Some mothers find it hard to relate to their new baby, and such failure may have long-term effects on the infant. This has been a neglected area of research. A new simple 8 item self-rating mother-to-infant bonding questionnaire has been designed to assess the feelings of a mother towards her new baby. A principal components and reliability analysis demonstrated an alpha score of 0.71. One hundred and sixty two women filled in the Kennerley Blues Scale, the Edinburgh Postnatal Depression Scale (EPDS) the Highs Scale and the new Mother to Infant Bonding Scale on day 3 postpartum. Twelve weeks later they were sent the EPDS and the Bonding scales again. One hundred and forty four returned all questionnaires. There was a strong correlation between the Bonding scores at 3 days and at 12 weeks (r(s)=0.54 p<0.001). Multiple regression analysis showed that those with raised Blues scores had worse, and those with raised Highs scores had better bonding at 3 days. Those with raised EPDS scores at 3 days (13 and over) had worse bonding scores in the “first few weeks” (median 4 versus 1, p = 0.028), as recalled at 12 weeks. This simple questionnaire is acceptable for use with mothers and gives significant correlations with their early mood.
INTRODUCTION:
The importance of early human relationships, especially with primary caregivers, for the cognitive and behavioural development of infants is well established. Bowlby said that “mother love in infancy and childhood is as important for mental health as are vitamins and proteins for physical health” (Rutter 1972). Much research in this area has focussed on attachment (Bowlby 1958; Maccoby & Masters 1970; Ainsworth & Bell 1970; Rajecki 1978), which has been defined as “a close emotional relationship between two persons, characterised by mutual affection and a desire to maintain proximity.” (Shaffer 1993), and “the dyadic regulation of emotion.” (Sroufe 1996). One key aspect of this is that attachments are reciprocal. The child must be old enough to exhibit emotion on separation and on being reunited as well as the ability to move to or from the primary carer. To assess and identify the type of attachment the Strange Situation Test was developed for use with 12 month old children (Ainsworth et al 1978). However the relationship between a mother and her baby starts at birth or even before. It is also important to study the very early relationship between a mother and her baby, and the mother’s early attitude to her baby. Robson and Kumar (1980) showed that many mothers were indifferent to their babies on first holding them, although most developed affection within the first week. However some mothers with depression have prolonged problems in developing a loving attitude to their child (Kumar 1997). There is now considerable research into perinatal disturbances of affect, and its effect on the development of the child (Hay et al 2001; Hoffman & Drotar 1991; Murray et al 2003). This work suggests that children of a mother who has had postnatal depression are more likely to have cognitive and emotional problems throughout development. Murray et al (1996), using video analysis 3methods, have shown that depressed mothers have problems interacting with their babies from about two months, These problems may well start even earlier in the postnatal period. In a recent review Brockington (2004) underlines the importance of disorders in the mother-infant relationship with regard to the long-term effect on the infant. He also emphasises that intervention methods are available, if the problem can be detected.Klaus, (Klaus et al 1972) suggested that mothers who spent even a few hours in skin to skin contact with their baby in the immediate postpartum period had a stronger relationship several months later. He described this as better bonding. This work, generally accepted at first, was later criticised (Lamb 1982; Myers 1984). Other research showed that there was little difference between mothers who had spent extended physical contact with their newborn infant and others in their later maternal behaviour (Carlsson 1979). The idea of an early sensitive period for mother-infant bonding became very controversial, and the concept of the importance of early mother to infant bonding fell into disrepute. This may have been premature. Kumar (1997) published a study of 44 women, contacted through the Association for Postnatal Illness, who described problems in the early relationship with their infants. They gave detailed accounts of absent affection, sometimes hate, rejection, neglect or impulses to harm. He described this as evidence for a disorder of mother-to-infant bonding. The work described here arose out of this study. The terms used in the questionnaire are based on common themes, which were identified in two ways. First, by analysis of the letters the women wrote describing their emotions and experiences, and secondly from data collected in response to a series of repeated questions detailing maternal feelings after the birth, 4changes over time and current feelings in relation to their infant. In the current study the term bonding is used to describe how the mother feels towards her infant, and is different from attachment, which includes the infant’s behaviour towards the mother. The aim was to devise a simple self rating questionnaire that could be used to screen the general population for problems in the mother’s feelings towards her new baby, initially for research purposes, and later possibly for clinical use. The only previous rating scales in this area are either very simple (3 item) (Fleming et al 1988), for use with a severely ill clinical population (Kumar and Hipwell 1996), or based on attitudes of mothers whose babies were in intensive care (Nagata et al 2000). The study described here was carried out with the aim of studying the acceptability of the scale with a general population, its internal consistency, its score distribution, and the relationship between its scores and early maternal mood
PATIENTS AND METHODS:
Mothers, who had given birth at Queen Charlotte’s and Chelsea Maternity Hospital, London, were approached on the wards on day 3 postpartum. This is a general maternity hospital and the women are typical of the population of West London. They were in general unselected, except that only those who had had a vaginal delivery were included; mothers of babies with medical problems or who were admitted to the special care baby unit, and those with multiple births were also excluded. One hundred and sixty two consented to participate in the study as approved by our institutional ethical committee. Mothers filled in self-rating questionnaires at 3 days and then by post at 12 weeks postpartum. At 3 days these were the Kennerley Blues 5Scale, (Kennerley and Gath 1989), the Edinburgh Postnatal Depression Scale (EPDS) (Cox et al. 1987), the Highs Scale, for the detection of mild sub clinical hypomania (Glover et al 1994), and the new Mother-to-Infant bonding scale. Both the Blues and Highs scales are well validated and ask how the subject is feeling today and include words such as “tearful” , “anxious” “over-emotional” for the Blues and “elated” “more talkative than usual”, “racing thoughts” for the Highs. The subject has to tick “Yes a lot”, “Yes a little” or “No” for each question. Elevated scores indicate more severe Blues or Highs. The Mother-to-Infant Bonding Scale and its scoring is shown in the Appendix. A high score indicates worse mother-to-infant bonding. All these scales are very simple, do not require a high reading level, and each takes only a few minutes to complete.At 12 weeks the same women were sent an EPDS and two Bonding questionnaires by post. For the latter, they were asked to recall how they had felt towards their baby “in the first few weeks”, and how they felt “now”. One hundred and forty four women returned fully completed questionnaires. The mean (SD) age of the mothers was 31.9 (± 4.6) years; the babies mean (SD) weight was 3.382 (± 0.527) kg. Sixty nine% of the mothers were primiparous and the babies were 51% male and 49% female.
STATISTICAL ANALYSIS:
The main outcome measure was the bonding score of the mother with her new child (Mother-to-Infant Bonding Scale). Data was collected using self-report measures administered at two time points. Blues score, EPDS score, Highs score and Mother-to-Infant Bonding scores were obtained prospectively at day 3 postpartum. EPDS and Mother-to-Infant Bonding were obtained prospectively at 12 weeks postpartum at which time participants also provided a retrospective report of Mother-to-Infant Bonding 6for the first few weeks postpartum. Zero was a possible response to the Blues, Highs, Mother-to-Infant Bonding and EPDS questionnaires. The distribution of EPDS scores was approximately normal with positive skew. All other continuous variables had non-normal, positively skewed distributions.The internal reliability of the new bonding instrument was examined using bivariate inter-item correlations (Pearson’s product moment correlation), Cronbach’s alpha, and principal components analysis. Spearman’s rank correlation was used to test the stability of bonding scores across the study. Scores on the bonding instrument approximated a Poisson distribution so relationships between bonding data and the results from the instruments rating maternal affective state (EPDS, Highs and Blues scales) were examined using Poisson regression calculated using Stata version 7 (Stata Corp., Texas).
RESULTS:
An initial reliability analysis performed on the nine-item mother-infant bonding scale gave an internal reliability of 0.66. Examination of inter-item correlations suggested the possibility of a multi-dimensional structure so principal components analysis was used to examine this. Two components emerged: items 1, 2, 3, 5, 6, 8 and 9 in the first and items 4 (possessive) and 7 (protective) in the second. During the study some participants had complained that item 4 (possessive) was difficult to interpret as positive or negative. This item weighted poorly on the first component and was therefore discarded. Although initially placed in the second component (r = .59), item 7 also weighted well on the first component (r = .52) so this was retained. The remaining eight items had an alpha score of 0.71, showing 7reasonable internal reliability. All mother-to-infant bonding scores were recalculated using this 8-item scale.The data was examined for differences related to the sex of the infant. Poisson regression indicated sex was not a significant predictor of any of the constructs measured. Therefore in this study sex of infant was not included in further analyses.Table 1 shows the median and ranges for the Bonding, the EPDS, Highs and Blues scores at the different time points, 3 days, first few weeks, and 12 weeks. The median Bonding score was 1 at each time, with a range of 0-13, 0-16 and 0-6 respectively; 18.5%, 22% and 8.9% scored 4 or more at each time point respectively. Figure 1 shows the distribution of bonding scores for day 3, the first few weeks, and at 12 weeks postpartum; Poisson regression was used to compare the distributions. Results showed average bonding scores at 12 weeks were 67% (95%CI 55% to 81%) lower than those recorded at 3 days, indicating mother-infant bonding scores had improved considerably between these two occasions. Table 2 shows the correlations between the mother-infant bonding scores at the three different time points, calculated using Spearman’s ranked correlation. Bonding scores at 3 days correlated significantly with those in the first few weeks and with those at week 12. There was also a significant correlation between the bonding scores in the first few weeks and at week 12, demonstrating the stability of the bonding instrument results over the period of the study. Spearman’s correlation was also used for an initial assessment of the relationship between bonding scores and the mother’s mood measured by the 8EPDS, Blues and Highs scales at each time-point of the study. The results are given in Table 3. In general there was a trend to positive correlation with the EPDS and the Blues and a negative one with the Highs. Thus negative affect correlated with poorer bonding and positive affect correlated with better bonding. The strongest correlations were between bonding in the first few weeks and EPDS day 3 and EPDS week 12. We also tested whether those scoring 13 or over on the EPDS had worse bonding scores in the first few weeks. Figure 2 shows that there was a significant difference in bonding scores in the first few weeks between those with high and low EPDS at 3 days (p = .028 two tailed Mann Whitney U test). A similar analysis comparing those scoring 13 or above on the EPDS, and those scoring below 13 at 12 weeks, and bonding in the first few weeks just failed to reach significance (p = .058).Further assessment of the relationships between bonding and concurrent measures of maternal affect were provided by Poisson regression (Table 4). At day 3, bonding was significantly predicted by concurrent measures of EPDS, Blues and Highs, and at week 12, bonding was significantly predicted by the concurrent EPDS score. Bonding in the first few weeks was predicted by EPDS and Blues at 3 days and as also related to the EPDS score at 12 weeks.
DISCUSSION:
The response to this prospective study was very good; of 162 women recruited 161 replied and 144 returned fully completed questionnaires for all items at 12 weeks. The questionnaire was quite acceptable to the mothers. The questions on the Mother-to-Infant Bonding Scale were re-scored using the 8 items shown to have good internal reliability. The subjects were unselected women from a maternity hospital, and the results suggest that even among such a population there is quite a widespread range of feeling towards the baby. However the distribution of the scores was very skewed (Fig1) suggesting that the questionnaire may be best for detecting the more abnormal responses, rather than subtle differences within the normal population. It may be noted that it is possible to score quite highly on the scale without the subject endorsing anything very negative. It is designed to detect a failure to bond, or a lack of feeling, although it also has questions regarding dislike, resentment and aggression.This study was designed and carried out in a prospective manner, with one exception, the recollection of the maternal bonding in the first few weeks. It was this score that showed the clearest simple relationship with the EPDS (Table 3) and Fig 2. There was a significant correlation between the Blues and EPDS scores at day 3 (rs = .69 p < .001), but they measure overlapping but different symptoms. It is of interest that it was the EPDS, which showed the greatest correlation with bonding (Table 3). This suggests that is depression rather than the lability of the Blues that is linked with worse bonding.It should be noted that in the study of Robson and Kumar (1980), it was striking that the maternal recollection of their initial feelings towards the baby, 10even after one year, were very close to those that had been expressed at the time (kappa = .78). It seems that the immediate postpartum period, and early feelings towards the baby, are particularly well remembered by the mother. However, it may also be easier for the mother to admit to an earlier detachment or resentment towards her newborn baby, at a later date. It may be that the new questionnaire should be used to assess bonding, as well as immediately after birth, also over a brief retrospective period also.The results shown here indicate that the maternal bonding developed progressively over the first 12 weeks postpartum. It has been shown by Matthey et al (2000) that adjustment to parenthood is related to the experiences the mother and father had as children and their own personality traits. In the study presented here 69% of the births were primiparous so the general trend towards better bonding over the early period of 12 weeks may be linked with an adjustment to the status of parenthood. The Mother-to-Infant Bonding Scale was designed for use from day one postpartum, offering the mother one-word descriptors of possible emotions towards her new child. It is quick and easy to use. It is of interest that women who had raised scores on the Highs, indicating mild sub clinical hypomania, actually had significantly better bonding than others. Hipwell et al (2000) have shown that a diagnosis of full mania in the postpartum period was associated with secure attachment whereas severe or psychotic depression was related to insecurity. It is possible that there is a biochemical basis for these findings and related to the hormonal profile of the woman at the time. It is known that cortisol is raised in the Blues and lowered in the Highs (Taylor et al 1994). It is also known that there is an inverse relationship between cortisol and the bonding hormone oxytocin (Carter 2003). Thus women with the Blues may 11have low oxytocin and those with the Highs have raised oxytocin. Further research is needed to examine this. This questionnaire is suitable for use with the general population, and is substantially different from the few other questionnaires that have been designed to study the feelings of a mother towards her young baby. It differs from the “Clinical rating scale to assess mother-infant interaction” devised by Kumar and Hipwell (1996) which is for use by nurses in psychiatric mother and baby units. An early questionnaire in this area was a 3 item scale devised by Fleming et al 1988 which included “thinking of baby makes me feel good” and “talk a lot about baby”; this found no relationship with maternal depression. The Japanese attachment scale devised by Nagata et al (2000) was based on words and phrases used by mothers with infants in neonatal intensive care. It has three dimensions, acceptance of child, anxiety regarding children and involvement with children. They did find that mothers with higher maternity blues scores had worse attachment. In conclusion, the mother-to-infant bonding scale reported here is novel, simple to use and shows links with early mood. It may aid future research into the causes of a mother’s failure to bond with her child and the results of such a failure.
READ FULL TEXT: DOWNLOAD