TOXOPLASMA GONDII INFECTION IN RURAL GUATEMALAN CHILDREN

INTRODUCTION

Toxoplasma gondii is a protozoan parasite that is widespread in humans and many warm-blooded animals, but completes its sexual cycle producing oocysts only in the intestinal tract of the cat.¹ Cats shed the environmentally resistant oocysts in their feces, which can lead to human infection when they are ingested in soil (via pica), uncooked fruits or vegetables contaminated with soil, or water. Humans may also become infected by ingesting undercooked meat from infected animals or by congenital transmission, which occurs when a woman becomes newly infected during pregnancy. Once infected, people are thought to produce IgG antibodies for life.² After acute infection, persons with normal immune systems may be asymptomatic or develop fever, malaise, and lymphadenopathy lasting several weeks to months. Less frequently, acute infection can lead to chorioretinitis and vision impairment or loss. When a pregnant woman becomes acutely infected, severe ocular and neurologic disease may occur in the fetus or later in the infant. In addition, reactivation of subclinical chronic T. gondii infection can lead to life-threatening disease, including encephalitis, in immune suppressed persons.

In the 1950s, T. gondii seroprevalence studies in Guatemala showed high rates of infection, for example, 94% among Mayan Indians 16–70 years of age from near Escuintla, and 50% among Mayan military recruits 15–26 years of age from the central highlands.³ In these populations, more than 75% of the accumulation in prevalence occurred by 16–19 years of age. In 1987, Sinibaldi and Ramirez⁴ found a seroprevalence of 44% among pregnant women with a median age of 25 in an urban Guatemalan hospital. However, to the best of our knowledge, no studies have examined sera from young children in Guatemala and documented the cumulative prevalence. In this report, we present the IgG antibody prevalence for T. gondii IgG in rural Guatemalan children ≤ 10 years of age and assess their risk factors for acquiring infection.

MATERIALS AND METHODS

The study protocol was reviewed and approved by the Ethics Committee Review Board at the Universidad del Valle de Guatemala and the Institutional Review Board at the Centers for Disease Control and Prevention (CDC). Researchers from CDC and the Medical Entomology Research Training Unit (MERTU) of the Universidad de Valle de Guatemala collected study data during two time periods from children in villages in San Juan Sacatepequez, a rural area approximately 32 km west of Guatemala City at an altitude of approximately 450 meters. This region has a temperate climate with a wet season from May through October, when ≥ 27 cm of rain can fall monthly (Guatemala National Institute of Weather, 2003), and a dry season from November through April with little rainfall. The methods for the survey in the first time period, September through December 1999, have been previously described in a report about potential waterborne disease (not including T. gondii).⁵ Briefly, a census had recently been conducted in 10 study villages and all children 6 months to 2 years of age were invited to participate. Participation was nearly universal. The child’s age was verified by checking the birth certificate or vaccination card. Trained local field workers interviewed the mother or caretaker of each child and completed a questionnaire. The questionnaire was designed for a water treatment study and included information about demographics, household composition and characteristics, feeding practices, water use, and pets. Caretakers were given a labeled stool collection cup one day in advance and asked to collect a stool sample from their child and keep it in the shade until the visit, then give it to the interviewers. The study staff attempted to obtain a 1-mL serum specimen from each participating child using a 23-gauge butterfly needle and a 3-mL syringe, then transferred to the blood to labeled Microtainer tubes (Becton Dickinson, Franklin Lakes, NJ). Sera were transported to a freezer and stored at −70°C. In 2003 and 2004, remaining aliquots of sera were tested for IgG antibodies to T. gondii at MERTU and CDC.

Data collection for the second time period occurred in February 2003. Participants were children 3–10 years old from families that volunteered to participate in three villages near those involved in the 1999 study. The culture and living conditions were very similar in these villages to those in the 1999 study. A questionnaire specifically designed to inquire about toxoplasmosis-related risk factors was used for the survey in February 2003 and included information about cats and cat feces exposure, pets, diet, soil exposure, and water exposure. Again, trained local field workers interviewed the mother or caretaker of each child and completed the questionnaire. Blood samples were drawn using methods similar to those described above for the survey in 1999.

RESULTS

DISCUSSION

We found that among rural Guatemalan children ≤ 10 years old, most of the increase in T. gondii antibody prevalence occurred by age 5 (to approximately 43%). The age curves among children six months to five years old start lower, but are steeper than found by Remington and others in cities in El Salvador in the 1960s using the Sabin-Feldman dye test.¹³ Our findings indicate that if these children were not exposed to soil-, water-, or food-related factors that lead to seroconversion at an early age, they were not very likely to seroconvert before adolescence. However, these findings also indicate that as these children mature, up to half of the young women will be susceptible to primary infection with T. gondii when they enter child-bearing age, putting their infants at risk of congenital toxoplasmosis. Rates of congenital toxoplasmosis have not been studied in this population, but would be an important area for future study. High rates of congenital toxoplasmosis were found in an urban Guatemalan hospital in 1987.¹⁴

In the Cochran-Mantel-Haenszel analysis of children 3–10 years old controlled for age, we found two known risk factors for T. gondii infection associated with T. gondii seropositivity, drinking goat’s milk and drinking local well water. Unpasteurized goat’s milk was associated with an outbreak of T. gondii infection in the United States¹⁵ and T. gondii infections thought to have occurred from goat’s milk have been reported in Great Britain¹⁶ and Brazil.¹⁷ Goat’s milk in these Guatemalan villages is generally unpasteurized. In multivariate analysis, although the risk associated with goat’s milk was elevated (RR = 1.6), the results were not statistically significant at the P = 0.05 level (P = 0.1). The second risk factor associated with T. gondii seropositivity, drinking well water, remained significant in multivariate analysis and is similar to a risk factor found in Brazil, drinking unfiltered water.¹⁸ Wells in this region of Guatemala are often shallow and open to the surface, and are likely to become contaminated with soil that may contain cat feces. A third risk factor associated with seropositivity, the mother not cleaning up cat feces, is likely to be an indicator of the presence of cat feces contaminating the soil around the household. Although the risks attributable to the two significant factors in the multivariate model were substantial for those drinking well water or not cleaning up cat feces, a relatively small proportion of the population had these risk factors.

Studies in other Central American countries have found several other risk factors associated with T. gondii seropositivity. Etheredge and Frenkel¹⁹ surveyed children 2–12 years of age in eastern Panama and found that floor type and having cats inside the home were risk factors for T. gondii seropositivity. We did not find these factors to be associated with risk in our study; however, as noted earlier, not cleaning up cat feces was a risk factor. It may be that risk was not associated with cat ownership in our survey (Table 2) because feral cats and those owned by neighbors posed as much risk in spreading oocysts as owned cats. Similar to the study by Etheredge and Frenkel, we did not find evidence for infection by tissue cysts in meat because meat was well cooked in the study villages.

Frenkel and others²⁰ evaluated a cohort of 500 children in Panama City, Panama in 1987–1992 and found a cumulative incidence rate of 12.6% in children between one and six years of age. Elevated relative risks of T. gondii transmission to children were predicted by contact with dogs, cats, flies, much garbage, and many roaches. In the study of Frenkel and others, some of the highest relative risks for seroconversion were associated with dog contact, suggesting the possibility that dogs, by eating and rolling in cat feces, were instrumental in mechanically transmitting T. gondii infection. We did not find questions about dog ownership to be associated with T. gondii seropositivity in the Guatemalan villages in our study.

Our study has several limitations. Inability of respondents to recall events could limit the accuracy of some responses, but the questionnaire was designed to ask about general behaviors (for example, if a food is eaten more or less than once per week, or the number of times in a usual week) rather than specific events in time so that recall would be less of a problem. Seropositivity of IgG for T. gondii does not necessarily represent recent infection with T. gondii, yet the questionnaire completed by caregivers may represent more recent behavioral patterns. However, because most of the children were quite young in our two studies, T. gondii seroconversion would have occurred more recently on the average for these children then it would have for a study of adults, and therefore these children’s risk factors are more closely associated in time to their T. gondii serologic status then would be the case for adults. In addition, since living conditions in these Guatemalan villages have been fairly constant over the past decade, food-, soil-, and cat-related behaviors have probably not changed much since the children in this study were born. One consideration regarding positive IgG antibody test results in infants 6–11 months old is that the antibodies may have been passively transferred from their mother or be due to congenital infection, and therefore would reflect behaviors leading to infection in the mother rather than the child.

It should also be noted when comparing the results for the 1999 and the 2003 studies that they were done in villages in the same area of Guatemala, but the three villages in the 2003 study were not in the 1999 study. Another important consideration is that our findings should not be generalized to other regions of Guatemala or Central America where topography, population density, water supplies, sanitation, soil exposure, and cat populations may differ. Nevertheless, we feel that our findings are a good representation of the children in the indigenous villages from which the samples were drawn. As for comparing participants to non-participants in the surveys, for the 1999 study, specific age or risk factors were not available for the non-participants, so it is not possible to compare epidemiologic data for these groups. However the T. gondii seropositivity rate was higher in the non-participants (31.4% [17 of 54]) versus the participants (10.3% [49 of 478]). In the 2003 study, age was collected for all 500 persons with a serologic sample. The 43 persons with age recorded, but no other data, were a little older (median age = 84.0 months, mean = 80.2 months, SD = 28.1 months) than the 457 persons with age and other data recorded (i.e., the participants who completed questionnaires, median age = 72 months, mean = 74.1 months, SD = 26.8 months), and had a similar but slightly higher T. gondii seropositivity rate (39.5% [17 of 43] versus 37.6% [172 of 457], respectively).

Our results raise many questions. How common is congenital toxoplasmosis in this region? What are the rates of ocular disease due to toxoplasmosis (both congenital and acquired) in Guatemalan children? Are the genetic types of T. gondii found in this region of Guatemala similar to those that cause high rates of ocular disease in Brazil.^21–23 Hopefully, our study will stimulate further research on this interesting organism and the diseases it causes in Guatemala and Central America.