Seroepidemiology of Toxocara Canis infection among primary schoolchildren in the capital area of the Republic of the Marshall Islands

The ascarids that cause human toxocariasis are Toxocara canis (T. canis) and, likely to a lesser extent, Toxocara cati (T. cati). The definitive hosts of T. canis and T. cati are dogs and cats, respectively; these ascarids inhabit the lumen of the small intestine [1]. Worldwide surveys of T. canis occurrence have indicated a prevalence ranging from 86% to 100% in pups and 1% to 45% in adult dogs [2, 3]. Humans are one of several accidental hosts, and are primarily infected by ingesting parasite eggs or, to a lesser extent, by consuming chicken or cow livers [4].

Although human infections with Toxocara spp. are typically asymptomatic, larval migration into the internal organs via the blood can cause various clinical syndromes including visceral larva migrans and ocular larva migrans. The manifestation of symptoms in human toxocariasis depends on multiple factors, including which organs are affected and the magnitude of the infection [3, 5].

Young children up to the age of 12 years appear to be the primary population susceptible to T. canis infection because of dirt pica, poor hygiene, or frequent contact with dogs [3, 6]. Multiple reports have indicated that child toxocariasis is associated with endomyocarditis, generalized lymphadenopathy, endophthalmitis, asthma, hepatosplenomegaly, and meningoencephalitis [7‐11]. Considerable interest has been directed toward the role of T. canis infection in epilepsy, and particularly in partial epilepsy [12‐14].

In humans, parasites cannot mature to the adult stage; thus, examining stool for parasites and eggs is not useful. Making a direct parasitological diagnosis by using biopsy is extremely difficult; thus, serological methods are the diagnostic mainstay. Serological diagnoses of toxocariasis primarily rely on a T. canis larval excretory-secretory (TcES) antigen-based enzyme-linked immunosorbent assay (ELISA) of T. canis[3, 5]. The seroprevalence of T. canis infection among children in various countries has been reported to range from 4% to 86% according to TcES-ELISA [15‐17]. No reports on the seroprevalence of T. canis infection in children in Micronesian areas are available, and its status remains unknown among children who live in the Republic of the Marshall Islands (RMI).

The sensitivity and specificity of TcES-ELISA, when 1:32 was used as the threshold titer for positivity, have recently been estimated at 78% and 92%, respectively [18, 19]; however, antigenic cross-reactivity (e.g., with Ascaris lumbricoides) reduces the usefulness of such assays, particularly in areas where polyparasitism is common [19]. Western blotting based on the fractionated, native, and excretory-secretory antigens of T. canis larvae (TcES-WB) can yield superior specificity levels, exhibiting reactivity to bands of low molecular weights (24–32 kDa) that were proven to be specific to T. canis infection [19]. In the present study, TcES-WB was used to detect T. canis-specific Immunoglobulin G (IgG) and estimate the seroprevalence of T. canis infection among primary schoolchildren (PSC) living in the capital area of Majuro of the RMI.

Of the 166 PSC serum samples, 144 (86.75%; 144/166) tested positive for Toxocara IgG antibodies according to TcES-WB. The overall seroprevalence was 83.33% (75/90) in boys and 90.79% (69/76) in girls (Table 1). The seroprevalence was highest (95.45%, 21/22) in Group 1 (<7 years), followed by 88.14% (52/59) in Group 3 (7–9 years), and 83.53% (71/85) in Group 3 (>9 years; Table 1). PSC who lived in urban and suburban areas demonstrated seropositive values for the IgG antibody against TcES of 90.32% (84/93) and 82.19% (60/73), respectively (Table 1). The seropositive rates among those whose parents were nonskilled or skilled workers were 89.91% (98/109) and 75.68% (28/37), respectively (Table 1). The risk factors analysis indicated that the seropositive rates were 88.05% (140/159) among PSC who exhibited histories of feeding dogs, 88.89% (56/63) among those who cleaned dog huts, 80.0% (8/10) among those who cleaned dog huts while wearing gloves, 87.74% (136/155) among those who played with soil, 87.41% (125/143) among those who consumed frozen or raw foods, 85.07% (114/134) among those who ate raw vegetables, and 88.72% (118/133) among those who drank untreated or unboiled water (Table 2). In the multivariate logistic regression analysis, gender, age, parental occupation, and risk factors were included in the regression model. As indicated by Table 3, gender was not significantly associated with seropositivity for the T. canis antibody when conducting the uni- or multivariate logistic regression analyses (p > 0.05). In addition, age was not a critical factor related to T. canis infection, because no significant differences in seroprevalence were observed among the age groups when conducting the uni- or multivariate logistic regression analyses (p > 0.05).

Urbanization level was not significantly associated with seropositivity for T. canis in the uni- or multivariate logistic regression analyses (p > 0.05). By contrast, parental occupation level was a crucial contributing factor because PSC whose parents were employed as semiskilled workers exhibited a significantly higher risk of T. canis infection than did those whose parents were employed as skilled workers in both the univariate (ORs = 2.86, 95% CIs = 1.08–7.60, p = 0.03) and multivariate logistic regression analyses (ORs = 3.83, 95% CIs = 1.20–12.22, p = 0.02). Among the analyzed risk factors, univariate regression indicated that only PSC who exhibited a history of feeding dogs demonstrated an increased risk of T. canis infection compared with those who lacked such a history (ORs = 5.53, 95% CIs = 1.15–26.61, p = 0.03); by contrast, these effects were nonsignificant in the multivariate logistic regression analysis (ORs = 2.64, 95% CIs = 0.48–14.65, p = 0.27).

The RMI is a tropical developing country; thus, its climatic and living conditions might favor various pathogens survival including parasites such as T. canis. However, few systemic studies have evaluated the prevalence of T. canis infection among PSC in the RMI.

PSC are particularly vulnerable to toxocariasis because of their habits of playing in water or soil; eating raw foods; or contacting pets, including cats and dogs; thus, PSC are an ideal target group for investigating the prevalence of toxocariasis. Data collected from the evaluated age groups can be used to assess whether toxocariasis threatens the health of school-aged children, and can serve as a reference when evaluating the need for community interventions [23].

In most laboratories, such as the Centers for Disease Control and Prevention (CDC) in Atlanta, the current methods of detecting T. canis infection in humans are typically based on TcES-ELISA [6, 24]. Although TcES-ELISA has been reported to yield reasonable levels of sensitivity (78%) and specificity (92%), when the threshold titer of positivity is set at 1:32 [18, 19], the specificity is excessively low to be reliable when assaying communities in which several species of intestinal cross-reacting helminths (e.g., Ascaris lumbricoides ) are common. Although no attempt was made to verify whether the current participants were host to helminths other than T. canis, it was assumed that infection with several species of helminth would be common in the RMI. Thus, western blotting [19], rather than a general TcES-ELISA, was employed in this study. The findings indicated that T. canis infection was common among the PSC living in the capital areas of the RMI, and most PCS (86.75%) tested seropositive. Making a valid comparison of the present seroprevalence data with those recorded in previous studies is hampered by the variation in detection methods (ELISA or WB) and cut-off titers employed and by the general difficulty in exploring the relationships among titers, infection, and clinical disease [3, 5, 19]. Serological surveys conducted primarily among children in developed countries have indicated T. canis seroprevalence levels of 0.7% in New Zealand, 1.6% in Japan, 2.4% in Denmark, 7.5% in Australia, 14% in the United States, and 15% in Poland [3, 5, 25]. By contrast, high levels of seroprevalence have been reported in less developed, or tropical countries in Africa (30% in Nigeria, 45% in Swaziland, and 93% in La Reunion), Asia (81% in Nepal, 63.2% in Indonesia, and 58% in Malaysia), and South America (36% in Brazil and 37% in Peru) [21, 26, 27]. Other studies involving TcES-ELISA assessments have reported markedly increased seroprevalence levels of 77% among indigenous children in Taiwan [22] and 86% among children in St. Lucia [28]. In the present study, a relatively high cut-off titer (1:64) was used and it seems likely that some and perhaps many of the seropositive PSC exhibited active T. canis infections when they were sampled. According to the CDC, when conducting TcES-ELISA, a titer of 1:32 is indicative of active Toxocara infection [19].

Although boys might be more likely to be infected than girls in certain communities in which boys have frequent contact with dogs [15, 26], the current results indicated that girls exhibited an elevated risk of infection. However, gender did not significantly affect seroprevalence in the present study, contrasting the results of studies in China, Iran, Nigeria, and Spain [16, 29‐31]. The reasons for this discrepancy warrant further investigation. Girls might often be involved in housework-related chores, such as washing clothes in water, or might have more frequent contact with dogs than boys do, increasing their opportunity to be exposed to T. canis; thus, subsequent studies are required.

The seroprevalence of T. canis infection in RMI PSC decreased as age increased; however, this trend was nonsignificant, possibly because of the small numbers of children in certain age groups. Seroprevalence might be cumulative in children because detectable titers of anti-Toxocara antibodies might persist over time postinfection [15, 19, 21]. No age-related increase in seroprevalence was observed among children in Argentina, Iran, Nigeria, or Spain [16, 30, 31].

Exposure to Toxocara is common because soil contamination is prevalent in peridomestic environments, and is exacerbated by poverty, poor hygiene, and the risk of contact with infected dogs [5, 6]. The results of the full regression analysis suggested that poverty is a critical risk factor because seropositivity levels were elevated among PSC whose parents were employed as nonskilled workers (e.g., factory workers). A recent study verified that Toxocara infection is closely related to poverty status in the United States, suggesting that those in certain occupations, such as farming, might be at an increased risk [32].

A crucial risk factor of T. canis infection is contact with dogs [3]. The results of the univariate analysis in the current study indicated that PSC who exhibited a history of feeding dogs demonstrated a higher risk of T. canis infection than those who did not; however, these effects were nonsignificant in the multivariate regression analysis. Substantial evidence has indicated that direct human-dog contact might be a route of human infection based on the retained infectivity and pathogenicity of embryonated T. canis eggs recovered from dog hair [33‐35]. Nagy et al. [36] located embryonated eggs on dog hair, but suggested that this route of transmission is rare. Nevertheless, a recent study indicated that embryonation is slow on dog hair but can occur; thus, transmission through direct contact, and even contact with well-groomed dogs, should not be ruled out [37]. Therefore, potential transmission through contact with embryonated T. canis egg contaminated in the environment such as soil should be not ignored also [3].

Children who have tested seropositive for T. canis have been linked to impaired cognitive development [25, 38], generating cause for concern. Walsh and Haseeb [39] reported that seropositive children in the United States attained statistically lower cognitive scores on both the Wechsler Intelligence Scale for Children-Revised and Wide Range Achievement Test-Revised than did seronegative children. This finding was independent of socioeconomic status, ethnicity, gender, rural residence, cytomegalovirus infection, and lead levels in blood. Growing evidence has also implied that T. canis infection is associated with epilepsy [13, 14]. Whether neurological deficits exist among PSC infected by T. canis in RMI warrants further comprehensive investigation.