Lipid Based Suppl Vs CSB Mod Underwt Malawi 2010
Lipid Based Suppl Vs CSB Mod Underwt Malawi 2010
Lipid Based Suppl Vs CSB Mod Underwt Malawi 2010
http://jn.nutrition.org/content/suppl/2010/10/20/jn.110.12249
9.DC1.html
Department of International Health, University of Tampere Medical School, Tampere 33014, Finland; College of Medicine, University of Malawi,
7
8
Blantyre P/Bag 360, Malawi; Singapore Clinical Research Institute, 138669, Singapore; Department of Paediatrics, Tampere University Hospital,
9
Tampere 33014, Finland; and National Public Health Institute, Helsinki 30, Finland
Abstract
Although widely used, there is little information concerning the efficacy of corn-soy blend (CSB) supplementation in the
treatment of moderate underweight in African children. Lipid-based nutrient supplements (LNS), which have proven to
be beneficial treatment for severely wasted children, could offer benefits to less severely affected individuals. We
conducted a clinical randomized trial to determine whether LNS or CSB supplementation improves weight gain of
moderately underweight children. A total of 182 underweight [weight-for-age Z-score (WAZ) , 22] 6- to 15-mo-old
children were randomized to receive for 12 wk a ration of 43 g/d LNS or 71 g/d CSB, providing 1189 and 921 kJ,
respectively, or no supplementation (control). The primary outcome was weight change; secondary outcomes included
changes in anthropometric indices, hemoglobin levels, and morbidity. The body weight increases (mean 6 SD) did not
differ and were 620 6 470, 510 6 350, and 470 6 350 g in the LNS, CSB, and control groups, respectively (P = 0.11).
Compared with controls, infants and children in the LNS group gained more weight [mean (95% CI) = 150 g (0300 g);
P = 0.05] and had a greater increase in WAZ [0.33 (20.020.65); P = 0.04]. Weight and WAZ changes did not differ
between the control and CSB groups. In exploratory stratified analysis, the weight increase was higher in the LNS
group compared with the control group among those with lower initial WAZ [250 g (60430 g; P = 0.01].
Supplementation with LNS but not CSB modestly increases weight gain among moderately underweight children and
the effect appears most pronounced among those with a lower initial WAZ. J. Nutr. 140: 20082013, 2010.
Introduction
Childhood undernutrition is an important public
health and development problem in low- and middleincome countries. Children who are mildly
underweight with weight-for-age Z-scores (WAZ) 10
between 21.0 and 22.0 have twice the risk of
1
Supported
by the Academy
of Finland
no. 109796).
Chrissie
Thakwalakwa
and John
Phuka (grant
received
personal
stipends
from
Nestle Foundation.
Nutriset
Ashorn,
J.until
C. conflicts
Phuka,
December,
Y.ofB.
2003.
Cheung,
C.T. funders
M.
Puumalainen,
Thakwalakwa,
P.
M. Maleta,
the
implementation,
no
analysis,
interest.
or reporting.
Trial
had no and
role K.
in
3
This
trial was
NCT00420368.
registered
at
clinicaltrials.gov
as
Methods
This was a single-center, randomized, controlled, investigator-blinded
clinical trial testing the growth-promoting effect and other health benefits to
infants and children of daily provision of LNS or CSB for 12 wk. The
primary outcome measure was weight change during the follow-up period.
Secondary outcomes included mean changes in length (mm), hemoglobin
(Hb) concentration (g/L), weight-for-length Z-score (WLZ), length-for-age
Z-score (LAZ), mid-upper arm circumference (MUAC), head
circumference, and incidence of adverse events (AE) or serious AE (SAE).
The researcher and data collector responsible for collecting the outcome
measures were unaware of participants group allocations but the other data
collectors were aware of the allocations.
Study area. The study was conducted in Lungwena, Mangochi district of
Malawi, southeastern Africa. Exclusive breastfeeding for 6 mo is almost
nonexistent in the study area and infant diets are typically complemented
with maize porridge (phala; 10% dry weight maize flour) already as soon
as 2 mo after birth. Underweight (WAZ , 22) and stunting (LAZ , 22) are
very common, with a prevalence of 40% and almost 80% by 18 mo of age,
respectively (3). The study area has 1 rainy season between December and
March during which the staple food, maize, and other crops are grown.
Enrollment in the trial was conducted during this growing season when food
levels were at the lowest, i.e. from December to February 2007. The 12-wk
follow-up of the last participant ended in May 2007.
Sample size. The target sample size was 63 infants and children per group
(189 in total) calculated from the expected difference in the primary
outcome, i.e. weight gain, among those provided either LNS or CSB and
those provided nothing. The expected difference was based on an assumed
weight gain (mean 6 SD) of 550 6 440 g among the control infants and
children and 800 6 440 g among those receiving either LNS or CSB
supplementation (14,17). This gave the trial 85% power and a type 1 error of
no more than 5% to detect a difference of $250 g in the mean weight gain
between the control and intervention groups and allowed a 10% loss to
follow-up.
Trial interventions. Participants in the control group did not receive any
food supplement during the trial period. Those in the first intervention group
received 500 g of CSB weekly and those in the second intervention group
received 300 g of LNS weekly for 12 wk. Food supplements were delivered
to their homes. The guardians were provided with spoons and advised to
give their infants twice daily either 3 spoonfuls of LNS or porridge made
from 5 spoonfuls of CSB. The mean daily dose was 43 g of LNS (921 kJ) or
71 g of CSB (1189 kJ). The LNS could be eaten alone or mixed with other
infant foods, whereas the CSB required processing, typically through
cooking into porridge. Mothers were encouraged to give the supplement in
addition to breastfeeding. Information about breastfeeding status was
collected monthly during health center visits. Rab Processors produced the
CSB locally in Malawi. LNS was produced at a Malawian nongovernmental
organization, Project Peanut Butter, from locally purchased peanut paste
(26%), dried skimmed milk (25%), vegetable oil (20%), icing sugar (27.5%),
and a premade mineral and vitamin mix (1.5%) from Nutriset. Both supplements were fortified with micronutrients, but the level of fortification varied
between the products. Table 1 provides the energy and nutrient content of a
daily ration of each supplement.
Measurement of outcome variables. Infant nude weights were measured
to the nearest 10 g using an electronic infant weighing scale (SECA 735,
Chasmors). Head circumference and MUAC were measured to the nearest 1
mm by a nonstretchable measuring tape (Lasoo-o-tape, Harlow Printing).
Length was measured to the nearest 0.1 cm using a high quality length board
(infantometer, Child Growth Foundation). All anthropometric measurements
were done in triplicate by 1 investigator (C.T.) assisted by 1 trained research
assistant, both of whom were unaware of the participants group allocations.
Anthropometric indices (WAZ, WLZ, LAZ) were calculated using Epi Info
3.3.2 software (CDC) based on the CDC 2000 growth reference (16),
because the WHO 2006 growth standards (18) became available only after
enrollment. As a sensitivity analysis to assess if the use of the WHO growth
standards would give similar results, we conducted exploratory analyses
based on the WHO 2006 standards.
During weekly home visits, trained research assistants asked if the
participants had experienced any problems eating the food supplements. If
the guardians spontaneously reported that the participant had suffered from
diarrhea, abdominal discomfort, vomiting, or skin rash, they were advised to
bring the trial participant to the local health center. At these and any
unprompted, nonscheduled visits, a medical assistant assessed and managed
the participant and recorded all information on structured forms. A trial
physician (J.P.) who was unaware of the participants group allocations
reviewed the data on suspected AE and determined and classified the
severity of AE and the likelihood of their association to the trial
interventions. All suspected SAE were reported to the trials Data Safety
Monitoring Board (DSMB).
Supplemental feeding in Malawi 2009
TABLE 1
Weight, g
Energy,1 kJ
Protein, g
Carbohydrates, g
Fat, g
Retinol, mg RE
Folate, mg
Niacin, mg
Pantothenic acid, mg
Riboflavin, mg
Thiamin, mg
Vitamin B-6, mg
Vitamin B-12, mg
Vitamin C, mg
Cholecalciferol, mg
Calcium, mg
Copper, mg
Iodine, mg
LNS or CSB
Iron, mg
Magnesium, mg
Zinc, mg
CSB
71
LNS
43
1189
10.4
921
6.0
NA2
3.1
139
43.2
3.5
NA2
11.9
13.5
400
160
6
2
0.3
0.13
0.3
0.9
48
NA2
0.5
0.5
0.5
0.9
30
5
72
NA2
NA2
366
0.4
135
5.46
NA2
8
60
3.6
8.4
Results
Of the 1304 infants and children who were initially screened,
951 (73%) did not meet inclusion criteria for further eligibility
assessment (.15 mo of age, WAZ . 21.7). Of the remaining
353 infants and children, 37 (10%) did not attend the enrollment session or refused participation and 124 (35%) were not
eligible for inclusion. The remaining 192 infants and children
enrolled and were
randomized into 1 of the
3 trial arms
( Supplemental Fig. 1). At enrollment, there were slightly more
children in the LNS (66 ) and CSB (67) groups compared with
the control group (59), but the groups were comparable for
baseline characteristics (Table 2).
A total of 188 participants (98%) completed the intervention,
i.e. underwent medical and anthropometric assessment after the
12-wk intervention. The deaths and/or losses to follow-up did not
differ among the 3 groups. All mothers reported that their children
readily ate the provided supplement and diversion of any portion
to anyone other than the intended beneficiary was reported at 2/
795 (0.3%) food delivery interviews in CSB and none in LNS
delivery interviews. From the weekly home visits during which
trial products were
checked, the percentage of
visits when
leftovers were found were 4/795 (0.5%) in the CSB and 4/780
(0.5%) in the LNS group. All infants and children except 1 from
the control group were breastfed during the entire study period.
During the 12-wk follow-up, the mean weight increase was
620 g in the LNS group, 510 g in the CSB group, and 470 g in the
control group but did not differ among the groups (P = 0.11).
The corresponding changes in WAZ were +0.02, 20.31, and
20.32 units, respectively (P = 0.03), and none of the participants
had a WLZ . 2.00 at the end of the trial period. Compared with
the controls, infants and children in the LNS group gained more
weight (150 g; P = 0.05) and had a greater increase in WAZ
(0.33; P = 0.04). Changes in weight and WAZ did not differ
between the control and CSB groups. Changes in length, head
circumference, MUAC, and Hb concentration did not differ
between the control and either of the intervention groups (Table
3). In exploratory analyses based on the WHO 2006 growth
standard, infants in the LNS group tended to have a greater
(12
wa
transform
ex
betw
te
b
g
AN
re
tory
aforeme
TABLE 3 Anthropometric outcomes among infants and children that received LNS, CSB, or no
1
supplement (control) for 12 wk
LNS vs. control
Outcome
Control
LNS
CSB
P2
ANOVA.
3
Difference (95% CI) P
t test.
0.05
0.51
0.23
0.99
0.04
0.13
0.21
0.60
3
Difference (95% CI) P
0.49
0.34
0.27
0.25
0.91
0.87
0.59
0.35
fortified
supplementary
foods among 6- to 18-moold underweight infants
and children. Participants
receiving LNS for 12 wk
gained significantly more
weight (150 g) than did
the
unsupplemented
controls. The differences
appeared more among the
most
undernourished
participants, i.e. those who
had lower initial WAZ
before the intervention. In
contrast, supple-mentation
with CSB was not
associated
with
significantly higher
TABLE 4 Anthropometric
outcomes among
infants and
children stratified
by WAZ at
enrollment that
received LNS,
CSB, or no
Supplemental
sup
ple
me
nt
(co
ntro
l)
for
12
wk
Length increase, cm
3.4 6 1.0
3.4 6 1.2
3.4 6 1.1 0.
WAZ change
20.36 6 0.58 20.18 6 0.97 20.29 6 0.44 0.
Hb change, g/L
23.1 6 12
25.6 6 23
22.6 6 11 0.
Values
are
mean
difference
(95%
CI). 6 SD or mean
ANOVA.
feeding in
TABLE 5 Proportion with confirmed AE and SAE among infants and children that received LNS, CSB, or
no supplement (control) for 12 wk
LNS vs. control
Outcome
Control
RR (95% CI)
3 (5)
0.20
5.4 (0.73.0)
0.08
2.6 (20.325.0)
13 (19)
0.25
1.7 (0.83.7)
0.14
1.4 (0.63.2)
LNS
CSB
59
66
67
SAE, n (%)
1 (2)
6 (9)
8 (14)
15 (23)
Participants, n
NA
0 (0)
6 (9)
2 (3)
0.03
2 (3)
8 (12)
6 (9)
0.21
2.6 (0.612.6)
Diarrhea, n (%)
4 (7)
9 (14)
11(16)
0.25
2.0 (0.66.2)
0.17
2.4 (0.87.2)
0 (0)
5 (8)
4 (6)
0.09
NA
0.04
5 (8)
5 (8)
2 (3)
0.40
0.9 (0.32.9)
0.56
0.02
P1
RR (95% CI)
NA
NA
0.8 (0.11.7)
0.3
6
0.2
6
0.2
9
0.3
6
0.0
8
0.0
8
0.1
7
Fishers test.
NA = No values computed.
t
h
T
h
T
h
Thakwalakwa et al.
2012
7.
reactions caused by allergy to the milk or peanut
component of LNS cannot be ruled out. Strong
allergic reactions seem uncommon altogether
among severely undernourished individ-uals (28)
and to our knowledge none have been associated
with LNS use in that target group. Less severely
affected children might have an intact immune
system, rendering them more prone to
immunological reactions. In the current sample,
all episodes of vomiting or rash were transient,
possibly suggesting a mechanism other than
allergy to explain the finding.
In conclusion, our results suggest that a 12 wklong supple-mental feeding with LNS can boost
weight gain in moderately underweight children
and may promote recovery from under-nutrition.
Further studies should assess if LNS
supplementation is associated with an increased
risk for allergic reaction or other symptoms in this
target group.
Acknowledgments
We thank Daniel Pondani, the medical assistant
for Lungwena Health Centre, for his support in all
the stages of the study; the DSMB members
(Kamija Phiri, Paul Ndebele, Tom Heikens) for
monitoring the trial; Laszlo Csonka for designing
the data entry program; and Matti Kataja for his
excellent technical advice on statistical analyses.
All authors designed the trial; P. A. wrote the
protocol; C.T. was responsible for data collection;
Y.B.C. designed the details of statistical analysis;
and C.T. conducted the analysis and wrote the
first draft of the manuscript under the supervision
of K.M., P.A., and Y.B.C. All authors commented
on the analysis and participated in writing of the
manuscript. C.T. had full access to all the data in
the study and takes responsibility for data
integrity and accuracy of the data analysis. All
authors read and approved the final manuscript.
Literature Cited
1.
2.
3.
4.
5.
6.
8.
9.
11.S
10.
a
n
d
i
g
e
H
,
N
d
e
k
h
a
M
J,
B
ri
e
n
d
A
,
A
s
h
o
r
n
P
,
M
a
n
a
r
y
M
J.
H
o
m
e
b
a
s
e
d
tr
e
a
t
m
e
n
t
o
f
m
a
l
n
o
u
ri
s
h
e
d
M
a
l
a
w
i
a
n
c
h
ildren with locally produced or imported ready-to-usefood. J Pediatr Gastroenterol Nutr. 2004;39: 1416.
12.
13.
14.
15.
16.
Kuczmarski RJ, Ogden CL, Guo SS, GrummerStrawn LM, Flegal KM, Mei Z, Wei R, Curtin LR,
Roche AF, et al. 2000 CDC growth charts for the
United States: methods and development. Vital Health
Stat 11. 2002;246:1190.
17.
18.
19.
20.
21.
22.
23.
24.
Rocquelin G, Tapsoba S, Kiffer J, EymardDuvernay S. Human milk fatty acids and growth of
infants in Brazzaville (The Congo) and Ouagadougou
(Burkina Faso). Public Health Nutr. 2003;6:2418.
25.
26.
27.
28.
Supplemental feeding
in Malawi 2013