To date there have been few reports on the impact of dietary intervention on the clinical course of acute shigellosis. Current management of acute shigellosis is primarily focused on antibiotic therapy with less emphasis on nutritional management. In a randomised clinical trial, we examined the role of an energy-dense diet on the clinical outcome in malnourished children with acute dysentery due to shigellosis. Seventy-five children aged 12–48 months with acute dysentery randomly received either a milk–cereal formula with an energy density of 4960 kJ/l (test group) or a milk–cereal formula with energy of 2480 kJ/l (control group) for 10 d in hospital. In both milk–cereal formulas, protein provided 11 % energy. In addition, the standard hospital diet was offered to all children and all children received an appropriate antibiotic for 5 d. The mean food intakes (g/kg per d) in the test and control groups were: 112 (SE 2·28) AND 116 (se 3·48) (P=0·16) on day 1; 118 (se 2·72) and 107 (se 3·13) (P=0·04) on day 5; 120 (se 2·25) and 100 (se 3·83) (P=0·04) on day 10. The mean energy intakes (kJ/kg per d) in the test and control groups respectively were: 622 (se 13·2) and 315 (se 11·3) (P<0·05) on day 1; 655 (se 15·1) and 311 (se 7·98) (P<0·05) on day 5; 672 (se 14·7) and 294 (se 11·1) (P<0·05) on day 10. The food and energy intakes were mostly from the milk–cereal diet. There was no difference between two groups in resolution of fever, dysenteric (bloody and or mucoid) stools, stool frequency and tenesmus. However, vomiting was more frequently observed among the test-group children during the first 5 d of intervention (67 % v. 41 %, P=0·04). There was an increase in the mean weight-for-age (%) in the test group compared with the control group after the 10 d of dietary intervention (6·2 (se 0·6) v. 2·7 (se 0·4), P<0·01). In addition, resolution of rectal prolapse was better (26 % v. 8 %, P=0·04) in the test group v. control group after 5 d, and 13 % v. 6 %, (P=0·08) after 10 d of dietary intervention. Supplementation with a high-energy diet does not have any adverse effect on clinical course of acute shigellosis and reduces the incidence of rectal prolapse in malnourished children.

Efficient isolation of individual lipid classes is a critical step in the analysis of plasma and lipoprotein fatty acid compositions. Whilst good separations of total lipid extracts are possible by TLC, this method is time consuming and a major rate-limiting step when processing large numbers of specimens. A method for rapid separation of phosphatidylcholine (PC), non-esterified fatty acids (NEFA), cholesterol ester (CE) and triacylglycerol (TAG) from total plasma lipid extracts by solid-phase extraction (SPE) using aminopropyl silica columns has been developed and validated. Following initial separation of polar and neutral lipids, individual classes were isolated by application of solvents with increasing polarity. Recoveries for combined plasma extraction with chloroform–methanol and SPE were (%): PC 74·2 (SD 7·5), NEFA 73·6 (sd 8·3), CE 84·9 (sd 4·9), and TAG 86·8 (sd 4·9), which were significantly greater for TAG and NEFA than by TLC (P<0·001). Both GC–flame ionisation detector and GC-MS analysis of fatty acid methyl esters demonstrated that there was no cross-contamination between lipid classes. Measurements of repeatability of fatty acid composition for TAG, PC, CE and NEFA fractions showed similar CV for each fatty acid. The magnitude of the CV appeared to be related inversely to the fractional fatty acid concentration, and was greatest at concentrations of less than 1 g/100 g total fatty acids. There was no evidence of selective elution of individual fatty acid or CE species. In conclusion, this method represents an efficient, rapid alternative to TLC for isolation of these lipid classes from plasma.