Breastfeeding, Physical Growth, and

Cognitive Development
Jordyn T. Wallenborn, MPH, PhD,a,b Gillian A. Levine, MPH, PhD,a,b Angélica Carreira dos Santos, MPH, PhD,c Sandra Grisi, PhD,c
Alexandra Brentani, PhD,c,* Günther Fink, PhDa,b,*

BACKGROUND AND OBJECTIVES: Breastfeeding

is an evidence-based recommendation for all countries, abstract
but breastfeeding rates have been declining in many middle-income settings. One reason
behind this decline is the perception that breastfeeding may not be necessary in modern
urban settings, where clean water is available and alternative foods are abundant. We
investigate the importance of breastfeeding for early childhood development in the modern
urban context of São Paulo, Brazil.
METHODS: Inour study, we used data from the ongoing prospective Western Region Birth cohort
in São Paulo, Brazil. Children were recruited at birth and managed for 3 years. Durations of
exclusive and mixed breastfeeding were our primary independent variables. Our secondary
independent variable was an indicator for compliance with World Health Organization (WHO)
breastfeeding recommendations. Our primary outcomes of interest were indicators of
children’s physical, cognitive, language, and social-emotional development at 3 years of age.
Adjusted estimates and 95% confidence intervals were calculated by using linear and logistic
regression.
RESULTS: Complying with WHO recommendations to exclusively breastfeed for 6 months
followed by complementary feeding until 2 years of age was associated with a 0.4-SD increase
in overall child development (b: .38; confidence limit = 0.23 to 0.53), a 0.6-SD increase in
height-for-age z score (b: .55; confidence limit = 0.31 to 0.79), and a 67% decrease in the odds
of stunting (odds ratio = 0.33; 95% confidence interval = 0.20 to 0.54).
CONCLUSIONS: Our results suggest that even in settings with easy access to complementary foods,
complying with WHO breastfeeding recommendations is important for healthy physical
growth and cognitive development.

a
Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, bUniversity of Basel, WHAT’S KNOWN ON THIS SUBJECT: As countries
Basel, Switzerland; and cDepartment of Pediatrics, Universidade de São Paulo, São Paulo, Brazil develop economically and reach middle-income levels,
breastfeeding often becomes more challenging, and
*Contributed equally as co-senior authors
rates of breastfeeding significantly drop. Evidence on
Dr Wallenborn conceptualized and designed the study, conducted the initial analyses, and drafted the importance of continued breastfeeding for child
the initial manuscript; Dr Fink conceptualized and designed the study, designed data collection development in upper middle-income countries with
instruments, and coordinated and supervised data collection; Dr Levine and Dr Carreira dos Santos abundant food access is limited.
reviewed the manuscript for important intellectual content; Dr Brentani conceptualized the cohort
design, designed data collection instruments, coordinated and supervised data collection, and WHAT THIS STUDY ADDS: Our results suggest that in
critically reviewed the manuscript for important intellectual content; Dr Grisi coordinated and the upper-middle–income settings of Brazil, complying
supervised data collection and critically reviewed the manuscript for important intellectual content; with the World Health Organization breastfeeding
and all authors reviewed and revised the manuscript, approved the final manuscript as submitted, recommendations is likely beneficial for children’s
and agree to be accountable for all aspects of the work. physical and cognitive development.
DOI: https://doi.org/10.1542/peds.2020-008029
To cite: Wallenborn JT, Levine GA, Carreira dos Santos A,
Accepted for publication Jan 15, 2021
et al. Breastfeeding, Physical Growth, and Cognitive
Development. Pediatrics. 2021;147(5):e2020008029

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

PEDIATRICS Volume 147, number 5, May 2021:e2020008029 ARTICLE
Currently, .250 million children such as the WHO Code of Marketing emotional development, (2) this
worldwide do not reach their full of Breast-milk Substitutes, may also association will be more apparent in
developmental potential.1 Two Lancet lead to lower breastfeeding rates and exclusively breastfed children, with
series on early childhood higher rates of complementary smaller estimates seen in children
development highlighted critical feeding with infant formula.20 who had mixed feeding, and (3)
consequences of delayed childhood height-for-age z score (HAZ) will be
In Brazil, the prevalence of
development and identified risk and smaller among exclusively breastfed
breastfeeding decreased drastically
protective factors that help children children compared with nonbreastfed
during the 1970s.21 Factors contributing
reach their full potential.2 or mixed-fed children. Post hoc, we
to these declines included increasingly
Breastfeeding is among the factors added childhood obesity, which had
affordable alternative feeding options,
that help children’s healthy physical been omitted in the original
changing social norms, and a rise in
and cognitive development3 and is preanalysis plan but was deemed an
mothers’ formal labor commitments
actively promoted by the World important outcome by local
without sufficient parental leave or
Health Organization (WHO) coinvestigators given the rising rates
breastfeeding support.22 The Brazilian
worldwide.4 In a recent systematic of child obesity in this setting.26
National Breastfeeding Program was
review, authors found breastfeeding
a major turning point for the prevalence
to be consistently associated with
of exclusive breastfeeding in infants ,6 METHODS
improved intelligence tests, schooling
months of age, increasing rates from 5%
performance, and adult income Data Source
in 1986 to 37% in 2013.21 However, the
earnings5; for social-emotional
prevalence of exclusive breastfeeding Data used for our prospective cohort
functioning, evidence appears more
varies, with the highest rates seen in study were collected as part of the
mixed.6 For physical growth, the
capital cities and higher socioeconomic São Paulo Western Region Birth
majority of studies from high-income
groups.23 Cohort (ROC) located in the Butantã-
countries reveal leaner growth and
Jaguaré region of São Paulo
slower weight gain trajectories in In this study, we aim to investigate
municipality, Brazil. The cohort
exclusively breastfed infants the association between
comprises all resident children born
compared with formula-fed breastfeeding and children’s physical,
at São Paulo’s university hospital
infants,7–11 implying that cognitive, language, and social-
between April 1, 2012, and March 31,
breastfeeding could also be a key emotional development in this
2014. Birth outcomes were obtained
protective factor for obesity and setting. Our study population of
from electronic medical records.
cardiovascular diseases.12,13 mothers living in a large metropolitan
Additional information on mother-
area of Brazil, a dynamic and rapidly
infant dyads was collected at 36
Today, the extent to which mothers growing modern middle-income
months postpartum by study staff at
engage in breastfeeding varies widely country, represents large urban
the child’s home through structured
across country income groups.14 In environments that are home to
interviews. Data were collected on
most low-income countries, a growing share of families in low-
socioeconomic status, health
breastfeeding is almost universal.15 and middle-income countries.24 Our
standing, breastfeeding practices and
However, as countries develop research directly responds to
other infant feeding behaviors, and
economically and reach middle- a previous Pediatrics study in which
childhood development indicators.
income levels, breastfeeding often authors called for research on a dose-
Additional details on the ROC can be
becomes more challenging or is response relationship between
found elsewhere.27
perceived as less necessary, and rates breastfeeding and infant development
of breastfeeding drop that could adequately control for The original study population
significantly.15–18 Although the confounders.25 We respond to this included 3620 mothers that were
critical importance of breastfeeding call by using a prospective cohort of interviewed at 3 years postpartum.
for early childhood development children growing up in poor urban Our study excluded mother-infant
seems obvious in low-income settings neighborhoods of São Paulo, Brazil, dyads who were not selected for the
where safe alternative foods are while controlling for essential 3-year breastfeeding module (n =
scarce,19 this is not necessarily true in confounding factors overlooked in 1239), had a multiple birth (ie, twins)
higher-income settings where previous studies (ie, home (n = 21), or had a child with
alternative feeding choices are stimulation). On the basis of our a malformation or disability (n = 72)
abundant, affordable, relatively safe, a priori data analysis plan, we (Fig 1). Our final study population
and easily accessible. A lack of hypothesized that (1) breastfeeding consisted of 2288 mother-infant
regulatory frameworks for sales and duration is associated with a higher dyads. The breastfeeding module that
marketing of breast milk substitutes, level of cognitive and social- consisted of 5 breastfeeding

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

2 WALLENBORN et al
 than the 95th percentile), and obese
($95th percentile). The total PRIDI
score (range: 0–61) was normalized
within the study sample to a mean of
0 and SD of 1. At 36 months, both the
mother and trained interviewer
measured the child’s height in
centimeters. Weight was measured by
trained staff only. HAZ and weight-
for-height z score were computed by
using the WHO’s Anthro software
package.30

Our secondary outcomes for social-

emotional development included z
scores on the Early Childhood
Behavior Questionnaire (ECBQ) and
the Strengths and Difficulties
Questionnaire (SDQ). The ECBQ is
a parent report of toddler (1.5–3
FIGURE 1
Consolidated Standards of Reporting Trials flow diagram for sample population. years) temperament consisting of 18
items, with higher scores
questions was added ∼6 months after defined as follows: does not comply representing better social-emotional
launching the 36-month follow-up. As with recommendations, only complies development.31 The SDQ is an
a result, ∼1000 mother-infant dyads with exclusive breastfeeding for at emotional and behavioral screening
were not administered the least 6 months, only complies with tool for children, comprising 25
breastfeeding module. providing breast milk for at least questions for caregivers. Scores range
24 months, and complies with both from 0 to 30, with lower scores
Breastfeeding (at least 6 months of exclusive representing fewer behavioral
Our primary exposure of interest, breastfeeding and total difficulties. We reversed the SDQ
breastfeeding duration, was breastfeeding duration of at least 24 scale so higher scores represent
parameterized multiple ways to months).28 better outcomes to facilitate
explore the mechanisms between comparability with the PRIDI
breastfeeding and our outcomes of Childhood Development Outcomes
estimates. In our a priori data
interest. We investigated both Our primary outcomes for cognitive analysis plan, we specified an
exclusive breastfeeding duration and and physical development were additional secondary outcome: the
total breastfeeding duration in children’s overall development as Caregiver Reported Early
months. Mothers self-reported assessed by the Regional Project on Development Instruments (CREDI).
breastfeeding duration by answering Child Development Indicators (PRIDI) However, CREDI was dropped as
the following questions: “For how (Engle scale) and HAZ. PRIDI is a tool
a secondary outcome because of the
many months did the child used to collect high-quality and
high rate of missing information
exclusively receive breast milk?” and regionally comparable data on the
(67.4%; n = 1542) and potential
“For how many months did the child overall development of children aged
biases created by parents reporting
receive any breast milk?” In addition 2 to 5 years in Latin America,
to the continuous measure, exclusive capturing cognitive, language, social- on their own children. All cognitive
breastfeeding was also categorized as emotional, and motor development.29 and social-emotional development
follows: exclusive breastfeeding for We also analyzed a dichotomous (yes indicators, including PRIDI, ECBQ,
#3 months, 4 to 5 months, and at or no) indicator of child stunting and SDQ, are indicators for overall
least 6 months. Lastly, we combined (HAZ ,22) and child weight status child development and do not directly
exclusive breastfeeding and any using BMI z scores, categorized into identify children with developmental
breastfeeding duration to create an underweight (less than the fifth difficulties. These indicators were
indicator that signifies accordance percentile), normal weight (fifth normalized to a mean of 0 and SD of 1
with WHO international percentile to less than the 85th to facilitate interpretation of
breastfeeding recommendations, percentile), overweight (85th to less estimated group differences.

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

PEDIATRICS Volume 147, number 5, May 2021 3
Potential Confounding Factors evaluate the risk of potential selection until 3 years of age. Similarly, among
A literature search was conducted to bias. We also examined differences all participants at 3 years
identify the following potential between mothers who completed the postpartum, those who did not
confounding factors32–35: household breastfeeding module and those who complete the breastfeeding module
food insecurity score, social support did not. Kernel density plots were had slightly higher levels of caregiver
score, caregiver’s highest educational created to display the empirical educational attainment, MICS home
attainment, income (in Brazilian real), distribution of PRIDI and HAZ by stimulation scores, low birth weight,
Multiple Indicator Cluster Surveys exclusive breastfeeding duration. stunting, and HAZ scores
(MICS) (home stimulation score), (Supplemental Table 6).
To investigate associations between
preterm birth, low birth weight, breastfeeding and physical growth Among study participants (N = 2288),
presence of father or father figure at and childhood development 4.9% of children were born low birth
home, hours per week the caregiver indicators, linear regression was used weight (,2500 g) and 8.1% were
works outside of the home, maternal to obtain b estimates and 95% born prematurely (,37 weeks’
age at birth, age at child assessment, confidence limits (CLs) for PRIDI, gestation). At 3 years of age, almost 1
child care attendance, child sex, ECBQ, SDQ, and HAZ. Logistic in 4 children were stunted (23.9%).
couples conflict, and the Edinburgh regression was used to obtain odds The majority of children attended
Postnatal Depression score. Couples ratios (OR) and 95% confidence child care at least once a week
conflict is the sum score of 4 domains intervals (CIs) for childhood stunting (82.6%) and had a father or father
(eg, assault, sexual coercion, injury, and obesity. Logistic regression was figure in the household (85.2%)
and psychological aggression) from also used in an additional analysis (Table 1). Characteristics stratified by
a revised couples conflict tactics focusing on children with a PRIDI reported exclusive breastfeeding
scale. Respondents could answer score .1 SD below the sample mean. duration categories are also displayed
between 0 and 3, from none of the Child sex and the Edinburgh in Table 1. We found significant
time to all of the time. The social Postnatal Depression score were differences between exclusive
support score consists of 4 domains tested for effect modification by using breastfeeding duration categories and
aimed at quantifying the level of an interaction term in the initial HAZ, stunted growth, PRIDI, postnatal
support for companionship, analysis. After finding no evidence of depression score, child age at
assistance, or other support systems: effect modification, we included both development assessment, levels of
(1) someone to confide in or talk to variables as confounders in our caregiver educational attainment,
about problems, (2) someone to take empirical models. Additionally, all social support score, MICS home
them to the doctor, (3) someone to models investigating mixed stimulation score, house presence of
help with daily chores if they are sick, breastfeeding controlled for father or father figure, and preterm
and (4) some to loan small amounts preceding exclusive breastfeeding birth.
of money if needed. We calculated duration. A P value of .05 signified
a sum score for the social support statistical significance. SAS version Supplemental Figures 2 and 3 display
scores (from 0, being no social 9.4 (SAS Institute, Inc, Cary, NC) was the distribution of exclusive and total
support, to 16, being the highest level used for all analyses. This study was breastfeeding duration. A quarter of
of social support). Categorization approved by the Faculdade de mothers exclusively breastfed 3 to 5
schematics are found in Table 1. Medicina da Universidade de São months. Almost half of women
Paulo Institutional Review Board (9 exclusively breastfed 6 months or
Statistical Analysis 01604312.1.0000.0065). more (Supplemental Fig 2). The
majority (∼55%) of mothers provided
Descriptive statistics were used to at least some breast milk for at least 6
describe the study population by months (Supplemental Fig 3).
using frequencies and percentages for RESULTS
categorical variables and means and From the ROC data, we identified The relationship between child
SDs for continuous variables. 2288 mother-infant dyads for our development indicators and exclusive
Characteristics were also described study. As shown in Supplemental breastfeeding duration categories is
by exclusive breastfeeding duration Table 5, overall, children not assessed shown in Table 2. Compared with
category coupled with a x2 or t test to in the 3-year survey had a slightly mother-infant dyads who exclusively
identify significant differences higher prevalence of low birth weight breastfed #3 months, infants
between groups. Because of the high (9.1% vs 6.1%) and preterm birth exclusively breastfed at least 6
attrition rate from baseline to 36 (10.7% vs 8.9%) as well as slightly months had a 0.3-SD higher PRIDI
months, we examined differences in more supportive home environments score (CL = 0.16 to 0.34) and 0.4
maternal-infant characteristics to compared with children managed higher HAZ score (CL = 0.16 to 0.54).

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

4 WALLENBORN et al
TABLE 1 Description of Sample Population Overall and By Exclusive Breastfeeding Duration
Overall (N = 2288) Exclusively Exclusively Exclusively P
Breastfed #3 mo Breastfed 4–5 mo Breastfed $6 mo
(n = 769) (n = 531) (n = 973)
Dependent variables
HAZa 20.7 (1.7) 21.0 (1.7) 20.8 (1.7) 20.5 (1.7) ,.0001b
Weight status .29
Underweight 251 (11.6) 71 (9.7) 71 (14.0) 108 (11.9) —
Normal weight 1173 (54.3) 416 (56.6) 267 (52.5) 483 (53.3) —
Overweight 350 (16.2) 121 (16.5) 85 (16.7) 143 (15.8) —
Obese 388 (18.0) 127 (17.3) 86 (16.9) 172 (19.0) —
Stunted growth ,.0001b
Yes 524 (23.9) 227 (30.8) 121 (23.5) 175 (18.8) —
PRIDIa 0.1 (1.0) 20.1 (1.0) 0.1 (1.0) 0.2 (1.0) ,.0001b
ECBQa 20.03 (1.0) 20.1 (1.0) 20.03 (1.0) 0.02 (1.0) .14
SDQa 0.03 (1.0) 20.02 (1.0) 0.05 (1.0) 0.1 (1.0) .16
Potential effect modifiers
Edinburgh Postnatal Depression scorea 6.9 (5.2) 7.4 (5.5) 7.1 (5.1) 6.5 (5.1) .01b
Child sex .16
Female 1220 (53.3) 390 (50.7) 286 (53.9) 538 (55.3) —
Male 1068 (46.7) 379 (49.3) 245 (46.1) 435 (44.7) —
Potential confounders
Maternal age at delivery, y .24
13–20 502 (21.9) 176 (22.9) 116 (21.9) 207 (21.3) —
21–25 643 (28.1) 233 (30.3) 155 (29.2) 250 (25.7) —
26–30 553 (24.2) 175 (22.8) 121 (22.8) 253 (26.0) —
.30 590 (25.8) 185 (24.1) 139 (26.2) 263 (27.0) —
Child age at assessmenta 3.5 (0.7) 3.5 (0.6) 3.5 (0.7) 3.4 (0.7) ,.01b
Caregiver highest grade completed ,.01b
None 67 (3.0) 27 (3.6) 16 (3.1) 24 (2.5) —
Elementary 967 (43.3) 367 (48.5) 232 (45.0) 367 (38.8) —
Middle 1086 (48.6) 333 (44.1) 249 (48.3) 493 (52.1) —
Upper 113 (5.1) 29 (3.8) 19 (3.7) 63 (6.7) —
Hours caregiver works outside the homea 17.9 (20.2) 17.6 (20.0) 18.3 (20.1) 17.9 (20.6) .83
Income, R$ .34
0–1000 523 (26.5) 196 (29.4) 109 (24.2) 217 (25.7) —
1001–1600 473 (4.0) 146 (21.9) 105 (23.3) 216 (25.6) —
1601–2250 471 (23.9) 161 (24.1) 111 (24.6) 197 (23.3) —
.2250 507 (25.7) 164 (24.6) 126 (27.9) 215 (25.4) —
Household food insecurity scorea 0.9 (1.6) 1.0 (1.7) 1.0 (1.6) 0.9 (1.6) .22
Social support scorea 12.6 (3.9) 12.6 (4.1) 12.5 (3.9) 12.6 (3.8) .91
MICS home stimulation scorea 4.9 (1.4) 4.8 (1.5) 5.0 (1.3) 5.0 (1.4) .01b
Maternal BMI .04b
Underweight, ,18.5 60 (2.8) 24 (3.3) 11 (2.2) 25 (2.7) —
Normal wt, 18.5–24.9 963 (44.4) 348 (48.1) 220 (43.5) 388 (41.7) —
Overweight, 25–30 738 (34.0) 213 (29.4) 185 (36.6) 339 (36.5) —
Obese, .30 410 (18.9) 139 (19.2) 90 (17.8) 178 (19.1) —
Low birth wt, ,2500 g .80
Yes 111 (4.9) 39 (5.1) 26 (4.9) 43 (4.4) —
Preterm birth, ,37 wk’ gestation .01b
Yes 186 (8.1) 80 (10.4) 43 (8.1) 62 (6.4) —
Child care attendance .05
Never attends 372 (17.4) 130 (18.2) 70 (13.9) 172 (18.9) —
Attends at least once per week 1768 (82.6) 584 (81.8) 432 (86.1) 738 (81.1) —
House presence of father or father figure .04b
Yes 1948 (85.2) 638 (83.1) 447 (84.3) 849 (87.3) —
Couples conflict scorea 1.9 (2.0) 2.0 (1.9) 1.8 (1.9) 1.8 (2.0) .14
Categorical variables are presented as n (column %). Because of rounding, percentages may not add to 100. R$, Brazilian real; —, not applicable.
a Continuous variable. Data are presented as mean (SD).
b Significant P value.

Similarly, the odds of child stunting at months (OR = 0.62; 95% CI = 0.45 to exclusively breastfed $6 months had
36 months were 38% lower with 0.84). Our exploratory analysis 44% lower odds (OR = 0.56; 95% CI =
exclusive breastfeeding for at least 6 confirms that children who were 0.39 to 0.81) of having a PRIDI score

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

PEDIATRICS Volume 147, number 5, May 2021 5
.1 SD below the sample mean complementary breast milk for at breastfeeding duration and child
compared with children exclusively least 24 months (b = .30; CL = 0.03 to development indicators are shown in
breastfed #3 months (results not 0.58). All other behavioral indicators Table 4. Our fully adjusted models
shown). We did not find evidence of revealed weaker associations with revealed an increase of 0.03 in the
a relationship between childhood breastfeeding (P . .05). For physical PRIDI standardized score for every
weight status, ECBQ, or SDQ. Density growth, complying with the month increase in exclusive
plots of PRIDI (Supplemental Fig 4) recommendation to exclusively breastfeeding (CL = 0.02 to 0.04). For
and HAZ (Supplemental Fig 5) breastfed for 6 months only was mixed breastfeeding, the estimate
suggest that the improvements in associated with a 0.7-SD increase (CL remained significant but slightly
these 2 outcomes affect all parts of = 0.44 to 0.99) in HAZ. Similar attenuated (b = .01; CL = 0.002 to
the distribution (ie, that average associations were found for 0.01). The same trend can be seen in
improvements are not driven by complying with both exclusive and physical growth outcomes. A 1-month
children with extremely positive or complementary feeding guidelines (b increase in exclusive breastfeeding
negative outcomes). = .55, CL = 0.31 to 0.79) or only resulted in a significant increase of
providing complementary breast milk 0.04 in our HAZ standardized score
The association between current for at least 24 months (b = .54; CL = (CL = 0.02 to 0.05), whereas a 1-
WHO breastfeeding 0.10 to 0.99). The odds of child month increase in any breastfeeding
recommendations and child stunting were lowest for children duration increased the HAZ
development indicators is shown in who were exclusively breastfed for 6 standardized score by 0.01 (CL = 0.01
Table 3. Compared with maternal- months and received breast milk for to 0.02). The odds of child stunting
infant pairs who did not comply with at least 24 months (OR = 0.33; 95% were lowest for exclusive
WHO recommendations, maternal- CI = 0.20 to 0.54); however, exclusive breastfeeding (OR = 0.93; 95% CI =
infant pairs only complying with the breastfeeding for at least 6 months 0.89 to 0.97) but attenuated for
recommendation to exclusively was also associated with reduced mixed breastfeeding (OR = 0.96; 95%
breastfeed for at least 6 months were odds of child stunting (OR = 0.49; CI = 0.95 to 0.98). We did not find
associated with a 0.4-SD increase in 95% CI = 0.28 to 0.85). No evidence to support a relationship
PRIDI score (b: .41; CL = 0.23 to associations were found between between breastfeeding duration and
0.58). Results were largely the same WHO breastfeeding SDQ or childhood weight status.
for compliance with exclusive recommendations and childhood Supplemental Table 7 provides
breastfeeding for the first 6 months weight status. support that each additional month of
followed by complementary feeding mixed breastfeeding after cessation of
until 2 years of age (b = .38; CL = 0.23 The adjusted associations between exclusive breastfeeding increased
to 0.53) and only providing months of exclusive and mixed PRIDI, ECBQ, SDQ, and HAZ scores

TABLE 2 Adjusted Associations Between Exclusive Breastfeeding and Child Outcomes

Exclusive Breastfeeding Duration, mo
#3 4–5 $6
Fully adjusted b (95% CL)
Cognitive and social-emotional development
PRIDI (continuous) Referent 0.12 (20.02 to 0.26) 0.28 (0.16 to 0.34)***
ECBQ Referent 20.01 (20.14 to 0.12) 20.02 (20.14 to 0.01)
SDQ Referent 0.02 (20.11 to 0.16) 0.04 (20.08 to 0.15)
Physical growth
HAZ Referent 0.10 (20.11 to 0.32) 0.35 (0.16 to 0.54)***
Fully adjusted OR (95% CI)
PRIDI z score ,21 Referent 0.76 (0.51 to 1.14) 0.56 (0.39 to 0.81)**
Weight statusa
Underweight Referent 0.89 (0.53 to 1.49) 1.13 (0.74 to 1.73)
Normal weight Referent Referent Referent
Overweight Referent 1.07 (0.72 to 1.59) 0.81 (0.57 to 1.16)
Obese Referent 1.21 (0.82 to 1.79) 1.38 (0.99 to 1.93)
Stunted Referent 0.80 (0.56 to 1.13) 0.62 (0.45 to 0.84)*
All models adjusted for child sex, maternal age at birth, caregiver highest educational attainment, income, presence of the father or father figure at home, preterm birth, low birth wt,
child care attendance, age at child assessment, household food insecurity score, social support score, couples conflict, hours caregiver works away from the home, and depression and
MICS stimulation score. —, not applicable.
a Adjusted also for maternal BMI.

* P , .05; *** P , .0001.

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

6 WALLENBORN et al
TABLE 3 Relationship Between WHO Breastfeeding Recommendations and Child Outcomes
Does Not Only Complies With Only Complies With Complies With Both
Comply With Exclusive Breastfeeding Providing Breast Milk Exclusive Breastfeeding
Recommendations for At Least 6 mo for At Least 24 mo for At Least 6 mo and Providing
Breast Milk Until At Least 24 mo
Fully adjusted b (95% CL)
Cognitive and social-emotional development
PRIDI Referent 0.41 (0.23 to 0.58)*** 0.30 (0.03 to 0.58)* 0.38 (0.23 to 0.53)***
ECBQ Referent 0.14 (20.03 to 0.31) 0.12 (20.15 to 0.39) 20.01 (20.16 to 0.13)
SDQ Referent 0.12 (20.05 to 0.29) 0.13 (20.14 to 0.40) 0.09 (20.05 to 0.24)
Physical growth
HAZ Referent 0.71 (0.44 to 0.99)*** 0.54 (0.10 to 0.99)* 0.55 (0.31 to 0.79)***
Fully adjusted OR (95% CI)
Weight statusa
Underweight Referent 1.12 (0.59 to 2.14) 2.02 (0.84 to 4.87) 1.48 (0.86 to 2.55)
Normal weight Referent Referent Referent Referent
Overweight Referent 0.77 (0.45 to 1.33) 1.17 (0.51 to 2.65) 0.63 (0.37 to 1.07)
Obese Referent 1.30 (0.80 to 2.09) 0.95 (0.41 to 2.22) 1.50 (0.99 to 2.28)
Stunted Referent 0.49 (0.28 to 0.85)* 0.49 (0.20 to 1.19) 0.33 (0.20 to 0.54)**
All models were adjusted for child sex, maternal age at birth, caregiver highest educational attainment, income, presence of the father or father figure at home, preterm birth, low birth
wt, child care attendance, age at child assessment, household food insecurity score, social support score, couples conflict, hours caregiver works away from the home, depression, and
MICS stimulation score.
a Adjusting for the additional confounder of maternal BMI.

* P , .05; ** P , .01; *** P , .0001.

and decreased the odds of child Our findings that breastfeeding is properly respond to situations that
stunting. No association was found associated with better child promote child development).44 In
between mixed breastfeeding development could be partially several studies, researchers also
duration and childhood weight status. explained through maternal-infant report that breastfeeding duration is
bonding rather than the nutritional linked to positive parenting practices
influence of breast milk alone. in later childhood.43 Authors of
DISCUSSION Research has revealed that children a study based in the United Kingdom
with strong maternal-infant bonding reported that formula use or short
The first 1000 days of life are
have better cognitive and social- breastfeeding duration was
fundamental for cognitive, social-
emotional development.37 In fact, an associated with low levels of
emotional, and physical
infant’s brain development has been nurturance,45 which is a critical
development.36 Our results support
linked to the parental attachment component of parental care that helps
existing evidence that exclusive and
relationship,38 which may be children achieve their full
mixed breastfeeding is a critical
promoted by breastfeeding. Research developmental potential.46
component in ensuring healthy
suggests that breastfeeding lowers
cognitive development and physical We also found evidence that
maternal levels of stress,39 increases
growth, even in a middle-income breastfeeding is associated with
bonding,40 and increases mother-
country. We investigated physical growth at 3 years of age.
infant relationships more generally.41
breastfeeding and early childhood Specifically, we found lower odds of
However, it is also plausible that
development indicators in a region child stunting (ie, higher HAZ score)
lower stress levels enable women to
that, like many middle-income among breastfed infants. Research
breastfeed longer and reach their
countries, has been struggling to has revealed a direct relationship
breastfeeding goals.42
improve breastfeeding rates. We between hormones and growth
provide evidence that exclusively Our evidence on increased factors found in breast milk and
breastfeeding for 6 months alone or breastfeeding duration and better healthy infant body composition,47
in combination with complementary child development outcomes may which could help explain our findings.
feeding for at least 24 months is also be explained through responsive The method of breast milk feeding
important for physical and cognitive feeding and parenting behavior. A may also relate to physical growth.
development. In addition, each systematic review found a consistent Emerging evidence suggests that
additional month of exclusive or relationship between prolonged feeding infants breast milk from
mixed breastfeeding appears to have breastfeeding and responsive feeding, a bottle has a weaker association with
a positive impact on early childhood which is an indicator for responsive healthy weight compared with
development. caregiving43 (ie, the ability to exclusive direct breastfeeding.48

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

PEDIATRICS Volume 147, number 5, May 2021 7
TABLE 4 Fully Adjusted Associations Between Months of Exclusive and Mixed Breastfeeding and Child Development Indicators
Exclusive Breastfeeding Mixed Breastfeedinga
Fully adjusted b (95% CL)
Cognitive and social-emotional development
PRIDI .03 (0.02 to 0.04)*** .01 (0.002 to 0.01)**
ECBQ .01 (20.001 to 0.01) 2.001 (20.01 to 0.004)
SDQ 2.001 (20.01 to 0.01) .01 (20.001 to 0.01)
Physical growth
HAZ .04 (0.02 to 0.05)*** .01 (0.01 to 0.02)**
Fully adjusted OR (95% CI)
Weight statusb
Underweight 1.01 (0.98 to 1.04) 1.02 (1.00 to 1.05)
Normal weight Referent Referent
Overweight .97 (0.93 to 1.01) 1.00 (0.98 to 1.02)
Obese 1.00 (0.97 to 1.03) 1.02 (1.00 to 1.04)
Stunted .93 (0.89 to 0.97)** .96 (0.95 to 0.98)**
All models were adjusted for child sex, maternal age at birth, caregiver highest educational attainment, income, presence of the father or father figure at home, preterm birth, low birth
wt, child care attendance, age at child assessment, household food insecurity score, social support score, couples conflict, hours caregiver works away from the home, depression, and
MICS stimulation score.
a All mixed breastfeeding models control for preceding exclusive breastfeeding duration.
b Also adjusting for the additional confounder of maternal BMI.

* P , .05; ** P , .01; *** P , .0001.

We found no association between all participants completed the 3-year however control for an extensive set of
breastfeeding duration and child breastfeeding module. Our study is also variables capturing home environments,
obesity. Yet, the estimates trended not representative of the entire which may at least partially capture
toward increased odds of overweight Brazilian population or other middle- these aspects.
or obesity for children who were income countries, although large urban
breastfed longer. This counterintuitive areas have become home to the majority
CONCLUSIONS
trend could be explained through the of children in many low- and middle-
high prevalence of overweight and income countries. Additionally, as with The results of this article suggest
obesity in our study population. It is any breastfeeding measure, report of large and robust associations
estimated that .50% of Brazilian breastfeeding is prone to recall and between both exclusive and
populations are overweight or obese.49 social desirability bias; nevertheless, nonexclusive breastfeeding and
Our trends between breastfeeding and recall of breastfeeding duration has been children’s cognitive and physical
child overweight or obesity may be shown to be reliable.53 Albeit relying on development. Further efforts are
a reflection of parental preference for a recall at 36 months, the families were needed to increase breastfeeding
heavier infants50,51 in a setting with managed since the child’s birth, which rates to support children’s healthy
rapidly growing obesity rates52 but are may result in a trusting relationship development.
concerning from a public health with research staff and, consequently,
perspective and warrant further less biased responses during the
research. interview. Emerging evidence suggests ABBREVIATIONS
To our knowledge, this study is the that feeding infants breast milk from CI: confidence interval
first used to investigate breastfeeding a bottle has a weaker association with CL: confidence limit
and early childhood development healthy weight compared with exclusive ECBQ: Early Childhood Behavior
indicators among a unique population direct breastfeeding.48 With our study, Questionnaire
in which middle- and high-income we could not consider direct HAZ: height-for-age z score
characteristics are blended. The breastfeeding compared with bottle- MICS: Multiple Indicator Cluster
prospective birth cohort enabled feeding of human milk, a food frequency Surveys
extensive data collection and allowed list for complementary breastfeeding OR: odds ratio
us to control for important including vitamin supplementation, PRIDI: Regional Project on Child
confounding factors, such as parent- parental height, the role of maternal- Development Indicators
child interactions and home infant bonding, responsive feeding, or ROC: São Paulo Western Region
stimulation, which are likely to parenting behavior as possible Birth Cohort
confound the general associations confounders or mediating factors SDQ: Strengths and Difficulties
between breastfeeding and child because these variables were not Questionnaire
outcomes. However, our study may collected; future research can hopefully WHO: World Health Organization
suffer from selection bias because not address these. In our study, we did

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

8 WALLENBORN et al
Address correspondence Jordyn T. Wallenborn, MPH, PhD, Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, University of
Basel, Socinstrasse 57, PO Box, 4002 Basel, Switzerland. E-mail: [email protected]
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright © 2021 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

REFERENCES
1. Black MM, Walker SP, Fernald LCH, et al.; 10. Patro-Goła˛ b B, Zalewski BM, Polaczek A, 19. Olufunlayo TF, Roberts AA, MacArthur C,
Lancet Early Childhood Development Szajewska H. Duration of Breastfeeding et al. Improving exclusive breastfeeding
Series Steering Committee. Early and early growth: a systematic review in low and middle-income countries:
childhood development coming of age: of current evidence. Breastfeed Med. a systematic review. Matern Child Nutr.
science through the life course. Lancet. 2019;14(4):218–229 2019;15(3):e12788
2017;389(10064):77–90 20. Piwoz EG, Huffman SL. The impact of
11. Ziegler EE. Growth of Breast-Fed and
2. Walker SP, Wachs TD, Grantham- Formula-Fed Infants: Protein and marketing of breast-milk substitutes on
McGregor S, et al. Inequality in early Energy Requirements in Infancy and WHO-recommended breastfeeding
childhood: risk and protective factors Childhood. Basel, Switzerland: Karger practices. Food Nutr Bull. 2015;36(4):
for early child development. Lancet. Publishers; 2006 373–386
2011;378(9799):1325–1338 21. Boccolini CS, Boccolini PMM, Monteiro
12. Carberry AE, Colditz PB, Lingwood BE.
3. Quigley MA, Hockley C, Carson C, Kelly Y, Body composition from birth to 4.5 FR, Venâncio SI, Giugliani ERJ.
Renfrew MJ, Sacker A. Breastfeeding is months in infants born to non- Breastfeeding indicators trends in
associated with improved child obese women. Pediatr Res. 2010;68(1): Brazil for three decades. Rev Saude
cognitive development: a population- 84–88 Publica. 2017;51:108
based cohort study. J Pediatr. 2012; 22. Cai X, Wardlaw T, Brown DW. Global
13. Wells JC, Chomtho S, Fewtrell MS.
160(1):25–32 trends in exclusive breastfeeding. Int
Programming of body composition by
4. World Health Organization. Global early growth and nutrition. Proc Nutr Breastfeed J. 2012;7(1):12
Strategy for Infant and Young Child Soc. 2007;66(3):423–434 23. Santos IS, Barros FC, Horta BL, et al.
Feeding. Geneva, Switzerland: World Breastfeeding exclusivity and duration:
14. Persson LÅ. Breastfeeding in low-
Health Organization; 2003 trends and inequalities in four
resource settings: not a “small matter”.
population-based birth cohorts in
5. Horta BL, Loret de Mola C, Victora CG. PLoS Med. 2018;15(8):e1002646
Pelotas, Brazil, 1982-2015. 2019;48(suppl
Breastfeeding and intelligence:
15. United Nations Children’s Fund. 1):i72–i79
a systematic review and meta-analysis.
Breastfeeding: A Mother’s Gift, for Every 24. United Nations Population Division.
Acta Paediatr. 2015;104(467):14–19
Child. New York, NY: United Nations World Urbanization Prospects: The 2018
6. Krol KM, Grossmann T. Psychological Children’s Fund; 2018 Revision. New York, NY:: United Nations;
effects of breastfeeding on children and 2018
16. Ndirangu MN, Gatimu SM, Mwinyi HM,
mothers. Bundesgesundheitsblatt
Kibiwott DC. Trends and factors 25. Sacker A, Quigley MA, Kelly YJ.
Gesundheitsforschung
associated with early initiation of Breastfeeding and developmental delay:
Gesundheitsschutz. 2018;61(8):977–985
breastfeeding in Namibia: analysis of findings from the millennium cohort
7. Péneau S, Hercberg S, Rolland-Cachera the Demographic and Health Surveys study. Pediatrics. 2006;118(3). Available
M-F. Breastfeeding, early nutrition, and 2000-2013. BMC Pregnancy Childbirth. at: www.pediatrics.org/cgi/content/full/
adult body fat. J Pediatr. 2014;164(6): 2018;18(1):171 118/3/e682
1363–1368
17. Ogbo FA, Dhami MV, Awosemo AO, et al. 26. Casagrande D, Waib PH, Sgarbi JA.
8. Singhal A, Lanigan J. Breastfeeding, Regional prevalence and determinants Increase in the prevalence of
early growth and later obesity. Obes of exclusive breastfeeding in India. Int abdominal obesity in Brazilian school
Rev. 2007;8(suppl 1):51–54 Breastfeed J. 2019;14(1):20 children (2000-2015). Int J Pediatr
9. Kramer MS, Guo T, Platt RW, et al. Infant 18. Liu P, Qiao L, Xu F, Zhang M, Wang Y, Adolesc Med. 2017;4(4):133–137
growth and health outcomes Binns CW. Factors associated with 27. Brentani A, Scoleze Ferrer AP, Brentani
associated with 3 compared with 6 mo breastfeeding duration: a 30-month H, et al. Cohort profile: São Paulo
of exclusive breastfeeding. Am J Clin cohort study in northwest China. J Hum Western Region Birth Cohort (ROC). Int
Nutr. 2003;78(2):291–295 Lact. 2013;29(2):253–259 J Epidemiol. 2020;49(5):1438–1438g

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

PEDIATRICS Volume 147, number 5, May 2021 9
 28. World Health Organization. Exclusive insecurity and breastfeeding in Canada. 45. Brown A, Arnott B. Breastfeeding
Breastfeeding for Six Months: Best for CMAJ. 2018;190(11):E312–E319 duration and early parenting
Babies Everywhere. Geneva, 36. Daelmans B, Darmstadt GL, Lombardi J, behaviour: the importance of an infant-
Switzerland: World Health Organization; et al.; Lancet Early Childhood led, responsive style. PLoS One. 2014;
2011 Development Series Steering 9(2):e83893
29. Verdisco A, Cueto S, Thompson J, et al. Committee. Early childhood 46. Britto PR, Lye SJ, Proulx K, et al.
Urgency and Possibility Results of PRIDI development: the foundation of Nurturing care: promoting early
A First Initiative to Create Regionally sustainable development. Lancet. 2017; childhood development. Lancet. 2017;
Comparative Data on Child 389(10064):9–11 389(10064):91–102
Development in Four Latin American 37. Johnson K. Maternal-infant bonding: 47. Fields DA, Demerath EW. Relationship of
Countries Technical Annex. Washington, a review of literature. Int J Childbirth insulin, glucose, leptin, IL-6 and TNF-a in
DC: Technical Annex Inter-American Educ. 2013;28(3):17–22 human breast milk with infant growth
Development Bank; 2009 and body composition. Pediatr Obes.
38. Winston R, Chicot R. The importance of
30. World Health Organization. WHO Child early bonding on the long-term mental 2012;7(4):304–312
Growth Standards: Length/Height-for- health and resilience of children. 48. Azad MB, Vehling L, Chan D, et al.; CHILD
Age, Weight-for-Age, Weight-for-Length, London J Prim Care (Abingdon). 2016; Study Investigators. Infant feeding and
Weight-for-Height and Body Mass Index- 8(1):12–14 weight gain: separating breast milk
for-Age: Methods and Development. from breastfeeding and formula from
39. Mezzacappa ES. Breastfeeding and
Geneva, Switzerland: World Health food. Pediatrics. 2018;142(4):e20181092
maternal stress response and health.
Organization; 2006
Nutr Rev. 2004;62(7 pt 1):261–268 49. Arbex AK, Rocha DR, Aizenberg M,
31. Putnam SP, Gartstein MA, Rothbart MK. Ciruzzi MS. Obesity epidemic in Brazil
40. Liu J, Leung P, Yang A. Breastfeeding
Measurement of fine-grained aspects of and Argentina: a public health concern.
and active bonding protects against
toddler temperament: the Early J Health Popul Nutr. 2014;32(2):327–334
children’s internalizing behavior
Childhood Behavior Questionnaire.
problems. Nutrients. 2013;6(1):76–89 50. Laraway KA, Birch LL, Shaffer ML, Paul
Infant Behav Dev. 2006;29(3):386–401
41. Else-Quest NM, Hyde JS, Clark R. IM. Parent perception of healthy infant
32. Wallenborn JT, Chambers GJ, Masho SW. Breastfeeding, bonding, and the and toddler growth. Clin Pediatr (Phila).
The role of paternity acknowledgment mother-infant relationship. Merrill- 2010;49(4):343–349
in breastfeeding noninitiation. J Hum Palmer Q. 2003;49(4):495–517
Lact. 2018;34(4):737–744 51. Redsell SA, Atkinson P, Nathan D,
42. Dozier AM, Nelson A, Brownell E. The Siriwardena AN, Swift JA, Glazebrook C.
33. Brand E, Kothari C, Stark MA. Factors relationship between life stress and Parents’ beliefs about appropriate
related to breastfeeding breastfeeding outcomes among low- infant size, growth and feeding
discontinuation between hospital income mothers. Adv Prev Med. 2012; behaviour: implications for the
discharge and 2 weeks postpartum. J 2012:902487 prevention of childhood obesity. BMC
Perinat Educ. 2011;20(1):36–44 Public Health. 2010;10(1):711
43. Ventura AK. Associations between
34. Wallenborn JT, Wheeler DC, Lu J, Perera breastfeeding and maternal 52. Monteiro CA, Conde WL, Popkin BM.
RA, Masho SW. Importance of familial responsiveness: a systematic review of Income-specific trends in obesity in
opinions on breastfeeding practices: the literature. Adv Nutr. 2017;8(3): Brazil: 1975-2003. Am J Public Health.
differences between father, mother, and 495–510 2007;97(10):1808–1812
mother-in-law. Breastfeed Med. 2019;
44. Black MM, Aboud FE. Responsive 53. Li R, Scanlon KS, Serdula MK. The
14(8):560–567
feeding is embedded in a theoretical validity and reliability of maternal
35. Orr SK, Dachner N, Frank L, Tarasuk V. framework of responsive parenting. J recall of breastfeeding practice. Nutr
Relation between household food Nutr. 2011;141(3):490–494 Rev. 2005;63(4):103–110

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

10 WALLENBORN et al
 Breastfeeding, Physical Growth, and Cognitive Development
Jordyn T. Wallenborn, Gillian A. Levine, Angélica Carreira dos Santos, Sandra Grisi,
Alexandra Brentani and Günther Fink
Pediatrics 2021;147;
DOI: 10.1542/peds.2020-008029 originally published online April 22, 2021;

Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/147/5/e2020008029
References This article cites 44 articles, 6 of which you can access for free at:
http://pediatrics.aappublications.org/content/147/5/e2020008029#BI
BL
Subspecialty Collections This article, along with others on similar topics, appears in the
following collection(s):
Growth/Development Milestones
http://www.aappublications.org/cgi/collection/growth:development_
milestones_sub
International Child Health
http://www.aappublications.org/cgi/collection/international_child_he
alth_sub
Breastfeeding
http://www.aappublications.org/cgi/collection/breastfeeding_sub
Permissions & Licensing Information about reproducing this article in parts (figures, tables) or
in its entirety can be found online at:
http://www.aappublications.org/site/misc/Permissions.xhtml
Reprints Information about ordering reprints can be found online:
http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021

 Breastfeeding, Physical Growth, and Cognitive Development
Jordyn T. Wallenborn, Gillian A. Levine, Angélica Carreira dos Santos, Sandra Grisi,
Alexandra Brentani and Günther Fink
Pediatrics 2021;147;
DOI: 10.1542/peds.2020-008029 originally published online April 22, 2021;

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/147/5/e2020008029

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it
has been published continuously since 1948. Pediatrics is owned, published, and trademarked by
the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2021
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news at Swets Blackwell Inc. on May 7, 2021