ABSTRACT
This paper aims to assess the clinical; nutritional; microbiological; environmental and socio-economic determinants that have a seasonal distribution; or are potential confounders of a seasonal association of previously described seasonal fluctuations in plasma sodium and potassium concentrations in children with dehydrating diarrhoea. A cross-sectional analytical study was conducted from 15 April 2002 to 14 April 2003 of 350 children aged six weeks to two years admitted to the diarrhoea rehydration unit of a children's hospital in Cape Town. Multiple linear regression analysis showed the plasma sodium levels to be statistically significantly associated with age [-0.3 mmol/l per month of age; 95 confidence interval (CI): -0.2; -0.4]; enterotoxigenic Escherichia coli infection [-2.1 mmol/l; 95 CI: -4; -0.2]; enteropathogenic E. coli infection [-5.1 mmol/l; 95 CI: -7.1; -3.1]; being breastfed [1.9 mmol/l; 95 CI: 0.4; 3.4] and living in a brick house [2.7 mmol/l; 95 CI: 0.8; 4.5]. Plasma potassium levels were associated with duration of diarrhoea before admission [-0.02 mmol/l per day of diarrhoea; 95 CI: -0.01; -0.04]; cryptosporidium infection [-0.3 mmol/l; 95 CI: -0.1; -0.6] and parental education [0.04 mmol/l per year of education; 95 CI: 0.1; 0.01]. Of these; enterotoxigenic and enteropathogenic E. coli and cryptosporidium infections followed seasonal distributions that were similar to those of the electrolyte concentrations. Seasonal fluctuations in plasma sodium and potassium levels are at least partly explained by both enterotoxigenic E. coli and cryptosporidium infections working together. Enterotoxigenic E. coli infection has a larger effect on plasma sodium levels and cryptosporidium infection has a larger effect on plasma potassium levels