Risk factors associated with seropositivity for Leishmania spp . and Trypanosoma cruzi in dogs in the state of Paraiba , Brazil

The aim of this survey was to determine the seropositivity and risk factors for Leishmania spp. and Trypanosoma cruzi in dogs in the State of Paraíba, Northeastern Brazil. A total of 1,043 dogs were tested, and the serological diagnoses of Chagas disease (CD) and canine visceral leishmaniasis (CVL) was performed by the indirect fluorescent antibody test (IFAT). Animals that tested seropositive for both diseases (by IFAT) were further subjected to ELISA. Of the 1,043 dogs 81 (7.8%; 95% CI = 6.1-9.4%) tested seropositive for Leishmania spp., while 83 were seropositive for T. cruzi (7.9%; 95% CI = 6.3-9.6%). Simultaneous serological reactions were detected in 49 animals (4.6%; 95% CI= 3.6-6.2%). Semi-domiciled housing (OR = 2.044), free housing (OR = 4.151), and soil (OR = 3.425) and soil/cement (OR = 3.065) environmental conditions were identified as risk factors for CVL seropositivity. The risk factors identified for CD seropositivity were semi-domiciled (OR = 2.353) or free housing (OR = 3.454), and contact with bovine (OR = 2.015). This study revealed the presence of dogs in the Paraíba State seropositive for CVL and CD, suggesting the need for revisiting and intensification of disease control measures through constant monitoring of the canine population.


Introduction
The zoonoses visceral leishmaniasis (VL) and Chagas disease (CD) have been known to significantly affect human health in Brazil, prompting the need for repeated medical assistance.These diseases are caused by the protozoans Leishmania infantum and Trypanosoma cruzi, respectively; both are known to be carried by hematophagous insect (Lutzomyia spp.and Triatoma spp.) vectors.Wild and domesticated canids have been identified as the reservoirs of these parasites (SIMÕES-MATTOS et al., 2005;LUCIANO et al., 2009).
Canine visceral leishmaniasis (CVL) is an important zoonosis, which is associated with rapid geographical expansion.This disease has been observed in 47 countries, and is caused by specie Leishmania (L.) infantum (KUHLS et al., 2011).The disease is widely distributed throughout Brazil, and presents specific geographical, climatic, and social characteristics.CVL, which was mostly diagnosed in rural areas in the past, has recently migrated to medium and large urban areas, which has led to changes in its epidemiological profile.In Brazil, CVL occurs in the central-western, southeastern, northern, and northeastern regions, however, the majority of cases have been reported in northeastern region (BRASIL, 2006).
Dogs are the major domestic reservoirs of VL, and play a major role in maintaining the disease cycle (MELO, 2004).Their relevance is attributed to the greater prevalence of VL in the canine than in the human population, since infections in humans are often preceded by infections in dogs.Furthermore, dogs carry a greater number of parasites on their skin than humans, which favors the infection of the vectors (CASTRO, 1996;SANTA ROSA & OLIVEIRA, 1997;BANETH, 2006).
CD, also known as American trypanosomiasis, is a major public health concern in several countries around the world (BORCHHARDT et al., 2010).According to the World Health Organization (WHO), over 6 million people from 21 countries are estimated to be infected with CD, with an annual incidence of 100,000 to 200,000 cases (WHO, 2015).With regard to Brazil, it is estimated that the number of infected individuals is around of three million, and this zoonotic disease is present in the list of neglected tropical diseases (DIAS, 2011).Despite the main reservoirs of CD being wild species, cats and dogs are known to get infected by the causative protozoan; this plays an important role in the ecology and epidemiology of this disease (GÜRTLER et al., 2007).The natural infection of dogs by T. cruzi occurs in a manner similar to human infections, occurring either through active transmission by the vector, or through contamination of the skin and/or conjunctiva by infected feces.However, Barr (2006) stated that the transmission frequently occurs through the ingestion of infected vectors or infected tissues from rodents or other wild animals found in around shelters/residences.
The phylogenetic proximity between the parasites, and the fact that both diseases are endemic to some regions of South America, necessitate the analysis of the two infections in parallel.These zoonoses must be monitored and controlled through surveys that combine serological and epidemiological approaches for each geographical location, as these strategies can lead to the allocation of specific funds that will allow policy-makers to organize and direct new policies towards strengthening public health as a whole.
The aforementioned reasons, the effect of CVL and CD on public health, the role of dogs as parasite reservoirs, and the scarce evidence-based data available in the State of Paraíba, have motivated this study, which aims to determine the seropositive and associated risk factors for Leishmania spp.and T. cruzi in the canine population of this region.

Ethics Committee
This work was approved by the Ethics Committee on Animal Use (CEUA/CESED), Faculty of Medical Sciences of Campina Grande-FCM, under the code 0041/280314.

Samples
Dogs older than three months were included in this study; the test subjects were identified by visiting the residences of their owners, and from those admitted to veterinary clinics and analysis laboratories.Animals from the João Pessoa, Campina Grande, Patos, Sousa, and Cajazeiras counties (Figure 1), five regional urban centers in the State of Paraíba situated along one of the major highways (BR-230; also known as the Trans-Amazonian highway), were included in this study.The sample size was determined according to the total population of dogs (141,863 animals) in these counties (6,843 dogs in Sousa,10,553 in Patos,78,073 in João Pessoa,6,103 in Cajazeiras,and 40,291 in Campina Grande).These numbers were estimated based on human population data for the year 2013, provided by the Brazilian Institute of Geography and Statistics (IBGE, 2013).The dog/human distribution in urban areas was calculated at a ratio of 1:10 ( WHO, 1990;REICHMANN et al., 1999).The sample size was determined for each county based on an estimated seropositivity of 50% (value adopted for sample maximization), a confidence level of 95%, and an error of 10% (THRUSFIELD, 2007); this provided the required sample size of at least 96 animals per county.Ultimately, 1,043 animals were included in this study (125 dogs in Sousa, 206 in Patos, 338 in João Pessoa, 125 in Cajazeiras, and 249 in Campina Grande).
Probabilistic criteria were not established for animal selection, i.e., inclusion of the animals depended on previous contact with the owners and their agreement to taking part in the study.Blood samples were collected from the external jugular or cephalic veins using 5 mL disposable syringes, in the period from January 2013 to June 2014; the serum samples were stored at -20 °C until serological tests.

Serology
Serum antibodies for both diseases were searched by indirect fluorescent antibody test (IFAT), using the protocol by Camargo (1966); the samples were diluted as follows: 1:40, 1:80, 1:160, 1:320, and 1:640.The antigen used to coat the slides for CVL diagnosis was prepared from L. major-like promastigotes, whereas the antigen for CD was prepared from T. cruzi (strain Y) epimastigotes; both cultures were maintained in LIT (Liver Infusion Triptose) and NNN (Neal, Novy, Nicolle) culture media.Positive and negative control sera, for both parasites, were provided by the Núcleo de Pesquisa em Zoonoses (NUPEZO), Universidade Estadual Paulista "Júlio de Mesquita Filho" (UNESP), Botucatu -SP.Based on the results obtained for the controls, the final antibody titer was determined to correspond to the highest dilution of the sera; under these conditions, the membranes of at least 50% of the promastigotes (CVL) and epimastigotes (CD) emitted readable fluorescence, with a cutoff of 40 or higher.
Animals that tested seropositive for both parasites (by IFAT) were further subjected to enzyme-linked immunosorbent assay (ELISA) using the ELISA S7® kit (Biogene Indústria & Comercio Ltda ME, Recife-PE, Brazil) for the sorological diagnosis of CVL.ELISA was performed as per the manufacturer protocols in order to minimize cross-reactivity and accurately detect possible co-infections.
Serum samples that showed positive results at the 1:40 or higher dilutions by IFAT, and which were also determined to be reactive by ELISA, were established to be positive for Leishmania spp.Dogs were determined to be positive for T. cruzi when results from the 1:40 or higher serum dilutions were detected by IFAT.Simultaneous serological reactions were diagnosed when the animal was seropositive for both parasites.

Epidemiological questionnaires
During the collection of blood samples from the dogs, their owners were also provided with an epidemiological questionnaire.This questionnaire requested information regarding a series of variables, in order to investigate certain behaviors and conditions that could act as risk factors for CVL and CD.The variables analyzed and respective categories were as follow: -Owner's information: county of origin (Sousa, Patos, João Pessoa, Cajazeiras, Campina Grande), level of education (illiterate,1 st degree, 2 nd degree, 3 th degree), and traveling with the dog (yes, no);  goat/sheep (yes, no), contact with pigs (yes, no), contact with wild animals (yes, no), environmental conditions (soil, cement, soil and cement), environmental hygiene (yes, no), presence of rodents (yes, no), and access to water dams (yes, no).

Risk factor analysis
The risk factors were analyzed from the data obtained by the epidemiological questionnaires, using univariable approaches.Two groups of animals, seropositive and seronegative dogs, were formed for univariable analysis; these were compared with the tested variables.Variables with p ≤ 0.2, determined by the chi-square or Fisher's exact tests (ZAR, 1999), were selected for multivariable analysis, using multiple logistic regression (HOSMER & LEMESHOW, 2000).The significance level considered to discarding a determined variable was 5%.The collinearity among independent variables was assessed using correlation analysis, and when two variables were highly collinear (correlation coefficient > 0.90), only one variable was likely to enter into analysis.In such situations, selection of which collinear variable to put into the model was guided by biological plausibility (DOHOO et al., 1997).The tests were performed using the SPSS software package, version 13.0 for Windows (SPSS, Inc., Chicago, IL, USA).

Discussion
Several surveys have been conducted in Brazil with the aim of establishing the prevalence of CVL; these studies have produced variable results depending on the characteristics of the study population and the methods used.Alves et al. (1998), Martins (2008), andBarboza et al. (2009), who conducted studies in Fortaleza, CE, Maceió, AL, and Salvador, BA, observed disease frequencies of 1.59%, 1.9%, and 0.7%, respectively; these disease frequencies were lower than the ones reported in this study.In contrast, similar disease frequency was reported by Azevedo et al. (2008) in a survey conducted in Poxoréo, MT (7.8%).The highest disease frequencies were identified by Matos et al. (2006)   (22.1%), and Araguaína, TO (51.35%), respectively.However, it should be emphasized that a majority of these studies were conducted in single locations, whereas the current study focused on five different urban hubs in the State of Paraíba, which could explain the differences in seropositivity.Rondon et al. (2008) reported that the distribution of CVL frequency suggested a seasonal variation, which was caused by the high and low peaks of the vector population, which could justify the variability of data for CVL prevalence throughout Brazil.
Paraíba is an endemic region for CVL, and according to the SINAN Health Information System (Sistema de Informação de Agravos de Notificação), 162 cases of human visceral leishmaniasis have been recorded between 2007and 2013(BRASIL, 2014) ) in Paraíba.Therefore, the measures developed for the control of CVL and VL must be revisited by the responsible policy-makers to facilitate constant monitoring of the canine population for the presence of anti-L.chagasi antibodies, in order to prevent transmission to humans.
It should be highlighted that the criterion adopted for positive serological results in this study (IFAT associated with ELISA, instead of IFAT alone) may have influenced the results.Because of the controversy and the fact that many owners are reluctant to euthanize their animals, since the adoption of that measure has no sufficient effectiveness in reducing the prevalence of cases of the disease, there is the need to carry out more than one test for the confirmatory diagnosis of this disease as well as a standardization of diagnostic techniques to be used both in surveys and in individual cases.This question becomes even more delicate when many of the animals are asymptomatic.
The housing conditions of the dogs, particularly semi-domiciled and free environments, have previously been identified as risk factors for L. chagasi infection by Oliveira & Araújo (2003),   Amora et al. (2006), andNaveda et al. (2006), in Feira de Santana, BA, Mossoró, RN, and Pedro Leopoldo, MG, respectively; these studies have strongly suggested the greater exposure of free animals to the vector.These results are also in agreement with the results obtained by Uchôa et al. (2001), who reported that the lack of organized human occupation (proximity to hillsides and/or forest areas) caused an environmental imbalance that favored the occurrence of disease cycle outside the forests, and closer to the urban areas.The variables related to the environmental conditions (soil, soil/cement), also identified as risk factors for CVL, suggested that a strong presence of organic materials contributes to the proliferation of synanthropic species, in addition to creating a favorable habitat for the spread of the vector (as eggs are usually laid on organic materials).
For DC, 7.9% of the animals were seropositive for T. cruzi.This value differs from that (22.7%) described by Souza et al. (2009), which analyzed the seroprevalence of T. cruzi infection in dogs from Mato Grosso do Sul using IFAT and ELISA.Mendes et al. (2013) found a prevalence of 4.08% using IFAT, ELISA and indirect hemagglutination (IHA), in Patos, PB.The higher frequencies could probably be attributed to the predominantly rural survey areas, which are the usual ecotopes of disease vectors.Silva & Fernandes (2013) utilized IFAT and ELISA and identified a prevalence for CD of 31% in domiciled dogs in São Domingos do Capim, PA.Silva et al. (2014) used IFAT to detect a CD prevalence of 22.2% in a rural area of Bragança, PA.These values, which are higher than the ones reported in this study, could be attributed to the prior instances of human CD infections in both locations.Despite the low positivity reported in this study, the general population and the authorities must be alerted to the presence of infectious agents, in order to plan and establish epidemiological strategies for the effective control of CD.
The housing conditions of the dogs, particularly semi-domiciled and free environments, have also been identified as risk factors for T. cruzi.This is a reason for concern; despite the major vector for this disease (Triatoma infestans) being no longer responsible for CD infections in Brazil (ARGOLO et al., 2008), other species could contribute to its spread in the country.Triatoma sp.insects are currently migrating to urban areas, and have contributed to the strengthening of the disease cycle outside the forest (and closer to urban areas), thereby infecting dogs (especially the ones in free environments) and humans.Contact with bovine is another risk factor for CD.Many wild and domestic mammals are known reservoirs of CD; although dogs are the major domestic reservoir for human infection, other animals can also contribute to CD ecology (DIAS & COURA, 1997).Furthermore, it is usual the occurrence of bovine roaming free in many urban areas of the State of Paraíba.Dias et al. (2000) have reported the decades-old existence of peri-urban disease foci, and have suggested that the constant migration from rural to urban areas, and the poverty and semi-rural characteristics of the periphery neighborhoods determine disease distribution.
According to the criteria established in this study, 4.6% (49/1043) of the dogs tested positive (by IFAT assay) for Leishmania spp.and T. cruzi, possibly indicating the occurrence of mixed (simultaneous) infections (UMEZAWA & SILVEIRA, 1999).These results corroborate that ones obtained by Luciano et al. (2009), who observed intense cross-reactions following higher differences in dog serum titers (tested by IFAT) for Leishmania spp.and T. cruzi antigens.It must be highlighted that the remaining T. cruzi titers varied between 40 and 640.In contrast, Souza et al. (2009) mentioned the difficulties in discriminating between infections by T. cruzi and Leishmania spp. in asymptomatic dogs using conventional diagnostic techniques.Several South American locations are endemic to CVL and CD, with a high possibility of mixed infections.Morais et al. (2013) have also observed the simultaneous infection of canines by L. chagasi and T. cruzi in Araguaína, TO; of the 111 samples tested 57 were observed to be positive (IFAT and ELISA) for CVL (51.35%).The same sera were also analyzed by TESA-blot, which suggested the possibility of CD infection in five animals (4.5%); among these, 3 tested ELISA-positive and IFAT-negative for leishmaniasis.
In conclusion, seropositive animals for Leishmania spp.and T. cruzi were detected in the canine population in the State of Paraíba, suggesting the need for revisiting and intensification of disease control measures through constant monitoring by the competent authorities.Based on the identified risk factors for CVL and CD, we propose the observation of certain measures when allowing dogs on the streets; in addition, the environmental and housing conditions that the dogs are subjected must be improved.Contact with bovine was identified as risk factor for CD, which emphasizes the need for future studies on the role of this species in the transmission cycle of the disease.

Figure 1 .
Figure 1.Geographical distribution of counties used.Detail shows the State of Paraíba within Brazil.

Table 1 .
Seropositivity for visceral leishmaniasis, Chagas disease and both diseases in dogs in the period from January 2013 to June 2014, in the State of Paraíba, Brazil.

Table 2 .
Univariable analysis for risk factors associated with the seropositivity for visceral leishmaniasis and Chagas disease in dogs in the period from January 2013 to June 2014, in the State of Paraíba, Brazil.
*Variables selected for the multiple analysis (p ≤ 0.2); p: probability of casual occurrence.

Table 3 .
Risk factors associated with the seropositivity for visceral leishmaniasis and Chagas disease in dogs in the period from January 2013 to June 2014, in the State of Paraíba, Brazil.
*Variables selected for the multiple analysis (p ≤ 0.2); p: probability of casual occurrence.