Seroprevalence and associated risk factors of Toxocara canis infection among primary schoolchildren from rural areas of southern Thailand

Human toxocariasis is a parasitic zoonosis caused by a parasite in the genus Toxocara and is transmitted mainly by accidental ingestion of embryonated Toxocara canis (dog round worm) or T. cati (cat round worm) eggs. Several studies reported that children were the main population at risk for T. canis infection. Currently, no reports on the seroprevalence of T. canis infection in Thailand are available, and its status remains unknown among children who live in rural areas of southern Thailand. This study aimed to investigate the seroprevalence of T. canis infection and its associated risk factors among primary schoolchildren in rural areas of Nakhon Si Thammarat Province, Thailand. The high rate of Toxocara seropositivity reected high levels of exposure to T. canis among schoolchildren in rural areas of southern Thailand. The results also provide baseline data regarding modiable risk behaviors for effective T. canis infection prevention strategies in southern Thailand, especially strengthening hand washing practices among schoolchildren.


Introduction
Human toxocariasis is one of most prevalent parasitic zoonoses worldwide; it is particularly present in subtropical and tropical regions and in developing countries [1]. Toxocariasis is caused by nematode parasites of the genus Toxocara, including dog roundworms (Toxocara canis) and cat roundworms (Toxocara cati), whose de nitive hosts are dogs and cats, respectively [2]. Humans acquire Toxocara spp. via the accidental consumption of Toxocara eggs, which are present in soil contaminated by dog/cat feces, or via ingestion of larvae in undercooked meat [2]. Clinical manifestations of human toxocariasis are mostly asymptomatic; however, severe disease can occur when larvae migrate within the body to internal organs or the eyes, causing visceral toxocariasis and ocular toxocariasis, respectively [1].
Humans are incidental hosts for Toxocara spp. The larvae cannot develop into adult worms in the human small intestine; thus, no Toxocara eggs are present in human stool [3]. The typical diagnosis of toxocariasis usually relies on serological tests. Enzyme-linked immunosorbent assay (ELISA) using Toxocara excretory-secretory (TES) antigens has been widely used and recommended by the Centers for Disease Control and Prevention to detect a Toxocara-speci c IgG [3,4]. The limitations of this test are its cross-reactivity with other parasites, especially the human roundworm, Ascaris lumbricoides [5], and its lack of ability to differentiate between active and past infections. However, ELISA remains useful in seroepidermiological surveys, as it is quick in determining the prevalence of asymptomatic toxocariasis and affordable.
In Thailand, previous reports revealed that Toxocara eggs were found in the stool of dogs and cats [20,21]. Furthermore, Toxocara eggs were identi ed in raw vegetables from the markets in southern Thailand [22]; however, its impact on humans has not been investigated. In this study, we aimed to assess the seroprevalence and associated risk factors of Toxocara canis infection among schoolchildren in rural areas of southern Thailand.

Study design and setting
This cross-sectional study was carried out from June to July 2019 in ve districts of Nakhon Si Thammarat Province, including Phrom Khiri, Tha Sala, Sichon, Khanom and Nopphitam. Nakhon Si Thammarat Province is located in

Study population and sample size
The study population consisted of primary schoolchildren from 7 to 13 years of age. The sample size was determined using the single proportion population formula: where p = prevalence of intestinal parasites from a previous study, d = margin of error, and Z = standard score, which corresponds to 1.96. This formula was calculated based on a prevalence rate of 86.75% from a previous study [7], with a margin of error of 0.05 and a con dence level of 95%. An appropriate sample size was determined to be 177. Students who had immune system disorders (such as autoimmune disorders and acquired immune de ciency syndrome), who were on steroid treatment for at least 3 months and who had acute diseases on the day of blood collection were excluded from the study. Three schools were randomly selected from each district for a total of fteen schools. Finally, the participants were selected from the school rosters using a systematic random sampling technique in which every 10 th person was included in the study.

Questionnaire survey
A structured questionnaire was developed and used to collect demographic data (i.e., age, sex, and parental education and occupation) and information on possible risk factors (habits of handwashing before a meal, after playing and after animal contact; owning pets (dogs or cats); eating fresh vegetables; and drinking untreated water). Agriculturists, farm laborers, housemaids, shing boat staff, and street vendors were considered nonskilled workers. The questionnaire was administered by two trained interviewers who conducted direct interviews with the participating schoolchildren.

Blood collection and preparation
Blood samples were collected from the antecubital vein by medical technologists and kept for 45 min at room temperature. After that, serum was separated by centrifugation at 2500 rpm for 10 minutes and stored at -80°C until measurement of IgG class antibodies against T. canis.

Detection of anti-T. canis IgG antibodies
Serum anti-T. canis IgG antibodies were detected with a commercial ELISA kit (Abcam, UK) according to the manufacturer's instructions. In brief, all samples were diluted to 1:100 with IgG sample diluent, and all controls (T. canis IgG-positive, T. canis IgG-negative and T. canis IgG cutoff (CO)) were prepared. A precoated plate was incubated (100 μL/well) with control or diluted samples and incubated for 1 hour at 37°C. After incubation with the serum samples, the plates were washed 3 times with 1x washing solution and incubated (100 μL/well) with protein A HRP conjugate for 30 minutes at room temperature. The plates were washed 3 times and incubated (100 μL/well) with TMB substrate solution for exactly 15 minutes at room temperature in the dark. The reaction was stopped (100 μL/well) with stop solution for 15 minutes at room temperature. The assay included negative and positive serum samples in addition to a blank (no serum sample). Absorbance at 492 nm was measured using an automatic microplate reader. For interpretation of the results, samples with an absorbance value of more than 10% above the CO were considered positive. If the absorbance value was between 10% above and 10% below the CO, it was considered inconclusive (in the gray zone), and a fresh sample was run again. If the results of the second test were less than 10% above or below the CO control value, the sample was considered negative.

Data analysis
Data were entered, cleaned, and analyzed using IBM SPSS Statistics for Windows, Version 23. Quantitative variables were described by medians and interquartile ranges (IQRs), and qualitative variables were described by frequencies (percentages). A chi-square test was used to compare the proportions of T. canis infection with sex, age group, level of education, districts, parental occupation and parental education. A univariate logistic regression model was constructed to investigate risk factors associated with T. canis infection. The variables with P-values less than 0.2 in the univariate logistic regression model were included in a multiple logistic regression model to adjust for confounding factors. Differences were considered to be statistically signi cant when the P-value obtained was less than 0.05.
Their parents' occupation and education level were mostly nonskilled workers and completion of high school or less, respectively ( Table 1).

Seroprevalence of Toxocara canis infection
The overall seroprevalence of T. canis infection among the participants was 58.2% (95% CI: 50.9-65.5). Boys had a higher seropositive rate (64.8%) than girls (51.2%). The rates of seropositivity (54.3% in Khanom, 52.9% in Sichon, 58.8% in Phrom Kiri, 55.6% in Nopphitam and 68.4% in Thasala) were not signi cantly different among the ve districts (P = 0.672). The highest seropositivity was observed among third-grade students (68.3%), followed by sixth-grade students (65.0%), while the lowest rate was observed among fourth-grade students (43.8%). Chi-square tests showed that there were no signi cant associations between T. canis infection and sex, age group, level of education, district, parents' occupation and education (Table 1).

Associated risk factors for Toxocara canis infections
The univariate analysis revealed that handwashing before a meal, hand washing after animal contact and drinking untreated water were signi cantly associated with seropositivity for T. canis. However, the multivariate analysis indicated that handwashing before a meal was the only factor associated with T. canis seropositivity. Children who did not practice handwashing before a meal were more likely to be infected with T. canis than those who did (AOR = 2.20, 95% CI: 1.11-4.34) ( Table 2).
When strati ed by age, the rates of seropositivity for T. canis were approximately equal among age groups (56.3%, 58.5% and 58.7% for children 6-8 years of age, 9-10 years of age and 11-13 years of age, respectively). Boys were more likely to be infected with T. canis than girls. Children who did not practice hand washing after playing as well as children who owned dogs and cats and who ate fresh vegetables were more likely to be infected with T. canis than those who did not. However, there was no statistical signi cance among these factors in both the univariate and multivariate analyses ( Table 2).

Discussion
Toxocariasis is a prevalent parasitic zoonosis worldwide, especially in the tropics and subtropics. This study was the rst serological investigation of T. canis infection among schoolchildren in southern Thailand, with a rate of 58.2%. The rate of T. canis infection in this study was slightly higher than those in other countries in Asia, including the Philippines (49.0%) [11], Taiwan (57.5%) [23], Turkey (45.9%) [10], Korea (51.2%) [24] and Isfahan, Iran (45.9%) [13], whereas the rate was lower than those in some other parts of the world, including Nigeria (86.1%) [6] and the Republic of the Marshall Islands (86.8%) [7]. In general, multiple factors, including environmental, geographic, cultural and socioeconomic factors, contribute to the magnitude of T. canis infection. Our study areas included ve districts of Nakhon Si Thammarat Province, which were considered rural, where most of the countryside contains rubber plantations and farmlands, and where dogs are commonly owned as pets; moreover, there is a great number of stray dogs in these rural districts. Nakhon Si Thammarat has a tropical rainforest climate, in which the temperature ranges from 24 to 34℃, which is suitable for the development of Toxocara eggs to the infective larval stage [25]. We hypothesized that climate, soil and the number of dogs in the areas in uenced the rate of seropositivity in this study.
Among the ve districts, Thasala showed the highest rate of Toxocara seropositivity (68.4%); however, there was no statistically signi cant difference among the ve districts. This might be explained by the similarities in climate, geography and culture among the districts. Children whose parents were nonskilled workers and had low education levels had an increased rate of Toxocara seropositivity. This trend was also observed in a previous study in the Republic of the Marshall Islands [7]. However, parents' education levels and occupations did not signi cantly affect the seroprevalence rate in this study. This might be due to a small number of children whose parents were skilled workers and obtained Bachelor's degrees.
Among the age groups, there was no signi cant difference in the rate of Toxocara seropositivity in the current study. The rates of Toxocara seropositivity were approximately similar across all age groups: 56.3% in children 6-8 years of age, 58.5% in children 9-10 years of age and 58.7% in children 11-13 years of age. Similar ndings were observed in the study from the Republic of the Marshall Islands [7]. However, this result was in contrast to results of several studies among schoolchildren that showed that older age was a signi cant risk factor for T. canis infection [6,12,18,26,27].
The detection of Toxocara via serology increases over time because of the persistence of serum IgG. The possible explanation for the similar seropositivity rates among the different age groups might be that the children might have been in contact with Toxocara eggs since a young age and the persistence of an IgG remains over a long period of time. Our study revealed that boys tended to be infected with Toxocara spp. more often than girls (64.8% in boys and 51.2% in girls). This trend was also observed in previous studies [8,18,19]; it has been hypothesized that boys engage in more outdoor activities than girls; thus, they are more likely to be in contact with soil and dogs. However, sex was not signi cantly associated with Toxocara seropositivity in either the univariate or multivariate analysis in the current study.
Regarding modi able risk behaviors, hand washing before a meal appeared to be the only signi cant risk factor associated with Toxocara infection in both the univariate and multivariate analyses. Schoolchildren who did not practice hand washing before a meal were 2.2 times more likely to be infected with T. canis than those who did. This effect was also observed in a previous study among Brazilian schoolchildren [28]. Furthermore, the univariate analysis in this study revealed that hand washing after animal contact and drinking untreated water were signi cantly associated with T. canis infection; however, these associations were not statistically signi cant in the multivariate analysis. Humans acquire T. canis by ingesting infective eggs. This study emphasizes the importance of hand hygiene to prevent parasite eggs from entering the body. Despite the report of Toxocara egg contamination in raw vegetables in Nakhon Si Thammarat Province in a previous study [22], eating fresh vegetables was not signi cantly associated with T. canis infection in the current study. This may be explained by the small number of schoolchildren who ate fresh vegetables 36 (20.3%); thus, contamination from vegetables was not a likely route of transmission among these children. Schoolchildren who engaged in risk behaviors, such as owning dogs and owning cats, also tended to have a higher rate of T. canis infection than those who did not; however, no signi cant associations were observed among these factors and T. canis infection.
This study had the following limitations. First, the method for detection of anti-T. canis IgG antibodies was the ELISA technique, which could yield false positive results due its cross-reactivity with other helminths, especially A. lumbricoides [5,29]. However, our study sites were not endemic areas for A. lumbricoides [30,31]; hence, the chance of false positivity was likely low. Second, the study design was cross-sectional, and seroprevalence and risk factors were evaluated simultaneously, not over a period of time; thus, true causes and effects might not be strongly demonstrated.

Conclusions
This is the rst serological investigation of T. canis infection among schoolchildren in Thailand. The high rate of Toxocara seropositivity re ected high levels of exposure to T. canis among schoolchildren in rural areas of southern Thailand. A lack of hand washing before meals appeared to be a signi cant risk factor for T. canis infection. Appropriate public health education on personal hygiene, especially strengthening hand hygiene practices among schoolchildren, should be implemented to prevent the transmission of T. canis.

List Of Abbreviations
AOR: adjusted odds ratio CI: con dence interval CO: cut-off COR: crude odds ratio IQR: interquartile range