Hypo-phosphataemia in children under five years with kwashiorkor and marasmic kwashiorkor.

BACKGROUND Severe malnutrition contributes up to 50% of childhood mortality in developing countries is frequently characterised by electrolyte depletion, including low total body phosphate. During therapeutic re-feeding, electrolyte shift from extracellular to intra-cellular compartments may induce hypo-phosphataemia (hypo-P) with resultant increased morbidity and mortality. This biochemical imbalance is under-recognised, and the frequency of this problem among African malnourished children is unclear. OBJECTIVES To determine the magnitude of hypo-phosphataemia in children under five years of age presenting to Kenyatta National Hospital with kwashiorkor and marasmic kwashiorkor and to evaluate the relationship between hypo-phosphataemia and nutritional intervention during the first five days of treatment. DESIGN Short longitudinal survey. SETTING The General Paediatric wards of the Kenyatta National Hospital (KNH), Nairobi. SUBJECTS Children under five years of age presenting with kwashiorkor or marasmic kwashiorkor at KNH were recruited into the study. MAIN OUTCOME MEASURES Low serum phosphate level (< 1.20 mmol/l) and patient outcome (survival or death) during the first five days of treatment. RESULTS One hundred and sixty five children were enrolled between June 2005 and February 2006 of which 107 (64%) had kwashiorkor and 58 (36%) had marasmic kwashiorkor. They were of mean age 20 months (range 3-60), and 95 (58%) were male. The prevalence of hypo-phosphataemia was 86% on admission, increased to 90% and 93% on day one and two respectively, and then declined to 90% by the fourth day. At admission 6% were hypo-phosphataemic, increasing to 18% and 22% on day one and two respectively, and declining to 11% by day four. On admission mean serum phosphate was below normal at 0.91 mmol/l, declined significantly to 0.67 mmol/l and to a nadir of 0.63 mmol/l after the first and second day of treatment respectively, then rose slightly to 0.75 mmol/l on the fourth day (p < 0.001 comparing each follow-up mean level with the admission level). There was a positive association between severity of nadir serum phosphate level and mortality (p = 0.028). There were no deaths among children with normal nadir serum phosphate levels. However, among children with mild, moderate and severe nadir hypo-phosphataemia, 8,14 and 21% died respectively. Children with dermatosis and hypomagnesaemia showed a trend for association with mortality (p = 0.082 and 0.099 respectively). CONCLUSION Hypo-phosphataemia is frequent among children with kwashiorkor and marasmic kwashiorkor presenting at KNH. Serum phosphate levels decline significantly during the first two days of nutritional intervention, and severity of