WASH and nutrition: trials and tribulations? 

How should we interpret the latest research findings on undernutrition and the role of water, sanitation and hygiene (WASH)? WaterAid UK’s Megan Wilson-Jones (Policy Analyst – Health and Hygiene), Om Prasad Gautam (Senior WASH Manager – Hygiene) and Kyla Smith (Research Manager) share their reflections.

Part of what we learned is that this problem of stunting is not going to be easily fixed by a little bit of attention to water, sanitation and hygiene… Modest efforts to marginally improve environments are not going to be sufficient. If we want children in the lowest-income, most resource-constrained environments to thrive, we’re going to need to make their environments radically cleaner. 

Stephen Luby, lead of WASH Benefits Bangladesh trial.

The long-awaited publication of the first findings from the WASH Benefits trials in Bangladesh and Kenya has been generating substantial debate among those working to improve nutrition and access to water, sanitation and hygiene (WASH) for the world’s most vulnerable people. The studies, which aimed to improve understanding of the impact on child growth of WASH combined with nutrition interventions, found that children who received household WASH interventions did not grow taller after two years compared with those who did not receive the interventions.

Although overall the interventions that combined WASH and nutrition had other positive health benefits, the findings on child growth have been particularly surprising; they seem to go against an otherwise growing body of evidence that suggests a strong link between poor WASH and stunting.

So what can these studies tell us? Is everything we thought we knew about poor WASH as a driver of undernutrition obsolete? Or are there particular aspects of the interventions’ design and delivery that help demystify this complex relationship, and indeed further our understanding about the type and quality of WASH interventions needed for nutrition impact? We think the latter, and you can read our full reflections on the studies here.

Here’s a summary of how we think the findings can be usefully interpreted…

1. All critical pathways of faecal-oral transmission need to be adequately blocked to achieve significant benefits.

There are multiple pathways by which a new host can ingest faecal pathogens. Although many of these are known, the relative importance of each remains largely unknown and is highly context specific. The two WASH Benefits studies did not measure the surrounding household and compound environments (with the exception of testing the water quality), and as such it was unknown whether any environmental transmission pathways were effectively interrupted. Furthermore, and other important pathways – including contamination via food-related pathways and from animal faeces – were not addressed fully. This means that the young children studied were probably still exposed to faecal pathogens through different pathways.

2. Coverage and quality of WASH interventions need to be higher.

The WASH interventions were delivered at the level of the household and compound of the target families. In Bangladesh, this meant that the sanitation intervention reached only 10% of village residents, and therefore environmental faecal contamination did not change substantially. Previous studies by Dean Spears et al (2013) highlight that stunting is not so much related to household access to a toilet as to levels of open defecation (per sq km). In other words, the health benefits of sanitation do not come from personal use of improved sanitation facilities but from improved community coverage and decreasing open defecation, so most of the benefits will occur when coverage becomes universal.

Other limitations of the intervention design and delivery include:

• The use of chlorine for water treatment, which is not effective against all microorganisms responsible for diarrhoea.
• The limited ability to conclude from the studies whether behaviour change actually happened, because the researchers only included proxy measures for handwashing, such as the presence of handwashing stations and soap, rather than measuring the actual practice of handwashing behaviours at critical times.

3. Diarrhoea is a complex outcome measure.

The authors measured diarrhoea prevalence by caregiver-reported diarrhoea during the preceding seven days. The fact that the intervention in Bangladesh did correlate with reduced diarrhoea rates but not with growth could suggest possible reporter bias, perhaps due to parents being grateful for the intervention and, out of courtesy, reporting fewer cases of diarrhoea. The fact that incidence of diarrhoea reduced only in the Bangladesh study, where promoter visits were more frequent than in Kenya, reinforces the bias concern. Self-reported diarrhoea is inherently biased and subjective in nature, open to both over- and under-reporting. Studies need better measures of impact, such as laboratory confirmation of pathogens, clinical admissions or physician diagnosis of diarrhoea.

4. More research is needed to understand the key underlying causes and pathways that lead to diarrhoea and stunting.

Significantly reducing complex outcomes such as stunting and diarrhoea may require improvements across multiple/all underlying causes, and a high level of compliance on all relevant behaviours, beyond handwashing. Although catch-up growth – in which faster growth compensates after a period of stunted development – is technically possible, it is likely that this would require a profound improvement in the health, nutrition and environment in which young children live. This would necessitate a comprehensive approach across multiple areas.

Where do we go from here?

• WASH and nutrition practitioners and researchers need to do better. Alone, the set of basic household-level WASH interventions used in both Bangladesh and Kenya were not ‘good enough’ to clean up a highly contaminated environment. There needs to be a dramatic shift in how we address poor WASH in these settings if we are to ensure children have the best chance to grow and develop to their full potential. This will require: moving towards universal coverage of WASH services in communities; greater priority to animal husbandry practices; and innovative ways of changing hygiene behaviours, moving beyond a cognition model to ensure behaviours become new social norms.
• Although randomised control trials (RCTs) are often considered the gold standard of research, it is important to interpret the findings in light of some of the limitations and in the context of the study settings. It is also vital to reflect on the quality of the programme, and the extent of public acceptance of the intervention.

These study findings are not generalisable to all WASH interventions on nutrition, as we allude to in our analysis. Those working to improve child health must continue to make public policy decisions based on the full body of available evidence, which underpins a strong link between poor WASH and nutrition. Furthermore, health and non-health benefits of basic WASH services are well known and proven elsewhere, so investing in basic WASH services will remain a fundamental pillar of development.

Read a more detailed summary of WaterAid’s analysis here.