Journal of Affective Disorders 131 (2011) 284–292

Contents lists available at ScienceDirect

Journal of Affective Disorders

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d

Research Report

Attitudes and adjustment to the parental role in mothers following

treatment for postnatal depression
Ming Wai Wan a,⁎, Deborah J. Sharp b, Louise M. Howard c, Kathryn M. Abel a
a
Psychiatry Research Group, University of Manchester, United Kingdom
b
Academic Unit of Primary Health Care, University of Bristol, United Kingdom
c
Institute of Psychiatry, King's College London, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Background: Few intervention studies of postnatal depression (PND) have evaluated
Received 17 June 2010 accompanying changes in parenting, in spite of mounting evidence that exposure to chronic
Received in revised form 7 October 2010 depression is detrimental to infant development. This study examined maternal attitudes and
Accepted 13 January 2011
adjustment over the first postnatal year within a treatment trial. The aim was to examine
Available online 23 February 2011
whether maternal adjustment improved with earlier remission, and with combined medical
and psychological treatment.
Keywords: Methods: As part of a multicentre pragmatic randomised controlled trial of treatment for PND,
Postpartum depression mothers completed a measure of maternal adjustment and attitudes and the Edinburgh
Maternal depression
Postnatal Depression Scale at an initial home visit (week 0) and three follow-ups (weeks 4, 18
Maternal adjustment
and 44).
Antidepressant
Counselling
Results: Maternal attitudes and adjustment improved with PND remission; earlier remission
conferred no additional benefit by 44-week follow-up. In line with previous studies, no
particular treatment modality (antidepressant or health-visitor delivered non-directive
counselling), or combination of treatments, was more effective for improving adjustment to
parenthood. However, the earlier start of antidepressant treatment provided a short-term
advantage for improving attitudes and reducing perceived stress.
Limitations: As a result of the study's pragmatic trial design, there was high treatment non-
compliance and no ‘pure’ control group. More depressed mothers may have been less likely to
complete the maternal adjustment and attitudes measure.
Conclusions: Effective treatment of PND is important not only for the mother's wellbeing but
also for healthy adjustment to parenthood. Provision of treatment choice and early
antidepressant treatment are suggested for optimising maternal attitudes and adjustment.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction ence postnatal depression (PND; Gaynes et al., 2005). There is

At no other time is someone so dependent on a caregiver
to meet their needs as in early infancy — for stimulation,
emotion regulation and optimisation of learning. However,
approximately 13% of newly delivered mothers will experi-
⁎ Corresponding author at: Community Based Medicine, Jean McFarlane
Building, University of Manchester, Oxford Road, Manchester, M13 9PL,
United Kingdom. Tel.: + 44 1 161 275 0731.
E-mail address: m.w.wan@manchester.ac.uk (M.W. Wan).
ample evidence that affected mothers are more likely to have
bonding difficulties (Brockington et al., 2001), increased
parenting-related stress and anxiety (Milgrom and McCloud,
1996), low parenting efficacy (Teti and Gelfand, 1991),
different interpretations of infant behaviour from objective
measures (Foreman and Henshaw, 2002), and more negative
interactions with their infant (Cornish et al., 2006; Nagata
et al., 2003; Righetti-Veltema et al., 2002; Rogosch et al.,
2004). These difficulties are reflected in the negative attitudes
of mothers with PND toward their children and adjustment

0165-0327/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jad.2011.01.009
 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292
to parenthood (Bernazzani et al., 2005; Webster et al., 1994).
Several prospective studies suggest that exposure to maternal
depression in the first year of a child's life is particularly
detrimental to their developmental progress (Essex et al.,
2001; Fihrer et al., 2009; Halligan et al., 2007; Hay et al., 2001,
2008; Murray, 1992; Murray et al., 1999). Treatment aims to
emphasise the need for early detection and intervention
(NICE Guidelines, 2007), one rationale being to prevent infant
problems.
Despite the relative ease with which maternal adjustment
and maternal attitudes can be measured, very few PND
intervention studies have evaluated changes related to
parenting. They have typically involved small samples and
have produced inconsistent results. Warner et al. (1997)
reported that in a postnatally depressed sample (N = 80),
negative cognitions related to motherhood were highly
associated with depressive symptoms, although cluster
analysis revealed a depressed subgroup (n = 35) without
such maladaptive attitudes. Among those who received
treatment (N = 41), negative cognitions decreased after
three months, concordant with symptom reduction and
irrespective of treatment type (drug/placebo or 1–6 counsel-
ling sessions). By contrast, in another sample (N = 44),
Fleming et al. (1992) reported that depressed mothers who
attended a support group described no differences in feelings
about parenthood compared with a depressed control group,
following intervention. However, the intervention did not
alleviate maternal depression either.
Other studies suggest that mothers with PND adjust to the
parenting role as mood is enhanced, irrespective of treatment
modality. Murray et al.'s (2003) randomised trial of counsel-
ling (N = 40), psychotherapy (N = 40) and behavioural
interaction guidance (N = 41) demonstrated a reduction in
parent-reported infant relationship problems as well as an
improvement in maternal mood in all treatment arms.
However, post-treatment relational difficulty was measured
only at four months, relatively early in the postpartum
period; and parenting attitudes may not necessarily be
enhanced when infant relationship problems diminish.
Indeed, such change did not translate to more sensitive
maternal behaviour in later observational assessments. The
only other PND trial to evaluate a parent-related variable
(Misri et al., 2006) reported that parenting stress was
alleviated by antidepressant treatment alone (N = 16) or
when combined with CBT (N = 19); mothers who remitted
reported lower post-treatment parenting stress than those
who did not. Moreover, at post-treatment, most of the stress
that mothers reported was related to the parenting role
rather than viewed as caused by the infant, which may be
more adaptive.
It remains unclear whether maternal adjustment to
parenting in the context of PND is improved by the effective
treatment of maternal mood and, if so, which treatment is
most effective at producing positive attitudinal change. We
attempted to address these questions within a large commu-
nity sample of mothers with moderately severe PND. The
RESPOND (Randomised Evaluation of antidepressants and
Support for women with POstNatal Depression) trial compared
the effectiveness of two of the most widely used treatments for
moderately severe PND, antidepressants and a community-
based psychosocial intervention (listening visits), in a multi-
285
centre, pragmatic randomised controlled trial. We examined
maternal adjustment to becoming a parent and attitudes to
parenthood at four time points during the first postnatal year.
The aims were: (1) to examine the effect of treatment
remission from PND, and the timing of remission, on maternal
adjustment to parenting and maternal attitudes at final follow-
up when the infant was one year; and (2) to compare the
relative effectiveness of the two treatments.
2. Methods
2.1. Sample and randomisation
The recruitment procedure is described in full in Sharp
et al. (2010). Eligible women were aged 18 years and over
who had recently delivered and were living with their new
baby aged b27 weeks old. Women were recruited between
January 2005 and August 2007 from 77 GP practices across
three UK centres (Bristol, Manchester and London) providing
a wide range of demographic backgrounds. Women with a
diagnosis of alcohol or drug abuse, psychosis or those already
receiving treatment for depression were excluded. Practices
sent out 10,666 invitations to eligible women and received
4173 (39%) non-duplicate responses with a completed
Edinburgh Postnatal Depression Scale (EPDS). Those who
scored 11+ (N = 989; 23%) were offered a ‘baseline assess-
ment’ home visit for trial eligibility. A diagnosis of PND was
confirmed if women scored 13+ on a further EPDS, and had a
diagnosis of depression as assessed by a computerised
version of the Clinical Interview Schedule — Revised (CIS-R;
Lewis, 1994). The home visit was usually conducted at about
8 weeks postpartum (2 weeks after initial EPDS screening),
although women could receive this visit up to 26 weeks
postpartum. At baseline (week 0), infants had a mean age of
11.54 weeks (SD: 4.50; range: 3.29–26.57 weeks). Of the 622
home visits conducted, 298 women scored b13 on the EPDS
and 54 scored N13 but did not have a CIS-R diagnosis, so were
ineligible, as was 1 woman whose baby was N26 weeks old.
Of the 269 eligible women, 7 were excluded by their GP/HV
and 8 declined randomisation. A total of 254 eligible women
were randomised and allocated to either antidepressants
(N = 129) or listening visits (N = 125). Socio-demographic
data from a self-report questionnaire showed some differ-
ences between the two randomised groups in terms of
diagnosis, number of children, breastfeeding and employ-
ment status (Sharp et al., 2010).
2.2. Interventions
Women randomised to antidepressants (ADs) were asked
to see their GP who prescribed according to a clinical practice
guideline based on the North of England Guidelines (Eccles
et al., 1999) and the British Association for Psychopharma-
cology 2000 Guidelines (Anderson et al., 2000). An SSRI was
recommended as first line treatment (fluoxetine if not breast-
feeding, sertraline, paroxetine, citalopram or escitalopram),
but a pragmatic approach was employed whereby the GP and
the patient agreed which medication would be prescribed.
The woman's GP and practice health visitor (PHV) were in-
formed of the AD group allocation and asked not to start any
form of psychological intervention aimed at alleviating the
286 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292
depressive symptoms. Treatment adherence information was
obtained through women's self-report at all follow-up points,
and by ascertaining prescribing information from her medical
record.
Women randomised to the psychological intervention,
listening visits (LVs), first underwent a four-week waiting
period comprising usual care (General Supportive Care; GSC)
from their GP and practice HV: i) as a comparison against
which to test the effectiveness of ADs; and ii) to replicate the
likely wait a woman might have for referral to a counsellor.
During this time, PHVs were asked not to provide any
psychological intervention aimed to alleviate depressive
symptoms. Experienced research health visitors (RHVs)
who were trained in the use of non-directive counselling for
postnatal depression delivered LVs in a series of up to eight
sessions in the woman's home.
LVs utilise a form of non-directive counselling similar to
that developed by Holden et al. (1989), originating from
client-centred psychotherapy (Rogers, 1951), aimed to
facilitate the expression of feelings in a respectful, empathic
and reflective manner (described in Turner et al., 2010). RHVs
contacted women two weeks after randomisation, initially
offering four LVs over 4–6 weeks, with the possibility of a
second set of four LVs over a further 4–6 weeks. Each visit was
typically one hour long, of which at least 30 min was active
listening. The RHV's role was exclusively to provide LVs that
focused on the woman's depression (discussion of the
practicalities of baby care was discouraged), while the PHV
continued to provide the family health visiting role.
2.3. Change to alternative intervention
The pragmatic element of the trial design allowed women
to receive the alternative intervention had they not
responded to their allocated intervention, or wished to
change to, or add in, the other intervention at any time
after four weeks. Women allocated ADs were reminded (as
was their GP) that they could also receive LVs (up to eight
visits) when the four-week assessment had been completed.
Women allocated LVs were able to visit their GP for ADs at
any time during the study, which was also discussed by the
RHV usually at the fourth LV, but GPs could not prescribe ADs
until 4 weeks unless absolutely necessary.
2.4. Measures
All questionnaires were administered during a home visit
at baseline by a research assistant, and were distributed by
post for self-completion at all subsequent follow-ups. Non-
responders at follow-up were visited at home where possible
to increase the response rate.
The Edinburgh Postnatal Depression Scale (EPDS), a 10
item self-report questionnaire (Cox et al., 1987), was used to
screen for postnatal depression and to monitor symptom
change at baseline and all follow-ups. The case definition for
probable depression was a score of ≥13 (Cox et al., 1987).
The Clinical Interview Schedule — Revised (CIS-R) is a
structured psychiatric assessment for 14 common symptoms
of depression and anxiety in the week before the interview
(Lewis, 1994). The self-administered computerised version has
been used widely (e.g. UK Psychiatric morbidity surveys —
Jenkins et al., 1997) and has an excellent agreement with the
interviewer administered version (Lewis, 1994).
The Maternal Adjustment and Maternal Attitudes Ques-
tionnaire — postpartum version (Kumar et al., 1984) is a
60-item instrument that assesses maternal attitudes during
pregnancy and after childbirth. Maternal attitudes to infant
were measured using the relevant 12-item sub-scale (atti-
tudes towards pregnancy and the baby, henceforth referred
to as the MAMA), which has demonstrated inter-rater
reliability (Kumar et al., 1984) and does not vary with
maternal age or parity (Windridge and Berryman, 1996).
Higher scores indicate more positive attitude or adjustment.
Low scores on the full instrument have been demonstrated to
predict higher symptomatology among depressed mothers
(Feldman et al., 2009).
2.5. Data analysis
2.5.1. MAMA data
Complete MAMA responses were obtained for most
women at week 0 and week 4 of the study period, but only
for N = 110 (43%) at all time points (Fig. 1). Incomplete or
unreturned MAMAs were excluded from total MAMA
analyses. MAMA non-completers had higher EPDS scores
at baseline (F[1, 252] = 5.51; p = 0.02) and 44 weeks (F[1,
252] = 5.16; p = 0.02) and, at 44 weeks, tended to be
younger (F[1, 216] = 8.82; p = 0.003), less likely to be living
with a partner (χ2 = 5.74; p = 0.02) and to be in employ-
ment (χ2 = 5.13; p = 0.02). MAMA non-completers at each
follow-up did not differ from completers in baseline MAMA,
except at 44 weeks when non-completers showed a trend
towards higher baseline scores (F[1, 240] = 3,52; p = 0.06).
To examine the relationship between the MAMA and EPDS,
Pearson correlations and a K-means cluster analysis were
conducted.
2.5.2. Data cleaning prior to main analyses
Prior to testing whether earlier remission (the time point
at which an EPDS score of b13 was obtained) led to better
maternal attitudes at 44 weeks, relapse cases (i.e. EPDS scores
≥12 following previous remission) were removed (N = 25;
9.8%). Similar to the non-relapsed group, MAMA scores in the
overall relapsed group improved at the first three time points,
but worsened at 44 weeks (repeated measures ANOVA
controlling for treatment arm; time: F[3, 321] = 8.68;
p b 0.001; relapse: F[1, 107] = 1.73; p N 0.05; time × relapse
interaction: F[3, 321] = 5.53; p = 0.001). The non-relapsed
sample was divided into 4 ‘remission’ groups (remission by:
4 weeks, 18 weeks or 44 weeks; and no remission). A
repeated-measures ANOVA was conducted with remission
group as the independent variable and 44-week MAMA as the
dependent variable. To minimise data loss for EPDS non-
completion (full EPDS data available at week 0), we assumed
‘no remission’ in missing EPDS data points that precede the
first recorded point of remission (if at all) and continued
remission in any subsequent missing data points.
Non-compliant cases (defined as having received b4 LVs
and/or reported (through questionnaire response or from GP
notes) at least ‘sometimes missed 1 dose’ of AD up to the time
point of interest) were removed before repeated-measures
ANOVA compared treatment effects on the MAMA. N = 85
 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292 287

Random allocation

Antidepressants (AD) Listening Visits (LV)

Randomised: N=129 Randomised: N=125
MAMA total: N=122 (94.57%) MAMA total: N=118 (94.40%)

Week 0: Woman commences AD prescription Week 0: General supportive care only until Week 4

4 Weeks 4 Weeks
MAMA total: N=97 (75.19%) MAMA total: N=106 (84.80%)

AD LV group LV only group LV AD group

AD only group
N=69 N=70 N=55
N=60 At any point, woman requests Four weekly visits, Woman requests ADs
LVs if required: Four weekly review, then 2nd block from GP if required
visits, review, then 2nd block of 4 visits if required.
of 4 visits if required.

18 Weeks

AD only AD LV LV only LV AD
MAMA total: N=33 (55.0%) N=59 (85.5%) N=55 (78.6%) N=46 (83.6%)

44 Weeks

AD only AD LV LV only LV AD
MAMA total: N=28 (46.67%) N=45 (65.22%) N=36 (51.4%) N=22 (70.9%)

Fig. 1. MAMA total completion rates throughout intervention and follow-up.

(33.5%) were ‘non-compliant’ at some point (the design
allowed women to switch treatment arm part way through
and thus not complete a full course of the randomised
treatment). Non-compliant women did not differ significant-
ly from their compliant counterparts on EPDS score at any
point. Separate analyses were conducted first for each time
point of comparison as a result of the staged study design and
to minimise sample size loss (due to higher rates of MAMA
non-completion at later time points). MAMA scores were
normally distributed at all time points.
2.5.3. Treatment group comparisons
At 0–4 weeks, the AD group was compared with GSC (i.e.
the ‘LV group’ pre-treatment); at 0–18 weeks and
18–44 weeks, the four treatment groups were compared (AD
only, AD→ LV, LV only, and LV→ AD). All analyses controlled
for baseline (week 0) MAMA, the point of randomisation
(subsequent treatment groups were to some extent self-
selected because of the pragmatic cross-over design). Infant
age (at latest measurement) was also adjusted for, because of
its correlation with early total MAMA (week 0: Pearson
r = 0.19; p = 0.003; week 4: r = 0.21; p = 0.003) and a
between-group (3–8 weeks (N= 24), 8–14 weeks (N= 68),
14+ weeks (N= 18)) trend of infant age (repeated-measures
ANOVA controlling for treatment arm: F[2, 106] = 2.38;
p = 0.1). Specifically, mothers of infants 14+ weeks at baseline
had slightly higher MAMA scores throughout. No other main
demographic effects emerged: socioeconomic status (profes-
sional-intermediate, semi-routine/routine; F[1, 77] = 0.01);
low wage (b£1000, £1000+ per month; F[1, 71]= 0.41);
employment status (F[1, 106]= 2.06); qualifications (none/
GCSE, further; F[1, 102] = 0.74); region/centre (Bristol, Man-
chester, London; F[2, 106]= 0.62); ethnicity (Caucasian, other;
F[1, 107] = 1.64); cohabitation status (F[1, 107] = 0.21);
marital status (married, unmarried; F[1, 101] = 0.01); primi-
parity (F[1, 107] = 0.04); or breastfeeding (F[1, 107]= 0.17).
Some factors interacted with time: mothers educated up to
GCSE (F[3, 306]= 3.16; p = 0.03) and those with a low income
288 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292

(F[3, 213] = 3.72; p = 0.01) started with better maternal
attitudes that worsened by 44 weeks. Married women (F[3,
303] = 2.93; p = 0.03) and women living with a partner (F[3,
321] = 3.27; p = 0.02) showed more improvement in their
MAMA score.
2.5.4. MAMA subscales and repeated measures analysis
To identify whether particular kinds of parent-related
attitude were more amenable to change, repeated-measures
ANOVA were conducted on MAMA subscales (maternal
feelings, parent stress, and parenting anxiety), after removing
missing data within the given subscale and controlling for
baseline subscale scores. The subscales were derived from
exploratory factor analyses (using principal component
analysis with eigenvalues N1 and Promax rotation method)
from which a ‘best fit’ across all time points was identified
(Table 1). In the parenting anxiety subscale, item 7 (the
appeal of having several children) was reverse-scored
because of its consistent negative loading relative to other
item loadings in the third factor. The original scoring was
retained for the total MAMA in line with previous validation
work (Kumar et al., 1984).
We also compared groups across all time points using
repeated-measures ANOVA — with a smaller sample because
of missing data and non-compliance. Remission (most recent
EPDS score b13) was added as an independent variable to
each total MAMA × treatment analysis to test whether the
effect of remission was stronger than that of treatment group/
allocation.
3. Results
3.1. Does PND remission and, in particular, earlier remission
lead to better maternal attitudes?
Overall maternal adjustment and attitudes to parenthood
improved with time (or treatment); mean total MAMA at 0, 4,
18 and 44 weeks respectively were 32.58 (SD: 5.36), 34.45
(SD: 5.81), 36.49 (SD: 5.00) and 37.47 (SD: 5.13). When EPDS
scores were correlated with MAMA scores, weak associations
emerged at baseline – likely to be an artefact of the narrow
EPDS score range (which were necessarily above clinical
threshold for eligibility) – but which increasingly strength-
ened with time (Table 2). A cluster analysis of the distribution
of EPDS and MAMA scores at 44 weeks identified 2 groups: a
‘persistent depression/low adjustment’ group (N = 47; mean
EPDS = 17; MAMA = 32.06), and a ‘remitted/high adjust-
ment’ group (N = 101; mean EPDS = 8; MAMA = 39.99).
Early remission from PND (at 4 weeks) conferred no
extra benefit to 44-week maternal attitudes over remission
at 18 or 44 weeks (Fig. 2). Although there were significant
between-group effects of remission time on total MAMA
score (F[3122] = 13.53; p b 0.001) and MAMA subscales
(maternal feelings: F[3125] = 6.96; p b 0.001; parent stress:
F[3124] = 6.00; p = 0.001; parenting anxiety: F[3123] =
7.02; p b 0.001), Bonferroni post-hoc tests revealed that
the differences were accounted for by the suboptimal
maternal attitudes of the ‘no remission’ group compared
with all other groups (p b 0.01). Thus, although maternal
adjustment improved with remission from PND, earlier
remission conferred no additional benefit at final (44-week)
follow-up.
3.2. Comparing maternal attitudes across treatment arms at
each follow-up
3.2.1. 0–4 weeks: AD and GSC
Maternal attitudes improved with time (F[1, 163]= 25.47;
p b 0.001), particularly in the AD group, although the interac-
tion effect did not reach statistical significance (F[1, 163]=
3.25; p = 0.07) (Fig. 3a). At MAMA subscale level, maternal
feelings improved over time in both groups (F[1 171]= 24.62;
p b 0.001) with ADs conferring a significant advantage (time × -
treatment allocation: F[1, 171]= 4.22; p = 0.04). ADs also

Table 1
Pattern matrix of rotated factors of MAMA at each time point a.

Factor Week 0 Week 4 Week 18 Week 44

1 2 3 1 2 3 1 2 3 4 1 2 3

Subscale 1: maternal feelings

4. Regretted having .67 .86 .60 .72
5. Proud mother .71 .76 75 .92
6. Happy have baby .84 .76 .86 .84
8. Disappointed by .61 .53 .55 .68
9. Enjoyed caring .74 .71 .91 .68
12. Enjoyed feeding .79 .67 .88 .77

Subscale 2: parent stress

3. Enough time .87 .78 .64 .82
11. Life difficult .61 .55 .77 .88

Subscale 3: parenting anxiety

1. Worry good .66 .62 .50 .41 .45
2. Worried hurt baby .51 .74 −.48 .92 .74
7. Several children .53 −.45 .75 .64 −.53 −.45
10. Healthy normal .73 .67 .85 .69
% of variance 33.62 8.63 11.75 38.39 12.07 8.55 34.53 11.25 8.97 8.39 37.67 13.28 9.30
a
Only factor loadings of above .40 were considered in determining the overall MAMA subscale.
 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292 289

Table 2 Treatment allocation

Pearson correlations: MAMA scores and depressed mood. AD
a LV
Week Total Maternal Parent Parenting 35.00
attitudes feelings stress anxiety

CIS-R
0 r −.14 ⁎ −.09 −.11 −.25 ⁎⁎ 34.50
N 240 246 252 248

MAMA score
34.00
EPDS:
0 r −.08 −.06 .02 −.13
N 240 246 252 248
33.50
4 r −.34 ⁎⁎ −.28 ⁎⁎ −.31 ⁎⁎ −.20 ⁎⁎
N 203 207 210 206
18 r −.46 ⁎⁎ −.33 ⁎⁎ −.36 ⁎⁎ −.23 ⁎⁎ 33.00
N 193 198 199 194
44 r −.63 ⁎⁎ −.45 ⁎⁎ −.51 ⁎⁎ −.31 ⁎⁎
N 148 152 151 149 32.50
⁎ p b 0.05.
0 4
⁎⁎ p b 0.005.
Study time point (weeks)

improved parenting stress (F[1, 177]= 5.67; p = 0.02), but not Treatmtment arm
parenting anxiety (F[1, 171] = 0.01; p N 0.05). When remission AD only AD to LV
LV only LV to AD
was adjusted for in the total MAMA analysis, the trend in b
treatment allocation disappeared (F[1, 161] = 0.48; p N 0.49). 39.00
Although the effect of remission was highly significant (F[1,
161]= 15.37; p b 0.001), there was no time × treatment× re- 38.00
mission interaction (F[1, 161] = 0.25).
37.00
MAMA score

3.2.2. 4–18 weeks: AD only, AD → LV, LV only, and LV → AD

Effects of treatment arm (F[3, 102]=2.63; p=0.05) and 36.00
time (F[1, 102]=26.13; pb 0.001) emerged, but did not interact
(F[3, 102]=0.95). All treatment groups showed improvement, 35.00
with the AD only group producing consistently higher MAMA
total scores even when controlling for baseline MAMA (Fig. 3b). 34.00
Subscale analyses showed that parenting stress particularly
decreased with time (F[1, 117]=6.20; p=0.01) as did parent 33.00

1 2
Time
Key
1. Remission by 4 weeks (n=35) Treatmtment arm
2. Remission by 18 weeks (n=37) AD only AD to LV
3. Remission by 44 weeks (n=21) LV only LV to AD
4. No remission at 44 weeks (n=33) c
40.00

39.00
40
MAMA score
MAMA follow-up score

38.00
30

37.00
20

36.00
10 1 2
Time

Fig. 3. Maternal attitude change (a) between 0 and 4 weeks: AD and GSC; (b)
1 2 3 4 between 4 and 18 weeks by treatment arm; and (c) between 0 and 44 weeks
Time point of remission by treatment arm.

Fig. 2. Maternal attitudes at 44 weeks according to time point of remission.

290 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292
anxiety to a lesser extent (F[1, 110]=3.66; p=0.06), although
in the latter with little change in the AD only group. By contrast,
maternal feelings became more negative (F[1, 109]=17.05;
pb 0.001), irrespective of treatment arm (treatment: F[3, 109]=
2.00); interaction: (F[3, 109]=0.21). When remission was
added as an independent variable to the total MAMA analysis,
the treatment effect disappeared (F[1, 98]=1.29; pN 0.05); the
effect of remission was highly significant (F[1, 98] =6.57;
p=0.01). There was an interaction trend of time×treatmen-
t × remission (F[3, 98] = 0.08). While all remitted groups
improved in MAMA total scores, among those still depressed,
the AD only group scores worsened and those who had switched
to LVs showed a larger degree of improvement.
3.2.3. 18–44 weeks: AD only, AD → LV, LV only, and LV → AD
Overall, no significant effects were found (treatment arm:
F[3, 74] = 0.85); time F[1, 74] = 0.72; time × treatment: F[3,
74] = 0.05) (Fig. 3c). A trend effect emerged of treatment on
parenting anxiety specifically, as the AD → LV group became
slightly more anxious (F[3, 80] = 2.38; p = 0.08). The addition
of remission as an independent variable in the total MAMA
analysis was also non-significant (F[1, 70] = 0.51), although
there was a time × remission interaction (F[1, 70] = 10.40;
p = 0.002). Those who remitted had increasing MAMA scores
whereas the scores of those who were still unwell tended to
remain stable or decrease slightly, suggesting a protective
effect of remission on maternal adjustment.
3.3. Comparing treatments across all time points
No significant effect of treatment arm (F[3, 66] =0.85) or
time×treatment arm interaction (F[9, 198] =0.95) was found;
all treatment groups improved with time (F[3, 198]= 14.09;
p b 0.001) (Fig. 4). The addition of remission (i.e. EPDS score b13
at week 44) as an independent variable was non-significant
(F[1, 70]= 0.51), but there was a time× remission interaction
(F[3, 186] = 4.98; p =0.002). Those who remitted by 44 weeks
showed steadily improving MAMA scores while those still
depressed at 44 weeks showed initial increases followed by
decreases at 18–44 weeks, irrespective of the treatment arm.
Treatment arm
AD only AD LV
LV only LV AD
38.00
MAMA score
36.00
34.00
32.00
0 4 18 44
Time point of study (weeks)
Fig. 4. Maternal attitude change between 0 and 44 weeks by treatment arm.
4. Discussion
Children of parents with mental illness may be at
particular risk of a poor early parental and social environment
(Wan and Green, 2009; Wan et al., 2008), which is
increasingly recognised by clinicians and policy makers
alike (DH, 2006; Parrott et al., 2008). Few treatment trials
have assessed aspects of early parenting in mothers with
PND. This represents the largest study to date following a
community sample of mothers with PND as they adjust to
parenthood. Our findings highlight the need for prompt and
effective treatment in order to minimise exposure of the
infant to maladaptive maternal cognitions, stress and anxiety.
Consistent with previous smaller intervention studies that
have evaluated parenting-related variables (Warner et al.,
1997; Murray et al., 2003; Misri et al., 2006), the effective
treatment of maternal depression led to a positive change in
parental adjustment and attitudes not seen in unremitted and
relapse cases. Mothers who were resistant to treatment
showed worsened attitudes after initial improvement, while
the attitudes of those who relapsed at any point also
worsened. These findings follow a similar pattern of effects
on child psychiatric symptoms reported in the STAR*D-Child
study in which child symptoms decreased as maternal
depression severity decreased with treatment, whereas
symptom count increased in children of non-remitting
mothers — who, as in our study, also tended to be on a low
income and not to be married or cohabiting (Pilowsky et al.,
2008).
We found that early remission of depressed mood did not
provide an added advantage to long-term maternal adjust-
ment, and – in line with previous similar studies (Misri et al.,
2006; Murray et al., 2003; Warner et al., 1997) – no benefit
associated with particular treatment modality or combination
of treatments. However, the earlier start of ADs provided a
short-term benefit (particularly for attitudes toward the
infant/parenthood and parent stress) at a potentially impor-
tant time in the early postnatal period. Moreover, the provision
of treatment choice may be important. Compared with other
treatment groups, the AD only group continued to make
attitudinal improvement (parenting stress specifically de-
creased) and showed higher MAMA scores at 18 weeks in
line with their remission. Among those who were still
depressed, continuing with AD only treatment seemed to
result in worsening attitudes (compared with other treatment
arms), while switching to, or adding in, LVs seemed to improve
attitudes in spite of continuing depressed mood. The impor-
tance of treatment choice is also notable when we consider
that the latter group, who was originally randomised to ADs,
was of a substantial number, which may reflect the negative
attitudes of PND women to taking ADs (Turner et al., 2008). No
such differential effects were found for those who received LVs
only or in combination with ADs. More LV sessions may be
needed to obtain measurable attitudinal change, as many
women reported that eight visits were insufficient to address
their symptoms (Turner et al., 2010), and many only received
four sessions. Moreover, the study design meant that LVs
started four weeks later than ADs and that those who took
combined treatment were self-selected (and so had possibly
been resistant to the original treatment to which they were
assigned), and there was no truly randomised LV only group
 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292
(since the later LV only group were those who did not wish to
move to or add in ADs).
The mothers in our study tended to adjust positively to
their infant with time and/or treatment for depression in the
first few months postnatally. These improvements, along
with the changes observed in maternal adjustment subscale
factor loadings across time, suggest a need in studies to
measure postnatal maternal adjustment at multiple time
points. However, the attitudes and adjustment of some
mothers are more likely to alter than others: women with a
partner are more likely to show improved attitudes, and less
educated and low income groups tended to worsen over
time/treatment. The increasing maladjustment or negativity
to parenting and/or their infant in the latter groups may
indeed contribute to or even exacerbate symptoms, as these
also constitute the risk factors for more chronic maternal
depression (Parrott et al., 2008; Pilowsky et al., 2008).
Our findings have a number of limitations. The pragmatic trial
design provided useful information on women's treatment
preferences and adherence but meant that there was no control
group against which to compare the four treatment arms.
Although treatment arms did not differ in severity of depressive
symptoms, the pragmatic design also led to a high rate of non-
compliance. The sample may also have excluded the more
depressed mothers since MAMA non-completers tended to be
more depressed at baseline and to be at higher social adversity at
44 weeks. This is likely to have diluted our findings of an
association between depression and maternal adjustment. As
may be expected in longitudinal research, attrition at 44 weeks
was high, but the MAMA subscale was sufficiently brief to enable
a large sample to complete it at four time points. Finally, it should
be noted that validation was based on the full MAMA, on a
version designed for prenatal use.
Although we found that earlier remission did not confer
added benefit to maternal adjustment in the longer-term,
more prolonged exposure to suboptimal attitudes and related
behaviour may adversely affect infants, perhaps particularly
in the more depressed mothers who may be underrepresent-
ed in our sample. It is well documented that infants of
mothers with chronic depression are less likely to have
secure, organised attachments (e.g. Campbell et al., 2004).
Longer follow-up is needed to study the stability of attitudinal
change and whether such changes are later reflected in
parenting behaviour and adverse infant outcomes.
Role of funding source
Funding for this study was provided by HTA, Department of Health
(grant number 02/7/04); the funder had no further role in study design; in
the collection, analysis and interpretation of data; in the writing of the
report; and in the decision to submit the paper for publication.
Conflict of interest
All authors confirm that we have no actual or potential conflict of
interest including any financial, personal or other relationships with other
people or organisations within three (3) years of beginning the work
submitted that could inappropriately influence, or be perceived to influence,
their work.
Acknowledgements
The study was funded by the HTA, Department of Health
(grant number 02/7/04). The views expressed in the paper
291
are not necessarily those of the funders. RESPOND trial:
Randomised Evaluation of antidepressants and Support for
women with POstNatal Depression. ISRCTN 16479414.
RESPOND team: Ian Anderson, Kathryn Abel, Carolyn
Chew-Graham, Elizabeth Chamberlain, Sandra Elliot, Liz
Folkes, Louise Howard, Glyn Lewis, Anne McCarthy, Anita
Mehay, Jean Mulligan, Tim Peters, Debbie Sharp, Morag
Turnbull, Katrina Turner, Andre Tylee, Alison Warburton.
We thank those primary care professionals and women
who participated in the trial.
Louise Howard is also affiliated with the NIHR Biomedical
Research Centre for Mental Health at the South London and
Maudsley NHS Foundation Trust and Institute of Psychiatry,
King's College London.
References
Anderson, I.M., Nutt, D.J., Deakin, J.F.W., 2000. Evidence-based guidelines for
treating depressive disorders with antidepressants: a revision of the
1993 British Association for Psychopharmacology guidelines. J. Psycho-
pharmacol. 14, 3–20.
Bernazzani, O., Marks, M.N., Bifulco, A., Siddle, K., Asten, P., Conroy, S., 2005.
Assessing psychosocial risk in pregnant/postpartum women using the
Contextual Assessment of Maternity Experience (CAME): recent life
adversity, social support and maternal feelings. Soc. Psychiatry Psychiatr.
Epidemiol. 40, 497–508.
Brockington, I.F., Oates, J., George, S., Turner, D., Vostanis, P., Sullivan, M., Loh,
C., Murdoch, C., 2001. A screening questionnaire for mother–infant
bonding disorders. Arch. Womens Ment. Health 3, 133–140.
Campbell, S.B., Brownell, C.A., Hungerford, A., Speiker, S.J., Mohan, R.,
Blessing, J.S., 2004. The course of maternal depressive symptoms and
maternal sensitivity as predictors of attachment security at 36 months.
Dev. Psychopathol. 16, 231–252.
Cornish, A.M., McMahon, C.A., Ungerer, J.A., Barnett, B., Kowalenko, N.,
Tennant, C., 2006. Maternal depression and the experience of parenting
in the second postnatal year. J. Reprod. Infant Psyc. 24, 121–132.
Cox, J.L., Holden, J.M., Sagovsky, R., 1987. Detection of postnatal depression:
development of the 10-item Edinburgh Postnatal Depression Scale. Br. J.
Psychiatry 150, 782–786.
Department of Health, 2006. Working together to safeguard children: a guide
to inter-agency working to safeguard and promote the welfare of
children. http://www.dcsf.gov.uk/everychildmatters/safeguardingand
socialcare/safeguardingchildren/workingtogether/workingtogetherto
safeguardchildren/2006.
Eccles, M., Freemantle, N., Mason, J., 1999. North of England evidence-based
guideline development project: summary version of guidelines for the
choice of antidepressants for depression in primary care. Fam. Pract. 16,
103–111.
Essex, M.J., Klein, M.H., Miech, R., Smider, N.A., 2001. Timing of initial
exposure to maternal major depression and children's mental health
symptoms in kindergarten. Br. J. Psychiatry 179, 151–156.
Feldman, R., Granat, A., Pariente, C., Kanety, H., Kuint, J., Gilboa-Schechtman,
E., 2009. Maternal depression and anxiety across the postpartum year
and infant social engagement, fear regulation, and stress reactivity. J. Am.
Acad. Child Adolesc. Psychiatry 48, 919–927.
Fihrer, I., McMahon, Catherine A., Taylor, Alan J., 2009. The impact of
postnatal and concurrent maternal depression on child behaviour during
the early school years. J. Affect. Disord. 119, 116–123.
Fleming, A.S., Klein, E., Corter, C., 1992. The effects of a social support group
on depression, maternal attitudes and behavior in new mothers. J. Child
Psychol. Psychiatry 33, 685–698.
Foreman, D.M., Henshaw, C., 2002. Objectivity and subjectivity in postnatally
depressed mothers' perceptions of their infants. Child Psychiatry Hum.
Dev. 32, 263–275.
Gaynes, B.N., Gavin, N., Meltzer-Brody, S., Lohr, K.N., Swinson, T., Gartlehner,
G., et al., 2005. Perinatal depression: prevalence, screening accuracy, and
screening outcomes. Evid. Rep. Technol. Assess. 119, 1–8.
Halligan, S.L., Murray, L., Martins, C., Cooper, P.J., 2007. Maternal depression
and psychiatric outcomes in adolescent offspring: a 13-year longitudinal
study. J. Affect. Disord. 97, 145–154.
Hay, D.F., Pawlby, S., Sharp, D., Asten, P., Mills, A., Kumar, R., 2001. Intellectual
problems shown by 11-year old children whose mothers had postnatal
depression. J. Child Psychol. Psychiatry 42, 871–889.
292 M.W. Wan et al. / Journal of Affective Disorders 131 (2011) 284–292
Hay, D.F., Pawlby, S., Waters, C., Sharp, D., 2008. Antepartum and postpartum
exposure to maternal depression: different effects on different adoles-
cent outcomes. J. Child Psychol. Psychiatry 49, 1079–1088.
Holden, J.M., Sagovsky, R., Cox, J.L., 1989. Counselling in a general practice
setting: controlled study of health visitor intervention in treatment of
postnatal depression. BMJ 298, 223–226.
Jenkins, R., Lewis, G., Bebbington, P., Brugha, T., Farrell, M., Gill, B., et al., 1997.
The National Psychiatric Morbidity Surveys of Great Britain — initial
findings from the household survey. Psychol. Med. 27, 775–789.
Kumar, R., Robson, K.M., Smith, A.M.R., 1984. Development of a self-
administered questionnaire to measure maternal adjustment and
maternal attitudes during pregnancy and after delivery. J. Psychosom.
Res. 28, 43–51.
Lewis, G., 1994. Assessing psychiatric disorder with a human interviewer or a
computer. J. Epidemiol. Community Health 48, 207–210.
Milgrom, J., McCloud, P., 1996. Parenting stress and postnatal depression.
Stress Med. 12, 177–186.
Misri, S., Reebye, P., Mills, L., Shah, S., 2006. The impact of treatment
intervention on parenting stress in postpartum depressed mothers: a
prospective study. Am. J. Orthpsychiatry 76, 115–119.
Murray, L., 1992. The impact of postnatal depression on infant development.
J. Child Psychol. Psychiatry 33, 543–561.
Murray, L., Sinclair, D., Cooper, P., Ducournau, P., Turner, P., 1999. The
socioemotional development of 5-year-old children of postnatally
depressed mothers. J. Child Psychol. Psychiatry 40, 1259–1271.
Murray, L., Cooper, P.J., Wilson, A., Romaniuk, H., 2003. Controlled trial of the
short and long-term effects of psychological treatment of post-partum
depression. 2. Impact on the mother–child relationship and child
outcome. Br. J. Psychiatry 182, 420–427.
Nagata, M., Nagai, Y., Sobajima, H., Ando, T., Honjo, S., 2003. Depression in the
mother and maternal attachment: results from a follow-up study at
1 year postpartum. Psychopathology 36, 142–151.
NICE Clinical Guideline, 2007. Antenatal and Postnatal Mental Health:
Clinical Management and Service Guidance. (Feb).
Parrott, L., Jacobs, G., Roberts, D., 2008. Research briefing: stress and
resilience factors in parents with mental health problems and their
children. Social Care Institute for Excellence. http://www.pmhcwn.org.
uk/files/scare23.pdf.
Pilowsky, D.J., Wickramaratne, P., Talati, A., Tang, M., Hughes, C.W., Garber, J.,
Malloy, E., King, C., Cerda, G., Sood, A.B., Alpert, J.E., Trivedi, M.H., Fava,
M., Rush, A.J., Wisniewski, S., Weissman, M.M., 2008. Children of
depressed mothers 1 year after the initiation of maternal treatment:
findings from the STAR*D-Child Study. Am. J. Psychiatry 165, 1136–1147.
Righetti-Veltema, M., Conne-Perréard, E., Bousquet, A., Manzano, J., 2002.
Postpartum depression and mother–infant relationship at 3 months old.
J. Affect. Disord. 70, 291–306.
Rogers, C.R., 1951. Client-Centered Therapy. Houghton Mifflin, Boston.
Rogosch, F.A., Cicchetti, D., Toth, S.L., 2004. Expressed emotion in multiple
subsystems of the families of toddlers with depressed mothers. Dev.
Psychopathol. 16, 689–709.
Sharp, D.J., Chew-Graham, C., Tylee, A., Lewis, G., Howard, L., Anderson, I., et
al., 2010. A pragmatic randomised controlled trial to compare anti-
depressants with a community based psychosocial intervention for the
treatment of women with postnatal depression: the RESPOND trial.
Health Technol. Assess. 14.
Teti, D.M., Gelfand, D.M., 1991. Behavioral competence among mothers of
infants in the first year: the mediational role of maternal self-efficacy.
Child Dev. 62, 918–929.
Turner, K., Sharp, D., Folkes, L., Chew-Graham, C.A., 2008. Women's views and
experiences of antidepressants as a treatment for postnatal depression: a
qualitative study. Fam. Pract. 25, 450–455.
Turner, K., Chew-Graham, C.A., Folkes, L., Sharp, D., 2010. Women's
experiences of health visitor delivered listening visits as a treatment
for postnatal depression: a qualitative study. Patient Educ. Couns. 78,
234–239.
Wan, M.W., Green, J., 2009. The impact of maternal psychopathology on
child–mother attachment. Arch. Womens Ment. Health 12, 123–134.
Wan, M.W., Abel, K.M., Green, J., 2008. The transmission of risk to children
from mothers with schizophrenia: a developmental psychopathology
model. Clin. Psychol. Rev. 28, 613–637.
Warner, R., Appleby, L., Whitton, A., Faragher, B., 1997. Attitudes toward
motherhood in postnatal depression: development of the maternal
attitudes questionnaire. J. Psychosom. Res. 43, 351–358.
Webster, M.L., Thompson, J.M.D., Mitchell, E.A., Werry, J.S., 1994. Postnatal
depression in a community cohort. Aust. NZ J. Psychiatry 28, 42–49.
Windridge, K.C., Berryman, J.C., 1996. Maternal adjustment and maternal
attitudes during pregnancy and early motherhood in women of 35 and
over. J. Reprod. Infant Psychol. 1414, 45–55.