Академический Документы
Профессиональный Документы
Культура Документы
aUnite
INSERM 593, Institut de Sante Publique Epidemiologie Developpement, Universite Victor Segalen, Bordeaux, France; bProjet ANRS 1201/1202 Ditrame Plus,
Programme PAC-CI, Centre Hospitalier Universitaire de Treichville, Abidjan, Cote dIvoire; cUnite de Surveillance et dEpidemiologie Nutritionnelles, Institut de Veille
Sanitaire, Conservatoire National des Arts et Metiers, Paris, France; dService de pediatrie, Centre Hospitalier Universitaire de Yopougon, Abidjan, Cote dIvoire
The authors have indicated they have no nancial relationships relevant to this article to disclose.
ABSTRACT
OBJECTIVE. In high HIV prevalence resource constrained settings, exclusive breast-
feeding with early cessation is one of the conceivable interventions aimed at the
www.pediatrics.org/cgi/doi/10.1542/
prevention of HIV through breast milk. Nevertheless, this intervention has poten- peds.2005-1911
tial adverse effects, such as the inappropriateness of complementary feeding to doi:10.1542/peds.2005-1911
take over breast milk. The purpose of our study first was to describe the nature and
Key Words
the ages of introduction of complementary feeding among early weaned breastfed breastfeeding, HIV, infant nutrition,
infants up to their first birthday and second was to assess the nutritional adequacy nutritional status
of these complementary foods by creating a child feeding index and to investigate Abbreviations
WHOWorld Health Organization
its association with child nutritional status. ANRSAgence Nationale de Recherches
sur le Sida
METHODS. A prospective cohort study in Abidjan, Cote dIvoire, was conducted in CI condence interval
HIV-infected pregnant women who were willing to breastfeed and had received a RRrelative risk
perinatal antiretroviral prophylaxis. They were requested to practice exclusive Accepted for publication Oct 13, 2005
breastfeeding and initiate early cessation of breastfeeding from the fourth month Address correspondence to Renaud Becquet,
PhD, Unite INSERM 593, Institut de Sante
to reduce breast milk HIV transmission. Nature and ages of introductory comple- Publique Epidemiologie et Developpement
mentary feeding were described in infants up to their first birthday by longitudinal (ISPED), Universite Victor Segalen Bordeaux 2,
146 rue Leo Saignat, 33076 Bordeaux Cedex,
compilation of 24-hour and 7-day recall histories. These recalls were done weekly France. E-mail: Renaud.Becquet@isped.u-
until 6 weeks of age, monthly until 9 months of age, and then quarterly. We bordeaux2.fr
PEDIATRICS (ISSN Numbers: Print, 0031-4005;
created an index to synthesize the nutritional adequacy of infant feeding practices Online, 1098-4275). Copyright 2006 by the
(in terms of quality of the source of milk, dietary diversity, food, and meal American Academy of Pediatrics
frequencies) ranging from 0 to 12. The association of this feeding index with
growth outcomes in children was investigated.
RESULTS. Among the 262 breastfed children included, complete cessation of breast-
feeding occurred in 77% by their first birthday, with a median duration of 4
months. Most of the complementary foods were introduced within the seventh
e702 BECQUET, et al
Downloaded from by guest on April 10, 2017
chosocial follow-up of both mothers and infants. Moth- cereal-based infant food enriched with powdered animal
erinfant pairs were seen on study sites at birth, 2 days milk. The term weaning food was used for all solid foods
after delivery, weekly until 6 weeks of age, monthly and/or any breast milk substitutes (eg, infant formula).
until 9 months of age, and every 3 months until the
second birthday. Services that were dispensed by the Child Feeding Index
study team were also available whenever needed be- To assess the nutritional adequacy of complementary
tween scheduled visits. All transport costs were reim- feeding, we created an index to synthesize multiple di-
bursed, and all care expenses related to any clinical mensions of child feeding practices on the basis of both
event were supported entirely by the project. current infant feeding recommendations2,10 and previous
Nutritionists counseled individually the women on work on the subject.23 This child feeding index was adapted
study sites about infant feeding practices whenever to the context of the Ditrame Plus study, in which women
needed. Collective sessions were organized to help were encouraged to breastfeed exclusively during 4
mothers to position their infant correctly to the breast, to months, then replace breast milk with formula feeding
reiterate the benefits of exclusive breastfeeding, how to until 9 months of age. From weaning initiation, women
prepare artificial feeding safely, to initiate weaning, to also were encouraged to provide milk sources to their
use appropriate complementary feeding, or to cook the infant through infant food enriched with powdered animal
infant food. At each scheduled visit, anthropometric milk and through dairy products. The scoring system that
measurements including height and weight were taken was used to create the child feeding index at ages 6, 9, and
by trained staff according to standard procedures.18 12 months is detailed in Table 1. The more positive the
nutritional practices were, the higher scores assigned were.
Collection of Infant Feeding Practices This index was a summation of 4 subscores that are de-
At each scheduled visit, infant feeding practices were tailed below and ranged from 0 to 12.
recorded via structured questionnaires by trained social A source of milk score was created on the basis of the
workers who were not involved in nutritional counsel- foods that contained milk and were consumed by the
ing. Women were asked whether their child had been child in the previous 24 hours. Nutritionally speaking,
given breast milk, artificial milk, or both since the last breastfeeding was the best practice, but breastfeeding
visit. Fluids and foods other than breast milk or artificial beyond 6 months of age is associated with an increased
milk were also documented using a 24-hour and a 7-day risk for postnatal transmission of HIV, which needs to be
recall history. Social workers went over a detailed list of taken into account in the appropriateness of this prac-
commonly used fluids or foods. Women were asked tice. A score of 1 therefore was assigned to breastfed
whether these fluids, foods, or some other items not children. In our context, in which nonbreastfeeding car-
listed had been given in the previous 7 days and, if so, ried a much lower risk for HIV transmission, we decided
how many times on the day before (24-hour recall his- to assign the same positive score to formula-fed children
tory) and how frequently in the past 7 days. but only when the women reported to have prepared
Infants were classified at each scheduled visit as ex- the correct amounts of feeds. Cereal-based infant foods
clusively or predominantly breastfed, mixed fed, or ar- that were enriched with powdered animal milk and
tificial fed using these recall histories.19 Being exclusively dairy products were considered as substantial sources of
breastfed from birth at a given time meant having been milk and therefore assigned positive scores.
classified in this category at all of the preceding visits A dietary diversity score was created on the basis of
since birth. We used the following WHO definitions to the number of food groups consumed by the child in the
allow a better comparability of results between studies. previous 24 hours. Emphasis was placed on animal prod-
Exclusive breastfeeding means giving a child no other ucts such as meat, fish, and eggs on the one hand and on
food or drink, including water, in addition to breastfeed- products that contained animal milk (dairy products and
ing with the exception of medicines, vitamin drops or infant food) on the other hand and constituted 2 food
syrups, and mineral supplements.20 Predominant breast- groups. Vegetables and fruits, which are important
feeding means breastfeeding a child but also giving small sources of vitamins and are rich in dietary fiber, and
amounts of water or water-based drinks. Neither food- tubers, grain, and starchy foods, which are staples of the
based fluid nor solid food is allowed under this defini- diet in this setting, constituted another 2 food groups.
tion.20 Artificial feeding means feeding a child on artifi- Considering that all of these food groups were essential
cial feeds (including infant formula and powdered to ensure a high dietary diversity, a score of 1 was
animal milk) and not breastfeeding at all.21 Mixed feed- assigned to each of them.
ing means breastfeeding while giving nonhuman milk A food frequency score was based on the number of
such as infant formula or food-based fluid or solid food.22 days the children consumed each of these food groups in
We defined the weaning process as the period from the previous week. The scoring depends on the age of
the introduction of the first weaning food until complete the child and is detailed in Table 1.
cessation of breastfeeding. We defined infant food as A meal frequency score was based on the number of
c Sum of subscores 1 to 4.
meals (complementary foods) in the previous 24 hours. for-age, and weight-for-length z scores were calculated
A maximum score of 2 was given to children who had on the basis of the gender- and age-specific growth chart
received complementary feeding at least twice a day at references that were developed by the National Center
age 6 months and at least 3 times a day at ages 9 and 12 for Health Statistics and the Centers for Disease Control
months. and Prevention and recommended for international use
by WHO.2426 The z score or SD unit is defined as the
Statistical Analysis difference between the value for an individual and the
The following analyses were conducted among women median value of the reference population for the same
whose live-born infant initially was classified as breast- age or height, divided by the SD of the reference popu-
fed using the recall history that was obtained at the day lation. The mean z scores were presented at ages 9, 12,
2 visit. The probability of being breastfed was calculated and 18 months and compared between children with a
from birth until 1 year of age, using the Kaplan-Meier low versus an average or high index at age 6 months and
method. We also detailed the proportion of children who with a low or average versus a high index at ages 9 and
were in each feeding category at given ages. 12 months.
The proportion of children who were ever fed each The cumulative probability of being stunted (defined
food item from birth up to 12 months of age and the as height-for-age z scores less than 2 SD) at least once
median ages of introduction of these food items were from age 7 months to 18 months was compared between
calculated. For each food item and for each monthly or children with a low versus an average or high index at 6
quarterly visit from birth up to 12 months of age, we months, using the Kaplan-Meier technique.27,28 Multi-
reported the proportion of children who had been given variate analysis used Coxs proportional hazard models.
this food item at least once in the previous week. This approach allowed for adjustment of this comparison
The mean and median values of the child feeding on potential confounding variables: maternal education,
index were calculated at ages 6, 9, and 12 months. At type of housing, low birth weight (2500 g), and pedi-
each of these ages, the index was grouped into tertiles to atric HIV status (time-dependent variable). All statistical
form 3 categories of child feeding practices (low, aver- analyses were conducted with the use of SAS software
age, or high) to assess the nutritional adequacy of com- (version 8.2; SAS Institute, Inc, Cary, NC).
plementary feeding.
The relationship between these 3 nutritional catego- Ethical Permissions
ries and long-term growth outcomes in children was also The ANRS 1201/1202 Ditrame Plus study was granted
investigated. For this purpose, weight-for-age, height- ethical permission in Cote dIvoire from the ethical com-
e704 BECQUET, et al
Downloaded from by guest on April 10, 2017
mittee of the National AIDS Control Program and in age, 83% of the mixed-fed children were in the process
France from the institutional review board of the French of being weaned and therefore were receiving both in-
ANRS. As part of the Ditrame Plus program, the study fant formula and breast milk. From 6 months of age,
presented here was included in the institutional review most of the infants were not breastfed any more and
board approval. received artificial feeding instead, whereas the remain-
ing breastfed infants essentially were mixed fed.
RESULTS Within our cohort, the cumulative probabilities of
being exclusively breastfed from birth were 0.18 (95%
Baseline Study Population Characteristics
confidence interval [CI]: 0.13 0.22), 0.10 (95% CI:
Among the 557 mothers who were included in the Dit-
0.06 0.13), and 0.01 (95% CI: 0 0.02) at ages 1, 3, and
rame Plus study and delivered a live birth, 262 (47%)
6 months, respectively. As detailed on Fig 2, this low
initiated breastfeeding and constituted the breastfeeding
prevalence of exclusive breastfeeding could be explained
group for the present analysis. Overall, 47% of these
by early common introduction of fluids such as water
breastfeeding women were illiterate, 70% lived with
(essentially tap water, but use of mineral water was also
their partner, all but 8 had electricity at home, and all
relatively common early in life). Indeed, 98% of the
had at least access to tap water in their yard. Three
infants had ever been given water from a median of 8
quarters of them lived in a typical shared housing with
days of age. Other fluids, such as herbal tea or fruit juice,
several houses organized around a yard, where inhabit-
were widely used but introduced later, ie, a median 12
ants live in crowded accommodation and share kitchen
weeks and 5 months after birth, respectively.
and restroom.
FIGURE 1
Kaplan-Meier probability of being breastfed (N
262).
83% and 74% of the children had received these food mean values of the source of milk and meal frequency
items by their first birthday. scores were satisfactory (1.63 of 2 and 1.41 of 2, respec-
tively), whereas the dietary diversity and food frequency
Child Feeding Index and Its Relation to Child Growth scores were low (1 of 4), leading to a relatively low
The values of the child feeding index scores at ages 6, 9, child feeding index score with a mean of approximately
and 12 months are detailed in Table 2. At all ages, all of 5 of 12. At 9 and 12 months of age, the dietary diversity
the 4 subscore values ranged from 0 to the maximum and food frequency were more adequate (2.5 of 4),
possible value, namely 2 or 4. At age 6 months, the resulting in a considerably improved child feeding index.
e706 BECQUET, et al
Downloaded from by guest on April 10, 2017
No statistically significant associations were found at any complementary feeding and detail the adequacy of these
of the 3 ages between maternal sociodemographic char- complementary foods and its implications on nutritional
acteristics and the child feeding index categories (low, status among early weaned breastfed children who were
average, or high). born to HIV-infected mothers in an urban African con-
As shown in Table 3, a low compared with an average text. Within the Ditrame Plus study, complete breast-
or high child feeding index score at age 6 months was feeding cessation was obtained at approximately the
associated with a significantly lower mean height-for- fourth month of age, after a short (median: 9 days)
age z score at ages 12 and 18 months and a lower mean transition period of mixed feeding when breast milk and
weight-for-age z score at ages 9, 12, and 18 months. No infant formula were given simultaneously to the in-
statistically significant associations were found between fant.13 After this weaning process, breastfeeding was re-
the values of the child feeding index at ages 9 and 12 placed by infant formula and infant food that was en-
months and the z score values in the subsequent riched with powdered animal milk, therefore covering
months. Very similar results were obtained when HIV- the nutritional requirements in terms of source of milk.
infected children were excluded (data not shown). However, the dietary diversity was not appropriate in
Given the relatively small number of HIV-infected chil- the first months after this weaning process. Indeed,
dren, they were not examined as a separate stratum. fruits, vegetables, and staple or animal products such as
The relationship between the child feeding index fish, meat, and eggs were introduced later, from a me-
score at age 6 months and the cumulative probability of dian of the seventh month of age. Moreover, infant
stunting in the following year was investigated further feeding practices during the critical period around the
and is detailed in Fig 3. Children with a low child feeding weaning process seemed to be a predictor of the childs
index score at age 6 months had a 37% increased risk for future nutritional status. Indeed, inadequate comple-
being stunted at least once from ages 7 to 18 months mentary feeding at age 6 months was strongly associated
compared with those with an average or high index (P with impaired growth and increased probability of stunt-
.03). This association was even stronger after adjustment ing during at least the next 12 months. This could indi-
on variables that potentially were linked to this growth cate a critical importance of this age (developmentally).
outcome. Indeed, in a multivariate analysis, the occur- After this crucial period of transition, the nutritional
rence of stunting was significantly associated with a low adequacy of complementary feeding was considerably
child feeding index at age 6 months (relative risk [RR]: improved to cover the nutritional needs of most of the
1.5; 95% CI: 1.12.0), the diagnosis of pediatric HIV children at ages 9 and 12 months. As a result, the values
infection (RR: 13.9; 95% CI: 10.319.0), and mothers of the child feeding index at these later ages were no
illiteracy (RR: 1.6; 95% CI: 1.22.1), but it was not longer associated with growth outcomes in the subse-
associated with low birth weight (RR: 1.1; 95% CI: 0.6 quent months. This nutritional adequacy improvement
2.0) or the with living in a typical shared housing (RR: could be explained by the continuous nutritional coun-
1.0; 95% CI: 0.9 1.3). seling provided by the study team but also by the fact
that women in Abidjan are more accustomed to weaning
DISCUSSION from 9 months of age rather than earlier.15 It also is
To our knowledge, this study is the first to describe possible that the child feeding index was not as sensitive
prospectively the nature and the ages of introduction of at detecting infants who were receiving inadequate com-
TABLE 3 Relationship Between Mean HAZ, WAZ, and WHZ and Child Feeding Index Scores, at Given Ages
Mean z Score at Age 9 Mo Mean z Score at Age 12 Mo Mean z Score at Age 18 Mo
HAZ WAZ WHZ HAZ WAZ WHZ HAZ WAZ WHZ
Child feeding index score at 6 mo
Low 1.10 1.13 0.04 1.11 1.62 0.57 1.07 1.67 0.77
Average or high 0.78 0.80 0.10 0.77 1.23 0.38 0.67 1.24 0.59
P valuea .10 .04 .41 .04 .03 .31 .01 .02 .30
Child feeding index score at 9 mo
Low or average 0.95 1.47 0.54 0.88 1.47 0.72
High 0.76 1.17 0.28 0.70 1.22 0.49
P valuea .27 .12 .17 .31 .19 .21
Child feeding index score at 12 mo
Low or average 0.88 1.47 0.70
High 0.79 1.33 0.60
P valuea .61 .50 .56
HAZ indicates height-for-age z score; WAZ, weight-for-age z score; WHZ, weight-for-height z score.
a Students t test on the equality of means.
plementary feeding at ages 9 and 12 months as it was at of the complementary feeding characteristics.29,30 That
age 6 months. the previous week of food consumption is not necessar-
We had previously reported that the women who ily representative of long-term usual feeding practices
were included in the Ditrame Plus cohort were repre- might constitute another limitation. Indeed, food could
sentative of the general population of Abidjan as they have been introduced in intervals that were not covered
had been recruited among all attendees of community- by the interviews, which could have overestimated the
run health facilities located in poor areas, with no other age at which this food was introduced. Nevertheless, all
selection criteria than having HIV infection, being at complementary foods were introduced during the first 9
least 18 years of age, and having accepted the study months of age, a period when interviews were con-
protocol.12,13 Given this resource-limited environment, ducted at least once a month, which contributed to
breast milk substitutes (infant formula) were provided minimization of this limitation. Moreover, the longitu-
for free from the initiation of the weaning process until dinal and regular compilation of several 24-hour and
9 months of age. This needs to be taken into account as 7-day recall histories tends to reflect reliably the feeding
it contributed to the nutritional accuracy of the source of pattern during the study period.
milk provided to the infants. Assessing the nutritional adequacy of complementary
This prospective study provided detailed information feeding is complex because qualitative (eg, food diver-
on infant feeding practices from birth with a reasonably sity, food frequency) as well as quantitative (eg, number
high level of precision. Indeed, emphasis was made on of meals, exact amount of each food group, nutrient
the collection of nutritional data with the use of stan- intakes, total energy intake, vitamin coverage) dimen-
dardized forms to perform the recall histories, the fre- sions of infant feeding practices need to be taken into
quent visits scheduled during the follow-up period, and account. The child feeding index that we used essentially
interviews that were conducted by trained health care was qualitative and could have been improved by assess-
workers other than those who counseled the women on ment of quantitative dimensions of child feeding prac-
infant feeding practices. This strategy minimized the ma- tices. Nevertheless, this evaluation would have been
ternal recall bias that could have impaired the estimation difficult in a context in which most of the mothers were
e708 BECQUET, et al
Downloaded from by guest on April 10, 2017
illiterate. Moreover, it would have been impractical be- decline with age and was probably attributable to both
cause the recall histories already lasted 30 min in a study lower breast milk intakes and inaccurate complementary
that had multiple judgment criteria and in which moth- feeding.32,34 Our study provides useful knowledge on this
ers also had to be interviewed by a clinician for their issue in the context of a nutritional intervention aimed
child and for their own health at each visit. at the prevention of HIV through breast milk, underlying
We believe that because of this child feeding index, that adequate feeding practices around the weaning pe-
our study provides a reliable longitudinal view of the riod seem to be crucial for achievement of optimal child
evolution of both characteristics and nutritional ade- growth. In resource-limited countries, HIV-infected
quacy of infant feeding practices among early weaned women therefore should turn to early cessation of
breastfed children. Moreover, this study highlights the breastfeeding only when they are counseled to provide
critical period when such an index could be a predictor to their child adequate complementary feeding to take
of the childs future growth outcomes. over breast milk. In this context, we strongly believe that
The relationship between the different categories of emphasis should be placed on innovative ways to coun-
the child feeding index and the nutritional status of the sel women properly on infant feeding so that the public
children was assessed using anthropometric indices. The health messages could be adapted to their individual
cumulative probability of low height for age was assessed situations.
as it reflects a process of failure to reach linear growth Ideally, the child feeding index presented here could
potential as a result of suboptimal health and/or nutri- be used routinely, especially around the weaning period,
tional conditions.28 As stunting is a severe event in low- to contribute to the assessment of the nutritional ade-
income countries, especially when it starts early in in- quacy of complementary feeding. This index therefore
fancy,31 our intent was to detect the proportion of infants could help to detect children who are at risk for malnu-
who were exposed at least once to this risk. This analysis trition and whose mothers need to receive appropriate
was coupled with the estimations of the mean z scores. and reinforced nutritional counseling. Nevertheless,
We believe that the analysis of the trajectory of height or other prospective studies are needed to assess fully the
weight for age would have been more difficult to inter- accuracy of this child feeding index to detect early chil-
pret. Indeed, a relatively high standard of care was pro- dren who are at risk for malnutrition in other settings
posed within our study: close clinical and nutritional and circumstances.
follow-up adapted to the child age and free provision of
care. Children who had a z score less than 2 SD were Composition of the ANRS 1201/1202 Ditrame Plus Study
expected consequently to be clinically and nutritionally Group
treated, which would have positive consequences on Principal investigators: Francois Dabis, Valeriane Leroy,
growth velocity. Marguerite Timite-Konan, Christiane Welffens-Ekra; co-
We assumed that poor growth was a consequence of ordination in Abidjan: Laurence Bequet, Didier K. Ek-
the nutritional inadequacy of the complementary feeds. ouevi, Besigin Tonwe-Gold, Ida Viho; methods, biosta-
Poor growth also could come from illness that was asso- tistics, and data management: Gerard Allou, Renaud
ciated with not receiving the immune protection from Becquet, Katia Castetbon, Laurence Dequae-Mer-
breast milk. However, the relationship between infant chadou, Charlotte Sakarovitch, Dominique Touchard;
feeding practices and the occurrence of interim illness is clinical team: Clarisse Amani-Bosse, Ignace Ayekoe,
difficult to interpret because of a reverse causation bias.32 Gedeon Bedikou, Nacoumba Coulibaly, Christine Danel,
In addition, poor growth could come from receiving Patricia Fassinou, Apollinaire Horo, Ruffin Likikouet,
contaminated foods. All of the women who were in- Hassan Toure; laboratory team: Andre Inwoley, Francois
cluded in the study had access to tap water, but because Rouet, Ramata Toure; psychosocial team: Helene Agbo,
two thirds of them lived in typical shared housing, the Hortense Aka-Dago, Hermann Brou, Annabel Desgrees-
tap mainly was outside home. It was reported previously du-Lou, Alphonse Sihe, Annick Tijou-Traore, Benjamin
that the quality of municipal water in Abidjan was good Zanou; Scientific Committee: Stephane Blanche, Jean-
but that household water storage was a common prac- Francois Delfraissy, Philippe Lepage, Laurent Mandel-
tice that contributed to contamination of drinking wa- brot, Christine Rouzioux, Roger Salamon.
ter.33 Within our study, women were encouraged to
avoid water storage, but one third of them reported ever ACKNOWLEDGMENTS
having given stored water to their child (Fig 2). Such a The primary sponsor of the ANRS 1201/1202 Ditrame
practice might have had adverse consequences on infant Plus study was the ANRS. Dr Becquet was a fellow of the
health. French Ministry of Education, Research and Technology
Several studies that were conducted in resource-con- and is now a postdoctoral fellow of the French charity
strained countries, where breastfeeding was prolonged SIDACTION. Dr Ekouevi was a fellow of the French
long term, had underlined that the protection against charity SIDACTION.
mortality that was provided by breast milk tended to We gratefully acknowledge the women and children
e710 BECQUET, et al
Downloaded from by guest on April 10, 2017
Complementary Feeding Adequacy in Relation to Nutritional Status Among
Early Weaned Breastfed Children Who Are Born to HIV-Infected Mothers:
ANRS 1201/1202 Ditrame Plus, Abidjan, Cte d'Ivoire
Renaud Becquet, Valriane Leroy, Didier K. Ekouevi, Ida Viho, Katia Castetbon,
Patricia Fassinou, Franois Dabis, Marguerite Timite-Konan and ANRS 1201/1202
Ditrame Plus Study Group
Pediatrics 2006;117;e701
DOI: 10.1542/peds.2005-1911
Updated Information & including high resolution figures, can be found at:
Services /content/117/4/e701.full.html
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
/content/117/4/e701.full.html