Bonding is the process of developing a relationship and attachment between mother and newborn child. In general, the maternal instinct begins when the mother meets her baby for the first time. A bonding disorder occurs when there is difficulty forming an attachment between mother and child. Bonding disorder has been acknowledged as an important problem in perinatal psychiatry.[1, 2] However, it has not been recognized as a nosological category either in the ICD-10 or the DSM-IV. Several factors can lead to the onset of bonding disorder, for example: death of a twin baby, stillbirth, painful delivery, or undesired pregnancy. It has also been reported that infants with diseases, handicaps, or delays in social interaction due to premature birth are risk factors for bonding disorder. In addition, the mother's severe anxiety, a compulsive temperament, or distress due to difficulty with feeding, calming the infant, and getting the child to sleep can interfere with the development of attachment and severe parenting problems, such as child neglect or abuse, may impact effective bonding.
Depression and low mood are common symptoms that may be present during gestation, the puerperal period, or the menopausal period. Therefore, biological factors, such as hormonal imbalance, and psychosocial factors, such as changing roles of home and society, might influence the onset of depression or influence mood. ‘Postpartum depression’ is used to describe a continuum of depressive symptoms and diagnoses that occurs several weeks after childbirth. Postpartum depression occurs in approximately 10–15% of women.[2, 5-10] A recent clinical study showed that estimated rates of depression during pregnancy vary from 7 to 15% in developed countries to 19 to 25% in developing countries. Moreover, more than half of women with postpartum depression suffer from depression before or during pregnancy. Studies focused on a Japanese population reported that about 16% of women experience depression or low mood during pregnancy and postpartum.[13, 14]
The gestation and puerperal periods are very important for developing mother–child bonding and the mother's depression during these critical periods may prevent effective bonding. Furthermore, the importance of bonding for the cognitive and behavioral development of infants is well known. When postpartum depression is prolonged, the adverse influences on mother–child bonding become more prominent. Specifically, the risk of abuse and neglect increase if the mother's depression is not treated. Moreover, long-term effects continue to impact the mother–child dyad. One study suggested that children of mothers who have had postpartum depression are more likely to have cognitive and emotional problems throughout development.
Several questionnaires have been used to assess bonding in severely ill clinical inpatients, mothers whose babies were in intensive care, and maternal representations of attachment during pregnancy.[19, 20] Moreover, many groups have studied the process of bonding between mother and child. Specifically, Brockington et al. reported on the use of the Postpartum Bonding Questionnaire, a self-rating instrument to assess the mother–infant relationship, as a screening instrument to detect bonding problems in obstetric and primary care services. Taylor et al. developed the Mother-to-Infant Bonding Scale (MIB) to screen the general British population for postpartum difficulties in relation to maternal emotions toward the baby.
There are two major hypotheses regarding bonding disorder. One hypothesis suggests that postpartum depression is the primary disorder and that the bonding disorder is caused by improper treatment of the mother's depression. The other theory suggests that the mother-to-infant relationship is the primary concern, and that bonding disorder can occur even if maternal depression is not present. Righetti-Veltema and colleagues reported that mother–infant relationship was correlated to the Edinburgh Postnatal Depression Scale (EPDS), but an intermediate or pathological score on mother–infant relationship tests was observed in 8.8–21.8% of mothers without depressive symptoms. In other words, there is a controversy about whether the mother's mood disorder, such as depression, secondarily causes trouble with relations with the child, or if bonding development and difficulty in caring for the infant influence the mother's mental health.
Kumar analyzed features of 44 women who showed serious maternal affective disorders. Results showed that mothers who did not have any mental disorders after childbirth did not have inadequate maternal emotions. However, he reported that he could not completely clarify the contexts because the onset of the mental disorder and bonding disorder were so close, and the progression of the mental disorder did not correspond completely with the clinical course of the bonding disorder. Condon studied bonding dynamics between mothers and unborn children during pregnancy. According to him, there is a pathological dyad between a mother and her fetus that starts before the baby is born, and he reported some cases of fetal abuse.[22, 23] A similar study was conducted in Japan, and results suggested an association between depressive tendency of a mother and attachment to the unborn child. However, there is no report that examines changes in the relation of depressive symptoms and bonding disorder from the pregnancy period to after childbirth using the same evaluation scale. Moreover, a large number of studies have examined and reported the rates of depression among women in the postpartum period,[2, 12] but in most of them, samples were collected after the onset of depression; therefore, the possibility of recall bias cannot be excluded.
From a preventive point of view, the potential links between maternal pre/postnatal mood and mother–infant bonding need to be investigated in greater detail. Therefore, we conducted a prospective study using self-rating questionnaires to investigate the correlation between bonding disorder and maternal depression during pregnancy and after childbirth. The objectives of this study were to analyze the course of infant–mother attachment, and to assess the association between bonding disorder and maternal depressive state during pregnancy and the postpartum period.
- Top of page
Table 1 shows the mean age and number of women in each of the four groups based on EPDS scores, after exclusion of one subject who was an extreme outlier in terms of EPDS and MIB scores. The mean age of subjects was 31.7 years (range, 20–44 years; SD, 4.3 years). Of the 388 women, 282 (72.7%) belonged to the non-depressive group; 39 (10.0%) were in the temporary gestational depressive group; 27 (7.0%) were in the continuous depressive group; and 40 (10.3%) were in the postpartum depressive group. Although uniparous mothers were more often found to be in the postpartum depressive group, there was no statistical difference regarding experiences of delivery among the four groups (Table 1). There were no differences regarding income and academic backgrounds.
Table 1. Women who completed the EPDS at all four periods were divided into groups‡ based on results of the EPDS scores
|Pregnancy||Postpartum||Groups†||n (%)||Age (years ± SD)||Parity|
|Early pregnancy||Late pregnancy||1 month||Uniparous||Biparous||Multiparous|
|−‡||−||−||Non-depressive||282 (72.7%)||31.8 ± 4.1||176||86||20|
|+§||−||−||Temporary gestational depressive||39 (10.0%)||31.9 ± 5.4||26||9||4|
|−||−||+||Postpartum depressive||40 (10.3%)||32.0 ± 4.1||34||4||2|
|+||−||+||Continuous depressive||27 (7.0%)||30.5 ± 4.4||16||9||2|
The hypothesis that maternal mood was related to bonding was tested with a Pearson correlation of EPDS measures. There were weak to moderate correlations (r = 0.24–0.39; P < 0.01) between the EPDS and MIB scores during early pregnancy, during late pregnancy, and at 1 month postpartum (Fig. 1a,b,d); the weakest correlation (r = 0.142) was observed at 5 days after delivery (Fig. 1c). The prospective design of our study gave us the opportunity to examine the difference in the developmental trend of the MIB score among the four EPDS groups.
Figure 1. Correlation between Mother-to-Infant Bonding Scale (MIB) and Edinburgh Postnatal Depression Scale (EPDS) scores. The data-points are shown as numbers, while each number represents count of subjects at each data-point. The data-density (count of subjects per data-point) is shown using the blue color gradient. [Correction added on 30 July 2014, after first online publication: Figure was redrawn and replaced to improve its clarity.]
Download figure to PowerPoint
The MIB score of the continuous depressive group was the highest in all groups during pregnancy, decreased after delivery, and then went up at 1 month postpartum (Fig. 2). The score of the postpartum depressive group was lower than the continuous depressive group's during pregnancy, but increased after childbirth (Fig. 2). The mean, SD, median, and quartile MIB score for each period of the current study are shown in Table 2. Four groups based on the EPDS scores and four groups based on the MIB scores were summarized in the cross table, and we calculated the proportion of the four groups based on MIB score by each group based on the EPDS score (Table 3). A majority of the subjects did not score over the MIB scale threshold. However, we detected tendencies that the gestational bonding disorder group made up a large proportion of the temporary gestational depressive group, and the continuous bonding disorder group represented a large proportion of the continuous depressive group. Analysis revealed a statistical association between the proportion of women in the four MIB groups and the proportion of women in the four groups based on EPDS (Fisher's exact test P < 0.0001).
Table 2. Mean and median of MIB scores
| ||Early pregnancy||Late pregnancy||At 5th day||At 1st month|
|Mean (SD)||2.09 (2.49)||2.15 (2.43)||0.89 (1.55)||1.28 (1.82)|
|Median (25–75%)||1 (0–3)||1 (0–3)||0 (0–1)||0 (0–2)|
Table 3. Number (%) of women in four groups based on MIB† by four groups based on EPDS‡
|Groups† based on EPDS score||Four groups based on MIB Score n (%)|
|No bonding disorder||Gestational bonding disorder||Postpartum bonding disorder||Continuous bonding disorder|
|Non-depressive||216 (55.7%)||44 (11.3%)||11 (2.8%)||11 (2.8%)|
|Temporary gestational depressive||26 (6.7%)||10 (2.6%)||2 (0.5%)||1 (0.3%)|
|Postpartum depressive||22 (5.7%)||6 (1.5%)||8 (2.1%)||4 (1.0%)|
|Continuous depressive||12 (3.1%)||6 (1.5%)||3 (0.8%)||6 (1.5%)|
- Top of page
In the current study, we observed a weak to moderate correlation between the EPDS scores and the MIB score during early pregnancy, during late pregnancy, at the 5th day after delivery, and at 1 month postpartum. There was a tendency that higher EPDS scores were associated with more severe bonding disorders. An interaction effect was observed in the developing pattern of the MIB between the four groups based on the EPDS scores. The MIB scores of the continuous depressive group were the highest in comparison with the other groups during pregnancy, decreased once after delivery, and then went up at 1 month postpartum. The scores of the postpartum depressive group were lower than the continuous depressive group's during pregnancy, but increased after childbirth. When we calculated the proportion of the four groups based on the MIB scores and of each group based on the EPDS, there were various patterns in the development of bonding. In other words, some of the mothers who had low MIB scores were not in a depressive state at 1 month postpartum, but others were. In Robson and Kumar's study, although some mothers were indifferent to their newborn babies 7 days after delivery, most mothers developed bonding within 1 week. Some studies indicated that women with low bonding in late pregnancy were at risk of poorer bonding with their infant postpartum.[19, 29] However, until this study, no report has examined the longitudinal trend of bonding from early pregnancy to postpartum using the same questionnaire.
In the current study we observed that the MIB scores after childbirth were lower than during pregnancy. Some mothers had not developed an appropriate maternal emotional response during pregnancy, but bonding gradually formed. The MIB score at 5 days after delivery was the lowest of the four time-points, and the score 1 month postpartum rose slightly. This differs from findings of Taylor et al., who showed that average bonding scores 12 weeks after childbirth were 67% lower than those recorded 3 days after childbirth, indicating that mother–infant bonding scores had improved considerably between these two occasions. In Japan, most women are still in the hospital 5 days after delivery. We believe that the MIB score at 5 days in our study was slightly lower than the score reported in the previous UK study because mothers did not realize the difficulty of child care and the associated lack of sleep that accompanies caring for an infant. The four groups, as classified by the EPDS, were different in terms of the development pattern of MIB. On the one hand, the MIB score in the postpartum depressive group was low during pregnancy, which was a similar pattern to the non-depressive group. On the other hand, the MIB score in the continuous depressive group was already high during pregnancy, which suggested an association between depressive state and bonding disorder. The continuous depressive group had high EPDS scores in the early pregnancy period; they might appear to have experienced low mood before pregnancy, but that was not confirmed in the current study.
It has been shown that depressive symptomatology during pregnancy can affect the development of mother-to-infant bonding. In reference to the controversy of whether depression or bonding disorder is the primary disorder, for mothers in the continuous depressive group, depression was regarded as a causal factor in previous studies, as the mothers had depressive mood during pregnancy. On the other hand, for mothers in the postpartum depressive group, the bonding disorder might be one of the factors influencing low mood. In a recent study, it was reported that poor attachment in late pregnancy predicted postpartum depression, but that study did not completely clarify the context of its findings. Our colleagues have observed different time sequences regarding the depressive state. These differences were not investigated in previous studies focused on the association between bonding disorder and maternal depression.[3, 15, 25, 32] Furthermore, we found that there were some women with a high MIB score even though they did not experience low mood. This result is consistent with previous studies,[3, 33] indicating that the progression of the depressive state does not correspond completely with the clinical course of bonding disorder, and suggesting that, in some cases, bonding disorder is not a direct consequence of the depressive state. Therefore, in order to investigate factors related to the development of bonding disorder that are not related to depression, research focused on the mother's social interactions and attachment style is needed.
The following cautionary details should be considered when interpreting the results of the current study. We did not use a psychiatric interview to establish the diagnosis of depression. We used self-administered questionnaires and we did not analyze comparative performance of the Japanese version of EPDS and other depression rating scales, such as the Beck Depression Inventory or the Japanese version of the Quick Inventory of Depressive Symptomatology. It is of note, however, that EPDS was validated in a Japanese sample for screening for depression. In addition to the high false-negative rate in the self-rating scale, due to fear of negative evaluation, it is important to consider that cognitive deviations due to depressive influences may influence the results of this type of questionnaire. Although the MIB is short and easy to use, the cut-off point in the present study was not verified in the clinical setting. Furthermore, the validity and reliability of the MIB for assessing attachment during pregnancy were not verified in the current study; this questionnaire was originally developed to assess mother-to-infant bonding after childbirth and its use during pregnancy has not been verified. Another limitation is that bonding disorder of women who were identified as belonging to the postpartum depressive group may have been influenced by various factors related to delivery (such as hormonal regulation, obstetric complications, and maternal blues), which were not considered in the current study. Furthermore, as we investigated bonding only from pregnancy to 1 month postpartum, additional research regarding bonding disorder after 1 month and its influence on cognitive development is needed. In addition, as significant changes have been found in physiological and psychological functions in different trimesters of pregnancy,[35, 36] a time-specific threshold might be appropriate. However, as these data have not been validated in the Japanese population, we used the same cut-off (8/9) in the present study.
Finding potential risk factors during pregnancy is extremely important not only for the mother but also for the newborn child. To clarify the risks related to bonding disorder, future studies should include a larger sample size, as well as additional information regarding social support from a woman's spouse and mother, the mother's temperament, and the mother's parenting experience.
- Top of page
Funding for this study was provided by research grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan and the Ministry of Health, Labor and Welfare of Japan, the intramural Research Grant (21B-2) for Neurological and Psychiatric Disorders of National Center for Neurology and Psychiatry and the specific research fund 2012 for the Great East Japan Earthquake Revival by the New Technology Development Foundation. We declare that we have no conflicts of interest. Authors' Contributions: S.G., S.M., A.K., T.M. and N.O. designed the study. H.O., N.I. and K.F. participated in collecting clinical information and samples. H.O., T.K., S.G., S.M., A.K., T.M., B.A. and N.O. analyzed, interpreted and discussed the results. H.O., T.K., B.A. and N.O. wrote the draft. H.O., T.K. and B.A. wrote the final manuscript.