The effect of zinc supplementation on intestinal permeability changes and protein loss was studied in 32 children aged between 1 Zinc is a micronutrient that is commonly deficient in children in developing countries. Loss of zinc occurs in infants with acute diarrhoea.7 Furthermore, hypozincaemia and low rectal mucosal zinc concentrations have been shown in cases of childhood chronic diarrhoea,8 9 although it was not clear whether the mucosal injury associated with these conditions resulted in such a loss. Zinc supplementation in patients with diarrhoea has been shown, however, to improve the mucosal integrity and is associated with significant reduction in diarrhoea attack rates, duration of diarrhoea, and stool volume, particularly when the patients are undernourished.6 The present study also aimed to explore the inter-relationship between gut permeability changes and enteric protein loss and to determine if zinc supplementation affects these changes.
MethodsForty three boys and girls aged 1-12 years with a history of bloody mucoid diarrhoea of less than three days' duration were initially enrolled in the study. Microscopic examination of the stools of these patients suggested shigellosis.10 Children with obvious systemic illness -for example, pneumonia, meningitis, leukaemoid reaction,11 severe malnutrition (<65% weight for age according to the National Center for Health Statistics), otitis media, and distension of the abdomen -were excluded from the study. All An intestinal permeability test using a freshly prepared solution containing 5 g lactulose with 0 5 g lactose (7.5 ml of Duphalac, Duphar Labs, Southampton, UK) and 1 g mannitol made up to 20 ml with 1% chloroform water was performed using the method described by Behrens et al,13 starting soon after admission when rehydration had been accomplished. All patients were hydrated and were made to empty their bladder before the test dose was offered. Adhesive paediatric urine bags (Downs Ltd, London, UK) containing a drop of 20% (vol/vol) chlorhexidine gluconate to prevent bacterial degradation of the probes were applied to the clean perineum to collect all urine samples over the next five hour period for subsequent measurement of the markers by an automated enzyme assay using Cobas-bio (Switzerland).After this test, two charcoal tablets (500 mg, homogenised in 15 ml water) were given orally. Once the charcoal marker had appeared in the stool and the patients were clinically settled after full rehydration, a 48 hour balance study (first balance study), as previously described,4 was begun with a diet of known compositions (Table I). Food was weighed to an accuracy of 0-1 g on a Toledo scale (Ohaus, Dial Og) and was offered freely. The amounts offered and left over were measured and the difference was recorded as the amount consumed. None of our patients was breast fed. All patients received nalidixic acid (55 mg/kg per day in four divided doses for five days), as Shigella species isolated from their stools or rectal swabs, or both, were sensiti...