MAJOR ARTICLE
A Large Community Outbreak of Blastomycosis in Wisconsin With Geographic and Ethnic Clustering Monika Roy,1,2 Kaitlin Benedict,2 Eszter Deak,2 Miles A. Kirby,3 Jena T. McNiel,4 Carrie J. Sickler,5 Eileen Eckardt,5 Ruth K. Marx,5 Richard T. Heffernan,3 Jennifer K. Meece,6 Bruce S. Klein,7 John R. Archer,4 Joan Theurer,5 Jeffrey P. Davis,4 and Benjamin J. Park2 1
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Epidemic Intelligence Service, Scientific Education and Professional Development Program Office, and 2Mycotic Diseases Branch, Centers for Disease Control and Prevention, Atlanta, Georgia; 3Bureau of Environmental and Occupational Health, and 4Bureau of Communicable Diseases and Emergency Response, Wisconsin Division of Public Health, Madison, 5Marathon County Health Department, Wausau, 6Marshfield Clinic Research Foundation; and 7 Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin
Background. Blastomycosis is a potentially life-threatening infection caused by the soil-based dimorphic fungus Blastomyces dermatitidis, which is endemic throughout much of the Midwestern United States. We investigated an increase in reported cases of blastomycosis that occurred during 2009–2010 in Marathon County, Wisconsin. Methods. Case detection was conducted using the Wisconsin Electronic Disease Surveillance System (WEDSS). WEDSS data were used to compare demographic, clinical, and exposure characteristics between outbreak-related and historical case patients, and to calculate blastomycosis incidence rates. Because initial mapping of outbreak case patients’ homes and recreational sites demonstrated unusual neighborhood and household case clustering, we conducted a 1:3 matched case-control study to identify factors associated with being in a geographic cluster. Results. Among the 55 patients with outbreak-related cases, 33 (70%) were hospitalized, 2 (5%) died, 30 (55%) had cluster-related cases, and 20 (45%) were Hmong. The overall incidence increased significantly since 2005 (average 11% increase per year, P < .001), and incidence during 2005–2010 was significantly higher among Asians than non-Asians (2010 incidence: 168 vs 13 per 100 000 population). Thirty of the outbreak cases grouped into 5 residential clusters. Outdoor activities were not risk factors for blastomycosis among cluster case patients or when comparing outbreak cases to historical cases. Conclusions. This outbreak of blastomycosis, the largest ever reported, was characterized by unique household and neighborhood clustering likely related to multifocal environmental sources. The reasons for the large number of Hmong affected are unclear, but may involve genetic predisposition. Keywords.
blastomycosis; outbreak; Wisconsin; Hmong.
Blastomycosis is an infection caused by Blastomyces dermatitidis, a soil-based dimorphic fungus that can infect immunocompetent and immunocompromised individuals. The most frequent clinical manifestations of
Received 8 January 2013; accepted 16 May 2013; electronically published 3 June 2013. Correspondence: Monika Roy, MD, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, MS C-09, Atlanta, GA 30307 (
[email protected]). Clinical Infectious Diseases 2013;57(5):655–62 Published by Oxford University Press on behalf of the Infectious Diseases Society of America 2013. This work is written by (a) US Government employee(s) and is in the public domain in the US. DOI: 10.1093/cid/cit366
blastomycosis are pulmonary disease, and, less frequently, cutaneous and disseminated disease [1]. Blastomyces dermatitidis infection is endemic in the United States, with most cases occurring in the Ohio and Mississippi river valleys, the southeastern states, and around the Great Lakes [1]. Investigation of prior outbreaks, the majority of which occurred in northern and central Wisconsin, have identified exposure to soil, yard waste, waterways, and outdoor activities such as camping, hiking, and fishing as risk factors for blastomycosis [1–4]. Blastomycosis is reportable in 5 states and has been reportable in Wisconsin since 1984.
Largest US Blastomycosis Outbreak
•
CID 2013:57 (1 September)
•
655
During February 2010, county health officials first noticed an increase in the number of reported cases of blastomycosis in Marathon County, in north-central Wisconsin, beginning in October 2009. Initial investigation revealed case clustering within neighborhoods and households, and a disproportionate number of cases occurring among persons of Hmong ethnicity. Hmong people, originally from Southeast Asia, resettled in the United States after the Vietnam War, and many reside in a small number of counties in Wisconsin, with Marathon County having the largest Hmong population in the state [5]. During June–July 2010, we conducted an epidemiologic investigation to examine the high rates of blastomycosis occurring among Marathon County residents and determine risk factors for B. dermatitidis infection, including exposures to potential common environmental sources.
of binomial proportions were used to examine differences in ageadjusted incidence rates between Asians and non-Asians. Case-Case Analyses
Case-case analyses were conducted to examine and compare characteristics of cases occurring among different ethnic and temporal groupings. Comparisons were made between the following groups of Marathon County cases: (1) outbreak-period cases versus those occurring during previous years (2005–2009), (2) 2005–2010 Asian cases versus 2005–2010 non-Asian cases, and (3) outbreak-period Asian cases versus 2005–2009 Asian cases. Case-case analyses were conducted using Asian race as a proxy because, with the exception of cluster cases during the outbreak period, Hmong-specific data were not available.
METHODS Case Finding and Case and Cluster Definitions
A case was defined as illness in a resident of Marathon County with a clinical specimen positive for Blastomyces by culture or morphology consistent with Blastomyces using microscopy. The outbreak period was defined as 1 September 2009 through 14 June 2010. Cases were identified through the Wisconsin Electronic Disease Surveillance System (WEDSS), which contains demographic, clinical, and exposure information from a detailed standardized Wisconsin Division of Public Health (WDPH) blastomycosis questionnaire. Exposures were defined as those occurring during the 90 days prior to the illness onset date. Additionally, we reviewed all cases of blastomycosis among Marathon County residents that were reported to the WDPH with illness onsets during calendar years 2000–2010. The residential addresses and locations of outdoor activities of outbreak-period Marathon County case patients were mapped using ArcGIS software (version 9.3), and a statistical space-time analysis was conducted using SaTScan software (version 9.1) to confirm the presence of an outbreak. Based on residential mapping, we defined a cluster as either ≥2 cases occurring in the same household with ≥1 other case occurring in the same neighborhood/subdivision, or ≥4 cases occurring in the same neighborhood/subdivision. A cluster case was defined as an outbreak case within a cluster. Incidence Rates
Annual age-adjusted incidence rates were calculated among Asian and non-Asian populations in Marathon County during 2000– 2010. Hmong-specific incidence rates were not available, but approximately 78% of Asians in Marathon County are estimated to be Hmong [5]. Denominator data obtained from the yearly US Census Bureau estimates for 2000–2010 were used to calculate incidence estimates. Age adjustment was performed using the Marathon County 2000 population as a standard reference. Comparison
656
•
CID 2013:57 (1 September)
•
Roy et al
To understand whether specific types of exposures were associated with cluster cases during the outbreak period, we conducted a 1:3 matched case-control study. A control was defined as a person with no clinical, laboratory, or radiological evidence of blastomycosis who was a resident of the same household or neighborhood as the matched case patient during 1 September 2009–14 June 2010. Controls were matched by age group (child aged <18 years vs adult) and were asked about their potential exposures during the same 90-day time period as their matched case patient. If household controls were unavailable, neighborhood controls were systematically selected by geographic proximity. A standardized questionnaire was administered in person in English, Spanish, or Hmong by trained staff. The questionnaire included questions regarding physical aspects of the home environment and potential exposures near the home, recreational and occupational outdoor activities, and medical information (ie, underlying conditions and course of illness). Data Analysis
Data were analyzed using SAS software (version 9.2). Categorical variables were compared using χ2 and Fisher exact tests, and continuous variables were compared using Student t tests. Univariate odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Two-sided P values of ≤.05 were considered statistically significant. In the cluster case-control analysis, variables with a P value of <.3 were considered for inclusion in a forward stepwise multivariate logistic regression model. RESULTS Outbreak and Cluster Delineation and Case Characteristics
The number of blastomycosis cases among Marathon County residents during 2000–2010 increased during May 2005–December 2006, November 2007–June 2008, and September 2009–July 2010 (Figure 1).
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Cluster Case-Control Study
We identified 55 blastomycosis cases with estimated illness onset dates during the outbreak period of 1 September 2009–14 June 2010. Among case patients, the median age was 31 years (range, 3–81 years); 36 (65%) were male, 20 (45%) were Asian, 33 (70%) were hospitalized, and 2 (5%) died. Three patients
(5%) had extrapulmonary infection: 1 patient, a renal transplant patient, had disseminated infection, 1 had bone disease, and 1 had skin lesions without pulmonary involvement. Eleven (25%) case patients were immunocompromised, 9 (20%) had a history of smoking, and 8 (18%) had other household members
Figure 2. Age-adjusted annual reported incidence rates of blastomycosis per 100 000 population among Asian and non-Asian residents, Marathon County, Wisconsin, 2000–2010.
Largest US Blastomycosis Outbreak
•
CID 2013:57 (1 September)
•
657
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Figure 1. Reported number of blastomycosis cases by month of illness onset, Marathon County, Wisconsin, January 2000–December 2010.
Table 1. Comparison of Characteristics of Patients With OutbreakAssociated Blastomycosis Illness Onsets During 1 September 2009– 14 June 2010 to Those of Patients With Blastomycosis Illness Onsets During 1 January 2005–31 August 2009 (Historical Cases), Marathon County, Wisconsin
Characteristic
Outbreak (n = 55)a
Historical (n = 108)a
OR (95% CI)
36 (65) 31 (3–81)
77 (71) 0.8 (.4–1.6) 37 (5–85) NA
.55 .18
Asian Hospitalized
20 (45) 33 (70)
27 (27) 58 (62)
2.3 (1.1–4.7) 1.5 (.7–3.1)
.03 .32
2 (5)
4 (4)
1.1 (.2–6.2)
1.00
29 (32) 37 (42)
0.7 (.3–1.6) 0.3 (.2–.8)
.41 .02
Immunocompromised 11 (25) Smoking history 9 (20) 8 (18)
2 (2)
9.3 (1.9–46.1) .001
18 (44)
34 (40)
1.2 (.6–2.5)
.64
9 (21)
13 (15)
1.6 (.6–4.1)
.33
Visited a cabin Camp/fish/hike
7 (21) 20 (54)
18 (21) 36 (44)
1.0 (.4–2.8) 1.5 (.7–3.3)
.95 .30
All-terrain vehicle use
10 (24)
11 (13)
2.0 (.8–5.3)
.14
Brush clearing Near excavation or construction site
10 (24) 8 (20)
24 (27) 29 (33)
0.9 (.4–2.0) 0.5 (.2–1.2)
Hunting
Gardening Occupational exposure Travel
Male Age, y, median (range)
32 (68) 68 (70) 0.9 (.4–1.9) 25 (3–81) 40 (5–85) NA
.81 .0001
Hospitalized
OR (95% CI)
P Value
27 (68)
57 (65)
1.1 (.5–2.5)
.76
Died Immunocompromised
2 (5) 8 (21)
3 (4) 31 (35)
1.5 (.2–9.2) 0.7 (.2–1.2)
.65 .10
Household member diagnosed with blastomycosis Smoking history
8 (21)
2 (2)
5 (13)
37 (43)
0.2 (.1–.8)
<.001
Own a dog
2 (6)
43 (51)
0.1 (.0–.3)
<.001
Hunting Visited a cabin
2 (5) 0 (0)
17 (20) 25 (30)
0.2 (.1–1.1) NA
.05 <.001
Camp/fish/hike
10.9 (2.2–54.4) .001
11 (35)
40 (49)
0.6 (.2–1.3)
.19
All-terrain vehicle use Brush clearing
0 (0) 2 (5)
18 (21) 28 (32)
0.1 (.0–.5) 0.1 (.0–.5)
.004 .001
.73 .13
Near excavation or construction site Gardening
3 (8)
30 (35)
0.2 (.0–.6)
.003
Occupational exposure Travel
16 (38) 3 (7)
33 (37) 27 (32)
1.0 (.5–2.2) 0.2 (.0–.6)
.91 .002
6 (15)
48 (59)
0.1 (0.0–.3)
<.0001
Data are presented as No. (%) unless otherwise specified. Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio. a
Characteristic
10 (27)
35 (41)
0.5 (.2–1.3)
.15
2 (6)
24 (29)
0.1 (.0–.6)
.004
8 (24)
43 (54)
0.3 (.1–.7)
.004
Data are presented as No. (%) unless otherwise specified. Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio. a
Denominators for specific characteristics vary because of missing data.
Denominators for specific characteristics vary because of missing data.
diagnosed with blastomycosis. Camping, fishing, or hiking was reported among the majority (54%) of case patients. Among the 55 outbreak blastomycosis cases, 30 (55%) cases were geographically clustered in 5 neighborhoods, with a range of 3–10 cases per cluster. When cluster cases were excluded from the epidemic curve, the distribution of cases not assigned to a cluster during the outbreak period was similar to that noted during previous years, with the exception of 2006, when a previously documented outbreak occurred in a neighboring county [4] (Figure 1). SaTScan analysis further demonstrated the existence of a statistically significant (P < .05) outbreak during 2009–2010 (data not shown). Mapping of outbreak case patients’ homes and recreational activities identified 5 distinct residential clusters containing a total of 30 cluster cases. Incidence Rates
The overall blastomycosis age-adjusted incidence rate (aIR) among Marathon County residents (Figure 2) has been increasing since 2005 (average increase per year, 11%), with peaks
658
•
CID 2013:57 (1 September)
•
Roy et al
occurring during outbreak years 2006 and 2010. Whereas aIRs were similarly low among Asians and non-Asians prior to 2005, the increased aIR observed among the total population since 2005 has disproportionally affected Asians, who had a 586% total increase, compared to non-Asians with a 9.2% decrease during the same time period. The aIR among Asians was significantly greater than among non-Asians during each individual year (P = .003 for 2005, and P < .001 for each year during 2006–2010). Case-Case Analyses
Outbreak Cases Versus Historical (2005–2009) Cases Compared to historical case patients, outbreak case patients were significantly more likely to be Asian (OR, 2.3; 95% CI, 1.1–4.7), and to have a household member diagnosed with blastomycosis (OR, 9.3; 95% CI, 1.9–46.1) (Table 1). Outbreak case patients were less likely than historical case patients to have smoking histories (OR, 0.3; 95% CI, .2–.8), occupational exposures (OR, 0.2; 95% CI, .0–.6), or travel within the 90 days prior to illness onset (OR, 0.1; 95% CI, .0–.3).
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Household member diagnosed with blastomycosis Own a dog
2005– 2010 NonAsian (n = 97)a
2005– 2010 Asian (n = 47)a
P Value
Male Age, y, median (range)
Died
Table 2. Comparison of Characteristics of Asian Patients to NonAsian Patients With Blastomycosis Illness Onsets During 1 January 2005–14 June 2010, Marathon County, Wisconsin
Table 3. Comparison of Characteristics of Asian Patients With Blastomycosis Onsets During 2010 to Characteristics of Asian Patients With Blastomycosis Onsets During 2005–2009, Marathon County, Wisconsin
Characteristic Male Age, y, median, (range) Hospitalized
2005–2010 2010 Asian Asian (n = 27)a (n = 20)a 14 (70)
18 (67)
OR (95% CI)
Cluster Case-Control Study P Value
1.2 (.3–4.1)
20.5 (3–81) 33 (5–73)
NA 0.6 (.2–2.6)
.81 .20
15 (71)
1 (6) 4 (21)
1 (5) 4 (20)
.58
Household member diagnosed with blastomycosis Smoking history
8 (42)
0 (0)
NA
3 (6)
2 (10)
1.7 (.2–11.4)
.66
Own a dog
2 (13)
0 (0)
NA
.21
Hunting Visited a cabin
0 (0) 0 (0)
2 (10) (0)
NA NA
.49 NA
Camp/fish/hike
4 (27)
7 (44)
0.5 (.1–2.1)
All-terrain vehicle use Brush clearing
0 (0) 1 (6)
0 (0) 1 (5)
NA NA 1.2 (.07–20.6) 1.00
Near excavation or construction site Gardening
3 (19)
0 (0)
3 (18)
7 (35)
Occupational exposure Travel
0 (0)
2 (11)
NA
.49
0 (0)
8 (44)
NA
.004
1.2 (.07–20.3) 1.00 1.1 (.2–5.0) 1.00
NA
.003
.32
Twenty-six of the 30 cluster case patients and 78 controls (29 household controls and 49 neighborhood controls) were interviewed. The median age was 28 years for case patients and 36 years for controls (Table 4). In bivariate analysis, case patients were more likely to be male (OR, 2.8; 95% CI, 1.0–8.4), Hmong (OR, 18.4; 95% CI, 2.4–828.1), have a chronic medical condition (OR, 5.4; 95% CI, 1.1–34.0), and report exposure to an excavation or construction site (OR, 4.2; 95% CI, 1.1–19.9). Other outdoor exposures or activities, including cultural events, did not differ between the 2 groups. In the multivariable logistic regression analysis, Hmong ethnicity (adjusted OR [AOR], 12.1; 95% CI, 1.3–611.9; P = .019), age (AOR, 0.94; 95% CI, .88–.99; P = .027), and having a chronic medical condition (AOR, 10.6; 95% CI, 1.5–132.6; P = .013) remained independently associated with cluster case status.
.08
DISCUSSION 0.4 (.1–1.9)
.29
Data are presented as No. (%) unless otherwise specified. Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio. Denominators for specific characteristics vary because of missing data.
2005–2010 Asian Cases Versus 2005–2010 Non-Asian Cases The median age among Asian case patients was significantly younger than non-Asian case patients (25 vs 40 years, P = .0001; Table 2). Asian case patients with illness onsets during 2005–2010 (n = 47) were significantly less likely to have certain occupational, recreational, and environmental exposures than non-Asian case patients with onsets during 2005–2010 (n = 97). These exposures include brush clearing (OR, 0.1; 95% CI, .0–.5), visiting a cabin (OR, 0.1; 95% CI, .0–.3), using an allterrain vehicle (OR, 0.1; 95% CI, .0–.5), owning a dog (OR, 0.1; 95% CI, .0–.3), and having exposure to an excavation or construction site (OR, 0.2; 95% CI, .0–.6). Twenty-one percent of 2005–2010 Asian case patients had another household member diagnosed with blastomycosis compared to 2% of non-Asian patients (OR, 10.9; 95% CI, 2.2–54.4). 2010 Asian Cases Versus 2005–2009 Asian Cases The 2010 Asian case patients were significantly more likely to have a household member diagnosed with blastomycosis
We investigated a large increase in reported cases of blastomycosis among Marathon County, Wisconsin, residents with cases geographically clustered within neighborhoods and households. A common outdoor recreational activity did not appear to be associated with the observed increase in cases, in contrast to descriptions of previous Wisconsin outbreaks [3, 6, 7]. A disproportionately high proportion of case patients were of Hmong ethnicity, and the blastomycosis incidence rate among Asians was approximately 12 times that for non-Asians in 2010. Of note, the incidence rate was similarly high among Asians in 2006, when an outbreak occurred in a neighboring county. The 2006 outbreak was attributed to exposure to airborne conidia originating from a central yard waste site [4], and the local environmental conditions favorable for B. dermatitidis growth and dispersal associated with this outbreak could have also contributed to increased sporadic cases in Marathon County. A similar phenomenon could account for the observed increase in 2010. This investigation did not yield evidence for a single common environmental point source of B. dermatitidis. The typical soil-related exposures that have been identified in previously reported outbreaks of blastomycosis are unlikely to explain the increase described here. Although several prior Wisconsin outbreaks were associated with recreational activities near waterways [3, 6–8], our analysis revealed that outbreak case patients were not more likely to engage in typical high-risk exposure activities including camping, fishing, and hiking, compared to case patients from previous years. Cluster case
Largest US Blastomycosis Outbreak
•
CID 2013:57 (1 September)
•
659
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
12 (63)
Died Immunocompromised
a
(P = .003) and were less likely to have traveled during the 90 days prior to illness onset (P = .004) than Asian case patients from previous years (Table 3). No other variables in this casecase comparison were statistically different.
Table 4. Univariate Analysis of Risk Factors for Blastomycosis Among Patients With Outbreak-Associated Cluster Cases and Matched Controls, Marathon County, Wisconsin, 2009–2010 Characteristic
Cases (n = 26)a
Controls (n = 78)a
16 (62) 28 (3–59)
28 (36) 36 (2–71)
OR (95% CI)
P Value
Demographic Male Age, y, median (range)
2.8 (1.0–8.4) NA
.05b .14b .001b
Race White Hmong
11 (42) 15 (58)
54 (69) 24 (31)
18.4 (2.4–828.1)
7 (27) 11 (42)
18 (23) 35 (45)
1.6 (.3–12.9) 0.8 (.2–4.0)
.85 .99
Neighborhood and residence Rural or wooded neighborhood Underdeveloped or partially developed yard
8 (42)
22 (39)
2.0 (.3–21.8)
.68
Wood-burning stove or fireplace Chop or gather firewood
3 (12) 4 (15)
16 (21) 21 (28)
0.2 (.0–2.1) 0.5 (.1–1.7)
.31 .31
Wood or brush clearing/cutting
1 (4)
12 (15
0.2 (0–1.5)
.13
Spend >20 h outdoors per wk near home Collect or transport yard waste
7 (28) 9 (35)
15 (19) 25 (32)
1.6 (.5–8.0) 1.2 (.3–4.0)
.35 .81
12 (46) 8 (31)
45 (58) 18 (23)
0.5 (.2–1.7) 1.6 (.5–5.5)
.36 .55
Hiking
6 (23)
12 (15)
1.8 (.4–7.1)
.52
Hunting Camping
1 (4) 1 (4)
6 (8) 1 (1)
0.9 (.0–4.5) 3.0 (.0–235.5)
.47 .88
Spend >20 h/wk away from home, school, work
8 (31)
18 (23)
1.6 (.5–5.2)
.57
Non-Western medical treatment Attend any group cultural events or gatherings
4 (15) 8 (31)
2 (3) 33 (42)
10.2 (.9–512.1) 0.5 (.13–1.7)
.06 .33
Attend any group religious services or ceremonies
7 (27)
25 (33)
0.8 (.22–2.3)
.77
Travel and recreational activities Travel to other WI counties Fishing
Other outdoor and occupational exposures Near excavation or construction site
14 (54)
26 (33)
4.2 (1.1–19.9)
Gardening, landscaping, or farming at home
7 (27)
21 (27)
1.0 (.2–3.9)
1.00
.04
Gardening, landscaping, or farming away from home Exposure to soil or wooded/boggy areas
4 (15) 6 (23)
6 (8) 25 (32)
2.3 (.4–12.9) 0.6 (.1–2.0)
.42 .50
Use homemade compost
3 (12)
6 (8)
1.6 (.2–10.2)
.80
Currently own a dog Underlying medical conditions
5 (19)
28 (36)
0.3 (.1–1.3)
.12
Chronic medical conditionc
6 (23)
4 (5)
5.4 (1.1–34.0)
.03b
Smoking history
2 (8)
9 (12)
0.6 (.1–3.7)
.73
Data are presented as No. (%) unless otherwise specified. Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio; WI, Wisconsin. a
Denominators for specific characteristics vary because of missing data.
Included in the multivariable model. Sex: adjusted OR (AOR), 0.45 (95% CI, .14–1.4), P = .19; Hmong: AOR,12.1 (95% CI, 1.3–611.9), P = .019; age: AOR, 0.94 (95% CI, .88–.99), P = .027; chronic medical conditions: AOR, 10.6 (95% CI, 1.5–132.6), P = .013. b
c
Emphysema, chronic obstructive pulmonary disease, chronic bronchitis, asthma, sarcoidosis, other lung disease, congestive heart failure, cancer, or diabetes.
patients were not more likely to engage in outdoor activities compared to control patients. Furthermore, we found that Asian case patients generally participated in fewer outdoor recreational or occupational activities (compared to non-Asian case patients) during the outbreak period and previous years. Previous epidemiologic investigations in other states, including Louisiana, Missouri, and Mississippi, have also demonstrated
660
•
CID 2013:57 (1 September)
•
Roy et al
that in outbreak settings, patients with blastomycosis were not always more likely to engage in occupational or recreational outdoor activities than controls [9–11]. Additionally, because many case patients had illness onsets during late winter and early spring of 2009, exposure to soil was unlikely because snow covered much of the soil in Wisconsin during these months.
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Live within 1/4-mile from a swamp or wetland
Genetic predisposition to blastomycosis as a result of impaired host–pathogen response is another possible explanation, and deserves further study [18]. Interleukin 12 receptor β1 deficiency and dominant interferon-γ receptor mutations have been reported in patients with disseminated coccidioidomycosis; the latter genetic deficiency has also been reported in patients with disseminated histoplasmosis and paracoccidioidomycosis [19, 20]. During this outbreak, the number of patients who developed disseminated blastomycosis was small; however, a similar genetic mechanism may contribute to susceptibility to acquiring pulmonary infection among Hmong individuals. Further evaluation of the role of potential immune deficits in blastomycosis acquisition is warranted. One limitation of the study is that a large number of statistical tests were performed without adjustment for multiple testing, and hence some of the findings may be spurious, as the type I error rate is inflated. Because there are no clear strategies to prevent blastomycosis and many of the signs and symptoms of blastomycosis are virtually indistinguishable from those of other lower respiratory infections, there is a need to increase community and healthcare provider awareness of this disease throughout endemic areas. Increased awareness of blastomycosis may facilitate earlier detection of cases and prompt initiation of appropriate treatment. Further research is needed to evaluate genetic differences in key proteins involved in the immune response to Blastomyces among different populations. Notes Acknowledgments. The authors thank Gordana Derado, PhD, Division of Foodborne, Waterborne, and Environmental Diseases, Centers for Disease Control and Prevention (CDC), for her statistical assistance. Financial support. This work was supported by the CDC. Disclaimer. The findings and conclusion of this article are those of the authors and do not necessarily represent the official position of the CDC. Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References 1. Bradsher RW, Chapman SW, Pappas PG. Blastomycosis. Infect Dis Clin North Am 2003; 17:21–40, vii. 2. Klein BS, Vergeront JM, Davis JP. Epidemiologic aspects of blastomycosis, the enigmatic systemic mycosis. Semin Respir Infect 1986; 1:29–39. 3. Klein BS, Vergeront JM, Weeks RJ, et al. Isolation of Blastomyces dermatitidis in soil associated with a large outbreak of blastomycosis in Wisconsin. N Engl J Med 1986; 314:529–34. 4. Pfister JR, Archer JR, Hersil S, et al. Non-rural point source blastomycosis outbreak near a yard waste collection site. Clin Med Res 2011; 9:57–65. 5. Karon J, Veroff D. Wisconsin’s Hmong population: census 2000 population and other demographic trends. Madison: University of Wisconsin Extension and Applied Population Laboratory. 6. Cockerill FR 3rd, Roberts GD, Rosenblatt JE, Utz JP, Utz DC. Epidemic of pulmonary blastomycosis (Namekagon fever) in Wisconsin Canoeists. Chest 1984; 86:688–92.
Largest US Blastomycosis Outbreak
•
CID 2013:57 (1 September)
•
661
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
Results of our epidemiologic investigation did not suggest transmission of B. dermatitidis from a single environmental source. Instead, it is possible that multiple environmental foci existed, and increased aerosolization of the organism may have occurred as a result of weather conditions, construction activities, or other factors. Transmission of B. dermatitidis conidia from multiple environmental foci could have resulted in the observed clustering within specific households and neighborhoods. Prior reports have demonstrated that B. dermatitidis infection may be acquired at home [12]. However, because our case-control study was designed to examine possible risk factors within neighborhoods, we are not able to determine whether specific neighborhoods were associated with increased risk. Whereas exposure near one’s home may explain the neighborhood clustering of blastomycosis demonstrated during this outbreak, we also observed household clustering. Household clustering may be attributed to the larger number of household members in Hmong households compared to non-Hmong households. [5] Our case-control study was not able to independently evaluate the role of household size as ethnicity would likely confound the relationship between household size and case status. This is the first report describing Asian race associated with an increased risk of acquiring symptomatic B. dermatitidis infection. Racial and ethnic differences in the occurrence of blastomycosis have been described; specifically, high incidence rates have been observed among persons of aboriginal ethnicity in Canada, and among black persons in some US states [11, 13, 14]. The reasons for the disproportionately high number of Hmong persons affected, both during this outbreak and in previous years, are still unclear. Asian case patients were not more likely to engage in the types of outdoor activities that have previously been associated with blastomycosis. A small number of studies suggest the Hmong may be at increased risk of developing certain illnesses, particularly cancer [15]. Glucose intolerance and undiagnosed diabetes mellitus have been associated with the Hmong population in Wisconsin [16]. In our study, only 6 cluster case patients reported a history of diabetes, 3 of whom were Hmong. The prevalence of glucose intolerance or undiagnosed diabetes among cluster case patients is not known. Because blastomycosis is not endemic in Southeast Asia, Hmong persons may lack naturally acquired immunity to this disease and may have subsequently been more susceptible to B. dermatitidis infection. However, in Marathon County, many of the Hmong case patients had been living in Wisconsin for more than a decade, and many cases occurred in Hmong children who were born in the United States; both groups thus probably had ample time to develop natural immunity to B. dermatitidis infection. Unmeasured sociocultural factors, including differences in behaviors regarding Western medical care (which may affect overall health status), may have also been a contributing factor [17].
7. Klein BS, Vergeront JM, DiSalvo AF, Kaufman L, Davis JP. Two outbreaks of blastomycosis along rivers in Wisconsin. Isolation of Blastomyces dermatitidis from riverbank soil and evidence of its transmission along waterways. Am Rev Respir Dis 1987; 136:1333–8. 8. Baumgardner DJ, Burdick JS. An outbreak of human and canine blastomycosis. Rev Infect Dis 1991; 13:898–905. 9. Cano MV, Ponce-de-Leon GF, Tippen S, Lindsley MD, Warwick M, Hajjeh RA. Blastomycosis in Missouri: epidemiology and risk factors for endemic disease. Epidemiol Infect 2003; 131:907–14. 10. Chapman SW, Lin AC, Hendricks KA, et al. Endemic blastomycosis in Mississippi: epidemiological and clinical studies. Semin Respir Infect 1997; 12:219–28. 11. Lowry PW, Kelso KY, McFarland LM. Blastomycosis in Washington Parish, Louisiana, 1976–1985. Am J Epidemiol 1989; 130:151–9. 12. Baumgardner DJ, Paretsky DP. Blastomycosis: more evidence for exposure near one’s domicile. WMJ 2001; 100:43–5. 13. Crampton TL, Light RB, Berg GM, et al. Epidemiology and clinical spectrum of blastomycosis diagnosed at Manitoba hospitals. Clin Infect Dis 2002; 34:1310–6.
14. Kralt D, Light B, Cheang M, et al. Clinical characteristics and outcomes in patients with pulmonary blastomycosis. Mycopathologia 2009; 167: 115–24. 15. Kiffmeyer WR, Langer E, Davies SM, Envall J, Robison LL, Ross JA. Genetic polymorphisms in the Hmong population: implications for cancer etiology and survival. Cancer 2004; 100:411–7. 16. Her C, Mundt M. Risk prevalence for type 2 diabetes mellitus in adult Hmong in Wisconsin: a pilot study. WMJ 2005; 104:70–7. 17. Fadiman A. The spirit catches you and you fall down. New York: Farrar, Straus and Giroux, 1997. 18. Klein BS. Molecular basis of pathogenicity in Blastomyces dermatitidis: the importance of adhesion. Curr Opin Microbiol 2000; 3:339–43. 19. Moraes-Vasconcelos D, Grumach AS, Yamaguti A, et al. Paracoccidioides brasiliensis disseminated disease in a patient with inherited deficiency in the beta1 subunit of the interleukin (IL)-12/IL-23 receptor. Clin Infect Dis 2005; 41:e31–7. 20. Zerbe CS, Holland SM. Disseminated histoplasmosis in persons with interferon-gamma receptor 1 deficiency. Clin Infect Dis 2005; 41: e38–41.
Downloaded from http://cid.oxfordjournals.org/ at University of Manchester on August 15, 2013
662
•
CID 2013:57 (1 September)
•
Roy et al
