Page 104 - 15Diarrhoeaandvomiting

Nutritional management

post-hydration diet. A further seven RCTs investigated which solid foods may be suitable for
the re-feeding period. For the last sub-question, there were three RCTs comparing lactose with
lactose-free cow’s milk feeds and three comparing soy formula with lactose-containing formula.
One RCT compared the effectiveness of soy formula in early and late re-feeding and two others
compared cow’s milk formula versus a special formula.

6.2.1 Early versus late re-introduction of feed
Evidence overview
Five studies (three RCTs and two quasi-randomised trials) compared the effectiveness of early
feeding with late feeding.
118
A multicentre RCT involving 12 European hospitals was conducted to compare the effect of
early or late feeding on the duration and severity of diarrhoea, weight gain and complications
in weaned infants (n = 230) younger than 3 years who were hospitalised with acute diarrhoea
(duration more than 1 but less than 5 days). Excluded were children with short gut syndrome,
chronic inflammatory bowel disease, ileus, previous treatment with antidiarrhoeal drugs,
associated hepatic or renal disease, and those already receiving ORS solution or on IV fluids.
After appropriate rehydration with ORS solution over a period of 4 hours, the children were
allocated by random numbers to receive either their usual diet (early feeding group n = 134) or
continue ORS solution only for 20 hours followed by their usual diet (late feeding group n = 96).
In addition, both groups were offered ORS solution (10 ml/kg) for each watery stool. Breastfed
children received ORS solution and diet in addition to breastfeeding. On comparing the baseline
characteristics of the two groups, it was found that there were statistically significant differences
between the two groups regarding the age of introduction of solid foods, proportion of children
presenting with blood in stools, proportion of children with rotavirus detected and incidence of
lactose intolerance, but it was not clear whether these factors had been adjusted during the final
data analysis. Moreover, limited data were given for the results and they were expressed mainly in
graphs. The authors found mean weight gain in the early feeding group to be significantly greater
compared with the late feeding group at both 24 hours (P = 0.01) and during hospitalisation
(P = 0.001), but the weight gain at day 5 and day 14 was similar in both the groups. There were
also no statistically significant differences between the two groups for the duration of diarrhoea
or the incidence of watery stools or vomiting on days 1–5. [EL = 1−]
In a four-armed RCT carried out in Peru, 119 138 hospitalised male children aged between 3 and
36 months with diarrhoea and dehydration (mild to severe) were recruited. Excluded from the
study were children receiving more than one breastfeeding per day, those who received more
than a single dose of antibiotics, those who had had an episode of diarrhoea within the previous
3 weeks, and those with poor nutritional status (weight-for-length < 2 SD below the national
standard). Children (n = 10) were also excluded from the data analysis if they did not remain
in the hospital for the study period. After initial rehydration for 2–4 hours with ORS solution
according to WHO guidelines (except Group 4 which received IVF for rehydration), children
were assigned to four dietary groups by a block randomisation procedure. These groups were:

• Group 1 (n = 31): full-strength formula based on casein, sucrose, dextrin, maltose, soybean
oil and cotton oil with a vitamin–mineral mix added
• Group 2 (n = 29): half-strength formula with the same composition as Group 1 for the first
48 hours, followed by full-strength formula
• Group 3 (n = 34): ORS solution continued for the first 48 hours followed by half-strength
formula for the next 48 hours and then full-strength formula
• Group 4 (n = 34): IV fluids (no oral fluids) for the first 48 hours followed by half-strength
formula for the next 48 hours and then full-strength formula.

Thus by day 5, children in all four groups received the same dietary therapy. The main outcome
measures were changes in body weight and duration of diarrhoea over a 2 week period. The
baseline characteristics of the four groups were similar but data on the outcomes were either
presented graphically or were not presented clearly. It was seen that, although all groups gained
weight during the first 12 hours in the hospital, only children in Group 1 were able to maintain
a positive weight trend after 24 hours. When the combined Groups 1 and 2 were compared with
combined Groups 3 and 4 at 2 weeks after admission, the difference in weight increments was

79