Abstract
Background
The burden of leptospirosis in Puerto Rico remains unclear due to underreporting.
Methods
A cross-sectional survey and rodent trapping was performed in a community within San Juan, Puerto Rico to determine the seroprevalence and risk factors for Leptospira infection. The microscopic agglutination test was used to detect anti-Leptospira antibodies as a marker of previous infection. We evaluated Leptospira carriage by quantitative polymerase chain reaction among rodents trapped at the community site.
Results
Of 202 study participants, 55 (27.2%) had Leptospira agglutinating antibodies. Among the 55 seropositive individuals, antibodies were directed most frequently against serogroups Icterohaemorrhagiae (22.0%) and Autumnalis (10.6%). Of 18 captured rodents, 11 (61.1%) carried pathogenic Leptospira (Leptospira borgpetersenii, 7 and Leptospira interrogans, 2). Four participants showed their highest titer against an isolate obtained from a rodent (serogroup Ballum). Increasing household distance to the canal that runs through the community was associated with decreased risk of infection (odds ratio = 0.934 per 10-meter increase; 95% confidence interval, .952–.992).
Conclusions
There are high levels of Leptospira exposure in an urban setting in Puerto Rico, for which rodents may be an important reservoir for transmission. Our findings indicate that prevention should focus on mitigating risk posed by infrastructure deficiencies such as the canal.
Keywords: leptospirosis, public health, Puerto Rico, rodent, seroprevalence
We found high levels of Leptospira seroprevalence (27.2% of 202 subjects) in an urban community of Puerto Rico for which rodents may be an important reservoir. Prevention should focus on mitigating risk posed by infrastructure deficiencies.
Leptospirosis is a worldwide distributed zoonotic disease with 1.03 million cases and 58 900 deaths estimated annually [1]. Pathogenic Leptospira thrive in the kidneys of rodents and other animals and are excreted in their urine, contaminating water and soil where they survive for weeks [2]. Infection occurs when Leptospira enter the body through abrasions on the skin or mucous membranes [3]. Leptospirosis manifestations range from asymptomatic infection to a self-limited acute febrile illness to life-threatening disease. An estimated 5%–15% of hospitalized leptospirosis patients die due to pulmonary hemorrhage, acute renal failure, multiorgan failure, or septic shock [4]. Leptospirosis has emerged as a global epidemic in impoverished urban communities [5–7] and is increasingly being recognized as an important cause of disaster-associated outbreaks exemplified by large epidemics after seasonal rainfall in Brazil [8] and hurricanes in Central America [9, 10] and Puerto Rico [11, 12].
After widespread damage after Hurricane Maria struck Puerto Rico in 2017, the large number of deaths attributable to this disaster [13] and the identification of leptospirosis as a cause of hurricane-associated mortality emphasize the importance of leptospirosis as a health problem in Puerto Rico. During 2014–2015, a total of 114 leptospirosis cases were reported to the Puerto Rico Department of Health (PRDH) [14]. It is noteworthy to mention that, during a concurrent dengue and leptospirosis epidemic in 2010, enhanced surveillance identified 26 fatal and 149 nonfatal leptospirosis cases throughout the island [15], although one third of the patients had not been diagnosed with leptospirosis and two thirds had not been reported to PRDH [12]. Therefore, leptospirosis transmission is likely to be widespread in impoverished communities within the island outside of the context of a disaster setting, but the true burden of the disease remains poorly characterized because of misdiagnosis and underreporting [11, 16–18]. In this context, there is a need for community-based studies to determine the transmission risk on the island and identify animal reservoirs and leptospiral strains implicated in spillover transmission.
In this study, we sought to better understand the prevalence of leptospirosis in Puerto Rico and performed a survey of Caño Martín Peña, a community in the city of San Juan that surrounds a canal that overflows during rainfall events [19]. The inadequate sewage discharge system and the illegal disposal of trash promote rodent infestation, which combined with frequent flooding and high population density, create ideal conditions for the transmission of leptospirosis [15]. We aimed to (1) estimate the seroprevalence of infection with Leptospira among residents of this community, (2) determine risk factors for human infection, and (3) investigate the carriage of pathogenic Leptospira among rodents.
METHODS
Study Site and Sampling Methodology
The Caño Martín Peña community (San Juan, Puerto Rico) has an area of 1.87 km2, and in 2010 it comprised 6638 households and 16 021 inhabitants (8567 inhabitants/km2). The neighborhood surrounds a naturally formed waterway that has become mostly stagnant due to trash and lack of sewage infrastructure leading to unsanitary conditions (Figure 1) [20]. Using geographic information system (GIS) mapping with 2010 Census Data, a study site was defined by selecting 4 census tracts close to the canal and affected by flooding (1.35 km2). The census tracts were divided in 160 blocks of similar size (approximately 25 houses per block), and 24 were randomly selected. Next, 50% of the houses in each selected block were randomly selected to aim for a sample size of 500 participants. Houses were geocoded and labeled using the mobile device application Map PDF (Avenza). During July–September 2015, study teams visited each selected house up to 3 times to attempt to offer residents enrollment. All household members aged 5 years or older who had slept at least 4 of the previous 7 nights in the sampled household and provided written informed consent to participate were enrolled in the study.
Figure 1.
(A) Satellite image of the city of San Juan, Puerto Rico, with a black rectangle delineating the Caño Martín Peña neighborhood. (B) Satellite image of the 4 sampled census tracts (purple) along the canal (blue). The number of subjects and the proportion of seropositive subjects within each studied block are indicated in colored circles. (C) An image of the canal showing accumulation of vegetation and debris and the close distance to households.
Head of Household and Individual Survey
The study team conducted interviews to obtain information on socioeconomic indicators, exposures to sources of environmental contamination, and potential reservoirs in and around the household. The head-of-household, defined as the member who self-identified as best able to answer questions about the household, was interviewed to determine total income for the household, the highest number of rats sighted, the presence of other animals, and frequency and severity of flooding in and around the household. In addition, each household member was interviewed individually to obtain information on demographic and socioeconomic indicators, employment and occupation, use of personal protective equipment, and exposures to potential sources of environmental contamination and reservoirs in the household and workplace.
Environmental Features
The area immediately surrounding the household was surveyed for signs of rodent infestation using an adapted version of a standardized form checking for open sewers, plant debris, animal food, standing water, rodent burrows, and refuse deposits [21]. The GIS coordinates of each household were collected and recorded during the household visit using Map PDF on mobile devices.
Serological Analysis
The study team collected blood from all volunteer participants. Human and rat serum were tested by microscopic agglutination test (MAT) to detect prior Leptospira infection using a panel of 28 reference strains (Supplementary Table S1). The panel included an isolate obtained from a rodent captured during the study. Screening was performed with serum dilutions of 1:50 and 1:100, and agglutination against a serum dilution at concentration of 1:50 or higher was considered a positive result. When agglutination was observed at a dilution of 1:100, the specimen was titrated to endpoint. The presumptive infecting serogroup was defined as the serogroup against which the highest agglutination titer was directed [22]. The serogroup was defined as mixed when specimens had the same highest agglutination titer for more than 1 pathogenic serogroup.
Rodent Trapping
Rodents were captured with medium-sized Tomahawk Live Traps (Tomahawk Live Traps) and small Sherman live traps (H.B. Sherman Traps) for 3 consecutive nights in August 2015. For the first night, 1 Tomahawk and 2 Sherman traps were set in the yards of 6 study participants households. For the subsequent 2 nights, traps were placed in communal trash areas and in a riparian zone adjacent to households. Traps were set at dusk in lines of 40 (80 traps total) and checked at dawn. Captured rodents were anesthetized with isoflurane and then euthanized by cervical dislocation. Rodent species were identified using morphologic characteristics in body shape and tail-to-body length ratios. Urine, blood, and kidney specimens were collected aseptically at necropsy. Kidney and blood specimens were stored at −80°C until analysis.
Isolation of Leptospira From Urine, Deoxyribonucleic Acid Extraction, and Quantitative Polymerase Chain Reaction
Urine specimens were inoculated after collection into 5 mL EMJH tubes and incubated at 29°C. Cultures were checked 3 times a week under dark-field microscopy for the presence of Leptospira for up to 28 days. For deoxyribonucleic acid (DNA) extraction, 10 mL of late-log cultures were harvested by centrifugation and extracted using the Maxwell 16 Cell DNA Purification Kit (Promega). For kidney specimens, DNA was prepared from 25 mg of previously frozen tissue, with the Maxwell 16 Tissue DNA Purification Kit. The concentration of pathogenic Leptospira in the rodent specimens was assessed using quantitative polymerase chain reaction (qPCR) targeting the lipL32 gene and reported as genome copies per gram of kidney (GC/g) [23]. An additional qPCR targeting a rodent housekeeping gene (gadph) was used to monitor for the presence of inhibitors [24].
Identification, Multilocus Sequence Typing, Serogrouping, and Virulence Testing
To identify Leptospira species, a fragment of the secY gene was amplified and sequenced from all qPCR-positive specimens and urine isolates using a modification of a previously described protocol [25]. A multilocus sequence typing (MLST) scheme for 7 loci was applied to the urine isolate and a selection of kidney extracts using a previously described method with some modifications [26]. The allelic profiles obtained were used to assign sequence types (STs) to all specimens using the Leptospira MLST website [27]. To determine the serogroup of the urine isolate, a MAT was conducted with 20 rabbit antisera for the major Leptospira serogroups. Finally, to determine the virulence of the urine isolate, we determined the LD50 (lethal dose, 50%) in Golden Syrian male hamsters via intraperitoneal and conjunctival infection (Supplementary Figure S1).
Statistical Methods
Statistical analyses were conducted using SAS Studio (SAS Institute Inc., Cary, NC) using P < .05 as the criterion for significance. Continuous variables were compared using Student’s t test. A generalized linear mixed model with a binomial distribution was built to identify the individual effect of factors on risk of infection and to account for clustering within blocks. Variables that were significant in univariate analysis were included in the multivariate regression. The outcome variable was seropositivity. Home flooding, home distance from canal, age, and sex were included as covariates. A random intercept for block grouping was included to capture unexplained variability in the data. The model was fit with PROC GLIMMIX. Average Leptospira concentrations in the kidney specimens were compared with a one-way analysis of variance with Tukey correction. Distance of each sampled household from the canal in Caño Martín Peña was calculated using the ArcGIS spatial analysis tools package.
Ethics Statements
Participants were enrolled according to written informed consent procedures approved by the Institutional Review Boards of the University of Puerto Rico and Yale University (protocol no. 1506016072). The protocol for collection and handling of wild rodents was approved by the Texas State University Institutional Animal Care and Use Committee (protocol no. 0514_0303_07), and collection was authorized by the Departamento de Recursos Naturales y Ambientales from Puerto Rico (protocol no. 2014-IC-063; to I. C.-A.). Access to private property was granted by landowners. Hamster protocols were approved by the Yale Institutional Animal Care and Use Committee guidelines (protocol no. 2014–11424).
RESULTS
Seroprevalence Survey
Among 11 963 eligible residents from the community site, 202 (1.7%) were enrolled in the study. Study participants had a higher proportion of females (60.1% vs 42.9%; P < .05) and older mean age (39.7 vs 48.1 years; P < .05) than eligible residents who did not participate in the study. Specifically, age groups 15–29 and 30–44 were slightly underrepresented (12.4% and 12.9% among enrolled subjects vs 20.6% and 19.6% among eligible residents). Participants’ median annual household income was $15 000, and 57.2% had not graduated from high school. Among 179 participants ≥18 years of age, 71 (39.7%) were employed. The enrollment rate was evenly distributed across the 4 selected census tracts (from 1.2% to 4.3% of the population), and each census tract comprised between 20 and 41 enrolled households with an average of 1.64 ± 0.15 individuals per household.
Among the 202 study participants, 55 (27.2%) had Leptospira agglutinating antibodies, as determined by the presence of MAT titer ≥1:50 (Figure 2). Of the 55 seropositive individuals, highest titers were directed against serogroups Icterohaemorrhagiae (n = 12, 22.0%), Autumnalis (n = 6, 10.9%), Mini (n = 4, 7.3%), the M2 urine isolate (subsequently identified as Leptospira borgpetersenii serogroup Ballum) (n = 4, 7.3%), Hebdomadis (n = 3, 5.5%), and Tarassovi (n = 3, 5.5%) (Figure 2). Of note, 3 participants had a MAT ≥1:400 against serogroup Autumnalis, which could be an indication of acute or recent infection [28]. For the 19 individuals (34.5%) who had highest titers against 2 or more pathogenic serogroups, MAT recognized Icterohaemorrhagiae (n = 11, 57.9%), Autumnalis (n = 6, 31.6%), and the M2 urine isolate (n = 5, 26.3%) as one of the serovars in the mixed titer. We did not observe differences regarding serovar distribution across census tracts because the most frequently identified serovars (Icterohaemorrhagiae, Autumnalis, and mixed) occurred in all census tracts. Likewise, no differences were observed between the median levels of antibody response between census tracts (P = .13).
Figure 2.
Distribution of microscopic agglutination test titers for 202 subjects from the Caño Martín Peña neighborhood. Labels above the bars indicate the number of subjects (% of total), according to their highest microscopic agglutination test (MAT) titer.
Risk Factors
Prevalence of Leptospira agglutinating antibodies was slightly but not statistically higher in males than females (32.5% vs 23.0%, respectively; P = .21) (Figure 3). Similarly, we did not observe significant differences in the prevalence of Leptospira antibodies across age groups (Table 1 and Figure 3). Increased household distance to the canal was associated with decreased risk of Leptospira infection both independently and after controlling for block effects, age, sex, and reported frequency of household flooding, with each 10 meters of distance from the canal associated with a 7.1% decrease in risk of infection (odds ratio [OR] = 0.934; 95% confidence interval [CI], .952–.992) (Table 2). Those who lived in a household that reported seldom, sometimes, often, or always flooding in the rain were 78% less likely to be exposed to Leptospira, compared to living in a household that never flooded (OR = 0.22; 95% CI, .08–.60) (Table 1). After adjusting for distance from the canal, block grouping, age, and sex, participants who reported flooding in their homes still had decreased risk of Leptospira exposure (OR = 0.12; 95% CI, .04–.37) (Table 2). In a generalized linear mixed model, participants living in one block of the 24 blocks sampled were significantly more likely to be seropositive (OR = 1.31; P = .017) than would be expected based on the covariates, demonstrating unexplained clustering of risk by block location.
Figure 3.
Seroprevalence of Leptospira antibodies in human participants in Caño Martín Peña in 2015 by age and sex, with 95% confidence intervals.
Table 1.
Associations Between Demographic and Behavioral Characteristics and Leptospira Seropositivity Among Residents of Caño Martín Peña, 2015
| MAT Result | |||||
|---|---|---|---|---|---|
| Seropositive | Seronegative | ||||
| Characteristic | Na | No. (%) | Na | No. (%) | OR (95% CI) |
| Individual Risk Factors | |||||
| Age (years) | 54 | 143 | |||
| 5–14 | 6 (11.1) | 13 (9.1) | ref. | ||
| 15–29 | 8 (14.8) | 17 (11.9) | 1.02 (.28–3.67) | ||
| 30–44 | 16 (29.6) | 20 (14.0) | 1.73 (.54–5.58) | ||
| 45–59 | 9 (16.7) | 40 (28.0) | 0.49 (.15–1.63) | ||
| ≥60 | 15 (27.8) | 53 (37.1) | 0.61 (.20–1.89) | ||
| Female Sex | 54 | 26 (48.2) | 148 | 54 (36.5) | 1.61 (.86–3.03) |
| Highest Education Level | 46 | 133 | 0.3494 | ||
| Elementary School or less | 11 (20.0) | 48 (32.2) | ref. | ||
| High school | 15 (27.3) | 28 (18.8) | 2.34 (.94–5.79) | ||
| College | 20 (36.4) | 57 (38.5) | 1.53 (.67–3.51) | ||
| Household-Related Factors | |||||
| Distance from canal (meters)b | 55 | 201 (172–226) | 146 | 247 (230–268) | 0.997 (.994–.999) |
| Contact with mud | 44 | 11 (25.0) | 134 | 24 (17.9) | 1.53 (.68–3.44) |
| Home flooding | 40 | 5 (12.5) | 119 | 47 (39.5) | 0.22 (.08–.60) |
| Street flooding | 50 | 22 (44.0) | 139 | 78 (73.5) | 0.61 (.32–1.18) |
| Household-Associated Behavioral Exposures | |||||
| Walk barefoot at home | 45 | 25 (55.6) | 134 | 69 (51.5) | 1.18 (.60–2.32) |
| Wear boots in floodwater | 18 | 10 (55.6) | 79 | 38 (48.1) | 1.35 (.48–3.77) |
| Contact with mud | 36 | 21 (58.3) | 93 | 53 (57.0) | 1.06 (.48–2.30) |
| Rodent-Related Exposures | |||||
| Rats sighted at home | 45 | 39 (86.7) | 133 | 111 (83.4) | 1.29 (.49–3.41) |
| Dog at home | 55 | 49 (89.1) | 148 | 132 (89.2) | 0.99 (.36–2.67) |
| Chicken coop at home | 40 | 15 (37.5) | 135 | 54 (40.0) | 0.90 (.44–1.86) |
| Occupation-Related Exposures | |||||
| Employed | 45 | 23 (51.1) | 123 | 44 (73.2) | 1.87 (.94–3.75) |
| Construction-related job | 55 | 5(9.1) | 147 | 5 (3.4) | 2.84 (.79–10.22) |
Abbreviations: CI, confidence interval; OR, odds ratio; ref., reference.
aThe N for each characteristic is variable because not all enrolled participants answered all questions.
bResults expressed as mean and 95% CI.
Table 2.
Multivariable Risk Factors for Leptospira Infection Adjusted for Block Groupings Among Residents of Caño Martín Peña, 2015
| Characteristic | OR (95% CI)a |
|---|---|
| Distance from canal (per 10-meter increase) | 0.934 (.952–.992) |
| Reported Household Flooding | |
| Yes | 0.12 (.04–.37) |
| No | ref. |
| Age (per 1-year increase) | 0.995 (.976–1.013) |
| Female sex | 0.85 (.37–1.95) |
Abbreviations: CI, confidence interval; OR, odds ratio; ref., reference.
aA generalized linear mixed model was fit in PROC GLIMMIX using a random intercept for block grouping and including home flooding, distance from canal, age, and sex as covariates.
Rodent Survey
A total of 18 rodents (5 Mus musculus, 10 Rattus norvegicus, and 3 Rattus rattus) were captured in Caño Martín Peña; pathogenic Leptospira were detected by qPCR in 11 (61.1%) (Table 3). The average concentration of Leptospira in positive kidneys in M musculus (2.17 × 108 GC/g) was significantly higher than the concentrations found in R rattus and R norvegicus (2.04 × 104 and 3.77 × 104 GC/g; P = .008 and P = .006, respectively).
Table 3.
Leptospira Concentration and Identification in Kidney Specimens From Rodents Trapped in Caño Martín Peña, 2015
| Rodent Species | Animal | Leptospira Concentration (Log10 GC/g) | secY Identification | MLST Type | MAT Highest Titer |
|---|---|---|---|---|---|
| Mus musculus | M1 | 8.93 | Leptospira borgpetersenii | ND | Mixed |
| M2 | 8.71 | L borgpetersenii | 149 | ND | |
| M3 | 8.50 | L borgpetersenii | 149 | Mixed | |
| M4 | 8.11 | L borgpetersenii | ND | Mixed | |
| M5 | 7.43 | L borgpetersenii | ND | Mixed | |
| Rattus norvegicus | N1 | 5.60 | L borgpetersenii | 149a | Icterohaemorrhagiae |
| N2 | 3.02 | Unsuccessful | ND | Icterohaemorrhagiae | |
| N3 | Negative | ND | ND | ND | |
| Rattus rattus | R1 | 3.16 | Unsuccessful | ND | ND |
| R2 | 4.34 | L borgpetersenii | 149a | Mixed | |
| R3 | 5.88 | Leptospira interrogans | Unsuccessful | Icterohaemorrhagiae | |
| R4 | 6.23 | L interrogans | 17a | Icterohaemorrhagiae | |
| R5 | Negative | ND | ND | Negative | |
| R6 | Negative | ND | ND | Mixed | |
| R7 | Negative | ND | ND | M2 isolate (Ballum) | |
| R8 | Negative | ND | ND | Icterohaemorrhagiae | |
| R9 | Negative | ND | ND | ND | |
| R10 | Negative | ND | ND | ND |
Abbreviations: GC, genome copies; MAT, microscopic agglutination test; MLST, multilocus sequence typing; ND, not determined.
aSequence type obtained from partial allelic profiles.
All M musculus, 1 R norvegicus (N1), and 1 R rattus (R2) secY sequences showed 99% identity to L borgpetersenii sequences in GenBank. The other 2 R rattus sequences (R3 and R4) were identical to Leptospira interrogans secY sequences (Table 3). The MLST results showed that the 2 mice analyzed had ST 149, which is shared by isolates identified as L borgpetersenii serogroup Ballum in the MLST database. In addition, 2 other rats (N1 and R2) showed partial similarity to ST 149 and 1 rat (R4) to ST 17, which is shared by L interrogans serogroup Icterohaemorrhagiae isolates (see accession numbers in Supplementary Table S2).
Leptospira Isolation
We obtained 1 isolate from a urine specimen obtained from M musculus M2, which was identified as L borgpetersenii serogroup Ballum based on the secY sequence (99% identity to L borgpetersenii sequences in GenBank), MLST ST (149, which is shared by other L borgpetersenii Ballum isolates), and serogrouping (maximum titer against antibodies from serogroup Ballum). The secY sequence and all the MLST genes were identical to the sequences from the corresponding kidney specimen. After testing the virulence of the M2 isolate in a hamster model of infection, we found that the LD50 was <5 bacteria via intraperitoneal infection and 2.37 × 106 bacteria via conjunctival infection, confirming that M2 isolate was a virulent strain (Supplementary Figure S1).
DISCUSSION
The burden of leptospirosis in Puerto Rico remains largely unknown, despite evidence that the disease may be underdiagnosed. We conducted a serosurvey in the Caño Martín Peña community of San Juan and found that a relatively high proportion (27.2%) of participating residents had been exposed to pathogenic Leptospira. Rodent trapping revealed that a high proportion of rodents sampled were carriers of pathogenic Leptospira, indicating their potential role as reservoirs of the pathogen. Proximity of residence to the canal and residence in a certain block area were associated with increased risk of exposure to Leptospira, suggesting that there are significant household factors in this community that are linked to infection.
Infection in Caño Martín Peña was common among both male and female children, adults, and the elderly, which is in contrast to the usual age and sex pattern of leptospirosis [28, 29]. Serogroup Icterohaemorrhagiae, which is common in other Caribbean countries [30], was the most frequently detected serovar. The 27.2% seroprevalence observed is among the highest reported in nonoccupational related serosurveys. By comparison, seroprevalence of 37% was observed among adult males in the Seychelles [31] and 38.2% and 23.9% of those living in flood-prone areas in Bangladesh and Laos, respectively [32, 33]. In Peru, up to 28.0% seroprevalence was reported in the Amazonian city of Iquitos [34], and in Trinidad seroprevalence ranged from 5% to 37% in rural villages [35]. However, unlike Puerto Rico, these studies were performed in developing countries. Taken together, our findings indicate that Leptospira infection is frequent among residents of Caño Martín Peña, which may pose a serious health burden in this community.
Rodents were identified as reservoirs of pathogenic Leptospira in Caño Martín Peña. This is in agreement with a recent study that found that rodents from rural areas of Puerto Rico (R norvegicus, R rattus, and M musculus) were carriers of L interrogans and L borgpetersenii [36]. Rodent infestation is a well recognized risk factor for urban leptospirosis [5, 29], and the identification of serogroups Icterohaemorrhagiae and Ballum in both humans and rodents suggest that rodents may be a source of spillover infection in this community. However, the specific contribution that each rodent species plays in the transmission of the pathogen in this setting deserves further investigation. Furthermore, a systematic characterization of other potential domestic and wild animal reservoirs is necessary to identify the sources of the other serogroups found in the study subjects. Moreover, this was the first documented isolation of a pathogenic Leptospira (L borgpetersenii Ballum strain M2) from any specimen, human or animal, in Puerto Rico in the past 50 years [37]. Positive MAT titers against this virulent strain are indicative of human exposure, although its role in human disease is yet to be determined.
The Caño Martín Peña canal was identified as a structural factor negatively associated with Leptospira infection in humans residing in this neighborhood. Increasing distance of residence away from the canal was a predictor of protection from infection both independently and after accounting for household flooding and block grouping effects (Table 2). Residents who live closer to the canal may be more likely to be exposed to the pathogen because of a higher degree of environmental contamination [38, 39] possibly related to higher rat infestation [40, 41], although this hypothesis needs further evaluation. In 2014, a health impact assessment determined that the canal posed substantial health risks and concluded that the Puerto Rico and federal governments should support dredging of the canal [20]; however, this activity is not yet underway. The results from our study contribute to the body of evidence suggesting that this structural risk should be mitigated.
Other household attributes and individual risk factors were not found to be associated with infection with Leptospira. However, reported household flooding was inversely associated with risk of Leptospira infection (Tables 1 and 2), in contrast to previous studies that point at flooding as a risk factor for Leptospira infection [32, 42]. Because flooding was measured through self-reporting, it is possible that the association observed represents a proxy of attitude toward flooding or that there is unmeasured confounding contributing to the observed effect. In addition, a cross-sectional design may not be appropriate to assess seasonal risk factors such as flooding, which vary in severity between months or years. Nevertheless, recent studies indicate that the association of leptospirosis and flooding may be more nuanced than previously thought. An inconsistent relationship between leptospirosis incidence and flooding was observed in Thailand [43], and prospective study in an urban slum in Brazil found that contact with mud, but not water, was a risk factor for infection [5]. These factors should be considered if serosurveys are repeated in Puerto Rico after natural diseases associated with flooding, such as hurricane Maria [13, 44].This study was limited by the cross-sectional design of the seroprevalence survey. Because risk exposures may change over time, seroprevalence observed herein may not be representative of the true burden of infection in the community. Now that infection with Leptospira has been confirmed as a serious problem in this community, a prospective study should be conducted to ascertain incidence of infection and disease and thereby obtain more reliable estimates of risk associations. Last, although this study only examined 1 community, other urban areas of Puerto Rico could face similarly high risk of leptospirosis. To identify modes of transmission in communities with high seroprevalence and develop relevant interventions, studies should be conducted in both rural and urban areas of Puerto Rico to identify region-specific risk factors for infection.
CONCLUSIONS
In conclusion, our study shows that there are high levels of exposure to Leptospira among Caño Martín Peña residents for which rodents may be an important reservoir, supporting prior observations that the burden of leptospirosis is likely underestimated in Puerto Rico. These findings should be used to inform public health policy, including increasing healthcare provider awareness about leptospirosis to improve diagnosis, as well as increasing efforts to dredge the canal. Such interventions would be expected to have a direct impact on reducing the burden of leptospirosis in this community.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments. We thank Danielle Bloch, Maria Terrón, Patricia Castro, Jonathan Chapman, Frances Luna, Fabiola Cruz, Pilar Torres, Joyce Meadows, Rossana Acosta, Madison Torres, and Joey Martinez for their support of the serosurvey. We also thank the G8 resident associations that comprise ENLACE of Caño Martín Peña and community leaders and residents, whose support was integral to completion of this study.
Disclaimer. The conclusions, findings, and opinions expressed by authors contributing to this journal do not necessarily reflect the official position of the US Department of Health and Human Services or the Centers for Disease Control and Prevention.
Financial support. This work was funded by the National Institutes of Health research grants (R01 AI052473, U01 AI088752, R25 TW009338, R01 TW009504, and R01 AI121207). E. A. B. was supported by the Yale University Downs Fellowship and the Yale School of Medicine Office of Student Research Summer Fellowship.
Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.
Presented in part: Yale University School of Public Health, Department of Epidemiology of Microbial Diseases Thesis Symposium, New Haven, CT, April 2016.
References
- 1. Costa F, Hagan JE, Calcagno J, et al. Global morbidity and mortality of leptospirosis: a systematic review. PLoS Negl Trop Dis 2015; 9:e0003898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Casanovas-Massana A, Pedra GG, Wunder EA, Diggle PJ, Begon M, Ko AI. Quantification of Leptospira interrogans survival in soil and water microcosms. Appl Environ Microbiol 2018; 84:e00507-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ko AI, Goarant C, Picardeau M. Leptospira: the dawn of the molecular genetics era for an emerging zoonotic pathogen. Nat Rev Microbiol 2009; 7:736–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Bharti AR, Nally JE, Ricaldi JN, et al. Leptospirosis: a zoonotic disease of global importance. Lancet Infect Dis 2003; 3:757–71. [DOI] [PubMed] [Google Scholar]
- 5. Hagan JE, Moraga P, Costa F, et al. Spatiotemporal determinants of urban leptospirosis transmission: four-year prospective cohort study of slum residents in Brazil. PLoS Negl Trop Dis 2016; 10:e0004275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kyobutungi C, Ziraba AK, Ezeh A, Yé Y. The burden of disease profile of residents of Nairobi’s slums: results from a demographic surveillance system. Popul Health Metr 2008; 6:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Riley LW, Ko AI, Unger A, Reis MG. Slum health: diseases of neglected populations. BMC Int Health Hum Rights 2007; 7:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ko AI, Galvão Reis M, Ribeiro Dourado CM, Johnson WD Jr, Riley LW. Urban epidemic of severe leptospirosis in Brazil. Salvador Leptospirosis Study Group. Lancet 1999; 354:820–5. [DOI] [PubMed] [Google Scholar]
- 9. Trevejo RT, Rigau-Perez JG, Ashford DA, et al. Epidemic leptospirosis associated with pulmonary hemorrhage-Nicaragua, 1995. J Infect Dis 1998; 178:1457–63. [DOI] [PubMed] [Google Scholar]
- 10. Naranjo M, Suárez M, Fernández C, et al. Study of a leptospirosis outbreak in honduras following hurricane mitch and prophylactic protection of the vax-SPIRAL® Vaccine. MEDICC Rev 2008; 10:38–42. [DOI] [PubMed] [Google Scholar]
- 11. Sanders EJ, Rigau-Pérez JG, Smits HL, et al. Increase of leptospirosis in dengue-negative patients after a hurricane in Puerto Rico in 1996 [correction of 1966]. Am J Trop Med Hyg 1999; 61:399–404. [DOI] [PubMed] [Google Scholar]
- 12. Sanders EJ, Rigau-Perez JG, Smits H, et al. Hurricane-related leptospirosis in dengue-negative patients, Puerto Rico, 1996. Clin Infect Dis 1997; 25:411. [Google Scholar]
- 13. Kishore N, Marqués D, Mahmud A, et al. Mortality in Puerto Rico after Hurricane Maria. N Engl J Med 2018; 379:162–70. [DOI] [PubMed] [Google Scholar]
- 14. Adams DA, Thomas KR, Jajosky RA, et al. Summary of notifiable infectious diseases and conditions - United States, 2015. MMWR Morb Mortal Wkly Rep 2017; 64:1–143. [DOI] [PubMed] [Google Scholar]
- 15. Sharp TM, Rivera García B, Pérez-Padilla J, et al. Early indicators of fatal leptospirosis during the 2010 epidemic in Puerto Rico. PLoS Negl Trop Dis 2016; 10:e0004482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bruce MG, Sanders EJ, Leake JA, et al. Leptospirosis among patients presenting with dengue-like illness in Puerto Rico. Acta Trop 2005; 96:36–46. [DOI] [PubMed] [Google Scholar]
- 17. Rivera B, Bower WA, Guerra M, et al. Investigation of Leptospirosis underreporting — Puerto Rico, 2010. MMWR Morb Mortal Wkly Rep 2012; 61:421. [PubMed] [Google Scholar]
- 18. Torres H, Garcia-Rivera EJ, Gorbea H, et al. Leptospirosis cases infected with uncommon serogroups, Puerto Rico, 2013–2015. Am J Trop Med Hyg 2018; 98:258–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Sheffield P, Rowe M, Agu D, Rodríguez L, Avilés K. Health impact assessments for environmental restoration: the case of Caño Martín Peña. Ann Glob Health 2014; 80:296–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Sheffield P, Rowe M, Agu D, Rodríguez L, Avilés K. Health impact assessments for environmental restoration: the case of Caño Martín Peña. Ann Glob Health 2014; 80:296–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Centers for Disease Control and Prevention. Integrated pest management: conducting urban rodent surveys. Atlanta: US Department of Health and Human Services, 2006. [Google Scholar]
- 22. World Health Organization. Human leptospirosis: guidance for diagnosis, surveillance and control. Malta: World Health Organization, 2003: pp 122. [Google Scholar]
- 23. Stoddard RA, Gee JE, Wilkins PP, McCaustland K, Hoffmaster AR. Detection of pathogenic Leptospira spp. through TaqMan polymerase chain reaction targeting the LipL32 gene. Diagn Microbiol Infect Dis 2009; 64:247–55. [DOI] [PubMed] [Google Scholar]
- 24. Costa F, Wunder EA Jr, De Oliveira D, et al. Patterns in leptospira shedding in Norway rats (Rattus norvegicus) from Brazilian slum communities at high risk of disease transmission. PLoS Negl Trop Dis 2015; 9:e0003819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Victoria B, Ahmed A, Zuerner RL, et al. Conservation of the S10-spc-alpha locus within otherwise highly plastic genomes provides phylogenetic insight into the genus Leptospira. PLoS One 2008; 3:e2752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Boonsilp S, Thaipadungpanit J, Amornchai P, et al. A single multilocus sequence typing (MLST) scheme for seven pathogenic Leptospira species. PLoS Negl Trop Dis 2013; 7:e1954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Jolley KA, Maiden MC. BIGSdb: scalable analysis of bacterial genome variation at the population level. BMC Bioinformatics 2010; 11:595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Lau CL, Watson CH, Lowry JH, et al. Human leptospirosis infection in Fiji: an eco-epidemiological approach to identifying risk factors and environmental drivers for transmission. PLoS Negl Trop Dis 2016; 10:e0004405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Felzemburgh RD, Ribeiro GS, Costa F, et al. Prospective study of leptospirosis transmission in an urban slum community: role of poor environment in repeated exposures to the Leptospira agent. PLoS Negl Trop Dis 2014; 8:e2927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Adesiyun AA, Baboolal S, Suepaul S, Dookeran S, Stewart-Johnson A. Human leptospirosis in the Caribbean, 1997–2005: characteristics and serotyping of clinical samples from 14 countries. Rev Panam Salud Publica 2011; 29:350–7. [PubMed] [Google Scholar]
- 31. Bovet P, Yersin C, Merien F, Davis CE, Perolat P. Factors associated with clinical leptospirosis: a population-based case-control study in the Seychelles (Indian Ocean). Int J Epidemiol 1999; 28:583–90. [DOI] [PubMed] [Google Scholar]
- 32. Kawaguchi L, Sengkeopraseuth B, Tsuyuoka R, et al. Seroprevalence of leptospirosis and risk factor analysis in flood-prone rural areas in Lao PDR. Am J Trop Med Hyg 2008; 78:957–61. [PubMed] [Google Scholar]
- 33. Morshed MG, Konishi H, Terada Y, Arimitsu Y, Nakazawa T. Seroprevalence of leptospirosis in a rural flood prone district of Bangladesh. Epidemiol Infect 1994; 112:527–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Johnson MA, Smith H, Joseph P, et al. Environmental exposure and leptospirosis, Peru. Emerg Infect Dis 2004; 10:1016–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Everard CO, Hayes RJ, Fraser-Chanpong GM. A serosurvey for leptospirosis in Trinidad among urban and rural dwellers and persons occupationally at risk. Trans R Soc Trop Med Hyg 1985; 79:96–105. [DOI] [PubMed] [Google Scholar]
- 36. Benavidez KM, Guerra T, Torres M, et al. The prevalence of Leptospira among invasive small mammals on Puerto Rican cattle farms. PLoS Negl Trop Dis 2019; 13:e0007236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Alexander AD, Benenson AS, Byrne RJ, et al. Leptospirosis in Puerto Rico. Zoonoses Res 1963; 2:152–227. [PubMed] [Google Scholar]
- 38. Casanovas-Massana A, Costa F, Riediger IN, et al. Spatial and temporal dynamics of pathogenic Leptospira in surface waters from the urban slum environment. Water Res 2018; 130:176–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Schneider AG, Casanovas-Massana A, Hacker KP, et al. Quantification of pathogenic Leptospira in the soils of a Brazilian urban slum. PLoS Negl Trop Dis 2018; 12:e0006415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Feng AY, Himsworth CG. The secret life of the city rat: a review of the ecology of urban Norway and black rats (Rattus norvegicus and Rattus rattus). Urban Ecosyst 2014; 17:149–62. [Google Scholar]
- 41. Panti-May JA, Carvalho-Pereira TS, Serrano S, et al. A two-year ecological study of Norway rats (Rattus norvegicus) in a Brazilian urban slum. PLoS One 2016; 11:e0152511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Goarant C, Laumond-Barny S, Perez J, Vernel-Pauillac F, Chanteau S, Guigon A. Outbreak of leptospirosis in New Caledonia: diagnosis issues and burden of disease. Trop Med Int Health 2009; 14:926–9. [DOI] [PubMed] [Google Scholar]
- 43. Suwanpakdee S, Kaewkungwal J, White LJ, et al. Spatio-temporal patterns of leptospirosis in Thailand: is flooding a risk factor? Epidemiol Infect 2015; 143:2106–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Concepción-Acevedo J, Patel A, Luna-Pinto C, et al. Initial public health laboratory response after Hurricane Maria - Puerto Rico, 2017. MMWR Morb Mortal Wkly Rep 2018; 67:333–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.