Acta Pñdiatr 88: 1327±32.

1999

Duration of breastfeeding and developmental milestones during the

latter half of infancy
M Vestergaard1, C Obel1, TB Henriksen1, HT Sørensen2, E Skajaa1 and J Østergaard3
Perinatal Epidemiological Research Unit1, Department of Obstetrics and Gynaecology, Department of Internal Medicine V2 and
Department of Paediatrics3, Aarhus University Hospital, Denmark

Vestergaard M, Obel C, Henriksen TB, Sørensen HT, Skajaa E, Østergaard J. Duration of

breastfeeding and developmental milestones during the latter half of infancy. Acta Pædiatr 1999;
88: 1327–32. Stockholm. ISSN 0803–5253
Several studies have suggested that breastfeeding has a long-term influence on brain development.
However, interpretation of these findings is complicated by the presence of many potential
confounding factors. Only a few studies have examined infants before 1 y of age, although very
early assessment might reduce the role of environmental influence. We investigated the association
between exclusive breastfeeding and three developmental milestones related to general and fine
motor skills and early language development at the age of 8 mo. We followed 1656 healthy,
singleton, term infants, with a birthweight of at least 2500 g, born between May 1991 and February
1992 in Aarhus, Denmark. Information was collected at 16 wk gestation, at delivery and when the
infant was 8 mo old. Motor skills were evaluated by measurement of crawling and pincer grip.
Early language development was defined as the ability to babble in polysyllables. The proportion
of infants who mastered the specific milestones increased consistently with increasing duration of
breastfeeding. The relative risk for the highest versus the lowest breastfeeding category was 1.3
(95% CI: 1.0–1.6) for crawling, 1.2 (95% CI: 1.1–1.3) for pincer grip and 1.5 (95% CI: 1.3–1.8)
for polysyllable babbling. Little change was found after adjustment for confounding. In conclusion,
our data support the hypothesis that breastfeeding benefits neurodevelopment. & Breastfeeding,
developmental milestones, infant development

M Vestergaard, Perinatal Epidemiological Research Unit, Department of Obstetrics and

Gynaecology, Aarhus University Hospital, DK-8200 Aarhus N, Denmark (Tel. ‡45 89496380, fax.
‡45 89496373)

Experiments in animals have indicated that prenatal and
early postnatal malnutrition affects the maturation and
function of the developing central nervous system
(1, 2). At the time of birth, the human brain experiences
a growth spurt (3), which continues for 12–18 mo
postpartum (4, 5), suggesting that optimal nutrition may
be particularly important during this period. Several
studies have examined the relationship between breast-
feeding and development. The majority found that
neurodevelopment was more advanced among indivi-
duals who were breastfed rather than formula-fed in
early life (6–14), but some found no difference (15–17).
However, the association between breastfeeding and
development of infants below 1 y of age has seldom
been studied. One would expect differences to be
present early in life if breastfeeding benefits neurode-
velopment directly. In two studies (9, 10) infants were
examined at various ages beginning at 6 mo, but no
differences were found until age 1 y (10) and 2 y (9).
Most investigators have measured infant development
with standardized tests such as “the Bayley Scales of
Infant Development”. These tests summarize different
fields of development into a common score. However, it
is also important to assess the association between
breastfeeding and specific aspects of neurodevelop-
ment.
Breastfed and formula-fed infants could differ due to
differences in parental skills, genetic potential or other
potential confounding factors. However, another possi-
ble influential factor is that human milk contains
substances essential for optimal neurodevelopment.
Most evidence is related to the long-chain polyunsatu-
rated fatty acids, particularly docosahexaenoic acid
(DHA), which is present in human milk, but not in
standard formulas (18). DHA is an important compo-
nent of cell membranes in the cortex and retina (19–22).
Randomized trials have shown that infants who were
fed formula supplemented with DHA had more ad-
vanced neural maturation than infants fed standard
formula (19, 20, 23). Furthermore, human milk contains
hormones, oligosaccharides and other trophic factors
(24, 25) that might influence brain growth and matura-
tion.
The aim of the present study was to examine the

 Scandinavian University Press 1999. ISSN 0803-5253

1328 A Vestergaard et al.
association between breastfeeding and three develop-
mental milestones related to fine motor skills, general
motor skills and early language development at the age
of 8 mo.
Patients and methods
The infants constituted a well-defined part of a prenatal
cohort established at the Department of Obstetrics,
Aarhus University Hospital, Denmark, from August
1989 to September 1991. The design of the study has
been described previously (26). Briefly, all women
attending the routine antenatal care programme were
invited to participate. The department, which is the only
maternity ward in the city, serves a well-defined
geographical area with a population of about 250 000.
At 16 wk gestation the women were asked to fill in a
basic questionnaire about medical and obstetric history,
and a study entry questionnaire providing information
about the sociodemographic conditions of the family.
The tending midwife filled in a specific registration
form providing information about the delivery and the
newborn (sex, weight, length, Apgar score and mal-
formations). A research midwife who reviewed all the
hospital records validated the obstetric data. Any
complication during the first days after delivery was
registered.
The infants were born between May 1991 and
February 1992. Inclusion criteria for the study were as
follows: singleton infants whose mothers were included
in the basic study and who were still living in the
municipality of Aarhus at the age of 8 mo; Danish-
speaking parents; gestational age 37–42 wk; birth-
weight 2500 g; Apgar score 7 at 5 min after birth; no
known handicaps; and no referral to the neonatal
department. Of the 2863 liveborn infants, 2108 were
eligible for the study. Of these, 1883 (89%) participated,
but because of missing information on breastfeeding
and milestone evaluation, only 1656 (79%) were
included in the analyses.
A health visitor interviewed the parents about the
duration of breastfeeding, periods of illness, the use of
day care, changes of cohabitation status, parents’ job
status and the parents’ experience with child rearing. A
questionnaire completed by the parents provided
information about mothers’ alcohol and smoking habits
during the first months postpartum. Milestone evalua-
tions of fine motor skills, general motor skills and
language development were tested according to well-
described methods (27, 28). The assessments were
made by the health visitors (n = 74) in the homes of
the infants in the presence of at least one parent. The
assessment was based on this single visit to the home,
which lasted between 1 and 1.5 h. Milestone evaluation
was exclusively based on observations of the health
visitors. Examinations were postponed when the infants
ACTA PÆDIATR 88 (1999)
were not feeling well. The health visitors were not
aware of the specific aim of this study.
Duration of breastfeeding was classified into four
categories based on the number of months of exclusive
breastfeeding, i.e. no food apart from mother’s milk,
water and vitamins. Information about the duration of
partial breastfeeding was available for the infants who
were never exclusively breastfed.
General motor skills were evaluated, i.e. infants
crawling with arms and legs moving in a diagonal
pattern, and with the stomach above the mat.
Fine motor skills were evaluated, i.e. infants demon-
strating the pincer grip, that is the ability to pick up a
tiny object between thumb and index finger with right or
left hand.
Early language development was defined as the
infant’s babbling abilities and use of syllables and
consonant change. The infants were categorized as
polysyllable babblers if they used syllables and changed
consonants within the same string, e.g. “nonono-
mamama-najnajnaj”.
Corrected examination age was defined as the age at
examination adjusted for gestational age at birth.
The research protocol was approved by the regional
committee for ethics in science and by the Danish
committee of registers.
We analysed fine and general motor skills and
language development using contingency tables in
relation to duration of breastfeeding. The results from
the bivariate analyses are presented as relative risks
(RR) with 95% confidence intervals (CI). We subse-
quently stratified the contingency tables according to
each of the potential confounding factors listed and
categorized as shown in Table 3. Adjustment for
multiple variables was carried out by logistic regression
analyses. A priori important variables, and variables
that changed the estimates of association, were chosen
for the final model (29, 30). Unadjusted and adjusted
odds ratios (OR) are presented in Table 4 because
logistic regression analyses do not provide RR.
Results
We found that 644 (38.8%) infants were able to crawl,
1243 (75.1%) had mastered the pincer grip with right or
left hand and 1021 (61.7%) were polysyllable babblers.
One thousand five hundred and fifty-two (93.7%)
women breastfed their infants exclusively for at least
1 mo, and 1088 (65.7%) continued for 4 mo, as
recommended by the WHO. The mean corrected
examination age was 35.7 wk, and 95% of the infants
were examined within the age interval 30.7–40.7 wk.
The proportion of infants who mastered the specific
milestones increased consistently with increasing dura-
tion of breastfeeding. The RR for the highest vs. lowest
breastfeeding category was 1.3 (95% CI: 1.0–1.6) for
ACTA PÆDIATR 88 (1999) Breastfeeding and developmental milestones 1329

Table 1. Developmental milestones according to the duration of exclusive breastfeeding among 1656 infants.

Crawling Pincer grip Polysyllable babblers

Exclusive Total
breastfeeding (mo) count (n) % RR (95% CI) % RR (95% CI) % RR (95% CI)
0–1 266 35.0 Reference 70.7 Reference 48.5 Reference
2–3 302 33.8 1.0 (0.8–1.2) 71.9 1.0 (0.9–1.1) 58.3 1.2 (1.0–1.4)
4–5 866 40.2 1.1 (1.0–1.4) 75.6 1.1 (1.0–1.2) 63.9 1.3 (1.2–1.5)
6‡ 222 45.5 1.3 (1.0–1.6) 82.4 1.2 (1.1–1.3) 73.4 1.5 (1.3–1.8)

crawling, 1.2 (95% CI: 1.1–1.3) for pincer grip and 1.5
(95% CI: 1.3–1.8) for polysyllable babbling (Table 1).
Table 2 shows the association between partial
breastfeeding and the three specific milestones for
infants never exclusively breastfed (n = 101). The
results failed to reach statistical significance. However,
the advantages of exclusive breastfeeding were even
more pronounced when we excluded infants who were
partially breastfed (data not shown).
Table 3 shows the proportion of infants who mastered
the developmental milestones according to potential
confounders. Mastery of the three developmental mile-
stones was more prevalent among infants with a higher
corrected age at examination. Furthermore, the motor
skills were associated with maternal education. Infants
delivered by Caesarean section were less likely to
crawl. Girls and infants with no or at least two siblings
at home were more likely to have mastered the pincer
grip and to be polysyllable babblers. The pincer grip
was more frequently found if gestational age at delivery
and birthweight were high. Infants were more likely to
be polysyllable babblers if their mothers were non-
smokers or smoked <9 cigarettes/d during pregnancy,
and if they spent time in day care at the time of the
examination.
Data were stratified according to the potential
confounders listed and categorized as shown in Table
3; there was little or no confounding from any single
variable. Multiple logistic regression models were
constructed, incorporating family’s social group,
mother’s education, smoking during pregnancy, birth-
weight, gestational age at delivery, corrected examina-
tion age, whether the infant was living with one or two
parents and the number of illnesses during the first 8 mo
of life. Table 4 shows that the unadjusted and adjusted
OR are almost identical. Thus, there was little aggregate
confounding from the factors mentioned.
Discussion
By contrast with two previous studies (9, 10), we found
an association between breastfeeding and neurodeve-
lopment in infants below 1 y of age, which persisted
after adjustment for confounding. Small sample sizes
may explain lack of association in the previous studies
(9, 10). The role of environmental confounding might
be reduced in our study because the infants were
examined very early in life, at a time before other
factors can have had much influence.
The loss to follow-up in our study of 21% was mainly
due to unwillingness to participate and to missing
information on breastfeeding or milestone develop-
ment. Selection bias would have been introduced if non-
responders were breastfed for a longer time and had less
advanced development than the infants who remained in
the study, or if non-responders were breastfed for a
shorter time and had more advanced development than
the study infants. Unfortunately, there was insufficient
information on non-responders.
Ninety-five percent of the infants were examined
within the age interval 30.7–40.7 wk. We found no
change in the estimate of interest when the data were
stratified according to the corrected examination age.
Thus, variation in the examination age did not introduce
bias. However, the corrected examination age was
included in the final logistic regression model.
Perinatal problems or illness of the infant may
interfere with the duration of breastfeeding. Children
nursed for a longer time might therefore be a selected
group of more healthy individuals. We carefully
selected a homogeneous cohort of healthy, full-term
babies with a birthweight of at least 2500 g in order to
minimize the risk that this “healthy child selection”
might possibly affect later development. In addition, we
included the number of periods with illnesses in the

Table 2. Developmental milestones according to the duration of partial breastfeeding for children never exclusively breastfed (n = 101).
Crawling Pincer grip Polysyllable babblers
Partial Total
breastfeeding (mo) count (n) % RR (95% CI) % RR (95% CI) % RR (95% CI)
0 25 20.0 Reference 72.0 Reference 32.0 Reference
1 27 22.2 1.1 (0.4–3.2) 66.7 0.9 (0.6–1.3) 63.0 2.0 (1.1–3.6)
2 14 35.7 1.8 (0.6–5.2) 71.4 1.0 (0.7–1.5) 57.1 1.8 (0.8–3.8)
3‡ 35 40.0 2.0 (0.9–4.6) 68.6 1.0 (0.7–1.3) 37.1 1.2 (0.6–2.4)

Information about duration of partial breastfeeding was missing for 3 children.

1330 A Vestergaard et al. ACTA PÆDIATR 88 (1999)

Table 3. Developmental milestones according to child and maternal characteristics, sociodemographic factors and maternal lifestyle factors
(n = 1656).
Variable n Crawling (%) Pincer grip (%) Polysyllable babblers (%)
Family’s social status
High 271 40.2 73.1 59.4
Middle 582 37.6 77.5 64.8
Low 676 39.2 73.1 58.9
Mother’s education (y)
7–9 114 34.2 65.8 56.1
10–11 400 33.8 73.8 59.3
12‡ 1013 41.5 76.2 62.9
Mother’s age (y)
<25 239 42.7 73.6 59.0
25–29 720 38.8 74.3 61.3
30–34 499 39.3 75.2 60.7
35‡ 198 33.8 79.3 68.7
Smoking during pregnancy (cigarettes/d)
0 1097 39.7 75.8 62.6
1–9 232 39.2 75.4 62.5
10‡ 259 35.5 71.4 54.1
Alcohol during pregnancy (units/wk)
<1 1032 39.9 75.1 62.4
1–2 387 37.0 73.9 59.4
3–4 136 34.6 73.5 61.8
5–14 38 34.2 68.4 63.2
Sex
M 852 39.2 70.2 58.9
F 804 38.6 80.2 64.6
Gestational age (wk)
37–38 190 33.7 71.1 58.9
39–40 991 39.4 73.8 60.5
41‡ 464 40.3 79.3 64.7
Birthweight (g)
2500–2999 180 38.3 66.1 60.0
3000–3499 504 39.5 75.4 61.5
3500–3999 663 37.7 76.2 61.4
4000‡ 306 41.2 77.5 63.7
Apgar score (5 min)
7–9 59 35.6 64.4 57.6
10‡ 1583 39.0 75.6 61.7
Delivery
Vaginal 1558 39.7 75.0 61.3
Caesarean section 98 26.5 75.5 67.3
Siblings at home
0 791 40.1 77.1 64.3
1 563 36.8 70.9 57.5
2‡ 172 37.8 76.2 59.9
Day care (h/wk)
0 538 39.4 75.8 56.7
1–21 219 43.4 78.5 65.8
22–50 834 37.9 74.3 64.7
Periods of illness (n/8 mo)
0–1 862 38.6 75.8 62.4
2–3 571 39.1 74.3 62.0
4‡ 181 38.7 76.2 57.5
Parental experience with child rearing
Yes 969 37.8 74.1 59.4
No 457 39.2 74.8 63.9
Infants living with
Two parents 1484 38.8 74.4 62.0
One parent 126 38.9 80.2 57.1
Mother’s working status
Employed 1179 38.6 75.0 63.0
Unemployed 460 39.3 74.8 58.3
Father’s working status
Employed 1402 38.9 74.8 61.6
Unemployed 186 39.2 74.7 62.9
Corrected examination age (wk)
<33 145 19.3 52.4 45.5
33–34 456 29.8 63.4 50.4
35–36 503 36.4 76.5 62.0
37‡ 552 53.8 89.5 74.8

Due to missing values the numbers do not always add up to 1656.

ACTA PÆDIATR 88 (1999) Breastfeeding and developmental milestones 1331

Table 4. Developmental milestones according to duration of exclusive breastfeeding for 1348 infants with no missing information on the
confounding factors.

OR (95% CI)
Crawling Pincer grip Polysyllable babblers
Exclusive
a a
breastfeeding (mo) Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted a

0–1 Reference Reference Reference Reference Reference Reference

2–3 0.9 (0.6–1.3) 0.7 (0.5–1.1) 1.2 (0.8–1.8) 1.1 (0.7–1.8) 1.2 (0.8–1.7) 1.1 (0.8–1.7)
4–5 1.3 (1.0–1.8) 1.2 (0.8–1.7) 1.5 (1.1–2.1) 1.4 (1.0–2.1) 1.7 (1.3–2.4) 1.6 (1.1–2.3)
6‡ 1.5 (1.0–2.3) 1.4 (0.9–2.1) 2.3 (1.5–3.7) 2.2 (1.3–3.7) 2.8 (1.8–4.2) 2.5 (1.6–3.9)
a
Adjusted for family’s social group, mother’s education and age, smoking during pregnancy, birthweight and gestational age at delivery, corrected
examination age, number of periods with illness and whether the child was living with one or two parents. Variables categorized as in Table 4 and
entered into the logistic regression model as dummy variables equal to the number of categories ÿ1.

regression model, but the association remained un-
changed.
Developmental milestones were used as outcome
measures to simplify evaluation and interpretation of
the results. Although the definitions were simple, and
the health visitors were trained to evaluate develop-
mental milestones, some variation in the judgements
must be expected because many health visitors partici-
pated in the study. The health visitors were not blinded
concerning breastfeeding, but they were unaware of the
specific aim of the study; thus, any misclassification is
probably non-differential, therefore causing bias to-
wards the null (30).
Although we adjusted for a variety of potential
confounding factors, residual confounding may still be
present. Some factors are very difficult to control for,
e.g. children’s genetic potentiality and environmental
stimulation. Three studies have examined the associa-
tion between breastfeeding and infant development
after adjustment for maternal intelligence. The associa-
tion remained significant in two (10, 14), but was
insignificant in the third (17). Information about
parental intelligence was not available in the present
study. However, the results remained unchanged after
adjustment for the main proxy variables, maternal
education and social class. Furthermore, adjustment
for parental intelligence might be less important when
studying the development of milestones than the
performance in mental tests. Breastfeeding might
intensify the early mother–infant relationship, which
again could influence the degree of stimulation offered
the infant. We acknowledge that our measures of
environmental stimulation might be insufficient.
A causal relationship between breastfeeding and
neurodevelopment has major public health implica-
tions. Future research should therefore explore causality
and try to reduce the role of genetic and environmental
confounding even further.
Acknowledgements.—We are indebted to Professor Niels Jørgen Secher
and Morten Hedegaard for their key roles in the initiation of the cohort,
and to the health visitors of the municipality of Aarhus for meticulous
data collection. The study was supported by grants from the Danish
Medical Research Council (j.nr. 12–1663), the Health Fund, Aarhus
University Research Fund, the Egmont Fund and the Municipality of
Aarhus.
References
1. Morgane PJ, Austin LaFrance R, Bronzino J, et al. Prenatal
malnutrition and development of the brain. Neurosci Biobehav
Rev 1993; 17: 91–128
2. Smart JL. Vulnerability of developing brain to undernutrition.
Ups J Med Sci 1990; 48 Suppl: 21–41
3. Klein PS, Forbes GB, Nader PR. Effects of starvation in infancy
(pyloric stenosis) on subsequent learning abilities. J Pediatr
1975; 87: 8–15
4. Dobbing J, Sands J. Timing of neuroblast multiplication in
developing human brain. Nature 1970; 226: 639–40
5. Dobbing J. Undernutrition and the developing brain. The
relevance of animal models to the human problem. Bibl Nutr
Dieta 1972; 17: 35–46
6. Lucas A, Morley R, Cole TJ, Lister G, Leeson Payne C. Breast
milk and subsequent intelligence quotient in children born
preterm. Lancet 1992; 339: 261–4
7. Lanting CI, Fidler V, Huisman M, Touwen BC, Boersma ER.
Neurological differences between 9-year-old children fed breast
milk or formula milk as babies. Lancet 1994; 344: 1319–22
8. Florey CDuV, Leech AM, Blackhall A. Infant feeding and
mental and motor development at 18 months of age in first born
singletons. Int J Epidemiol 1995; 24: S21–6
9. Rogan WJ, Gladen BC. Breast-feeding and cognitive develop-
ment. Early Hum Dev 1993; 31: 181–93
10. Morrow Tlucak M, Haude RH, Ernhart CB. Breastfeeding and
cognitive development in the first 2 years of life. Soc Sci Med
1988; 26: 635–9
11. Pollock JI. Long-term associations with infant feeding in a
clinically advantaged population of babies. Dev Med Child
Neurol 1994; 36: 429–40
12. Niemelä A, Järvenpää AL. Is breastfeeding beneficial and
maternal smoking harmful to the cognitive development of
children? Acta Paediatr 1996; 85: 1202–6
13. Temboury MC, Otero A, Polanco I, Arribas E. Influence of
breast-feeding on the infant’s intellectual development. J Pediatr
Gastroenterol Nutr 1994; 18: 32–6
14. Fergusson DM, Beautrais AL, Silva PA. Breast-feeding and
cognitive development in the first seven years of life. Soc Sci
Med 1982; 16: 1705–8
15. Silva PA, Buckfield P, Spears GF. Some maternal and child
developmental characteristics associated with breast feeding: a
report from the Dunedin Multidisciplinary Child Development
Study. Aust Paediatr J 1978; 14: 265–8
1332 A Vestergaard et al.
16. Gale CR, Martyn CN. Breastfeeding, dummy use, and adult
intelligence. Lancet 1996; 347: 1072–5
17. Jacobson SW, Jacobson JL. Breastfeeding and intelligence
[letter]. Lancet 1992; 339: 926
18. Jorgensen MH, Hernell O, Lund P, Holmer G, Michaelsen KF.
Visual acuity and erythrocyte docosahexaenoic acid status in
breast-fed and formula-fed term infants during the first four
months of life. Lipids 1996; 31: 99–105
19. Agostoni C, Trojan S, Bellù R, Riva E, Giovannini M.
Neurodevelopmental quotient of healthy term infants at 4 months
and feeding practice: the role of long-chain polyunsaturated fatty
acids. Pediatr Res 1995; 38: 262–6
20. Makrides M, Neumann M, Simmer K, Pater J, Gibson R. Are
long-chain polyunsaturated fatty acids essential nutrients in
infancy? Lancet 1995; 345: 1463–8
21. Neuringer M, Anderson GJ, Connor WE. The essentiality of n-3
fatty acids for the development and function of the retina and
brain. Ann Rev Nutr 1988; 8: 517–41
22. Farquharson J, Cockburn F, Patrick WA, Jamieson EC, Logan
RW. Infant cerebral cortex phospholipid fatty-acid composition
and diet. Lancet 1992; 340: 810–3
ACTA PÆDIATR 88 (1999)
23. Willatts P, Forsyth JS, DiModugno MK, Varma S, Colvin M.
Effect of long-chain polyunsaturated fatty acids in infant formula
on problem solving at 10 months of age. Lancet 1998; 352: 688–
91
24. Donovan SM, Odle J. Growth factors in milk as mediators of
infant development. Ann Rev Nutr 1994; 14: 147–67
25. Uauy R, De Andraca I. Human milk and breast feeding for
optimal mental development. J Nutr 1995; 125: S2278–80
26. Hedegaard M, Henriksen TB, Sabroe S, Secher NJ. Psychologi-
cal distress in pregnancy and preterm delivery. Br Med J 1993;
307: 234–9
27. Illingworth RS. The development of the infant and young child.
Edinburgh: Churchill Livingstone, 1987
28. Holle B. Motorisk—perceptuel udvikling. Copenhagen: Munks-
gaard, 1977
29. Greenland S. Modeling and variable selection in epidemiologic
analysis. Am J Public Health 1989; 79: 340–9
30. Rothman KJ. Modern epidemiology. Boston: Little, Brown, 1986
Received Aug. 31, 1998; revision received Nov. 27, 1998 and June
28, 1999; accepted July 21, 1999