Introduction

One-third of the world’s population harbors at least one species of intestinal parasite with school-age children bearing the greatest burden in terms of morbidity and mortality [1,2]. Parasitic infections are an important cause of nutritional and energetic stress. The most significant cause of nutritional stress resulting from parasitic infections is hookworm-associated iron deficiency anemia. In a recently performed survey in Kenya including 16,941 children aged 5 to 16 years, the prevalence of anemia was as high as 35% and low hemoglobin levels were significantly associated with hookworm infections [3]. Eosinophilia can also occur secondary to parasitic infections. Infection with intestinal helminths results in immune responses involving cytokines produced by T helper 2 (Th2) cells and consequently the development of eosinophilia [4]. Even in the absence of overt disease, long-lived helminth and protozoan infections may cause immune alterations and nutritional stress associated with immunosuppression, poor growth patterns and chronic malnutrition, i.e. stunting [5-7]. For intestinal helminths, the morbidity severity depends on the intensity of infection [8-10]. Stunting is associated with impaired physical growth and cognitive development [11]. Intestinal parasitic infections have also been associated with acute malnutrition, also known as wasting [12-14]. Wasting results from inadequate nutrition over a shorter period and carries an immediate increased risk of morbidity and mortality [15]. Stunting and wasting can have different determinants and respond to different interventions. Therefore, consideration of stunting and wasting as anthropometric indicators is more useful than consideration of underweight, which encompasses both stunting and wasting [16].

Parasitic infections typically afflict the poorest population segments and prevalence rates of both parasitic infections as well as malnutrition are generally higher in rural compared to urban settings [14,17,18]. There has been an increased awareness of the public health importance of parasitic infections, but much of the interest has been focused on chemotherapy as a means of control while the study of social, cultural and economic factors underlying infection risk has been relatively neglected [19]. Recently, Gazzinelli and co-authors outlined a research agenda for the socioeconomic and health systems research required for the development of sustainable helminth control programs. They concluded that research on social and environmental determinants can contribute significantly to the prevention and control of helminth disease [20]. We present the results of a cross-sectional survey estimating the prevalence of intestinal parasitic infections in school-age children residing in three rural regions of Venezuela, a country with a rich cultural display and a wide variety of climate regions. Two of the three regions are inhabited by Amerindian populations whereas the third region is inhabited by a rural non-indigenous (Creole) population. In addition to providing an overview of helminth and Giardia lamblia infection prevalence rates in all three regions, we also studied the association of intestinal parasitic infections and anemia with acute and chronic malnutrition. To our knowledge, this is the first study comparing prevalence rates of intestinal parasites and malnutrition between three ethnically different populations inhabiting geographically dispersed areas in a South American country.

Methods

Socio-environmental characteristics of study regions

This cross-sectional study was conducted in children aged ≥4 and <17 years from three regions of Venezuela: the Amazon Region, Carabobo State (the village of El 25) and the Orinoco Delta (Figure 1). In the Amazon Region and in Carabobo State, children were included during the dry season, respectively in November 2011 and in March 2010. Door-to-door visits were made to include age-eligible children present in the communities. If an individual was not home during the door-to-door visits, return visits were made within two weeks to include children 4-16 years of age living in the communities at the time of survey for whom parental written informed consent was obtained. In the Orinoco Delta, children were included over a period of 12 months, from May 2010 to May 2011.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Location of study regions in Venezuela and prevalence of intestinal parasites per region.

https://doi.org/10.1371/journal.pone.0077581.g001

The Yanomami Amerindians residing in the Amazon rainforest live in large communal houses where the entire village lives under a common roof of palm leaves called the shabono. Within the shabono, family units are incompletely separated by wooden poles. Households lack toilets and communities have no sewage facilities. The sources of drinking water in Yanomami villages are unprotected wells and river streams. Basic primary health centers are located in some of the communities, but most villages depend on visits of medical teams and do not have a constant access to healthcare. Yanomami children included in this study resided in 11 geographically dispersed villages in the High Orinoco Region of southern Venezuela, an area of perennial tropical forest intersected by a dense river system.

El 25 is a rural community in the mountains south of Lake Valencia in Carabobo State in north-central Venezuela, inhabited by a non-indigenous (Creole) population. People in El 25 live in brick houses with toilets. However, toilet facilities are not used permanently because indoor water supply is not regular, causing a lot of people to defecate in the bushes close to their homes. Sources of drinking water are an untreated underground well and, sometimes, surface water. A small rural hospital provides villagers with basic health care facilities.

The Orinoco Delta in northeastern Venezuela is resided by the Warao Amerindians living in wooden houses raised on stilts in small communities along the Orinoco River banks. There is no sewerage system in the Warao villages. The river is used for bathing and also for drinking and cooking. Medical help is not readily available in Warao communities; in most villages, only a small poorly equipped health post is present. Canoe trips to and from the nearest hospital can take more than 36 hours.

Results

Stool samples from 390 Venezuelan children aged 4-16 years residing in three regions of Venezuela were collected. A total of 264 children (68%) were infected with at least one type of parasite. Helminth infections were present in more than half of the included children in each region (Figure 1, Table 1). The prevalence rate of helminths infecting the host through skin penetration, hookworm and S. stercoralis, was higher in Yanomami children from the Amazon rainforest compared with the other two regions (all p<0.01). In contrast, prevalence rates of helminths that are transmitted through the fecal-oral route, A. lumbricoides and Trichuris trichiura, were lower in the Yanomami compared to the Warao children (p=0.018 and p=0.020, respectively) and the children from El 25 (p<0.01 and p=0.035, respectively). Hymenolepis nana was only present in the Orinoco Delta (8%). The prevalence of G. lamblia infections was 17%, 21% and 16% in the Amazon Region, the Carabobo State and the Orinoco Delta respectively and was not significantly different between the three regions (p=0.56, Table 1).

The inter-observer variability of microscopy reading for parasitic infections was assessed on a subset of 30 fecal samples that were blindly re-read by study microscopists from the three study regions. The parasites present in the 30 randomly selected samples were hookworm, A. lumbricoides, T. trichiura and G. lamblia. The median kappa (κ) values for pairwise comparisons between observers showed excellent agreement for all parasites present in the samples with κ ranging from 0.81 for T. trichiura to 0.93 for A. lumbricoides (all p<0.001).

Anthropometric measurements were obtained from 375 (96%) of the included children. The percentage of malnourished children was highest in the Amazon Region (84%) compared with both the Orinoco Delta (30%, p<0.01) and the Carabobo State (13%, p<0.01, Table 1). Multivariate linear regression analysis showed that the mean height-for-age Z score was lowest in the Amazon Region compared to the other two regions (Table 2). Males had significantly higher height-for-age Z scores than females (0.30, 95% CI 0.01-0.59). Furthermore, the height-for-age Z score was significantly negatively associated with age in years (-0.06, 95% CI -0.10 - -0.01) and G. lamblia infection (-0.39, 95% CI -0.76 - -0.01). Hemoglobin level was significantly positively associated with mean height-for-age Z score in multivariate analysis (0.11, 95% CI 0.02 - 0.20). Helminth infection was not significantly related to height-for-age Z score (Table 2). The indicators of acute malnutrition, weight-for-height and BMI-for-age Z score in respectively children under 5 years of age and those aged 5 years and above, also showed the lowest values in children from the Amazon Region (Table 2). Males showed a significantly higher weight-for-height or BMI-for-age Z score than females (0.22, 95% CI 0.01 - 0.44). Furthermore, the weight-for-height or BMI-for-age Z score was significantly negatively associated with the presence of a helminth infection (-0.24, 95% CI -0.46 - -0.01). G. lamblia infection, hemoglobin levels and age were not significantly associated with the weight-for-height/BMI-for-age Z score (Table 2).

Blood samples were collected from 111 of the 133 included Yanomami children in the Amazon Region (83%) and from 150 of the 152 included Warao children in the Orinoco Delta (99%). Almost a quarter (24%) of the Yanomami and 9% of the Warao children were anemic (p<0.01, Table 1). Hookworm infection was significantly more present in anemic children compared to non-anemic children (51% vs. 35%, p=0.043). The prevalence rates of other helminth or G. lamblia infections were not significantly different between children with and without anemia. Eosinophilia was significantly more often observed in children with a helminth infection compared to uninfected children (72% vs. 28%, p=0.011). In particular, children with S. stercoralis or hookworm infection showed significantly higher rates of eosinophilia than children not infected with these helminths (90% vs. 70%, p=0.022 and 85% vs. 65%, p<0.01, respectively).

Since acute nutritional status and anemia were associated with intestinal helminths, we investigated their relationship with the intensity of helminth infection (Table 3). Most children (65%) infected with hookworm, A. lumbricoides or S. stercoralis had light-intensity infections. The mean hemoglobin level in children infected with hookworm was 12.8 (95% CI 12.5 - 13.2) compared to 14.0 (95% CI 13.6 - 14.3) in children not infected with hookworm. There were no significant differences in hemoglobin level between children with light-, moderate- and heavy-intensity hookworm infections (Table 3). For A. lumbricoides and S. stercoralis, no significant differences in hemoglobin levels between infected and uninfected children were observed. Children infected with hookworm had a significantly lower mean Z score for acute nutritional status compared to uninfected children (-0.25, 95% CI -0.45 - -0.06 vs. 0.12, 95% CI 0.00 - 0.25). This seemed to be due to the low mean Z score observed in children with light- and moderate-intensity hookworm infections as children with heavy-intensity hookworm infections did not show a lower mean Z score compared to uninfected children (Table 3). However, the low number of children with moderate and heavy-intensity hookworm infections in our study population (n=6 and n=9 respectively) precludes firm conclusions on the association between hookworm infection intensity and morbidity. Children with an A. lumbricoides infection showed a lower mean Z score for acute nutritional status compared to uninfected children, but this was only significant for moderate-intensity A. lumbricoides infection (mean Z score uninfected vs. moderate-intensity infection 0.07 (95% CI -0.06 - 0.20) vs. -0.54 (95% CI -1.00 - -0.07)).

Of the 390 included children, 142 (36%) were infected with multiple parasite species. Eighty-nine children (23%) were infected with two different parasite species, 46 children (12%) were infected with three species and seven children (2%) were infected with four different parasite species. In multivariate analysis including age, sex, indicators of acute and chronic malnutrition, region and previously administered anti-helminthic treatment as independent variables, the prevalence of polyparasitism was highest in the Amazon Region compared with the Orinoco Delta (OR 0.47, 95% CI 0.28 - 0.78) and the Carabobo State (OR 0.50, 95% CI 0.29 - 0.89). Multivariate logistic regression analysis did not show significant associations of polyparasitism with age, sex, nutritional status or the administration of anti-helminthic treatment in the past.

Discussion

This large survey performed in school-age children from one non-indigenous and two indigenous populations in rural Venezuela showed high prevalence rates of parasitic infections in all three regions. The frequency of hookworm and S. stercoralis infections was highest in the Amazon Region while Warao and Carabobo State children showed more A. lumbricoides and T. trichiura infections. The prevalence of G. lamblia was around 20% in all three regions. The prevalence of malnutrition varied from 13% in the Carabobo State to 84% in the Amazon Region. In multivariate analysis including all three rural regions, acute nutritional status was significantly associated with helminth infections while chronic nutritional status was significantly associated with the presence of G. lamblia. Hemoglobin levels were significantly decreased in children with a hookworm infection and a significant positive association of hemoglobin levels with the height-for-age Z score, an indicator of chronic malnutrition, was observed in multivariate analysis.

Studies performed in school-age children residing in other indigenous and non-indigenous rural areas of South America also showed high intestinal parasitic infection prevalence rates of 50% to 70% [36-39]. Differences in prevalence rates of intestinal parasitic infections between populations can be explained by behavioral, geographical and host genetic factors [40-42]. A behavioral risk factor found to be associated with intestinal parasitic infections is open field defecation, which is practiced in 40% of the rural Venezuelan population [43]. In particular, high A. lumbricoides and T. trichiura prevalence rates have been observed in communities where people defecate in open fields, bushes or bodies of water around their houses [44,45]. In the Carabobo State and Orinoco Delta, people defecate in the bushes or water surrounding their houses which could be an explanation for the observed higher A. lumbricoides and T. trichiura prevalence in these regions compared to the Amazon Region, where people defecate in specific areas further away from their shabonos. In urban areas of Venezuela, improved sanitation has led to a lower prevalence of helminth infections [18]. Besides improvement in household hygiene, health education adjusted to local ethnomedical knowledge and practices is also important for the reduction of parasitic infections, as demonstrated by a recently performed study in an Amerindian population from the Bolivian Amazon [37]. Wearing clothes and footwear can protect against skin penetration of helminths. In the Peruvian Amazon Region, S. stercoralis infection was significantly associated with not wearing shoes (OR 1.89, 95% CI 1.10-3.27) [46]. The high prevalence of skin-penetrating helminth infections we observed in the Amazon Region could be related to the fact that Yanomami children do not wear clothes or shoes, while Warao and Carabobo State children generally wear clothes and often wear shoes.

Environmental factors can also affect the dissemination and distribution of intestinal parasites in human communities. A. lumbricoides and T. trichiura are sensitive to high land surface temperatures and a lower prevalence of these helminths has been observed in regions where the maximum land surface temperature exceeds 40°C. In contrast, hookworm infection remains prevalent throughout the upper end of the thermal range [47,48]. The Amazon rainforest has a tropical climate with temperatures generally higher than those recorded in the Carabobo State and the Orinoco Delta, where winds decrease the temperature, especially in the afternoon and during the night. Similar to our observations in Yanomami children from the Venezuelan Amazon, in indigenous children from the Bolivian and Peruvian Amazon, hookworm infection rates were at least 3-fold higher than infection rates of A. lumbricoides and T. trichiura [37,49]. Hookworm seems to survive better in the continuously hot climate of the Amazon rainforest than A. lumbricoides and T. trichiura. This may be explained in part by the ability of mobile larvae to migrate to more suitable thermal conditions. Whereas hookworm larvae stages have some limited motility and can move downward into the soil, thereby avoiding desiccation, the ova of A. lumbricoides and T. trichiura are non-motile, and high surface temperatures will result in ova dying from desiccation [50]. The relationship between environmental conditions and survival of parasites may lead to seasonal variation in the prevalence of intestinal parasitic infections. In general, more children are infected with soil-transmitted helminths and G. lamblia in the rainy season while less acquisition of infections takes place in the dry season [51-53].Seasonal variation in parasitic infection prevalence could have influenced our study results as children in the Amazon Region and Carabobo State were included during the dry season while children in the Orinoco Delta were included over a period of 12 months. The inclusion of children in both the dry and the rainy season in the Orinoco Delta may have led to a relatively higher recorded rate of infection with soil-transmitted helminths and G. lamblia in children from this area. However, it is likely that an all-year-round transmission with little distinction between dry and rainy seasons occurs in this river delta as infective stages of intestinal parasites are present in the pools of river water that surround houses during the dry as well as the rainy season. An additional analysis including only children from the Orinoco Delta did not show significant differences between prevalence rates in children included during the rainy season compared to those included during the dry season for any of the soil-transmitted helminths or G. lamblia (median p-value 0.33, all p>0.05). Year-round transmission of hookworm was also observed in the riverine coastal communities in the delta of the Niger River in Nigeria [54].

As data demonstrating differences in parasitic infection prevalence between ethnically different populations are gathered, the need to better understand the role of human genetics and host immune response in susceptibility to parasitic infections becomes clear. Results of epidemiological studies suggest that up to 44% of the variance in worm burden is explicable by genetic effects, compared to 3% to 14% of variance explicable by environmental effects [55]. Ramsey et al. identified polymorphisms associated with A. lumbricoides infection in Venezuelan children from Coche Island [56]. In Yanomami and Warao children, a genetic susceptibility to tuberculosis associated with an altered immune response has been suggested [57,58]. Whether genetic factors play a role in the region-specific pattern of intestinal parasitic infections in Venezuela remains to be investigated.

Besides behavioral, geographical and genetic factors, variation in the detection methods used could have played a role in observed differences in parasitic infection prevalence rates between regions. Chhakda et al. used three different fixatives, namely Kato-Katz, Baermann and SAF techniques, in the same set of single fecal samples taken from Cambodian school-age children [59]. The percentage of children infected with S. stercoralis detected by Baermann and SAF fixation in their study was 20% and 3%, respectively. According to the World Gastroenterology Organization (WGO), the Baermann technique is regarded as the gold standard for the detection of S. stercoralis in stool [60]. Therefore, the difference in S. stercoralis prevalence between the three regions in our study could be due to a difference in technical methods, as Baermann was only used in the Amazon Region. However, in an earlier study performed in the Orinoco Delta, the prevalence of S. stercoralis infection assessed by the Baermann technique was very low [61]. In El 25, the Baermann technique has not been used in previous surveys. Hookworm prevalence in the Amazon Region was more than ten times higher than in the other two regions. As only a 1.3-fold difference was observed between Kato-Katz and SAF analyzed samples in the Cambodian study [59], it is very unlikely that the observed difference in our study was solely due to methodological issues. Another methodological difference to take into account in the interpretation of the results of our study was the difference in the number of samples taken from each child. In the Yanomami children only one fecal sample was collected while two consecutive samples were collected from the Warao and Carabobo State children. The examination of two instead of only one Kato-Katz smear increased the observed prevalence of hookworm with 160% in schoolchildren in Tanzania while the percentage of detected T. trichiura and A. lumbricoides showed a less than 1.5-fold increase when two samples were taken [62]. Since we observed the highest prevalence of hookworm in the Amazon Region where only one fecal sample was collected from each child, it is likely that the observed difference in hookworm prevalence between regions reflects a true difference. Also, the difference in A. lumbricoides prevalence between children from the Amazon Region and the Carabobo State was more than 2-fold and therefore probably not solely due to the collection of two samples in the Carabobo State. For G. lamblia infections, no differences in prevalence rates when one, two or three samples were collected were seen in a study including children from Colombia [63].

We observed significantly higher rates of eosinophilia in children infected with helminths, in particular S. stercoralis and hookworm, compared to uninfected children. Eosinophils are elicited by Th2 effector cells and involved in the initiation and regulation of Th2 immunity [64]. Intestinal helminths are reported to induce a Th2 type immunity in the host and evidence suggests that the Th2 immune response may play a crucial role in reducing the severity of acute disease upon helminth infection [4]. Up-regulation of Th2 responses including eosinophilia by helminth infection can suppress the production of a Th1 immune response which is important to combat intracellular pathogens such as Mycobacterium tuberculosis and the Human Immunodeficiency Virus (HIV) [65,66]. The association between helminth infections and eosinophilia in rural Venezuelan children indicates an effect of helminths on host immunity. As prevalence rates of tuberculosis and HIV are rising in children residing in rural areas of Venezuela [67-69], it is important to consider the immunomodulatory role of helminth infections in these diseases.

Acknowledgments

The authors thank the participating families and children for their cooperation. For their assistance in the laboratory and field work in Carabobo State we thank Esmirna Colmenares, Milagros Armas, Carlos Sequera, Luis Salas, Marisabel Salazar, Paola Santos, María Alejandra Márquez, María Simons, Gabriela Giménez, Yolinda Méndez, Vianny Moreno, Nikki ten Kulve and Robert Ramak. For their assistance in the recruitment of participants in the Orinoco Delta we thank Mailis Maes and Julian Villalba as well as the medical students of the Escuela de Medicina José Maria Vargas of the Universidad Central de Venezuela. Furthermore, we thank Berenice del Nogal for her help in the coordination of the work in the Orinoco Delta and Titia Kortbeek and Jeroen Roelfsema for facilitating study logistics at the RIVM in Bilthoven, The Netherlands.