Keywords perinatal depression - brooding - maternal-fetal bonding - adaptation to pregnancy
Being pregnant can be a very exciting yet also complex experience for many women.
The hormonal changes and adjustment difficulties that are typical of pregnancy can
cause frequent mood disturbances, which, if exacerbated, can lead to the onset of
mood disorders.[1 ] Specifically, perinatal depression has notably high incidence rates, particularly
during the third trimester of pregnancy.[2 ]
Perinatal depression comprises the presence of major depressive disorder symptoms,
with onset during pregnancy, lasting up to 3 months postpartum, after which it is
termed as postpartum depression.[2 ] The prevalence rate of perinatal depression is between 10 and 20% and rising to
26% in adolescent mothers.[3 ]
[4 ] However, despite this high prevalence, 14 to 25% of pregnant women who meet diagnostic
criteria for this problem are not diagnosed or receive psychological treatment.[2 ] This is due to the lack of training of health professionals to detect mental health
problems and the stigmatization of maternal depression, with depressive symptomatology
being attributed to the organic condition of being pregnant.[5 ]
Developing effective ways to detect and prevent the occurrence of depressive symptoms
during pregnancy is highly relevant, not only because of the medical and psychological
consequences that depression has on the mother and the baby, but also because of its
influence on the quality of the relationship that the mother will establish with him/her.[6 ] Studies show that depressive symptomatology in pregnancy, when not properly treated,
is associated with higher risk of premature birth, as well as with lower birth weight
and more heightened irritability in the newborn.[7 ]
[8 ]
[9 ] In addition, depressive symptomatology in pregnancy is considered a risk factor
for developing adequate maternal–fetal bonding. Several studies show a relationship
between depressive symptomatology during pregnancy, lower quality of breastfeeding
2 months after childbirth, and a poorer experience of motherhood 1 year after childbirth.[10 ]
In view of all these issues and given the relevance of establishing adequate models
for the prevention of perinatal depression, this study aimed to identify possible
mechanisms and paths influencing its development, going beyond the risk factors that
have been previously established in the literature, as detailed below.
Risk Factors for Perinatal Depression
Risk Factors for Perinatal Depression
Research has focused on the better understanding of depressive symptomatology during
pregnancy and its prevention. Different risk factors contributing to its development
have been established. The exhaustive study by Lancaster et al[11 ] established the main sociodemographic risk factors for the development of perinatal
depression as: low socioeconomic level, use of private medical insurance (the review
was conducted in a population within the American health system), among others. In
addition, as shown in other investigations, psychological or psychosocial factors
include the presence of anxiety during pregnancy,[12 ] having a previous history of depression,[13 ]
[14 ] the experience of negative stressful life events,[15 ]
[16 ]
[17 ] and a lower educational level.[18 ] Moreover, the lack of support from a significant other such as one's partner is
linked to an elevated risk of developing depressive symptoms during the third trimester
of pregnancy.[11 ]
[17 ]
[19 ]
[20 ]
Despite advances in understanding these risk factors, there is still considerable
room for improvement in predicting and detecting perinatal depression during pregnancy,
as discussed above. The most recent research highlights the need to better understand
the role of other psychosocial processes, both individual and relational, that have
not yet been fully considered in this field.[2 ]
[11 ] From this view, several key aspects that must be considered include: (1) how the
difficulties women have in adapting to their pregnancy, as well as (2) the way in
which they regulate the negative emotions this produces in them and (3) the quality
of the bonding they establish with their baby, can all relate and contribute to the
development of perinatal depressive symptoms.
Adaptation to Pregnancy
The transition to motherhood involves major adjustments and changes to all aspects
of a woman's life.[21 ] Adaptation to pregnancy is thus defined as the woman's acceptance responses to the
changes produced by her baby's development and the satisfaction of being pregnant.[22 ] It is fundamental to having a healthy experience of pregnancy.[23 ] A certain ambivalence is common among pregnant women, since total and complete adaptation
to pregnancy is unusual.[22 ]
[24 ] Nevertheless, insufficient adaptation to pregnancy is linked to lower tolerance
of changes and physical discomfort, a greater fear and anxiety, and a tendency to
experience low mood and/or despair that can lead to conditions of clinical depression
during pregnancy.[22 ]
[25 ]
The mechanisms involved in the contribution of low adaptation to pregnancy to perinatal
depression onset may not be particularly clear, a link could be mediated by various
factors: individual factor, by the way that negative emotions produced by the adaptation
difficulties of pregnant women are managed (brooding); and relational factor, by the
quality of the maternal–fetal bond that the mother is establishing with her baby as
a result of these adaptation problems. Both factors could be two of the main ways
through which depressive symptoms during the third trimester of pregnancy develop
because of low adaptation to pregnancy.
Brooding
Individual cognitive processes, such as how a pregnant woman interprets emotionally
ambiguous situations in her environment, as well as the thinking styles used to cope
with negative moods[26 ] can facilitate the development of perinatal depressive symptomatology.[27 ] In this sense, a ruminative style, characterized by higher attention to negative
cognitions,[28 ] could predict and maintain the negative mood states typical of depressive symptoms[29 ]
[30 ] acting as downward spirals.[31 ]
Rumination comprises the occurrence of recurrent thoughts that focus the individual's
attention on negative moods and their consequences.[32 ] Specifically, brooding involves persistent passive thinking in response to sad moods,
focused on past problems.[32 ]
[33 ]
Numerous studies show that higher levels of brooding are generally associated with
greater vulnerability to develop depressive episodes as well as with increases in
their duration.[34 ] In the case of pregnancy, higher levels of brooding have been linked to higher levels
of depressive symptomatology.[27 ] Specifically in the third trimester of pregnancy, brooding is a predictor of depressive
symptomatology[27 ]
[35 ]
[36 ] and thus an important subject of study as a possible mechanism involved in the development
of perinatal depression because of problems in adapting to pregnancy.
Maternal–Fetal Bonding
Poorer quality of bonding during pregnancy, because of pregnancy adaptation issues,
should also be considered as a clear predictor of perinatal depressive symptoms. Maternal–fetal
bonding is the mental representation a mother develops regarding her baby and her
care behaviors toward it (i.e., having good nutrition, trying to maintain a stable
state of mind, the experience of early motherhood or nesting).[37 ] An adequate bonding must be established to ensure the further development of the
child and its relationship with the mother. Good quality bonding during pregnancy
is associated with higher quality of mother–baby interactions 12 weeks after giving
birth.[38 ]
On the one hand, poor adaptation to pregnancy is one of the risk factors for the development
of low-quality maternal bonding.[23 ] Women who have more difficulty in adjusting to their pregnancy experience greater
ambivalence toward pregnancy and dislike of their physical image, exhibit fewer care
behaviors, and feel less attached to their baby,[39 ] resulting in poorer quality bonding.[23 ] On the other hand, poorer quality maternal–fetal bonding is related to the development
of depressive symptomatology.[1 ]
[40 ]
[41 ] Bonding can be considered as a protective factor during pregnancy against difficulties
and/or tensions that may arise in the adjustment to pregnancy, including in the later
postpartum stage.[1 ]
Given the clear evidence of adaptation to pregnancy problems as direct contributors
to the establishment of worse bonding[23 ] and the well-established connection between pregnancy adaptation issues and the
development of perinatal depression,[22 ] the possible mediating role of low bonding in the latter relationship between adaptation
to pregnancy and perinatal depression must be considered.
Study Aims
The objective of this study was to test the association between symptoms of perinatal
depression and poorer adaptation to pregnancy, in which brooding and bonding act as
mediating variables in the relationship between poorer adaptation to pregnancy and
the development of perinatal depression. Thus, we predicted that poor adaptation to
pregnancy would predict elevated levels of brooding and low levels of bonding, which
in turn would lead to higher levels of depressive symptoms during the third trimester
of pregnancy.
First, previous literature shows that women with poor adaptation to pregnancy may
present negative mood states that could lead to depressive symptoms.[22 ] It has been suggested that the appearance of such depressive emotional states would
be mediated by the mother's ruminative response style, as a maladaptive emotion regulation
strategy in response to problems of adaptation.[26 ] Research has also shown that depressive rumination is a predictor of depressive
symptomatology in the last trimester of pregnancy.[27 ]
[34 ]
[35 ]
[36 ] For this reason, a path model was conducted to analyze the hypothesis that poorer
adaptation to pregnancy would play an indirect predictive role in perinatal depressive
symptom levels through its direct influence on brooding rumination (hypothesis 1,
i.e., adaptation to pregnancy → brooding → perinatal depressive symptoms).
Second, studies indicate that mothers with poorer adaptation to pregnancy have greater
difficulties in establishing a bonding of quality with their babies.[23 ] In turn, this mother–baby bond established during pregnancy has been considered
a protective factor against the difficulties and/or emotional tensions that could
appear during pregnancy or postpartum.[1 ] Therefore, mothers establishing poorer quality bonding due to the occurrence of
problems in adapting to pregnancy would present a greater risk of suffering from depressive
symptoms. Based on this, we used the path model to analyze the hypothesis that a poorer
adaptation to pregnancy would play an indirect predictive role in perinatal depressive
symptoms through its direct influence on a worser maternal–fetal bonding (hypothesis
2, i.e., adaptation to pregnancy → bonding → perinatal depressive symptoms).
Materials and Methods
Design : an online survey was created which included informed consent, and assessment of
socio-demographic and psychological variables. For dissemination, the study information
and the link to access the survey were shared via social networks. This way, a nondiscriminatory
exponential chain sampling procedure (i.e., snowball sampling technique) was used.
The inclusion criteria for the study were women over 18 years of age, of Spanish nationality,
and in the third trimester of pregnancy. The sample was recruited between November
2021 and February 2022.
Participants : a nondiscriminatory exponential chain-referral sampling procedure (i.e., snowball
sampling technique) was performed to recruit the sample between November 2021 and
February 2022. The dissemination of the questionnaires was performed through social
networks. The inclusion criteria to participate in the study were women over 18 years
of age and in the third trimester of pregnancy. An initial sample of 648 was recruited,
of which 54 subjects were discarded for not meeting the inclusion criteria. The final
sample consisted of 594 women in the third trimester of pregnancy, with 75% of them
being primiparous. All the participants were of Spanish nationality. The average number
of weeks of pregnancy at the time of participation in the study was 33.26 (standard
deviation [SD] = 3.88). The mean age was 32.40 years old (SD = 4.19). Descriptive
data of the sociodemographic characteristics of the participants and dichotomous psychological
measures are shown in detail in [Table 1 ]. All participating women signed an informed consent. The institutional ethics committees
approved the study protocol (Universidad Pontificia Comillas). No type of remuneration
or incentive was offered to the subjects for their participation.
Table 1
Descriptive data of demographic characteristics and dichotomous psychological measures
Variable
Total sample (N = 594)
N
%
Number of children
0
444
74.6
1
133
22.5
2
14
2.4
3
3
0.5
Educational level (%)
Primary
23
3.4
Secondary
113
19.0
University graduate
458
77.4
Marital status (%)
Single
19
3.2
Married or living together
575
96.8
Type of medical care (%)
Public health care
386
64.8
Health insurance
122
20.7
Private
86
14.5
Previous abortions (%)
No
585
98.3
Yes
9
1.7
Pregnancy complications (%)
No
455
76.9
Yes
139
23.1
Instruments : first, the survey included questions about sociodemographic data: age, week of gestation,
type of medical care received, marital status, educational level, and nationality.
Second, the following questionnaires were used to evaluate the main variables under
study (perinatal depression, adaptation to pregnancy, brooding, and bonding quality).
Mean and standard deviations of the psychological measures are depicted in [Table 2 ] (see in the Results section)
Table 2
Mean, standard deviation, and bivariate correlations of psychological variables
Measure
r (p -value)
M (SD)
Adaptation to pregnancy
Bonding
Brooding
Perinatal depression
Adaptation to pregnancy
37.17 (4.58)
1
Bonding
29.61 (4.64)
0.545[a ]
1
Brooding
3.79 (3.06)
−0.369[a ]
−0.130[a ]
1
Perinatal depression
8.54 (4.99)
−0.375[a ]
−0.111[a ]
0.528[a ]
1
Abbreviations: M, mean; SD, standard deviation.
a Correlations significant at p < 0.01.
Perinatal depression : the Edinburgh Postnatal Depression Scale[42 ] has been validated for use in the perinatal stage. This is a 10-item scale and items
are rated on a 4-point Likert scale. The cutoff score of ≥13 indicates an elevated
risk of depression. The Cronbach's α of the Spanish adaptation of the instrument was
0.91[43 ] and in this study it was α = 0.86.
Adaptation to pregnancy and the maternal–fetal bonding : these variables were assessed using the Affective Bonding and Prenatal Adaptation
Scale,[44 ] which has been adapted for its use with pregnant women from the second trimester
of pregnancy onward.[45 ] It is a 21-item scale (12 items assessing adaptation to pregnancy, 9 items assessing
bonding). Items are rated on a 5-point Likert scale. This version presents α coefficients
of 0.74 for the adaptation subscale and 0.74 for the bonding subscale. The α coefficients
in this study were α = 0.56 for the adaptation subscale and α = 0.73 for the bonding
subscale.
Ruminative responses : we used the Ruminative Response Scale,[46 ] which allows the evaluation of individual ruminative response styles, specifically
brooding rumination. It is a 22-item scale, with items rated on a 4-point Likert scale.
The Spanish adaptation shows a Cronbach's α of 0.93 on the total scale.[47 ] The α coefficients in this study were α = 0.93 for the total scale and α = 0.80
for the brooding subscale.
Other psychological risk factors identified in previous literature (i.e., experience
of stressful life events, history of previous depression, general anxiety and social
support) were also assessed and controlled for as covariates in the study. All information
regarding these variables can be found in the [Supplementary Appendix ] (available in the online version).
The Total Scores of Each Instrument Were Used to Compare the Variables with Each Other
The Total Scores of Each Instrument Were Used to Compare the Variables with Each Other
Statistical Analyses
Statistical analyses were performed using IBM SPSS Statistics version 22 software.
First, descriptive analyses of sociodemographic data and psychological measures of
the participants were executed. Second, Pearson bivariate correlation analyses were
performed to test the association between the main variables of the model (i.e., adaptation
to pregnancy, bonding, brooding, and perinatal depression) and the rest of the dimensional
risk factors under study (i.e., general anxiety and social support levels). In accordance
with Cohen's[48 ] criteria, the following magnitudes were applied in interpreting the results: between ± 0.10
and ± 0.29 low; between ± 0.30 and ± 0.49 medium or moderate; between ± 0.50 and ± 1.0
high correlation..
Finally, we tested a path model using a structural equation that included the full
set of variables that were significantly correlated (i.e., adaptation to pregnancy
→ brooding/bonding → perinatal depression). In this model, adaptation to pregnancy
acted as an exogenous variable, predicting perinatal depression symptomatology levels
directly and indirectly through the influence of brooding (i.e., mediator 1) and bonding
quality levels (i.e., mediator 2). The estimation of standardized parameters of the
path model was conducted using the full information maximum likelihood estimation
method. To test model fit, we used standard criteria[49 ]: (1) χ2: nonsignificant value; (2) χ
2 /gL: values lower than 2; (3) CFI (comparative fit index) and TLI (Tucker–Lewis index):
values ≥0.95; (4) RMSEA (root mean square error of approximation): values ≤0.05. The
hypothesized mediation pathways (i.e., adaptation to pregnancy → brooding → perinatal
depression; adaptation to pregnancy → bonding → perinatal depression) were tested
via estimation of indirect effects within the full path model. Structural equation
models and resulting path analyses were conducted using AMOS v18.0 (SPSS).
Results
Descriptive data of the sociodemographic characteristics of the participants and dichotomous
psychological measures are shown in [Table 1 ]. [Table 2 ] shows the mean (SD) obtained by the sample in each instrument.
Bivariate Correlations
Bivariate correlation analyses showed significant correlations between the main variables
under study (i.e., adaptation to pregnancy, brooding, bonding, and symptomatology
of perinatal depression), and other of the further risk factors included in the study
(general anxiety and social support levels). The full set of correlation results is
shown in [Table 2 ].
Structural Equation Model
We tested a structural equation model (see [Fig. 1 ]) including the main variables under study (adaptation to pregnancy, brooding, bonding,
and symptomatology of perinatal depression), and modeling the hypothesized relations
between them (i.e., indirect effect paths), if they were all significantly correlated
in the previous analyses. In this model, a worse adaptation to pregnancy would predict
the presence of perinatal depressive symptoms in the third trimester of pregnancy
indirectly through two paths of mediation: higher levels of brooding (hypothesis 1)
and a worse quality of maternal–fetal bonding (hypothesis 2). All the goodness-of-fit
indicators were good, as shown in [Table 3 ].
Table 3
Goodness-of-fit indices for the tested path model
ꭓ
2 (gL)
p
ꭓ
2 /gL
CFI
TLI
RMSEA (90% CI)
Model
1.214 (1)
0.271
1.214
1.000
0.997
0.019 (0.000/0.113)
Abbreviations: CI, confidence interval; CFI, comparative fit index; RMSEA, root mean
square error of approximation; TLI, Tucker–Lewis index.
Fig. 1 Path model of the relation between adaptation to pregnancy, brooding, bonding and
perinatal depression.
Indirect effects were then tested using bias-corrected bootstrap estimations (2,000
bootstrap samples with 95% CI). A significant indirect effect of lower adaptation
to pregnancy on higher perinatal depressive symptoms via individual differences in
brooding (p = 0.001; standard error [SE] = − 0.172; 95% confidence interval [CI}: −0.215 to −0.137)
was found, supporting the first hypothesis. Additionally, the indirect effect of lower
adaptation to pregnancy on higher perinatal depressive symptoms, via lower maternal–fetal
bonding (p = 0.044; SE = 0.039; 95% CI: 0.007–0.072), was statistically significant, thus also
supporting the second hypothesis. A graphical representation of the entire path model
is provided in [Fig. 1 ].
Discussion
This research aimed to study the pathways through which risk factors explain depressive
symptoms in the third trimester of pregnancy, especially the quality with which a
woman adapts to her pregnancy. Previous research has focused on the study of sociodemographic
and psychological risk factors implicated in the development of perinatal depressive
symptoms, such as low socioeconomic status, private medical insurance, and not living
with the significant other[11 ]; presence of maternal anxiety[12 ]; previous history of depression[13 ]
[14 ]; experience of negative stressful life events[15 ]
[16 ]
[17 ]; and lack of adequate social support network.[11 ]
[17 ]
[19 ]
[20 ] However, these previous studies have not considered the further possible pathways
for the development of depression, through both individual and relational mechanisms,
that can result from the difficulties experienced in adapting to pregnancy. Special
attention in this study was therefore paid to the difficulties that some mothers may
have in adapting to the changes that their pregnancy entails, since this can influence
their use of emotion regulation strategies and the quality of the bond they establish
with their babies, both potentially mediating the development of depressive symptoms.
Although the data collected in this study are cross-sectional, which makes it difficult
to establish the temporal order of the influences between these factors, in keeping
with the main objective, a path was formulated to capture the relationship between
these variables. The results obtained in the path model show a clear indirect effect
of adjustment to pregnancy on perinatal depressive symptomatology through the mediation
of both parenting (hypothesis 1) and low-quality attachment (hypothesis 2).
First, better adaptation to pregnancy was related to women's lower levels of brooding
and depressive symptomatology, in line with previous research.[22 ] In turn, more brooding was linked to higher levels of depressive symptomatology
in the third trimester of pregnancy, congruent with other studies.[27 ]
[35 ]
[36 ] Path analysis confirmed these relationships, as well as the mediating role of brooding
in the relationship between poorer adaptation to pregnancy and higher levels of perinatal
depressive symptoms. These results are in line with cognitive theories regarding basic
processes leading to depression,[50 ] suggesting that a low tolerance to bodily changes and physical discomfort during
pregnancy and a significant increase in fears and anxiety in this period (i.e., problems
of adaptation to pregnancy) could be equivalent to experiencing pregnancy as a stressful
life event or a set of adverse experiences. Such experiences would be associated with
the use of a ruminative coping style, in line with the theory of response styles and
the empirical research derived from it[27 ]
[28 ]
[34 ]
[35 ]
[36 ] that would ultimately contribute to higher levels of perinatal depressive symptoms.
In turn, brooding could act as a feedback factor, maintaining the symptoms throughout
the final trimester of pregnancy, as well as beyond it in the postnatal period.[29 ]
[30 ]
Second, the results also supported the conclusions of Hidalgo and Menéndez,[23 ] that a more satisfactory adaptation to pregnancy is related to a higher quality
of bonding during pregnancy. Thus, results suggest that if the mother accepts the
changes she experiences during pregnancy, she will develop an adequate maternal response
toward the baby from her earliest interactions.[51 ] Higher levels of bonding were associated with lower levels of brooding and depressive
symptomatology. Path analysis confirmed these relationships, as well as the mediating
role of lower quality bonding in accounting for the relationship between poorer adaptation
to pregnancy and higher levels of perinatal depressive symptoms. However, the direction
of the relationship between bonding and perinatal depressive symptoms would be inverse
to that considered in other research that have rather considered bonding problems
as a consequence of perinatal depression,[10 ]
[40 ]
[41 ] thus suggesting an alternative model in which the quality of bonding could also
be considered a protective factor during pregnancy against the difficulties and/or
tensions that may arise, including increases in perinatal depression in the third
trimester of pregnancy.[1 ]
Limitations and Strengths
Limitations and Strengths
This study has some limitations. First, the measurement instruments used were self-report
questionnaires, and doing it online made it impossible to guarantee that participants
completed them in optimal circumstances and times. Moreover, the internal consistency
of the adaptation to pregnancy subscale was relatively low (obtaining a Cronbach's
α of 0.58). Given the importance of this risk factor, this makes it important to replicate
our findings with other instruments assessing adaptation to pregnancy that have better
psychometric properties. It is also important to consider the global pandemic situation
caused by COVID-19 occurring at the time of data collection for this study and how
this may have influenced the characteristics of the sample. Indeed, the prevalence
of depressive symptomatology in our sample, representative of the population studied,
was twice as high in pregnant women compared with prepandemic data.[8 ]
As for the strengths of the study, the sample size was high considering that this
type of population is usually less likely to participate in studies of this nature
given the state of gestation. In turn, the results obtained belong to a representative
sample of the population, as can be seen in the percentage of participants with depression
during pregnancy (21% of the sample exceeded the cut-off point for perinatal depression)
and can therefore be generalized to the entire population of pregnant women in the
third trimester of pregnancy. Finally, the results support an alternative direction
to the one typically considered in the literature, in the relation between the quality
of bonding and perinatal depressive symptoms, such as assessing the influence of individual
process variables in the development of depressive symptomatology in pregnancy, as
in the case of brooding.
Implications
These findings are highly informative and can be taken as a starting point for the
development of longitudinal studies, with the aim of confirming the temporal relationship
between these variables to improve knowledge about the risk factors present in the
final period of pregnancy and their subsequent influence on postpartum mood and adaptation
to motherhood.
Continuing in this direction, correct identification and early detection are framed
as fundamental aspects for future prevention and treatment.
Similarly, the development of new strategies to reduce the occurrence of early mood
problems during pregnancy should be considered essential, particularly through the
development of preventive interventions against melancholic rumination and to prevent
difficulties in establishing maternal–fetal bonding.
Our results show that it is also important in this type of interventions to consider
the style of the woman's attachment bond with her own mother, since this is an essential
component for the development of maternal identity.
Conclusion
The negative feelings a woman may have in adapting to the changes produced by her
pregnancy, leading to an unsatisfactory experience of it, can act as a risk factor
for the onset of perinatal depressive symptoms in the third trimester of pregnancy.
This relationship would be mediated by two factors, the presence of brooding and a
low quality of the maternal–fetal bond. Poor adaptation to pregnancy could predict
a maladaptive coping style in the mother, comprising a higher use of brooding rumination,
in turn predicting depressive symptoms. Lower adaptation to pregnancy would also influence
the adequate establishment of good-quality bonding, reducing its protective effect
against the development of perinatal depressive symptoms. This model provides a new
direction of interpretation of the relationship between bonding and perinatal depressive
symptoms, as well as identifies new risk factors that can improve the adequate detection
and prevention of perinatal depressive symptoms in the third trimester of pregnancy.
