The Impact of Maternal Depression in Pregnancy on Early Child Development

Article (PDF Available)inBJOG An International Journal of Obstetrics & Gynaecology 115(8):1043-51 · July 2008with396 Reads
DOI: 10.1111/j.1471-0528.2008.01752.x · Source: PubMed
Abstract
Postpartum depression in mothers is associated with developmental problems in their children. Many women who are depressed following childbirth are also depressed during pregnancy. The aim of this study was to examine the associations between maternal depressive symptoms during pregnancy and child development at 18 months of age. A prospective cohort study, Avon Longitudinal Study of Parents and Children. The former county of Avon, southwest England. All pregnant women in the defined area with delivery dates between April 1991 and December 1992, 9244 women and their children. Data were collected antenatally, at 18 and 32 weeks of gestation and at 8 weeks and 8 months postnatally, through postal questionnaires, including a self-report measure of depression (Edinburgh Postnatal Depression Scale [EPDS]). By the time their child was 18 months old, women completed five further questionnaires about their children's health and development. Child development at 18 months using a modified Denver Developmental Screening Test (modified DDST). Applying the standard 12/13 cutoff, 1565 (14%) women were depressed antenatally but not at either time-points postnatally. Employing the modified DDST, 893 (9%) children were developmentally delayed at 18 months of age. Persistent depression (EPDS > or = 10 at both time-points) is associated with developmental delay (adjusted OR 1.34, 95% CI 1.11-1.62). Applying the 12/13 and 14/15 cutoffs gave similar results. After further adjustment for postnatal depression, the effect sizes were slightly attenuated. These findings highlight the importance of depression in pregnancy. Some effects on child development attributed to postpartum depression are caused in part by depressive symptoms during pregnancy.

Figures

The impact of maternal depression in pregnancy
on early child development
T Deave,
a
J Heron,
b
J Evans,
c
A Emond
d
a
Centre for Child and Adolescent Health, University of the West of England, Bristol, UK
b
Department of Social Medicine, ALSPAC, University
of Bristol, Bristol, UK
c
Academic Unit of Psychiatry and
d
Centre for Child and Adolescent Health, University of Bristol, Bristol, UK
Correspondence: Dr T Deave, Centre for Child and Adolescent Health, University of the West of England, Bristol, Hampton House, Cotham Hill,
Bristol BS6 6JS, UK. Email toity.deave@uwe.ac.uk
Accepted 20 March 2008.
Objective Postpartum depression in mothers is associated with
developmental problems in their children. Many women who are
depressed following childbirth are also depressed during
pregnancy. The aim of this study was to examine the associations
between maternal depressive symptoms during pregnancy and
child development at 18 months of age.
Design A prospective cohort study, Avon Longitudinal Study of
Parents and Children.
Setting The former county of Avon, southwest England.
Population All pregnant women in the defined area with delivery
dates between April 1991 and December 1992, 9244 women and
their children.
Methods Data were collected antenatally, at 18 and 32 weeks of
gestation and at 8 weeks and 8 months postnatally, through postal
questionnaires, including a self-report measure of depression
(Edinburgh Postnatal Depression Scale [EPDS]). By the time their
child was 18 months old, women completed five further
questionnaires about their children’s health and development.
Main outcome measure Child development at 18 months using
a modified Denver Developmental Screening Test (modified
DDST).
Results Applying the standard 12/13 cutoff, 1565 (14%) women
were depressed antenatally but not at either time-points
postnatally. Employing the modified DDST, 893 (9%) children
were developmentally delayed at 18 months of age. Persistent
depression (EPDS 10 at both time-points) is associated with
developmental delay (adjusted OR 1.34, 95% CI 1.11–1.62).
Applying the 12/13 and 14/15 cutoffs gave similar results. After
further adjustment for postnatal depression, the effect sizes were
slightly attenuated.
Conclusions These findings highlight the importance of depression
in pregnancy. Some effects on child development attributed to
postpartum depression are caused in part by depressive symptoms
during pregnancy.
Keywords ALSPAC, antenatal depression, child development,
pregnancy, postnatal depression.
Please cite this paper as: Deave T, Heron J, Evans J, Emond A. The impact of maternal depression in pregnancy on early child development. BJOG 2008;115:
1043–1051.
Background
It is widely recognised that depression is common during the
postpartum period and that maternal postnatal depression
can have a detrimental effect on child development (the term
‘depression’ and ‘depressed’ is used as a short-hand for those
women who scored above the specified cutoff on the Edin-
burgh Postnatal Depression Scale [EPDS]. The EPDS is not
a diagnostic tool and the women had not been assessed for-
mally, therefore they had not been diagnosed as clinically
depressed).
1–7
Investigations of the inter-relationships
between maternal wellbeing and child development have
mostly been conducted after childbirth with few studies col-
lecting data prospectively, starting during pregnancy.
8–10
There is increasing evidence that the mother’s mood during
pregnancy is important. For example, uterine blood flow has
been observed to change in association with anxiety during
pregnancy,
11–15
and antenatal anxiety in late pregnancy is
independently associated with children’s behavioural/emo-
tional problems at 4 years of age.
16
The children of women
in socio-economically deprived families, who are at risk of
depression antenatally, are at greater risk of developmental
delay at 2 years of age.
17
The strongest predictor of postnatal
depression is depression in pregnancy; however, antenatal
depression may be more common than postnatal depres-
sion.
18,19
Although severity and chronicity of maternal depres-
sion are related to increased developmental problems in their
children,
20–22
less is known about the importance of timing of
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1043
DOI: 10.1111/j.1471-0528.2008.01752.x
www.blackwellpublishing.com/bjog
General obstetrics
the exposure to maternal depression, in particular whether
pregnancy is a sensitive period.
Closing the gap in the evidence about associations between
parental depression and child health outcomes has important
implications for both families and healthcare practitioners.
The aim of this study was to assess the association between
maternal depression in pregnancy and child development.
Although fathers’ mood influences child development,
23
the
focus of this paper is on depressive symptoms in mothers. We
hypothesised that antenatal depression scores have an impact
on early childhood development that is independent of post-
natal depression. This was tested, first, by investigating the
association between maternal antenatal depression and a mea-
sure of child development and second, by examining the
independent effects of antenatal depression. Data used to test
this hypothesis were derived from the Avon Longitudinal
Study of Parents and Children (ALSPAC),
24
a large commu-
nity sample in England that has been followed prospectively
since early pregnancy.
Methods
Samples
The ALSPAC study design included all pregnant women liv-
ing in the geographical area of Avon, England, who expected
to deliver their baby between April 1991 and December 1992,
resulting in 14 062 live births. Recruitment, dropout and
other methodologies have been described elsewhere.
24
Base-
line data, including socio-demographic and family details,
were collected from pregnant woman and their partners at
18 and 32 weeks of gestation using postal questionnaires.
Those women who did not complete the antenatal question-
naires at both 18 and 32 weeks of gestation were excluded
from these analyses, as were those children from multiple
births.
Measures
Exposures
The exposure of maternal depression was assessed using the
EPDS.
25
Although developed as a screening tool for depres-
sion following childbirth, this scale has been validated during
pregnancy as well as outside the postpartum period.
26,27
The
EPDS is a widely used 10-item self-rating questionnaire on
which women rate their feelings over the previous 7 days,
giving a score ranging from 0 to 30.
27
The EPDS has been
used in many studies of depression in childbearing women,
most typically in the first year postpartum. Women com-
pleted the EPDS at 18 and 32 weeks of gestation and then
again at 8 weeks and 8 months postpartum. Women were
categorised into three ‘antenatal depression’ groups: those
with a score below the EPDS cutoff indicating the absence
of depression and those with EPDS scores at or above the
cutoff indicating depression either at 18 or 32 weeks (once),
or at 18 and 32 weeks (twice); the latter as a measure of
more persistent depression. A priori, to reflect the continuous
nature of the data, there were three stages of the analysis
undertaken. Each stage applied a different cutoff on the
EPDS: 9/10, 12/13 and 14/15; 12/13 is the standard cutoff.
Post hoc, another method was also applied in which women
were categorised at both 18 and 32 weeks of gestation into
four groups depending on their EPDS scores (0–9, 10–12, 13–
14 and 15–30) and data were scored as given in Table 1,
creating an ordinal scale from 0 to 6. This method combined
both levels of intensity (EPDS scores) and persistence
(depressed once, twice or at both time-points during preg-
nancy) and aimed to better characterise any dose–response
relationship, or threshold effect, between persistent antenatal
depression and developmental delay.
Outcomes
A modification of the Denver Developmental Screening Test
(modified DDST) was used as the child outcome. This is
a screening questionnaire designed to identify cognitive and
behavioural problems in preschool children. The data are
presented as age norms. The more items a child fails to per-
form passed by 90% of his/her peers, the more likely the child
manifests a significant developmental deviation. The items
used in the screening test were from the Denver II that has
been shown to be predictive of developmental delay.
28
It had
been adapted for parental report through discussions with
focus groups and piloting, using the same cohort as this study
and found to significantly relate to the Griffiths Scales classi-
fication
29
for general development.
30
Parents completed a ques-
tionnaire at 18 months postpartum. The developmentally
delayed group were those who failed two or more items, which
was the case for 10% of the ALSPAC cohort.
Data analysis
Descriptive statistics and frequencies were used to examine
the characteristics of the study population. The three cutoffs
Table 1 . Categories allocated to each woman according to their
antenatal EPDS scores. A woman’s category at each time-point was
then summed to indicate the severity and persistence of her
depression
EPDS score at
18 weeks of
gestation
EPDS score at 32 weeks of gestation
EPDS 0–9 EPDS 10–12 EPDS 13–14 EPDS 15
EPDS 0–9 0 1 2 3
EPDS 10–12 1 2 3 4
EPDS 13–14 2 3 4 5
EPDS 15 3 4 5 6
Category 0, EPDS 0–9; category 1, EPDS 10–12; category 2,
EPDS 13–14; category 3, EPDS 15.
Deave et al.
1044 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
on the women’s EPDS scores (9/10, 12/13 and 14/15) were
applied at three different stages of the analysis. Bivariable tests
of association and multivariable regression analyses were
undertaken for each stage to quantify the individual relation-
ships between maternal depression, potential confounding
factors and developmental delay. Potential confounding fac-
tors entered into the analysis included demographic details
for both parents, previous maternal depression, maternal anx-
iety, paternal depression and anxiety, life events in the pre-
vious year, gestation, gender and ethnicity of child, feeding
method, postnatal mood of parents and life events postnatally
(Table 2). Individual unadjusted odds ratios were calculated
for each potential confounding and explanatory variable. For
the multivariable analysis, because of the complex inter-
relationship between antenatal depression and child deve-
lopment, a conceptual framework was used to describe the
hierarchical relationships between risk factors (Table 3).
31
This framework uses findings from previous research and
guides the decision about which variables to include in the
analysis. A multivariable regression analysis was undertaken
applying the steps of the conceptual framework one by one
starting with step 1. Following the inclusion of each step,
variables were removed if, individually, they had minimal
impact on the relationship between women with antenatal
depression and developmental delay (less than 5% change
in the odds ratio).
32
Those variables remaining in the model
were carried forward and included together with the variables
in the next step of the framework. Each variable was then
individually re-entered at the end. The evidence for any neg-
ative effect of young maternal age on developmental delay is
conflicting, therefore maternal age was retained.
33
This may
also be a proxy variable for some other factor that has not
been, or cannot be, measured. Due to missing data for the
exposure and for potential confounding variables, a sensitivity
analysis was undertaken using imputed data. In the multivari-
able models of our primary analyses, missing data on con-
founding variables resulted in a loss of almost 10% of the
sample when the fully adjusted models were derived. To avoid
any potential bias that might result when incorporating these
confounding variables into the model, a missing data impu-
tation technique was employed (missing imputation for
chained equations)
34
using the procedure in STATA known
as ice.
35
Imputation was restricted to confounding variables
(no imputation of the outcome variables was performed). The
results from the regression using imputed data were similar to
those using unimputed data. This suggests that the missing
data were not substantially biasing the findings and therefore
only the analysis using the unimputed data is reported. Data
Table 2. Details of the potential confounding factors that were
included in the conceptual framework and analysis
Potential
confounding
factors
Not depressed
women EPDS < 13
(n 5 8799),
n (%)*
Depressed women
EPDS 14
(n 5 2299),
n (%)*
Socio-economic
Primiparous women 3911 (44.5) 922 (40.1)
Owned/mortgaged housing 6943 (78.9) 1417 (61.6)
Overcrowding 1272 (14.5) 591 (25.7)
No use of car 1749 (19.9) 771 (33.5)
No smoking first trimester 7035 (79.9) 1464 (63.7)
No alcohol first trimester 3938 (44.8) 1038 (45.2)
Parental
Mother less than 20 years 234 (2.7) 146 (6.4)
Father less than 20 years 52 (0.6) 53 (2.6)
Mother: attained
O’levels/equivalent
3091 (35.3) 802 (35.1)
Father unemployed 495 (6.4) 233 (12.6)
Have partner 8372 (98.4) 2040 (95.6)
Preconceptional/pregnancy
Mother: previous depression 463 (5.3) 454 (19.7)
Mother: anxious 920 (10.5) 1562 (67.8)
Father: depressed 185 (2.7) 135 (8.2)
Father: anxious 274 (3.1) 118 (5.1)
,3 life events in past year 2135 (24.5) 189 (8.4)
Birth
Preterm 462 (5.3) 150 (6.5)
Ever breastfed 6163 (77.9) 1344 (70.3)
Male child 4510 (51.3) 1194 (51.9)
Ethnicity: white 8545 (97.1) 2183 (94.9)
Postnatal
Mother: no low mood 7176 (81.5) 1282 (55.9)
Father: no low mood 4378 (49.8) 876 (38.1)
,3 life events in past year 2456 (31.1) 1052 (54.7)
*These percentages represent the percentage response for the total
number of answers for that particular question.
Table 3. Conceptual hierarchical framework of risk factors for
child development
Step Label for steps
Antenatal factors
1 Socio-economic Parity, housing tenure, overcrowding,
use of car, mother’s smoking habit,
mother’s alcohol intake
2 Parental Mother’s age, partner’s age, mother’s
education, partner’s employment,
have partner
3 Preconceptional/
pregnancy
Previous maternal depression, maternal
anxiety, life events in previous year,
paternal depression and anxiety
Postnatal factors
4 Birth Gestation, breastfeeding, gender,
ethnicity of child
5 Postnatal Postnatal mood of parents, life events
postnatally
Modified from Victora et al.
31
Antenatal depression: impact on child development
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1045
analysis was performed using STATA (STATA version 8)
(Stata Corporation, College Station, TX, USA).
Ethics approval for the study was obtained from the
ALSPAC Law and Ethics Committee and the Local Research
Ethics Committees.
Results
Demographic characteristics
Complete antenatal EPDS data (18 and 32 weeks of gestation)
were available for 11 098 women, of whom 4833 (44%) were
expecting their first child. Women with incomplete antenatal
EPDS data were excluded from the study. There were few
demographic differences between these women and those
retained in the analysis. The main differences were that, com-
pared with the women who were excluded, a higher percent of
included women had a partner, a lower percent achieved
O’level or equivalent qualifications or had partners who were
unemployed (Table 4).
Child developmental outcomes
Parental report on the modified DDST (n = 10 125) identified
893 (9%) children who were developmentally delayed. The
children were mainly White British (97.7%), 612 (5.5%) were
preterm, 5704 (51.4%) were male children and 7505 (76.4%)
ever breastfed. There were just 36 (0.4%) children whose
parents reported that their child was usually unwell.
Antenatal depression
Using the standard 12/13 cutoff, 8262 (74.4%) women were
not depressed either antenatally or postnatally, 1565 (14.1%)
women were depressed on at least one occasion antenatally
but at neither postnatally. Five hundred and thirty-seven
(4.8%) women were depressed on at least one of the two
measurement occasions postnatally but at neither antenatally
(Table 5). There were just 156 (1.4%) women who were per-
sistently depressed antenatally and postnatally.
When combined with the modified DDST data, of the
11 098 women with complete EPDS scores, antenatal data
were available on 9244 (83%) women and children.
Multivariable analysis
For the first stage of the analysis, complete case analyses for
the EPDS 9/10 cutoff were applied. Those women whose
EPDS scores were <10 at both 18 and 32 weeks of gestation
formed the baseline comparison group. The final model pro-
vided adjusted estimates of the odds of developmental delay
of tobacco smoked in the first trimester, mothers’ age and life
events at 8 months. The odds ratio for developmental delay
associated with antenatal depression at both time-points was
1.24 (95% CI 1.04–1.49) and when adjusted for smoking,
maternal age and life events was 1.34 (95% CI 1.11–1.62).
For the second stage, the EPDS 12/13 cutoff was applied
and the results followed a similar pattern, with a 50% increase
in the odds of developmental delay (adjusted odds ratio 1.50;
95% CI 1.15–1.96) (Table 6). For the third stage, applying the
14/15 cutoffs gave similar results, although with wider and
nonsignificant confidence intervals. For each cutoff, after fur-
ther adjustment for postnatal depression, the effect sizes were
similar to those of the unadjusted models for each cutoff
(Table 6).
Applying the conceptual framework, paternal depression
and antenatal anxiety were included in the analysis. However,
as they both had minimal impact on the relationship between
antenatal depression and the odds of developmental delay,
they were removed from the model.
Post hoc analysis
In the post hoc analysis, an ordinal scale combining both
intensity (EPDS scores) and persistence (above EPDS cutoff
at both 18 and 32 weeks of gestation) of symptoms was
applied. The odds ratios and 95% CIs show a consistent pat-
tern between the multivariable models without and with
adjustment (Figure 1). There is a suggestion of a threshold
Table 4. Demographic descriptions for the women who were retained in the analysis and those who were excluded from the study
Included women
(n 5 11 098), n (%)
Excluded women
(n 5 3370), n (%)
Primiparous [missing data for variable] 4833 (44.3) [180 (1.6)] 971 (47.3.) [1317 (39.1)]
Average age (SD) 28 (4.8) 27 (5.3)
Age in years (range) 15–44 15–44
Achieved O’levels/equivalent [missing data for variable] 6002 (54.4) [60 (0.5)] 892 (63.2) [1958 (58.1)]
No partner [missing data for variable] 227 (2.1) [459 (4.1)] 63 (5.7) [2267 (67.3)]
Partner unemployed [missing data for variable] 728 (7.4) [1228 (11.1)] 215 (14.1) [1840 (54.6)]
Each variable have missing data and therefore percentages are presented for both missing data and for the variable in question. For the variable
itself (e.g. no partner) what is taken as 100% 5 11 098 (or 3370)—missing data. For ‘partner unemployed’, the data for no partner are excluded
from the missing data figures.
Deave et al.
1046 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
effect between the groups who scored 3 and 4 on the ordinal
scale, indicating an underlying effect of antenatal depression,
although none of the results are statistically significant and
the wide confidence intervals arise from the small numbers
involved, especially in the severe and persistent depression
groups.
Discussion
Summary of the study findings
This study set out to test the hypothesis that antenatal depres-
sion has an adverse impact on early childhood development.
This was tested, first, by investigating the association between
women’s antenatal depression and a child development
measure and second, by examining the effects of antenatal
depression independent of postnatal depression. Applying
the modified DDST at 18 months of age as a measure of
development, an association was found between persistent
depression during pregnancy and developmental delay with
a 50% increase in the odds of developmental delay associated
with persistent antenatal depression. While the presence of
postnatal depression had a modifying effect on the odds,
we found evidence of an independent and statistically signifi-
cant 34% increase in the odds of developmental delay when
the standard 12/13 cutoff on the EPDS was used to indicate
depression.
However, postnatal depression may be either a factor on
the causal pathway between antenatal depression and child
development, in which case it would be inappropriate to
adjust for its effects, or it may act as an independent influence
on child development, in which case it would be appropriate
to adjust for its effects. This study set out to examine specific
effects of antenatal depression on child development, inde-
pendent of postnatal depression, rather than to investigate
any potential link between the two. The observed threshold
effect is difficult to explain by confounding factors; moreover,
Table 5. Women’s depression scores during pregnancy (18 and 32 weeks of gestation) and postnatally (8 weeks and 8 months) using
the EPDS and the standard 12/13 cutoff (n = 11 098)
Antenatally PN EPDS <13,
n (%)*,***
PN EPDS at
8 weeks, n (%)*,***
PN EPDS at 8 months,
n (%)*,***
PN EPDS at 8 weeks
and 8 months, n (%)*,***
Total, n (%)**
AN EPDS , 13** 8262 (74.4) 247 215 75 8799 (79.3)
AN EPDS at 18 weeks** 518 69 55 33 675 (6.1)
AN EPDS at 32 weeks** 603 88 73 81 845 (7.6)
AN EPDS at 18 and 32 weeks** 444 103 76 156 779 (7)
Total 9827 (88.5) 507 (4.6) 419 (3.8) 345 (3.1) 11 098 (100)
*Percentages given relative to total sample.
**AN EPDS: antenatal EPDS scores.
***PN EPDS: postnatal EPDS scores.
Table 6. Unadjusted and adjusted odds ratios of the relationship between antenatal depression and developmental delay
Variables Unadjusted Adjusted* Adjusted* 1 PN depression**
OR (95% CI) P OR (95% CI) P OR (95% CI) P
Stage 1
EPDS 10 at 18 or 32 weeks 0.99 (0.84–1.19) 0.996 1.06 (0. 88–1.27) 0.557 1.11 (0.85–1.23) 0.819
EPDS 10 at 18 and 32 weeks 1.24 (1.04–1.49) 0.023 1.34 (1.11–1.62) 0.003 1.23 (1.00–1.51) 0.047
Stage 2
EPDS 12 at 18 or 32 weeks 1.04 (0.86–1.29) 0.641 1.14 (0.92–1.40) 0.231 1.07 (0.86–1.33) 0.530
EPDS 12 at 18 and 32 weeks 1.35 (1.04–1.76) 0.025 1.50 (1.15–1.96) 0.003 1.34 (1.01–1.78) 0.043
Stage 3
EPDS 14 at 18 or 32 weeks 1.16 (0.92–1.47) 0.218 1.26 (0.99–1.60) 0.065 1.16 (0.90–1.49) 0.245
EPDS 14 at 18 and 32 weeks 1.34 (0.91–1.98) 0.140 1.49 (1.00–2.20) 0.050 1.27 (0.84–1.92) 0.254
PN, postnatal.
*Adjusted for tobacco used in first 3/12 of pregnancy, mothers’ age, life events at 8 months.
**Adjusted for postnatal depression as well.
Antenatal depression: impact on child development
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1047
adjustment for suspected confounders did not markedly
change the estimates for the effect of cumulative exposure
to antenatal depression. So, although not statistically signifi-
cant, the consistent association between antenatal depression
and child development seems credible. While the link between
antenatal and postnatal depression is widely recognised,
8,9,18
aspects of both antenatal and postnatal depression in mothers
are likely to be important in relation to early child develop-
ment. Any persistent depression during pregnancy has the
potential to be an important risk factor for developmental
delay in childhood. On the basis of the threshold effect, these
findings suggest that, for those women who reach this thres-
hold during pregnancy, there is a risk of developmental delay
in early childhood.
Strengths and limitations
Strengths
This study uses a large prospective, longitudinal design and
recruited a large community sample that is broadly represen-
tative of the UK population when compared with the 1991
census.
24
Many potential confounding variables were con-
trolled for during the process of the analysis. However, all
but three were removed by the confounder selection strategy
for the final model. The evidence for an adverse affect of
young maternal age on child development is conflicting and
therefore maternal age was included.
33
It may also be a proxy
variable for some other factor that has not been, or cannot be,
measured. Nevertheless, there remains a possibility of residual
confounding of some unmeasured variables, in particular
mother–child interaction, a measure of which was not avail-
able for these analyses. This has been shown to be related to
both maternal depression and lead to poor cognitive devel-
opment in children.
22,36
A conceptual framework was used to manage the complex
hierarchical inter-relationship between antenatal depression
and child development and to describe the relationships be-
tween risk factors. The framework guided the multivariable ana-
lysis and was thus not based on purely statistical association.
31
To reflect the continuous nature of the data and to explore
the possibility of a dose–response effect between antenatal
depression and developmental delay, separate stages of the
analysis used different cutoffs on the EPDS: 9/10, 12/13 and
14/15. Using these three cutoffs do not necessarily deal with
the problem of misclassification of cases but rather shifts the
balance of false positives (women classified as having depres-
sion but who are not depressed) and false negatives (women
classified as not having depression but who are depressed) by
a particular cutoff. A post hoc analysis tried to address this by
creating an ordinal scale combining both intensity and per-
sistence of symptoms. Analysing the data in this way allowed
us to explore the possibility of a dose–response relationship to
be investigated further. However, underlying this method is
the assumption that the association with developmental delay
is the same for women in each category on the ordinal scale.
.7 .8 .9 1 1.2 1.5 2
Odds ratio
± 95% CI
Unadjusted Adjusted Adjusted + PN EPDS
0123456 0123456 0123456
EPDS 7 category scale
0 = No depressive symptoms.
6 = Greatest intensity and persistence of symptoms.
*Adjusted for tobacco used in first trimester of pregnancy, mothers’ age and life events at 8 months.
Figure 1. Association of antenatal depression (using an ordinal scale) and developmental delay (post hoc analysis). Results for the unadjusted
analyses, adjusted for confounding factors* and further adjusted for postnatal depression (PN EPDS) are illustrated.
Deave et al.
1048 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
Limitations
Despite the overall size of the cohort because of the relatively
small number of women with severe and persistent depres-
sion, our analysis has limited power to detect modest but
clinically important effects. The modified DDST is a binary
outcome and, by the nature of any population, there were few
women who experienced more severe depression that further
limits the power.
Maternal low mood may make completion of the modified
DDST difficult:
37
women who successfully completed the ques-
tions relating to their child’s development may be less likely to
experience depression than those nonrespondents. However,
this would probably result in an underestimation of any asso-
ciation. Moreover, the identification of those women with
depression was undertaken prior to the analysis of the out-
come. Typically, within large-scale studies, most of the data
relied on maternal (or parental) report. There is always the
potential for some effect of reporter bias, e.g. depressed
mothers not recognising their children’s abilities. However,
the ndings are broadly consistent with an earlier study that
looked at associations between women at risk of antenatal
depression and child development in a socio-economically
deprived population, independent of postpartum depression,
where child development was assessed using a trained assessor,
not by parental completion as in this study.
17
A sensitivity
analysis was undertaken imputing missing data, which sug-
gested missing data were not biasing the results.
Generalisability of the study findings
This was a large community-based population sample that
was followed-up for 18 months. The prevalence estimates
of probable depression, as assessed by a screening tool both
antenatally and postpartum, are similar to those reported in
other studies.
5,22,38
Comparison is difficult because few studies
have been undertaken on prospectively collected antenatal
depression data. Those that have, found no association with
antenatal depression and preschool children’s development
after adjusting for postpartum depression.
2,6
However, the
number of women with antenatal depression in these studies
was small. As mentioned above, the findings are broadly con-
sistent with an earlier study that looked at associations
between women at risk of antenatal depression and child
development in a socio-economically deprived population,
independent of postpartum depression.
17
Our results also fit
with other studies that reported associations between post-
partum depression and child development
5,8,22
and, using the
ALSPAC data, between antenatal anxiety and children’s emo-
tional and behavioural problems.
16
Our findings are also consistent with studies on rodents and
nonhuman primates, which indicate that maternal anxiety
during pregnancy can influence the developing fetus, resulting
in delay in motor and cognitive development and impaired
adaptation to stressful situations.
39
Moreover, studies have
shown a link between antenatal anxiety and cognitive, behav-
ioural and emotional problems in children, after adjusting for
postpartum factors.
13
These results support a fetal program-
ming hypothesis whereby biological systems adapt to environ-
mental input during sensitive periods of development, e.g. the
fetal period;
11
anxiety experienced by the mother has a direct
influence on the development of the hypothalamic–pituitary–
adrenal axis in the fetus.
12
Chronic maternal distress compro-
mises the normal regulations of hormonal activity during
pregnancy. Additionally, children of depressed mothers may
inherit directly a vulnerability to depression.
40
It has also been
suggested that the magnitude of the long-term effects of ante-
natal maternal anxiety in the child could be substantial.
41
Postpartum depression is a known risk factor for behav-
ioural/emotional problems in children, but it is less clear
whether postnatal depression leads to cognitive delay in chil-
dren unless it is prolonged or severe.
36
When postpartum
depression was included in these analyses, persistent antenatal
depression remained an important risk factor for develop-
mental delay and this is likely to be magnified for those
who might be diagnosed as clinically depressed. This implies
that this association is unlikely to be fully explained by raised
postnatal depression in those women who experienced ante-
natal depression. Therefore, some effects on child develop-
ment attributed to postpartum depression may be caused in
part by depression in pregnancy.
Implications of the study
These findings add to the growing body of research, suggest-
ing that maternal psychological wellbeing during pregnancy
has important consequences for child development. Women
who are between 15 and 44 years form a group in which
depression is the leading cause of disease burden worldwide.
42
Therefore, obstetricians, midwives and GPs can play an active
role in assessing and identifying maternal depression.
43–45
Family and personal histories of depression and postpartum
depression are known risk factors, and therefore, the use of
these by healthcare professionals may be helpful in identifying
women at risk. This is highly relevant for practice because
maternal mental health problems are associated with the
health of women’s partners, their children’s health, develop-
ment and use of health services.
46,47
Acknowledgements
We are extremely grateful to all the families who took part in
this study, the midwives for their help in recruiting them and
the whole ALSPAC team, which includes interviewers, com-
puter and laboratory technicians, clerical workers, research
scientists, volunteers, managers, receptionists and nurses.
The UK Medical Research Council, the Wellcome Trust and
the University of Bristol provide core support for ALSPAC.
Antenatal depression: impact on child development
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1049
This publication is the work of the authors and will serve as
guarantors for the contents of this paper. T.D. is funded by
a Higher Education Funding Council for England postdoc-
toral fellowship from the University of the West of England,
Bristol.
j
References
1 Murray L, Cooper P. Postpartum depression and child development.
Psychol Med 1997;77:253–60.
2 Murray L. The impact of postnatal depression on infant development.
J Child Psychol Psychiatr y 1992;33:543–61.
3 Stein A, Gath DH, Bucher J, Bond A, Day A, Cooper PJ. The relationship
between post-natal depression and mother-child interaction. Br J Psy-
chiatry 1991;158:46–52.
4 Hay DF, Kumar R. Interpreting the effects of mothers’ postnatal de-
pression on children’s intelligence: a critique and re-analysis. Child
Psychiatry Hum Dev 1995;253:165–81.
5 Sharp D, Hay D, Pawlby S, Schmucker G, Allen H, Kumar R. The impact
of postnatal depression on boys’ intellectual development. J Child
Psychol Psychiatry 1995;36:1315–36.
6 Cogill SR, Caplan HL, Alexandra H, Robson K, Kumar R. Impact of
maternal depression on cognitive development in young children.
Br Med J 1986;292:1165–7.
7 Dawson G, Ashman SB, Panagiotides H, Hessl D, Slef J, Yamada E, et al.
Preschool outcomes of children of depressed mothers: role of maternal
behavior, contextual risk, and children’s brain activity. Child Dev 2003;
74:1158–75.
8 Murray L, Cooper P. The role of infant and maternal factors in post-
partum depression, mother-infant interactions, and infant outcome. In:
Murray, L, Cooper, P, editors. Postpartum Depression and Child Devel-
opment. New York, NY: Guilford Press; 1997: pp. 111–35.
9 Green JM. Postnatal depression or perinatal dysphoria? Findings from
a longitudinal community-based study using the Edinburgh postnatal
depression scale. J Reprod Infant Psychol 1998;16:143–55.
10 Webster J, Linnane JWJ, Dibley LM, Hinson JK, Starrenburg SE, Roberts
JA. Measuring Social Support in Pregnancy: can it be simple and mean-
ingful? Birth 2000;27:97–101.
11 Barker DJ. In utero programming of chronic disease. Clin Sci 1998;95:
115–28.
12 Schneider M, Moore CF. Effect of prenatal stress on development:
a nonhuman primate model. In: Nelson C, editor. Minnesota Sympo-
sium on Child Psychology. New Jersey: Erlbaum; 2000; pp. 201–43.
13 Van den Bergh BR, Mulder EJ, Mennes M, Glover V. Antenatal maternal
anxiety and stress and the neurobehavioural development of the fetus
and child: links and possible mechanisms. A review. Neurosci Biobeh av
Rev 2005;29:237–58.
14 Weinstock M. The potential influence of maternal stress hormones on
development and mental health of the offspring. Brain Behav Immun
2005;19:296–308.
15 Teixeira JMA, Fisk NM, Glover V. Association between maternal anxiety
in pregnancy and increased uterine artery resistance index: cohort
based study. BMJ 1999;318:153–7.
16 O’Connor TG, Heron J, Golding J, Beveridge M, Glover V. Maternal
anxiety and children’s behavioural/emotional problems at 4 years. Br J
Psychiatry 2002;180:502–8.
17 Deave T. Maternal associations between child development and
women’s attitudes to pregnancy and motherhood. J Reprod Infant
Psychol 2005;32:1–14.
18 Llewellyn AM, Stowe ZN, Nemeroff CB. Depression during pre-
gnancy and the puerperium. J Clin Psychiatry 1997;58(Suppl 2):26–32.
19 Evans J, Heron J, Francomb H, Oke S, Golding J. Cohort study of
depressed mood during pregnancy and after childbirth. BMJ 2001;
323:257–60.
20 Brennan PA, Hammen C, Andersen MJ, Bor W, Najman JM, Williams
GM. Chronicity severity and timing of maternal depressive symptoms:
relationships with child outcomes at age 5. Dev Psychol 2000;36:
759–66.
21 Essex MJ, Klein MH, Miech R, Smider NA. Timing of initial exposure to
maternal major depression and children’s mental health symptoms in
kindergarten. Br J Psychiatry 2001;179:151–6.
22 Murray L, Fiori-Cowley A, Hooper R, Cooper PJ. The impact of
postnatal depression and associated adversity on early mother-
infant interactions and later infant outcome. Child Dev 1996;67:
2512–16.
23 Ramchandani P, Stein A, Evans J, O’Connor TG. Paternal depression in
the postnatal period and child development: a prospective population
study. Lancet 2005;365:2201–5.
24 Golding J, Pembrey M, Jones R, ALSPAC Study Team. ALSPAC—the
Avon Longitudinal Study of Parents and Children 1. Study methodol-
ogy. Paediatr Perinat Epidemiol 2001;15:74–87.
25 Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression:
development of the 10-item Edinburgh Postnatal Depression Scale.
Br J Psychiatry 1987;150:782–6.
26 Murray L, Carothers AD. The validation of the EPDS on a community
sample. Br J Psychiatry 1990;157:288–90.
27 Murray L, Cox JL. Screening for depression during pregnancy with the
Edinburgh Postnatal Depression Scale (EPDS). J Reprod Infant Psychol
1990;8:99–107.
28 Frankenburg WK, Dodds J, Archer P, Shapiro H, Bresnick B. The Denver
II: a major revision and restandardization of the Denver Developmental
Screening Test. Pediatrics 1992;89:91–7.
29 Griffiths R. The Abilities of Babies. New York, NY: McGraw Hill; 1954.
30 Emond A, Bell C, Heron J; The ALSPAC Study Team. Using parental
questionnaires to identify developmental delay (Letter). Dev Med Child
Neurol 2003;45:844–6.
31 Victora CG, Huttly SR, Fuchs SC, Olinto MTA. The concept of concep-
tual frameworks in epidemiological analysis: a hierarchical approach.
Int J Epidemiol 1997;26:224–7.
32 Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY:
Wiley, 1989.
33 Geronimus AT, Koreman S, Hillemeier MM. Does young maternal age
adversely affect child development? Evidence from cousin comparisons
in the United States. Popul Dev Rev 1994;20:585–609.
34 van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing
blood pressure covariates in survival analysis. Stat Med 1999;18:
681–94.
35 Royston P. Multiple imputation of missing values: update. Stata J 2005;
5:188–201.
36 Sohr-Preston SL, Scaramella LV. Implications of timing of maternal
depressive symptoms for early cognitive and language development.
Clin Child Fam Psychol Rev 2006;9:65–83.
37 Harding EK, Paul ES, Mende M. Cognitive bias and affective state.
Nature 2004;427:312.
38 O’Hara MW, Swain AM. Rates and risk of postpartum depression:
a meta-analysis. Int Rev Psychiatry 1996;8:37–54.
39 Buitelaar JK, Huizink AC, Mulder EJ, de Medina PG, Visser GH. Prenatal
stress and cognitive development and temperament in infants (Com-
ment). Neurobiol Aging 2003;24(Suppl. 1):S61–5; discussion S67–8.
40 De Weerth C, Buitleaar JK. Physiological stress reactivity in human
pregnancy—a review. Neurosci Biobehav Rev 2005;29:295–312.
41 Goodman SH, Gotlib IH. Risk of psychopathology in the children of
depressed mothers: a developmental model for understanding mech-
anisms of transmission. Psychol Rev 1999;106:458–90.
Deave et al.
1050 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
42 Murray C, Lopez A. The Global Burden of Disease. Cambridge, MA:
Harvard University Press; 1996.
43 NICE. Antenatal Care: routine care for the healthy pregnant woman.
London: National Collaborating Centre for Women and Children’s
Health, National Institute for Clinical Excellence; 2003.
44 NICE. Antenatal and postnatal mental health. London: National Col-
laborating Centre for Mental Health, National Institute for Health and
Clinical Excellence; 2006.
45 Hillard PJA. Well care for women. Curr Probl Obstet Gynecol Fertil
1996;21:159–75.
46 Andersson L, Sundstrom-Poromaa I, Wulff M, Astrom M, Bixo M. Impli-
cations of antenatal depression and anxiety for obstetric outcome.
Obstet Gynecol 2004;104:467–76.
47 Sills MR, Shetterly S, Xu S, Magid D, Kempe A. Association between
parental depression and children’s health care use. Pediatrics 2007;
119:829–36.
Antenatal depression: impact on child development
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1051
    • "Whilst areas of lower socioeconomic status (SES) have not been identified as a risk factor in all studies [5], they have been shown to contribute to the presence of a greater number of individual psychosocial risks [6] and consequently may result in increased vulnerability. Morbid consequences include a significantly increased risk of obstetric complications [1, 4,7891011 and post-natal depression in mothers [2, 12], as well as alterations in growth, development [11, 13], autonomic neuroendocrine function [4,141516 and mental health [4, 14, 17] in offspring. Treatment recommendations range from psychotherapy or anti-depressants for moderate cases to electroconvulsive therapy or a tricyclic combined with an anti-psychotic for depression with psychosis [18]. "
    [Show abstract] [Hide abstract] ABSTRACT: Depressed pregnant women face difficulty navigating a course between the potentially serious consequences of leaving depression untreated and significant limitations associated with conventional therapies, such as foetal toxicity and teratogenicity. Preliminary evidence is suggestive that acupuncture may provide a safe and effective alternative treatment option for antenatal depression; however, additional research is required. The purpose of this study is to further investigate this treatment possibility, with an additional examination of a potential biomechanistic acupuncture effect. In this pragmatic randomised controlled trial, we will compare individually tailored, flexible antenatal depression-oriented acupuncture with equivalent attention progressive muscle relaxation and routine antenatal depression hospital care. Eligible women at 24 weeks of gestation with Edinburgh Postnatal Depression Scale scores of 13 or more will be recruited from 2 antenatal clinics in South Western Sydney, Australia. The recruitment goal of 96 is powered to demonstrate a significant difference in Edinburgh Postnatal Depression Scale score severity between acupuncture and usual care, with intervention groups receiving weekly 1-h treatments for 8 weeks from 24 to 31 weeks of gestation. Mental health and quality-of-life assessments will occur at study commencement, intervention weeks 4 and 8 and 6 weeks post-natally via the collection of completed Edinburgh Postnatal Depression Scale scores, Depression, Stress and Anxiety Scale scores and World Health Organisation Quality of Life Scale scores. Adjustment to mothering will also be evaluated at 6 weeks post-natally using the Being a Mother Scale. A putative biomechanistic effect of acupuncture on the oxytocinergic system will additionally be examined by comparing baseline salivary hormone levels with those measured at intervention weeks 4 and 8, as well as leucocyte oxytocin receptor expression at baseline and intervention week 8. Ethical approval was received in February 2015, and recruitment is underway and expected to be completed in July 2016. Trial registration Australian New Zealand Clinical Trials Registry ACTRN12615000250538, Registered on 19 March 2015.
    Full-text · Article · Dec 2016
    • "Social stressors such as maternal mental state and stress level – which we refer to as " non-chemical toxicants " [24] – can also have a significant clinical impact on the developing fetus [25][26][27][28][29][30]. Women who experience toxic stress, anxiety, or depression while pregnant have children at elevated risk for certain physical, cognitive, and behavioral difficulties [28][29][30][31][32]. To our knowledge, no study has attempted to address whether concurrent prenatal exposure to chemical and non-chemical stressors interact to predict maladaptive temperament phenotypes. "
    [Show abstract] [Hide abstract] ABSTRACT: Background Temperament is a psychological construct that reflects both personality and an infant’s reaction to social stimuli. It can be assessed early in life and is stable over time Temperament predicts many later life behaviors and illnesses, including impulsivity, emotional regulation and obesity. Early life exposure to neurotoxicants often results in developmental deficits in attention, social function, and IQ, but environmental predictors of infant temperament are largely unknown. We propose that prenatal exposure to both chemical and non-chemical environmental toxicants impacts the development of temperament, which can itself be used as a marker of risk for maladaptive neurobehavior in later life. In this study, we assessed associations among prenatal and early life exposure to lead, mercury, poverty, maternal depression and toddler temperament. Methods A prospective cohort of women living in the Mexico City area were followed longitudinally beginning in the second trimester of pregnancy. Prenatal exposure to lead (blood, bone), mercury, and maternal depression were assessed repeatedly and the Toddler Temperament Scale (TTS) was completed when the child was 24 months old. The association between each measure of prenatal exposure and performance on individual TTS subscales was evaluated by multivariable linear regression. Latent profile analysis was used to classify subjects by TTS performance. Multinomial regression models were used to estimate the prospective association between prenatal exposures and TTS performance. Results 500 mother-child pairs completed the TTS and had complete data on exposures and covariates. Three latent profiles were identified and categorized as predominantly difficult, intermediate, or easy temperament. Prenatal exposure to maternal depression predicted increasing probability of difficult toddler temperament. Maternal bone lead, a marker of cumulative exposure, also predicted difficult temperament. Prenatal lead exposure modified this association, suggesting that joint exposure in pregnancy to both was most toxic. Conclusions Maternal depression predicts difficult temperament and concurrent prenatal exposure to maternal depression and lead predicts a more difficult temperament phenotype in 2 year olds. The role of temperament as an intermediate variable in the path from prenatal exposures to neurobehavioral deficits and other health effects deserves further study.
    Full-text · Article · Dec 2016
    • "The perinatal period is a sensitive time for children, with substantial neural, cognitive and socio-emotional developments taking place. Research indicates that parental mental health during this time can leave a far-reaching shadow (for reviews, see Deave et al, 2008; O'Connor et al, 2014et al, 2015). Like other mental health difficulties, those that occur during pregnancy and the postnatal period can range from mild depression and anxiety to more severe presentations, such as puerperal psychosis, and everything in between. "
    Full-text · Article · Aug 2016
Show more