be due, in part, to lower tick abundance
during 2010 (K.C.S., unpublished data), which was
also reßected in 29% fewer conÞrmed Lyme disease
cases in Connecticut when compared with the previous year (Ertel et al. 2011). The number of cases
reported in this study also represented another Þvefold increase from the 10-yr period from 1991 to 2000,
when only 296 cases were reported to the CTÐDPH
(Cartter and Ertel 2001). While babesiosis is no longer
a coastal, mainly southeastern concern, Windham and
Tolland counties had the third and fourth highest
incidence of human infection, with the majority of
cases still reported in residents of New London
County. LitchÞeld County showed a nearly sixfold
increase in incidence between the 5-yr periods, again
demonstrating the expansion of disease. Similarly,
there were signiÞcant increases or levels in the seroprevalence of antibodies to B. microti in the mice in
FairÞeld, LitchÞeld, and Tolland counties. Clearly, the
incidence of human disease and evidence of infection
in the reservoir host have expanded to other areas of
the state.
Human risk is reßected by recovery of the pathogen
in earlier Connecticut studies, reported cases of human disease, human seroprevalence studies, and the
presence of the parasite, as measured by serological
reactivity, in the Peromyscus reservoir host in this
study. While children and adults often experience
symptomatic disease, some people infected with B.
microti are asymptomatic (Krause et al. 2003). Like
any tick-associated disease, the number of reported
human babesiosis cases might be inßuenced by a number of factors, such as tick abundance, infectivity rates,
peopleÕs activity levels (exposure), media attention,
surveillance methods and physician reporting, number of blood transfusions in the case of babesiosis, and
the clinical information needed for conÞrmation. One
of the principal factors for ßuctuations in reported
Lyme disease cases has been changes in surveillance
250
JOURNAL OF MEDICAL ENTOMOLOGY
method over time (Ertel et al. 2012), although this isnÕt
the case for babesiosis in Connecticut. However, human incidence of babesiosis was high in the coastal
and eastern counties, where we detected high B. microti seroprevalence in reservoir mice. Additional evidence that B. microti is prevalent in rodents in coastal
Connecticut has been provided by the recovery and
isolation in inoculated laboratory hamsters of the protozoan parasite from mice in New London and FairÞeld counties (Anderson et al. 1991, Anderson and
Magnarelli 2004). Similarly, Johnson et al. (2009)
found that Middlesex and New London counties had
signiÞcantly higher human seroprevalence rates for
babesiosis than all other counties among sampled
blood donors; there were seropositive donations from
residents in all eight Connecticut counties.
In contrast to a general pattern of increased reservoir infection and corresponding human infection,
New Haven County had the highest proportion of
mice testing positive for antibodies to B. microti, yet
had the second lowest incidence (1.4 cases per 100,000
population) in the state. This may be attributable, in
part, to the largely urban and suburban population of
New Haven County and partly to the rural, forested
landscape of the study site that may be more characteristic of many areas of Middlesex County, where
disease incidence was relatively high (12.71 cases per
100,000 population) during the same period. Increased
forest density was associated with increased risk of
babesiosis in the state of New York, but the largest
forest patch areas were not associated with babesiosis
(Walsh 2013). As already noted, the North Branford
property had moderate to heavy stands of mature
Japanese barberry, Berberis thunbergii, which provides habitat conducive to higher tick densities (Williams et al. 2009, Williams and Ward 2010), but it is not
a residential area and public access is restricted. In
addition, the New London data were collected earlier
than those from New Haven County. Unlike Tolland
County, this study was limited by the lack of reservoir
host data for Windham and Middlesex counties and
tick prevalence data. It is unclear how many of the
apparent differences are due to differences between
sites rather than changes through time. Additional
studies may show a high prevalence of the parasite in
both the reservoir host and tick in Windham County.
The seroprevalence for B. microti in the mice at the
Tolland County site was similar to that of the other
emerging areas of the state, that is, FairÞeld and LitchÞeld counties. However, the proportion of mice
with antibodies to B. microti in recaptured mice from
a subset from this study in 2007 and 2008 was generally
high in New Haven County (range, 81Ð 84%) but
lower (range, 15Ð33%) in Tolland County (Magnarelli
et al. 2013).
The distribution of human babesiosis remains geographically more constrained than that of Lyme disease, where some of the highest rates in Connecticut
over the past decade have been in LitchÞeld County
(Ertel and Nelson 2003). While the incidence of babesiosis has increased in the northwest region of the
state, it appears less than that reported for the neigh-
Vol. 51, no. 1
boring Duchess County, NY. In the Lower Hudson
River Valley, human babesiosis increased from ⬇2.9
cases per 100,000 during 2001Ð2005 to ⬇16.9 cases per
100,000 during 2006 Ð2008 (Joseph et al. 2011), a rate
comparable with that observed in some areas of Connecticut but still higher than reported for the adjacent
LitchÞeld County. There were 1,610 human cases of
babesiosis in New York (not including New York City)
from 2000 to 2010, a number similar to the 10-yr period
of this study in Connecticut. Nevertheless, the significant increase in B. microti antibodies and evidence of
infection in the mice in LitchÞeld County should provide for increased transmission to I. scapularis, resulting in increased human disease. New London County
in the southeastern corner of the state still had the
highest number of reported cases and documented
incidence of human babesiosis, as well as a high rate
of infection in the P. leucopus reservoir based on seropositivity to B. microti.
National surveillance for human babesiosis began in
18 states and 1 city in January 2011, and 97% of the
1,124 conÞrmed and probable cases were from 7 states
(Connecticut, Massachusetts, Minnesota, New Jersey,
New York [including New York City], Rhode Island,
and Wisconsin; Herwaldt et al. 2012). Connecticut
accounted for only 6.6% of the reported cases, reßecting expansion of the disease through the northeastern
United States However, human babesiosis is underreported and the true incidence in some endemic areas
may be closer to that of the more prevalent Lyme
disease (Krause et al. 2003). In Connecticut, serological screening from endemic (southeastern) and nonendemic (north central) areas of the state in 1999
revealed that of the 1,745 human blood samples, 1.4
and 0.3%, respectively, were conÞrmed as positive for
antibodies to B. microti (Leiby et al. 2005). In the
Lyme region of New London County, we found that
reactivity of P. leucopus sera to the Lyme disease spirochete, Borrelia burgdorferi, was identical to that for
B. microti in the mice (76.9% of 26 mice; Stafford et al.
1999). This again suggests that the proportion of Babesia-infected reservoir mice has been relatively constant in the southeastern portion of the state since at
least the late 1990s, which contributes to the greater
incidence of human disease in New London County.
In the adjacent Rhode Island, the potential risk for
human babesiosis was greatest along the coast where
B. microti is present and nymphal tick densities for I.
scapularis are high (Mather et al. 1996). The possible
reasons for the slow expansion of human babesiosis
relative to Lyme disease in the state are multifold and
may include the following: tick infection rates with B.
microti are generally lower than those of Borrelia burgdorferi and may require greater tick abundance for
transmission, the number of Babesia delivered by a
feeding tick (Mather et al. 1990), the establishment
and prevalence of infection in the reservoir host, and
the more limited reservoir host dispersal capabilities.
Babesiosis was not detected in Connecticut until 1988,
a decade after Lyme disease. In contrast, during the
same period, there were 32,911 reported conÞrmed or
January 2014
STAFFORD ET AL.: ZOONOTIC BABESIOSIS AND HUMAN CASES IN CONNECTICUT
probable Lyme disease cases, for which I. scapularis is
also the principal vector.
There are many factors in the emergence of infectious disease (Smolinski et al. 2003), but the emergence of Lyme disease (and therefore to a similar
extent, that of babesiosis) has been related to reforestation and an increase of white-tailed deer (Barbour
and Fish 1993). More speciÞcally, the emergence of
human babesisosis, like any zoonotic, vector-borne
disease, is a result of the prevalence of infection in
reservoir species, the sufÞcient presence of vectors,
and adequate humanÐvector interaction to result in
infection. Human cases, infected reservoir animals,
and I. scapularis have been identiÞed throughout Connecticut. Babesiosis is highly endemic in southeastern
and coastal areas but still appears to be an emerging,
even endemic, albeit still increasing, infectious disease
in northern and western regions of the state. Clinicians
should consider babesiosis in their diagnoses of persons demonstrating appropriate clinical symptoms
with potential exposure to ticks, and be aware that the
disease is not conÞned to long-established endemic or
coastal areas of the state.
Acknowledgments
We appreciate the assistance of CAES technicians Heidi R.
Stuber, Joseph P. Barsky, Tia M. Blevins, and Michael R.
Short; Department of Public Health epidemiologist Brenda
L. Esponda; and a succession of summer research assistants
with the small mammal trapping, collection of blood and
serum samples, and serological analyses. This study would
not have been possible without the support of private homeowners, the Ledge Light Health District, the Westport
Weston Health District, the Torrington Area Health District,
the Aquarion Water Company, the University of Connecticut, Town of MansÞeld, and the South Central Connecticut
Regional Water Authority. Trapping was conducted under
the Institutional Animal Care and Use Committee Protocols
(P01-99, -05, -09) and wildlife permits from the Connecticut
Department of Energy and Environmental Protection. Various parts of the work in this study were supported by the
Centers for Disease Control and Prevention, The Nature
Conservancy, Natural Resources Conservation Service, Propane Education Research Council, and Hatch funds. This
paper is dedicated to coauthor Louis A. Magnarelli, who
passed away 11 July 2013 after a long illness and whose
comments on the Þnal drafts were, as always, insightful and
constructive.
References Cited
Anderson, J. F., and L. A. Magnarelli. 2004. Babesiosis in
FairÞeld County, Connecticut [letter]. Emerg. Infect.
Dis. 10: 545Ð546.
Anderson, J. F., E. D. Mintz, J. J. Gadbaw, and L. A. Magnarelli. 1991. Babesia microti, human babesiosis and Borrelia burgdorferi in Connecticut. J. Clin. Microbiol. 29:
2779 Ð2783.
Barbour, A. G., and D. Fish. 1993. The biological and social
phenomenon of Lyme disease. Science 260: 1610 Ð1616.
Cartter, M. L., and S.-H. Ertel. 2001. Babesiosis-Connecticut, 1991Ð2000. Conn. Epidemol. Newsl. (serial online)
21: 11Ð12.
251
[CDC] Centers for Disease Control. 1989. Babesiosis Connecticut. Morb. Mortal Wkly Rep. 38: 649 Ð 650.
Dolan, M. C., G. O. Maupin, B. S. Schneider, C. Denatale, N.
Hamon, C. Cole, N. S. Zeidner, and K. C. Stafford III.
2004. Control of immature Ixodes scapularis (Acari: Ixodidae) on rodent reservoirs of Borrelia burgdorferi in a
residential community of southeastern Connecticut.
J. Med. Entomol. 41: 1043Ð1054.
Ertel, S.-H., and R. Nelson. 2003. Lyme disease - Connecticut, 2002. Conn. Epidemiol. 23: 9 Ð10.
Ertel, S.-H., B. Esponda, and R. Nelson. 2003. Babesiosis Connecticut, 2002. Conn. Epidemiol. 23: 10.
Ertel, S.-H., J. Sun, and R. S. Nelson. 2011. Lyme disease Connecticut, 2010. Conn. Epidemol. Newsl. (serial online) 31: 25Ð28.
Ertel, S.-H., R. S. Nelson, and M. L. Cartter. 2012. Effect of
surveillance method on reported characteristics of Lyme
disease, Connecticut, 1996 Ð2007. Emerg. Infect. Dis. 18:
242Ð247.
Herwaldt, B. L., P. C. McGovern, M. P. Gerwel, R. M. Easton,
and R. R. MacGregor. 2003. Endemic babesiosis in another eastern state: New Jersey. Emerg. Infect. Dis. 9:
184 Ð188.
Herwaldt, B., S. Montgomery, D. Woodhall, and E. A. Bosserman. 2012. Babesiosis surveillance - 18 states, 2011.
Morb. Mortal Wkly Rep. 61: 505Ð509.
Homer, M. J., I. Aguilar-Delfin, S. R. Telford III, P. J. Krause,
and D. H. Persing. 2000. Babesiosis. Clin. Microbiol.
Rev. 13: 451Ð 469.
Johnson, S. T., R. G. Cable, L. Tonnetti, B. Spencer, J. Rios,
and D. A. Leiby. 2009. Seroprevalence of Babesia microti
in blood donors from Babesia-endemic areas of the northeastern United States: 2000 through 2007. Transfusion 49:
2574 Ð2582.
Joseph, J. T., S. S. Roy, N. Shams, P. Visintainer, R. B. Nadelman, S. Hosur, J. Nelson, and G. P. Wormser. 2011.
Babesiosis in lower Hudson Valley, New York, USA.
Emerg. Infect. Dis. 17: 843Ð 847.
Kilpatrick, H. J., and A. M. LaBonte. 2003. Deer hunting in
a residential community: the communityÕs perspective.
Wildl. Soc. Bull. 31: 340 Ð348.
Kogut, S. J., C. D. Thill, M. A. Prusinski, J.-H. Lee, P. B.
Backenson, J. L. Coleman, M. Anand, and D. J. White.
2005. Babesia microti, Upstate New York. Emerg. Infect.
Dis. 11: 476 Ð 478.
Krause, P. J. 2002. Babesiosis. Med. Clin. North Am. 86:
361Ð373.
Krause, P. J., K. McKay, J. Gadbaw, D. Christianson, L.
Closter, T. Lepore, S. R. Telford, 3rd, V. Sikand, R. Ryan,
D. Persing, et al. 2003. Increasing health burden of human babesiosis in endemic sites. Am. J. Trop. Med. Hyg.
68: 431Ð 436.
Leiby, D. A., A.P.S. Chung, J. E. Gill, R. L. Houghton, D. H.
Persing, S. Badon, and R. G. Cable. 2005. Demonstrable
parasitemia among Connecticut blood donors with antibodies to Babesia microti. Transfusion 45: 1804 Ð1810.
Magnarelli, L. A., J. F. Anderson, K. C. Stafford, III, and J. S.
Dumler. 1997. Antibodies to multiple tick-borne pathogens of babesiosis, ehrlichiosis, and Lyme borreliosis in
white-footed mice. J. Wildl. Dis. 33: 466 Ð 473.
Magnarelli, L. A., S. C. Williams, S. J. Norris, and E. Fikrig.
2013. Serum antibodies to Borrelia burgdorferi, Anaplasma phagocytophilum, and Babesia microti in recaptured white-footed mice. J. Wildl. Dis. 49: 294 Ð302.
Mather, T. N., S.R.T. III, S. I. Moore, and A. Spielman. 1990.
Borrelia burgdorferi and Babesia microti: efÞciency of
transmission from reservoirs to vector ticks (Ixodes dammini). Exp. Parasitol. 70:55Ð 61
252
JOURNAL OF MEDICAL ENTOMOLOGY
Mather, T. N., M. C. Nicholson, R. Hu, and N. J. Miller. 1996.
Entomological correlates of Babesia microti prevalence in
an area where Ixodes scapularis (Acari: Ixodidae) in endemic. J. Med. Entomol. 33: 866 Ð 870.
Menis, M., S. A. Anderson, H. S. Izurieta, S. Kumar, D. R.
Burwen, J. Gibbs, G. Kropp, T. Erten, T. E. MaCurdy,
C. M. Worrall, et al. 2012. Babesiosis among elderly
medicare beneÞciaries, United States, 2006 Ð2008. Emerg.
Infect. Dis. 18: 128 Ð131.
Perez Acosta, M. E., P. T. Ender, E. M. Smith, and J. A. Jahre.
2013. Babesia microti infection, Eastern Pennsylvania,
USA. Emerg. Infect. Dis. 19: 1105Ð1107.
Smolinski, M. S., M. A. Hamburg, J. Lederberg, and Committee on Emerging Microbial Threats to Health in the
21st Century. 2003. Microbial Threats to Health: Emergence, Detection, and Response. National Academy
Press, Washington, DC.
Spielman, A., P. Etkind, J. Piesman, T. K. Ruebush II, D. D.
Juranek, and M. S. Jacobs. 1981. Reservoir hosts of human babesiosis on Nantucket Island. Am. J. Trop. Med.
Hyg. 30: 560 Ð565.
Stafford, K. C., III, R. F. Massung, L. A. Magnarelli, J. W. Lido,
and J. F. Anderson. 1999. Infection with agents of human
granulocytic ehrlichiosis, Lyme disease, and babesiosis in
Vol. 51, no. 1
wild white-footed mice (Peromyscus leucopus) in Connecticut. J. Clin. Microbiol. 37: 2887Ð2892.
Systat Software Inc. 2008. SigmaPlot 11 userÕs guide part 2 Statistics, vol. 2. Systat Software, Inc., San Jose, CA.
Vannier, E., and P. J. Krause. 2012. Human babesiosis (review). N. Engl. J. Med. 366: 2397Ð2407.
Walsh, M. G. 2013. The relevance of forest fragmentation
on the incidence of human babesiosis: investigating the
landscape epidemiology of an emerging tick-borne disease. Vector-borne Zoonotic Dis. 13: 250 Ð255.
Williams, S. C., and J. S. Ward. 2010. Effects of Japanese
barberry (Ranunclales: Berberidaceae) removed and resulting microclimate changes on Ixodes scapularis (Acari:
Ixodidae) abundances in Connecticut, USA. Environ. Entomol. 39: 1911Ð1921.
Williams, S. C., J. S. Ward, T. E. Worthley, and K. C. Stafford
III. 2009. Managing Japanese barberry (Ranunculales:
Berberidaceae) infestations reduces blacklegged tick
(Acari: Ixodidae) abundance and infection prevalence
with Borrelia burgdorferi (Spirochaetales: Spirochaetaceae). J. Med. Entomol. 38: 977Ð984.
Received 6 August 2013; accepted 12 November 2013.