Globally, stunting affects 26% (165 million) of under-5-year children, underlies 15-17% of their mortality and leads to long-term cognitive deficits, fewer years and poorer performance in school, lower adult economic productivity, and a higher risk that their own children will also be stunted, perpetuating the problem into future generations. Stunting begins antenatally and peaks at 18-24 months of postnatal life, when mean length-for-age Z-score (LAZ) is about -2.0 among children living in Africa and Asia. Improving the diets of young children can reduce stunting, though, at best, only by about one-third. Frequent diarrheal illness has also been implicated. However, the effect of diarrhea on permanent stunting is relatively small, maybe because children grow at "catch-up" rates between illness episodes.

The Sanitation Hygiene Infant Nutrition Efficacy (SHINE) trial is motivated by a 2-part premise:

• A major cause of child stunting and anemia is Environmental Enteric Dysfunction (EED). EED is a subclinical disorder of the small intestine, which is virtually ubiquitous among asymptomatic people living in low-income settings throughout the world. EED is characterized by increased permeability which facilitates microbial translocation into the systemic circulation and triggers chronic immune activation.
• The primary cause of EED is infant ingestion of fecal microbes due to living in conditions of poor quality and quantity of water, sanitation, and hygiene (WASH).

The Sanitation Hygiene Infant Nutrition Efficacy ("SHINE") trial will test the effects of two packages of interventions: 1) improved water, sanitation and hygiene (WASH) and 2) improved infant and young child feeding (IYCF) on child stunting and anemia in the first 18 months of life. The trial will be conducted in rural Zimbabwe where WASH is poor, food insecurity high, and where about 15% of pregnant women are infected with HIV. The study will enroll 5282 women early in pregnancy and follow them and their infants until 18 months after delivery. The study will be a cluster-randomized controlled trial: two entire districts in central Zimbabwe have been divided into 212 geographic areas, each of about 100 households. The areas will be randomly allocated (that is, assigned by according to chance like the flip of a coin) to one of four interventions:

1. Improved WASH (a ventilated pit latrine, hand washing facilities with soap, drinking water treatment, a protected play space and health lessons to adopt improved hygiene behaviors)
2. Improved Infant Nutrition (health lessons on best infant feeding practices and a nutritional supplement (Nutributter) to be fed daily to babies from 6 to 18 months).
3. Improved WASH and Infant Nutrition (both interventions)
4. Standard of Care

All women living in the two districts who become pregnant during the recruitment period of the study will be invited to enroll. They will receive one of the 4 packages of interventions according to the area where they live. Health lessons will be given by Village Health Workers. Latrines and hand washing facilities will be constructed by building teams. Mothers will be followed up by research nurses at 7 months gestation, and at 1, 3, 6, 12, and 18 months after delivery. Primary outcomes are infant height and hemoglobin at 18 months of age.

Within SHINE we will measure two causal pathways: the biomedical pathway and the program impact pathway.

The biomedical pathway comprises the infant biologic responses to the WASH and IYCF interventions that ultimately result in attained stature and hemoglobin concentration at 18 months of age; it will be elucidated by measuring biomarkers of intestinal structure and function (inflammation, regeneration, absorption and permeability); microbial translocation; systemic inflammation; and hormonal determinants of growth and anemia among a subgroup of infants enrolled in an EED substudy. The investigators will also ask these mothers to record daily any episodes of diarrhea; blood/mucus in the stool; cough; fast or difficult breathing; fever; and lethargy preventing breastfeeding, that the child has between 1 month and 18 months of age. A subgroup of infants will also have stool samples collected during diarrhoeal episodes to evaluate reductions in pathogen-specific diarrhoea following WASH interventions.

Since the mothers enrolled in SHINE will have lived in unsanitary living conditions throughout their lives, it is anticipated that most will have some degree of EED themselves. It is hypothesized that resulting chronic inflammation contributes to adverse birth outcomes, such as prematurity and low birth weight. This question will be investigated through an observational design. For all mothers enrolled in SHINE, the sugar absorption test described above will be conducted and specimens of saliva, stool and blood collected and archived at the 10-12 week gestation visit for subsequent assessment of EED biomarkers. The association of severity of EED with risk of adverse birth outcomes (low birth length and weight; miscarriage, stillbirth, and premature delivery) will be assessed.

The program impact pathway comprises the series of processes and behaviors linking implementation of the interventions with the two child health outcomes; it will be modeled using measures of fidelity of intervention delivery and household uptake of promoted behaviors and practices. We will also measure a range of household and individual characteristics, social interactions, and maternal capabilities for childcare, which we hypothesize will explain heterogeneity along these pathways.

Study participants will be women who are rural residents of Chirumanzu or Shurugwi districts in Zimbabwe and who become pregnant during the enrollment period of the trial and are identified and consent to participation during pregnancy, and their live born infants. A total of 5280 women will be enrolled.

Inclusion Criteria:

Pregnant women residing in the study districts, whose pregnancy is confirmed by a urine pregnancy test.

Exclusion Criteria:

• Women residing in the study districts who become pregnant during the enrollment period but do not consent to join the trial
• Women who reside in urban areas of these two districts
• Infants with major non-fatal abnormalities will not be excluded from study procedures, but will be excluded from the final analytic sample if the abnormality is likely to directly affect gut health/function or stature (e.g. neural tube defects, cerebral palsy, Down syndrome)

Given the nature of the intervention, masking to treatment group is not possible.