Vol.5, No.2, 237-244 (2013) Health
http://dx.doi.org/10.4236/health.2013.52032
Impact of postnatal maternal depressive symptoms
and infant’s sex on mother-infant interaction among
Bangladeshi women
Maigun Edhborg1*, Beatrice Hogg1#, Hashima-E-Nasreen2, Zarina Nahar Kabir1
1Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden;
*Corresponding Author:
2Research and Evaluation Division, BRAC, Dhaka, Bangladesh
Received 14 September 2012; revised 15 October 2012; accepted 22 October 2012
z
ABSTRACT
Aim: To investigate the impact of postnatal de-
pressive symptoms and infant sex on perceived
and observed mother-infant interaction among
rural Bangladeshi women. Metho ds: Fifty women
with depressive symptoms and their infants at 2
- 3 months were compared with 50 women
without depressed symptoms and their infants,
matched on geographic areas, parity and infant
sex. The Edinburgh Postnatal Depression Scale
assessed depressive symptoms, the Postpartum
Bonding Questionnaire assessed the mother’s
perception of bonding with the infant and
mother-infant interactions were videotaped and
analyzed with the Global Rating Scale. Results:
Mothers with depressive symptoms were poorer,
were less educated and rated lower infant bond-
ing than mothers without depressive symptoms
(p = 0.03), yet objective observation revealed no
difference between the two groups regarding
maternal interactive behavior (p = 0.57). How-
ever, infants, particularly boys (p = 0.002), of
mothers with depressive symptoms fretted more
in mother-infant interaction than infants of moth-
ers without depressive symptoms (p = 0.009).
Conclusion: Although mothers with depressive
symptoms did not show less sensitivity in in-
teractive behavior at 2 - 3 months than those
without depressive symptoms, our results indi-
cate that infants, particularly boys, of mothers
with depressive symptoms may be negatively
influenced by depressiv e symptoms.
Keyw ords: Postpartum Depressive Symptoms;
Mother-Infant Interaction; Bonding; Bangladesh
1. INTRODUCTION
Maternal depressive symptoms related to childbirth
are common in both high- and low income countries [1,
2]. However, socially disadvantaged populations show
higher depression rates than wealthy industrial nations
[3]. A systematic review from low- and lower-middle
income countries showed a prevalence of 19.8% of com-
mon mental disorder postpartum [4]. In South Asia, the
prevalence of depression has been reported to be 12% -
28% postpartum, including the low prevalence of 12%
found in Nepal [5-8].
Depression is a disabling disorder with symptoms such
as low mood, lack of energy and interest, poor concen-
tration and sometimes thoughts of suicide or death, all of
which are likely to interfere with a mother’s parenting
capacity. Indeed, there is convincing evidence that ma-
ternal depression during the postnatal period is associ-
ated with negative effects on the infant’s mental, cogni-
tive and socio-emotional development [9-11]. An expla-
nation for these developmental adversities could be dis-
turbances in the early mother-infant relationship [12].
The relational behavior of depressed mothers in high-
income countries has been characterized by low sensitiv-
ity, a restricted range of affective expression and incon-
sistent support of the infant’s growing social engagement
[13]. A mother’s emotional communicative skill is very
important during the first postpartum period to regulate
the infant’s states of arousal and emotion. Ineffective
regulation generates extreme states of arousal and affect
that could disrupt the infant’s engagement with people
and the environment [3]. Although there is a mismatch of
emotions and intentions in all mother-infant interaction,
the mismatch is quickly repaired in normal interactions.
Mismatches generate negative affect in the infant, whereas
restoration to a matching state is associated with positive
affect, leading to a sense of mastery and control [3]. Stein
#Formerly Department of Neurobiology, Care Sciences and Society,
Karolinska Institutet, Stockholm, Sweden.

M. Edhborg et al. / Health 5 (2013) 237-244
238
et al. [13] suggest that maternal preoccupation is a factor
that contributes to the adverse effect on development
among infants of mothers with depressive symptoms.
Preoccupation refers to the process of recurrent negative
thinking that is characteristic of depression and interferes
with the mother’s attention and responsiveness towards
the infant [13]. A study in Bangladesh exploring mater-
nal depressive symptoms and infant development re-
ported that maternal depressive symptoms interfered with
the mothers’ sensitivity and ability to provide stimulating
play materials and thus, give the infant few opportunities
for exploration and social engagement at home [14].
Only a few studies have investigated mother-infant
interactions related to maternal depressive symptoms in
low-income countries. In South Africa, Cooper et al. [15]
found that depressed mothers in Khayelitsha, a periurban
setting in the outskirts of Cape Town with high levels of
poverty, were less sensitive to their infants in early face-
to-face interactions, and their infants were less positively
engaged with their mothers, than infants of non-de-
pressed mothers.
Although Field [16] claims that interaction distur-
bances among depressed mothers and their infants appear
to be universal across cultures and socioeconomic groups,
the pressure on a mother in conditions of extreme po-
verty could differ from those facing a mother in high-
income countries. Since several studies have reported on
gender preference in favor of boys in some Asian cul-
tures [17,18] and how giving birth to a female baby is a
risk factor for depressive symptoms [19], it is important
to explore if the sex of the infant influences the mother-
infant interaction. Therefore, this study aims to investi-
gate the relationship between maternal depressive symp-
toms and the infant’s sex 2 - 3 months postpartum on
perceived and observed mother-infant interaction in rural
Bangladesh.
2. METHOD
2.1. Design and Setting
This is a nested-case control study of a prospective-
cohort study of postnatal depressive symptoms and infant
development carried out in two rural subdistricts of the
Mymensingh district, 120 km north of the capital city of
Dhaka in Bangladesh. This district is typical of deltaic
Bangladesh and is a predominantly agricultural commu-
nity characterized by homogeneity in terms of ethnicity,
culture and language. For the prospective-cohort study,
720 women were recruited during their third trimester of
pregnancy, and the women and their infants were fol-
lowed up at birth, 2 - 3 months and 6 - 8 months post-
partum. At 2 - 3 months postpartum, 673 mothers with
infants remained in the study. The reasons for attrition
(6.5%) were outmigration (n = 9), maternal death at
childbirth (n = 2), stillbirth (n = 25) and neonatal death
(n = 11). As in other rural areas of Bangladesh, the ma-
jority of the women was poor and had domestic work
and child care as their primary occupations. Approxi-
mately 40 per cent of the population lived below the
poverty line [20].
2.2. Participants
Of the 673 mothers remaining in the cohort study at 2
- 3 months postpartum, 95mothers (14%) showed de-
pressive symptoms (i.e., scored 10 or more on the Edin-
burgh Postnatal Depression Scale, or EPDS). As soon as
a mother was identified as having depressive symptoms
at 2 - 3 months, she was approached and asked to par-
ticipate together with her infant in a video observation of
a mother-infant interaction. In case of consent, a mother
without depressive symptoms (control, i.e., scored less
than 10 on the EPDS) was identified and requested to
participate with her infant. Thus, 50 mothers and their
infants were consecutively selected as cases, and 50 as
controls. The control mothers were matched with the
cases by geographic area, parity and the infant’s sex. All
100 mothers were informed of the aim of the study, pro-
cedures of the video recording and that participation was
voluntary. None of the mothers declined participation.
The sample size was chosen at a significance level of 5
per cent, a power of 80 per cent and an effect size of 60
per cent to detect differences in the mother-infant inter-
action between cases and controls.
2.3. Data Collection and Procedures
The background data were taken from the prospective
cohort study, which was collected by trained, female
interviewers through structured interviews in the respon-
dents’ homes. These female interviewers were trained
over two weeks; all questionnaires were discussed and
pretested outside the study site. Information was col-
lected during pregnancy on demographics, including
maternal age and socioeconomic status (SES) indicated
by educational level, primary occupation, involvement in
income-generating activities, land ownership by house-
hold and daily expenditure on food in the respondent’s
household.
Obstetric and child information were measured by
parity, mean number of living children and having more
than four children. Family characteristics and support in-
cluded the women’s perceived relationship with husband
and mother-in-law and practical help received during
pregnancy. Intimate partner violence encompassed phy-
sical violence ever and/or during pregnancy and forced
sex. At 2 - 3 months postpartum, information was ob-
tained on low birth weight of the infant (≤2.5 kg), pre-
term delivery (≤37 pregnancy weeks), exclusive breast-

M. Edhborg et al. / Health 5 (2013) 237-244 239
feeding, the mother’s perception of the infant’s care
(easy, in-between, difficult), the mother’s bonding with
her infant and maternal depressive symptoms.
Mother-infant interactions were video recorded. The
recording took place at an office in the local setting by
two specially trained Bangladeshi social science re-
searchers. The mothers were offered transportation and
food for their participation. The video sessions lasted for
five minutes when the infants were alert and not hungry.
The mothers were asked to interact or play with the in-
fant as they would do at home. The infants were between
3 - 9 months old when the video recording took place. A
camera on a tripod was placed in the room to the right
side of a big mirror. The mothers were seated in front of
the mirror and had a large pillow on their outstretched
legs on which to put the baby, so that the infant’s face
and body would be seen in the camera. The camera was
to the right of the mother, recording the mother’s face
and body to the hips and the infant’s face and body. The
mothers were left alone with their infants during the
video recording. The Bangladesh Medical Research
Council, Bangladesh, and the Regional Ethical Review
Board at Karolinska Institute, Sweden, approved the
study.
2.4. Instruments
The Edinburgh Postnatal Depressive Scale (EPDS)
[21] was used to detect depressive symptoms. The EPDS
is a 10-item questionnaire that has recently been vali-
dated in Bangladesh (EPDS-B), showing a sensitivity of
89 per cent, a specificity of 87 percent and a cutoff score
of 9/10 [18]. For this study we used the same cutoff as
Gausia et al. [22]. The items of the EPDS were rated on
a scale of 0 - 3, and a high score indicated more depres-
sive symptoms. Cronbach’s alpha was 0.80 at 2 - 3
months postpartum.
The Postpartum Bonding Questionnaire (PBQ) [23,24]
was used to assess mothers’ perceptions of their emo-
tional bonding with their infants. The PBQ consists of 25
items, rated on a scale of 0 - 5, high scores indicating
greater bonding disturbances. Four subscales are in-
cluded. Subscale 1 indicates impaired bonding between
mother and infant (12 items) with a cutoff score of 11/12
identifying mild bonding problems. Subscale 2 indicates
rejection and anger towards the infant (7 items) with a
cutoff score of 12/13 identifying mothers with threatened
rejection and 16/17 identifying those with established
rejection, or more severe bonding problems. Subscale 3
indicates anxiety about the care of the infant (4 items),
with a cutoff of 10/11. At subscale 4, there is a risk of
abuse of the infant by the mother (2 items). In this study,
the Cronbach’s alpha on the total PBQ was 0.82, and on
the subscales of impaired bonding it was 0.65, rejection
and anger 0.60, and anxiety about care 0.44—results that
are similar to those reported by Brockington et al. [23].
Risk of abuse showed an excessively low Cronbach’s
alpha and was excluded from the analyses. The instru-
ment was translated from English into Bangla and back
from Bangla into English by two bilingual persons [25].
PBQ has been validated in China and was found to have
satisfactory face and content validity and to be sensitive
in detecting impairments of the mother-infant relation-
ship as perceived by the mother with the cut-off scores
suggested by Brockington et al. [26].
Analysis of mother-infant interaction
The Global Rating Scale (GRS) of infant-mother in-
teraction was used to assess infant-mother interaction
[27]. This method has been found to be valid cross-cul-
turally in four European countries [28] and in South Af-
rica [15]. Two of the authors (ME and BH) were trained
in the method at a research centre in Reading, England.
The analyses were done blind as to the depressive status
of the mothers, based on videos and English verbatim
transcripts, which were translated from Bangla to Eng-
lish, English to Bangla and back again to English by two
bilingual persons to ensure the reliability of the transla-
tion.
The mothers’ behavior was assessed on 13 items, the
infants’ on 7 and dyadic behavior on 5. The items are
scored on a 5-point scale from 1 (poor) to 5 (good) and
clustered by summing up and form dimensions.
1) The mother’s good-poor dimension (sensitivity) was
measured by five items: warmth, acceptance, respon-
siveness, non-demanding behavior and sensitivity.
2) The mother’s intrusive dimension was measured by
her intrusive or nonintrusive behavior and speech.
3) The mother’s remote dimension was assessed by her
remote or non-remote and silent or non-silent behavior.
4) The mother’s affective behavior (depression) was
assessed by happiness, energy to engage with the infant,
absorption in the infant and whether she was relaxed or
tense during the interaction.
The infant’s interactive behavior was also measured.
1) The infant’s good-poor dimension (engagement) was
measured by attentiveness towards the mother, active
communication with body and voice towards the mother
and positive vocalization.
2) The infant inert-lively dimension (liveliness) meas-
ured the infant’s engagement with the environment,
whether he/she was self-absorbed or not and whether or
not he/she was lively or inert.
3) The infant distress dimension (fretfulness) was in-
dicated by happy or distressed and fretful or non-fretful.
Dyadic behavior:
4) The mother-infant interaction dimension was as-
sessed by smooth and easy or difficult, fun or serious,
mutually satisfying or unsatisfying, involving much en-

M. Edhborg et al. / Health 5 (2013) 237-244
240
gagement or no engagement and excited engagement or
quiet engagement [24].
Two of the authors (BH, ME) rated the videos. Forty-
eight videos were rated by consensus and fifty-two indi-
vidually. Of those individually rated, nine were coded for
internal reliability, with 92 percent of the ratings being
found to be in agreement.
2.5. Statistical Analyses
The statistical analyses were done using SPSS for
Windows, version 18. Differences between the mothers
with and without depressive symptoms, infant sex and
the PBQ were calculated by two-way multivariate analy-
sis of variance (MANOVA). The GRS (assessing mother
and infant behavior) was similarly analyzed by two-way
MANOVAs, and dyadic behavior was analyzed with a
two-way analysis of variance (ANOVA). Cases and con-
trols were compared according to the background vari-
ables by independent t-test and chi-square.
3. RESULTS
3.1. Description of the Sample
As seen in Table 1, compared to mothers without,
mothers with depressive symptoms were less educated
and poorer in terms of land owned by the household and
daily expenditure on food per household member. More
mothers with depressive symptoms reported experienc-
ing physical violence ever in life and during current
pregnancy and had infants with low birth weight com-
pared to mothers without depressive symptoms. Mothers
Table 1. Information on demographic, socio-economic, obstetric, child, family support and intimate partner violence of the sample
(N = 100).
Mothers with depressive
symptoms N = 50
Mothers without depressive
symptoms N = 50 p
Age (M ± SD) 26.6 (6.32) 25.6 (6.09) 0.432
Educational level
Years of schooling (M ± SD) 1.86 (2.86) 3.22 (3.07) 0.024
Illiterate [N (%)] 38 (78) 10 (28) 0.001
Primary occupation [N (%)]
Domestic work 45 (90) 49 (98) 0.092
Land owned by HH [N (%)]
1 - 49 decimal 46 (92) 34 (68)
≥50 decimal 4 (8) 16 (32) 0.003
Per daily household expenditure
(in Taka)* on food (M ± SD) 27.4 (11.1) 36.0 (17.3) 0.004
Obstetric and infant data
First time mothers [N (%)] 6 (12) 7 (14) 0.766
Number of children (M ± SD) 3.0 (2.3) 2.4 (1.8) 0.130
More than 4 children [N (%)] 19 (38) 12 (24) 0.130
Low birth weight (<2.5 kg) [N (%)] 15 (30) 5 (10) 0.012
Preterm (<37 weeks) [N (%)] 8 (16) 8 (16) 1.000
Still breastfeeding exclusively [N (%)] 16 (32%) 25 (50%) 0.067
Difficult to care for the infant [N (%)] 34 (68%) 20 (40%) 0.019
Family characteristic and support
Poor relationship—husband [N (%)] 7 (14) 0 0.012
Poor relationship—mother-in-law [N (%)]** 5 (12.5) 6 (14.3) 0.813
Did not have practical support [N (%)] 11 (22) 16 (32) 0.260
Intimate partner violence
Forced sex [N (%)] 44 (88) 37 (74) 0.074
Physical violence ever [N (%)] 46 (92) 34 (68) 0.003
Physical violence during current pregnancy [N (%)] 18 (36) 7 (14) 0.011
*1 US-dollar = approximately 71 Bangladeshi Taka (BDT); **Calculation computed on 82 mothers-in-law.

M. Edhborg et al. / Health 5 (2013) 237-244 241
with depressive symptoms reported their infants at 2 - 3
months as being significantly more difficult to care for
compared to the mothers in the control group (Table 1).
The mean EPDS at 2 - 3 months was 11.5 (SD 1.4)
among the cases and 4.3 (SD 1.4) among the controls.
3.2. Maternal Perception of Bonding with
Infants According to the PBQ
A two-way MANOVA showed a main effect of ma-
ternal depressive symptoms (F(3,94) = 3.06; p = 0.03) but
no effect of infant sex (F(3,94) = 1.96; p = 0.13) on the
PBQ, which measures mothers’ perceptions of bonding
with their infants. Follow-up ANOVAs indicated that
mothers with depressive symptoms reported more im-
paired bonding and anxiety about infant care than mo-
ther’s without depressive symptoms, but they did not
show more rejection and anger towards the infant indi-
vidually. Of those individually rated, nine were coded for
internal reliability, with 92 percent of the ratings being
found to be in agreement.
A tendency of interaction effect was observed between
depressive symptoms and infant sex (F(3,94) = 2.71; p =
0.051) indicating that mothers without depressive symp-
toms expressed more rejection and anger to their daugh-
ters than to their sons on the PBQ. However, mothers
with depressive symptoms rated rejection and anger
equally, regardless of sex of the infant (Table 2). 12% of
the mother reported mild impaired bonding to their in-
fants. Four mothers reported threatened rejection and one
establish rejection i.e. more severe bonding disturbances.
The last sentence has disappeared in the text. It is only a
descriptive clarification of how many of the mothers who
showed bonding disturbances. The whole sentence could
be taken away if you want.
3.3. The Mother-Infant Interaction According
to the GRS
A two-way MANOVA showed that neither maternal
depressive symptoms (F(4,93) = 0.74; p = 0.57) nor infant
sex (F(4,93) = 0.37; p = 0.83) had any significant main
effect on maternal interacting behavior on the GRS. On
the other hand, on the GRS, maternal depressive symp-
toms (F(3,94) = 3.52; p = 0.02) as well as infant sex (F(3,94)
= 3.91; p = 0.01) showed main effects on infant interac-
tive behavior.
Follow-up ANOVAs indicated that infants, particu-
larly boys of mothers with depressive symptoms, were
more distressed and fretful, but not less engaged or lively,
during the interaction than infants of mothers without
depressive symptoms (Table 3).
No significant interaction between maternal depressive
symptoms and infant sex was found (F(3,94) = 0.69; p =
0.56). Finally, on the GRS, a two-way ANOVA showed
no effects of maternal depressive symptoms (F(1,96) =
0.45; p = 0.51) or infant sex (F(1,96) = 0.88; p = 0.35) on
dyadic behavior.
4. DISCUSSION
The main finding of this study is that mothers with
depressive symptoms rated their bonding with their in-
fants significantly lower on a self-reported scale than did
mothers without depressive symptoms in rural Bangla-
desh. However, objective observation of mother-infant
interaction did not indicate any significant differences in
the interactive behavior of mothers with and without de-
pressive symptoms. On the other hand, observation of
infant behavior indicated that infants of mothers with
depressive symptoms were more distressed and fretful
during the interaction than infants of mothers without
depressive symptoms. This behavior was more pro-
nounced in boys than in girls.
The finding that mothers with depressive symptoms
rated their bonding lower than mothers without depres-
sive symptoms, while their observed interactive behavior
did not indicate any difference, was consistent with re-
search by Frankel and Harmon [29]. They also found that
depressed women reported more negative evaluations of
themselves as parents than did non-depressed women,
even if no difference was found in observed maternal in-
teractive behavior. Similarly, Hornstein et al. [30] found
Table 2. Comparisons between mothers with and without depressive symptoms and sex of infant according to parental bonding (N =
100) (Mean, SD).
Mothers with
depressive symptoms
N = 50
Mothers without
depressive symptoms
N = 50 p-value p-value p-value
Boys
N = 27
Girls
N = 23
Boys
N = 28
Girls
N = 22
Difference depressive/non
depressive symptoms
Difference
infant’s sex
Interaction maternal
depr/infant’s sex
PBQ-subscale 1
Impaired bonding 8.2 (4.3) 8.5 (4.0) 5.6 (2.8) 6.8 (3.6) 0.005 0.301 0.594
PBQ-subscale 2
Rejection and anger 5.3 (3.1) 5.1 (3.4) 3.1 (2.4) 5.5 (4.1) 0.143 0.088 0.051
PBQ-subscale 3
Anxiety about care 3.9 (2.2) 3.9 (2.4) 3.0 (2.2) 2.8 (2.2) 0.032 0.823 0.719

M. Edhborg et al. / Health 5 (2013) 237-244
242
Ta bl e 3 . Comparisons between mothers with and without depressive symptoms and sex of the infant according to the dimensions of
Global Ratings Scale (N = 100) (Mean, SD).
Mothers with depressive
symptoms (N = 50) Mothers without depressive symptoms (N = 50) p-value p-value
Boy (N = 27) Girl (N = 23)Boy (N = 28)Girl (N = 22)Difference depressive/non
depr symptoms
Difference infant’s
sex
Mother’s interactive behaviour
Mother’s sensitivity 3.0 (0.8) 3.4 (0.9) 3.4 (0.9) 3.3 (0.9) 0.338 0.372
Maternal intrusiveness 3.4 (1.1) 3.6 (1.3) 3.2 (1.3) 3.6 (1.0) 0.815 0.224
Maternal remoteness 3.3 (1.0) 3.7 (1.0) 3.8 (1.0) 3.5 (1.3) 0.367 0.885
Maternal affective behaviour
(depression) 3.0 (1.0) 3.3 (1.0) 3.6 (0.9) 3.3 (1.2) 0.145 0.778
Infant’s interactive behaviour
Infant engagement 2.9 (1.0) 3.0 (0.9) 2.9 (0.9) 3.1 (1.1) 0.818 0.355
Infant liveliness 3.0 (0.8) 3.3 (0.9) 3.2 (0.8) 3.4 (1.0) 0.371 0.182
Infant fretfulness 3.4 (0.8) 3.8 (0.8) 3.7 (0.8) 4.3 (0.6) 0.009 0.002
Dyadic behaviour
Mother—infant interaction 3.0 (1.0) 3.3 (0.9) 3.3 (0.8) 3.3 (1.2) 0.505 0.350
that depressed mothers rated their emotional bonding
lower than psychotic mothers did, but no differences
were found in observed mother-infant interaction. This
finding was explained by the fact that depressed mothers,
contrary to psychotic mothers, may have negative thoughts
and low self-esteem due to their depressive symptoms,
which might negatively influence their ratings. It is
known that depressed mothers appear to focus mostly on
the negative aspects of their infants [31], which might
influence the mothers’ ratings of bonding. In our Bang-
ladeshi sample, mothers with depressive symptoms were
less educated and poorer than mothers without depres-
sive symptoms, which might have contributed to low
self-confidence in mothers with depressive symptoms
and the ratings of the PBQ. Low self-esteem and feelings
of worthlessness have been associated with both depres-
sion and limited literacy [32].
Mothers with depressive symptoms often know that
they have not been available enough for their infants, and
therefore could try hard to do everything right during the
observations. Thus, it may be questioned if a short video-
observation is a valid representation of mother-infant
interaction. A Finnish study demonstrated that a video-
recording of only 5 minutes provided similar information
about mother-infant interaction observed, once a week,
over a year in the families’ homes during the infant’s
first year [33].
Even if a mother with depressive symptoms did eve-
rything right during the observation, she might not al-
ways have been sensitive to her infant’s cues and as a
result the infant might have accumulated negative affect
and stress that was later demonstrated by the infant’s
distress and fretfulness during the mother-infant interac-
tion observation. Another explanation to the distress and
fretfulness in the infants of mothers with depressive
symptoms could be that the infants of these mothers have
a more difficult temperament than infants of mothers
without depressive symptoms as reported in several
studies [14,34]. In this study, this is indicated by the fact
that mothers with depressive symptoms rated their in-
fants as more difficult to care for than the mothers in the
control group. Although mothers with and without de-
pressive symptoms in this sample were rated as equally
sensitive, a temperamentally difficult and fussy infant
may have been too challenging for the mothers with de-
pressive symptoms in everyday life, particularly if the
infant was a boy. In a study from Bangladesh, it was
found that mothers with depressive symptoms who per
ceived their infants to be irritable were less sensitive and
provided their infants a less stimulating environment at
home than mothers without depressive symptoms who
did not perceive their infants as irritable [14].
The finding of no observed difference in interactive
behavior with their infants between mothers with and
without depressive symptoms is inconsistent with a study
by Cooper et al. [15] from a disadvantaged suburban
area in South Africa. An explanation of the difference in
findings may be that mothers in rural Bangladesh, unlike
those in South Africa, had relatively stable situations in
terms of their living arrangements in joint families and in
a society with low migration. This indicates a social
network that is relatively stable and safe for the infants,
who can be looked after by relatives and neighbours.
Another explanation might be that the depressive symp-
toms in this study were self-reported and the cutoff score
(9/10) of EPDS rather low in contrast to the clinical di-
agnoses according to DSM-IV conducted by Cooper et al.
[15], indicating that the Bangladeshi mothers may have

M. Edhborg et al. / Health 5 (2013) 237-244 243
had fewer depressive symptoms than the South African
mothers. Although the families in our sample live in
poverty, the relatively stable situation in the family and
the transient maternal depressive symptoms might ex-
plain the positive behaviour through the mother-infant
interaction.
5. STRENGTHS AND LIMITATIONS
As this case-control study of mother-infant interaction
was nested in a cohort, the risk of bias imposed by retro-
spective recall was reduced. The instrument, EPDS, used
to assess depressive symptoms is validated in Bangla-
desh [22]. Another assessment instrument, PBQ, used in
this study is validated in several countries including in
Asia, e.g. in China [26] and has shown good ability to
discriminate between good and poor bonding, even in
Bangladesh [25]. The Global Rating Scale (GRS) was
chosen as it was developed to assess differences in
mother-infant interaction between groups of women with
and without postnatal depressive symptoms and has been
successfully been used in cross-cultural settings [28].
The coders in the study did not know Bangla, which
meant that verbal communications in the video record-
ings had to be translated from Bangla into English. This
procedure could have led to misunderstandings and in-
terpretation problems. However, all translations have
been checked against the videos by two bilingual (Bangla
and English speaking) researchers (HN and ZNK). Half
of the videos were coded in consensus, and differences in
the ratings were discussed in detail. Trained Bangladeshi
interviewers collected the background data and two spe-
cially trained Bangladeshi social science researchers
videotaped the mother-infant interactions.
6. CONCLUSION
Although the mothers with depressive symptoms were
poorer, were less educated and rated the emotional
bonding with their infants lower than mothers without
depressive symptoms, they were not objectively rated as
less sensitive and responsive towards their infants than
the control group. However, it should be noted that in-
fants in the case group, particularly boys, were more dis-
tressed and fretful in the observed interaction than in-
fants in the control group. Mothers who were cases also
rated themselves less confident as a mother compared to
the mothers who were controls. Thus, the results indicate
that it is important to identify mothers with depressive
symptoms early postpartum and give them support and
confirmation to increase their self-confidence as a mother,
since depressive symptoms in mothers may have a nega-
tive effect on the affective states of their infants. Particu-
larly boys seem to be less able to self-regulate their af-
fective states during the early interactions.
7. ACKNOWLEDGEMENTS
The study was supported by grants from the Swedish Research Link
(2007-25292-51983-33) to Karolinska Institute and the School of Pub-
lic Health at BRAC University. We appreciate the help of BRAC in
Bangladesh in carrying out the study. We would also like to thank all
the women who participated in the study with their infants for gener-
ously giving their time and energy.
REFERENCES
[1] Evans, J., Heron, J., Francomb, H., Oke, S. and Golding,
J. (2001) Cohort study of depressed mood during preg-
nancy and after childbirth. British Medical Journal, 32,
257-260. doi:10.1136/bmj.323.7307.257
[2] Rahman, A. and Creed, F. (2007) Outcome of prenatal
depression and risk factors associated with persistence in
the first postnatal year: Prospective study from Rawal-
pindi, Pakistan. Journal of Affective Disorders, 100, 115-
121. doi:10.1016/j.jad.2006.10.004
[3] Tronick, E. and Reck, C. (2009) Infants of depressed
mothers. Harvard Review of Psychiatry, 17, 14-156.
doi:10.1080/10673220902899714
[4] Fisher, J., de Mello, C., Patel, V., Raham, A., Tran, T.,
Holton, S. and Holmes W. (2012) Prevalence and deter-
minants of common perinatal mental disorders in women
in low- and lower-middle income countries: A systematic
review. Bulletin of the World Health Organization, 90,
139-149. doi:10.2471/BLT.11.091850
[5] Patel, V., Rodrigues, M. and De Souza, N. (2002) Gender,
poverty, and postnatal depression: A study of mothers in
Goa, India. American Journal of Psychiatry, 159, 43-47.
doi:10.1176/appi.ajp.159.1.43
[6] Rahman, A., Iqbal, Z. and Harrington, R. (2003) Life
events, social support and depression in childbirth: Per-
spectives from a rural community in the developing
world. Psychological Medicine, 33, 1161-1167.
doi:10.1017/S0033291703008286
[7] Gausia, K., Fisher, C., Ali, M. and Oosthuizen, J. (2009)
Magnitude and contributory factors of postnatal depres-
sion: A community-based cohort study from a rural sub
district of Bangladesh. Psychological Medicine, 39, 999-
1007. doi:10.1017/S0033291708004455
[8] Regmi, S., Sligl, W., Carter, D., Grut, W. and Seear, M.
(2002) A controlled study of postpartum depression among
Nepalese women: Validation of the Edinburgh Postpar-
tum Depression Scale in Kathmandu. Tropical Medicine
and International Health, 7, 378-382.
doi:10.1046/j.1365-3156.2002.00866.x
[9] Lyons-Ruth, K., Zoll, D., Connell, D. and Grunebaum,
H.U. (1986) The depressed mother and her one-year old
infant: Environment, interaction, attachment, and infant
development. New Directions for Child and Adolescent
Development, 34, 6-82.
[10] Murray, L. and Cooper, P. (1997) Effects of postnatal
depression on infant development. Archives of Disease
Childhood, 77, 99-101. doi:10.1136/adc.77.2.99
[11] Murray, L. (1992) The impact of postnatal depression on

M. Edhborg et al. / Health 5 (2013) 237-244
244
infant development. Journal of Child Psychology and Psy-
chiatry, 33, 543-561.
doi:10.1111/j.1469-7610.1992.tb00890.x
[12] Feldman, R., Granat, A., Pariente, C., Kanety, H., Kuint, J.
and Gilboa-Schechtman, E. (2009) Maternal depression
and anxiety across the postpartum year and infant social
engagement, fear regulation, and stress reactivity. Journal
of the American Academy of Child and Adolescent Psy-
chiatry, 48, 919-927.
doi:10.1097/CHI.0b013e3181b21651
[13] Stein, A., Lehtonen, A., Harvey, A.G., Nicol-Harper, R.
and Craske, M. (2009) The influence of postnatal psychi-
atric disorder on child development. Psychopathology, 42,
11-21. doi:10.1159/000173699
[14] Black, M., Bagui, A.H., Zaman, K., McNary, S.W., Le, K.,
El Arifeen, S., Hamadani, J.D., Parveen, M., Yunus, M.
and Black, R.E. (2007) Depressive symptoms among ru-
ral Bangladeshi mothers: Implications for infant devel-
opment. Journal of Child Psychology and Psychiatry, 48,
764-772. doi:10.1111/j.1469-7610.2007.01752.x
[15] Cooper, P.J., Tomlinson, M., Swartz, L., Woolgar, M.,
Murray, L. and Molteno, C. (1999) Post-partum depres-
sion and the mother-infant relationship in a South African
peri-urban settlement. British Journal of Psychiatry, 175,
554-558. doi:10.1192/bjp.175.6.554
[16] Field, T. (2010) Postpartum depression effects on early
interactions, parenting and safety practices: A review. In-
fant Behavior and Development, 33, 1-6.
[17] Patel, V., DeSouza, N. and Rodrigues, M. (2003) Postna-
tal depression and infant growth and development in low
income countries: A cohort study from Goa, India. Ar-
chives of Disease in Childhood, 88, 34-37.
doi:10.1136/adc.88.1.34
[18] Klainin, P. and Arthur, D.G. (2009) Postpartum depres-
sion in Asian culture: A literature review. International
Journal of Nursing Studies, 46, 1355-1373.
doi:10.1016/j.ijnurstu.2009.02.012
[19] Gausia, K., Fisher, C., Ali, M. and Oosthuizen, J. (2009)
Antenatal depression and suicidal ideation among rural
Bangladeshi women: A community-based study. Archives
of Women’s Mental Health, 12, 351-358.
doi:10.1007/s00737-009-0080-7
[20] World Bank (2011) World Development Report, 2011.
http://wdr2011.worldbank.org/sites/default/files/WDR201
1_Indicators.pdf
[21] Cox, J.L., Holden, J.M. and Sagovsky, R. (1987) Detec-
tion of postnatal depression. Development of the 10-item
Edinburgh postnatal depression scale. British Journal of
Psychiatry, 50, 782-786. doi:10.1192/bjp.150.6.782
[22] Gausia, K., Hamadani, J.D., Islam, M.M., Ali, M., Algin,
S., Yunus, M., Fisher, C. and Oosthuizen, J. (2007) Bangla
translation, adaptation and piloting of Edinburgh postna-
tal depression scale. Bangladesh Medical Research Coun-
cil Bulletin, 33, 81-87.
[23] Brockington, I.F., Oates, J., George, S., Turner, D., Vostanis,
P., Sullivan, M., Loh, C. and Murdoch, C. (2001) A screen-
ing questionnaire for mother-infant bonding disorders. Ar-
chives of Women’s Mental Health, 3, 133-140.
doi:10.1007/s007370170010
[24] Brockington, I.F., Fraser, C. and Wilson, D. (2006) The
postpartum bonding questionnaire: A validation. Archives
of Women’s Mental Health, 9, 233-242.
doi:10.1007/s00737-006-0132-1
[25] Edhborg, M., Nasreen, H.E. and Kabir, Z.N. (2011) Im-
pact of postpartum depressive and anxiety symptoms on
mothers’ emotional tie to their infants 2 - 3 months post-
partum: A population based study in rural Bangladesh.
Archives of Women’s Mental Health , 14, 307-316.
doi:10.1007/s00737-011-0221-7
[26] Siu, B.W.-M., Ip, P., Chow, H.M.-T., Kwok, S.S.-P., Li,
O.-L., Koo, M.-L., Cheung, E.F.C., Yeung, T.M.-H. and
Hung, S.-H. (2010) Impairment of mother-infant relation-
ship. Validation of the Chinese version of postpartum bond-
ing questionnaire. Journal of Nervous and Mental Dis-
ease, 198, 174-179.
doi:10.1097/NMD.0b013e3181d14154
[27] Murray, L., Fiori-Cowley, A., Hooper, R. and Cooper, P.
(1996) The impact of postnatal depression on early mother-
infant interaction and later infant outcome. Child Devel-
opment, 67, 2512-2526. doi:10.2307/1131637
[28] Gunning, M., Conroy, S., Valoriani, V., Figueiredo, B.,
Kammerer, M.H., Muzik, M., Glatigny-Dallay, E. and
Murray, L. (2004) Measurement of mother-infant interac-
tions and the home environment in a European setting:
Preliminary results from a cross-cultural study. British
Journal of Psychiatry, 184, 38-44.
doi:10.1192/bjp.184.46.s38
[29] Frankel, K.A. and Harmon, R.J. (1996) Depressed moth-
ers: They don’t always look as bad as they feel. Journal
of the Academy of Child Adolescent Psychiatry, 35, 289-
298. doi:10.1097/00004583-199603000-00009
[30] Hornstein, C.H., Trautman-Villa, P., Hohn, E., Rave, E.,
Wortmann-Fleischer, S. and Schwarz, M. (2006) Mater-
nal bond and mother-child interaction in severe postpar-
tum psychiatric disorders: Is there a link? Archives of
Women’s Mental Health, 9, 279-284.
doi:10.1007/s00737-006-0148-6
[31] Field, T. (1995) Infants of depressed mothers. Infant Be-
havior and Development, 18, 1-13.
doi:10.1016/0163-6383(95)90003-9
[32] Francis, L., Weiss, B.D., Senf, J.H., Heist, K. and Har-
graves, R. (2007) Does literacy education improve symp-
toms of depression and self-efficacy in individuals with
low literacy and depressive symptoms? A preliminary in-
vestigation. Journal of the American Board of Family
Medicine, 20, 23-27. doi:10.3122/jabfm.2007.01.060058
[33] Kemppinen, K., Kumpulainen, K., Räsänen, E., Moilanen,
I., Ebeling, H., Hiltunen, P. and Kunelius, A. (2005)
Mother-child interaction on video compared with infant
observation: Is five minutes enough time for assessment.
Infant Mental Health Journal, 26, 69-81.
doi:10.1002/imhj.20031
[34] Edhborg, M., Seimyr, L., Lundh, W. and Widström, A.-M.
(2000) Fussy child—Difficult parenthood? Comparisons
between families with a “depressed” mother and non-de-
pressed mother two months postpartum. Journal of Re-
productive and Infant Psychology, 18, 225-238.
doi:10.1080/713683036