Efficacy and safety of new complementary feeding guidelines with an
emphasis on red meat consumption: a randomized trial in Bogota,
Colombia1–3
Gilma A Olaya, Margaret Lawson, and Mary S Fewtrell
ABSTRACT
Background: Iron deficiency and poor linear growth are common
in infants from deprived socioeconomic backgrounds and may be
associated with inadequate complementary feeding (CF) practices.
Objective: We tested the hypothesis that new CF guidelines em-
phasizing meat as a source of iron and zinc would improve linear
growth, iron, and zinc status in infants living in poor socioeconomic
circumstances in Bogota, Colombia.
Design: A total of 85 term infants who were exclusively breastfed
for $4 mo were randomly assigned at 6 mo of age to a control
group [CG (n = 43); current advice] or intervention group (new
guidelines group [NGG (n = 42); with counseling to 1) continue
breastfeeding, 2) offer red meat $3 d/wk, and 3) offer fruit and
vegetables daily]). Main outcomes were 1) linear growth from 6 to
12 mo of age; 2) hemoglobin, hematocrit, iron [serum ferritin (SF)],
and zinc status at 12 mo of age; and 3) meat intake at 12 mo of age
(by using a food-frequency questionnaire).
Results: A total of 38 infants/group provided data at 12 mo of age.
NGG infants had significantly higher red meat intake [mean 6 SD:
5.4 6 1.8 compared with 3.5 6 1.7 d/wk at 12 mo of age; P ,
0.001), higher hemoglobin and hematocrit at 12 mo of age, and
a significantly greater increase in hemoglobin (mean 6 SD change:
0.41 6 0.8 compared with 20.13 6 1.0; P = 0.01) and hematocrit
(1.04 6 2.2 compared with 20.15 6 2.4; P = 0.03) from 6 to 12 mo
of age than those in CG infants. There were no significant differ-
ences in linear growth from 6 to 12 mo of age or in SF or zinc.
Conclusions: The new guidelines showed efficacy with higher red
meat intake and positive effects on hemoglobin and hematocrit. The
intervention was acceptable and affordable for most mothers. These
preliminary results suggest that the intervention merits investigation
in a larger cohort with longer-term follow-up. This trial was regis-
tered at http://isrctn.org as ISRCTN57733004. Am J Clin Nutr
2013;98:983–93.
INTRODUCTION
Complementary foods are introduced into the infant diet to
complement breast milk at a stage when sufficient amounts of
certain nutrients, notably iron and zinc, can no longer be provided
by breast milk alone. The WHO recommends that complementary
feeding (CF)4 should start from 6 mo of age (1, 2). However,
despite much discussion about the timing of the introduction of
CF, there has been little research on whether the type of com-
plementary foods introduced influences infant growth, nutritional
status, and health outcomes.
Am J Clin Nutr 2013;98:983–93. Printed in USA. Ó 2013 American Society for Nutrition
Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278
by guest
on 05 August 2018
Iron deficiency is common during the CF period, particularly
in vulnerable populations with a high prevalence of maternal iron
deficiency, low birth weight, infection, and poor dietary intakes
and may be associated with adverse effects on cognitive outcome
(3, 4). Zinc deficiency is also common and is associated with poor
linear growth and increased infection risk (5–9). Both micro-
nutrients are present in low concentrations in breast milk, albeit
with high bioavailability, and it is recognized that infants who
are exclusively breastfed for w6 mo require a good source of
these nutrients during CF, either directly from foods or as sup-
plements (10–12). Strategies for improving micronutrient intake
in low-income populations include food fortification, sprinkles,
and micronutrient supplements (13–20). Interventions that used
educational strategies to improve CF have shown positive effects
on meat consumption but have not focused on the effect on iron
and zinc status (21–27). Meat is a good source of bioavailable
zinc and iron as well as protein. In a randomized trial in the
United States, meat was acceptable to mothers and infants as an
early CF and associated with a trend toward better behavioral
outcomes at 12 mo of age (28), whereas Danish infants ran-
domly assigned to receive a higher meat intake had a smaller
decrease in hemoglobin in late infancy (29). A cohort study in
the United Kingdom reported positive associations between
meat intake from 4 to 12 mo of age and weight gain at #12 mo
of age and psychomotor development at 22 mo of age (30).
1
From the Nutrition and Biochemistry Department, Faculty of Sciences,
Pontificia Universidad Javeriana, Bogota, Colombia (GAO), and the Child-
hood Nutrition Research Centre, University College London Institute of
Child Health, London, United Kingdom (GAO, MSF, and ML).
2
Supported by the Childhood Nutrition Research Centre, University Col-
lege London Institute of Child Health, and Pontificia Universidad Javeriana.
Tommee Tippee (United Kingdom) donated the feeding spoons, cups, and
beakers used in the study.
3
Address correspondence to MS Fewtrell, Childhood Nutrition Research
Centre, University College London Institute of Child Health, 30 Guilford
Street, London WC1N 1EH. E-mail: m.fewtrell@ucl.ac.uk.
4
Abbreviations used: CF, complementary feeding; CG, control group;
CRP, C-reactive protein; LAZ, length-for-age z score; MCV, mean corpus-
cular volume; MUAC, midupper arm circumference; MUACZ, midupper
arm circumference z score; NGG, new guidelines group; RCT, randomized
controlled trial; SF, serum ferritin; WAZ, weight-for-age z score; WLZ,
weight-for-length z score.
Received January 25, 2013. Accepted for publication July 9, 2013.
First published online August 14, 2013; doi: 10.3945/ajcn.112.053595.
983
 984 OLAYA ET AL
Nevertheless, in most populations, meat is not introduced early
and is generally used in small amounts.
Our unpublished research with infants living in poor socio-
economic conditions in Bogota, Colombia (a group at known risk
of stunting and iron deficiency), identified several undesirable
practices during the CF period, including the early introduction of
infant formula and low intakes of iron and vitamin A. On the basis
of these data, we developed new CF guidelines to meet the es-
timated nutritional requirements of infants aged 6–12 mo of age
(WHO), with a focus on the use of red meat, including cheaper
and widely available iron-rich meats such as chicken liver, to
improve iron and zinc status. In this randomized controlled trial,
we tested the efficacy and safety of the new CF guidelines in the
same population. Our specific hypothesis was that the new
guidelines would result in 1) increased red meat intake, 2) im-
proved iron and zinc status, and 3) improved linear growth,
without adverse effects on adiposity or breastfeeding. We also
investigated the acceptability and affordability of the new guide-
lines and tolerance of the complementary foods recommended.
SUBJECTS AND METHODS
Subjects
Mothers of healthy term infants with birth weight .2500 g
who were participating in the growth-monitoring program at 2
hospitals in Bogota, Colombia (Fontibon and Suba), and who
were exclusively breastfeeding when their infants were 4 mo of
age were approached and given information about the study.
Both hospitals serve populations with low socioeconomic status;
family incomes come mainly from informal and part-time jobs,
most families live in rented or shared households, and w29% of
the population experience food insecurity (31), although 97% of
the population has public services coverage including tap water,
electricity, and waste disposal. Parasitic infections are uncommon
in infants in this population, and malaria does not occur in Bogota.
Mothers who were willing to participate and who were still
breastfeeding when their infants were 6 mo of age gave written
informed consent. A baseline hemoglobin measurement was
performed; infants with a hemoglobin concentration ,11 g/dL
(the cutoff used to define anemia in Colombia) were excluded
from additional participation in the study and were referred to
a physician for treatment. Eligible subjects were randomly as-
signed to either the new guidelines group (NGG) or control
group (CG). Random assignment was stratified by hospital and
whether or not CF had been introduced to infants between 4 and
6 mo of age. Randomization assignments were prepared by
using randomized blocks of permuted length by a member of the
team who had no contact with study subjects and were stored in
sealed opaque envelopes. It was not possible to blind researchers
who collected anthropometric and food-intake data, but labo-
ratory measurements were blinded. The study was approved by
the research ethics committees at University College London
and Pontificia Universidad Javariana. This trial was registered at
http://isrctn.org as ISRCTN57733004.
Treatment of subjects
Subjects were seen at clinic visits when infants were 6, 8, 10,
and 12 mo of age. Infants who were randomly assigned to the
Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278
by guest
on 05 August 2018
NGG received individual nutrition counseling with face-to-face
sessions and detailed verbal and written guidance from re-
searchers (see supplemental material under “Supplemental data”
in the online issue for examples). Guidelines focused on the
following 3 main messages that were emphasized at all study
visits: 1) the importance of continuing breastfeeding alongside
CF; 2) the importance of including red meat as a source of iron
to prevent anemia; and 3) the importance of fruit and vegetables
as part of a healthy diet. Mothers were offered specific advice
on the number of portions of meat that should be given ($5
portions/wk, including red meat $3 times/wk); mothers were
also advised to include chicken liver and heart as affordable
forms of meat, and suggestions were given for the preparation of
recommended foods. Mothers were also advised to give fruit and
vegetables daily. Menu plans were provided at each study visit,
including suggestions for the number of meals, types of com-
plementary foods, suggested amounts to be offered, recipes, and
advice on food hygiene. Mothers also received a leaflet with
additional general recommendations and tips to improve their
knowledge and confidence in feeding their babies (see online
supplementary materials 1 and 2 under “Supplemental data” in
the online issue).
Mothers randomly assigned to the CG received the standard
advice on CF from health care professionals in the growth-
monitoring program. This information included advice on
breastfeeding, general recommendations on suitable comple-
mentary foods including meat, food hygiene, and food prepa-
ration; however, no specific advice was given on the frequency or
amount of foods that should be offered.
Where possible, visits were scheduled at the same time as
those that were part of the growth-monitoring scheme. Mothers
were reimbursed for their travel expenses. All participants re-
ceived a weaning set consisting of a bowl and spoon as a gift for
participating, and these sets were also used to standardize the
assessment of food portions. At the end of the study, the mother
also received an infant-feeding beaker.
Outcome measures were recorded as follows
1) Anthropometric measures were performed at each visit.
Length was measured by using an infantometer with
a fixed headboard and movable footboard to the nearest
1 mm. The mean of 3 measurements was used. Infants
were weighed naked on an electronic scale (Tanita). Head
circumference and midupper arm circumference were
measured to the nearest 1 mm by using a nonstretchable
measuring tape. Results were converted to SD scores by
using WHO 2007 growth-standard data (32).
2) A blood sample was obtained at 6 and 12 mo of age to
measure hemoglobin, hematocrit, serum ferritin (SF), and
zinc. Hemoglobin was measured by using spectropho-
tometry; mean corpuscular volume (MCV) and hemato-
crit were measured by using automated flow cytometry.
SF was measured by using a 2-site ELISA (DALTIS-
Dizar). Assays were calibrated from 0 to 1000 ng/mL
by using Quimiolab controls (Colombian external quality-
control program). Sensitivity was 1.0 ng/mL, and specific-
ity was measured by using controls from human liver
ferritin, spleen ferritin, and cardiac ferritin from Quimiolab
 EFFICACY OF NEW COMPLEMENTARY FEEDING GUIDELINES 985

FIGURE 1. Flowchart of subject progress through the study. CONSORT, Consolidated Standards of Reporting Trials.

controls. Serum zinc samples were collected in the morning
(nonfasting) by following the protocol recommended by
the International Zinc Nutrition Consultative Group in their
technical document (33); samples were processed by using
zinc-free needles, syringes, centrifuge tubes, storage vials,
and transfer pipettes, with avoidance of hemolysis and con-
tamination. Blood was stored in the refrigerator and centri-
fuged to separate the serum, which was frozen before
analysis by using atomic absorption spectrophotometry.
C-reactive protein (CRP) was measured by using a 2-site
ELISA (ADALTIS-Dizar) and used as a biomarker of in-
flammation with CRP concentrations .6 mg/L considered
abnormal.
4) The acceptability, tolerance and affordability of recom-
For the analysis of iron and zinc status the following cutoffs
recommended by the WHO were used: hemoglobin concen-
tration ,11 g/dL, hematocrit ,33%, and SF concentration
,12 mg/L (34); MCV ,70 fL (35); and serum zinc concen-
tration ,65 mg/dL (34–36).
3) The intake of foods specifically highlighted by the new
guidelines (meat, red meat, milk other than breast milk,
fruit, and vegetables) was recorded at 6, 8, 10, and 12 mo
of age by using a semiquantitative food-frequency ques-
tionnaire. The frequency and number of portions of each
food consumed on each occasion were recorded; fre-
quency was defined as follows: daily denoted food con-
sumed on all 7 d of the week, weekly denoted foods
consumed $4 d/wk (excluding foods that were consumed
daily), and monthly denoted food consumed $2/mo.
Breast-milk consumption was assessed by taking into
account 1) the number of breastfeeds per day, 2) the time
required for each breastfeed, and 3) the number of breast-
feeds per night.
mendations in the NGG were assessed by using a ques-
tionnaire completed by mothers when infants were 8, 10,
and 12 mo of age. Acceptability was assessed by using
a scale from 1 to 3 [1 = disliked (baby did not accept
food), 2 = liked (baby accepted the food), and 3 = liked
very much (baby accepted and enjoyed the food)]. Com-
plementary food tolerance was defined as tolerated and

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 986 OLAYA ET AL
TABLE 1
Baseline characteristics of mothers and infants according to randomly assigned group1
Variable NGG (n = 38) CG (n = 38) P

Mother’s age (y) 24.1 6 6.4 2

22.9 6 5.9 0.4
Father’s age (y) 28.8 6 9.5 25.1 6 5.6 0.04
No. of children 1.6 6 0.70 1.5 6 0.86 0.7
Family size 5.3 6 1.3 5.7 6 2.3 0.4
No. of rooms in home 2.3 6 0.9 2.7 6 1.4 0.2
Mother completed high school [n (%)] 20 (52.6) 25 (67.6) 0.2
Sex [n (%)]
F 19 (50) 19 (50) 1.0
M 19 (50) 19 (50) —
Gestational age (wk) [n (%)] 39.0 (1.0) 39.2 (1.0) 0.5
Delivery [n (%)]
Vaginal 25 (65.8) 30 (78.9) 0.3
Cesarean 13 (34.2) 8 (21.1) —
Birth
Weight (kg) 3.03 6 0.28 3.1 6 0.39 0.2
Head circumference (cm) 33.9 6 1.3 33.8 6 1.3 0.8
WAZ 20.58 6 0.6 20.34 6 0.8 0.2
Feeding practices [n (%)]
CF at 4–6 mo of age 17 (44.7) 17 (44.7) 1.0
CF at .6 mo of age 21 (55.3) 21 (55.3) —
.1 bottle-feed at 6 mo of age [n (%)] 9 (23.7) 7 (18.4) 0.6
Consuming cow milk at 6 mo of age [n (%)] 9 (23.7) 6 (15.8) 0.4
Eating red meat at 6 mo of age [n (%)] 9 (23.7) 14 (36.8) 0.4
At 6 mo of age
Weight (kg) 7.25 6 0.73 7.86 6 0.99 0.003
Length (cm) 64.9 6 2.38 65.6 6 2.05 0.2
Head circumference (cm) 43.0 6 1.14 43.4 6 1.34 0.2
MUAC (cm) 14.3 6 0.91 14.9 6 1.15 0.02
1
Baseline data are shown for infants with outcome data at 12 mo of age. Mean values were compared by using
Student’s t test, and categorical variables were compared by using the chi-square test. CF, complementary feeding; CG,
control group; NGG, new guidelines group; MUAC, midupper arm circumference; WAZ, weight-for-age z score.
2
Mean 6 SD (all such values).

not tolerated [1 = tolerated (baby took and ate the com-
plementary food) and 2 = not tolerated (baby did not eat
the CF or ate it but it caused distress, vomiting, or other
symptoms)]. Affordability was assessed by using a scale
of 1 or 2 (1 = affordable (if the mother had been able to
buy the CFs recommended $3 times/wk, and 2 = not
affordable (if the mother could not buy some of the
CFs recommended)].
Statistics
Sample size
The sample size of 64 infants/group was calculated to detect
a 0.5-SD difference in outcome measures at a = 0.05 with 80%
power, which we considered to be biologically relevant and
plausible given our previous studies that used nutritional in-
terventions in infants. The sample size was considered to be
achievable within the timeframe for the study given local figures
on infant-feeding practices, notably reported rates of exclusive
breastfeeding at 4 mo of age (37). To allow for losses to follow-
up, we planned to randomly assign 90 infants/group.
Data analyses
Data were analyzed with SPSS software (version 18; IBM).
Main outcomes were compared between randomly assigned
groups by using the t test for continuous variables (hemoglobin,
hematocrit, MCV, zinc, and anthropometric variables) and the
chi-square test for categorical variables (cutoffs for hemoglobin,
SF, zinc, and food-consumption variables). SF concentrations at
6 and 12 mo of age were not normally distributed and were log
transformed for analysis. For anthropometric variables and
markers of iron and zinc status, we compared the values at 12 mo
of age and the change from 6 to 12 mo of age between randomly
assigned groups. ANCOVA was also performed to examine
differences in the change in anthropometric or biochemical
variables from 6 to 12 mo of age between randomly assigned
groups with adjustment for baseline values and, where appro-
priate, potential confounders and to test for interactions between
baseline values and randomly assigned groups on outcomes.
RESULTS
Effect of the randomized intervention
The planned sample size could not be achieved within the
timeframe of the study, mainly because the prevalence of ex-
clusive breastfeeding in Bogota had fallen, which reduced the
number of eligible infants. A total of 353 mothers who were
exclusively breastfeeding when their infants were 4 mo of age
were approached (Figure 1); 168 mothers were eligible for the
study, and 116 mothers were willing to participate. A total of

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 EFFICACY OF NEW COMPLEMENTARY FEEDING GUIDELINES
FIGURE 2. Comparison of numbers of portions per week of selected
foods consumed from 10 to 12 mo of age according to the randomly assigned
group. Box-and-whisker plots of numbers of portions of different foods
consumed per week are shown according to the randomly assigned group.
Boxes represent 25th–75th percentiles, with medians shown as lines. Bars
represent highest and lowest values. Comparisons between randomly as-
signed groups were made by using the Mann-Whitney U test. The NGG
had a significantly higher consumption of meat, red meat, fruit, and vegeta-
bles. The CG had a significantly higher consumption of milk, sugar, and
sweetened foods. CG, control group; NGG, new guidelines group.
110 infants attended the screening blood test, and 22 (20%) of
these infants had a hemoglobin concentration ,11 g/dL; these
infants were treated with iron supplements and were not enrolled
in the study. Eighty-five eligible infants were randomly assigned
after the blood test, 42 infants to the NGG (20 boys) and 43
infants to the CG (21 boys). Four NGG and 5 CG infants were
lost to follow-up between baseline (6 mo of age) and 12 mo of
age. Baseline data for infants who also provided data at 12 mo of
age are shown in Table 1 according to randomly assigned
groups. For all infants randomly assigned, those in the CG were
significantly heavier with higher midupper arm circumference
(MUAC), weight-for-age z score (WAZ), weight-for-length z
score (WLZ), and MUAC z score (MUACZ) at baseline (6 mo of
age); for infants with data at 12 mo of age, CG infants were also
heavier with higher MUAC at baseline. By design, similar
proportions of infants from the 2 groups had started CF before 6
mo of age [19 infants (45%) in the NGG and 20 infants (51%) in
TABLE 2
Comparison of breastfeeding practices at 12 mo of age between the NGG and CG by using data from a 24-h recall1
Variable NGG (n = 38)
No. of breastfeeds/d 3.7 6 2.33
Time/breastfeed (min) 16.12 6 9.6
No. of breastfeeds/night 2.08 6 1.79
Breast milk intake (mL) 450.7 6 209.3
1
CG, control group; NGG, new guidelines group.
2
All values are means; 95% CIs in parentheses.
3
Mean 6 SD (all such values).
Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278
by guest
on 05 August 2018
987
the CG)] Surprisingly, 23.8% of infants in the NGG and 16.3%
of infants in the CG were already consuming cow milk at 6 mo
of age.
To prevent iron and vitamin A deficiency, the Colombian
government recommends iron supplementation [2 mg/kg weight
(38)] and vitamin A supplementation [100,000 UI (39)] at 6 and
12 mo of age. However, compliance with iron and vitamin A
supplements was very low; 4 infants (10.8%) in the NGG and 6
infants (15.8%) in the CG received a first dose of iron at 6 mo of
age; 3 infants (7.9%) in the CG and no infants (0%) in the NGG
received the second dose at 12 mo of age. Four infants (10.8%) in
the NGG and 7 infants (18.4%) in the CG received vitamin A
supplementation at 6 mo of age, and 2 infants (5.4%) in the NGG
and 4 infants (10.8%) in the CG received the second dose at
12 mo of age.
Compliance with, and acceptability of, the intervention
Twenty-eight mothers (74%) mothers in the NGG followed the
new recommendations completely during the whole period of the
intervention. Reasons given for not following the recommen-
dations completely were family pressure (17%), grandmother’s
opinion (50%) and mother’s own decision (33%). As concerns
the recommendations given in the NCG, 23 mothers (60.5%)
reported that they were very good, 14 mothers (36.8%) that they
were good, and one mother reported that they were not helpful.
The educational material given was considered understandable
and practical for all the mothers. At 12 mo, 31 NGG mothers
(83.8%) reported that they could afford to include red meat $3
times/wk, whereas 6 mothers found it too expensive to include
red meat, and 1 mother did not give a response. One-hundred
percent of mothers in the NGG reported that their infants tol-
erated meat and all kinds of red meat offered, and there were no
adverse effects (in terms of gastrointestinal symptoms such as
distress or vomiting) of consuming these foods during the whole
period of intervention.
Intake of recommended foods
There was no significant difference between groups in the
proportion of infants who consumed meat, red meat, or fruit at
baseline (Figure 2, Tables 2 and 3). At 12 mo of age, infants in
the NGG were significantly more likely to have daily con-
sumption of meat, red meat, fruit, and vegetables, whereas in-
fants in the CG were significantly more likely to have daily
consumption of follow-on formula and cow milk. The frequency
of red meat consumption per week was significantly greater in
the NGG over the whole intervention period. The mean number
of portions of meat, red meat, fruit, vegetables, and legumes
CG (n = 38) Difference2
3.9 6 1.95 20.18 (21.17, 0.82)
17.13 6 8.9 20.96 (25.4, 3.47)
2.35 6 1.7 20.26 (21.01, 0.56)
501.5 6 171.5 250.9 (2142.1, 40.4)
 988 OLAYA ET AL
TABLE 3
Comparison of feeding practices at 12 mo of age between the NGG and CG by using data from a food-frequency
questionnaire and 24-h recall1
Variable NGG (n = 38) CG (n = 38) P

Proportion of infants consuming selected food groups daily [n (%)]

Dairy products (yogurt, cheese) 29 (76.3) 32 (84.2) 0.39
Milk2 19 (50.0) 31 (83.8) 0.002
Meat 28 (73.7) 17 (45.9) 0.014
Red meat 37 (97.4) 25 (67.6) 0.001
Fruit 35 (92.1) 22 (59.5) 0.001
Vegetables 27 (71.1) 11 (29.7) 0.001
Proportion of infants consuming selected recommended food
groups at the recommended frequency [n (%)]3
Meat (all types) 28 (73.7) 17 (45.9) 0.01
Red meat 37 (97.4) 25 (67.1) 0.001
Vegetables 27 (71.1) 11 (29.7) 0.001
Fruit 35 (92.1) 22 (59.5) 0.001
Legumes 23 (69.5) 16 (43.2) 0.13
1
CG, control group; NGG, new guidelines group.
2
Milk refers to any other milk than breast milk (ie, follow-on formula and cow milk).
3
Recommended food groups and frequencies: meat, 5 times/wk; red meat, 3 times/wk; vegetables, 7 times/wk; fruit, 7
times/wk; and legumes, 3 times/wk.

consumed per week was also significantly higher in the NGG,
whereas infants in the CG consumed significantly more portions
per week of sweetened foods (defined as sugar, jelly, chocolate,
and sweets), milk, and cow milk (Figure 2). Significantly more
infants in the NGG consumed recommended food portions per
week at 12 mo of age, whereas infants in the CG had signifi-
cantly more bottle-feeds than those of infants in the NGG,
consumed cow milk more than 1 time/wk, and had ,2 main
meals/d There was no significant difference between groups in
the number of breastfeeds per day, time per breastfeed, and
number of breastfeeds per night.
Linear growth
The length-for-age z score (LAZ) at 6 and 12 mo of age and
the change in LAZ were not significantly different between
groups (Table 4). The number and proportion of infants with
a LAZ less than 22 SDs was significantly higher at 6 mo of age
in the NGG [10 infants (23.8%)] than in the CG [2 infants
(4.7%)] (P = 0.02). At 12 mo of age, there was no significant
difference between groups, but the number and proportion of
infants with a LAZ less than 22 SDs in the CG increased from 2
infants (4.7%) at 6 mo of age to 8 infants (21.1%) at 12 mo of
age, whereas the number and proportion in the NGG remained
similar at 12 mo of age to those at 6 mo of age. After adjustment
for baseline LAZ and socioeconomic factors, the change in LAZ
from 6 to 12 mo of age was not predicted by the randomly as-
signed group, and there was no interaction between the LAZ at 6
mo of age and the randomly assigned group on the change in
LAZ from 6 to 12 mo of age (P = 0.9).
Iron status
Values of hemoglobin, hematocrit, MCV, and SF at 6 and 12
mo of age and the change between 6 and 12 mo of age for the
NGG and CG are shown in Table 5. At 6 mo of age, there was no
significant difference in these indicators between groups. Hemo-
globin (P = 0.009) and hematocrit (P = 0.02) were significantly
higher in the NGG than CG at 12 mo of age. These differences
remained after adjustment for confounding factors such as sex,
birth weight ,3 or $3 kg, weight gain from 0 to 6 mo of age, and
weight at 6 mo of age. SF at 12 mo of age was not significantly
different between groups, and this finding was unchanged when 5
infants (2 infants in the NGG and 3 infants in the CG) with CRP
.6 were excluded from the analysis.
Changes in hemoglobin and hematocrit from 6 to 12 mo of age
were positive and were significantly higher in the NGG than CG
[mean 6 SD change in hemoglobin 0.41 6 0.8 g/dL compared
with 20.13 6 1.0 (P = 0.01) and hematocrit 1.04 6 2.2%
compared with 20.15 6 2.4) (P = 0.03)]. After adjustment for
socioeconomic factors, the change in hemoglobin from 6 to 12
mo of age was negatively predicted by the hemoglobin at 6 mo
of age (coefficient: 20.58; 95% CI: 20.82, 20.33; P # 0.001),
and there was a positive effect of the randomly assigned group
(coefficient: 0.56; 95% CI: 0.19, 0.92; P = 0.003). There was no
interaction between hemoglobin at 6 mo of age and the ran-
domly assigned group on the change of hemoglobin from 6 to 12
mo of age (P = 0.42). The change in hematocrit from 6 to 12 mo
of age was not related to baseline hematocrit after adjustment for
socioeconomic factors, and there was no interaction between
hematocrit at 6 mo of age and the randomly assigned group on
the change in hematocrit from 6 to 12 mo of age (P = 0.73).
SF decreased from 6 to 12 mo of age in both groups with no
significant difference between groups (227.23 6 44.8 mg/L in
the CG compared with 213.62 6 37.3 mg/L in the NGG; 95%
CI for difference in decline: 20.63, 33.5; P = 0.2). After ad-
justment for socioeconomic factors, baseline SF was not asso-
ciated with the change in SF from 6 to 12 mo of age, and there
was no interaction between SF at 6 mo of age and the randomly
assigned group on the change in SF from 6 to 12 mo of age (P =
0.64).
The proportion of infants who had a hemoglobin concentration
,11 g/dL and hematocrit ,33% at 12 mo of age was 4 infants
(11%) in the CG and 0 infants (0%) in the NGG (P = 0.1). The

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 EFFICACY OF NEW COMPLEMENTARY FEEDING GUIDELINES 989
proportion of infants with a SF concentration ,12 mg/L at 6 mo

0.18 (20.02, 0.39)

0.059 (20.26, 0.38)

Anthropometric data at 6 mo of age are shown for those infants who also had measurements at 12 mo of age. a,bStudent’s t test: aP , 0.05, bP , 0.01. CG, control group; HCZ, head circumference–for-age
0.09 (20.23, 0.42)
0.11 (20.09, 0.31)
0.13 (20–16, 0.43)
of age was significantly higher [5 infants (12.2%)] in the NGG]

Difference2
than in the CG (0 infants; P = 0.03) and increased at 12 mo of
age to 29% in the NGG and 37.1% in the CG (P = 0.9)

Zinc status
Serum zinc increased from 6 to 12 mo of age in both groups.
0.44
0.67
0.66

0.71
The mean increase was higher in the CG than in the NGG, but

0.3
CG (n = 38)

z score; LAZ, length-for-age z score; MUACZ, midupper arm circumference z score; NGG, new guidelines group; WAZ, weight-for-age z score; WLZ, weight-for-length z score.
there was no significant difference between groups [mean 6 SD
Change from 6 to 12 mo

6
6
6
6
6
change: 31.01 6 25.7 mg/dL in the CG compared with 20.33 6
20.006
20.19
20.27
20.26

0.28
37.3 mg/dL in the NGG (Table 5)]. The proportion of infants
with low zinc status by using a cutoff of ,65 mg/dL decreased
between 6 and 12 mo of age in both groups [36 7.7% in the
NGG (n = 38)

0.45
0.60
0.76
0.52
0.70

NGG at 6 mo of age compared with 1 6 2.9% in the CG; and 1

6 2.9% in the NGG compared with 0 6 0% in the CG at 12 mo
6
6
6
6
6
20.00
20.13
20.16
0.11
0.34

of age; NS].
Comparison of anthropometric measurements at 6 and 12 mo of age and the change from 6 to 12 mo of age between randomly assigned groups1

Other anthropometric variables

20.12)b

The WAZ was significantly higher at baseline in CG than in

0.00)a

0.00)a
0.39)

0.28)

NGG infants, and similar results were observed for the WLZ and
Difference2

MUACZ at both 6 and 12 mo of age. However, changes in the

(20.93,

(20.96,

(20.84,
(20.59,

(20.57,

WAZ, WLZ, head circumference–for-age z score, and MUACZ

from 6 to 12 mo of age were not significantly different between
20.46

–0.54

20.42
20.98

20.15

groups (Table 4). After adjustment for baseline differences, the

WAZ at 12 mo of age was strongly related to the WAZ at 6 mo
of age (coefficient: 0.946; 95% CI: 0.83, 1.057; P = 0.0001).
1.0
1.0
1.0
0.9
0.9
CG (n = 38)

When socioeconomic factors and sex were included in the

6
6
6
6
6

model with randomly assigned groups by using stepwise re-

20.12
21.02
0.49
0.33
0.97

gression models, the WAZ at 6 mo of age remained related to the

12 mo of age

WAZ at 12 mo of age, with no significant effect of the randomly

assigned group. Similar results were shown for the WLZ and
MUACZ at 12 mo of age.
0.91
NGG (n = 38)

0.9
1.1
0.9

0.9

After adjustment for socioeconomic factors, neither baseline

6
6
6
6
6

values nor randomly assigned group were significant predictors

20.59
21.12

0.18
0.55
20.043

of changes in the WAZ or WLZ from 6 to 12 mo of age, and there

were no significant interactions between baseline measurements
and randomly assigned groups. Similar results were shown for the
20.63 (21.1, 20.20)a

–0.47 (20.9, 20.06)a

change in MUACZ from 6 to 12 mo of age.

-0.64 (21.0, 20.2)a

20.26 (20.64, 0.12)

20.23 (20.7, 0.23)
Difference2

Meat intake and iron and zinc status

After adjustment for socioeconomic factors, the frequency
(number of times per week) of red meat consumption from 8 to 10
mo of age was positively correlated with hemoglobin (r = 0.25,
All values are means; 95% CIs in parentheses.

P = 0.03) and hematocrit (r = 0.27, P = 0.02) at 12 mo of age.

0.87

The same positive correlation for this period was observed with
1–0
CG (n = 38)

0.9
1.0

0.9

changes in hemoglobin, hematocrit, and SF. The frequency of

6
6
6
6
6

red meat consumption from 6 to 8 mo of age also showed

0.06
20.76
0.76
0.34
0.69

a positive correlation with the change in hemoglobin from 6 to

6 mo of age

Mean 6 SD (all such values).

12 mo of age (r = 0.24, P = 0.05; Table 6).

Infants with red meat consumption $3 times/wk from 6 to 8
NGG (n = 38)

0.84
0.80
0.93
1.1

0.8

mo of age had significantly higher MCV at 12 mo of age than did

infants with red meat consumption ,3 times/wk (72.5 6 4.4 fL)
6
6
6
6
6

compared with 69.6 6 3.7; P = 0.02 ]and a greater increase in

20.59
20.98
0.12
0.08
0.22

MCV from 6 to 12 mo of age (P , 0.05), whereas infants with

red meat consumption $3 times/wk from 10 to 12 mo of age had
TABLE 4

higher hemoglobin concentrations (P = 0.016) and hematocrit

MUACZ
Variable

2
3

(P = 0.03) at 12 mo of age and a significantly greater increase in

WLZ
WAZ

HCZ
LAZ

hemoglobin from 6 to 12 mo of age (P , 0.01).

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 990 OLAYA ET AL

There was no correlation between the mean frequency of red

(227.0, 5.7)
Student’s t- test: aP , 0.05, bP , 0.001. CG, control group; MCV, mean corpuscular volume; NGG, new
(20.42, 3.0)
(26.3, 33.5)
(0.11, 0.9)b
(0.12, 2.3)a
meat consumption per week and serum zinc at 12 mo of age or the

Difference2
change in serum zinc between 6 and 12 mo of age. However,
infants with red meat consumption ,3 times/wk from 6 to 8 and

210.7
0.5

13.6
1.2
1.3
10 to 12 mo of age had significantly greater increase in zinc
between 6 and 12 mo of age.

44.8
25.7
1.0
2.4
3.9
DISCUSSION
6
6
6
6
6
CG

20.13
20.15
20.20
227.2
31.0
Change from 6 to 12 mo

The new CF guidelines tested in this trial were designed to

provide clear, specific, and practical advice for mothers living in
poor socioeconomic conditions in Bogota. The guidelines were
37
37
36
30
30
n

developed following WHO 2003 recommendations (12) but also

33.8
37.3
by taking into account the findings from our earlier studies that
0.8
2.2
3.3

described current CF practices in Bogota so that the guidelines

NGG

6
6
6
6
6

would be acceptable in terms of culture and food availability. The

0.41
1.04
1.07
213.6
20.3

3 main messages, emphasized at each study visit, both verbally

and in writing, emphasized the importance of 1) continuing
36
36
36
33
32

breastfeeding alongside CF, 2) red meat as a source of iron and

n

zinc, and 3) fruit and vegetables. Our preliminary results con-

(217.9, 9.4)
(22.7, 15.4)
(21.08, 3.0)
(0.13, 0.9)b
(0.19, 2.1)a

firmed the feasibility of the intervention; the new guidelines

Comparison of iron and zinc status at 6 and 12 mo of age and the change from 6 to 12 mo of age between randomly assigned groups1

Difference2

were acceptable and affordable by the majority of mothers,

despite living in conditions of socioeconomic constraint. Effi-
1.13

cacy was also shown with improved consumption of red meat,

0.52

24.23
0.97
6.31

fruit, and vegetables. The transition to CF was successfully at-

tained by infants in the NGG who, compared with infants in the
14.7
23.1
0.9
2.1
4.8

CG, had a better meal pattern and feeding practices including

a lower consumption of cow milk and formula with less bottle-
CG

6
6
6
6
6
11.9
35.5
71.4

114.9
17.3

feeds and consumption of sweetened foods. The intervention

had no adverse effect on breastfeeding, which was consistent
12 mo of age

with results from cluster randomized studies of interventions to

37
37
37
35
37
n

a,b

improve CF practices in Malawi (21) and China (24).

22.3
34.0
Data at 6 mo of age are shown for infants who also had measurements at 12 mo of age.

The intervention was associated with evidence of improved

0.7
1.9
3.9
NGG

iron status, with higher hemoglobin and hematocrit and positive

6
6
6
6
6

changes in hemoglobin and hematocrit from 6 to 12 mo of age.

110.6
12.5
36.6
72.4
23.7

Although there was no effect of the intervention on SF, the

frequency of red meat consumption from 8 to 10 mo of age was
36
36
36
34
34
n

positively associated with the change in SF from 6 to 12 mo of

20.06 (21.66, 1.54)
24.69 (221.8, 12.4)
0.003 (20.34, 0.34)
0.03 (20.89, 0.95)

age. We could not directly compare our findings with those from
3.65 (28.8, 16.0)

other studies which have used supplementation or interventions

Difference2

with meat or a combination of nutrition education with recom-

mendations to improve CF. We gave specific advice on the type,
amount, and frequencies of foods to be given to the infant to-
gether with education on the importance of certain foods;
however, foods were not provided because we considered that this
All values are means; 95% CIs in parentheses.
CG (n = 38)

40.5
19.3

approach would have a greater likelihood of a sustained impact

0.8
2.1
3.1

on nutritional intakes and dietary habits. Previous randomized

6
6
6
6
6
12.1
35.6
71.4

controlled trials (RCTs) (28, 29, 40) that investigated meat or red
43.4
88.7
6 mo of age

meat consumption were conducted in developed countries, and

meat was provided during the period of intervention. One study
Mean 6 SD (all such values).
NGG (n = 38)

30.3
29.7
0.73
1.9
3.8

reported an effect of higher meat intake on hemoglobin but not on

6
6
6
6
6

SF, which was consistent with our findings, whereas a study in

12.1
35.7
71.4
38.7
92.4

German infants showed that the consumption of commercial

weaning foods with a low meat content increased risk of de-
Hemoglobin (g/dL)

veloping marginal iron status at 10 mo of age in infants who were

guidelines group.

exclusively breastfed for 4–6 compared with the consumption of

Hematocrit (%)

Ferritin (mg/L)
Variable

Zinc (mg/dL)

weaning foods with a higher meat intake (40). A longitudinal

MCV (fL)
TABLE 5

cohort study in the United Kingdom (41) also reported that

1

2
3

a higher meat intake during the first year was associated with
a higher hemoglobin concentration. Some cluster randomized

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 EFFICACY OF NEW COMPLEMENTARY FEEDING GUIDELINES 991
TABLE 6
Unadjusted correlations between frequency (mean no. of times/wk) of red meat consumption and change in iron status from 6 to 12 mo of age1
Change in hemoglobin Change in hematocrit Change in MCV from Change in SF from
Variable from 6 to 12 mo of age from 6 to 12 mo of age 6 to 12 mo of age 6 to 12 mo of age

Frequency of red meat/wk

from 6 to 8 mo of age
n 70 70 69 61
r 0.24 0.16 0.12 0.07
P 0.05 0.20 0.34 0.56
Frequency of red meat/wk
from 8 to 10 mo of age
n 73 73 72 63
r 0.32 0.34 0.11 0.30
P 0.006 0.003 0.34 0.02
Frequency of red meat/wk
from 10 to 12 of age
n 72 72 71 62
r 0.21 0.18 0.063 0.14
P 0.07 0.13 0.60 0.29
Frequency of red meat/wk
from 6 to 12 mo of age
n 73 73 72 63
r 0.307 0.27 0.11 0.15
P 0.008 0.02 0.345 0.23
1
r, Pearson’s correlation coefficient for changes in hemoglobin, hematocrit, and MCV and Spearman’s correlation for changes in SF. MCV, mean
corpuscular volume; SF, serum ferritin.

trials that used nutrition education combined with recommen-
dations to improve CF (21–27) have reported significantly in-
creased meat consumption but did not focus on the effect of
meat consumption on iron and zinc status.
Other strategies such as iron supplementation and sprinkles
have been used to improve iron status in infants with conflicting
results, and RCTs that used sprinkles were generally testing the
treatment rather than primary prevention (20, 42). In a blinded
RCT in Honduras and Sweden, iron supplementation from 4 to 9
mo of age resulted in a significant increase in hemoglobin and
reduced iron deficiency anemia in Honduras but not in Sweden
(16), where impaired linear growth was observed (17). A study in
Indian infants also reported a negative effect of iron supple-
mentation on growth (43). In addition, compliance with iron
supplementation is often very low (15, 44).
We showed no significant effect of the intervention on zinc
status at 12 mo of age or a change in zinc status from 6 to 12 mo of
age. Serum zinc concentrations increased in both groups and
were not correlated with the frequency of red meat consumption
per week, which were results similar those in a randomized trial
of meat compared with micronutrient fortified cereals in US
infants (28). However, we showed that those infants who ate red
meat ,3 times/wk had greater and more-positive changes in
zinc status compared with those who ate red meat more fre-
quently, which suggested potential negative effects of increased
red meat consumption on zinc status. It is plausible that iron in
red meat could interfere with zinc absorption. However, there is
a need for additional research in this area to examine in-
teractions between iron and zinc intakes during CF.
The intervention had no significant effect on anthropometric
measures at 12 mo of age or on changes in z scores between 6 and
12 mo of age. A greater proportion of infants in the NGG had
linear stunting at 6 mo of age, but because this proportion did
not increase by 12 mo of age in the NGG, whereas it increased
in the CG, suggested a possible protective effect of the in-
tervention. Our findings were in agreement with those from 3
RCTs that investigated meat as a CF food, albeit by using dif-
ferent indicators (28, 29, 40), and with those from a cluster
randomized study in Pelotas, Brazil (26). Conversely, studies in
Pakistan (25), China (23), and Peru (22) reported positive effects
of interventions on growth.
Although the approach used to deliver the messages recom-
mended in the new CF guidelines was feasible and successful in
the setting of this study, the implications in terms of staffing and
resources need to be evaluated before the guidelines can be more
widely used. The messages were delivered by researchers, and
each session lasted w45 min, which was longer than the time
currently allocated for routine clinic visits. It would also be
important to examine in more detail which parts of the guide-
lines and their delivery were most important for success because
this might allow them to be implemented in a more-efficient
way. This issue was not examined in this study, but several of the
following possible factors may have contributed to the effective
uptake of the guidelines: 1) the guidelines were designed on the
basis of local culture, food availability, environment, and infant
characteristics; 2) individual counseling sessions allowed edu-
cational messages to be adapted to individual needs and un-
derstandings; 3) the 3 main messages were repeated and reinforced
by written materials; and 4) emphasis was given to the importance
of the recommendations for the infant’s health. Our observation
that grandmothers played an important role in the feeding de-
cisions of mother suggested that it might be worth including this
group in any future intervention.
The main strength of the study was the randomized design,
which allowed for causal relations between the intervention and
outcomes to be established. Compliance with the protocol was
good, and there was low attrition. The main limitation of the study
was the relatively small sample size, which limited the power for

Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278

by guest
on 05 August 2018
 992 OLAYA ET AL
some analyses. With 38 subjects per group, we had 80% power to
detect a difference of a 0.69 SD at 5% significance and could have
missed a smaller, although perhaps clinically relevant, effect on
outcomes. The small sample size also resulted in a baseline
imbalance in anthropometric measures between groups, which
we addressed by comparing differences in outcome variables
adjusted for the baseline value and testing for interactions be-
tween the baseline value and randomly assigned group. We chose
not to impute missing outcome data, and this may have in-
troduced some bias, although when we compared baseline
characteristics by group, the only difference was baseline dif-
ferences in WAZ, WLZ, and MUACZ were significant for all
infants randomly assigned but not for those with follow-up data.
We made no formal statistical correction for the multiple com-
parisons made during the study, and the possibility that this lack
resulted in a type II error with significant differences apparent
due to chance should be considered when interpreting the
findings. Also, the study was also not blinded, which may have
influenced certain outcomes such as dietary intakes and an-
thropometric measures, although this was unlikely because the
laboratory measurements as blood sampling and analyses were
performed by staff with no connection to the study or knowledge
of the protocol. Finally, infants were not followed beyond the
period of intervention, and thus, we could not investigate the
sustainability of the observed effects.
In conclusion, these preliminary data suggest that the use of CF
guidelines emphasizing red meat consumption might be a prac-
tical strategy for improving both CF practices and iron status,
especially in countries such as Colombia where many strategies
aimed at improving iron status have already been tried without
great success. Theoretically, red meat may be a good first
complementary food for all breastfed infants, regardless of their
environment, although this strategy would not be suitable in all
settings. A larger trial would be needed to evaluate the new CF
guidelines in other settings to test the generalizability and po-
tential for adapting the guidelines according to local circum-
stances. Such a trial should have a sufficient sample size to allow
subgroup analyses according to sex and mode of feeding before 6
mo of age to detect smaller differences in outcome measures
between groups and to further investigate associations between
meat intake and measures of iron status. Ideally, a longer follow-
up should be incorporated to examine whether dietary effects of
the intervention are maintained and associated with longer-term
effects on outcomes such as growth and developmental outcomes.
We thank the mothers and infants who participated in the study and the staff
in Fontibon and Suba Hospitals who helped during the period of data collec-
tion. We also thank Angela Umaña (Research Office at the Pontificia Uni-
versidad Javeriana, Bogota, Colombia), Ingrid Schuller (Faculty of Sciences,
Pontificia Universidad Javeriana), and Wilson Mejia for supporting the sub-
mission of the project at the Pontificia Universidad Javeriana; Maria del
Pillar Rodriguez (Nutrition Department, Suba Hospital) and Cristina Mendez
(Suba Hospital) for helping with study preparations in the hospitals; and
Luisa Fernanda Tovar (Nutrition and Biochemistry Department) for her
contribution with taking pictures of CFs used in the brochures and all edu-
cational material developed for the study, We also thank JAVEGRAF (Pon-
tificia Universidad Javeriana) for their contribution in the designing and
printing of the educational material, Maria Fernanda Ramirez and Luz Ma-
rina Fernandez of the laboratory in the Suba Hospital for undertaking the
biochemical analyses, and Deisy Fonseca for her help in data collection.
The authors’ responsibilities were as follows—GAO: study concept and
design, design of new CF guidelines, conduction of the study, statistical
Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278
by guest
on 05 August 2018
analyses, manuscript writing, and critical reading of the manuscript; and
ML and MSF: design of new CF guidelines and trial, data interpretation,
and critical reading and revision of manuscript. None of the authors declared
a conflict of interest following the guidelines of the International Committee
of Medical Journal Editors.
REFERENCES
1. WHO. The optimal duration of exclusive breastfeeding. Report of an
Expert Consultation Group. 2001. Available from: http://www.who.int/
nutrition/publications/optimal_duration_of_exc_bfeeding_report_eng.
pdf (cited 20 January 2012).
2. WHO. Global strategy for infant and young child feeding. 2003.
Available from: http://whqlibdocwhoint/publications/2003/9241562218pdf
(cited 20 January 2012).
3. Lozoff B. Early iron deficiency has brain and behavior effects con-
sistent with dopaminergic dysfunction. J Nutr 2011;141:740S–6S.
4. Madan N, Rusia U, Sikka M, Sharma S, Shankar N. Developmental
and neurophysiologic deficits in iron deficiency in children. Indian J
Pediatr 2011;78:58–64.
5. Krebs NF, Reidinger CJ, Robertson AD, Hambidge KM. Growth and
intakes of energy and zinc in infants fed human-milk. J Pediatr 1994;
124:32–9.
6. Gibson RS, Ferguson EL, Lehrfeld J. Complementary foods for infant
feeding in developing countries: their nutrient adequacy and im-
provement. Eur J Clin Nutr 1998;52:764–70.
7. Brown KH, Wuehler SE, Peerson JM. The importance of zinc in human
nutrition and estimation of the global prevalence of zinc deficiency.
Food Nutr Bull 2001;22:113–25.
8. Salgueiro MJZ, Lysionek MB, Caro AE, Weill RA, Boccio R. J. R. The
role of zinc in the growth and development of children. Nutrition 2002;
18:510–9.
9. FAO. Vitamin and mineral requirements in human nutrition. 2001.
Available from: http://wwwfaoorg/docrep/007/y5686e/y5686e00htm
(cited 20 April 2012),
10. WHO. Complementary feeding: family foods for breastfed children.
2000. WHO reference no. WHO/NHD/00.1, WHO/FCH/CAH/00.6.
Available from: http://wwwwhoint/nutrition/publications/infantfeed-
ing/WHO_NHD_001/en/index (cited 22 January 2012).
11. WHO. Breastfeeding in the “global strategy for infant and young child
feeding”. Anales de Pediatria 2003;58:208–10.
12. Dewey KG, Brown KH. Update on technical issues concerning com-
plementary feeding of young children in developing countries and
implications for intervention programs. Food Nutr Bull 2003;24:5–28.
13. WHO. Guidelines for the use of iron supplements to prevent and treat
iron deficiency anemia. 1998. Available from: http://wwwwhoint/nutri-
tion/publications/micronutrients/anemia_iron_deficiency/ (cited 2 April
2012).
14. Iannotti LL, Tielsch JM, Black MM, Black RE. Iron supplementation
in early childhood: health benefits and risks. Am J Clin Nutr 2006;84:
1261–76.
15. Amsel S, Boaz M, Ballin A, Filk D. Low compliance of iron supple-
mentation in infancy and relation to socioeconomic status in Israel.
Pediatrics 2002;110:410–1.
16. Domellöf M, Lonnerdal B, Abrams SA, Hernell O. Iron absorption in
breast-fed infants: effects of age, iron status, iron supplements, and
complementary foods. Am J Clin Nutr 2002;76:198–204.
17. Dewey KG, Domellof M, Cohen RJ, Rivera LL, Hernell O, Lonnerdal
B. Iron supplementation affects growth and morbidity of breast-fed
infants: results of a randomized trial in Sweden and Honduras. J Nutr
2002;132:3249–55.
18. Lutter CK, Dewey KG. Proposed nutrient composition for fortified
complementary foods. J Nutr 2003;133:3011S–20S.
19. Domellöf M, Lind T, Lonnerdal B, Persson LA, Dewey KG, Hernell O.
Effects of mode of oral iron administration on serum ferritin and
haemoglobin in infants. Acta Paediatr 2008;97:1055–60.
20. Zlotkin S, Antwi KY, Schauer C, Yeung G. Use of microencapsulated
iron(II) fumarate sprinkles to prevent recurrence of anaemia in infants
and young children at high risk. Bull World Health Organ 2003;81:
108–15.
21. Hotz C, Gibson RS. Participatory nutrition education and adoption of
new feeding practices are associated with improved adequacy of
complementary diets among rural Malawian children: a pilot study. Eur
J Clin Nutr 2005;59:226–37.
 EFFICACY OF NEW COMPLEMENTARY FEEDING GUIDELINES
22. Penny ME, Creed-Kanashiro HM, Robert RC, Narro MR, Caulfield
LE, Black RE. Effectiveness of an educational intervention delivered
through the health services to improve nutrition in young children:
a cluster-randomised controlled trial. Lancet 2005;365:1863–72.
23. Guldan GS, Fan HC, Ma X, Ni ZZ, Xiang X, Tang MZ. Culturally
appropriate nutrition education improves infant feeding and growth in
rural Sichuan, China. J Nutr 2000;130:1204–11.
24. Shi L, Zhang J, Wang Y, Caulfield LE, Guyer B. Effectiveness of an
educational intervention on complementary feeding practices and
growth in rural China: a cluster randomised controlled trial. Public
Health Nutr 2010;13:556–65.
25. Zaman S, Ashraf RN, Martines J. Training in complementary feeding
counselling of healthcare workers and its influence on maternal be-
haviours and child growth: A cluster-randomized controlled trial in
Lahore, Pakistan. J Health Popul Nutr 2008;26:210–22.
26. Santos I, Victora CG, Martines J, Gonçalves H, Gigante DP, Valle NJ,
Pelto G. Nutrition counselling increases weight gain among Brazilian
children. J Nutr 2001;131:2866–73.
27. Bhandari N, Mazumder S, Bahl R, Martines J, Black RE, Bhan MK;
Infant Feeding Study Group. An educational intervention to promote
appropriate complementary feeding practices and physical growth in in-
fants and young children in rural Haryana, India. J Nutr 2004;134:2342–8.
28. Krebs NFWJBNRCBMHM. Meat as first complementary food for
breastfed infants: feasibility and impact on zinc intake and status.
J Pediatr Gastroenterol Nutr 2006;42:207–14.
29. Engelmann MDM, Sandstrom B, Michaelsen KF. Meat intake and iron
status in late infancy: an intervention study. J Pediatr Gastroenterol
Nutr 1998;26:26–33.
30. Morgan J, Taylor A, Fewtrell T. Meat consumption is positively as-
sociated with psychomotor outcome in children up to 24 months of age.
J Pediatr Gastroenterol Nutr 2004;39:493–8.
31. Republica de Colombia MdPS, Instituto Colombiano de Bienestar
Familiar. Encuesta Nacional de la Situacion Alimentaria y Nutricional
(ENSIN) [Nutritional National Survey 2010.] 2011;84–91, 146–9,
216–24. (in Spanish). Available from: http://www.icbf.gov.co/portal/
page/portal/PortalICBF/NormatividadGestion/ENSIN1/ENSIN2010.
32. WHO. WHO Anthro (version 3.2.2, January 2011). Available from:
http://wwwwhoint/childgrowth/software/en/.
33. de Benoist B, Darnton-Hill I, Davidsson L, Fontaine O. Report of
a WHO/UNICEF/IAEA/IZiNCG interagency meeting on zinc status
indicators, held in IAEA headquarters, Vienna, December 9, 2005 -
Foreword. Food Nutr Bull 2007;28:S401–2.
Downloaded from https://academic.oup.com/ajcn/article-abstract/98/4/983/4577278
by guest
on 05 August 2018
993
34. WHO/UNICEF/UNU. iron deficiency anaemia: assessment, prevention
and control, a guide for programme managers.2001. Available from:
http://wwwwhoint/nutr/publications/micronutrients/anaemia_iron_
deficiency/WHO_NHD_013/en/indexhtml (cited 20 April 2012).
35. Male C, Persson LA, Freeman V, Guerra A, van’t Hof MA, Haschke F,
and the Euro-Growth Iron Study. Prevalence of iron deficiency in 12-
mo-old infants from 11 European areas and influence of dietary
factors on iron status (Euro-Growth study). Acta Paediatr 2001;90:
492–8.
36. WHO/CDC. Assessing the iron status of population: report of a joint
WHO/CDC technical consultation on the assessment of iron status at
the population level. Geneva Switzerland: WHO, 2004.
37. Olaya G, Borrero M. Propuesta para la formulación de pautas para la
alimentación complementaria del niño lactante de 6 a 12 meses.
[Guideline proposal of complementary feeding for infants 6-12 months
of age.] Perspectivas en Nutricion Humana. 2009;11:139–51 (in
Spanish).
38. Secretaria de Salud, Bogota, Colombia. Normas para la suplementa-
ción con hierro. SDS Resolución 657 de 1998. [Guide for iron sup-
plementation SDS. Resolution 657/ 1998.] Bogota, Colombia:
Secretaria de Salud Press, 1998 (in Spanish).
39. Secretaria de Salud, Bogota, Colombia. Documento Técnico para la
implementación de la suplementación con vitamina A. [Technical re-
port on vitamin A supplementation.] Bogota, Colombia: Secretaria de
Salud Press, 2003 (in Spanish).
40. Dube K, Schwartz J, Mueller MJ, Kalhoff H, Kersting M. Comple-
mentary food with low (8%) or high (12%) meat content as source of
dietary iron: a double-blinded randomized controlled trial. Eur J Nutr
2010;49:11–8.
41. Taylor A, Redworth EW, Morgan JB. Influence of diet on iron, copper,
and zinc status in children under 24 months of age. Biol Trace Elem
Res 2004;97:197–214.
42. Zlotkin S, Arthur P, Antwi KY, Yeung G. Treatment of anemia with
microencapsulated ferrous fumarate plus ascorbic acid supplied as
sprinkles to complementary (weaning) foods. Am J Clin Nutr 2001;74:
791–5.
43. Dewey KG, Chaparro CM. Iron status of breast-fed infants. Proc Nutr
Soc 2007;66:412–22.
44. Fox MK, Devaney B, Reidy K, Razafindrakoto C, Ziegler P. Re-
lationship between portion size and energy intake among infants and
toddlers: Evidence of self-regulation. J Am Diet Assoc 2006;106:
S77–83.