The village of Lak Sip (Luang Prabang district, Luang Prabang Province) is located in northern Lao PDR at 430 m above the sea level and 15 km distance from the Luang Prabang town (Fig. 1). The village covers an area of about 2 ha and was established in 1962 for the resettlement of Laotian highland communities. The annual rainfall is 1,400 mm and the runoff feeds several streams that run through the village territory, with Houai Xon the main stream traversing through the village. The livelihoods of Ban Lak Sip residents depend on a variety of on-farm activities with upland rice production being the primary agricultural activity in all village households. Most of the population belongs to the Khmu ethnic group. Most houses were of traditional construction, made from wood, and the streets were not paved.

Water sampling was performed at 3 different sites in and near by Ban Lak Sip village with respect to the direction of water flow from the village. The first site was Lak Sip village with the map reference at latitude 19°50′55.615″ and longitude 102°10′4.266″; the second site was upstream of the village (limited human activity) at latitude 19°51′8.781″ and longitude 102°10′7.669″; and the third site was downstream of the village (extensive human activity) at latitude 19°50′54.268″ and longitude 102°8′45.51″. All samplings were completed during the monsoon period in May 2012 because the rainy season is preferred to the drier months due to rising water tables and increased run-off which cause a spike in waterborne pathogens and parasitosis.

At each site, 2 bags of surface water (3 L each) were collected and concentrated using either the tangential flow filtration (TFF) method or the Lifesaver bottle (http://www.lifesaversystems.com, Lifesaver Systems, Great Britain). For the TFF method, the 3 L water samples were passed through a 100 kDa hollow fiber cartridge using a peristaltic pump to trap both bacteria and viruses. The trapped micro-organisms were then flushed out of the cartridge and collected in 100 ml of water. Similarly, the ultra-filtration membranes encased within the Lifesaver bottle have been shown to be able to trap both bacteria and viruses. Water samples were poured into the bottle and mechanical pumping created the pressure required to force the dirty water through the ultra-filtration cartridge, where micro-organisms were trapped. Then, 100 ml of clean water was poured into the bottle which was then shaken vigorously to dislodge the microorganisms from the cartridge before the water was collected. Water samples collected from Lak Sip village and the upstream site were concentrated using the Lifesaver bottle and the TFF respectively. For the water samples collected from the downstream site, 1 bag (3 L) was concentrated using the TFF and the other was concentrated using the Lifesaver bottle. An aliquot of the concentrated water samples collected from each site was plated on trypticase soy agar (TSA). Viable colonies obtained from the TSA plates were pooled. DNA was extracted from the concentrated raw water or from pooled bacterial colonies using a FastDNA Spin Kit (MP Biomedicals, Solon, Ohio, USA) according to the manufacturer’s protocol. The extracted nucleic acids were tested for the presence of enteric pathogens using a commercially available multiplex PCR kit, Seeplex ACE Diarrhea Detection kit (Seegene, Seoul, Korea), according to the manufacturer’s protocol.

Pathogens that could be detected by this kit include Salmonella spp. (S. bongori and S. enterica), Shigella spp. (S. flexneri, S. boydii, S. sonnei, and S. dysenteriae), Vibrio spp. (V. cholera, V. parahaemolyticus, and V. vulnificus), Clostridium difficile toxin B, Campylobacter spp. (C. jejuni and C. coli), Clostridium perfringens toxin, Yersinia enterolitica, Aeromonas spp. (A. salmonicida, A. sobria, A. bivalvium, and A. hydrophila), Escherichia coli O157:H7, verocytotoxin-producing E. coli (VTEC), human enteric adenovirus, Group A rotavirus, norovirus GI/GII, and astrovirus.

Human fecal samples were collected every morning in plastic containers (around 5 g) which had been distributed to villagers by the help from local health volunteers and returned to the field laboratory for further examination within the same day. Geographical coordinates of each participant were recorded at the individual households using GPS device Garmin Montana 600 (Garmin International Inc., Olathe, Kansas, USA).

The fecal samples were firstly checked for parasitic infections using the direct smear technique to initially screen for helminth and protozoan infections. Lugol’s iodine was added in simple smears, which helps to stain and produce better visualization of some features for protists identification. Intestinal protist identification was performed by consulting the taxonomic literature [21,22]. Subsequently, a concentration method, modified-Katz’s thick smear technique [23] was followed to identify and quantify helminth eggs or larvae presented in fecal samples. The slides were coated with an equal amount of sieved fecal samples (around 43 mg) through wire-net filter, and then covered with a cellophane strip pre-soaked in the Katz’s solution. The prepared slides were left at room temperature for 30 min allowing fecal contents to dry and clear, then the slides were examined and all helminth eggs were counted in the whole smear under light microscope by experienced parasitologists. Duplication of both modified-Katz’s thick smear and simple smear was performed for each fecal sample. To assess the intensity of helminthic infections, the number of eggs per gram (EPG) of soil-transmitted helminths species (i.e., Ascaris lumbricoides, hookworm, and Trichuris trichiura) was calculated following Katz et al. [23].

The study procedures concerning human sample collection, laboratory investigation, interviewing, and questionnaire were reviewed and approved by National Ethics Committee for Health Research, Lao Ministry of Health (reference no. 51/NECHR). First, the investigator team conducted the meeting with local public health officers, village chief and local health volunteers to inform the details of the research study. Then, community amplifying system was used to spread the research activity in local language in every household. All participants who agreed to join the study were asked to read through the participant information sheet explaining about the objectives, procedures, possible risks, and benefits of the research project. The signed consent/assent forms were consequently obtained from either participant themselves or children’s guardians prior to fecal sample submission and personal information taking through interviewing and questionnaire. All participants infected with intestinal parasites (helminths and protists) were informed of the diagnostic results and subjected for treatment under control by either qualified clinicians from the investigator team or local doctors from Luang Prabang Hospital, Lao PDR.

All participants providing fecal samples were interviewed and completed a questionnaire to enable the collection of epidemiological data, i.e., sanitary conditions, possession of pet at home, food consumption habits, source of water consumption, and anti-parasitic treatment history. These data were merged with the personal and demographic informations (sex, age, address, educational level, and occupation) extracted from record of local primary health care units (PHCU).

Principal Component Analyses (PCA) were performed on (1) individual parasitic infections (helminths and protozoans) and (2) individual questionnaire (socio-economics and habits) of the villagers. These analyses were followed by a Canonical Analysis (CA) representing the correlations among the 2 sets of variables, i.e., socio-economics/habit and parasitic infections. Wilk’s lambda (Wilk’s L) was used to test how well each independent variable contributes to the statistical multivariate model. The scale of Wilk’s L ranges from 0 (full discrimination) to 1 (no discrimination). These analyzes were performed using the packages MASS version 7.3–45 [24] and CCA version 1–2 embedded in R freeware [25].

The interviews and informations obtained from the PHCU in the village of Lak Sip allowed data to be obtained from 305 participants who had also been screened for parasitological data. Of these, 159 (52.1%) were males, and data was missing for 2 of the respondents. The average age of the population was 28.8± 19.9 years, reflecting the fact that it was mostly a young population with only a few individuals older than 60 years. In the analysis performed, we considered 4 age classes, i.e., (1) less than 15 years, (2) between 16 to 30 years, (3) between 31 to 45 years, and (4) greater than 45 years. The level of education achieved by most individuals was to the primary/early college level (69.2%), with only 13.4% having achieved a college qualification and 18% of the population being uneducated. Students comprised 29.2% of the respondents, 18.7% were farmers, 20% were dealers or merchants, and 21% were without paid employment. There was a significant association between gender and level of education, with men being formally educated to a higher level than women (chi-square= 10.05, P < 0.05). The food consumption habits of the villagers of Lak Sip are summarized in Table 1. In relation to sanitary conditions and preventive measures, 4.9% of the individuals did not use toilets, 10.5% did not systematically wear shoes, and 43.0% of the inhabitants had pets.

Our analysis indicated the presence of enteric pathogens in the water samples taken from the 3 water sampling sites in the vicinity of Lak Sip village as shown in Table 2. Upon concentrating water samples and culturing on TSA media, we only detected Aeromonas spp. at the upstream site. However, additional pathogens were found at the Lak Sip village site, including, Aeromonas spp., E. coli H7, E. coli O157:H7, VTEC, and Vibrio spp., whereas the downstream site was positive for Aeromonas spp., E. coli H7, Shigella spp., VTEC, Vibrio spp., and enteric adenovirus.

The water samples were negative for the other enteric pathogens detectable by the Seeplex ACE Diarrhea Detection kit, i.e., Salmonella spp., Clostridium difficile, C. perfringens, Campylobacter spp., Y. enterolitica, Group A rotavirus, norovirus GI/GII, and astrovirus.

The overall prevalence of helminth infections was 68.9% and that of protist infections was 27.2%. Eight helminth species were identified, including 2 tapeworm species (Taenia sp. and Hymenolepis diminuta), 1 trematode (Opisthorchis sp.), and 5 nematodes (Ascaris lumbricoides, Trichuris trichiura, Strongyloides stercoralis, trichostrongylids, and hookworms). The dominant parasite was hookworms which were present in 63.3% of the population. The intensities of soil-transmitted helminths, A. lumbricoides, T. trichiura, and hookworms were assessed through calculation of the mean number of EPG as 1,132 (min= 23, max= 11,960), 148 (min= 23, max= 2,047), and 691 (min= 23, max= 6,279), respectively. Six species of protists were identified; Blastocystis hominis, Cyclospora spp., Endolimax nana, Entamoeba histolytica/E. dispar, Entamoeba coli, and Giardia lamblia. Blastocystis hominis was the most prevalent (14.8%). The details are given in Table 3.

From the PCA results, the villagers of Lak Sip appeared very similar in their level and diversity of infection with intestinal helminths and protists, as well as for their socio-economics and/or habits using questionnaire responses (Fig. 2). The bip-lot shows no clear clustering pattern (scattered directions of parameters or vector arrows pointing in) for both intestinal parasite infection of individual inhabitants (Fig. 2A) and distribution of socioeconomic data/attitude among the inhabitants (Fig. 2B). Parasite infections showed a weak correlation with socio-economic as only the first canonical dimension being significant (Wilks L= 0.004, Fig. 3). Since Wilk’s L value is closed to zero revealing association between S. stercoralis infection with the lack of footwear and the age, i.e., higher infection in medium age classes (16–45 years old) than the younger class (< 15 years) and elderly (> 45 years).