Inpatient treatment for complicated severe acute malnutrition (SAM) continues to have a high mortality in Africa. This is partly because children are commonly brought for admission because they are seriously ill, rather than being brought to hospital because of malnutrition alone. Mortality rates are especially high where SAM is complicated by HIV or TB. The early phase of inpatient nutritional treatment for severe acute malnutrition is based on a low-protein milk known as F75, which is given to improve metabolic homeostasis prior to the re-feeding to achieve catch-up growth. F75 provides a high proportion of energy from carbohydrates, including sucrose, lactose and maltodextrin. However, malabsorption of different types of carbohydrates, but lactose in particular, is known to occur in SAM and may lead to osmotic diarrhoea. Diarrhoea is common in children with SAM and is associated with increased mortality. Furthermore, switching from a catabolic state to a high energy diet that consists of predominantly carbohydrates can lead to 're-feeding syndrome' that may lead to severe electrolyte abnormalities and multiple organ dysfunction.

The aim of this trial is to determine whether reducing the carbohydrate content of F75, and removing lactose, improves the stabilisation of severely malnourished children. The trial will involve randomising children who are eligible to receive F75 milk to either the current formulation or a revised formulation. Both formulations will be given according to current recommendations regarding frequency of feeding and caloric value. Since the purpose of F75 is to stabilise the child metabolically and biochemically, the primary endpoint of the trial will be time to stabilisation (the end of the first phase of treatment for severe acute malnutrition). Blood and stool samples at admission and after three days will be used to determine the effects on carbohydrate and fat malabsorption and evidence of the re-feeding syndrome. Children will be followed up until discharge from hospital. The project has been planned in consultation with the World Health Organisation (WHO) and, if the revised formulation of F75 results in improved outcomes, will lead to a global change in recommendations for its formulation.

Admission to hospital with complicated severe acute malnutrition (SAM) in Africa commonly has a case fatality of 10-30%. Importantly, children are usually admitted to hospital because they are severely ill rather than for malnutrition alone. Mortality may be improved to some extent by adherence to WHO recommended management, but the direct application of these guidelines to all contexts is controversial. The WHO guidelines were designed on the basis of historical data on nutrient requirements and medical treatments and almost none of the recommendations are supported by evidence from clinical trials.

For the treatment of complicated inpatient SAM, the guidelines consist of three distinct phases of treatment: phase 1 or stabilization phase where a low protein, liquid diet (F75) is introduced with a reduced energy intake (80-100 kcal/kg/day). Once a child has stabilised, there is a 'transition phase' consisting of either ready to use therapeutic foods (RUTF) with supplemental F75 or alternatively another milk formula known as F100. RUTF is a peanut based, energy dense supplement used to obtain catch-up growth. F100 is a liquid formula with a higher energy density and protein content than F75. The caloric intake is increased daily to a maximum of 130 kcal/kg/day. Finally, 'Phase 3' is the recovery phase during which the aim is to achieve catch-up growth with either RUTF or F100.

Typically, the highest mortality rate is found in the early phases of treatment. Children who fail treatment early often have profuse diarrhoea, signs of circulatory insufficiency which is hard to treat. The most recent reports from Zambia and Kenya note a prevalence of diarrhoea of more than 60% amongst children with SAM and it is associated with increased mortality. Currently at Kilifi County Hospital and Coast Provincial General Hospital, ~20% of children are given either a diluted or lactose-free milk feed during rehabilitation of SAM, although there are no specific guidelines for this. Diarrhoea may be caused by a viral or bacterial gastroenteritis, sepsis, or may be nutritionally induced (osmotic). There are no routinely available tests to distinguish osmotic from infective or other causes of diarrhoea in hospitals in sub-Saharan Africa.

The proportion of energy derived from carbohydrates in F75 is high. The carbohydrates in F75 milk (as well as F100 and RUTF) consist of a mixture of maltodextrin, sucrose and lactose. Disaccharides such as maltose, lactose or sucrose are normally hydrolysed into monosaccharides by disaccharidases localized at the tip of small intestinal villi. The monosaccharides such as glucose and galactose can then be transported across the apical membrane through Na+ dependent glucose transporter, whilst fructose makes use of a facilitative fructose transporter. There is limited information on the intestinal function and intestinal carbohydrate absorption in malnourished children. However, data from Jamaica and South Africa suggest that there is impaired absorption of disaccharides (lactose and sucrose), regardless of the presence or suspicion of gastroenteritis. Limited histological evidence has shown intestinal atrophy in children with SAM. These data are consistent with clinical signs of lactose malabsorption found in children with severe malnutrition. Recently, evidence of impaired absorption of monosaccharide glucose in children with SAM was reported in Malawi.

Apart from diarrhoea, early deterioration may also be related to severe metabolic derangements due to a sudden change from a catabolic to an anabolic state, resulting in refeeding syndrome. Refeeding syndrome is characterized by hypophosphataemia, hypokalaemia, hypomagnesaemia and sodium retention. These severe electrolyte disturbances can lead to impaired cardiac, pulmonary and neurological function and are often hard to treat. By receiving energy predominantly from sugars, pancreatic insulin secretion is increased which induces uptake of electrolytes including phosphate and potassium, into cells. Furthermore, as protein synthesis is stimulated, increased production of adenosine tri-phosphate (ATP) leads to a higher cellular demand for phosphate. Although insulin secretion appears to be partially impaired in the early stages of refeeding, hypophosphatemia is a common feature during refeeding of malnourished children, and is associated with mortality.

In this trial, the investigators aim to evaluate the outcome of using a revised formulation of F75 milk with reduced carbohydrate composition and without lactose, compared to the current formulation of F75 during the initial stabilisation period amongst children with severe acute malnutrition. In the new formulation, more will be provided by lipids and the total energy provided will be unchanged. The trial will be undertaken in two hospitals in Kenya and one hospital in Malawi.

Enrolment and follow up was completed in December 2015, laboratory analysis of plasma and faecal samples in ongoing.

Inclusion Criteria:

Age 6 months to 13 years

Severe malnutrition defined as: mid upper arm circumference (MUAC) <11.5cm if less than 5 years old;19 or weight for height Z score <-3; or kwashiorkor as defined in the current Kenyan and WHO guidelines.

Admitted to hospital because of medical complications or failure of an appetite test as defined in the current WHO guidelines.

Eligible to start F75 milk by current WHO guidelines.

Exclusion Criteria:

Declined to give informed consent.

Known allergy to milk products.

Any other reason the consenting investigator thinks that in the child's best interests it inappropriate for them to take part.