"cessful use of a chicken-based diet for the

: m e n t of severely malnourished children with

persistent diarrhea: A prospective, randomized study
iiiii!ii~

Samuel Nur/co, AID, dosdAlberto Garchz-Aranda, AID,Euyenia Fiahbein, RN, and

AIartha In& Pdrez-Ztiffiya, RD

Objective: To evaluate the efficacy of a chicken-based diet for the treatment of persistent diarrhea in severely malnourished children.

Stud}" design: Prospective, randomized, double-blind study that compared a chickenbased diet with elemental (Vivonex) and soy (Nursoy) diets. Hospitalized children with
third-degree malnutrition and persistent diarrhea, aged 3 to 36 months, were included.
Diets were isocaloric and given nasogastrically at 150 ml/kg per day in progressively
increasing concentrations.
Results: Fifty-six children were included (18 received Vivonex, 19 Nursoy, 19 chicken). They had a mean age of 6.4 _+4.4 months, a mean weight of 3604 _+1232 gm, and a
mean weight-for-age percentage of 51.4% _+7.2%. Sixty-four percent had associated
conditions on admission to the hospital. Forty-one children (73.2%) were successfully
treated (13 Vivonex, 13 Nursoy, 15 chicken). There were no differences in diarrheal
outcomes, and all groups had significant weight gain. Failure was independent of the
diet and was associated with the presence of infection on admission. There was a significantly higher nitrogen balance in the children from the chicken group (358.2 _+13
mg/kg per day) than in those receiving Vivonex (226.6 +_61) or Nursoy (291.4 _+111.6;
p < 0.05) groups.
C o n c l u s i o n s : The chicken-based diet was as effective as Vivonex or Nursoy. It is well
tolerated, inexpensive, and widely available and thus represents an effective and inexpensive alternative to the treatment of severely malnourished children with persistent
diarrhea. (J Pediatr 1997;131:405-12)

From the Departnwat of Pediatric GaaO'oenterologyand Nutrition Hospital l@mtil de Mdrico Federico Gdme;4Mexico CitN
d/fexico.
Supported in part b2- the Applied Diarrheal Disease Research Project at Harvard University, by means of a
cooperative agreement with the U.S. Agency for International Development, and in part by National Institutes
of Health grant T32-DK 07703.
Submitted for publication April 3, 1996; accepted Dec. 17, 1996.
Reprint requests: Samuel Nurko, MD, Pediatric Gastroenterology, Children's Hospital, 300 Longwood Ave.,
Boston, MA 02115.
~Dr. Nurko is now in the Combined Program in Pediatric Gastroenterology and Nutrifon, Children's Hospital,
Boston, Mass.
Copyright 9 1997 by Mosby-Year Book, Inc.
0022-3476/97/$5.00 + 0 9/21/79985

Persistent diarrhea confnues to be a major

health problem in developing countries 1"3
and is often associated with a deterioration
in nutritional state. 1'3'4The nutritional rehabilitation of children with PD and severe
malnutrition is difficult and usually requires
hospitalization 4 and specialized care. l's
Initially the introduction of parenteral nutrition improved the outcome of PD in these
children. 6'7 Recent studies have shown that
specialized enteral feeding during the diarrhea] episode results in improved nutritional outcomes. 1'4'8 Therefore the enteral administration of elemental and semielementat
diets, with supplementation with parenteral
nutrition when needed, has become the
standard therapy. 5-z'9 These specialized enteral feedings, however, are very expensive,
usually unpalatable, and not readily available in many areas of the world. 1,4,8

Recent work suggests that malnourished patients with PD may be capable of

tolerating more complex diets, 1'4'8 so efforts are being undertaken to find inexpensive, available, and culturally acceptable diets.l'8' 10,11 Because milk is not well
tolerated by children with PD when given
as a full diet, 1'4'12 alternatives have been
suggested. 1"4'8"10'1I Soy-based formulas
are still used extensively, but their efficacy continues to be controversial. 4'8
Chicken-based diets have been empirical-

405

THE JOURNAL OF PEDIATRICS

SEPTEMBER 1997

NURKO ETAL.

Table I. Composition of the diets at the maximum concentration

Study Design
DIETS

ly and successfully used for the treatment

of malnourished infants with PD when elemental or soy diets have not been available. 4'13-1s Chicken has the advantage of
being considered by mothers and health
personnel in Mexico 16'17 and other areas
of the world as a safe food for children
with diarrhea or malnutrition. 4'1a'14
Because the optimal nutritional therapy
for severely malnourished patients with
PD is still controversial, 1'4 we evaluated
the efficacy of a chicken-based diet for the
treatment of PD in severely malnourished
children.

METHODS
Patients
In this prospective, randomized, double-blind study a local chicken-based diet
was compared with both an elemental diet
(Vivonex Standard; Norwich Eaton) and
a soy-based formula (Nursoy; Wyeth
Laboratories) in the treatment of severely
malnourished hospitalized children with
PD.

406

The study was performed at the

Hospital Infantil de 1Vidxico Federico
Gdmez in Mexico City. Patients between
3 and 36 months of age hospitalized with
third-degree malnutrition of the marasmatic type with PD were included. Thirddegree malnutrition was defined by using
the Gdmez criteria for weight for age
(<60% of the National Center for Health
Statistics 50th percenfle), is PD was defined as three or more loose stools for 14
days or longer. 2
Patients with the following characteristics were excluded: exclusively breast
fed, chronic illness (e.g., acquired immunodeficieney syndrome, tuberculosis), congenital malformation, an abdominal condition that would preclude
enteral feedings, a severe condition requiring intensive care, or lack of parental
consent.
The protocol was approved by the local
ethical review committee and by the
Harvard School of Public Health Committee on the Use of Human Subjects in
Research. Informed consent was obtained
from all parents.

The mainstay of therapy for PD and severe malnutrition at the Hospital Infantil
de Mgxico has been the elemental diet
Vivonex Standard. 5 Vivonex contains
crystalline amino acids, glucose and glucose oligosaccharides, a small amount of
highly purified safflower oil, electrolytes,
minerals, micronutrients, and vitamins, s'9
For use in children with PD, we and others have shown that Vivonex is effective if
it is given in progressively increasing concentrations, starting at 150 ml/kg per day
in a concentration that provides 47.8
kcal/dl (12.5% weight/volume) and advancing slowly by 2.5% per day to a maximum concentration of 85.6 kcal/dl
(22.5% weight/volume) s'9 (Table I).
Sodium chloride and potassium chloride
were also added to the formula to ensure
administration of sodium, 4 mEq/kg per
day, and potassium, 3 mEq/kg per day. 5
The chicken-based diet was designed with
the use of tables of food composition 19'20
and consists of easily available and simple
ingredients: cooking oil, boiled chicken
breast, table sugar, and minerals. To prepare the chicken-based diet, we calculated
the total volume needed per day (150
ml/kg). At the maximum concentration
(Table I) the following ingredients per
deciliter of diet were used: 8 gm boiled,
comminuted chicken breast; 3 ml vegetable cooking oil; and 10.5 gm table
sugar. After these components were
blended together, the following minerals
were added: 5 ml calcium gluconate (10%
solution, PISA), 2.7 ml of dibasic sodium
phosphate (PISA, dibasic sodium phosphate), and 1.7 ml of magnesium sulfate
(10% solution, PISA). Sodium chloride
(0.gN solution) to achieve 4 mEq of sodium per kilogram of body weight per day
and potassium chloride to achieve 3
mEq/kg/day were also added. Finally,
boiled water was added to achieve the
total volume required.
The soyformula used was Nursoy, which
contains soy protein, coconut, safflower
and soy oils, sucrose, minerals, and vitamins. All diets were prepared in the pedi-

NURKO ET AL.

THE JOURNAL OF PEDIATRICS

Volume 13 I, Number 3
atric nutrition kitchen of the hospital
under the supervision of a trained nutritionist.
The study was designed to use Vivonex
as the standard against which the other
formulas were compared, so all three diets
were given nasogastrically at progressive
isocalorie concentrations. The maximum
concentration of the diets is shown in
Table I. Because of intrinsic differences in
diet composition, the percentage of total
calories provided as protein, carbohydrate, or fat varied (Table I).

PROTOCOL
Patients were randomly assigned to
treatment by using a table of random
numbers. Only the nutritionist who prepared the formula was aware of group assignment. The investigators, nurses, and
residents remained masked to the type of
diet because aluminum foil was used to
cover the formula bag and tubing.
On admission to the hospital, patients
were hydrated according to World Health
O r g a n i z a t i o n / U N I C E F guidelines with
the use of a standard glucose-electrolyte
solution. 21 Patients were then fasted
overnight. Hydration was maintained
during that time with intravenously administered fluids. The next morning the
assigned diet was started if the patient
was well hydrated and there were no
other contraindications to feeding. The
nasogastric tube was inserted by trained
nursing staff. The diet was started at the
lowest concentration at a volume of 150
ml/kg per day, and concentrations were
advanced every 48 hours. If no intolerance occurred, full concentration was
achieved by the ninth day (Table I). If
there was evidence of intolerance, the diet
concentration was either maintained or
decreased as follows: (1) It was kept unchanged if there was evidence of 2% or
3% positive reducing substances (before
or after hydrolysis) or if there was an increase in stool output of more than 50%
(>20 ml/kg). (2) It was reduced if Clinitest results showed 4% or if there was an
increase of 75% or more in the stool output (>20 m]/kg).
When full concentration of the diet was

achieved, it was maintained for an additional 7 days. Daily supplementation with

1 mg folie acid, 1 ml multivitamin (PolyVi-Sol), and elemental iron, 6 mg/kg, was
begun when the maximum concentration
was achieved. After 7 days of the maximum diet concentration, patients underwent a challenge with whole cow milk: we
administered half-strength whole cow
milk, 10 ml/kg, and advanced to fullstrength milk if tolerated. Milk-tolerant
patients continued their rehabilitation
with lactose-contalning formula or whole
milk, depending on the age. If patients
showed evidence of lactose malabsorption, as manifested by return of liquid
stools, with p H less than 5 and greater
than 2% reducing substances in the stool,
a milk-flee diet was instituted. After the
milk challenge, all patients restarted a
complete age-appropriate, complex-balanced diet, which was continued until discharge.
Cessation of diarrhea was defined as the
passage of formed stool not followed by
liquid stools for at least 24 hours.
Successful treatment was declared if the formula could be advanced to the highest
concentration and there was cessation of
the diarrhea at the end of the study. The
onset of nutritional recoverywas considered
to be when the diarrhea ceased and there
was consistent weight gain for at least 48
hours. Treatmentfai[are was declared if the
patient had 5% or more dehydration during the administration of the diet, if there
was clinical deterioration that precluded
further enteral therapy, if diarrhea persisted at the end of the study, or if the formula could not be advanced to full concentration.
When treatment was declared a failure,
the code was broken. If patients had been
receiving Nursoy or chicken, they were
started on a regimen of Vivonex. If the patients with treatment failure had originally been receiving Vivonex or were unable
to continue with enteral feedings, total
parenteral nutrition alone was initiated
and was then continued u n t l the patient
was stabilized and gaining weight. Continuous enteral feedings with Vivonex
were then added to the total parenteral

nutrition and advanced every 24 hours as

tolerated. Once patients achieved full enteral feedings, they continued to receive
Vivonex for another 2 weeks and nutritional rehabilitation continued as outlined
previously.

CLINICAL PROCEDURES
Nude weight was obtained on admission and daily thereafter. The posthydration weight, obtained on the morning of
the start of the feedings, was considered
the baseline weight. Weights were obtained at the same time every morning
with an electronic scale (ScalesTronLx,
Wheaton, Ill.) and were accurate to at
least 10 gm. Recumbent length was obtained with a specially designed board on
admission, at the end of 2 weeks, and before discharge. All measurements were
obtained by trained nutritionists, and
their accuracy was validated before the
beginning of the study.
All patients had baseline laboratory values obtained at admission; laboratory
studies included complete blood cell
count, electrolyte concentrations, D-xylose concentration, stool and urine cultures, and stool tests for ova and parasites.
Blood culture specimens were obtained
only if indicated.
All intake and output were recorded.
Patients, both male and female, were
placed on metabolic beds or cots for separation of stool from urine. 5 To confirm
Successful separation of stool and urine in
girls, we performed a separate analysis
for all the variables associated with the
stool collection at the end of the study.
No differences between sexes were found
(data not shown), so all data were pooled.
A 72-hour nitrogen balance test was performed at the end of the second week,
starting 4 days after the maximum diet
concentration had been achieved. The
beginning and end of the stool collection
time were marked by the fecal excretion
of orally administered activated charcoal. s The nitrogen balance was measured by the micro Kjeldahl method. 22
Tests for p H and reducing and nonreducing substances in stool were performed
daily.

407

NURKO ET AL.

THE JOURNAL OF PEDIATRICS

SEPTEMBER 1997

Table II. Patient characteristics at time of admission

INFECTION ON ADMISSION OR
DURING HOSPITALIZATION

Systemic infection was suspected and

treated with broad-spectrum intravenously
administered antibiotics if there was a general ill appearance with any of the following signs: temperature instability, hypotension, hypoglycemia, or acidosis despite
adequate hydration. Otitis media, urinary
tract infections, and pneumonia were treated with appropriate antibiotics. Children
with dysentery received trimethoprim-sulfamethoxazole. 4 Children infected with
Giardia larr~lia received metronidazole.

Statistical analysis
We calculated that a sample size of 20
children per group would be needed if we

408

assumed a power of 0.80, an alpha of 0.05,

and a difference of 30% in the duration of
diarrhea. The statistical analysis was performed with the use of SPSS/PC software
and Epi-Info software, version 5.01.
Significance was assumed when p was less
than 0.05. Descriptive analyses were used
to define the presenting characteristics.
Multivariate and repeated-measures
analyses of variance were used to establish differences between the three groups.
W h e n necessary, transformation of the
data was done to fulfill the assumption of
normally distributed residuals. Survival
analysis was used to compare the duration
of the diarrhea. Chi-square tests were
used for categorical variables, and the
Fisher Exact Test was used whenever

there were cells with small sizes. VaLues

are expressed as mean _+SD.

RESULTS
Admission Characteristics
A total of 60 patients were initially enrolled in the study. Four were later excluded: two in the Vivonex group (one
with acquired immunodeficiency syndrome, one who died before initiation of
feedings), one in the Nursoy group (the
patient had primary renal insufficiency),
and one in the chicken group (the patient
had acute renal failure as a result of dehydration soon after admission). For the
other 56 patients, Vivonex was given to

THE JOURNAL OF PEDIATRICS

NURKO ET AL.

Volume 13 I, Number 3
18, Nursoy to 19, and chicken to 19. Their
initial clinical and laboratory characteristics are Shown :in Table II. Sixty-four percent of the patients had associated conditions at the time of admission (Table II).
Fifty percent had a nongastrointestinal infection, and 14.3% had a gastrointestinal
infection.
There were no significant differences
between the three groups.

Table I I L Main outcome characteristics for the 41 patients who successfully completed

the stud;/

Outcome
A successful outcome was seen in 41 patients (73.2%): 13 (72.2%) with Vivonex,
13 (68.4%) with Nursoy, and 15 (78.9%)
with chicken (not significant). During the
study, 34 (60.7%) of the 56 patients had
some evidence of formula intolerance: 14
(77.8%) of the patients receiving Vivonex,
11 (57.9%) Nursoy, and 9 (47.4%) chicken (NS). The intolerance was transient in
19 (56%) of 34 patients. The other 15
(44%) (5 receiving Vivonex, 6 Nursoy,
and 4 chicken) had treatment failure. The
mean time from initiation of the diet to
failure was 85.6 72 hours (60.6 45.7
hours with Vivonex, 98.5 99.9 with
Nursoy, and 97.5 99.9 with chicken)
(NS). Intestinal pneumatosis developed
in 7.14% of the patients (2 patients receiving Vivonex, 1 Nursoy, and 1 chicken). One of the failures in the Nursoy
group was shown to be a result of allergy
to the formula.
Five patients (8.9%) died: two who had
been receiving Vivonex, 1 Nursoy, and 2
chicken (NS). The patients died of intestinal pneumatosis (2), central line-associated sepsis (2), and bacterial sepsis
(K[ebsfe/[apneanwn/ae)early in the hospital
course (1).
The other 10 patients with treatment
failure were successfully managed, and
their mean stay was 50 30 days. Total
parenteral nutrition was required in 7 of
the 10 patients, and 9 were eventually discharged home on a milk-containing diet
regimen. The other was discharged on a
soy- and milk-free diet regimen because of
allergy.

Diarrhea
The mean fecal output per kilogram of
body weight and the number of bowel

movements per kilogram per day in the

first 24 hours were similar in all groups
(Table II). There were also no differences
in the mean stool output per kilogram per
day or in the day of cessation of diarrhea
in comparison with the 41 patients who
successfully completed the study (Table
III).
The Figure shows the results of the survival analysis done to compare the daily
probability of continuing with diarrhea
among the three groups. There were no
differences between groups, and the median duration of diarrhea, estimated by
the analysis, was 8.8 days for Vivonex,
5.67 days for Nursoy, and 7.3 days for
chicken.

Nutritional Outcome
The mean number of total calories per
kilogram of body weight per day ingested
by each group after the full diet was tolerated was similar: 115.2 8.3, 111.3 9.1,
and 116.0 9.6 for the Vivonex, Nursoy,
and chicken groups, respectively. There
was a significant difference in the amount

of protein per kilogram per clay ingested

after the full diet was tolerated: 2.4 0.2
gm/kg per day, 3.4 0.3 gm/kg, and 3.5 _+
0.4 gm/kg with Vivonex, Nursoy, and
chicken, respectively (p < 0.05).
Table III shows the outcome characteristics of those patients successfully treated. There was a significant weight gain in
all groups and no differences between
groups.
All patients in each group had an apparent positive nitrogen balance and a
similar percentage of absorption, percentage of retention, and biologic values.
There was a statistically significant higher
nitrogen balance (p < 0.02) and a tendency toward a higher number of children
with nutritional recovery in the chicken
group (NS).

Laboratory Tests
The serum albumin concentration decreased significantly in the Nursoy group
(from 3.5 0.6 to 3,l
grrdcll;p < 0.05),
whereas it did not change significantly in
the other groups: 3.3 0.6 to 3.2 0.5

409

THE JOURNALOF PEDIATRICS

NURKO ET AL.

SEPTEMBER 1 9 9 7

A sodium concentration less than 130

mmol/L (relative risk, 3.07; 95% confidence limits, 1.41 to 6.65) and the presence of associated infections (RR, 3.61;
95% CL, 1.1 to 14.42), particularly infection with Cryptospor~ium (RR, 4.15; 95%
CL, 1.58 to 6.67) or pneumonia (RR,
3.25; 95% CL, 1.53 to 6.9), were identified
as important factors associated with treatment failure.

1.0
VIVONEX
"'. \ \

O.B

NUNSOY

\
.

CHICKEN

\\
-

0.6
\

DISCUSSION

\\

"0
~

0.4

\
m

oe-4
o,

~ ~

0.2

10

0.0
0

""

12

14

16

Days since admission

F~ure. Probabilityof continuingdiarrhea since admission:a comparisonbetween the diets.

gngdl with Vivonex, and 3.0 + 0.7 to 3.3 +

0.3 gm/dl with chicken. There were no
electrolyte abnormalities noted in children
of either group, and there were .no other
significant differences in laborato W values between formulas (data not shown).
At the end of the study, children receiving Vivonex had a significantly higher DxNlose concentration (34.6 _+ 13.7 mg/dl)
than those receiving Nursoy (23.8 _+ 10.1
mg/dl) or chicken (23.1 _+11.5 mg/dl) (0 <

0.05)
Milk Tolerance Test
Intolerance was present in 7 patients
(17%): in none of the 13 patients in the
Vivonex group, in 3 (23.07%) of 13 in the
Nursoy group, and in 4 (21.1%) of 15 in
the chicken group. Those with milk intolerance had a lower admission weight
(2900.83 _+ 289.24 vs 3659.28 _+ 1258.91
gm; p < 0.004), a tendency to be younger
(3.91 +- 2.80 vs 6.64 + 4.1 months; p <

410

0.07), and a lower D-xylose concentration

at baseline (17.00 ~_2.09 vs 23.28 10.6
mg/dl; p < 0.004) than those who tolerated
milk. There was no difference in the D-xylose level at the end of the study when
both groups were compared (20.66 _+8.61
vs 27.89 +_13.2 mg/dl).

Risk Factors
There were significant differences (o <
0.05) between patients with treatment
success and those with treatment failure
with regard to the following admission
characteristics: albumin concentration
(3.2 _+0.6 vs 2.9 _+0,4 gm/dl), sodium concentration (138.4 _+ 6.2 vs 133.5 _+ 7.9
mmol/L), and the incidence of associated
infections (56.1% vs 86.7%). There were
also differences in stool output on the second (20.9 _+19.8 vs 47.4 +_33.9 ml/kg) and
third (16.7 +_ 16.5 vs 54.0 _+40.3 ml/kg)
days. No other significant differences
were found.

We have shown in this study that the

use of a locally available chicken-based
diet is at least as effective as elemental and
soy-based diets in the hospital treatment
of severely malnourished children with
PD. The main components of the chicken-based diet are easily available, are culturally acceptable, and are inexpensive. 16'17 Chicken-based diets were also
previously used as an alternative for the
treatment and rehabilitation of children
with malnutrition 1'10'13'14 and acute diarrhea. 15'16 However, use of this diet in children with P D has been limited. 1'11
This study included a difficult population that is frequently excluded from
other clinical trials. 1'8'13 Severely malnourished children with P D usually have
a high mortality rate and high treatment
failure rates. There can be up to a 17-fold
increase in the mortality rate, l'a'la'ya and it
has been suggested that approximately
49% (range, 23% to 62%) of diarrhea-associated deaths result from P D and malnutrition. 1'a'13 The mortality rate found in
this study (8.9%) compares favorably
with rates previously reported in the literature, la'93 We also did not find any differences in outcome when comparing the
treatment failure rates of those children
ha~ng the most extreme levels of malnutrition (weight-for-age percentage, <40%)
with the rates of less malnourished children. We confirmed that children with superimposed infections are at a higher risk
of treatment failure, which emphasizes
the need to look for and control superimposed infections at the time of admission
and nutritional rehabilitation. 13'25
As in previous studies, the diets were

NURKO ET AL.

THE JOURNAL OF PEDIATRICS

Volume 13 I, Number 3
given at a fixed volume and the caloric
density was advanced slowly. 5'6'9'2a'24
Sixty percent of children acquired some
signs of intolerance while the diet regimens were being advanced. These signs
were most likely related to carbohydrate
malabsorption, 12 and a good correlation
between fecal carbohydrates and total
fecal output has been shown. 25 O f the
three diets tested, Vivonex has a much
higher carbohydrate concentration (Table
I), mainly of oligosaecharides, which
probably accounts for the higher incidence of transient intolerance seen in
those patients. &25 Transient intolerance
was also seen in children receiving chicken or soy, which suggests that malnourished infants with PD frequently have
transient intolerance to other sugars. 25
Caution should be exercised in the treatment of those patients in whom increasing
stool outputs appear in the first and second day, because they may be at risk of
failure. In those children, a slower advancement of the dietary regimen may be
necessary.
A potential shortcoming of this study is
the difference in the macronutrient composition of the diets (Table I). In Vivonex, the
majority of the calories are provided by
carbohydrate and the protein content is
lower, &26"27providing only about 8% of
total calories. Previous studies confirm that
children receiving 6.70/0 of energy as protein achieved a slow compensatory
growth, 24"27and it has been shown recently that, in recovering malnourished infants,
there were no differences in growth when
formulas with 5.5%, 6.70/0, and 8.0% protein calories were compared. 28 Furthermore other studies have documented the
adequacy of Vivonex for growth and for
treatment of PD. 5'9 It is then possible that
the hydrolyzed amino acids are better absorbed. 5'9 This difference in protein content may partially explain why children on
the chicken-based diet regimen had a significantly higher nitrogen balance than
those receiving Vivonex.
The protein and caloric intakes were
similar in children receiving chicken or
soy feedings. The higher nitrogen balance
in those receiving chicken indicates that
chicken protein has a higher biologic

value than soy. Chicken has a low osmolarity, a better amino acid score, and a
higher degree of digestibility and bioavailability. 11'14'24We also found a significant
decrease in the serum albumin concentration in patients who received Nursoy despite a positive nitrogen balance. These
data suggest that the protein status and
lean body mass of malnourished patients
fed soy formulas may be deteriorating
slowly despite apparently adequate nitrogen retention. 29 It is therefore possible
that protein intake with Nursoy was inadequate to allow more rapid accretion of nitrogen at higher energy intakes. 29 Other
problems have also been associated with
the use of soy formulas in these patients.
Some authors have found that nearly 50%
of hospital-referred patients with PD do
not recover from diarrhea a week after the
introduction of a soy formula. 8 It has also
been suggested that soy-containing diets
may produce transient sensitivity and
subtle mucosal abnormalities in the intestinal mucosa of children with diarrhea,
with the potential for increasing the severity of their illness.16
Like Vivonex, 5 the chicken diet requires the addition of minerals. These
mineral additions make preparation suitable only in health care facilities, a factor
that does not represent a major obstacle
for severely malnourished children with
PD, who usually require hospitalization. 4
In the community the treatment of children with PD needs to include continued
feeding with locally available, inexpensive, and effective nutrients. I It is possible
that the chicken-based diet may be another alternative once the usual therapies
have failed.
All diets were administered continuously via nasogastric infusion, a method of
feeding that has been shown to have a
beneficial effect in control of diarrhea, nutrient absorption, nitrogen balance, and
weight gain in children with PD. <8'11'1a
Two main limitations of the nasogastric
route need to be mentioned. The technical
aspects of the placement and management
of the feedings requires specialized personnel and equipment, s More importantly the child has no control over the
amount of food that is being ingested,

which increases the risk of overzealous

refeeding or intolerance 27 and, later, the
risk of limiting the amount of nutrients in
comparison with ad lib oral intake.
Although the use of milk as the sole nutrient for children with P D has been
shown to be deleterious, 12 the question
that remains unanswered is when milk
can safely be reintroduced into the diet of
these children. We found that 83% of patients who successfully finished the study
were able to tolerate a full milk load at
least after 2 weeks of nutritional rehabilitation. Those children who were milk intolerant at the end of the study had an initial lower admission weight, were
younger, and had a lower D-xylose concentration, which suggests that their initial mueosal damage was greater. 6 Most
likely the intolerance was related to lactose malabsorption, although we cannot
exclude the possibility of intolerance to
milk protein.
In summary, severely malnourished
children with P D can be successfully
managed with a chicken-based diet.
These children were able to tolerate a
complex diet, achieve positive nitrogen
balance, and show weight gain. There was
no advantage to the use of an elemental or
a soy-based diet. Clear benefits of the
chicken-based diet include good tolerance, low cost, availability, and cultural
acceptance. Therefore the chicken-based
diet represents a good alternative for the
treatment of hospitalized children with severe malnutrition and PD.
We express special thanks to the nurses~ residents, and laboratmy personnel of the nutrition
~vardat the Hospital lnfantil de ~Ilxico for their

help in the pe,formanee o/this study, in p~rtieula6 we thank Gina Toussaint, Dr. Liliana
Worona, Dr. Alejandra Consnelo, Rosaura
P~'ez, Sarah Arvizu, and Aloniea Covarrubias.
We are indebted to all the personnel of the
Applied Dialv'heal Disease Research Projectfor
rhea" support during the poformance of the
study. We also thank DI: Laurie Fishman and
DI: Alan Leiehtnerfor their o'itical reviewof the
manuscript andfor lheir helpful su~gestions.

REFERENCES
1. Bhutta ZA, Hendricks KM. Nutritional
management of persistent diarrhea in child-

411

THE JOURNALOF PEDIATRICS

SIFPTEI"IBI:R1997

NURKO ET AL.

2.

3.

4.

5.

6.

7.

8.

9.

10.

412

hood: a perspective from the developing

world, d Pediatr Gastroenterol Nutr 1996;
22:17-37.
Persistent diarrhea in children in developing countries. Memorandum from a World
Health Organization meeting. Bull WHO
1988;66:707-17.
Fauvean V, Henry FJ, Briend A, Yunus M,
Charkraborty d. Persistent diarrhea as a
cause of childhood mortality in rural
Bangladesh. Acta Pediatr 1992;381(Suppl):
12~4.
Bhan MK, Arora NK, Singh KD. Management of persistent diarrhea during infancy
in clinical practice. Indian d Pediatr 1991;
58:769-74.
Vega Franco L, Carbajal-Ouzmgm A,
Garcfa-Aranda JA. Alimentacidn enteral
continua en nifios lactantes empleando una
dicta elemental. Bol Med Hosp Infant Mex
1982;39:651-8.
Orenstein S. Enteral vs parenteral therapy
for intractable diarrhea of infancy: a
prospective randomized trial. J Pediatr
1986; 109:277-86.
Kleinman RE, Galeano NE Ghishan E
Lebenthal E, Sutphen J, Ulshen MH.
Nutritional management of chronic diarrhea and/or malabsorption. J Pediatr
Gastroenterol Nutr 1989;9:407-15.
Bhutta ZA, Molla AM, Issani Z,
Badruddin S, Hendrieks K, Snyder JD.
Dietary management of persistent diarrhea
and malnutrition: comparison of a traditional rice-lentil-based diet with soy formula.
Pediatrics 1991;88:1010-8.
Sherman ,JO, Hamty CA, Khachadurian
AK. Use of an oral elemental diet in infants
with severe intractable diarrhea. J Pediatr
1975;86:518-23.
Roy SK, Haider R, Akbar MS, Alam AN,
Khatum M, Eeckels R. Persistent diarrhea:
clinical efficacy and nutrient absorption
with a rice-based diet. Arch Dis Child 1990;
65:294-7,

11. Godard C, Bustos M, Mufioz M, Nussle D.

Value of a chicken-based formula for
refeeding of children with protracted diarrhea and malnutrition in a developing country. J Pediatr Gastroenterol Nutr 1989;9:
473-90.
12. Penny ME, Paredes E Brown KH. Clinical
and nutritional consequences of lactose
feeding during persistent postenterifis diarrhea. Pediatrics 1989;84:835-44.
13. Maffei HV, Padula NN, Annicchino GP,
Ferrari GF, Goldberg TB. Nutritional management and weight changes during hospitalization of Brazilian infants with diarrhea:
primary reliance on oral feeding or continuous nasogastrie drip with locally made,
modulated minced chicken formula, d Trop
Pediatr 1990;56:240-6.
14. Larcher E Shepherd R, Francis DEM,
Harries JT. Protracted diarrhea in infancy.
Arch Dis Child 1977;52:597-605.
15. Romer H, Guerra M, Pina ,JM,
Urrestarazu MI, Garcia D, Blanco ME.
Realimentafion of dehydrated children with
acute diarrhea: comparison of cow's milk to
a chicken-based formula, d Pediatr Gastroenterol Nutr 1991;13:46-51.
16. Maulrn-Radovgm I, Brown KH, Acosta
MA, Fern~indez-Varela H. Comparison of a
rice-based, mixed diet versus a lactose-free,
soy-protein isolate formula for young children with acute diarrhea. ,J Pediatr
1994; 125:699-706.
17. Martfnez-Salgado H, Saueedo G. Mothers'
perceptions about childhood diarrhea in
rural Mexico. d Diarrheal Dis Res
1991;9:235-43.
18. Gomez E Ramos Galvan R, Cravioto J,
Frenk S. Malnutrition in infancy and childhood, with special reference to kwashiorkor. Adv Pediatr 1955;7:131-55.
19. Chavez M, Hernandez M, Roldan S.
Tablas de valor nutritivo de los alimentos
de mayor consumo en Mexico. In:
Comision National de Alimentaeion, edi-

20.

21.

22.

25.

24.

25.

26.

27.

28.

29.

tors. Institnto Nacinnal de la Nutricion;

1992.
Composition of Foods [Agricultural
Handbook 8]. Washington (DC): U.S.
Department of Agriculture; Oct 1975.
The management of diarrhea and use of
oral rehydration therapy. Geneva: World
Health Organization; 1985.
Henry RJ. Clinical chemistry principles
and techniques. New York: Harper & Row;
1964.
MaeLean WC, Romafia GL, Massa E,
Graham GG. Nutritional management of
chronic diarrhea and malnutrition: primary
reliance on oral feeding. J Pediatr 1980;
97:316-23.
Brown KH. Appropriate diets for the rehabilitation of malnourished children in the
community setting. Acta Pediatr Scand
1991;374(Suppl):151-9.
Lifschitz C, Carrazza F. Effect of formula
carbohydrate concentration on tolerance
and maeronutrient absorption in infants
with severe chronic diarrhea. J Pediatr
1990; 117:578-85.
Field CR, Schoeller DA, Brown KH. Body
composition of children recovering from severe protein-energy malnutrition at two
rates of catch-up growth. Am ,J Clin Nntr
1989;50:1266-75.
Whitehead RG. Protein and energy requirements of young children living in the
developing countries to allow for catch-up
growth after infections. Am J Ctin Nutr
1977;50:1645-7.
Graham GG, MacLean WC, Brown KH,
Morales E, Lembcke ,J, Gastanaduy A.
Protein requirement of infants and children: growth during recovery from malnutrition. Pediatrics 1996;97:499-505.
MacLean WC, Graham GG. The effect of
energy intake on nitrogen content of
weight gained by recovering malnourished infants. Am J Clin Nutr 1980;55:
905-9.