Content uploaded by Henning Tiemeier
Author content
All content in this area was uploaded by Henning Tiemeier on Dec 16, 2013
Content may be subject to copyright.
Post-term birth and the risk of behavioural
and emotional problems in early childhood
Hanan El Marroun,
1,2
* Mijke Zeegers,
1,3
Eric AP Steegers,
4
Jan van der Ende,
1
Jacqueline J Schenk,
5
Albert Hofman,
6
Vincent WV Jaddoe,
2,6,7
Frank C Verhulst
1
and Henning Tiemeier
1,6,8
1
Department of Child and Adolescent Psychiatry, Erasmus MC—Sophia, Rotterdam, The Netherlands,
2
The Generation R Study
Group, Erasmus MC, Rotterdam, The Netherlands,
3
Pallas Health Research and Consultancy, Rotterdam, The Netherlands,
4
Department of Obstetrics and Gynecology, Erasmus MC, Rotterdam, The Netherlands,
5
Department of Psychology, Erasmus
University Rotterdam, The Netherlands,
6
Department of Epidemiology, Erasmus MC, Rotterdam, The Netherlands,
7
Department of
Pediatrics, Erasmus MC, Rotterdam, The Netherlands and
8
Department of Psychiatry, Erasmus MC, Rotterdam, The Netherlands
*Corresponding author. Department of Child and Adolescent Psychiatry, Erasmus MC—Sophia, P.O. Box 2060, 3000 CB Rotterdam,
The Netherlands. Email: h.marrounel@erasmusmc.nl
Accepted 23 February 2012
Background Post-term birth, defined as birth after pregnancy duration of
42 weeks, is associated with increased neonatal morbidity and mor-
tality. The long-term consequences of post-term birth are unknown.
We assessed the association of post-term birth with problem behav-
iour in early childhood.
Methods The study was performed in a large population-based prospective
cohort study in Rotterdam, The Netherlands. Pregnant mothers en-
rolled between 2001 and 2005. Of a cohort of 5145 children,
382 (7%) were born post-term, and 226 (4%) were born preterm.
Parents completed a standardized and validated behavioural check-
list (Child Behavior Checklist, CBCL/1.5–5) when their children
were 1.5 and 3 years old. We examined the relation between ges-
tational age (GA) at birth, based on early fetal ultrasound examin-
ation, and problem behaviour with regression analyses, adjusting
for socio-economic and pregnancy-related confounders.
Results A quadratic relationship between GA at birth and problem behav-
iour indicates that both preterm and post-term children have higher
behavioural and emotional problem scores than the term born chil-
dren. Compared with term born children, post-term born children
had a higher risk for overall problem behaviour [odds ratio
(OR) ¼2.10, 95% confidence interval (CI) ¼1.32–3.36] and were
almost two and a half times as likely to have attention deficit /
hyperactivity problem behaviour (OR ¼2.44, 95% CI ¼1.38–4.32).
Conclusions Post-term birth was associated with more behavioural and emo-
tional problems in early childhood, especially attention deficit /
hyperactivity problem behaviour. When considering expectant man-
agement, this aspect of post-term pregnancy should be taken into
account.
Keywords Post-term birth, preterm birth, behavioural and emotional problems,
childhood
Published by Oxford University Press on behalf of the International Epidemiological Association
ßThe Author 2012; all rights reserved.
International Journal of Epidemiology 2012;1–9
doi:10.1093/ije/dys043
1
Int. J. Epidemiol. Advance Access published May 2, 2012
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
Introduction
Timely onset of labour is important for peri- and
post-natal health. Both preterm (<37 weeks of gesta-
tion) and post-term birth (542 weeks of gestation)
are associated with neonatal morbidity and mortal-
ity.
1–3
Local management protocols with regard to
elective caesarean delivery and labour induction
affect the prevalence of post-term birth. Overall,
labour induction before or at 42 weeks of gestation
has increased,
1
but post-term births still occur rela-
tively frequently (up to 5–10%), even in industrialized
countries.
3,4
Accurate pregnancy dating is critical to
the diagnosis of post-term births.
3,4
Routine use of
ultrasound to confirm pregnancy dating can decrease
occurrence of post-term birth.
5
Common risk factors
for post-term birth include obesity, nulliparity and
prior post-term birth and rare causes include placental
sulphatase deficiency (an X-linked recessive disorder
characterized by low estriol levels), fetal adrenal
hypoplasia or insufficiency and trisomy 16
and 18.
1,2,6,7
The long-term problems associated with preterm
birth, such as increased incidence of cerebral palsy,
sensory impairments and behavioural problems are
well known.
8
The studies investigating effects of
post-term birth have focused on the risks during preg-
nancy and delivery.
9
Post-term birth increased the
risk of neonatal encephalopathy and death during
the first year of life,
5,10
but the long-term conse-
quences are unclear. One of the few studies per-
formed found that post-term born infants did not
differ from controls at age 2 years regarding general
intelligence, physical milestones and illnesses.
11
However, a recent study using referral to a neurologist
or psychologist as indicator of developmental prob-
lems found that 13% of children born post-term had
a neurological or developmental disorder at the age of
5 years.
12
In this population-based prospective study, we hy-
pothesize that post-term birth is related to behav-
ioural and emotional problems in preschool children.
In order to examine the specificity of the association
between post-term birth and problem behaviour, we
examined specific behavioural and emotional prob-
lems including attention deficit / hyperactivity dis-
order problems (ADHD), affective problems and
pervasive developmental problems.
Materials and methods
This study was embedded within the Generation R
Study, a population-based cohort from fetal life on-
wards.
13
Briefly, pregnant women who were resident
in Rotterdam, The Netherlands, and whose delivery
dates were between April 2002 and January 2006,
were asked by their midwives and gynaecologists to
participate. In the post-natal follow-up of the
Generation R cohort, 7484 live born children and
their prenatally recruited mothers participated.
Post-natally, 38 children died. The remaining 7446
children were eligible for the study. Mothers of 477
children withdrew consent, and mothers of 410 chil-
dren gave restricted consent (i.e. no participation in
questionnaire studies). The remaining mothers of
6559 children gave full consent for post-natal
follow-up. We excluded twin pregnancies, leaving
6422 children who could be contacted. Information
on child behavioural and emotional problems at 18
and/or 36 months was available for 5145 children (re-
sponse rate of 78%). Maternal report at both 18 and
36 months was available for 3840 children, 812
mothers reported at 18 months only and 493 mothers
reported at 36 months only. The Medical Ethical
Review Board of the Erasmus Medical Centre,
Rotterdam approved the study protocol. All parents
of participating children gave written informed
consent.
Ultrasound during the first visit determined gesta-
tional age (GA) to the nearest day, which will be ex-
pressed in our analyses in weeks. In total, 4132
women (80%) had their first ultrasound examination
in early pregnancy (median 13.1 weeks, range 5.1–
18.0), 868 women (17%) had it in mid-pregnancy
(20.4 weeks, 18.1–25.0) and only 145 women (3%)
had their first ultrasound examination in late preg-
nancy (30.2 weeks, 25.1–39.2). Crown–rump length
was used for pregnancy dating until a GA of 12
weeks and 5 days (crown–rump length <65 mm),
and biparietal diameter was used for pregnancy
dating thereafter (GA from 12 weeks and 5 days on-
wards, biparietal diameter 423 mm). Methods for es-
tablishing GA and standard ultrasound planes for
fetal measurements have been described previously.
14
Inter- and intra-observer intra-class correlation coef-
ficients were all 40.98.
14
Preterm birth was defined as birth before 37 weeks
gestation (N¼226) and post-term birth was defined
as birth after 542 weeks gestation (N¼382). As an
additional comparison group, we defined a group of
children born before 35 weeks of gestation (N¼78)
which is normally included in cohorts of preterm
babies.
The Child Behavior Checklist for toddlers (CBCL/
1.5–5) was used to obtain standardized parental re-
ports of children’s behavioural and emotional prob-
lems.
15,16
The CBCL was a postal questionnaire and
sent to be filled out by the mother when the child
was 18 months old and again when the child was 36
months old. At 36 months of age, we also asked the
father to fill out the CBCL. Each item (99 items in
total) is scored on a three-point scale (0 ¼not true,
1¼somewhat or sometimes true and 2 ¼very true or
often true), based on the child’s behaviour during the
preceding 2 months. The sum of all problem items is
the Total Problems score. There are five Diagnostic
and Statistical Manual of Mental Disorders (DSM)-
oriented scales: anxiety problems, affective problems,
2INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
pervasive developmental problems, ADHD and oppos-
itional defiant problems. It has been shown that these
DSM-oriented scales provide accurate and supplemen-
tary information on clinical diagnoses.
17
Also, good
reliability and validity have been reported for the
CBCL.
16
We used the clinical cut-off scores (91st per-
centile for the Total Problems score and 98th percent-
ile for the syndrome scales) to classify children as
having behavioural problems in the clinical range.
17
When parents filled out the questionnaire, they were
not aware of our research question exploring the re-
lation between post-term birth and behavioural prob-
lems, but parents generally are aware of the GA of
their child and the risks associated with preterm
birth. The maternal CBCL Total Problems ratings at
18 months and 36 months were correlated (r¼0.58,
P< 0.001). Maternal and paternal CBCL Total
Problems ratings at 36 months were correlated
(r¼0.56, P< 0.001).
Several covariates were considered in the analyses
and were chosen based on the existing literature
and effect estimate changes. Maternal weight and
height were measured at intake. We used postal ques-
tionnaires to obtain information on mother’s parity,
ethnicity and family income. Maternal ethnicity was
defined according to the classification of Statistics
Netherlands.
18
Educational level was categorized into
three levels: primary, secondary and higher educa-
tion.
19
Information about maternal smoking and alco-
hol use was obtained by questionnaires in each
trimester. Based on these questionnaires, maternal
smoking or drinking were categorized into ‘no’,
‘until pregnancy was known’ and ‘continued during
pregnancy’ as described previously.
20
The Brief
Symptom Inventory (BSI) was used to assess mater-
nal psychopathology in mid-pregnancy; the BSI is a
validated self-report questionnaire, which defines a
spectrum of psychiatric symptoms.
21
Registries pro-
vided information on obstetric variables such as in-
duction, birthweight, mode of delivery, umbilical
artery pH and Apgar scores after 1 and 5 min. The
post-natal questionnaire administered at age 6 and
12 months was used to gather information on breast-
feeding and frequency of day care use.
For descriptive analyses, children were categorized
in three groups based on GA: (i) born after 37 weeks
of gestation up to and including 41 weeks and 6 days
(term, reference group); (ii) born after <37 weeks of
gestation (preterm); and (iii) born after 42 þ0 weeks
of gestation or more (post-term). Chi-square and
t-tests were used to compare maternal and child char-
acteristics. To test the associations between GA and
behavioural problems, we used linear regression
models with GA as a continuous variable. We used
the generalized estimating equation (GEE) to analyse
the relation of GA with the behavioural and emo-
tional outcomes measured at different time points.
GEE adjusts for auto-correlation within the same
subject. We used an unstructured correlation matrix,
and thus no assumptions were made about the cor-
relations. The GEE procedure provides a more precise
effect estimate and reduces the error derived from
multiple comparisons (Type I error). A quadratic
term was added to the linear regression models to
test whether the associations between GA and behav-
ioural problems were curvilinear. We conducted the
primary analyses in all children, thus also including
the children whose GA was assessed in the second
and third trimester. This was done to reduce the
risk of potential selection bias. Furthermore, we
reran the analyses including only those children
with a GA dating in early pregnancy (N¼4132), be-
cause GA dating by ultrasound is assumed to be more
accurate in early pregnancy.
5
Moreover, we performed linear regression analyses
for maternal ratings at 18 and 36 months separately
to assess whether the quadratic association between
GA at birth and child total problems was present at
both ages. We also performed the same linear regres-
sion analyses for paternal ratings at 36 months. The
results of these analyses can be found in the
Supplementary data, available at IJE online.
Both linear and quadratic analyses were rerun after
exclusion of the preterm children, to ascertain that
the relationship between GA and behavioural prob-
lems was not solely driven by the preterm children.
In addition, we restricted the analyses to the children
born after 39 weeks of gestation. To check whether
results were not unduly influenced by ethnicity, we
reran analyses regarding the Total Problems score
including only the indigenous Dutch children.
Gender-specific estimates for the quadratic association
between GA at birth and child behavioural problems
are provided in the Supplementary Table S1, available
as Supplementary data at IJE online.
For logistic regression analyses, scales were dichot-
omized using the clinical cut-offs. We further
explored the nature of the association between GA
and behavioural problems with the GEE approach,
and calculated the odds ratios (ORs) of clinical prob-
lem behaviour for pre- and post-term born children.
We analysed only those scales on which 40.5% of the
participants were classified as having clinical prob-
lems; these were the ADHD, affective problems and
pervasive developmental problems scales.
Potential confounders were chosen based on the lit-
erature and effect estimate changes. Both linear and
logistic regression models were adjusted for child
gender, maternal age, education, ethnicity, psycho-
pathology, smoking and drinking during pregnancy,
family income and age of the child at the time of
assessments of the CBCL. Maternal weight, height,
parity, breastfeeding and day care did not change
the effect estimates (<5%). Maternal age, psycho-
pathology and age of the child were used as continu-
ous variables. Maternal education, ethnicity, smoking,
drinking and family income were used as categorical
variables in the analyses.
POST-TERM BIRTH AND CHILD PROBLEM BEHAVIOUR 3
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
Several post-hoc analyses were performed,
including only post-term children without induction,
without assisted extraction or no high birthweight
(44000 g) to test if effects were driven by these
birth characteristics.
Not all variables were available for each participant,
the mean proportion of missing values was 5.1% and
these were imputed. Variables were centred and miss-
ing data were imputed with the mean or, for categor-
ical variables, dummy variables were made. The
association between GA at birth and child behaviour
problems of the imputed and non-imputed data set
were compared, and these associations were similar.
Therefore, we only report results of analyses with the
imputed data.
For the non-response analysis, we compared mater-
nal and child characteristics of included partici-
pants with participants from whom we did not
obtain behavioural data. Non-responders were lower
educated (14.7% primary education vs 6.3%,
P< 0.001), younger (maternal age 28.1 vs 31.2
years, P< 0.001), more likely to be non-Dutch (62.0
vs 35.4%, P< 0.001). Excluded infants had a lower
birthweight (3313 vs 3431 g, P< 0.001) and were
born after a shorter period of gestation (39.5 vs 39.8
weeks, P< 0.001), compared with included infants.
The proportion of children born post-term was lower
in the non-response group than in the response group
(5.8 vs 6.9%, P< 0.001).
Results
Table 1 compares demographic and birth characteris-
tics of 5145 children of whom 4537 (88.2%) were
born at term, 382 were born post-term (7.4%) and
226 were born preterm (4.4%).
In Figure 1, the unadjusted associations between the
Total Problems scale and GA at age 18 and 36 months
are shown. The curves show a nadir of the Total
Problems score in children born with a GA around
40 weeks, whereas the mean problem scores are
higher in children who are born more preterm or
more post-term. In addition, we present a scatterplot
of the correlation between GA at birth and Total
Problems score in the Supplementary data, available
at IJE online.
For continuous scores on the total problems, ADHD,
affective problems scales and pervasive developmen-
tal problems, linear regression analyses showed a
curvilinear relation between GA and behavioural prob-
lems, indicating that children with shorter or longer
gestation had higher behavioural problem scores com-
pared with children born at term (Table 2). After exclu-
sion of the preterm born children, the curvilinear
relations between GA and behavioural problems re-
mained, showing that mean problem scores were
higher in children with a longer GA [Total Problems
score GA
2
¼0.34, 95% confidence interval (CI) ¼
0.14–0.54]. When we restricted the analyses to the
children born after 39 weeks of gestation (n¼4115),
we still observed a linear association between GA at
birth and total child behavioural and emotional prob-
lems (data not shown).
Supplementary analyses demonstrated that results
were similar when children with second or third tri-
mester GA dating were excluded (Total Problems
score GA
2
¼0.12, 95% CI ¼0.06–0.18). Moreover,
analyses on the Total Problems score were rerun in
a smaller subset of Dutch children; the results were
somewhat similar (GA
2
¼0.08, 95% CI ¼0.01–0.12).
In addition, there was no interaction between GA and
the two time points of CBCL measurement indicating
that effects remained stable over early childhood
(data not shown).
Separate linear regression analyses using the two
maternal ratings each demonstrated that the quad-
ratic association between GA at birth and child be-
havioural problems was present at 18 and 36 months
(Supplementary Table S2, available as Supplementary
data at IJE online). Analyses using the paternal rat-
ings of child behavioural and emotional problems also
demonstrated a quadratic relationship between GA at
birth and child emotional and behavioural outcomes
(Supplementary Table S3, available as Supplementary
data at IJE online).
Table 3 shows that post-term children were almost
twice as likely as term born children to have clinical
problem behaviour on the Total Problems scale and
were more likely to have problems in the clinical
range on the ADHD scale. Compared with term born
children, post-term children did not have higher ORs
on the affective problems or pervasive developmental
problems scales.
The preterm group showed considerably more be-
havioural or emotional problems compared with the
term group; this group was more likely to develop
problems on the Total Problems scale and the ADHD
scale. With a more stringent cut-off for preterm birth
(<35 weeks), we found that these children were more
likely to have problems in the clinical range on the
Total Problems scale, ADHD scale and the pervasive
developmental problems scale.
In addition, some groups of children were excluded
from the analyses. These exclusions did not change
the results. The risk for developing total problems in
children born post-term remained present after
excluding children with induction (OR ¼1.77, 95%
CI ¼1.01–3.10) or after excluding children 44000 g
of birth weight (OR ¼1.83, 95% CI ¼1.06–3.15).
Discussion
Our study demonstrated that children born post-term
were more likely than their term born peers to have
emotional and behavioural problems at both 18 and
36 months after birth.
Post-term delivery and behavioural problems could
be explained in several pathways. First, a larger baby
4INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
Table 1 Maternal and child characteristics in the study population
Term
a
N¼4537 Post-term
a
N¼382 Preterm
a
N¼226
Mean (SD) or % Mean (SD) or % P(to ref) Mean (SD) or % P(to ref)
Maternal characteristics
Age, mean years (SD) 31 (5) 31 (5) 0.34 31 (5) 0.55
Education level, %
Primary education 6.9 5.8 0.46 7.1 0.89
Secondary education 37.8 31.4 0.01 39.8 0.57
Higher education 52.0 59.2 <0.01 47.3 0.20
Ethnicity, %
Dutch 59.3 63.4 0.13 52.7 0.05
Other Western 8.6 10.5 0.22 8.0 0.90
Turkish, Moroccan 11.1 11.5 0.80 10.6 0.91
Surinamese, Antillean 8.1 5.9 <0.01 10.6 0.17
Psychopathology, mean score (SD) 0.25 (0.3) 0.23 (0.3) 0.43 0.28 (0.4) 0.15
Marital status, % single 9.4 7.8 0.35 14.8 0.01
Parity 41, % 12.2 10.2 0.15 8.4 0.09
Family income, %
<E1200 11.5 10.5 0.62 18.1 <0.01
4E1200 and <E2000 14.7 12.0 0.17 14.6 1.00
4E2000 63.4 67.5 0.11 53.5 <0.01
Smoking throughout pregnancy % 14.6 14.4 1.0 18.6 0.15
Alcohol use throughout pregnancy, % 46.8 52.9 0.03 44.9 0.62
Breastfeeding at 2 months, % 68.8 71.5 0.31 63.5 0.12
Child characteristics
GA, mean weeks (SD) 40.0 (1.1) 42.3 (0.3) <0.001 35.0 (2.0) <0.001
Range 37.0–41.9 42.0–43.4 26.7–36.9
Birthweight, mean grams (SD) 3481 (486) 3819 (456) <0.001 2403 (604) <0.001
Boys, % 49.1 58.4 <0.001 46.0 0.15
Suspected fetal distress, % 6.9 15.2 <0.001 13.5 <0.001
Hospital delivery, % 79.1 89.5 <0.001 96.4 <0.001
Labour induction, % 9.7 37.1 <0.001 12.9 0.16
Meconium in amniotic fluid, % 15.8 18.0 0.15 0.9 <0.001
Assisted delivery, % 8.6 30.6 <0.001 13.4 0.033
Caesarean delivery, % 11.6 18.2 <0.001 25.3 <0.001
Apgar 5 min <7, % 1.0 0.8 0.51 0.9 1.00
pH umbilical artery <7.1, % 3.8 3.3 0.45 2.9 0.68
Behavioural scores measured with CBCL
b
Total problems
Score at 18 months 23.53 (15.66) 24.47 (16.71) 0.29 28.35 (17.73) <0.001
Score at 36 months 20.52 (15.07) 22.29 (15.02) 0.04 24.10 (17.34) 0.002
ADHD
Mean score (SD) at 18 months 3.86 (2.45) 4.00 (2.45) 0.33 4.51 (2.56) <0.001
Mean score (SD) at 36 months 2.95 (2.32) 3.08 (2.31) 0.36 3.43 (2.63) 0.007
Affective problems
Mean score (SD) at 18 months 1.56 (1.69) 1.73 (1.90) 0.12 2.09 (1.91) <0.001
Mean score (SD) at 36 months 1.41 (1.58) 1.55 (1.62) 0.12 1.78 (1.76) 0.002
Pervasive developmental problems
Mean score (SD) at 18 months 1.73 (2.05) 1.86 (2.15) 0.28 2.27 (2.67) <0.001
Mean score (SD) at 36 months 1.99 (2.26) 2.23 (2.32) 0.06 2.46 (2.87) 0.006
P-values are derived from t-tests for continuous variables and chi-square tests for categorical variables.
a
Categorization was based on GA at birth, term (GA from 37 weeks up to 42 weeks), preterm (GA <37 weeks) and post-term (GA of 42 weeks or more).
b
Behavioural and emotional problem scores were measured using the CBCL reported by mother at 18 and 36 months.
POST-TERM BIRTH AND CHILD PROBLEM BEHAVIOUR 5
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
typically has a higher risk for perinatal problems.
Prolonged labour, cephalopelvic disproportion and
shoulder dystocia are increased in post-term chil-
dren.
2
A perinatal lack of oxygen has been associated
with behavioural problems.
22
However, our results did
not suggest increased fetal stress in the post-term
children, as indicated by low Apgar score, low umbil-
ical pH or meconium-stained amniotic fluid. We con-
trolled for several birth characteristics. Moreover,
exclusion of post-term children with induction and
44000 g of birth weight did not change results. A
second explanation is uteroplacental insufficiency: a
non-optimal ‘old’ placenta offers fewer nutrients and
less oxygen than a full term fetus requires.
1
The lack
of nutrients and oxygen may predispose to abnormal
fetal development and this may lead to abnormal
emotional and behavioural development.
23
In our
study, we could not distinguish possible effects of
uteroplacental insufficiency from perinatal problems.
Thirdly, it is possible that a disturbance of the
Figure 1 The unadjusted association between GA at birth and total behavioural and emotional problem score
Table 2 Association between GA at birth and behavioural and emotional problem score (continuous)
Total problems ADHD
Affective
problems
Pervasive
developmental
problems
(95% CI) (95% CI) (95% CI) (95% CI)
Model I
Linear model; GA 0.24 (0.48 to 0.01) 0.04 (0.08 to 0.01) 0.04 (0.06 to 0.10) 0.05 (0.09 to 0.02)
Quadratic model; GA 0.05 (0.21 to 0.32) 0.01 (0.05 to 0.04) 0.00 (0.03 to 0.06) 0.03 (0.06 to 0.01)
GA
2
0.13 (0.06 to 0.20) 0.01 (0.00 to 0.03) 0.02 (0.01 to 0.02) 0.01 (0.00 to 0.03)
Model II
Linear model; GA 0.15 (0.16 to 0.46) 0.01 (0.05 to 0.06) 0.08 (0.03 to 0.04) 0.02 (0.06 to 0.02)
Quadratic model; GA 0.06 (0.39 to 0.28) 0.02 (0.08 to 0.03) 0.01 (0.05 to 0.02) 0.05 (0.09 to 0.01)
GA
2
0.34 (0.14 to 0.54) 0.05 (0.02 to 0.08) 0.03 (0.01 to 0.05) 0.04 (0.02 to 0.07)
Behavioural and emotional problem scores were measured using the CBCL reported by mother at 18 and 36 months.
Model I: included all children.
Model II: included all children born after 37 weeks.
All models were adjusted for maternal age, education, ethnicity, psychopathology, smoking and drinking during pregnancy,
family income, gender of the child and age of the child at the assessment of the CBCL.
gives the estimate of increase in CBCL score per week increase of the centred GA. Bold values represent findings that were
considered statistically significant (P<0.05).
6INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
‘placental clock’, which controls the length of preg-
nancy, is involved. A marker of this clock is the pla-
cental secretion of corticotrophin-releasing hormone
(CRH), which is lower in women who deliver
post-term than in women delivering at term.
24
CRH
is the principal regulator of the maternal and fetal
hypothalamic–pituitary–adrenal (HPA) axis.
25
It has
been suggested that placental endocrine malfunction-
ing or maternal stress at critical times during fetal
development may influence the fetal HPA axis, lead-
ing to neuroendocrine abnormalities that could in-
crease the child’s vulnerability to emotional and
behavioural problems later in life.
26
Finally, the asso-
ciation between post-term birth and childhood behav-
ioural problems could be explained by underlying
causes of being born post-term. In other words, the
cause for post-term could also be the cause for having
behavioural problems, for example neurodevelopmen-
tal factors related to behavioural problems could be
involved in the complex process of birth.
This is a population-based study including many
post-term children. We measured problem behaviour
with the same validated instrument (CBCL/1.5–5) at
two time points. As ultrasound gestational dating is
thought to be superior to last menstrual period-based
gestational dating,
5
we decided to use primarily ultra-
sound dating. Eighty percent of our sample was dated
with ultrasound assessment in early pregnancy.
However, some limitations must be discussed.
Firstly, mothers were not formally blinded for the
GA of their children and they might perceive more
behavioural problems in post-term children.
However, the notion that a post-term birth may
signal at-risk babies is largely non-existent in the
medical profession and absent in the public debate.
Secondly, in the current study, we relied on the CBCL,
as it was not feasible to obtain clinical diagnoses in
such a large number of children. Moreover, these chil-
dren were too young to be assessed by teachers or
other informants, thus we had to rely on parental
ratings that may be biased. Moreover, the CBCL is
not a clinical instrument and cannot provide diag-
noses, but addresses continuous traits in children.
However, the DSM-oriented scales provide accurate
information
17
and good reliability and validity have
been reported.
16
Finally, although we controlled for a large number
of confounders, including maternal smoking, psycho-
pathology and socio-economic characteristics, residual
confounding, for example maternal malnutrition
during pregnancy, cannot be ruled out.
Management of prolonged pregnancy follows two
approaches: proposing induction before 42 weeks of
gestation or close monitoring of pregnancy after 41
weeks with selective induction in case of fetal distress
or a favourable Bishop score.
4
Pregnancy and peri-
natal care are criticized in The Netherlands, as peri-
natal mortality ranks as the third worst in Europe.
27
Until mid-2008, a woman with a low-risk pregnancy
at 42 weeks was referred to a gynaecologist for close
monitoring only. The current revised policy requires a
referral at 41 weeks. Although the rate of post-term
births went down after introducing first trimester
ultrasound dating of GA,
5
post-term delivery remains
common.
4
In conclusion, post-term children have a consider-
ably higher risk of clinically relevant problem behav-
iour. They are more than twice as likely as term born
children to have clinical ADHD. Further research is
needed to determine the causes of post-term birth
to reduce post-term birth rates and to minimize
long-term consequences. Also, longer follow-up is
Table 3 Association between GA at birth and behavioural and emotional problems (clinical cut-off)
Total problems ADHD
Affective
problems
Pervasive
developmental
problems
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Term birth N¼4537 1.0 1.0 1.0 1.0
Post-term birth N¼382 1.83 (1.17–2.85) 2.04 (1.18–3.55) 1.48 (0.88–2.51) 1.84 (0.95–3.55)
Preterm birth <37 weeks N¼226 2.35 (1.43–3.88) 2.28 (1.21–4.28) 1.51 (0.83–2.76) 1.83 (0.84–3.97)
Preterm birth <35 weeks N¼78 3.00 (1.48–6.09) 3.42 (1.41–8.32) 1.92 (0.74–4.99) 3.95 (1.47–10.6)
Prevalence of behavioural and emotional problems
Total
problems (%)
ADHD
(%)
Affective
problems (%)
Pervasive developmental
problems (%)
Problems at 18 months 4.2 2.5 3.5 1.8
Problems at 36 months 3.1 1.2 2.5 2.0
Behavioural and emotional problem scores were measured using the CBCL reported by mother at 18 and 36 months.
All models were adjusted for maternal age, education, ethnicity, psychopathology, smoking and drinking during pregnancy, family
income, gender of the child and age of the child at the assessment of the CBCL.
Categorization was based on GA at birth, term (GA from 37 weeks up to 42 weeks), post-term (GA of 42 weeks or more) and
preterm (GA <37 weeks or <35 weeks). Bold values represent findings that were considered statistically significant (P<0.05).
POST-TERM BIRTH AND CHILD PROBLEM BEHAVIOUR 7
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
necessary to establish whether the relationship be-
tween post-term birth and behavioural problems will
persist.
Supplementary Data
Supplementary Data are available at IJE online.
Funding
The Sophia Children’s Hospital Fund (project number
553) and the WH Kro
¨ger Foundation. The first phase
of the Generation R Study is made possible by
financial support from the Erasmus Medical Centre,
the Erasmus University and The Netherlands
Organization for Health Research and Development
(Zon MW, grant ZonMW Geestkracht 10.000.1003).
Acknowledgments
The Generation R Study is conducted by the Erasmus
Medical Centre in close collaboration with the School
of Law and Faculty of Social Sciences of the Erasmus
University Rotterdam, the Municipal Health Service
Rotterdam area, the Rotterdam Homecare
Foundation and the Stichting Trombosedienst &
Artsenlaboratorium Rijnmond (STAR), Rotterdam.
We gratefully acknowledge the contribution of general
practitioners, hospitals, midwives and pharmacies in
Rotterdam. H.E.M. checked the references used in this
article for accuracy and completeness. H.T. will act as
guarantor for the article. Someone with an excellent
mastery of the English language has carefully edited
the article. This article represents original material
and has not been published previously in whole or
in part. In addition, no similar paper is in press or
under review elsewhere.
Conflict of interest: F.C.V. is author and head of the
Department of Child and Adolescent Psychiatry at
Erasmus MC, which publishes the Achenbach
System of Empirically Based Assessment (ASEBA)
and from which he receives remuneration. All other
authors report no conflicts of interest.
KEY MESSAGES
GA at birth and behavioural and emotional problems in early childhood show a non-linear quadratic
relation indicating that both preterm and post-term children are at higher risk for problems.
Children born post-term were twice as likely as their term-born peers to have ADHD in early
childhood.
Our results suggest that children born post-term have a neurodevelopmental delay. However, further
research is needed to demonstrate a causal relation.
References
1
Caughey AB, Snegovskikh VV, Norwitz ER. Postterm
pregnancy: how can we improve outcomes? Obstet
Gynecol Surv 2008;63:715–24.
2
Norwitz ER, Snegovskikh VV, Caughey AB. Prolonged
pregnancy: when should we intervene? Clin Obstet
Gynecol 2007;50:547–57.
3
Shea KM, Wilcox AJ, Little RE. Postterm delivery: a chal-
lenge for epidemiologic research. Epidemiology 1998;9:
199–204.
4
Zeitlin J, Blondel B, Alexander S, Breart G. Variation in
rates of postterm birth in Europe: reality or artefact?
BJOG 2007;114:1097–103.
5
Savitz DA, Terry JW Jr, Dole N, Thorp JM Jr, Siega-
Riz AM, Herring AH. Comparison of pregnancy dating
by last menstrual period, ultrasound scanning, and their
combination. Am J Obstet Gynecol 2002;187:1660–66.
6
Caughey AB, Stotland NE, Washington AE, Escobar GJ.
Who is at risk for prolonged and postterm pregnancy? Am
J Obstet Gynecol 2009;200:683.e1–e5.
7
Roos N, Sahlin L, Ekman-Ordeberg G, Kieler H,
Stephansson O. Maternal risk factors for postterm preg-
nancy and cesarean delivery following labor induction.
Acta Obstet Gynecol Scand 2010;89:1003–10.
8
Allen MC. Preterm outcomes research: a critical compo-
nent of neonatal intensive care. Ment Retard Dev Disabil Res
Rev 2002;8:221–33.
9
Olesen AW, Westergaard JG, Olsen J. Perinatal and ma-
ternal complications related to postterm delivery: a na-
tional register-based study, 1978–1993. Am J Obstet
Gynecol 2003;189:222–27.
10
Mogren I, Stenlund H, Hogberg U. Recurrence of pro-
longed pregnancy. Int J Epidemiol 1999;28:253–57.
11
Shime J, Librach CL, Gare DJ, Cook CJ. The influence of
prolonged pregnancy on infant development at one and
two years of age: a prospective controlled study. Am J
Obstet Gynecol 1986;154:341–45.
12
Lindstrom K, Fernell E, Westgren M. Developmental data
in preschool children born after prolonged pregnancy.
Acta Paediatr 2005;94:1192–97.
13
Jaddoe VW, van Duijn CM, van der Heijden AJ et al. The
Generation R Study: design and cohort update 2010. Eur
J Epidemiol 2010;25:823–41.
14
Verburg BO, Mulder PG, Hofman A, Jaddoe VW,
Witteman JC, Steegers EA. Intra- and interobserver repro-
ducibility study of early fetal growth parameters. Prenat
Diagn 2008;28:323–31.
15
Reijneveld SA, de Kleine MJ, van Baar AL et al.
Behavioural and emotional problems in very preterm
8INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
and very low birthweight infants at age 5 years. Arch Dis
Child Fetal Neonatal Ed 2006;91:F423–28.
16
Achenbach TM, Rescorla LA. Manual for the ASEBA
Pre-School Forms and Profiles. Burlington, VT: Research
Center for Children, Youth and Families, University of
Vermont, 2000.
17
Nakamura BJ, Ebesutani C, Bernstein A, Chorpita BF. A
psychometric analysis of the child behavior checklist
DSM-oriented scales. J Psychopathol Behav 2009;31:178–89.
18
Statistics Nederland. Allochtonen in Nederland 2004.
Heerlen: Voorburg, 2004.
19
Statistics Nederland. Standaard onderwijsindeling 2003.
Heerlen: Voorburg, 2004.
20
Roza SJ, Verburg BO, Jaddoe VW et al. Effects of mater-
nal smoking in pregnancy on prenatal brain development:
the Generation R Study. Eur J Neurosci 2007;25:611–17.
21
Derogatis LR. Brief Symptom Inventory (BSI): Administration,
Scoring and Procedures Manual. 3rd edn. Minneapolis, MN:
National Computer Systems, 1993.
22
van Handel M, Swaab H, de Vries LS, Jongmans MJ.
Long-term cognitive and behavioral consequences of neo-
natal encephalopathy following perinatal asphyxia: a
review. Eur J Pediatr 2007;166:645–54.
23
Courchesne E, Townsend J, Chase C. Neurodevelopmental
principles guide research on developmental psychopathol-
ogies. In: Cicchetti D, Cohen DJ (eds). Developmental
Psychopathology. New York: Wiley, 1995, pp. 195–220.
24
McLean M, Bisits A, Davies J, Woods R, Lowry P,
Smith R. A placental clock controlling the length of
human pregnancy. Nat Med 1995;1:460–63.
25
Smith R. Parturition. N Engl J Med 2007;356:271–83.
26
Austin MP, Leader LR, Reilly N. Prenatal stress, the hypo-
thalamic–pituitary–adrenal axis, and fetal and infant
neurobehaviour. Early Hum Dev 2005;81:917–26.
27
Sheldon T. Perinatal mortality in Netherlands third worst
in Europe. BMJ 2008;337:a3118.
POST-TERM BIRTH AND CHILD PROBLEM BEHAVIOUR 9
at The University of British Colombia Library on May 3, 2012http://ije.oxfordjournals.org/Downloaded from
