The Journal of Nutrition

Infant Feeding and the Development of Obesity: What Does the Science Tell Us?
A Randomized Breast-feeding Promotion
Intervention Did Not Reduce Child Obesity
in Belarus
13
Michael S. Kramer,
4,5
* Lidia Matush,
6
Irina Vanilovich,
6
Robert W. Platt,
4,5
Natalia Bogdanovich,
6
Zinaida Sevkovskaya,
6
Irina Dzikovich,
6
Gyorgy Shishko,
6
Jean-Paul Collet,
5
Richard M. Martin,
7
George Davey Smith,
7
Matthew W. Gillman,
8
Beverley Chalmers,
9
Ellen Hodnett,
10
and Stanley Shapiro
5
Departments of
4
Pediatrics and of
5
Epidemiology and Biostatistics, McGill University Faculty of Medicine, Montreal, QC H3G 1Y6,
Canada;
6
National Research and Applied Medicine Mother and Child Centre, 220053 Minsk, Belarus;
7
MRC Centre for Causal Analysis
in Translational Epidemiology and Department of Social Medicine, University of Bristol, Bristol BS8 1TH, UK;
8
Department of
Ambulatory Care and Prevention, Harvard Medical School/Harvard Pilgrim Health Care, Boston, MA 02215;
9
Ottawa Health Research
Institute, University of Ottawa, Ottawa, ON, Canada; and
10
Faculty of Nursing, University of Toronto, Toronto, ON M5T 1P8, Canada
Abstract
The evidence that breast-feeding protects against obesity is based on observational studies, with potential for confounding
and selection bias. This article summarizes a previously published study in which we assessed whether an intervention
designed to promote exclusive and prolonged breast-feeding affects childrens height, weight, adiposity, and blood
pressure (BP) at age 6.5 y. The Promotion of Breastfeeding Intervention Trial (PROBIT) is a cluster-randomized trial of a
breast-feeding promotion intervention based on the WHO/UNICEF Baby-Friendly Hospital Initiative. A total of 17,046
healthy breast-fed infants were enrolled from 31 Belarussian maternity hospitals and afliated clinics, of whom 13,889
(81.5%) were followed up at 6.5 y with duplicate measurements of height, weight, waist circumference, triceps and
subscapular skinfold thicknesses, systolic and diastolic BP. Analysis was based on intention to treat, with statistical
adjustment for clustering within hospitals/clinics to permit inferences at the individual level. The experimental intervention
led to a large increase in exclusive breast-feeding at 3 mo (43.3% vs. 6.4%, P , 0.001) and a signicantly higher
prevalence of any breast-feeding throughout infancy. No signicant intervention effects were observed on height, BMI,
adiposity measures, or BP. The breast-feeding promotion intervention resulted in substantial increases in the duration and
exclusivity of breast-feeding yet did not reduce measures of adiposity at age 6.5 y. Previous reports of protective effects
against obesity may reect uncontrolled bias caused by confounding and selection. J. Nutr. 139: 417S421S, 2009.
Breast-feeding and obesity
This article summarizes a previously published study (1). If
breast-feeding has an effect on growth and obesity, it could be
mediated by several factors. Components of either breast milk
[e.g., leptin (2,3) or insulin (4)] or infant formula [protein (5)]
may mediate the effect. Another factor could be feeding style,
e.g., nishing the bottle if formula feeding (6). An additional
important and often overlooked potential explanation is selec-
tion bias (7,8). Many studies reporting dose-response effects
(reduced risks of obesity with either more exclusive or more
prolonged breast-feeding) may be biased by the fact that slower-
growing infants are satised by continued exclusive breast-
feeding. Infants who grow more rapidly have higher energy
demands and may cry or be hungry more often. Their mothers
may be either unwilling or unable to increase the frequency of
feeding and instead decide to supplement with formula. Alterna-
tively, family, friends, or physicians may encourage formula
feeding in such infants even if the mother is willing and able
to increase her breast-feeding frequency. In either case, the
faster-growing infant may cause the supplementation, rather
1
Published as a supplement to The Journal of Nutrition. Presented at the
conference Infant Feeding and the Development of Obesity: What Does the
Science Tell Us? held in San Diego, CA, April 9, 2008. The conference was
sponsored by The International Formula Council (IFC), Atlanta, GA. The contents
are the sole responsibility of the authors. The articles comprising this supple-
ment were developed independently, and the conclusions drawn do not
represent the ofcial views of IFC. The mention of trade names, commercial
products, or organizations does not imply endorsement by IFC. The Supplement
Coordinator was Heather Gorby, Life Sciences Research Ofce, Bethesda, MD.
Supplement Coordinator disclosure: H. Gorby is an employee of Life Sciences
Research Ofce and received compensation for services performed as
Supplement Coordinator. There are no other pending nancial interests.
2
Supported by a grant from the Canadian Institutes of Health Research. Dr. Platt
is a career investigator (chercheur-boursier) of the Fonds de la recherche en
sante du Que bec (FRSQ). Drs. Kramer, Platt, and Mr. Dahhou are members of
the Research Institute of the McGill University Health Centre, which is supported
in part by the FRSQ.
3
Author disclosures: M. Kramer, L. Matush, I. Vanilovich, R. Platt, N.
Bogdanovich, Z. Sevkovskaya, I. Dzikovich, G. Shishko, J-P. Collet, R. Martin,
G. Smith, M. Gillman, B. Chalmers, E. Hodnett, and S. Shapiro, no conicts of
interest.
* To whom correspondence should be addressed. E-mail: michael.kramer@
mcgill.ca.
0022-3166/08 $8.00 2009 American Society for Nutrition.
First published online December 23, 2008; doi:10.3945/jn.108.097675. 417S

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
,

2
0
1
4
j
n
.
n
u
t
r
i
t
i
o
n
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

than the supplementation causing the faster growth. This has been
anunderappreciatedprobleminprevious studies. Issues of residual
confounding have been covered by others in this symposium.
One potential solution to selection bias and residual con-
founding is the use of a randomized controlled trial (RCT)
11
design. However, randomization to breast- versus articial
feeding, or even to different durations or degrees of breast-
feeding, is infeasible and probably unethical. However, it is both
ethical and feasible to randomize mothers to a breast-feeding
promotion intervention. The question remains whether to attempt
to inuence the initial feeding choice, i.e., initiation of breast-
feeding versus formula feeding, or the duration and exclusivity of
breast-feeding. The former is very difcult because mothers often
decide even before pregnancy whether or not to breast-feed. For
feasibility reasons, it is easier to inuence a mother who has
already decided to initiate breast-feeding by helping her to breast-
feed exclusively and for a longer duration. This article summa-
rizes the methods and results of a previously published follow-up
of childrenenrolledat birthinanRCTof aninterventiondesigned
to promote breast-feeding exclusivity and duration (1).
The Promotion of Breast-feeding Intervention Trial
The RCT to be presented is the Promotion of Breastfeeding
Intervention Trial (PROBIT), a cluster-randomized trial in the
Republic of Belarus (9). Detailedinformationonthe trial has been
published previously (1,10). The intervention used was based on
the Baby-Friendly Hospital Initiative, an intervention developed
in the early 1990s by WHO and UNICEF. The initiative was
designed primarily for maternity hospitals to help mothers who
begin breast-feeding to succeed in doing so exclusively and for a
longer duration.
Baby-friendly hospital initiative. The Baby-Friendly Hospital
Initiative comprises 10 recommended steps for maternity hos-
pitals (11):
1. Have a written breast-feeding policy
2. Train staff to implement policy
3. Inform mothers about breast-feeding benets
4. Help mothers begin breast-feeding within 30 min of birth
5. Show mothers how to breast-feed and to maintain breast-
feeding
6. Give healthy newborns breast milk only
7. Practice rooming-in 24 h/d
8. Encourage breast-feeding on demand
9. Give no paciers to breast-feeding infants
10. Foster and refer mothers to breast-feeding support groups
Many of these steps had already been shown in RCT to be
effective in increasing duration and exclusivity of breast-feeding,
especially providing help with breast-feeding technique (Step 5),
breast-feeding on demand (Step 8), and postnatal support (Step
10) (10,12,13). We chose this intervention package because it
was already being promoted throughout the world, and the
available evidence strongly suggested that success in achieving
more prolonged and exclusive breast-feeding was likely.
Originally, we planned to implement PROBITin Canada and
the United States. However, most maternity hospitals in both
countries, as well as in Western Europe, had already imple-
mented many changes making them more baby-friendly. Al-
though neither Canada nor the United States had any hospitals
ofcially designated as baby-friendly at the time we designed
PROBIT, recently changed practices in both countries (such as
routine rooming-in and breast-feeding on demand, training of
nursing staff, and hiring of lactation consultants) would have
negatively impacted on our ability to make a difference between
experimental and control sites, resulting in little difference
between groups. In contrast, maternity hospital practices in the
former Soviet Union, which had recently split up when we
started planning the trial in the early 1990s, resembled those in
North America several decades earlier. Basic health services were
already widely available and of high quality in Belarus, however,
along with high vaccination rates, low infant mortality rates,
and high literacy rates. The highly centralized Belarusian health
care system greatly aided the implementation of the intervention
in the experimental hospitals. The Belarus Ministry of Health
was also eager to collaborate with us.
PROBIT resulted in 2 cohorts that differed substantially in
the exclusivity and duration of breast-feeding. These cohorts
were created by randomization rather than by individual choice,
which enables strong causal inferences with respect to breast-
feedings effects on long-term outcomes, thus providing a unique
scientic opportunity. A follow-up (PROBIT II) at age 6.5 y was
carried out at the polyclinics between December 2002 and April
2005. We were able to follow up 81.5% (n 13,889) of the
originally randomized 17,046 infants.
The objective of the 6.5-y follow-up was to examine the
causal relation between prolonged and exclusive breast-feeding
and a number of important health outcomes at early school age.
The growth (including height and adiposity) and blood pressure
(BP) outcomes have been previously published (1) and are
summarized here. We also assessed and have published the
results bearing on allergies and asthma (14), cognitive ability
(15), behavior (16), and dental health (17).
Methods
Study design. The RCTwas carried out using cluster randomization so
that the intervention was the same among all women giving birth at the
same maternity hospital. Randomizing individual women to different
interventions within the same room or even within the same maternity
ward would have led to extensive treatment contamination. To fa-
cilitate follow-up, we restrictedenrollment tochildrenwhowere followed
up at a single pediatric outpatient clinic (polyclinic) that was afliated
with each maternity hospital.
Thirty-four hospitals were originally included in the trial, but
1 experimental and 1 control hospital did not accept their randomiza-
tion, and 1 polyclinic falsied their follow-up data and was therefore
subsequently withdrawn from the trial, thus leaving 31 participating
sites. A total of 17,046 healthy breast-fed newborns born at term and
weighing at least 2500 g were enrolled from these 31 sites. Experimental
and control hospitals and polyclinics were spread throughout the
country of Belarus, excluding the extreme southeast portion of the
country just over the border from Chernobyl, because mothers living
there were often advised not to breast-feed for fear of radionuclide
contamination of their breast milk.
In the PROBIT II follow-up, anthropometric measurements and BP
measures were taken in duplicate and averaged. Anthropometric mea-
surements included height using a wall-mounted stadiometer; weight
using an electronic digital scale (Bella 840; Seca Corporation, Hamburg,
Germany); waist circumference using a nonstretchable cloth tape; and
triceps and subscapular skinfold (SF) thicknesses using Lange SF calipers
(Beta Technology, Santa Cruz, CA).
BP was measured using a digital oscillometric device (Omron M1,
Omron Heathcare, Milton Keynes, UK). There were some technical
problems withthe OmronM1, andtherefore sphygmomanometers hadto
be used in some children. Training and standardization of pediatricians
11
BP, blood pressure; RCT, randomized controlled trial; SF, skinfold.
418S Supplement

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
,

2
0
1
4
j
n
.
n
u
t
r
i
t
i
o
n
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

were carried out during a week-long workshop involving a convenience
sample of school-aged children living in a residential facility near Minsk,
Belarus. Printed materials (in Russian) and a training video fromthe U.S.
National Health and Nutrition Examination Survey were provided to
each participating pediatrician.
The nature of the study prevented blinding of pediatricians, as they
were the same ones who had delivered the original breast-feeding
intervention to the mothers in the experimental polyclinics. As a data
quality measure, we randomly selected 190 children to be audited
(reexamined and remeasured) from the 38 pediatricians participating in
PROBIT II, i.e., 5 children per pediatrician. (In 7 of the 31 polyclinics, the
heavy workload was shared between 2 pediatricians.) To ensure that all
participating children were eligible for selection to be audited, the audit
visit took place after all the study visits had been completed, an average of
18 mo after the study visit. The audit was carried out by 1 of our Minsk-
based pediatrician coinvestigators, all of whom were blinded to the
measures obtained at the primary study visit.
Statistical analysis. The analysis of the PROBIT trial was based on
intention to treat, i.e., as randomized. As expected, there was a large
overlap in actual breast-feeding behavior between the 2 randomized
groups. Some women in the experimental group weaned their infants
immediately, whereas some in the control group breast-fed exclusively
and for a prolonged duration. But with a large sample size and a potent
intervention, health effects should nonetheless be demonstrable.
In addition to analyses based on the continuous anthropometric and
BP measurements, we compared the proportions of children $85th and
$95th percentiles for BMI, based on the CDC 2000 standards (18).
Statistical analyses were based on MIXED for continuous outcomes and
GLIMMIX for dichotomous outcomes in SAS (version 8.2). These are
multilevel models that account for the clustered randomization and thus
permit inferences at the individual level. We also examined multivariate
models that included both stratum-level (geographic region, urban vs.
rural) and the following individual-level covariates: child age at follow-
up, sex, birth weight, maternal education, maternal and paternal height
(for child height), and maternal and paternal BMI (for child BMI and
BP). Because the multivariate results were virtually identical to those
based on the simpler models adjusted only for clustering (which was
expected, given the randomdistribution of covariates), only the latter are
presented below.
Results
PROBIT I. The results for breast-feeding during the rst year of
life are shown in Figure 1. All infants initiated breast-feeding (an
inclusion criterion for the trial). Rapid separation occurred
between the randomized groups. Exclusive breast-feeding was 7
times as frequent at 3 mo in the experimental group (43.3% vs.
6.4%, P ,0.001), but infrequent in both groups at 6 mo (7.9 vs.
0.6%, P , 0.01).
PROBIT II. Baseline factors (maternal age and education,
number of older siblings, maternal smoking, and child sex and
birth weight) between the experimental and control groups were
similar at follow-up (1). Follow-up rates were also similar in the
2 groups: 80.2% in the experimental group; 82.9% in the
control group.
The audit results are summarized in Table 1. Height and BMI
in particular had very good correlations between the clinic visit
and the audit measurements. Waist circumference was also
highly correlated. Measurements that had lower degrees of cor-
relation were the triceps SF and BP measurements.
The means and cluster-adjusted differences (with their 95%
CI) between the experimental and control groups for the
anthropometry and BP results are shown in Table 2. The means
for each of the outcomes were very similar in the 2 groups; no
signicant differences were observed. For BMI, the CI was
extremely narrow. The CI were wide, however, for the triceps SF
and both BP measurements, largely because of clustering of
measurements within polyclinics (i.e., the tendency of some
pediatricians to measure systematically higher or lower than
other pediatricians, even within the same treatment group).
The overweight (BMI $85th percentile) and obesity (BMI $
95th percentile) prevalences in Belarus were far less than those
currently observed in the United States: 13.4% and 5.9%,
respectively, in the experimental group and 12.2% and 5.0% in
the control group. No signicant effect of the experimental
intervention was observed on the risk of the BMI $ 85th
percentile [cluster-adjusted odds ratio (OR) 1.1; 95% CI: 0.8,
1.4] or $ 95
th
percentile [cluster-adjusted OR 1.2 (0.8, 1.6)].
In conclusion, we found no effect of prolonged and exclusive
breast-feeding on height, adiposity, or BP in Belarusian early
school-age children. The fact that our results are based on a
randomized trial design lends increasing weight to our ndings.
This Belarusian population has a low obesity prevalence,
however, and the results may not be generalizable to other
settings. Previous studies are based almost entirely on observa-
tional studies, with potential for selection bias and confounding.
PROBIT III will follow up the children at age 11 y to address
type 2 diabetes and coronary heart disease risk factors based on
fasting blood specimens.
Question and answer session
[Q1]: Your design allows you to look at the effect of breast-
feeding promotion on later obesity. I am curious as to whether
FIGURE 1 Duration of breast-feeding. Reprinted from Kramer et al.
(9), with permission from JAMA.
TABLE 1 Correlation of outcome measures at audit and at
initial visit
1
Anthropometry/BP r (95% CI)
Height 0.84 (0.79, 0.88)
BMI 0.89 (0.85, 0.91)
Waist circumference 0.84 (0.80, 0.88)
Triceps SF 0.59 (0.48, 0.67)
Subscapular SF 0.65 (0.55, 0.72)
Systolic BP 0.55 (0.44, 0.64)
Diastolic BP 0.45 (0.32, 0.55)
1
Adapted from Kramer et al. (1). Reprinted with permission from the American
Society for Nutrition.
Breast-feeding and child obesity in Belarus 419S

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
,

2
0
1
4
j
n
.
n
u
t
r
i
t
i
o
n
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

you were able to measure anything else about what that
intervention might have done in terms of changing other kinds
of health practices? Particularly as to whether the promotion
itself, besides changing breast-feeding, changed other things that
had longer-term effects?
[Dr. Kramer]: Not that we know of. The intervention was
focused on breast-feeding and was not a generalized intervention
on infant or child health. We were not able to collect much
additional information, because data collection was already a
huge workload on top of the regular work of the pediatricians.
On the other hand, the feeding differences we observed between
the experimental and control groups were large enough to have a
causal might effect on some of the outcomes we measured, albeit
not on the ones I showed you today. Now, I think a caveat here
has to be that the effects of the intervention might be less in
Belarus than they would in the United States, for example.
Another caveat is the width of the condence intervals for some
of the outcome measures, which are too wide to exclude small
effects. We are pretty condent in inferring no large effects, but
given some of the measurement issues, a study even as large as
ours cannot exclude small effects.
[Q2]: In your data, there does not appear to be an obesity
epidemic in Belarus. If you had known that about Belarus,
would you have been less anxious to test this outcome there?
And I have a second question: Why did these women stay in the
polyclinic so long?
[Dr. Kramer]: When you say polyclinics, I think you are
referring to the maternity hospitals. The routine postpartumstay
is ;6 or 7 d for a normal vaginal delivery, which is quite a bit
longer than it is in most Western countries. So you are rightI
think that certainly helped us; maybe that was why we were able
to achieve as big a difference as we did in breast-feeding
behavior. But in terms of the causal effect of that behavior on
obesity, that was an advantage rather than a disadvantage.
But I would like to make a point about the obesity epidemic.
If breast-feeding had a potent effect on preventing obesity,
would we be living the obesity epidemic at the same time as we
have witnessed a renaissance in breast-feeding? When parents
started putting their babies on their backs to sleep, the SIDS rate
went down; when people stopped smoking, lung cancer rates
went down; when people started using seatbelts, deaths from
automobile accidents went down. No one ever said that breast-
feeding is the cure-all preventive measure for obesity. But if it
were having a potent effect, would we have seen the epidemic we
are witnessing now? Another way of rephrasing that is, whether
or not breast-feeding has a small effect in any country, countries
such as ours that are experiencing this epidemic have to come up
with better ways of controlling it.
[Q3]: Perhaps successful promotion of breast-feeding would
provide the largest benet where there is the highest risk, such as
situations where the 2 parents are obese. In situations where
there is very little risk, breast-feeding has no effect.
[Dr. Kramer]: To rephrase your comment, Pima Indians were
not always fat, right? Pimas were thin for thousands of years;
their genetic tendency for obesity was probably adaptive. So
genes are obviously playing a role, and there may be some
environmental factors that interact with genes; but the genes
were there before the obesity epidemic. I think the emphasis has
to be on the environment and what public health measures we
can introduce to make it harder to get fat.
Other articles in this symposium include references (1921).
Literature Cited
1. Kramer MS, Matush L, Vanilovich I, Platt RW, Bogdanovich N,
Sevkovskaya Z, Dzikovich I, Shishko G, Collet JP, et al. Effects of
prolonged and exclusive breastfeeding on child height, weight, adipos-
ity, and blood pressure at age 6.5 y: evidence from a large randomized
trial. Am J Clin Nutr. 2007;86:171721.
2. Lyle RE, Kincaid SC, Bryant JC, Prince AM, McGehee RE, Jr. Human
milk contains detectable levels of immunoreactive leptin. Adv Exp Med
Biol. 2001;501:8792.
3. Miralles O, Sanchez J, Palou A, Pico C. A physiological role of breast
milk leptin in body weight control in developing infants. Obesity (Silver
Spring) 2006;14:13711377.
4. Shehadeh N, Khaesh-Goldberg E, Shamir R, Perlman R, Sujov P,
Tamir A, Makhoul IR. Insulin in human milk: postpartum changes
and effect of gestational age. Arch Dis Child Fetal Neonatal Ed. 2003;
88:F2146.
5. Hoppe C, Udam TR, Lauritzen L, Molgaard C, Juul A, Michaelsen
KF. Animal protein intake, serum insulin-like growth factor I, and
growth in healthy 2.5-y-old Danish children. Am J Clin Nutr. 2004;
80:44752.
6. Blissett J, Farrow C. Predictors of maternal control of feeding at 1 and 2
years of age. Int J Obes (Lond). 2007;31:15206.
7. Hill AB. A Short Textbook of Medical Statistics. London: Hodder &
Stoughton; 1977.
8. Sauls HS. Potential effect of demographic and other variables in studies
comparing morbidity of breast-fed and bottle-fed infants. Pediatrics.
1979;64:5237.
9. Kramer MS, Chalmers B, Hodnett ED, Sevkovskaya Z, Dzikovich I,
Shapiro S, Collet JP, Vanilovich I, Mezen I, et al. Promotion of
Breastfeeding Intervention Trial (PROBIT): a randomized trial in the
Republic of Belarus. JAMA. 2001;285:41320.
10. Renfrew MJ, Lang S, Woolridge MW. Early versus delayed initiation of
breastfeeding. Cochrane Database Syst Rev. 2000;CD000043.
11. World Health Organization, UNICEF. Protecting, Promoting and Sup-
porting Breast-feeding: The Special Role of Maternity Services. A Joint
WHO/UNICEF Statement Nonserial Publication. Geneva: World Health
Organization; 1989.
12. Renfrew MJ, Lang S, Martin L, Woolridge MW. Feeding schedules in
hospitals for newborn infants. Cochrane Database Syst Rev. 2000;
CD000090.
13. Sikorski J, Renfrew MJ, Pindoria S, Wade A. Support for breastfeeding
mothers. Cochrane Database Syst Rev. 2002;CD001141.
14. Kramer MS, Matush L, Vanilovich I, Platt R, Bogdanovich N,
Sevkovskaya Z, Dzikovich I, Shishko G, Mazer B. Effect of prolonged
and exclusive breast feeding on risk of allergy and asthma: cluster
randomised trial. BMJ. 2007;335:815.
15. Kramer MS, Aboud F, Mironova E, Vanilovich I, Platt RW, Matush L,
Igumnov S, Fombonne E, Bogdanovich N, et al. Breastfeeding and child
cognitive development: new evidence from a large randomized trial.
Arch Gen Psychiatry. 2008;65:57884.
16. Kramer MS, Fombonne E, Igumnov S, Vanilovich I, Matush L,
Mironova E, Bogdanovich N, Tremblay RE, Chalmers B, et al. Effects
of prolonged and exclusive breastfeeding on child behavior and
TABLE 2 Cluster-adjusted differences in anthropometry and
BP results
1
Outcome Experimental Control Difference (95% CI)
Height, cm 121.1 120.2 10.7 (20.3, 11.7)
BMI, kg/m
2
15.6 15.6 10.1 (20.2, 10.3)
Waist circumference, cm 54.6 54.2 10.3 (20.8, 11.4)
Triceps SF, mm 9.9 10.0 20.4 (21.8, 11.0)
Subscapular SF, mm 5.9 5.8 0.0 (20.4, 10.5)
Systolic BP, mm 97.8 96.7 10.2 (22.9, 13.3)
Diastolic BP, mm 57.3 57.8 10.2 (21.8, 12.2)
1
Adapted from Kramer et al. (1). Reprinted with permission from the American
Society for Nutrition.
420S Supplement

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
,

2
0
1
4
j
n
.
n
u
t
r
i
t
i
o
n
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

maternal adjustment: evidence from a large, randomized trial. Pediatrics.
2008;121:e43540.
17. Kramer MS, Vanilovich I, Matush L, Bogdanovich N, Zhang X,
Shishko G, Muller-Bolla M, Platt RW. The effect of prolonged and
exclusive breast-feeding on dental caries in early school-age children.
New evidence from a large randomized trial. Caries Res. 2007;41:
4848.
18. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS,
Wei R, et al. CDC Growth Charts. From Vital and Health Statistics of
the Centers for Disease Control and Prevention/National Center for
Health Statistics. United States 2000, 314.
19. Butte NF. Impact of infant feeding practices on childhood obesity.
J Nutr. 2009;139:4126.
20. Adair LS. Methods appropriate for studying the relationship of breast-
feeding to obesity. J Nutr. 2009;139:40811.
21. OTierney PF, Barker DJP, Osmond C, Kajantie E, Eriksson JG.
Duration of breast-feeding and adiposity in adult life. J Nutr.
2009;139:4225.
Breast-feeding and child obesity in Belarus 421S

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
,

2
0
1
4
j
n
.
n
u
t
r
i
t
i
o
n
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m