Page 108 - 15Diarrhoeaandvomiting

Nutritional management

infants were randomised to receive white grape juice (WGJ) twice daily and 30 infants were
randomised to receive coloured, flavoured water (WA) twice daily. The WA was coloured and
flavoured to resemble juice and the investigators were unaware of the group allocation, but the
process of randomisation was not explained. Plain water was offered ad libitum between meals to
all infants. Children in all three groups were similar in age, duration and severity of diarrhoea and
presence of vomiting, and there were no differences between the groups for serum electrolyte
levels or haematocrit values. Although the total energy intake was higher in the juice-fed groups
compared with the water group, the duration of diarrhoea after randomisation was significantly
lower in the water group compared with the juices group (49.4 ± 32.6 hours AJ group versus
47.5 ± 38.9 hours WGJ group versus 26.5 ± 27.4 hours WA group; P < 0.05). The mean weight
gain was higher in the juice groups but the difference was not statistically significant. [EL = 1+]

Evidence summary
Six trials compared the introduction of full-strength feeding after rehydration with graded re-
feeding but there were differences between the trials regarding the method of graded re-feeding
and outcomes measured. Five of these trials were of poor quality (with EL = 1−) and had small
sample sizes. However, the evidence was consistent in that there was no harm in giving immediate
full-strength re-feeding with cow’s milk formula following rehydration and also no benefit of
graded re-feeding over immediate full-strength re-feeding. Two trials found evidence of increased
weight gain with full-strength formula but the differences were not statistically significant.
Results from another trial with EL = 1+ suggested that giving juices to children after rehydration
prolonged the duration of diarrhoea compared with water. Although children receiving juices
had a higher weight gain, the difference was not statistically significant.

6.2.3 Reintroduction of solid foods

Evidence overview
There were seven RCTs that describe mixed diets. All the studies involved 100 participants or
fewer (range 46–95) and focused on culture-specific foods.
Two of the RCTs were direct comparisons of two local diets.
The first RCT, conducted in Pakistan, 127 recruited 78 children aged between 6 and 36 months
admitted to hospital after presenting to outpatients with a history of acute non-bloody diarrhoea
of less than 7 days’ duration. Children were excluded from the study if they were exclusively
breastfed, had a temperature of over 102 °F, had any systemic illness (including pneumonia,
meningitis or convulsions), were comatose after rehydration, had seizures, had paralytic ileus
or had third-degree malnutrition (Gomez classification). Children were rehydrated if necessary
with either ORS solution or IV Ringer’s lactate and then randomly assigned (computer-generated
sequence and allocation of treatment by drawing lots) to either the dowdo diet (wholewheat
flour, cow’s milk, oil, salt and water) (n = 39) or the khitchri diet (rice, lentils, oil, salt and
water) (n = 39). One child from each group was later withdrawn from the study and there were
three treatment failures (two in the khitchri group and one in the dowdo group). The two groups
had similar characteristics at the start of the study. Outcomes were assessed over 5 days. No
statistically significant differences were seen between the two diets in stool weight (males
only), stool frequency, caloric consumption/kg, total weight change or duration of hospital stay.
However, mothers reported that the children preferred the dowdo diet (27/36) to the khitchri diet
(19/34) (P = 0.2) and therefore were more likely to use it at home. [EL = 1+]
The second study, conducted in Peru, 128 recruited 46 male infants aged between 6 and 24 months
if they had acute diarrhoea (more than three stools per 24 hours) of less than 96 hours’ duration
prior to hospital admission. Children were excluded if they were breastfed more than once a
day, had significant wasting, oedema or systemic illness affecting enteral feeding, or had had
a diarrhoeal episode in the previous fortnight. The children were randomly assigned (fixed
interval, block randomisation to allow for age and dehydration confounders) to either a rice
bean diet (Phaseolis vulgaris, ‘frijol canario’, soybean oil, cotton seed oil) (n = 25) or rice soy diet

(rice, soy protein isolate, corn syrup, soybean oil, cotton seed oil) (n = 21). Rehydration therapy
(according to WHO guidelines) and subsequent hydration maintenance was given as necessary.

83