The association of water carriage, water supply and sanitation usage with maternal and child health. A combined analysis of 49 Multiple Indicator Cluster Surveys from 41 countries

Abstract

Background: Millions of people carry water home from off-plot sources each day and lack improved sanitation. Research on the health outcomes associated with water fetching is limited, and with usage of improved sanitation is inconclusive.

Objectives: To analyse the association of water fetching, unimproved water supplies, and usage of improved sanitation facilities with indicators of women's and children's health.

Methods: 49 Multiple Indicator Cluster Surveys from 41 countries were merged, creating a data set of 2,740,855 people from 539,915 households. Multilevel, multivariable analyses were conducted, using logistic regression for binary outcomes, negative binomial regression for count data and ordinary linear regression for linear data. We adjusted for confounding factors and accounted for clustering at survey, cluster and household level.

Results: Compared to households in which no-one collects water, water fetching by any household member is associated with reduced odds of a woman giving birth in a health care facility (OR 0.88 to 0.90). Adults collecting water is associated with increased relative risk of childhood death (RR 1.04 to 1.05), children collecting water is associated with increased odds of diarrheal disease (OR 1.10 to 1.13) and women or girls collecting water is associated with reduced uptake of antenatal care (β-0.04 to -0.06) and increased odds of leaving a child under five alone for one or more hours, one or more days per week (OR 1.07 to 1.16). Unimproved water supply is associated with childhood diarhhoea (OR 1.05), but not child deaths, or growth scores. When the percentage of people using improved sanitation is more than 80% an association with reduced childhood death and stunting was observed, and when more than 60%, usage of improved sanitation was associated with reduction of diarhhoea and acute undernutrition.

Conclusion: Fetching water is associated with poorer maternal and child health outcomes, depending on who collects water. The percentage of people using improved sanitation seems to be more important than type of toilet facility, and must be high to observe an association with reduced child deaths and diarhhoea. Water access on premises, and near universal usage of improved sanitation, is associated with improvements to maternal and child health.

Keywords: Child health; Maternal health; Mortality; Multi-level modelling; Sanitation; Water.