The Sustainable Development Goal drinking water indicators include microbiological safety measures, whereas the Millennium Development Goal indicator ‘improved water’ may be microbiologically unsafe. In rural Vanuatu, we undertook household surveys, child anthropometry, and tested stored drinking water, to investigate relationships between water and undernutrition. Using Escherichia coli most probable number, we categorized results according to Compartment Bag Test drinking water cutoffs: <1/100 mL (safe), 1–10/100 mL (intermediate risk), >10–100/100 mL (high risk), and >100/100 mL (very high risk). Of 201 households, 191 (95%) had microbiologically unsafe drinking water, regardless of ‘improved’ status. We investigated cross-sectional associations between households with microbiologically safer drinking water (≤10 E. coli/100 mL) versus ‘improved water’ and undernutrition among children. Of children under 5, 145 (48.8%, 95% CI: 42.8, 54.8) were stunted and 59 (19.1%, 95% CI: 14.4, 23.8) were underweight. Among households with ‘improved water’, the adjusted prevalence ratio (95% CI) of stunting was 0.61 (0.46, 0.80) and underweight was 0.46 (0.29, 0.73) compared with ‘unimproved water’. However, we found no association between having drinking water with ≤10 E. coli/100 mL at one point in time and undernutrition. Longer-term variations in water quality and unmeasured conditions beyond water may have contributed to these associations.

The Sustainable Development Goal (SDG) drinking water indicators incorporated microbiological testing to address concerns that the Millennium Development Goal's (MDG's) ‘improved water’ indicator did not equate with microbiological safety (Clasen 2012; Shaheed et al. 2014; Heitzinger et al. 2015). The MDGs' definition of ‘improved water’ is piped water into dwelling, yard, or plot; piped water to public standpipes; tubewell or borehole; protected dug well; protected spring; or rainwater collection (Box 1). It has been suggested that the health benefits associated with ‘improved water’ coverage have been overestimated (Clasen 2012). In response, the Joint Monitoring Programme (JMP) for Water Supply and Sanitation of the World Health Organization (WHO) and the United Nations International Children's Emergency Fund (UNICEF) developed new indicators for measuring access to safe drinking water for SDG monitoring (SDG 6.1 Global Indicator 6.1.1) (WHO/UNICEF 2017). These SDG indicators recognize that safe drinking water should be compliant with the WHO fecal contamination standard of less than one Escherichia coli colony-forming unit per 100 mL (equivalent to the most probable number, MPN) and compliant with WHO chemical contamination standards (WHO/UNICEF 2017). With the SDG indicators, ‘safely managed drinking water’ is defined as an improved drinking water source that is located on the premises, available when needed, and free of fecal, arsenic, and fluoride contamination (Box 1). It is expected that the SDG water indicators will be more closely related to health risk (WHO 2016).

Box 1
Definitions of SDG and MDG water indicators
SDGsMDGs
Safely managed service A basic/improved drinking water source, located on premises, available when needed and free of fecal (<1 E. coli/100 mL) and priority chemical contamination Improved drinking water source
Piped water into dwelling, yard or plot; piped water to a public standpipes; tubewell or borehole; protected dug well; protected spring; or rainwater collection 
Basic service An improved water point provided collection time is <30 min for a roundtrip 
Limited service Drinking water from unprotected springs or dug wells, carts with small tank/drum, tanker trucks or basic/improved sources with a total collection time of >30 min for a roundtrip Unimproved drinking water source
unprotected springs or dug wells, carts with small tank/drum, tanker trucks, surface water, or bottled water 
No service Drinking water coming from surface water 
SDGsMDGs
Safely managed service A basic/improved drinking water source, located on premises, available when needed and free of fecal (<1 E. coli/100 mL) and priority chemical contamination Improved drinking water source
Piped water into dwelling, yard or plot; piped water to a public standpipes; tubewell or borehole; protected dug well; protected spring; or rainwater collection 
Basic service An improved water point provided collection time is <30 min for a roundtrip 
Limited service Drinking water from unprotected springs or dug wells, carts with small tank/drum, tanker trucks or basic/improved sources with a total collection time of >30 min for a roundtrip Unimproved drinking water source
unprotected springs or dug wells, carts with small tank/drum, tanker trucks, surface water, or bottled water 
No service Drinking water coming from surface water 

Adapted from Ps-Eau (2016).

Child undernutrition is multifactorial and is a recognized consequence of poor nutrition and exposure to fecal contamination. Substandard water, sanitation, and hygiene (WASH) practices can lead to growth faltering through repeated bouts of diarrhea (Guerrant et al. 2013). Another proposed mechanism is the ingestion of large quantities of microbes in water contaminated with feces, which alters gut structure and function, leading to environmental enteric dysfunction (EED). EED is postulated to impair linear growth through insufficient nutrient absorption and chronic systemic inflammation that is independent of diarrhea occurrences (Keusch et al. 2014; Crane et al. 2015). Many studies have described associations between WASH shortcomings and undernutrition, in particular, stunting (Dangour et al. 2013; Vilcins et al. 2018). However, although there is reasonably good evidence for an association between poor sanitation and undernutrition, the association with unsafe drinking water is unclear. Studies show differing relationships depending upon the population and water measure used (Vilcins et al. 2018). Water measures that accurately reflect the water risk are likely to better elucidate this relationship, such as a recent study by Lauer et al. (2018).

WHO recommends an integrated approach to improving child nutrition outcomes that combines both nutrition and WASH interventions. Recently, large clustered randomized control trials (RCTs), the WASH benefits studies (Stewart et al. 2018; Tofail et al. 2018), and the Sanitation Hygiene Infant Nutrition Efficacy (SHINE) trial (Prendergast et al. 2019) concluded that WASH interventions did not improve growth without parallel nutrition interventions. However, the authors noted that uptake of WASH interventions was low and their conclusions may not be generalizable to populations where baseline characteristics differ.

In the Republic of Vanuatu, the proportion of the population with reported access to ‘improved water’ in rural areas is 87% (WHO/UNICEF 2017). However, groundwater in Vanuatu is frequently contaminated with E. coli (Foster et al. 2019). Despite improved water and food security provided by subsistence agriculture (World Vision 2013), the proportion of children under 5 years of age with poor nutrition outcomes in rural Vanuatu is high. The Vanuatu (2014) Demographic and Health Survey reported that in rural populations isolated from the main centers, 32% of children under 5 years were stunted (Vanuatu Ministry of Health 2014). Concurrently, there were just 12% of households reporting child diarrheal diseases in the previous two weeks (Vanuatu Ministry of Health 2014).

We sought to investigate the microbiological status of ‘improved water’ compared with ‘unimproved water’ in Vanuatu and whether there were associations between water microbiological quality and nutrition outcomes in children under 5 years of age in the context of a food secure population with suspected widespread undernutrition concerns. A cross-sectional study was undertaken, involving household questionnaires, observations of WASH facilities, microbiologically testing of stored household drinking water, and child anthropometry in a small, rural population of Vanuatu.

Survey setting and design

We conducted a cross-sectional survey from 19 through 29 January 2015 in an area participating in a three year World Vision Integrated WASH and Nutrition Project in southwest Tanna, an island in Tafea Province, Vanuatu (Figure 1). The overall goal of the World Vision Integrated WASH Project was to reduce child undernutrition, with short- and medium-term objectives to increase WASH infrastructure and improve caregiver WASH and nutrition knowledge and practices. Our survey took place at the end of the second year of the project when the sanitation and hygiene components of the project had been completed, but the water infrastructure component was yet to begin. The study population consisted of 28 villages with approximately 2,500 individuals in total. There was a total of 233 households with at least one child under 5 years of age, and complete enumeration of households and their children under 5 was attempted. A total of 13 households were not included: 12 were not present at the time of the survey as they were visiting non-study communities and one refused because of sick children in the household. A total of 220 households with 320 children under 5 years of age were included in our study. Figure 2 describes the numbers of households and children involved in each component of the study and where data were missing. All households in the study population were remote and rural, and participants were estimated to be in the lowest wealth quintiles (Vanuatu Ministry of Health 2014).

Figure 1

Map of project area within Tanna, Vanuatu. Edited Google Maps locations: Vanuatu and Tanna, Vanuatu.

Figure 1

Map of project area within Tanna, Vanuatu. Edited Google Maps locations: Vanuatu and Tanna, Vanuatu.

Close modal
Figure 2

Numbers of households and children involved in each of the study components in Tanna, 2015. Microbiological water categories according to the Aquagenx CBT method (Aquagenx 2013), based on WHO Guidelines for Drinking Water Quality, 4th ed.

Figure 2

Numbers of households and children involved in each of the study components in Tanna, 2015. Microbiological water categories according to the Aquagenx CBT method (Aquagenx 2013), based on WHO Guidelines for Drinking Water Quality, 4th ed.

Close modal

Following informed consent, interviews were conducted with the primary caregiver of the child and direct observations of WASH facilities were made. All interview and observation responses were recorded directly into smartphones (Samsung Group, Seoul, South Korea) using the survey platform Fieldtask, developed by Smap Consulting (Neil Penman, VIC, Australia).

WASH facilities

Improved water’ and ‘unimproved water’ definitions are based on WHO/UNICEF definitions described in Box 1. In our household questionnaire, adequate water treatment was defined as a treatment method that included a step to effectively remove most pathogens, such as boiling or ceramic filtration (WHO 2013). In observations of WASH facilities, adequate drinking water storage was defined as having drinking water separate from water used for cleaning and other purposes, covered with a lid and with a narrow mouth or a spigot (WHO 2013). An improved latrine was defined as a ventilated improved pit (VIP) latrine or a simple pit latrine with a concrete slab and further defined into SDG indicators ‘basic service’ or ‘limited service’ based on shared status (United Nations Water 2016). An operational handwashing facility was defined as one which was able to be used when study observations took place.

Water sampling and microbiology

We sampled drinking water that was stored at the household in drinking water storage containers. No households had water piped directly into the dwelling. The sources of the stored water included ‘improved water’ from piped water to public standpipes from protected or unprotected sources, and rainwater; and ‘unimproved water’ from unprotected springs, and surface water.

At each household, we collected 100 mL of drinking water into sterile bottles. Any ornamental or other objects were removed from the spigot or mouth of the container before sampling. We used the Compartment Bag Test (CBT) (Aquagenx, USA) method to estimate an MPN of E. coli based on color changes. We processed samples according to the manufacturer's instructions within 6 h of collection (Aquagenx 2013). We recorded color changes 48 h after processing unless all compartments, indicating varying E. coli concentrations, had become positive before 48 h. Positive and negative E. coli controls for each batch of CBTs were tested prior to data collection. Negative controls were also tested during data collection. We categorized results according to Aquagenx CBT recommended categories (Aquagenx 2013), which are based on WHO Guidelines for Drinking Water Quality, 4th Edition (WHO 2011): MPN of <1 E. coli/100 mL (safe), 1–10 E. coli/100 mL (intermediate risk), >10–100 E. coli/100 mL (high risk), and >100 E. coli/100 mL (very high risk).

Anthropometry

Growth measures were used as a proxy for child undernutrition. Anthropometric measurements were carried out on children under 5 years of age and their mothers. Measurements were conducted by trained personnel using calibrated equipment, standardized procedures, and conventional protocols (Cogill 2013). We used WHO recommended cutoffs: Z-score <− 2 was used for stunting, wasting, and underweight; and Z-score <− 3 was used for severe stunting, wasting, and underweight. Anthropometry measurements were excluded if they had missing information or if the Z-score was 5 or more standard deviations from the mean, indicating possible measurement error.

Covariates

Known confounders of the relationship between WASH and undernutrition were investigated, including child age, child gender, maternal height, and education of the primary caregiver. Maternal height was measured objectively at the household, and education was measured by self-report in the household questionnaire. We had insufficient household asset data to generate a wealth index.

Potential confounders of the relationship between water indicators and undernutrition were also investigated and included the WASH facilities described above, minimum dietary diversity (MDD), and diarrhea. Data for MDD were collected according to the WHO Infant and Young Child Feeding (IYCF) recommendation that children 6–23 months of age should be receiving food from at least four of the seven IYCF food groups. The IYCF food groups are grains, roots, and tubers; legumes and nuts; dairy products; flesh foods; eggs; vitamin-A rich fruits and vegetables; and other fruits and vegetables (WHO 2008). Caregivers were asked if each child under 5 years of age had experienced diarrhea (defined as three or more loose stools in the previous 24 h) in the past 7 days.

Statistical analysis

All data were analyzed using Stata 13.1. We used univariate and multivariate analyses to investigate any associations between households having MDG ‘improved water’ and water that tested <1 E. coli/100 mL or >100 E. coli/100 mL. For these analyses, we used a logistic model to generate odds ratios (ORs) because of symmetry with respect to outcome and exposure. We also controlled for village-level variation using clustering.

We then investigated any associations between ‘improved water’ and child undernutrition, and microbiological quality and undernutrition. Because of the low numbers of household drinking water with <1 E. coli/100 mL, these analyses used a dichotomous microbiological indicator of ≤10 E. coli/100 mL (safe or intermediate risk) compared to >10 E. coli/100 mL (high or very high risk).

We estimated unadjusted and adjusted associations using generalized estimating equation (GEE) models to obtain estimates with 95% confidence intervals and P-values which allowed for the clustering. As multiple children under five were surveyed at some households, two levels of data nesting, household and village, were used when any child measure was the outcome. We used independent working correlations with robust standard errors. When weight-for-height Z (WHZ) scores, height-for-age Z (HAZ) scores, or weight-for-age Z (WAZ) scores were the outcomes, we used a Gaussian GEE model to estimate the mean difference (MD) in Z-score. Where wasting, stunting, and underweight were the outcomes, we used a Poisson model (log link with Poisson errors) to generate prevalence ratios (PRs). The robust standard errors also account for the fact that the data were binary rather than Poisson.

The Zanthro Stata (StatsCorp, TX, USA) (Vidmar et al. 2004) module was used to generate HAZ scores (stunting), WAZ scores (underweight), and WHZ scores (wasting), based on 2006 child growth standards (WHO Multicentre Growth Reference Study Group 2006).

Known confounders of relationships were included in multivariate models regardless of significance in our univariate analysis. Potential confounders were included in the multivariate model after investigation by univariate analysis showed plausible confounding associations with p-values ≤0.1. A P-value of ≤0.05 was considered significant in all final analyses. Complete case analysis was used, so differences in the total numbers of children or households between tables are due to missing outcomes for some study components.

Research ethics

Free and informed consent of the participants or their legal representatives was obtained, and the study protocol was approved by the Human Ethics Committee (Health), University of Otago, New Zealand, Number 14/224, January 2015. We also obtained consent from the Ministry of Health, Vanuatu.

Demographics and WASH

Participation in the study components is summarized in Figure 2. Of the 233 households in the study population with children under 5 years of age, 220 (94.4%) were eligible and agreed to participate. The 220 households included 320 children under 5 years of age. Of the 208 household questionnaires completed, the mother was the main caregiver respondent and the median age (range) was 28 (16–69) years. Half of the caregivers reported that they had not received any schooling (Table 1).

Table 1

Participant, caregiver, and WASH characteristics by improved water status and household drinking water E. coli MPN in Tanna, 2015

Improved status
Household water E. coli MPN
Improved
Unimproved
≤10/100 mL
>10/100 mL
n(%)n(%)P-valuean(%)n(%)P-valuea
Children (N = 101) (N = 213)  (N = 37) (N = 253)  
Under 2 years of age 50 (49.5) 97 (45.5)  17 (46.0) 118 (46.6)  
Age in months: Medium (range) 26.7 (0.3–59.1) 26.3 (0.1–59.9) 0.894 27.7 (0.3–53.6) 25.4 (0.1–59.9) 0.749 
Gender, female 48 (47.5) 110 (51.6) 0.528 20 (54.1) 117 (47.6) 0.367 
Diarrheac, yes 14/95 (14.7) 21 (10.1) 0.239 (18.9) 26/235 (11.1) 0.187 
Child MDDb (ages 6–23 months only) 16/30 (53.3) 43/65 (66.2) 0.882 5/12 (41.7) 48/73 (65.8) 0.447 
Primary caregiver (N = 62) (N = 146)  (N = 22) (N = 167)  
Mother was respondent 59 (95.2) 136 (93.2)  22 (100) 157 (94.0)  
Education           
Any primary or above 47 (75.8) 54 (37.0) 0.003 18 (81.8) 74 (44.3) 0.001 
Secondary 13 (21.0) 16 (11.0) 0.048 (36.4) 19 (11.4) 0.020 
Height in cm: Median (range) 158.9 (147.6–186.7) 155.7 (146.2–167.7) 0.608 156.2 (147.6–186.7) 157.0 (147.4–170.9) 0.291 
Age in years: Median (range) 27 (18–49) 28 (16–69) 0.318 25.5 (16–42) 28 (17–69) 0.251 
WASH indicatorsd           
Self reported from household questionnaire (N = 62) (N = 146)  (N = 22) (N = 167)  
Improved water 62 (100.0) (0.0)  13 (59.1) 44 (26.4) 0.004 
Pubic standpipe 46 (74.2)    (27.3) 35 (21.0) 0.479 
Rainwater catchment 16 (25.8)    (31.8) (5.4) 0.011 
Adequate water treatment self-report 20 (32.3) 43 (29.5) 0.693 11 (50.0) 48 (28.9) 0.021 
Observed (N = 61) (N = 135)  (N = 22) (N = 174)  
Water storage adequate (9.8) (3.8) 0.048 (4.5) (5.2) 0.861 
Latrine improved (total) 57 (93.4) 133 (98.5) 0.081 21 (95.5) 157/162 (96.9) 0.540 
 Basic service (not shared34 (55.7) 82 (60.7) 0.078 11 (50.0) 98/162 (60.5) 0.522 
 Limited service (shared23 (37.7) 51 (37.8)  10 (45.5) 59/162 (36.4)  
Operational handwashing facility 31 (50.8) 72 (53.3) 0.787 (40.9) 89 (51.2) 0.462 
Improved status
Household water E. coli MPN
Improved
Unimproved
≤10/100 mL
>10/100 mL
n(%)n(%)P-valuean(%)n(%)P-valuea
Children (N = 101) (N = 213)  (N = 37) (N = 253)  
Under 2 years of age 50 (49.5) 97 (45.5)  17 (46.0) 118 (46.6)  
Age in months: Medium (range) 26.7 (0.3–59.1) 26.3 (0.1–59.9) 0.894 27.7 (0.3–53.6) 25.4 (0.1–59.9) 0.749 
Gender, female 48 (47.5) 110 (51.6) 0.528 20 (54.1) 117 (47.6) 0.367 
Diarrheac, yes 14/95 (14.7) 21 (10.1) 0.239 (18.9) 26/235 (11.1) 0.187 
Child MDDb (ages 6–23 months only) 16/30 (53.3) 43/65 (66.2) 0.882 5/12 (41.7) 48/73 (65.8) 0.447 
Primary caregiver (N = 62) (N = 146)  (N = 22) (N = 167)  
Mother was respondent 59 (95.2) 136 (93.2)  22 (100) 157 (94.0)  
Education           
Any primary or above 47 (75.8) 54 (37.0) 0.003 18 (81.8) 74 (44.3) 0.001 
Secondary 13 (21.0) 16 (11.0) 0.048 (36.4) 19 (11.4) 0.020 
Height in cm: Median (range) 158.9 (147.6–186.7) 155.7 (146.2–167.7) 0.608 156.2 (147.6–186.7) 157.0 (147.4–170.9) 0.291 
Age in years: Median (range) 27 (18–49) 28 (16–69) 0.318 25.5 (16–42) 28 (17–69) 0.251 
WASH indicatorsd           
Self reported from household questionnaire (N = 62) (N = 146)  (N = 22) (N = 167)  
Improved water 62 (100.0) (0.0)  13 (59.1) 44 (26.4) 0.004 
Pubic standpipe 46 (74.2)    (27.3) 35 (21.0) 0.479 
Rainwater catchment 16 (25.8)    (31.8) (5.4) 0.011 
Adequate water treatment self-report 20 (32.3) 43 (29.5) 0.693 11 (50.0) 48 (28.9) 0.021 
Observed (N = 61) (N = 135)  (N = 22) (N = 174)  
Water storage adequate (9.8) (3.8) 0.048 (4.5) (5.2) 0.861 
Latrine improved (total) 57 (93.4) 133 (98.5) 0.081 21 (95.5) 157/162 (96.9) 0.540 
 Basic service (not shared34 (55.7) 82 (60.7) 0.078 11 (50.0) 98/162 (60.5) 0.522 
 Limited service (shared23 (37.7) 51 (37.8)  10 (45.5) 59/162 (36.4)  
Operational handwashing facility 31 (50.8) 72 (53.3) 0.787 (40.9) 89 (51.2) 0.462 

Differing Ns for differing survey components are specified (child questionnaire, household questionnaire, observations). n (%) may not add to final N because of missing values. Missing <5% unless denominator specified.

aP-values adjusted for clustering by household and village (child inicators) or village (household indicators).

bIYCF MDD – World Health Organization Infant and Young Child Feeding Minimum Dietary Diversity.

cDiarrhea in last 7 days.

dWASH – water, sanitation, and hygiene.

Of the 208 households participating in the household survey, 63 (30.3%) self-reported adequate treatment of drinking water, with boiling water being the only adequate treatment method reported (Table 1). Of the 208 households participating in the WASH observation survey, 11 (5.3%) had adequate water storage; 106 (51.0%) had operational handwashing facilities, consisting of tippy tap variations using plastic containers or bamboo shoots to hold water; and 202 (97.1%) had improved latrines in the form of VIP latrines or improved simple pit latrines. When information from both observations and questionnaires were combined, 116 (59.2%) of 196 met the SDG sanitation definition of ‘basic service’, based on shared status.

Drinking water microbiological quality and relationship to water indicators

Of the 201 households that had stored water tested, 10 (5.0%) had water that met WHO fecal contamination standards for drinking water, <1 E. coli/100 mL (Figure 2). A further 12 (6.0%) households had drinking water with 1–10 E. coli/100 mL; 34 (16.9%) with >10–100 E. coli/100 mL; and 145 (72.1%) with >100 E. coli/100 mL.

Of 208 households, 62 (29.8%) reported collecting drinking water from ‘improved water’ sources, which consisted of 46 (74.2%) gravity fed systems to public standpipes and 16 (25.8%) rainwater catchments (Table 1). When we classified household drinking water according to SDG indicators (Box 1), 37 (17.8%) of the 208 households met the requirements for ‘basic drinking water’, and 5 (2.4%) met the requirements for ‘safely managed drinking water’.

Table 2 compares MDG ‘improved water’ with microbiological quality. Of the 57 households with improved water in this analysis, 50 (87.7%) had >1 E. coli/100 mL. In our univariate analysis, the odds of having drinking water <1 E. coli/100 mL was higher in households with ‘improved water’ (unadjusted OR: 6.02, 95% CI: 1.25, 29.10) than households with unimproved water. However, this was no longer significant when other WASH factors were adjusted for, including water treatment. The odds of having water with >100 E. coli/100 mL was lower in households with ‘improved water’, and this association remained significant when adjusted for WASH factors (adjusted OR: 0.22, 95% CI: 0.08, 0.65).

Table 2

Household water MPN and improved water status in Tanna, 2015

Household improved drinking water source
Improved
Unimproved
Unadjusted OR(95% CI)P-valueAdjustedb OR(95% CI)P-value
n(%)n(%)
Household water MPNa (N = 57) (N = 132)       
<1 E. coli/100 mL (12.3) (2.3) 6.02 (1.25, 29.10) 0.025 2.80 (0.55, 16.08) 0.205 
1–10 E. coli/100 mL (10.5) (4.6)       
>10–100 E. coli/100 mL 18 (31.6) 14 (10.6)       
>100 E. coli/100 mL 26 (45.6) 109 (82.6) 0.18 (0.06, 0.49) 0.001 0.22 (0.08, 0.65) 0.006 
Household improved drinking water source
Improved
Unimproved
Unadjusted OR(95% CI)P-valueAdjustedb OR(95% CI)P-value
n(%)n(%)
Household water MPNa (N = 57) (N = 132)       
<1 E. coli/100 mL (12.3) (2.3) 6.02 (1.25, 29.10) 0.025 2.80 (0.55, 16.08) 0.205 
1–10 E. coli/100 mL (10.5) (4.6)       
>10–100 E. coli/100 mL 18 (31.6) 14 (10.6)       
>100 E. coli/100 mL 26 (45.6) 109 (82.6) 0.18 (0.06, 0.49) 0.001 0.22 (0.08, 0.65) 0.006 

aAquagenx CBT categories, based on WHO Guidelines for Drinking Water Quality, 4th ed.

bAdjusted for water treatment, water storage, and having an improved latrine.

Variables associated with household drinking water ≤10 E. coli/100 included the caregiver having any primary or above education (P= 0.001), secondary education (P= 0.02), and self-reported adequate water treatment (P = 0.021) (Table 1). Household drinking water ≤10 E. coli/100 was also associated with having rainwater as the main drinking water source (P= 0.011) but not water from public standpipes (P = 0.479) (Table 1).

Variables associated with households using ‘improved water’ were having primary or above education (P= 0.003), secondary education (P= 0.048), and adequate water storage (P= 0.048) (Table 1).

Child anthropometry indicators

A total of 312 children under 5 years of age were available to have anthropometry measurements taken. The numbers in the analyses vary due to missing information (height/length, weight, or age) or data excluded for suspected measurement error. Of the 297 children in the height-for-age (HAZ) analysis, 145 (48.8%, 95% CI: 42.8, 54.8) were stunted (<−2z), including 52 (17.5%) severely stunted (<−3z). Of the 309 children in the weight-for-age (WAZ) analysis, 59 (19.1%, 95% CI: 14.4, 23.8) were underweight (<−2z), including 19 (6.2%) severely underweight (<−3z). Of the 303 children in the weight-for-height (WHZ) analysis, 12 (4.0%, 95% CI: 1.8, 6.1) were wasted (<−2z), including 2 (0.7%) severely wasted (<−3z).

Associations between water indicators and undernutrition

Table 3 describes the proportion of child nutrition outcomes according to household drinking water microbiological quality (≤10 E. coli/100 mL or >10 E. coli/100 mL) and household drinking water status (‘improved water’ or ‘unimproved’). Table 4 shows our investigation of associations between drinking water microbiological quality and child undernutrition using univariate and multivariate analysis. We found no association between having drinking water with ≤10 E. coli/100 mL and HAZ, WAZ, WHZ, stunting, or underweight in our analyses. No children in households with drinking water ≤10 E. coli/100 mL were wasted, therefore this could not be investigated in our model.

Table 3

Undernutrition among children under 5 years of age by improved water status and household drinking water E. coli MPN in Tanna, 2015

Improved status
Household water E. coli MPN
Improved
Unimproved
≤10/100 mL
>10/100 mL
n(%)n(%)n(%)n(%)
Stunting (N = 96) (N = 196) (N = 34) (N = 232) 
Normal (≥−2z) 65 (67.7) 83 (42.4) 20 (58.8) 116 (50.0) 
Stunted (<−2z) 31 (32.3) 113 (57.7) 14 (41.2) 72 (31.0) 
Severely stunted (<−3z) (8.3) 44 (22.5) (14.7) 44 (19.0) 
HAZa score (Median (range)) −1.62 (−4.47, 4.50) −2.19 (−4.96, 1.79) −1.73 (−4.23, 4.50) −2.02 (−4.96, 2.46) 
Underweight (N = 100) (N = 203) (N = 37) (N = 239) 
Normal (≥−2z) 90 (90.0) 154 (75.9) 31 (83.8) 191 (79.9) 
Underweight (<−2z) 10 (10.0) 49 (24.1) (16.2) 30 (12.6) 
Severely Underweight (<−3z) (0.0) 19 (9.4) (0.0) 18 (7.5) 
WAZa score (Median (range)) −0.71 (−2.95, 2.26) −1.09 (−4.87, 3.51) −1.20 (−2.82, 3.51) 0.95 (−4.86, 2.82) 
Wasting (N = 96) (N = 202) (N = 35) (N = 235) 
Normal (≥−2z) 95 (99.0) 191 (94.6) 35 (100.0) 224 (95.3) 
Wasted (<−2z) (1.0) 11 (5.5) (0.0) (3.8) 
Severely wasted (<−3z) (0.0) (1.0) (0.0) (0.85) 
WHZa score (Median (range)) 0.25 (−2.32, 2.26) 0.03 (−4.14, 2.71) −0.19 (−1.88, 1.41) 0.15 (−4.14, 2.71) 
Improved status
Household water E. coli MPN
Improved
Unimproved
≤10/100 mL
>10/100 mL
n(%)n(%)n(%)n(%)
Stunting (N = 96) (N = 196) (N = 34) (N = 232) 
Normal (≥−2z) 65 (67.7) 83 (42.4) 20 (58.8) 116 (50.0) 
Stunted (<−2z) 31 (32.3) 113 (57.7) 14 (41.2) 72 (31.0) 
Severely stunted (<−3z) (8.3) 44 (22.5) (14.7) 44 (19.0) 
HAZa score (Median (range)) −1.62 (−4.47, 4.50) −2.19 (−4.96, 1.79) −1.73 (−4.23, 4.50) −2.02 (−4.96, 2.46) 
Underweight (N = 100) (N = 203) (N = 37) (N = 239) 
Normal (≥−2z) 90 (90.0) 154 (75.9) 31 (83.8) 191 (79.9) 
Underweight (<−2z) 10 (10.0) 49 (24.1) (16.2) 30 (12.6) 
Severely Underweight (<−3z) (0.0) 19 (9.4) (0.0) 18 (7.5) 
WAZa score (Median (range)) −0.71 (−2.95, 2.26) −1.09 (−4.87, 3.51) −1.20 (−2.82, 3.51) 0.95 (−4.86, 2.82) 
Wasting (N = 96) (N = 202) (N = 35) (N = 235) 
Normal (≥−2z) 95 (99.0) 191 (94.6) 35 (100.0) 224 (95.3) 
Wasted (<−2z) (1.0) 11 (5.5) (0.0) (3.8) 
Severely wasted (<−3z) (0.0) (1.0) (0.0) (0.85) 
WHZa score (Median (range)) 0.25 (−2.32, 2.26) 0.03 (−4.14, 2.71) −0.19 (−1.88, 1.41) 0.15 (−4.14, 2.71) 

aWHZ – weight-for-height Z-score, HAZ – height-for-age Z-score, and WAZ – weight-for-age Z-score.

Table 4

Unadjusted and adjusted estimations of associations between undernutrition and water with an E. coli MPN ≤10/100 mL in Tanna, 2015

Household water MPN ≤10/100 mLa
Unadjusted MDb(95% CI)P-valueAdjustedc MD(95% CI)P-value
HAZ-score 0.20 (−0.32, 0.72) 0.437 0.07 (−0.42, 0.56) 0.770 
WAZ-score 0.10 (−0.27, 0.47) 0.586 −0.09 (−0.45, 0.27) 0.625 
WHZ-score −0.25 (−0.02, −0.47) 0.033 −0.33 (−0.67, −0.01) 0.057 
Unadjusted PRb(95% CI)P-valueAdjustedc PR(95% CI)P-value
Stunting (<−2z) 1.02 (0.53, 1.95) 0.959 0.96 (0.49, 1.90) 0.904 
Severe stunting (<−3z) 0.92 (0.40, 2.14) 0.849 1.04 (0.48, 2.24) 0.929 
Underweight (<−2z) 1.13 (0.63, 2.02) 0.682 1.19 (0.58, 2.45) 0.642 
Household water MPN ≤10/100 mLa
Unadjusted MDb(95% CI)P-valueAdjustedc MD(95% CI)P-value
HAZ-score 0.20 (−0.32, 0.72) 0.437 0.07 (−0.42, 0.56) 0.770 
WAZ-score 0.10 (−0.27, 0.47) 0.586 −0.09 (−0.45, 0.27) 0.625 
WHZ-score −0.25 (−0.02, −0.47) 0.033 −0.33 (−0.67, −0.01) 0.057 
Unadjusted PRb(95% CI)P-valueAdjustedc PR(95% CI)P-value
Stunting (<−2z) 1.02 (0.53, 1.95) 0.959 0.96 (0.49, 1.90) 0.904 
Severe stunting (<−3z) 0.92 (0.40, 2.14) 0.849 1.04 (0.48, 2.24) 0.929 
Underweight (<−2z) 1.13 (0.63, 2.02) 0.682 1.19 (0.58, 2.45) 0.642 

aMicrobiological groupings are amalgamated Aquagenx CBT categories: ≤10 E. coli/100 mL (safe and intermediate risk) compared to >10/100 mL (high and very high risk).

bMD – mean difference and PR – prevalence ratio.

cAssociations adjusted for child age, child gender, maternal height, and caregiver education.

Table 5 shows our investigation of associations between having ‘improved water’ and child undernutrition. In multivariate analysis, the adjusted MD in child HAZ and WAZ was 0.43 (95% CI: 0.20, 0.67) and 0.36 (95% CI: 0.12, 0.59) higher, respectively, in households using ‘improved water’ compared to ‘unimproved water’. We found no association between WHZ and households with ‘improved water’.

Table 5

Unadjusted and adjusted estimations of associations between undernutrition and improved water in Tanna, 2015

Unadjusted MDa(95% CI)P-valueAdjustedb MD(95% CI)P-value
HAZ-score 0.60 (0.27, 0.94) 0.001 0.43 (0.20, 0.67) 0.001 
WAZ-score 0.52 (0.20, 0.84) 0.002 0.36 (0.12, 0.59) 0.004 
WHZ-score 0.23 (0.02, 0.45) 0.036 0.17 (−0.17, 0.50) 0.313 
Unadjusted PRa(95% CI)P-valueAdjustedb PR(95% CI)P-value
Stunting (< −2z) 0.55 (0.43, 0.70) <0.001 0.61 (0.46, 0.80) <0.001 
Severe stunting (< −3z) 0.37 (0.19, 0.72) 0.003 0.41 (0.19, 0.90) 0.025 
Underweight (< −2z) 0.43 (0.24, 0.78) 0.006 0.46 (0.29, 0.73) 0.001 
Wasting (< −2z) 0.17 (0.02, 1.58) 0.119 0.20 (0.03, 1.51) 0.119 
Unadjusted MDa(95% CI)P-valueAdjustedb MD(95% CI)P-value
HAZ-score 0.60 (0.27, 0.94) 0.001 0.43 (0.20, 0.67) 0.001 
WAZ-score 0.52 (0.20, 0.84) 0.002 0.36 (0.12, 0.59) 0.004 
WHZ-score 0.23 (0.02, 0.45) 0.036 0.17 (−0.17, 0.50) 0.313 
Unadjusted PRa(95% CI)P-valueAdjustedb PR(95% CI)P-value
Stunting (< −2z) 0.55 (0.43, 0.70) <0.001 0.61 (0.46, 0.80) <0.001 
Severe stunting (< −3z) 0.37 (0.19, 0.72) 0.003 0.41 (0.19, 0.90) 0.025 
Underweight (< −2z) 0.43 (0.24, 0.78) 0.006 0.46 (0.29, 0.73) 0.001 
Wasting (< −2z) 0.17 (0.02, 1.58) 0.119 0.20 (0.03, 1.51) 0.119 

aMD – mean difference and PR – prevalence ratio.

bAssociations adjusted for: child age, child gender, maternal height, caregiver education, adequate water storage, and having an improved latrine (VIP latrine or improved pit latrine).

In multivariable analysis, the adjusted prevalence of stunted children was 39% lower in households using ‘improved water’ compared with those using ‘unimproved water’ (PR: 0.61, 95% CI 0.46, 0.80), and the prevalence of severe stunting was 59% lower (PR: 0.41, 95% CI 0.19, 0.90). The prevalence of underweight children was 54% lower in households using ‘improved water’ than households using ‘unimproved water’ (PR: 0.46, 95% CI 0.29, 0.73) in multivariate analysis. There was no statistical difference in the prevalence of wasted children among households with ‘improved water’ compared to ‘unimproved water’ although no children from households with ‘improved water’ were severely wasted or severely underweight.

We found that the majority of stored water in southwest Tanna, Vanuatu, at this single point in time had microbiological contamination, regardless of MDG ‘improved water’ status. Concurrently, according to WHO classifications, the population had a very high prevalence of stunted children and a medium prevalence of underweight children (WHO/UNICEF 2019). Compared with households with MDG ‘unimproved water’, those households with ‘improved water’ were likely to have fewer children being stunted and severely stunted, and underweight. However, there was no association between having lower drinking water contamination (≤10 E. coli/100 mL) at the household and undernutrition.

Although we found that ‘improved water’ was less likely to have >100 E. coli/100 mL compared with ‘unimproved water’, the majority of ‘improved water’ had microbiological contamination at the time of testing. This finding is consistent with other studies (Clasen 2012; Shaheed et al. 2014; Heitzinger et al. 2015). Despite ‘improved water’ not equating to having water <1 E. coli/100 mL, we cannot rule out that ‘improved water’ may have lower microbiological contamination over the long term rather than a single time point tested here, as, by the nature of their design and construction, they offer some protection from external contamination (WHO/UNICEF 2017). Our finding of an association between ‘improved water’ and fewer poor nutrition outcomes supports this hypothesis, particularly for stunting which occurs over long exposure periods (Prendergast & Humphrey 2014).

Microbiologically testing drinking water at one point in time may not reflect long-term exposure to contaminated drinking water and may not provide an accurate representation of long-term water quality because of fluctuation in response to environmental conditions (Csuros 1994; WHO 2011). This is supported by Geen et al. (2011), who reported that fecal contamination varied by up to 50% from season to season in Bangladesh. In Tanna, variation in microbiological water quality is likely driven by heavy rainfall and flooding in the wet season, when our study took place, and turbidity from volcanic ash (Nath et al. 2006).

To our knowledge, no previous studies have directly compared MDG ‘improved water’ and microbiological quality to undernutrition outcomes in the same population. Previous studies have reported associations between drinking water and child undernutrition (Huttly et al. 1990; Esrey 1996; Merchant et al. 2003; Ellis & Schoenberger 2017; Lauer et al. 2018). However, a 2018 systematic review concluded that the evidence of a casual association was inconclusive because the simplified water indicators did not take into account household water storage and handling, and a lack of studies testing the actual water quality (Vilcins et al. 2018). A recent study also using the CBT method found that microbiologically unsafe drinking water at one point in time was related to both EED and stunting in Uganda (Lauer et al. 2018), supporting the proposed mechanism between water, EED, and growth (Ngure et al. 2014). However, recently published large RCTs have emphasized that WASH interventions, as they are currently delivered, may be limited in their impact on child growth (Stewart et al. 2018; Tofail et al. 2018; Prendergast et al. 2019).

Using CBTs to measure water contamination was a strength of our study as it has high sensitivity and specificity when compared with membrane filtration (Stauber et al. 2013), did not require laboratory staff or equipment, and was relatively inexpensive. The increasing accessibility of water microbiological quality testing kits that can be used in field situations has the potential to greatly enhance community water development projects. Regular water testing by communities would allow for community awareness of water risk and ownership of interventions to mitigate risk where necessary as well as making SDG monitoring more reliable. However, WHO and UNICEF have recognized that attaining water microbiological quality results that are truly representative is a challenge for SDG monitoring (WHO/UNICEF 2012).

There are a number of limitations in our study. Although our population of south-west Tanna represents a distinct population, the results of recent RCT studies encourage us to be cautious with our findings, as they may be influenced in part by the cross-sectional nature of our study. In particular, causality cannot be inferred because poor nutritional status may have preceded water source use, as it was unknown exactly how long households had been using the water sources. In addition, unknown confounders were not able to be controlled for in our multivariate analyses. For example, we lacked sufficient information to generate a wealth index, so could not examine this potential confounder between wealth, ‘improved water’, and health. As all households in our study were homogenously in the lowest wealth quintile of Vanuatu and geographically isolated from main centers (World Vision 2013), we do not believe this affected our conclusions. Our study may also have lacked power to detect relevant associations. Low numbers of household with drinking water <1 E. coli/100 mL meant we could not investigate this category for associations with child undernutrition. At the time of the study, food security was reported to be achieved via self-subsistence agriculture (World Vision 2013). Although a variety of fruits, vegetables, and fish are available all year round in Vanuatu, we cannot be certain that children's dietary requirements were always upheld, and thus, poor diets prior to survey may be influencing our results. Self-report is also known to be unreliable in surveys, due to social desirability bias. To mitigate this, whenever possible in our study, WASH facilities were observed. Furthermore, our results may not be generalizable to other populations where water sources and environmental conditions differ, or to food insecure populations.

Our study sought to investigate the relationship between MDG ‘improved water’, microbiological water quality, and child undernutrition in a rural Vanuatu population underrepresented in the literature. The finding of extensive water contamination, regardless of improved status, and a high prevalence of child undernutrition emphasizes the need for continued community-led projects to address these problems. We found that ‘improved water’ was associated with a lower prevalence of undernutrition, whereas having drinking water with ≤10 E. coli/100 mL, measured at one point in time, was not. It is possible that long-term variations in water microbiological quality and unmeasured conditions beyond water quality drove these associations. Therefore, it would be beneficial for future studies to tease out the true nature of these relationships by looking at microbiological water quality and undernutrition over time in a population. Future studies should also aim to evaluate the SDG ‘safely managed drinking water’ indicator that combines the ‘improved water’ outcome with water microbiological quality and accessibility, as very few households in our study met these requirements. Establishing water indicators that can reflect health consequences is important for identifying whether poor drinking water is a risk factor for child morbidity in a population, and for monitoring interventions more efficiently.

We would like to express gratitude to the families in southwest Tanna for their participation in this study. We would also like to thank the World Vision Tanna team members who worked hard preparing and carrying out data collection. Team members included Norolyn Nimbtik, Anita Napua, Rosilyn Mojuju, Harold Nakao, David Nalau, Rosinel Harry, Maurice Manu, and Joseph Kiet. Additionally, thanks goes to Dr Raul Schneider (World Vision Regional Health Advisor) and Elise Bryce-Johnson (former World Vision New Zealand, Grants Programme Manager, Papua New Guinea and Vanuatu) for assisting in the planning of the data collection in Tanna.

Data for this study were from the second year evaluation of the World Vision Vanuatu WASH Project, which is funded by World Vision New Zealand. The evaluation was funded in part by World Vision New Zealand (donated CBTs, flights, and accommodation of researchers) and in part by the New Zealand Aid Programme through a Postgraduate Field Research Scholarship. The corresponding author received support from the University of Otago through the Department for Preventive and Social Medicine and an Otago University Masters Scholarship. World Vision Vanuatu team members were involved in design and data collection of the WASH evaluation but World Vision, both New Zealand and Vanuatu, were not involved in data analysis or interpretation, decision to publish, or preparation of the manuscript. The New Zealand Aid Programme had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

The findings and conclusions in this report are that of the authors alone, and do not necessarily represent the position of the University of Otago, World Vision, or the New Zealand Aid Programme.

Aquagenx
2013
Compartment Bag Test Instructions
.
Clasen
T.
2012
Millennium Development Goals water target claim exaggerates achievement
.
Tropical Medicine & International Health
17
(
10
),
1178
1180
.
doi:10.1111/j.1365-3156.2012.03052.x
.
Cogill
B.
2013
Anthropometric Indicators Measurement Guide
.
Food and Nutrition Technical Assistance (FANTA) Project, USAID
,
Washington, DC
.
Crane
R. J.
Jones
K. D. J.
Berkley
J. A.
2015
Environmental enteric dysfunction: an overview
.
Food and Nutrition Bulletin
36
(
10
),
76
87
.
doi: 10.1177/15648265150361S113
.
Csuros
M.
1994
Environmental Sampling and Analysis for Technicians
.
CRC Press
,
USA
.
Dangour
A. D.
Watson
L.
Cumming
O.
Boisson
S.
Che
Y.
Velleman
Y.
Cavill
S.
Allen
E.
Uauy
R.
2013
Interventions to improve water quality and supply, sanitation and hygiene practices, and their effects on the nutritional status of children
.
The Cochrane Database of Systematic Reviews
8
,
CD009382
CD009382
.
Esrey
S. A.
1996
Water, waste, and well-being: a multicountry study
.
American Journal of Epidemiology
143
(
6
),
603
623
.
https://doi.org/10.1093/oxfordjournals.aje.a008791
.
Foster
T.
Willetts
J.
Kotra
K. K.
2019
Faecal contamination of groundwater in rural Vanuatu: prevalence and predictors
.
Journal of Water & Health
.
https://doi.org/10.2166/wh.2019.016
.
Geen
A. v.
Ahmed
K. M.
Akita
Y.
Alam
M. J.
Culligan
P. J.
Emch
M.
Escamilla
V.
Feighery
J.
Ferguson
A. S.
Knappett
P.
Layton
A. C.
Mailloux
B. J.
McKay
L. D.
Mey
J. L.
Serre
M. L.
Streatfield
P. K.
Wu
J.
Yunus
M.
2011
Fecal contamination of shallow tubewells in Bangladesh inversely related to arsenic
.
Environmental Science & Technology
45
(
4
),
1199
1205
.
doi:10.1021/es103192b
.
Guerrant
R. L.
DeBoer
M. D.
Moore
S. R.
Scharf
R. J.
Lima
A. A. M.
2013
The impoverished gut – a triple burden of diarrhoea, stunting and chronic disease
.
Nature Reviews Gastroenterology & Hepatology
10
(
4
),
220
229
.
doi:10.1038/nrgastro.2012.239
.
Heitzinger
K.
Rocha
C.
Quick
R.
Montano
S.
Tilley
D.
Mock
C.
Carrasco A
C. R.
S
H.
2015
‘Improved’ but not necessarily safe: an assessment of fecal contamination of household drinking water in rural Peru
.
American Journal of Tropical Medicine and Hygiene
93
(
3
),
501
508
.
doi:10.4269/ajtmh.14-0802
.
Huttly
S. R. A.
Blum
D.
Kirkwood
B. R.
Emeh
R. N.
Okeke
N.
Ajala
M.
Smith
G. S.
Carson
D. C.
Dosunmu-Ogunbi
O.
Feachem
R. G.
1990
The Imo State (Nigeria) Drinking Water Supply and Sanitation Project, 2. Impact on dracunculiasis, diarrhoea and nutritional status
.
Transactions of the Royal Society of Tropical Medicine and Hygiene
84
(
2
),
316
321
.
https://doi.org/10.1016/0035-9203(90)90300-4
.
Keusch
G. T.
Denno
D. M.
Black
R. E.
Duggan
C.
Guerrant
R. L.
Lavery
J. V.
Nataro
J. P.
Rosenberg
I. H.
Ryan
E. T.
Tarr
P. I.
Ward
H.
Bhutta
Z. A.
2014
Environmental enteric dysfunction: pathogenesis, diagnosis, and clinical consequences
.
Clinical Infectious Diseases
59
(
4
),
207
212
.
Lauer
J. M.
Duggan
C. P.
Ausman
L. M.
Griffiths
J. K.
Webb
P.
Bashaasha
B.
Agaba
E.
Turyashemererwa
F. M.
Ghosh
S.
2018
Unsafe drinking water is associated with environmental enteric dysfunction and poor growth outcomes in young children in rural southwestern Uganda
.
The American Journal of Tropical Medicine and Hygiene
99
(
6
),
1606
1612
.
doi:10.4269/ajtmh.18-0143
.
Merchant
A. T.
Jones
C.
Kiure
A.
Kupka
R.
Fitzmaurice
G.
Herrera
M. G.
Fawzi
W. W.
2003
Water and sanitation associated with improved child growth
.
European Journal of Clinical Nutrition
57
(
12
),
1562
1568
.
doi:10.1038/sj.ejcn.1601725
.
Nath
D.
Mudaliar
M.
Ioan
C.
2006
Water Safety Plan: Water Supply Description Assessment Vanuatu
.
Pacific Water Safety Plans Programme
,
Vanuatu
.
Ngure
F. M.
Reid
B. M.
Humphrey
J. H.
Mduduzi
M. N.
Pelto
G.
Stoltzfus
R. J.
2014
Water, sanitation, and hygiene (WASH), environmental enteropathy, nutrition, and early child development: making the links
.
Annals of the New York Academy of Sciences
1308
,
118
128
.
doi:10.1111/nyas.12330
.
Prendergast
A. J.
Humphrey
J. H.
2014
The stunting syndrome in developing countries
.
Paediatrics and International Child Health
34
(
4
),
250
265
.
doi:10.1179/2046905514Y.0000000158
.
Prendergast
A. J.
Chasekwa
B.
Evans
C.
Mutasa
K.
Mbuya
M. N. N.
Stoltzfus
R. J.
Smith
L. E.
Majo
F. D.
Tavengwa
N. V.
Mutasa
B.
Chasokela
G. T. M. C. M.
Moulton
A. C. L. H.
Ntozini
R.
Humphrey
J. H.
2019
Independent and combined effects of improved water, sanitation, and hygiene, and improved complementary feeding, on stunting and anaemia among HIV-exposed children in rural Zimbabwe: a cluster-randomised controlled trial
.
The Lancet Child and Adolescent Health
3
(
2
),
77
90
.
https://doi.org/10.1016/S2352-4642(18)30340-7
.
Ps-Eau
2016
WASH Services in the Sustainable Development Goals
.
Working Document, Version – October 2016. Available from: http://www.pseau.org/en/agenda-2030 (retrieved 10 May 2017
).
Shaheed
A.
Orgill
J.
Montgomery
M. A.
Jeuland
M. A.
Brown
J.
2014
Why ‘improved’ water sources are not always safe
.
Bulletin of the World Health Organization
92
,
283
289
.
Stauber
C.
Miller
C.
Cantrell
B.
Kroell
K.
2013
Evaluation of the Compartment Bag Test for the Detection of Escherichia coli in Drinking Water
.
Georgia State University, Atlanta
.
Stewart
C. P.
Kariger
P.
Fernald
L.
Pickering
A. J.
Arnold
C. D.
Arnold
B. F.
Hubbard
A. E.
Dentz
H. N.
Lin
A.
Meerkerk
T. J.
Milner
E.
Swarthout
J.
Colford
J. M.
Null
C.
2018
Effects of water quality, sanitation, handwashing, and nutritional interventions on child development in rural Kenya (WASH Benefits Kenya): a cluster-randomised controlled trial
.
The Lancet Child and Adolescent Health
2
(
4
),
269
280
.
https://doi.org/10.1016/S2352-4642(18)30025-7
.
Tofail
F.
Fernald
L. C.
Das
K. K.
Rahman
M.
Ahmed
T.
Jannat
K. K.
Unicomb
L.
Arnold
B. F.
Ashraf
S.
Winch
P. J.
Kariger
P.
Stewart
C. P.
Colford
J. M.
Luby
S. P.
2018
Effect of water quality, sanitation, hand washing, and nutritional interventions on child development in rural Bangladesh (WASH Benefits Bangladesh): a cluster-randomised controlled trial
.
The Lancet Child and Adolescent Health
2
(
4
),
255
268
.
https://doi.org/10.1016/S2352-4642(18)30031-2
.
United Nations Water
.
2016
Integrated Monitoring Guide for SDG 6 Targets and Global Indicators
.
United Nations Water
,
Geneva
,
Switzerland
.
Vanuatu Ministry of Health
2014
Vanuatu Demographic and Health Survey 2013
.
Vanuatu National Statistics Office and the Secretariat of the Pacific Community
,
Vanuatu
.
Vilcins
D.
Sly
P. D.
Jagals
P.
2018
Environmental risk factors associated with child stunting: a systematic review of the literature
.
Annals of Global Health
84
(
4
),
551
562
.
http://doi.org/10.29024/aogh.2361
.
WHO
2008
Indicators for assessing infant and young child feeding practices: Part 1: definitions
. In
Conclusions of a Consensus Meeting Held
,
November 6–8, 2007
Washington, D.C
.
World Health Organization
,
Geneva
,
Switzerland
.
WHO
2011
Guidelines for Drinking-Water Quality
, 4th edn.
World Health Organization
,
Geneva
,
Switzerland
.
WHO
2013
Household Water Treatment and Safe Storage (HWTS): Manual for the Participant
.
World Health Organization
,
Geneva
,
Switzerland
.
WHO
2016
World Health Statistics 2016: Monitoring Health for the SDGs
.
Chapter 6: SDG health and health-related targets
,
World Health Organization
,
Geneva
,
Switzerland
.
WHO Multicentre Growth Reference Study Group
2006
WHO Child Growth Standards: Length/Height-for-Age, Weight-for-Age, Weight-for-Height and Body Mass Index-for-Age: Methods and Development
.
World Health Organization
,
Geneva
,
Switzerland
.
WHO/UNICEF
2012
Report of a Technical Consultation on the Measurability of Global WASH Indicators for Post-2015 Monitoring
.
World Health Organization and United Nations Children's Fund
,
New York
,
USA
.
WHO/UNICEF
2017
Progress on Drinking Water, Sanitation and Hygiene 2017 Update and SDG baselines. World Health Organization (WHO) and the United Nations Children's Fund (UNICEF), Geneva. Licence: CC BY-NC-SA 3.0 IGO.
WHO/UNICEF
2019
Levels and Trends in Child Malnutrition
.
Available from: http://www.who.int/nutgrowthdb/en/ (retrieved 20 February 2016
).
World Vision
2013
Tanna Integrated WASH Project Baseline Report
.
World Vision Vanuatu
,
Port Villa
,
Vanuatu
.