Serological Evidence of Exposure to Leptospira spp. in Veterinary Students and Other University Students in Trinidad and Tobago

The study compared the serological evidence of leptospirosis in 212 students in four schools (veterinary, dental, advanced nursing education and pharmacy) of the University of the West Indies (UWI), by testing for IgG immunoglobulins to Leptospira spp. using the enzyme-linked immunosorbent assay (ELISA) and the microscopic agglutination test (MAT). Overall, of 212 students tested by the ELISA, 12 (5.7%) and 31 (14.6%) were positive and borderline, respectively. Amongst the 113 veterinary students 11 (9.7%) and 19 (16.8%) were seropositive and borderline respectively compared with nonveterinary students with corresponding values of 1 (1.0%) and 12 (12.1%). The frequency of serological evidence of leptospirosis by the ELISA was statistically significantly (P < 0.05; χ 2) higher in veterinary students, 26.5% (30 of 113) than in nonveterinary students, 13.1% (13 of 99). By the MAT, the seropositivity for leptospirosis was similar for veterinary students, 7.1% (8 of 113) and nonveterinary students, 7.1% (7 of 99). For veterinary students, the prevalent infecting serovar was Icterohaemorrhagiae Copenhageni while amongst nonveterinary students, the prevalent serovar was Australis Rachmati. Being a veterinary student was the only risk factor that was significantly associated with Leptospira infection indicating that veterinary students need to be cognizant and to practise preventive measures for leptospirosis.


Introduction
Leptospirosis is a bacterial zoonosis with global distribution, although it has been documented in developing and developed as well as temperate and tropical countries [1,2]. is disease occurs predominantly as a subclinical infection although cases of clinical disease with numerous nonpathognomonic signs and symptoms have been reported [1][2][3][4]. It is therefore responsible for morbidities and mortalities worldwide [2,5,6].
It has been established that rodents are primary reservoirs for human and animal infections by Leptospira spp. which is one of the reasons that the prevalence of leptospirosis is higher in tropical environments with high rainfall and humidity and prevalent poor sanitary conditions which support the proliferation of rodents [7][8][9]. e distribution of the primary reservoir and the ability of the pathogen to infect animals (livestock, wildlife, pet animals, and others) have made leptospirosis an occupational disease [9][10][11].
High risk individuals include livestock farmers, animal handlers, veterinarians, slaughter house workers, sewerage or environmental sanitation workers, sugar cane, and rice �eld workers, compared to members of the general population [9,11,12]. Human infections are known to result from direct and indirect contact with urine of rodents or other animals containing high numbers of viable leptospires or following consumption of contaminated food or water [13,14]. A number of factors, in addition to occupational exposure, have been reported to affect exposure potential in humans to leptospirosis. ese factors include the age, gender, season of the year, and geographical locations and have been known to affect the infection rate in humans [15,16].
Diagnosis of human leptospirosis can be achieved through the demonstration of the microorganism itself or by the detection of antibodies produced against the pathogen following infection [13,17]. e organism can be demonstrated by culture in growth media, special staining of infected tissues, and the use of dark-�eld microscopy [13,18].
ere is a wide variety of serological tests which can detect IgG, IgM, and IgA [19]. Some of these tests include the enzyme-linked immunosorbent assay (ELISA) [20] and microscopic agglutination test (MAT) [21] amongst others. e MAT is considered the "gold standard" for the serological diagnosis of leptospirosis and also for the serotyping of Leptospira isolates [21,22]. e advantages and disadvantages of the MAT as a diagnostic test are well documented in the literature [21][22][23]. e ELISA is easy to perform, rapid, with a high sensitivity, and amenable to be standardized but the speci�city is low and it is genus-speci�c and, therefore, unlike the MAT, cannot be used to serotype infecting leptospires [24,25].
In Trinidad and Tobago, several reports exist on the prevalence of leptospirosis with the detection or isolation in school children [26], apparently healthy individuals [27] and in piggery farm workers [28]. Mohan et al. [29] conducted a retrospective study which determined the average annual incidence rate of leptospirosis and indicated that rate was affected by season, gender and age. In a study conducted on sugarcane �eld workers, Adesiyun et al. [30] reported the seroprevalence of leptospirosis to be 0.7%. To date there is no report on the occurrence of leptospirosis in students at tertiary institutions in the country.
e speci�c objectives of the current study were to compare the frequency of serological evidence of leptospirosis in veterinary students with nonveterinary students in the Faculty of Medical Sciences, to determine the infecting serovars of Leptospira spp. and to investigate the important risk factors for leptospirosis in these students.

Study
Design. e study design involved the sampling of students of each of the schools who volunteered to participate and then comparing the serological evidence of veterinary students with those of nonveterinary students.

Determination of Sample
Size. e sample size for the student population will be determined using the formula = 2 × 2 [31], where = , = precision at type 1 error of 0.04, = prevalence = 11% [28], and = estimated sample size. Since this equation is based on an in�nite population, was adjusted to suit a de�nite population of 698 student population in four schools (SOD, SVM, SANE, and SOP) in the FMS, using the formula adj = × , where adj was the number of humans required to estimate prevalence at the same absolute precision as the �rst equation. e estimated sample size used in the current study was therefore: = 2 ×0 0 0 04 2 = 3 84 × 0 0 7 0 00 = 235 and the minimum adjusted sample size was adj = 235 × 8 235 8 = 7 .

Administration of Questionnaire.
A questionnaire was administered to each participant in order to obtain information including the programme of study, year in the programme, age, gender, place of residence, ownership of pet animals (dogs, cats, and rodents), association with livestock, farming activities outside the FMS, and other risk factors. Each participant was allowed to pick a random number which was used to identify the samples and the test results of the participant. To maintain the con�dentiality of the study, email addresses were obtained to convey the results of the study to each participant.

Collection of Samples. �uali�ed phlebotomists and
nurses from the SANE assisted in blood collection from participants. Approximately 5 mL of blood was drawn from either the median cubital or cephalic veins of the arm using a 21-gauge one and a half inch needle attached to a 5 mL syringe. e blood was then placed into tubes without anticoagulant. Collected blood was refrigerated overnight at 4 ∘ C aer which it was centrifuged and serum harvested and stored at 20 ∘ C until tested. is was a serological study to detect exposure experience of Leptospira spp. amongst the students studied using both the ELISA and MAT and therefore no attempt was made to culture the blood samples for Leptospira spp. e MAT used consists of two parts, qualitative to detect the serogroups present in the antisera and the quantitative aspect to determine the titres of antibodies present. For the qualitative aspect, which followed the protocol described by World Health Organization (13), the panel of 26 serovars were used. e antigens were subcultured in Bijou bottles weekly, incubated at 30 ∘ C and checked for the attainment of a density of 1-2 × 10 8 aer 5-7 days. For determination of titres (the quantitative part), serial 2-fold serum dilutions were made with any sera showing agglutination to any of the antigens used. Where there were 2 serovars with the same titre, these were considered as mixed infections. is procedure was done following standard protocol as described by the Royal Tropical Institute in the manual for the international course on laboratory methods for the diagnosis of leptospirosis (21). A titre of 1 : 20 and higher was considered as evidence of Leptospira spp. infection as recommended for apparently healthy humans [32].

Approval of Study by Ethics Committee. e Ethics
Committee of the Faculty of Medical Sciences approved the study prior to commencement. e objectives and protocol for the study were explained to all the participants and a written consent was obtained.

Data
Analyses. e Statistical Package for Social Sciences (SPSS) version 15.0 was used to produce 2 × 2 contingency tables. Chi-square analysis was conducted to determine risk factors for leptospiral infection in humans and level of signi�cance will be �xed at alpha � 0.05. Table 1 shows the demographic data of students from the four schools in the FMS, University of the West Indies, Trinidad and Tobago, who volunteered to participate in the study. Overall, a total of 212 students participated in the study comprising 113 veterinary students and 99 nonveterinary (46 dental, 39 nursing, and 14 pharmacy) students. Based on data generated on students' characteristics and risk factors for leptospirosis, a majority (54.7%) of students sampled were in years 2 and 3 of their respective programmes, aged 25 years and older (37.3%), and female (73.1%) and reside at homes (53.3%). Amongst the participants across the four schools, none (0.0%) had prior diagnosis of leptospirosis, 69.8% kept pet animals (dogs/cats) with the highest frequency amongst veterinary students (80.5%). A majority (41.0%) kept animals strictly as home or pet animals; the most (36.8%) maintained low levels of contact with their animals with the exception of veterinary students where a majority (46.9%) mentioned that they have close contact with their pets. A high frequency (98.6%) of pet animals did not have recent diagnosis of leptospirosis. Contact with livestock was highest (31.0%) amongst veterinary students. Also the highest frequency of farming (10.6%) outside the faculty was practised by veterinary students of which most were livestock farming (9.7%). A total of 83 (39.2%) of 212 students stated that they encountered rodent problem in the premises where they lived but only 3.3% keep rodents as pets.

Comparison of the ELISA and MAT Results.
A comparison of both serological tests used in the study showed that the frequency of serological evidence in all students was signi�cantly ( ; 2 ) higher, 20.3% (43 of 212), by the ELISA compared with the 7.1% (15 of 212) detected by the MAT.

Risk Factors for Seropositivity for Immunoglobulins to
Leptospira . Regarding seropositivity by risk factors, enrolment as a veterinary student was the only factor that was statistically signi�cantly ( ; 2 ) associated with serological evidence of leptospirosis and this was detected only using the ELISA. Notably, the other risk factors studied (year in the programme, age, gender, place of residence, recent diagnosis of leptospirosis in pet animals or their owners, ownership of pets (dogs, cats, and rodents), rodent problem at/around homes, class of dogs owned (strictly pet, strictly guard, or both), and contact with pet animals or livestock (low, medium, or high) farming (livestock, rice, or sugarcane) were not signi�cantly ( ; 2 ) associated with seropositivity for Leptospira spp. infection.

Discussion
A seroprevalence of 9.7% detected amongst apparently healthy veterinary students using the IgG ELISA in the current study is comparable to the 8.14% reported for veterinary students in Spain [33] where the IgG ELISA was similarly used. However, in that study it was established that the rate of infection by Leptospira spp. increased with the number of years enrolled in the veterinary programme contrary to what was found in our study where the seropositivity rate was not signi�cantly di�erent for students by class in the programme. is may be due, in part, to the fact that from the �rst year and throughout the programme veterinary students are exposed to skills training and an externship programme which bring them in close contact to animals. Animal handling and exposure have been reported to be important in contracting Leptospira spp. infection [9,11,12,33,34]. e �nding, by the use of the ELISA that the serological evidence of exposure to Leptospira spp. in veterinary students was statistically signi�cantly higher than found in nonveterinary students therefore did not come as a surprise. In the current study, being a veterinary student was in fact the only risk factor that was determined to be signi�cantly associated with exposure to Leptospira spp. e relatively low titres (mostly 1 : 20 and 1 : 40) detected by MAT known to have high speci�city and low sensitivity [21][22][23] may have been responsible for the low seropositivity rate detected by the MAT. In apparently healthy individuals, low titres ranging from 1 : 20 to 1 : 40 have been used to determine seropositivity for leptospirosis [32,33].
Of diagnostic signi�cance is the �nding that the ELISA detected signi�cantly higher frequency (20.3%) of serological evidence of exposure experience of Leptospira spp. in all the students studied compared with the MAT (7.1%). is �nding agrees with published reports where the ELISA has been documented to have higher sensitivity and lower speci�city than the MAT [20,24,25]. Another reason that may be responsible, in part, for the lower sensitivity of the MAT is the number and type of serovars of Leptospira spp. in the panel used for testing [22,35]. It has been reported that the use of serovars prevalent in a particular geographical area in the screening panel of serovars increases the sensitivity of MAT [21][22][23]. A major advantage of the MAT over the ELISA is however the fact that it is able to determine the serovars of the infecting Leptospira spp. while the ELISA is genus speci�c. It is also pertinent to mention that crossreactions may occur across serovars but it is also known that individuals may be exposed to multiple serovars of Leptospira spp. and therefore may not be necessarily due to crossreaction amongst serovars.
Of epidemiological signi�cance was the �nding that the predominant Leptospira serovar to which veterinary students have been exposed was Icterohaemorrhagiae Copenhageni while, for nonveterinary students, serovar Australis Rachmati was the most frequently detected. is is relevant because most recent studies using isolation and serological techniques demonstrated that serovar Icterohaemorrhagiae Copenhageni was most prevalent in dogs (cases of clinical leptospirosis, apparently healthy stray and pet dogs), wild  20  40  80  160  320  20  40  80 160  320  Icterohaemorrhagiae Copenhageni 3 (2.7) 1 (0.9) 1 (0.9) - rats, and livestock [35][36][37]. It is therefore obvious that this serovar which is important in causing clinical leptospirosis and subclinical infections in apparently healthy animals is also most likely responsible for the seropositivity for Leptospira spp. in veterinary students in the country, emphasizing the zoonotic signi�cance of leptospirosis. It was of interest that serovar Australis Rachmati, not detected in veterinary students, was most common in nonveterinary students from other schools. e implication of this �nding is not readily apparent because the serovar has not been reported in human clinical leptospirosis or subclinical infection in Trinidad and Tobago but it has been documented in other countries [7,36]. It is however pertinent to mention that all several panels of serovars used for MAT prior to the current study did not contain serovar Rachmati.

Conclusions
It was concluded that although the frequency of detection of serological evidence of exposure experience of Leptospira spp. in the students tested is relatively low, 20.3% by the ELISA and 7.1% by the MAT, in addition to the low titres (mostly 20 and 80) detected in MAT-positive samples, veterinary students have a signi�cantly higher risk of becoming exposed to Leptospira spp. than nonveterinary students. None of the risks factors (age, year in programme, and gender amongst others) had any signi�cant e�ect on infection by Leptospira spp. Serovar Icterohaemorrhagiae Copenhageni was the predominant serovar to which veterinary students were exposed while serovar Australis Rachmati was most frequent amongst nonveterinary students, with possible epidemiological implications.