An evaluation of water , sanitation , and hygiene status and household assets and their associations with soil-transmitted helminthiasis and reported diarrhea in Nueva Santa Rosa , Guatemala

Soil-transmitted helminth (STH) infections and diarrheal illness affect billions of people yearly. We conducted a cross-sectional survey in Nueva Santa Rosa, Guatemala to identify factors associated with STH infections and diarrhea using univariable and multivariable logistic regression models. On multivariable analyses, we found associations between STH infections and two factors: school-aged children (odds ratio (OR) vs. adults: 2.35, 95% CI 1.10–4.99) and household drinking water supply classified as ‘other improved’ (OR vs. ‘improved’: 7.00, CI 1.22–40.14). Finished floors in the household vs. natural floors were highly protective (OR 0.16, CI 0.05–0.50) for STH infection. In crowded households (>2.5 people/bedroom), observing water present at handwashing stations was also protective (OR 0.32, CI 0.11–0.98). When adjusted for drying hands, diarrhea was associated with preschool-age children (OR vs. adults: 3.33, CI 1.83–6.04), spending >10 min per round trip collecting water (OR 1.90, CI 1.02–3.56), and having a handwashing station 10 m near a sanitation facility (OR 3.69, CI 1.33–10.21). Our study indicates that familiar WASH interventions, such as increasing drinking water quantity and water at handwashing stations in crowded homes, coupled with a hygiene intervention like finished flooring may hold promise for STH and diarrhea control programs.


INTRODUCTION
The World Health Organization (WHO) estimates that 1.5 billion people are afflicted with soil-transmitted helminth (STH) infections (WHO b, ). STH infections are a category of parasites that commonly include roundworm (Ascaris lumbricoides), whipworm (Trichuris trichiura), and hookworm (Necator americanus and Ancylostoma duodenale). In 2010, an estimated 5.3 billion people worldwide lived in areas with stable transmission of at least one STH species (Pullan & Brooker ). STH infections are treated with anthelminthic drugs, which are commonly distributed in mass drug administration (MDA) campaigns to preschool-aged children (PSAC) and school-aged children (SAC). In addition to anthelminthic treatment, recent studies also demonstrate the combination of safe drinking water, appropriate sanitation facilities, and good hygiene practices (WASH) along with STH programs (i.e., MDA campaigns) reduce STH reinfection (Strunz et al. ).
The WHO estimates 1.7 billion cases of childhood diarrhea occur globally each year and that diarrhea is the fifth Additional evidence is needed to better understand the factors for STH infection and diarrhea and to help guide programmatic decision making focused on the integration of WASH interventions. The aim of this study was to explore the WASH and demographic factors associated with diarrhea and STH infections, in communities of southeastern Guatemala. We conducted a cross-sectional survey in 2010 to collect information about household WASH condition, knowledge, attitudes, practices related to WASH, demographics, animal ownership, and examined associations with laboratory-confirmed STH infections and selfreported diarrhea.

Study design and sampling methods
In 2010, we conducted a cross-sectional survey in Nueva Santa Rosa (NSR) municipio (county), southeast of Guatemala City, Guatemala to determine multiple outcomes, including the population prevalence/incidence of and health-seeking behaviors for soil-transmitted helminthiasis (STH), diarrhea, and influenza-like illness (ILI). A secondary objective of this study was to complete a pilot study to assess the status of WASH in NSR and to evaluate the impact WASH status had on these diseases. The study included PSAC, SAC, women of childbearing age (WCBA), and other adults. Children <1 year of age were not included in the STH analyses due to lack of WHO deworming dosage recommendations for this age group (Montresor et al. ; WHO a). The household sampling frame was determined using aerial photos overlaid by 200 m × 200 m grid cells. The potential residential-associated (PRA) roofs were identified and geolocated (n ¼ 10,770), and a subset of PRA roofs was randomly selected (n ¼ 387). Sample size calculations were performed using Cochran's formula (Cochran ), and sample sizes were calculated for each of the multiple primary outcomes, of which health-seeking behavior for ILI required the largest sample size of 387 PRA roofs (Matanock et al. ). This sample size was insufficient for the WASH component of this study, which was underpowered 2-3 fold but was a pilot test of the survey instruments for future investigations.
Resources were unavailable to increase the size of the study to fully power the WASH pilot at that time. The sample size was converted using modifiers from the number of individuals needed to the number of PRA roofs needed.
These modifiers included design effect, estimated number of persons per household, nonresponse rate, estimated percentage of non-residential roofs in the sample, estimated percentage of roofs that would not be found in the field, and estimated proportion of roofs in the sample that would belong to households already surveyed. PRA roof is not an exact proxy for household because, in this area of Guatemala, dwellings are constructed such that rooms serving different functions are often constructed in separate buildings with separate roofs. Therefore, a single household may have multiple PRA roofs. For example, depending upon the socioeconomic status (SES) and location of an individual or family, a single household may have several buildings (e.g., kitchen, sleeping quarters, and main building). Study staff identified each randomly selected PRA roof in the field with the aid of maps made from the aerial photos and GPS coordinates, and then identified the household associated with each roof, if applicable, to try to recruit the occupants and administer the questionnaire. Further details on the sampling frame design can be found in Matanock et al. ().

Consent and data collection
A cross-sectional household survey and observational environment assessment were conducted to evaluate demographic data, WASH infrastructure and conditions, health status, and health-seeking behaviors of household members. Written consent for study participation was sought from each household member present at the time of interview: adult consent, parental consent, and assent from children aged 7-17 years (see Supplementary Material, Appendix 1-3). The consent form included permission to ask questions about household members both present and not present at the time of interview and permission to collect and test stool specimens and water samples. Then, an adult household spokesperson was identified to complete the questionnaires. In order of preference, this spokesperson was (1) the adult female head of household, (2) the eldest teenage daughter in the presence of an adult household member, (3) the adult male head of household, and (4) any consenting adult for the household. This order of preference reflected the likely level of knowledge about the household WASH status. All persons present during the interview were informed that the spokesperson would be asked the questions but if they wish to refuse disclosing or provide information they could. At the conclusion of the interview, the household spokesperson was given one stool collection kit for every person in the household 1 year of age for STH (Montresor et al. ; WHO a). Household members who were not present at the time of the interview were asked to include their signed consent form with their stool specimens. Stool specimens provided without signed consent forms were not accepted. Diarrhea was self-reported and defined as loose stool in the past 7 days.
Stool specimens were tested using the Mini Parasep ® Fecal Parasite Concentrator (FPC) method (Apacor, Berkshire, UK). Individuals who tested positive for one or more of Ascaris sp., Trichuris sp., or hookworm were considered to be STH-infected. For the environmental assessment, the household spokesperson was asked to show the interviewer the main handwashing place for the household and to demonstrate how she/he washed her/his hands. Additionally, a member of the family took a field staff member to observe and evaluate the toilet or latrine used most often by household members and observed where the family collects their drinking water. During the interview and the environmental assessment, the condition of the house, yard, and household surroundings was also observed using a standardized observational worksheet. A modified definition of the WHO/JMP Water Ladder was used for all analyses comparing 'improved' water (piped household water connection located inside the dwelling, yard, or plot); 'other improved' (public tap, borehole, protected dug well, protected spring, rainwater collection), bottled water; and 'unimproved' water sources (unprotected dug well, unprotected spring, water cart, tanker truck, surface water) (WHO ).

Data management and statistical analysis
Logistic regression models were used to assess the association between the candidate factors and two outcomes: Univariable logistic models of the association between WASH and demographic factors and STH infection were constructed using survey methods to take household clustering and weight by the inverse number of roofs per household. The variable 'age/sex' combined PSAC, SAC, WCBA, and other adults with distinct age groups (1-4 years old, 5-14 years old, 15-44 years old, and >15 years old, respectively). These specific age/sex categories were defined based on at-risk population groups for STH infections and commonly used across deworming programs (WHO ). The age group category for self-reported diarrhea did not differentiate between WCBA and other adults.
Variables initially considered for the multivariable models were identified from a subset of WASH factors chosen a priori, taking into consideration missingness  Trichuris only, and one (1.5%) hookworm only infections.
In Supplementary Material, Table S1, univariable STH results showed key hygiene associations, which included water used for handwashing from a container (OR 11.36, *SES variables were determined by factor analysis of 18 variables concerning family member possessions (including possession of animals and animal types), education level of the mother, monthly household income, and household wall construction material. The analysis generated two factors: the first dealing with household possessions, mother's education, monthly income, and wall construction, and the second dealing with household possession of animals. Only the first factor was statistically significant in univariable analysis. † Finished floors include wood, vinyl, ceramic tiles, cement, carpet, and brick. Natural earthen floors include sand, dung, straw, and sawdust.

CONCLUSIONS
Our study found strong associations between STH infections and two factors: observed household drinking water supplies classified as 'other improved' and SAC. We found a highly protective effect against STH infections with finished floors and a protective effect in crowded households with observed water at handwashing stations.
For diarrhea, we found associations between illness and three factors: preschool age (compared with adults), spending more than 10 min collecting water, and having a handwashing station within 10 m of a sanitation facility.

DECLARATIONS OF INTEREST
None.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.