Worldwide severe acute malnutrition (SAM) affects millions of children and considerably contributes to under-five mortality, mainly in low-income settings. Among children with SAM, deaths occur largely in those with oedema and during early phase of treatment often aggravated by infection. Treatment outcome could be improved by enhancing monitoring of body hydration (the proportion of water in tissues) during treatment.

We studied 351children between 0.5 and 14 years admitted to Jimma University Specialized Hospital with SAM. We recorded weight, height and grade of oedema. Further we measured bioimpedance which is the resistance of the body to an imperceptible current of electricity and in healthy individuals this method is used to estimate total body water. Finally we estimated amount of total body water (TBW) using deuterium dilution method on a subset of 35 children.

There were two important findings in this study. First, by comparing the bioimpedance estimated TBW with that estimated by deuterium, we found that bioimpedance is unreliable for measurement of TBW in children with SAM. Second, using the changes in raw bioimpedance values over time it was possible to differentiate tissue-related weight changes from hydration-related weight changes. This new method can improve monitoring of nutritional oedema and dehydration during treatment of SAM.