Prevalence and associated risk factors of intestinal parasites among schoolchildren in Ecuador, with emphasis on the molecular diversity of Giardia duodenalis, Blastocystis sp. and Enterocytozoon bieneusi

Background Intestinal helminths, including Soil-Transmitted Helminth (STH), and Gastrointestinal Protist (GP) infections are major contributors to the global burden of disease, particularly in low-income countries such Ecuador. Their epidemiology in these settings is largely unknown. Methodology This prospective cross-sectional study investigates the carriage of intestinal helminths, including STH, and GP in asymptomatic schoolchildren (3–11 years) in the Chimborazo and Guayas provinces, Ecuador. Single stool samples (n = 372) and epidemiological questionnaires on demographics and potential risk factors were collected from participating schoolchildren. Conventional microscopy examination was used as screening method, and molecular (PCR and Sanger sequencing) assays were used to further investigate the epidemiology of some GP. A multivariate logistic regression analysis was used to evaluate the strength of the association of suspected risk factors with the presence of helminths and GP. Principal findings At least one intestinal parasite species was observed by microscopy in 63.2% (235/372) of the participating schoolchildren. Enterobius vermicularis (16.7%, 62/372; 95% CI: 13.0–20.9) and Blastocystis sp. (39.2%, 146/372; 95% CI: 34.2–44.2) were the most prevalent among helminths and GP, respectively. Assemblages A (50.0%), B (37.5%) and A+B (12.5%) were detected within Giardia duodenalis and ST3 (28.6%), ST1 and ST2 (26.2% each), and ST4 (14.3%) within Blastocystis sp. Three genotypes, two known (A: 66.7%; KB-1: 16.7%) and a novel (HhEcEb1, 16.7%) were identified within Enterocytozoon bieneusi. Municipality of origin, household overcrowding, and poor sanitation and personal hygiene habits were risk factors for childhood intestinal parasites colonization. Conclusions/Significance Despite massive government drug administration programs, STH and GP infection remain a public health concern in paediatric populations living in poor-resource settings. Molecular analytical methods are required to better understand the epidemiology of these intestinal parasites. This study provides novel information on the occurrence of Blastocystis sp. and E. bieneusi genetic variants circulating in Ecuadorian human populations.


Introduction
Intestinal helminths, including Soil-Transmitted Helminth (STH), and Gastrointestinal Protist (GP) infections are major contributors to the global burden of disease, particularly in resource-deprived, low-income countries from tropical and subtropical regions [1][2][3]. In these endemic areas, concomitant infections are common, and their synergistic effect might exacerbate detrimental health outcomes in infected individuals [4]. Chronic or repeated STH and GP infections during childhood have a strong link with stunting, the most common form of malnutrition in children aged under five years [5]. In paediatric populations, stunting or chronic malnutrition refers to both reduced physical growth and cognitive impairment, representing a major public health concern leading to lifelong adverse health, education, and economic outcomes [6]. Because of their chronic rather than acute nature, STH infections are only partially addressed in the Global Burden of Disease (GBD) studies [7], whereas diarrhoea-causing GP infections are not formally considered as neglected tropical diseases [8]. These limitations impair our understanding of the epidemiology of STH and GP in endemic areas.
Among GP, the protozoa Cryptosporidium spp., Giardia duodenalis, and Entamoeba histolytica are regarded as relevant diarrhoea-causing pathogens globally. They are faecal-orally transmitted either directly through contact with infected humans and other animals, or indirectly via ingestion of contaminated food or water. Cryptosporidiosis is the second cause of childhood mortality after rotaviral enteritis in sub-Saharan African and southwest Asian countries [13] causing an estimated 48 million DALYs [14]. Although rarely mortal, giardiasis affects 200 million people globally every year, mainly children aged between two and five years [15]. Entamoeba histolytica has been linked to an increased risk of death in infants and toddlers with moderate-to-severe diarrhoea in sub-Saharan Africa [16]. Other less frequent GP, but still of public health relevance, include the microsporidia Enterocytozoon bieneusi and the Stramenopile Blastocystis sp. The former is an opportunistic pathogen primarily infecting immunocompromised individuals [17], whereas the latter has been increasingly linked to a variety of intestinal (diarrhoea, irritable bowel syndrome) and extra-intestinal (urticaria) clinical manifestations [18]. All the GPs mentioned above are featured by a large intra-species molecular diversity that influence their host range and specificity, pathogenicity, virulence, and zoonotic potential. Therefore, assessing the species/subtypes/genotypes involved is essential to characterise the epidemiology of these parasites [19][20][21].
As a shared feature, intestinal helminths and GP infections are highly endemic amongst people who are poor with little or no access to basic services and infrastructures. Under these circumstances, environmental improvements including access to safe drinking water, basic sanitation, and hygiene have been demonstrated effective to reduce the burden of those infections [12, [22][23][24].
This aim of this study was to investigate the prevalence and associated risk factors of intestinal helminths and GP in schoolchildren from the Chimborazo and Guayas provinces, two regions were little or no information at all was available on the epidemiology of these pathogens. Microscopy was conducted for all samples collected and PCR-based methods for molecular characterization of some relevant GPs.  18 May 2020) to meet the principles established by the Helsinki Declaration and by the Spanish/Ecuadorian legislations regarding biomedical research and personal data protection. Formal written consent was obtained from each parent/guardian of the participants.

Study design and setting
A prospective cross-sectional study was conducted in asymptomatic schoolchildren (3-11 years) in the Ecuadorian municipalities of Penipe and Pallatanga (in the Sierra region of the Chimborazo province) and General Antonio Elizalde (GAE, in the Coast region of the Guayas province). Stool samples were collected from participating schoolchildren from June to August 2020 and analysed for the presence of intestinal parasites by microscopy. In parallel, molecular (PCR and Sanger sequencing) methods were used to detect and genotype the main diarrhoeacausing protists species (Entamoeba spp., G. duodenalis, E. bieneusi, Blastocystis sp., and Cryptosporidium spp.).
In order to contextualise the reality of each municipality analysed and according to the available data, it should be noted that Pallatanga has a Gini coefficient, which is a measure to represent income inequality within a social group, of 0.37; Penipe of 0.35; and GAE of 0.30 [41]. The mean number of years of schooling of the population in Pallatanga is 6.2 years; in Penipe, it is 7.7 years; and in GAE, it is 8.3 years [41,42]. The percentage of households with inadequate sanitation facilities in Pallatanga is 14.9%; in Penipe 11.4%, and in GAE 5.7%. The percentage of homes with a public water supply in Pallatanga is of 44.8%, in Penipe of 76.6%, and in GAE of 63.2% [43].

Sample collection
Schools were approached in collaboration with the local university (Escuela Superior Politécnica de Chimborazo, ESPOCH). School heads were contacted and invited to participate in the study. Three public primary schools agreed to participate each with 355-510 schoolchildren. There were 372 schoolchildren enrolled, aged 3-11 years (median: 8.0 years), of which 115 were enrolled in GAE, 161 in Pallatanga, and 96 in Penipe. The rate ratio of boys/girls was 0.86. All recruited schoolchildren were from communities of low-to-medium socioeconomic status.
Informative meetings were personally held with teachers, who explained the goals and procedures of the project to schoolchildren. Parents/legal guardians were informed by formal letter. Participating schoolchildren were provided with uniquely labelled sampling kits (10 mL polystyrene plastic tube, disposable spatula, Graham tape, and instructions) to obtain individual samples. Parents/legal guardians assisted in the collection of samples from consenting schoolchildren and brought the samples to school the next day. A group of 10 students of the ESPOCH were trained to assist in the sample collection. Collected samples were immediately transported at the Chemistry Laboratory of ESPOCH and stored at 4˚C until further processing.

Epidemiological questionnaire
A standardized questionnaire (S3 Table) was provided as part of the sampling kit to be completed by the children's parents/legal guardians. Questions included: (i) demographic characteristics (age, gender, municipality of residence, number of siblings, number of relatives residing at home); (ii) behavioural habits (consumption of unsafe water, hand and fruit/vegetable washing, contact with domestic animals, playing outdoors); and (iii) clinical manifestations (diarrhoea, constipation, abdominal pain, vomiting, weight loss, allergic manifestations, teeth grinding, anal itching, abdominal distention, and flatulence). Completed questionnaires and written informed consents signed by the parent/guardian were returned for collection by each participating schoolchild as described above.

Parasitological assessment
An aliquot of fresh stool samples from each participant was analysed at the Chemistry Laboratory of ESPOCH using the Kato-Katz technique for the detection, identification, and quantitation of intestinal helminths. The remaining faecal material was preserved in 70% ethanol and, together with the Graham tapes, shipped to the Parasitology Laboratory of the Faculty of Pharmacy, University of Valencia (Spain) for further analyses.
A 3-gram stool sample from each participant was concentrated and filtered for 5 min at 2,500 rpm using Midi Parasep filter devices (Apacor Ltd., Wokingham, UK). The sediment obtained was divided in two aliquots. One was used for DNA extraction (see below); the other was fixed with 10% formalin in a 1:3 proportion for microscopic examination; and with a small amount of the remaining sediment thin smears were also prepared and stained with the Modified Ziehl-Neelsen (MZN) for the detection and identification of Cryptosporidium spp. and other coccidian. Graham tapes were examined by 10x microscopy for detecting eggs of Enterobius vermicularis. A sample was considered negative when no parasite structures (eggs, larvae cysts, oocysts or trophozoites) were observed.

DNA extraction and purification
Genomic DNA was isolated from 200 mg of each stool sample using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Extracted and purified DNA was eluted in 100 μL of PCR-grade water and kept at 4˚C until further molecular analysis.

Molecular detection and characterization of intestinal protists
Molecular studies focused on the most prominent GP species studied as a cause of diarrhoea and those under discussion or emergency. The presence of G. duodenalis was investigated using a real-time PCR (qPCR) method targeting a 62-bp region of the small subunit of the rRNA (ssu rRNA) gene of the parasite as the initial screening method [44]. For assessing the molecular diversity of this protozoa at the assemblage and sub-assemblage levels, a sequencebased multilocus genotyping (MLST) scheme targeting the genes encoding for the glutamate dehydrogenase (gdh), β-giardin (bg), and triose phosphate isomerase (tpi) proteins was adopted. Only samples that yielded cycle threshold (C T ) values <32 in qPCR were assessed under this MLST scheme. A semi-nested PCR was used to amplify a 432-bp fragment of the gdh gene [45], and nested PCRs were used to amplify 511 and 530 bp fragments of the bg and tpi genes, respectively [46,47]. Detection and differential diagnosis between pathogenic E. histolytica and non-pathogenic E. dispar was carried out by a qPCR method targeting a 172-bp fragment of the ssu rRNA gene of the E. histolytica/E. dispar complex [48,49]. The presence of Cryptosporidium spp. was assessed using a nested-PCR protocol to amplify a 587 bp fragment of the ssu rRNA gene of the parasite [50]. Identification of Blastocystis sp. was achieved by a direct PCR protocol targeting a 600 bp fragment of the ssu rRNA gene of the protist [51]. Detection of E. bieneusi was conducted by a nested PCR protocol to amplify the internal transcribed spacer (ITS) region as well as portions of the flanking large and small subunit of the ribosomal RNA gene as previously described [52]. Detailed information on the oligonucleotides and the PCR conditions used for the molecular identification and/or characterisation of the unicellular parasites investigated in the present study is presented in S4 and S5 Tables, respectively.
All the qPCR protocols described above were conducted on a Corbett Rotor Gene 6000 real-time PCR system (Qiagen). Reaction mixes always included 2x TaqMan Gene Expression Master Mix (Applied Biosystems, CA, USA). All the direct, semi-nested, and nested PCR protocols described above were conducted on a 2720 Thermal Cycler (Applied Biosystems). Reaction mixes always included 2.5 units of MyTAQ DNA polymerase (Bioline GmbH, Luckenwalde, Germany), and 5x MyTAQ Reaction Buffer containing 5 mM dNTPs and 15 mM MgCl 2 . Laboratory-confirmed positive and negative DNA samples of human origin for each parasitic species investigated were routinely used as controls and included in each round of PCR. PCR amplicons were visualized on 1.5-2% agarose gels (Conda, Madrid, Spain) stained with Pronasafe (Conda) nucleic acid staining solution. A 100 bp DNA ladder (Boehringer Mannheim GmbH, Baden-Wurttemberg, Germany) was used for the sizing of obtained amplicons.

Sequence analyses
PCR products of the expected size on agarose gel were directly sequenced in both directions using appropriate internal primer sets (S4 Table). DNA sequencing was conducted by capillary electrophoresis using the BigDye Terminator chemistry (Applied Biosystems) on an on ABI PRISM 3130 automated DNA sequencer.
Raw sequencing data in both forward and reverse directions were viewed using the Chromas Lite version 2.1 (Technelysium Pty Ltd., South Brisbane, Australia) sequence analysis program (https://technelysium.com.au/wp/chromas/). The BLAST tool (http://blast.ncbi.nlm.nih. gov/Blast.cgi) was used to compare nucleotide sequences with sequences retrieved from the NCBI GenBank database. Generated DNA consensus sequences were aligned to appropriate reference sequences using the MEGA version 10 software [53] to identify Giardia species and assemblages/sub-assemblages, Cryptosporidium species and E. bieneusi genotypes.

Phylogenetic analysis
Nucleotide sequences obtained in this study and E. bieneusi nucleotide sequences for genotypes previously identified in humans and animals as well as appropriate reference sequences to include all E. bieneusi groups retrieved from GenBank were aligned with the Clustal W algorithm using MEGA X [53]. Phylogenetic inference was carried out by the Neighbor-Joining (NJ) method as previously described [54]. Genetic distance was calculated with the Kimura parameter-2 model using MEGA X [53].

Data analysis
Descriptive statistics were calculated, including measures of central tendency (mean and median), measures of dispersion (standard deviation, range, and coefficient of variation) and measures of shape (asymmetry and pointing) for quantitative variables, as well as the absolute and relative frequencies for the qualitative variables. Association analyses were performed, stratifying by gender and age to observe possible heterogeneity in the results according to these factors. When data were stratified and the sample size was small (n < 15), non-parametric tests were used (Fisher's exact test and Mann-Whitney U test). To test the null hypotheses of no association between intestinal parasitism and demographic and socioeconomic factors and lifestyle habits controlling by confounders or effect modifiers, multiple logistic regression analyses with dummy variables for categorical terms, were applied. The magnitude of the association was expressed as adjusted odds ratios (AORs) with a 95% confidence interval (CI). A P-value �0.05 was considered statistically significant. All the variables were analysed using SPSS software version 26.0 (Statistical Package for Social Sciences, Chicago, IL, USA).

Occurrence of intestinal parasites by optical microscopy
Overall, 63.2% (235/372) of the participating schoolchildren were infected/colonized by at least one intestinal parasite. Global prevalence rates for each sampling site varied from 41.7% (48/115) in GAE, 71.9% (69/96) in Penipe and 73.3% (118/161) in Pallatanga. The frequency of intestinal helminths and protists by optical microscopy in the surveyed schoolchildren by municipalities is shown in Table 1.
Schoolchildren in Pallatanga were more likely to harbour helminths including E. vermicularis, H. nana, A. lumbricoides and T. trichiura. Schoolchildren in GAE were significantly less
Out of the six gdh sequences, one (16.7%) and two (33.3%) were assigned to the sub-assemblages AI and AII, respectively, showing 100% identity with reference sequences L40509 and L40510. Assemblages BIII and BIV were identified in two (33.3%) and one (16.7%) isolates, respectively. The two BIII sequences differed by 3-7 SNPs with reference sequence AF069059.  The only BIV sequence identified varied by a single SNP from reference sequence L40508 (Table 4). Out of the four bg sequences, one (25.0%) was identified as sub-assemblage AII and was identical to reference sequence AY072723. The remaining three sequences (75.0%) were identified as assemblage B and differed from 2-3 SNPs from reference sequence AY072727. Because bg is unsuitable for intra-assemblage discrimination for assemblage B, no information at the sub-assemblage level was available at this locus (Table 4).
Finally, out of the seven tpi sequences, four (57.1%) were identified as sub-assemblage AII and were identical to reference sequence U57897. The remaining three sequences (42.9%) differed by 1-5 SNPs with reference sequence AF069561 (Table 4).

Molecular characterization of Blastocystis sp
A total of 92 faecal DNA samples yielded amplicons of the expected size on agarose gels compatible with Blastocystis sp. Of them, 91.3% (84/92) were successfully subtyped. The remaining eight isolates produced unreadable or poor-quality Sanger sequences (associated to faint bands on agarose gels) and were conservatively not considered as true Blastocystis-positive samples.

Molecular characterization of Enterocytozoon bieneusi
In the case of microsporidia, detection was performed by nested-PCR and only for Enterocytozoon bieneusi that is the most prevalent species causing human microsporidiosis [17]. Out of the six Enterocytozoon-positive isolates, four (66.7%, 4/6) were identified as genotype A, and one (16.7%, 1/6) as genotype KB-1, both showing 100% identity with reference sequences AF101197 and MN136778, respectively. The remaining isolate (16.7%, 1/6) corresponded to a novel genotype named HhEcEb1. The identity of this novel E. bieneusi genotype were confirmed in two independent PCR reactions. Novel HhEcEb1 differed from known genotype A by a single SNP (T to C) in position 158 of reference sequence AF101197. Phylogenetic analysis revealed that novel HhEcEb1clustered within zoonotic Group 1 (Fig 2).

Risk factors of intestinal parasites
Logistic regression analyses were conducted to assess the risk of intestinal parasitism based on the demographic, socioeconomic, and behavioural features of the surveyed schoolchildren population. Multivariate logistic regression analysis computing the whole study population showed that residing in a particular municipality was the only variable determining the presence of intestinal parasites (S6 Table). Schoolchildren living in Pallatanga or Penipe were 2.2-fold more likely to harbour helminths and/or GP infections than those living in GAE (AOR = 2.296; 95% CI: 1.226-4.301, P = 0.009; AOR = 2.231; 95% CI: 1.067-4.665, P = 0.033).
When data were stratified by municipality of origin, other variables reached statistical significance and they could be identified as risk factors for intestinal parasites, including more than five persons residing in the same house in GAE (AOR = 22.634; 95% CI: 1.778-288.093, P = 0.016); and no handwashing after bathing (AOR = 0.54, 95% CI = 0.005-0.541, P = 0.013)

PLOS NEGLECTED TROPICAL DISEASES
Prevalence, molecular diversity and risk factors of intestinal parasites in Ecuador and consumption of unsafe water (AOR = 0.127, 95% CI: 0.220-0.743, P = 0.022) in Penipe (S6 Table). None of the variables considered in the study increased the likelihood of being infected by intestinal parasites in Pallatanga.

Fig 2. Phylogenetic relationships among Enterocytozoon bieneusi genotypes identified in this study.
All genotypes identified in humans and animals, and genotypes to cover all groups of E. bieneusi were included for comparative purposes. Analyses were inferred by a Neighbor-Joining method of the entire ITS of rRNA gene based on genetic distances calculated by the Kimura two-parameter model (MEGA X software). All nucleotide sequences include host information with the GenBank accession number in parenthesis. Nucleotide sequences determined in this study are identified with red circles before the genotype name. https://doi.org/10.1371/journal.pntd.0011339.g002

Prevalence, molecular diversity and risk factors of intestinal parasites in Ecuador
Despite the fact that the children are school-going and apparently asymptomatic, the most prevalent clinical manifestations reported by the participating children were abdominal pain (40.6%, 151/372), flatulence (29.0%, 108/372), and diarrhoea (26.1%, 97/372).

Discussion
We studied the epidemiology and risk factors for helminths and GP infections in schoolchildren from the Chimborazo and Guayas provinces, Ecuador. Although microscopy was used as screening test, molecular (PCR and Sanger sequencing) methods were also used for in-depth investigation of the main parasitic intestinal protist species. Six out of 10 children had at least one intestinal parasite species. E. vermicularis, followed by A. lumbricoides and T. trichiura were the most common species among helminths, and Blastocystis sp., E. coli and G. duodenalis (12.6%) among GP. Regarding molecular diversity, assemblages A and B were detected within G. duodenalis, and subtypes ST1-ST4 within Blastocystis sp. Three genotypes, two known (A, KB-1) and a novel (HhEcEb1) were identified within E. bieneusi. Several household and behavioural factors indicative of greater poverty and social marginalisation were associated with helminths and GP infection risk. Household overcrowding, consumption of unsafe water, and poor personal hygiene habits were risk factors for childhood intestinal parasitic diseases.
Strengths of this survey include a large sample size representative of three epidemiological scenarios with distinct environmental, climatic and socioeconomic conditions. The use of PCR-based methods for GP detection and genotyping allowed for i) improved prevalence estimates compared to more traditional (microscopy) methods, and ii) assessing potential sources of infection and transmission pathways. This is one of the main contributions of the study, as GP molecular data is particularly scarce in Ecuador. This study provides also the first description of microsporidiosis by E. bieneusi in the country.
The roundworm A. lumbricoides and the whipworm T. trichiura have been consistently reported as the most dominant STH present in Ecuadorian human populations, with infection varying from 7-45% (A. lumbricoides) and 0.5-25% (T. trichiura) (S1 Table). In contrast, we identified the presence of A. lumbricoides and T. trichiura at comparatively lower (9.4% and 5.1%, respectively) infection rates. This result is probably associated to Mass Drug Administration (MDA) campaigns with albendazol by the Ministry of Health. Remarkably, the pinworm E. vermicularis was the most prevalent nematode detected by the Graham test in the surveyed paediatric population (16.7%), a much higher rate than that (0.5%) identified in indigenous communities in Santo Domingo de los Tsáchilas province by conventional coproparasitological procedures [31]. The Graham's perianal swab method is the recommended procedure for the detection of E. vermicularis in young children [55], these data suggest that the true prevalence of this helminth is likely much higher than initially anticipated.
Recent studies investigating the epidemiology of STH in Ecuador have provided interesting data. A longitudinal birth cohort study (new-borns to 8 years of age) has shown that infections with A. lumbricoides and T. trichiura peak between 3 and 5 years [56]. Marked differences in STH prevalence rates have also been documented nationwide. Thus, comparatively fewer A. lumbricoides and T. trichiura infections occur in coastal and highland areas than in the Amazon region where greater poverty and inadequate sanitation likely favour transmission [57,58]. Local spatial clustering of STH infection has been observed among indigenous Shuar inhabiting two regions of Amazonian Ecuador [59]. This seems to be also the case of our study, where schoolchildren in Pallatanga or Penipe were at higher risk of infection by STH and/or GP than their counterparts in GAE.
Several environmental, sociodemographic, and behavioural factors have been linked to an increased risk of STH infection. T. trichiura infection intensity were lower among Shuar indigenous people living in houses with wood floors than those with dirt floors [60]. Indeed, household dust samples have been shown to contain DNA for a variety of STH including S. stercoralis (52%), A. lumbricoides (39%), Toxocara spp. (42%), hookworm (18%) and T. trichiura (8%) [61]. In the present study, determinants of infection risk included household overcrowded conditions, consumption of unsafe water, and inappropriate personal hygiene practices, highlighting the strong link between poverty and STH infections.
Regarding GPs, microscopy examination of faecal specimens allowed the identification of a well-recognized pathogen (G. duodenalis), two species of uncertain pathogenicity (Blastocystis sp. and D. fragilis), and commensals of the genera Entamoeba (E. coli, E. dispar, E. hartmanni), Endolimax nana and Chilomastix mesnili. G. duodenalis is widely present in Latin American countries, where microscopy-based prevalence rates in humans have been typically reported in the range of 10-50%, mostly in paediatric populations [62].
In Ecuador, G. duodenalis has been identified at prevalence rates of 26-33% in children with clinical manifestation, of 16-40% in asymptomatic children, and of 4-20% in rural and urban community-based surveys (S2 Table). In the present study, overall G. duodenalis prevalence was of 12.6% and of 26.6% by microscopy examination and qPCR, respectively. This finding confirm i) the superior diagnostic performance of molecular methods over conventional assays, and ii) the suitability of qPCR for the detection of subclinical Giardia spp. infections in endemic areas. G. duodenalis is a species complex consisting of eight assemblages (A-H), with assemblages A and B the dominant assemblages in humans, C and D in canids, E in hoofed animals, F in felids, G in rodents and H in marine mammals [21]. In the present study, most (82.8%, 82/99) of the Giardia-positive samples by qPCR yielded C T values �32, indicative of moderate-to-low parasite burdens. This fact explains the low amplification rates obtained at the gdh, bg, and tpi markers, all of them single-copy genes with limited sensitivity compared with the multiple-copy ssu rRNA gene used in qPCR for detection purposes. Our genotyping analyses revealed that assemblage A was more prevalent than assemblage B (50.0% vs. 37.5%). Similar frequencies (60% vs. 20%) have been previously identified in apparently healthy children in Pichincha province [37], but the opposite trend (32% vs. 61%) was observed in children with and without clinical manifestations in Esmeralda's province [34]. Mixed A+B infections have been found in 7-12% of isolates (including this study), suggestive of epidemiological scenarios where infection/re-infection events are frequent. Overall, the vast majority of human cases of giardiasis characterised in Ecuador are due to assemblages A and B. This finding supports the notion that human G. duodenalis infections are primarily of anthropic nature. However, and unknown fraction of these infections might be of animal origin, as both assemblages A and B are zoonotic. Although probably infrequent, zoonotic transmission events are possible, as demonstrated by the occasional presence of canine-adapted assemblage C in two asymptomatic children in Pichincha province [37].
In the present study, Blastocystis sp. carriage rates were estimated at 39.2% by microscopy and at 22.6% by ssu-PCR. One explanation for this discrepancy could be an incorrect diagnosis, because microscopy detection and identification are challenging [63]. The epidemiology of Blastocystis sp. in Ecuador is largely unknown. The protist has been identified in 7-20% of asymptomatic individuals of all age groups both in urban and rural settings in Esmeraldas, Pichincha, and Santo Domingo de los Tsáchilas provinces by conventional microscopy, but a carriage rate as high as 81.5% was identified by PCR in similar populations in Esmeraldas and Manabí provinces (S2 Table). These data seem to suggest that Blastocystis sp. is ubiquitous in the Ecuadorian population. A total of 34 Blastocystis sp. subtypes (STs) are currently recognized, of which ST1-ST10, ST12, ST14, ST16, ST23, and ST35 have been reported in humans [20,64]. ST1 (14.4%), ST2 (10.1%), ST3(19.2%), and ST4 (2.3%) are the most frequent Blastocystis sp. subtypes identified in humans in the Americas [65]. Interestingly, in the present study ST1-ST3 were found at very similar frequencies (27-30%), with ST4 being identified at a lower rate (15.0%). Of note, this is the first report of ST4 in Ecuador. In the only previous molecular epidemiological survey conducted in the country, ST1, ST2 and ST3 (alone or in combination) were identified in asymptomatic individuals in Esmeralda and Manabí provinces [39].
The fungi-related E. bieneusi is an emerging parasite responsible for 90% of human microsporidiosis reported worldwide [19]. It occurs most frequently in immunocompromised individuals [66], but it has also been reported in apparently healthy populations that might act as unnoticed disseminators [67]. More than 600 E. bieneusi genotypes distributed in 11 distinct phylogenetic groups have been described to date, which Group 1 and Group 2 comprising those with zoonotic potential [19]. Perhaps the most important contribution of this survey is the detection of E. bieneusi for the first time in Ecuador. Three different genotypes of E. bieneusi were identified, two known genotypes (A, KB-1) and a novel genotype (HhEcEb1). This finding demonstrated that an unexpected high diversity of E. bieneusi genetic variants were circulating in Ecuadorian asymptomatic children at low (but still significant) frequency rates. Remarkably, the identification of KB-1 represents also the second report of this genotype in humans globally after its initial description in a Chinese toddler attending kindergarten [68]. Previously, KB-1 was only described in a captive baboon in Kenya [69]. In contrast, the genotype A (syn. Peru1) of E. bieneusi is a common finding in human populations worldwide, particularly in immunocompromised patients [70], and has also been reported in other animals such non-human primates, dogs, or birds [19]. In a previous microscopy-based study conducted in Guayas province, microsporidial spores (of unknown species) were detected in 25% of HIV+ patients presenting with diarrhoea [35]. Our data contribute to improve our knowledge on the epidemiology of E. bieneusi in South America, a geographical region where this information is particularly scarce.
Previous studies conducted in Ecuador have reported the presence of members of the Entamoeba complex (E. histolytica/E. dispar/E. moshkovskii/E. bangladeshi) in 13-57% of the faecal samples examined by microscopy in different human populations (S2 Table). Because the transmission forms of these protists are morphologically indistinguishable, diagnostic differentiation of pathogenic E. histolytica and non-pathogenic Entamoeba species must be based on molecular methods. When PCR was used for this purpose, E. histolytica was detected at a prevalence rate of 2.8% (compared with the 69.8% of E. dispar) in rural and urban dwellers in Esmeraldas and Pichincha provinces [32]. In the present survey E. histolytica was not detected, whereas 14.5% of faecal DNA samples tested positive for E. dispar by qPCR.
Remarkably, no Cryptosporidium oocysts were detected using specific MZN staining. This apicomplexan protist has been identified at prevalence rates of 3-14% using microscopy as detection method in several Ecuadorian provinces including Azuay, Chimborazo, Esmeraldas, Pichincha, and Santo Domingo de los Tsáchilas provinces (S2 Table). As far as we know, no molecular-based studies have addressed the genetic diversity of Cryptosporidium spp. of human origin in Ecuador, this being a pending task that should be completed in future epidemiological surveys.
Our multivariate logistic regression analysis revealed marked differences of helminths and GP infection according to the municipality of origin, likely reflecting differences in socioeconomic status and sanitary conditions. When epidemiological data were stratified by municipality, variables associated with an increased risk of infection by intestinal parasite included overcrowded conditions (more than five people per household), limited or no access to safe drinking water, and inappropriate personal hygiene practices. These results are in line with those obtained in previous community surveys carried out in Ecuador [56,59,60]. Taken together, these findings demonstrate the strong link present between poverty and intestinal parasitic infections. Interventions directed to improve drinking water quality, household flooring and toilet facilities and personal hygiene practices are greatly needed to minimize helminths and GPs transmission to humans.
Several limitations might have compromised the accuracy of the results obtained and the conclusions reached in this study. First, epidemiological data generated here might not be representative of the whole Ecuadorian scenario; second, microscopy examination of a single stool sample per participating child was used as screening method, likely sub-estimating the true prevalence rates of helminths and GP and, in the case of the latter, reducing the chances of genotyping analyses; third, concentration devices involving centrifugation steps for the recovery of labile parasitic forms (e.g., trophozoites) might compromise the quality/quantity of template DNA used in PCR testing; fourth, the detection of D. fragilis trophozoites by direct microscopy would have required faecal smears stained with trichrome or another permanent stain and not, as in our case, directly on droplets of the sediment, therefore our microscopic data may be underestimating the real situation; fifth, insufficient number of participants in some locations might have compromised the robustness of the statistical analyses conducted, particularly when available data were stratified by municipality of origin; sixth, questionnaire data may have resulted in some recall/recording bias; and seventh, this surveys has focused on the human reservoir of helminths and GP, but the animal and environmental reservoirs have not been considered.
To conclude, this study shows that helminths and protists remain at high prevalence rates in Ecuadorian schoolchildren living in poor-resource settings, despite the MDA campaigns regularly conducted on this specific population. Our results indicate that albendazol mass administration, specifically designed to target STH, is insufficient to effectively control these pathogens and must be complemented with interventions directed to improve household quality and living conditions, sanitation, and health education for schoolchildren and families. Together with access to safe drinking water, these measures would contribute to the reduction of intestinal parasites in poor communities in Ecuador. This study also provided relevant molecular data, including the first description in Ecuadorian human populations of i) the genetic diversity of E. bieneusi, including a novel genotype, and ii) the presence of Blastocystis ST4, for the first time described in Ecuador.
Molecular analytical methods are therefore powerful complementary tools to microscopy, and essential to unravel the epidemiology of intestinal parasites (particularly GP) including sources of infection, transmission routes, zoonotic potential, and diagnostic differentiation between pathogenic and commensal species. More research should be conducted to ascertain the role of domestic and wild animals and the environment (source waters, soil, grass and fresh produce) in the epidemiology of STH and GP in endemic areas.
Supporting information S1