Mothers’ Expectations of Parenthood: The Impact of Prenatal Expectations on Self-Esteem, Depression, Anxiety, and Stress Post Birth

 

Kathryn Lazarus and Pieter Rossouw

The University of Queensland

bigstock-Mother-And-Baby-med

IJNPT_Vol3_issue2pp102-cover

 

 Download article from the international journal of neuropsychotherapy

Abstract

Pregnant women often connect the transition to motherhood with optimistic expectations, yet studies have shown that postnatal experiences that do not match or exceed prenatal expectations can have a large impact on the development of postnatal depression. The question whether prenatal expectations can influence postnatal emotional states and self-esteem was investigated via an online survey completed by 176 women. This survey explored expectation and experience scores regarding the infant, level of support, and sense of self as a mother (of one, two, or three babies) as predictors of depression, anxiety, stress, and self-esteem. As hypothesized, prenatal expectations that surpassed actual experience were predictive of lower levels of self-esteem and higher levels of depression, anxiety, and stress. Furthermore, self-expectation and experience scores significantly mediated the relationship between both infant and support expectation and experience scores, and each of the outcome variables. These findings highlight the value of educating women about the role of societal and self-expectations prior to having a child, and how these expectations can influence levels of self-esteem, depression, anxiety, and stress.

Keywords: postnatal depression, expectations, anxiety, stress, self-esteem


Postnatal depression (PND) is a mental health issue that affects a significant number of new mothers (Ogrodniczuk, 2004); however, as estimates vary, the precise figure remains up for debate. The Australian Institute of Health and Welfare (2012) found that approximately 10% of mothers were diagnosed with depression during the period from birth to the child’s first birthday, while the DSM-5 notes that between 3% and 6% of mothers will experience a major depressive episode during pregnancy or in the weeks or months post birth (American Psychiatric Association, 2013). Other estimates suggest that between 10% and 20% of new mothers are impacted by antenatal and/or postnatal depression (Goodman, 2009; Whitton, Warner, & Appleby, 1996).

PND can be defined as a form of clinical depression that occurs following the birth of a child, with symptoms generally arising any time during the first year (National Health and Medical Research Council, 2000; Women’s Health Queensland Wide, 2009). Symptoms of PND can be many and varied. Some women may have persistent feelings of sadness, anxiety, and irritability accompanied by rapid mood swings, difficulty sleeping, changes in eating habits, lack of energy, and loss of sexual interest. They may also experience panic attacks, migraines, and headaches, have difficulty concentrating and difficulty looking after self and baby, and experience overwhelming exhaustion and suicidal thoughts (Buultjens & Liamputtong, 2007; Mauthner, 2002; Women’s Health Queensland Wide, 2009). In the DSM-5, PND is not recognized as a separate diagnostic entity; instead it is contained within the category of Major Depressive Disorder under the specifier “with Peripartum Onset”. This specifier can be applied to an episode of major depression if mood symptoms occur in the antenatal period or within the first 4 weeks after birth (American Psychiatric Association, 2013).

The Influence of Prenatal Expectations on the Transition to Motherhood

The impending birth of a child is often a time in a woman’s life that is anticipated with excitement and enthusiasm (Delmore-Ko, Pancer, Hunsberger, & Pratt, 2000) along with expectations of happiness, fulfilment, and contentment (Nicolson, 1999). Many studies have shown that prenatal expectations play an important role in the transition to becoming a mother (Lawrence, Nylen, & Cobb, 2007; Read, Crockett, & Mason, 2012). Such expectations allow individuals to be guided by a sense of understanding and control over events in their lives, but when these expectations are inaccurate they can lead to significant difficulties in adapting to actual situations (Lawrence et al., 2007). Disconfirmed expectations have often been associated with a more difficult adjustment to parenthood and development of emotional problems (Harwood, McLean, & Durkin, 2007; National Health and Medical Research Council, 2000; Read et al., 2012). A qualitative study by Staneva and Wittkowski (2013) found that the expectations of participants were strongly influenced by myths surrounding motherhood and the belief that they were capable of being a superwoman. On reflection, the participants realized that their expectations were often not met, which led them to feel unprepared and overwhelmed in their new role as a mother. In another study, Choi, Henshaw, Baker, and Tree (2005) found that women’s expectations of parenthood were powerfully influenced by popular cultural images of happy families, and that upon giving birth they were faced with the realization that being a mother was indeed very different to what they had expected. This discrepancy between myth and reality (expectations and experience), especially in the absence of alternative motherhood discourses, can leave a mother feeling unable to make sense of her experience, leading to feelings of desperation, a sense of being trapped, and extreme guilt (Choi, Henshaw, Baker, & Tree, 2005).

Quantitative studies have also revealed the importance of expectations in the transition to motherhood. For example, Harwood, McLean, and Durkin (2007) found that mothers’ negative parenting experiences as compared to their expectations were significantly associated with poorer adjustment to parenthood. This study also investigated women’s expectations about caring for their infant, partner expectations, their own wellbeing (self-expectations) and their relationship with others (social expectations), and how these expectations influenced postnatal mood and relationship adjustment. The findings were complex: On the one hand, no significant differences between partner and social expectations and actual experience were found; however, infant and self-expectations were found to be significantly more positive than actual experience. Interestingly, the study also showed that 64.8% of women’s experiences were more positive than expected, contra the hypothesis that overall experiences of parenthood would not be as positive as expectations. Concerning depression, prenatal mood and postnatal self-efficacy beliefs were analysed against the dependent variable of postnatal mood, using the predictors of expectations and actual experiences (calculated using an overall composite score comprising infant, partner, self-, and social change scores). It was found that when actual experience exceeded prenatal expectations, there was a decrease in depression scores. Additionally, no relationship was found between self-efficacy beliefs and depression; however, a moderator effect between self-efficacy and expectation scores was found (again contra the authors’ predictions) indicating no relationship between mood and disconfirmed expectations for mothers with lower self-efficacy, whilst a significant relationship was found between expectations that exceeded experience and higher levels of depression for women with higher self-efficacy scores.

More recently, Flykt et al. (2009) examined prenatal expectations regarding the infant–parent relationship in the first year, with stress as the outcome variable. The findings of this study revealed that when parents’ infant relationship expectations were more positive that actual experience, they were more likely to experience higher parenting stress. Similar to Harwood and colleagues (2007), Flykt et al. also found that postnatal experience scores were in general either greater or equal to expectations.

In a later study, Flykt et al. (2014) examined whether parents’ mental health and marital relationship, delivery experience (birth experience, unplanned Caesarean, analgesia), and infant issues (health problems, temperament, parent worry) predicted violated expectations of the infant-parent relationship. Couples filled out an expectation questionnaire prior to the birth and a measure of the above predictors two months post birth. The results showed that, for mothers, associations between infant, marriage, and delivery difficulties and violated expectations were mediated by mental health problems (depression, anxiety, sleeping difficulties and social dysfunction based on participant state over the previous few weeks).

A review of these studies shows marked differences in variables and measures and highlights the complexity involved in studying the constructs of expectations and the transition to motherhood. At this stage there appears little consensus as to which factors may be the most salient to researching this topic; however, by taking a broad view of the variables most often investigated in parenthood transition research, it appears that expectations—with respect to the infant, the levels of support around the mother, the mother’s view of herself, and her experience—may play a critical role in predicting the development of PND.

The Social Expectations of Motherhood

As already mentioned, the expectations surrounding motherhood for most pregnant women are overwhelmingly positive. Western society’s socially constructed ideology of motherhood is one of fulfilment, contentment, and excitement, with an accompanying increase in overall happiness (Buultjens & Liamputtong, 2007; Mauthner, 2002). These cultural representations of women effortlessly transitioning to motherhood serve to reinforce the dominant myth that women are natural mothers who possess an innate ability to immediately bond with their baby and quickly become a selfless and caring nurturer (Choi et al., 2005). The ideology that mothering is instinctive and effortless was evident in a qualitative study by Read and colleagues (2012) who found that the female participants had a belief prior to having their baby that they would be able to draw on an innate coping ability as a mother, and this belief led them in turn to expect that they would easily cope with the demands of motherhood.

Interestingly, an exploration of motherhood and social ideals by Mauthner (2002) found that women often go to great lengths to protect the cultural ideology of motherhood by hiding their vulnerabilities and difficulties from other mothers. Staneva and Wittkowski (2013) found this in their study of maternal expectations, where two women expressed the feeling that other mothers had “kept secret” the reality of motherhood, with one stating: “I wish friends had warned me. It seems like there is this thing that women feel uncomfortable discussing that a child could be burdensome” (p. 263). Similarly, Choi and colleagues (2005) found that women who perceived themselves as failing to meet such ideological criteria, rather than talking to others about their difficulties, chose instead to work harder at appearing to be the ideal mother. The very strong taboo against mothers admitting to having difficulties leads women to mask their true feelings by feigning an appearance of strength and coping. This often unconscious acceptance of these socially constructed ideals can result in intense feelings of loneliness, isolation, and withdrawal from the outside world (Mauthner, 2002).

The ideology of motherhood is what society, women included, use to measure who is a “good” mother and who is a “bad” mother (Choi et al., 2005, p. 168). When a new mother fails to achieve such ideals and instead experiences anxiety and/or depression after birth, society condemns her—giving rise to feelings of guilt, shame, inadequacy, and worthlessness, along with a profound sense of failure (Mauthner, 2002). Mauthner (2002) argues that depression after childbirth is associated with the way in which mothers perceive and internalize such cultural expectations of motherhood.

Grief and Loss Associated With the Birth of a Child

Loss associated with successful childbirth is not generally acknowledged: It is counterintuitive to associate loss with the creation of new life and the optimism linked to motherhood (Nicolson, 1999). Nevertheless, for many women, there may indeed be an intense sense of grief and loss associated with the birth of a child, as Nicolson (1999) found. Subthemes of loss that emerged from Nicolson’s study included: loss of autonomy and time, loss of appearance, loss of femininity and sexuality, and loss of occupational identity. All transitions—be it death or in this case birth—involve significant change (Hunter, 2007). Having a baby brings about intense change, encompassing a succession of complex losses together with the gain and responsibility of a baby and new identity as a mother (Nicolson, 1999). Such acute change in a woman’s life may also contribute to the violation of positive motherhood expectations, resulting in a perhaps damaged ideal of what motherhood actually means.

For many women, motherhood is a central feature of female identity that is incorporated into a mental representation from a very early age (Nicolson, 1999). From an attachment theory perspective, Bowlby introduced the concept of inner working models (internal mental representations) of attachment relationships (Bowlby, 1980) in order to explain why physical proximity is not necessary for adolescents and adults to feel securely attached (Shear & Shair, 2005). This internal representation actually provides a connection to someone to whom the individual feels a close relationship even when they are not together. As there are many types of attachments in people’s lives, inner working models can fall into two categories: specific and generalized forms of attachment (Shear & Shair, 2005). Specific inner working models are linked to individual attachment figures, whilst generalized inner working models are permanent templates regarding expectations of close relationships (Shear & Shair, 2005). In the case of a generalized inner working model of motherhood, it could be suggested that women form an early mental representation of themselves as a mother (expectation) along with the type of close relationship they anticipate will be developed with their unborn infant. In the context of grief and loss, if, after the baby is born, the reality of the situation does not conform to the inner working model, the working model is unable to immediately incorporate or understand such severe change, similar to the death of a loved one. This inconsistency between the newly perceived reality and the inner working model leads to significant distress for the mother because her original inner working model of motherhood is now rendered inoperable. Once an inner working model has been established, it can be modified, but only slowly, through new repetitive patterns of behavior formed within the ongoing relationship between mother, baby, and the environment (Shear & Shair, 2005). But unlike the disruption of the inner working model following the death of a loved one, new mothers are firmly prevented from grieving their losses due not only to the socially imposed restrictions surrounding the ideology of motherhood but also the unconscious acceptance of these restrictions. Mothers are unable to grieve their losses, and, if they do, they are often pathologized for it (Nicolson, 1999).

The Impact of Anxiety, Stress, Depression, and Low Self-Esteem

Persistent maternal anxiety, stress, and depression can have long term debilitating consequences for all family members involved, both during pregnancy and in the postnatal period (Parent–Infant Research Institute, 2010). For women, severe depression and/or anxiety during pregnancy is linked to increased obstetric complications, low birth weight, stillbirth, postnatal specialist care, and suicide attempts (National Collaborating Centre for Mental Health, 2007). In the fifth report of the Confidential Enquiries into Maternal Deaths, Why Mothers Die 1997–1999 (2001), suicide was found to be the leading cause of maternal deaths. A follow-up report, Why Mothers Die 2000–2002 (Confidential Enquiry into Maternal and Child Health, 2004), again found suicide to be the leading cause of maternal death, with a significant number caused by psychiatric illness. This enquiry also found that out of 155 indirect deaths (i.e., indirectly related to pregnancy), four cases involved infanticide, and two of these involved the mother additionally taking the life of an older child (CEMACH, 2004). Suicidal thoughts and attempts may also be prevalent after the birth of a child. As reported by Mauthner (2002, p. 18), one woman experienced serious thoughts of suicide to the point of wanting “to just lie on the motorway … under this lorry … and I just felt really calm about it. . . . I just wanted to lie there and everything will be over.” Such statistics and qualitative data illuminate the potentially very serious consequences that high levels of anxiety, stress, and depression can have on a mother (and her children), and the importance of researching postnatal emotional states. A study by Delmore-Ko and colleagues (2000) revealed that women’s levels of stress naturally increase during the transition to motherhood, but that women who had realistic expectations and a sense of mastery over their situation were able to cope with stressors more effectively, were less susceptible to depression, and felt more capable in the early stages of motherhood. Notably, however, the authors of this study also found that significant numbers of women experienced worry and stress in the later stages of their pregnancy, which led to them having a sense of incompetence regarding their ability to be an effective mother.

With regard to children, it has been reported that infants of depressed, anxious, and stressed mothers experience significant cognitive, emotional, and behavioral deficits (such as attentional, language, and social difficulties) that may continue into later childhood (Parent–Infant Research Institute, 2009). Postnatal depression, anxiety, and stress may also play a large role in compromising the mother’s ability to nurture and bond with her infant, with intense feelings of guilt and worthlessness, as well as fatigue and low mood, significantly contributing to this issue (Parent–Infant Research Institute, 2010). Like the children of depressed mothers, fathers are also more likely to develop depression, with a study showing that maternal depression was the strongest predictor of paternal depression within the first year postpartum (Goodman, 2004).

Another characteristic of depression is low self-esteem, which has been viewed as a significant causative predictor of PND (Franck & De Raedt, 2007; Wylie et al., 2011). For example, Ogrodniczuk (2004) found that when pregnant women with low self-esteem and their partners (the empathy group) were provided with additional information on motherhood, they felt more competent as mothers and experienced significantly lower levels of distress and depression as compared to mothers with low self-esteem in the control group. All groups (empathy, baby play, and controls) were provided six weekly information sessions on breastfeeding, pregnancy, and birth, and brief information on PND. The empathy group also received information on psychosocial issues related to becoming new parents, such as strategies for dealing with being a new parent and assistance with feelings of stress, isolation, and lack of confidence (Ogrodniczuk, 2004). In addition, a qualitative study by Buultjens and Liamputtong (2007) found that many of the participants said they had experienced a lack of self-esteem and their confidence had fallen since becoming a mother. In contrast, Delmore-Ko et al. (2000) found that when women felt prepared as mothers, they experienced lower levels of perceived stress and a higher sense of self-esteem. This evidence highlights the importance of self-esteem levels for new mothers, and that confidence in their own ability and level of competence may be an essential component of a smooth transition to motherhood.

The Study

The present study was designed to investigate the discrepancy between a mother’s expectations regarding her infant, levels of support, and sense of self prior to having a baby, and her actual experiences post birth, along with the impact these discrepancies have on her self-esteem, and how this relates to the development of depression, stress, and anxiety in the postnatal period from birth to 6 months. To the authors’ knowledge, this is the first study to explore not just women’s experiences of becoming a first-time mother but their experiences with second and third children as well, and with an unrestricted time-frame between births. As most previous studies have focused on self-report questionnaires for first-time mothers, generally undertaken within the first 12 to 18 months after the birth of the baby (Delmore-Ko et al., 2000; Flykt et al., 2009; Flykt et al., 2014; Harwood et al., 2007; Lawrence et al., 2007; Pancer, Pratt, Hunsberger, & Gallant, 2000; Sayil, Güre, & Uçanok, 2006), it was hypothesized that the lengthier retrospective nature of this study could potentially allow mothers to feel less restricted by the constraints of the ideology of motherhood. Lewis and Nicolson (1998, cited in Choi et al., 2005) stated that there may be value in investigating the difficulties experienced in motherhood sometime after the birth, thereby allowing women to be reflexive in their responses. In addition, it may be expected that responses from mothers who are no longer in the throes of caring for a new infant can be more reflective of their experience by means of hindsight; and, furthermore, that the anonymous nature of the study may compel women to reveal a level of honesty they may not have felt comfortable revealing to others in the past.

It was hypothesized that:

  1. When a mother’s expectations with regard to her infant are significantly compromised, she will be more likely to exhibit low levels of self-esteem, along with higher levels of depression, anxiety, and stress.
  2. When a mother’s expectations with regard to her support networks are significantly compromised, she will be more likely to exhibit low levels of self-esteem, along with higher levels of depression, anxiety, and stress.
  3. When a mother’s expectations with regard to her personal sense of self are significantly compromised, she will be more likely to exhibit low levels of self-esteem, along with higher levels of depression, anxiety, and stress.
  4. Women’s expectations of motherhood will exceed their actual experience.

Method

Participants

The participants consisted of 257 adult women who had given birth to at least one child. A total of 176 respondents remained after those with missing data were removed. The majority of the sample were married (69.9%), whilst 16.5% were currently living in a de facto relationship. The age range for the majority of respondents was 31–40 years (46.6%), with 61.3% of participants having completed university-level education. The participants voluntarily completed an online questionnaire in response to notices placed on Facebook, via text messages, and in emails.

Procedure

Participants were given a link to a website comprising the questionnaire. Each participant was initially provided with an information sheet with details such as the purpose of the study and the risks involved, confidentiality, and security. Ethical clearance was applied for and granted via the University of Queensland Human Research Ethics Committee (Ethical Clearance Number: 14-PSYCH-MAP-11-TS).

Measures

To ensure content validity for each measure, an empirical literature review on expectations and the transition to motherhood was conducted. An existing scale, the Parenting Expectations Measure (Harwood, 2004) was found; however, as participants would be required to complete a questionnaire for each child, this 55-item measure was considered possibly too onerous for respondents to fill out several times. In order to maximize return rate, therefore, it was decided to keep survey items to a minimum, and so selected items from the Parenting Expectations Measure were used together with items from Bates, Freeland, and Lounsbury’s (1979) Infant Characteristics Questionnaire-6 Month Form, as these related specifically to infant, support, and self-expectations. Additional items were written by the first author, based on her personal experience and feedback from an expert panel of women with extensive experience as mothers. Each of the measures described below were administered with respondents taking a retrospective view of their expectations prior to the birth of each baby, and their actual experience up to 6 months post birth.

Mothers’ Expectations and Experience Measure. This measure contains six subscales: Infant Expectations and Experience; Support Expectations and Experience; and Self Expectations and Experience, and was used to assess expectations and experience for up to 3 children. The Infant subscale assesses the mother’s expectations about her infant (e.g., “Prior to having your first baby, to what extent did you expect that your baby would sleep for long periods?”) and experience in relation to caring for her infant (e.g., “After having your first baby, to what extent did baby sleep for long periods?”). The Support Expectations and Experience subscale measured mothers’ expectations (e.g., “Prior to having your first baby, to what extent did you expect that your family and friends would help out after the baby was born?”) and experience relating to the amount of perceived support the mother received (e.g., “After having your first baby, to what extent did your family and friends help out after the baby was born?”). The Self Expectations and Experience subscale pertained to respondents’ expectations (e.g., “Prior to having your first baby, to what extent did you expect that you would be a competent mother?”) and actual experience relating to her physical and psychological wellbeing (e.g., “After having your first baby, to what extent did you feel that you were a competent mother?”). All items were answered on a 4-point Likert scale ranging from 1 (not at all) to 4 (completely). Higher scores indicated high expectations as well as greater endorsement of actual experience. Internal consistency scores for the Expectations subscales for up to three children were: Infant Expectations: first baby α = .70, second baby α = .75, and third baby α = .81; Support Expectations: first baby α = .66, second baby α = .69, and third baby α = .74; and Self Expectations: first baby α = .66, second baby α = .67, and third baby α = .58. Internal consistency scores for the Experience subscales for up to three children were: Infant Experience: first baby α = .73, second baby α = .78, and third baby α = .73; Support Experience: first baby α = .75, second baby α = .79, and third baby α = .70; and Self Experience: first baby α = .81, second baby α = .80, and third baby α = .85.

Depression Anxiety Stress Scale (DASS). The 21-item version of the DASS (Lovibond & Lovibond, 1995) assessed respondents’ experience of depression, anxiety, and stress during the first 6 months after the birth of each baby. Items include three 7-item subscales which are answered on a 4-point Likert scale ranging from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). The Depression scale measures low positive affect, low self-esteem, and hopelessness, whilst the Anxiety scale evaluates physiological hyperarousal, autonomic arousal, and subjective feelings of fear. The Stress scale assesses agitation, tension, and negative affect (Gloster et al., 2008). Studies have shown that the DASS-21 demonstrates excellent internal consistency. For example, Henry and Crawford (2005) tested the construct validity of the 21-item measure, and found Cronbach’s alphas of: Depression α = .82; Anxiety α = .90; and Stress α = .93. Internal reliability in the current study for the first baby is: Depression α = .93; Anxiety α = .79; and Stress α = .89. In each scale, an aggregate across each 7-item subscale was created.

Rosenberg Self-Esteem Scale. This 10-item scale was used to assess respondents’ level of self-esteem during the first 6 months after the birth of each baby. The scale is a measure of global self-esteem, which was defined by the developer (Rosenberg, 1965) as an overall sense of worth as an individual (Sinclair et al., 2010). All items are measured on a 4-point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree). Studies have shown that the Rosenberg Self-Esteem Scale demonstrates excellent internal consistency. For example, Sinclair et al. (2010) tested the construct validity of the 21-item measure, and found Cronbach’s alphas of α = .91. To score this scale, an aggregate across all 10 items was created, with an obtained internal consistency score in the current study for the first baby of α = .94.

Sociodemographic factors. Age, marital status, education level, diagnosis of PND, and whether the mother personally felt she had PND were also surveyed.

Results

Prior to undertaking the main analyses, each variable of interest was examined through the SPSS 22.0 program for accuracy of data entry, missing values, applicable assumptions, and outliers.

Table 1 presents the descriptive statistics for each nominal variable, and Table 2 outlines the descriptive statistics and t-test results for the continuous variables. Prenatal and postnatal means for the expectation and experience variables were compared using t-tests (Table 2). It was hypothesized that mothers’ expectations would exceed their postnatal experiences. Scores were coded so that high expectation and experience scores reflected high expectations and positive experiences respectively. Regarding Baby 1, no significant differences between infant expectation and experience scores were found; however, both support and self-expectation scores were significantly higher than experience scores. For Baby 2, infant expectation scores were found to be significantly lower than experience scores, but no differences between support and self-expectation scores were found; and for Baby 3, no significant differences were found between infant, support, self-expectation or experience scores.

The outcome variables of depression, anxiety, stress, and self-esteem were also compared using t-tests (Table 2). Depression, anxiety, stress, and self-esteem scores for Baby 1 were compared to those scores for Baby 2; Baby 2 outcome scores were compared with Baby 3; and Baby 1 scores were compared with Baby 3 scores. There were no significant differences found between the outcome scores for Baby 1 and 2, Baby 2 and 3, or Baby 1 and 3.

Table 3 presents the zero-order Pearson’s correlations between all variables of interest. For Baby 1, all variables are significantly associated with each other. The predictor variables are positively related to each other, along with self-esteem, whilst the outcome variables of depression, anxiety, and stress are negatively correlated with the predictors and self-esteem. Regarding Baby 2, all variables are associated, except support expectations and anxiety, and stress. Concerning Baby 3, effect sizes are smaller; however, this is to be expected due to the small sample size (n = 21). Hierarchical multiple regression analyses were used to test the hypotheses that compromised expectations regarding infant, support, and self-experiences would predict higher levels of depression, anxiety, and stress, as well as low levels of self-esteem. In order to investigate these hypotheses, analysis of the discrepancy score between expectations and experience was necessary. According to Harwood (2004) and Cohen and Cohen (1983, cited in Harwood et al., 2007), discrepancy scores attained by subtracting expectation scores from experience scores are problematic because it is not possible to determine whether the variance is coming from the expectation score, the experience score, or a combination of both. As such, it was recommended that standardized residual scores be created using a regression analysis with the expectation score as the predictor variable and experience score as the outcome variable (Harwood et al., 2007). The resulting infant, support, and self-residualized discrepancy scores were then used as the predictor variables in subsequent analyses. Positive residualized discrepancy scores indicated that expectation scores were lower than experience scores, and negative residualized discrepancy scores denoted that expectations exceeded experience, resulting in compromised expectations.

The first hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting depression (Table 4). Infant and support residualized discrepancy scores were entered first into the model, and both were significantly negatively associated with depression (β = −.30, p < .001 and β = −.37, p < .001 respectively). After adding the residualized self-discrepancy score to the model in Block 2, the direct relationship between infant and depression became insignificant (β = .06, p = ns), indicating a full mediation, whilst the association between support and depression was largely reduced but remained significant (β = −.19, p < .01), implying partial mediation. The results shown in Table 4 indicate that the model was significant, F(3, 172) = 57.00, p < .001, and accounted for 50% (R2=.499) of the variance of depression scores. To test the significance of the mediations, several Sobel tests were undertaken. These results showed that self-residualized discrepancy scores significantly mediated the relationship between infant residualized discrepancy scores (z = −7.55, p < .001), support residualized discrepancy scores (z = −4.43, p < .001), and depression. See Figure 1.

A second hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting anxiety (Table 5). Infant and support residualized discrepancy scores were entered first, and both were significantly negatively associated with anxiety (β = −.23, p < .01 and β = −.22, p < .01 respectively). The self-residualized discrepancy scores were then added to the model in Step 2, resulting in the direct relationship between infant and anxiety and support and anxiety becoming insignificant (β = .12 and β = −.05 respectively), indicating a full mediation of both infant and support predictor variables. The results shown in Table 5 indicate that the model was significant, F(3, 172) = 28.26, p < .001, and accounted for 33% (R2=.330) of the variance of anxiety scores. Results of Sobel tests revealed that self-residualized discrepancy scores significantly mediated the relationship between infant residualized discrepancy scores and anxiety (z = −4.20, p < .001) and support residualized discrepancy scores and anxiety (z = −6.58, p < .001). See Figure 1.

The third hierarchical regression investigated the effect of compromised infant, support and self-expectations on predicting stress (Table 6). Infant and support residualized discrepancy scores were entered first into the model, and both were significantly negatively associated with stress (β = −.40, p < .001 and β = −.24, p < .001 respectively). The residualized self-discrepancy score was then added to the model, resulting in the direct relationship between infant and stress and support and stress becoming insignificant (β = .04 and β = −.06 respectively), suggesting a full mediation of both infant and support predictor variables by self-expectations. The results shown in Table 6 indicate that the model was significant, F(3, 172) = 57.680, p < .001, and accounted for 50.2% (R2=.502) of the variance of stress scores. A Sobel test showed that self-residualized discrepancy scores significantly mediated the direct relationship between infant residualized discrepancy scores and stress (z = −.773, p < .001) and support residualized discrepancy scores and stress (z = −4.46, p < .001). See Figure 1.

A fourth hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting levels of self-esteem (Table 7). Infant and support residualized discrepancy scores were entered first, and both were significantly positively associated with self-esteem (β = .47, p < .001 and β = .30, p < .001 respectively). In Step 2, the residualized self-discrepancy score was added to the model, resulting in infant residualized discrepancy scores becoming insignificant due to the reduced beta coefficient of .013 (β = .15, p = .020). Support remained a significant predictor of self-esteem; however, the beta coefficient for support dropped (β = .15, p < .05), suggesting a partial mediation for this predictor. The results shown in Table 7 indicate that the model was significant, F(3, 172) = 72.64, p < .001, and accounted for 55.9% (R2=.559) of the variance of self-esteem scores. A Sobel test confirmed that self-residualized discrepancy scores significantly mediated the relationship between infant residualized discrepancy scores and self-esteem (z = 6.19, p < .001) and support residualized discrepancy scores and self-esteem (z = 4.09, p < .001). See Figure 1.

With regard to Baby 2, a fifth hierarchical regression investigated the effect of compromised infant, support and self-expectations on predicting depression (Table 8). In Step 1 of the model, infant and support residualized discrepancy scores were entered first. Infant expectations significantly negatively predicted levels of depression (β = −.31, p < .001); however, support expectations did not. When self-expectations were added to the model in Block 2, the relationship between infant and depression became insignificant (β = .12), evidencing full mediation. The results shown in Table 8 indicate that the model was significant, F(3, 85) = 9.23, p < .001, and accounted for 24.6% (R2=.246) of the variance of depression scores. A Sobel test confirmed that self-expectations significantly indirectly mediated the relationship between infant expectations and depression (z = −3.78, p < .001). In addition, when further exploring the mediation effect of self on support expectations and depression scores, although the initial total effect of support on depression was insignificant (Rucker, Preacher, Tormala, & Petty, 2011), self-expectation scores were found to significantly mediate the effect of support on depression (z = −2.91, p = .002). See Figure 2.

A sixth hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting anxiety for Baby 2 (Table 9). In the first block of the model neither infant nor support expectation scores predicted levels of anxiety. When self-expectations were added to the model, the relationship between infant and anxiety reduced (β = .05), suggesting possible mediation. The results shown in Table 9 indicate that the model was significant, F(3, 85) = 7.17, p < .001, and accounted for 20.2% (R2=.202) of the variance of anxiety scores. A Sobel test confirmed that self-expectations significantly mediated the relationship between infant expectations and anxiety levels (z = −3.46, p < .001). As with depression, although the initial total effect of support on anxiety was insignificant, self-expectations were found to indirectly significantly mediate the effect of support on anxiety levels (z = −2.76, p = .003). See Figure 2.

A seventh hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting stress for Baby 2 (Table 10). In the first block of the model, infant expectation scores significantly negatively predicted levels of stress (β = −.36, p < .001); however, support expectations did not predict stress levels. When self-expectations were added to the model, the relationship between infant and stress became insignificant (β = .15), signifying a full mediation. The results shown in Table 10 indicate that the model was significant, F(3, 85) = 9.85, p = .001, and accounted for 25.8% (R2=.258) of the variance of stress scores. A Sobel test confirmed that self-expectations significantly mediated the relationship between infant expectations and stress levels (z = −3.79, p < .001). As with depression and anxiety for Baby 2, support was also found to be significantly related to self-expectations by indirectly predicting stress scores (z = −2.92, p = .002); however, there was no direct relationship between support expectations and stress. See Figure 2.

An eighth hierarchical regression investigated the effect of compromised infant, support, and self-expectations with regard to predicting self-esteem levels for Baby 2 (Table 11). In Step 1 of the model, infant expectation scores significantly positively predicted levels of self-esteem (β = .38, p < .001); however, support expectations did not, due to the reduced significance level in order to reduce Type I errors. When self-expectations were added to the model, the relationship between infant and self-esteem became insignificant (β = .18), evidencing full mediation. The results shown in Table 11 indicate that the model was significant, F(3, 85) = 14.81, p < .001, and accounted for 34.3% (R2=.343) of the variance of self-esteem scores. A Sobel test confirmed that self-expectations significantly mediated the relationship between infant expectations and self-esteem levels (z = 4.11, p < .001). As with depression, anxiety, and stress, support was also found to be related to self-expectations by indirectly predicting self-esteem levels (z = −2.91, p = .002); however, there was no direct effect between support expectations and self-esteem. See Figure 2.

Pertaining to Baby 3, a ninth hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting depression levels (Table 12). In Step 1 of the model, neither infant nor support expectation scores predicted levels of depression. In Step 2 of the model, self-expectations significantly negatively predicted levels of depression (β = −.55, p < .01). See Figure 3. The results shown in Table 12 indicate that the model was significant, F(3, 25) = 6.83, p < .01, and accounted for 45% (R2=.450) of the variance of depression scores. A tenth hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting anxiety levels (Table 13). In Step 1 of the model, neither infant nor support expectation scores predicted levels of anxiety. In Step 2 of the model, self-expectations did not predict levels of anxiety due to the reduced significance level to reduce the Type I error rate. See Figure 3.

            An eleventh hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting stress levels (Table 14). In Step 1 of the model, neither infant nor support expectation scores predicted levels of stress. In Step 2 of the model, due to the reduced significance level, self-expectations did not predict levels of stress either. See Figure 3.

            A twelfth hierarchical regression investigated the effect of compromised infant, support, and self-expectations on predicting self-esteem levels (Table 15). In Step 1 of the model, neither infant nor support expectation scores predicted levels of self-esteem. In Step 2 of the model, self-expectations significantly negatively predicted levels of self-esteem (β = −.55, p < .01). The results shown in Table 15 indicate that the model was significant, F(3, 25) = 4.70, p = .01, and accounted for 36.1% (R2=.361) of the variance of self-esteem scores. See Figure 3.

Further examination of two individual items in the self-expectations and experience scale revealed some particularly important findings. A comparison between the questions concerning the first baby, “To what extent did you expect that you would be disturbed by feelings you may have toward your baby?” and “To what extent were you disturbed by feelings you had toward your baby?” revealed that a total of 2.3% of respondents expected either to “Mostly” or “Completely” experience disturbing feelings toward their first baby, but when actual experience was examined, this increased to 7.3%. A paired samples t-test revealed that this increase in disturbing feelings was significant t (175) = −2.35, p = .02 (Bonferroni correction = .025). Additionally, a comparison between the questions, “To what extent did you expect that you may have disturbing thoughts (however fleeting) that you may do something extreme to quieten/harm your baby?” and “To what extent did you have disturbing thoughts (however fleeting) that you may do something extreme to quieten/harm your baby?” revealed that nil respondents expected to either “Mostly” or “Completely” deal with thoughts of harming their first baby; however, when experience was examined, this increased to 2.8%. A paired samples t-test revealed that this increase in thoughts of harming the baby was also significant t (175) = −2.26, p = .025 (Bonferroni correction = .025).

Discussion

This study measured the difference between mothers’ expectations and experiences of motherhood and investigated the relationship between these discrepancies and levels of self-esteem, depression, anxiety, and stress. The hypothesis that compromised prenatal expectations regarding infant, support, and self would predict lower levels of self-esteem and higher levels of depression, anxiety, and stress was supported, but in an unexpected manner. In the case of both Baby 1 and Baby 2, when self-residualized discrepancy scores (representing the discrepancy between self-expectation and experience scores) were considered in the relationship between the independent variables of infant and support residualized discrepancy scores (representing the discrepancy between infant and support expectation and experience scores) and the dependent variables of depression, anxiety, stress, and self-esteem, a mediating effect was apparent, with infant and support residualized discrepancy scores either remaining significant, but with largely reduced coefficients, or becoming insignificant. Therefore, self-residualized discrepancy scores at least partially (and at times fully) explained the relationship between infant residualized discrepancy scores and support residualized discrepancy scores, and each of the outcome variables of depression, anxiety, stress, and self-esteem. These findings suggest that the way in which a mother’s expectations reflect her actual experience with her infant and/or her level of support, has a significant effect on how she feels about herself, which subsequently influences her level of self-esteem, depression, anxiety, and stress. Regarding Baby 3, there were mixed results. Infant and support residualized discrepancy scores were not found to be predictors of any of the outcome variables; however, self-residualized discrepancy scores remained a significant predictor for both depression scores and self-esteem levels. Due to the reduction in the significance level to reduce Type I errors, self-expectation and experience scores were reduced to non-significant levels for anxiety (p = .03) and stress (p = .05); however, this may be a result of the small sample size for Baby 3 (n = 21). Interestingly, even with this small sample size, there still appears to be a pattern of self-expectations and experience having a large bearing on a mother’s level of self-esteem and the potential development of depression, anxiety, and stress.

Additionally, the hypothesis that expectations would significantly exceed actual experience was supported, but only for Baby 1. In this case, expectations were significantly higher than actual experience regarding level of support and sense of self; however, no significant differences were found between infant expectations and actual experience. For Baby 2, no significant differences were found between expectations and experiences of support or sense of self, whilst a significant difference was found between infant expectations and experience, with experience exceeding expectation. This finding suggests that a mother’s expectations of support and sense of self for her second baby tend to be more realistic when compared to her expectations regarding the first child. This could be due to the mother’s expectations being more accurate the second time around as a result of their initial learning experience with their first child. Another explanation could also be that, with second children, mothers are feeling more confident within themselves and less reliant on others. This is an interesting finding, which has important implications for future prevention programs aimed at educating women—particularly, but not limited to, first-time mothers—about the risk factors of PND. Regarding the infant discrepancy scores for Baby 2, this finding is somewhat in agreement with Harwood et al. (2007) and Flykt et al. (2009) in that participants reported that their expectations matched or exceeded their experience; however, whilst Harwood et al. found highly positive responses, in this study infant scores regarding expectations and experience were moderate. As previously hypothesized, this result could possibly be due to the lengthier retrospective nature of the study, which allowed participants to be more reflective in their answers concerning the difficulties they experienced, rather than immediately opting for the most socially desirable response. Furthermore, a possible explanation for the low rate of compromised infant expectations and experiences, particularly for Baby 2, could be a consequence of the social ideology of motherhood and the strength of the unconscious restrictions in speaking negatively about baby, even when decades have passed after the birth. Future research could also explore whether the discrepancy between expectations and experience impacts on a mothers’ decision to have more than one child.

Overall, these results support the quantitative findings of Harwood et al. (2007) that when actual experience fails to live up to expectations, an increase in depression scores was found. Additionally, these results concur with Flykt et al. (2009), who also found a significant relationship between negative infant expectations and higher parenting stress. In contrast, Harwood et al. (2007) found no relationship between self-efficacy levels and depression—whereas, arguably, the most important finding of this study is that how a woman perceives herself as a mother is the key to how well or poorly she transitions into her new role. (It is also noteworthy that this difference in findings may be due to the measures of self being somewhat different.) The importance of self-efficacy expectations and experiences concerning motherhood has also been soundly established in prior qualitative studies (Choi et al., 2005; Mauthner, 2002; Staneva & Wittkowski, 2013) and is firmly corroborated by the results of this study. How a mother feels about herself appears to be a critical variable in her adaptation to becoming a mother. When a woman perceives herself as not living up to the social ideals and expectations of motherhood, she may begin to doubt whether she is a “good” mother, and may, at the same time, feel extremely uncomfortable about expressing these concerns to others, fearing that they may judge her in a negative way. This concern of women about their competence is not just restricted to those who have had children, however, as Delmore-Ko and colleagues (2000) found that pregnant women also were often anxious about their ability to mother effectively.

As this is one of the first quantitative studies to highlight the importance of expectations and experience regarding how a woman feels about herself as a mother, future research should continue to explore this variable within the context of the social ideology of motherhood. Mothers can feel overwhelmed by the messages of others, be it friends, family, or the media saying that they should not feel sadness or ambivalence toward their child, and then, if they do, being labelled a bad mother (Mauthner, 2002). Rarely is it acknowledged that being a mother is a difficult and demanding job and that, like any job, at times there can be a sense of uncertainty, disappointment, and regret about the job they are doing. Compared to other positions of responsibility, society regards motherhood very differently, and this differentiation between career-driven responsibility and motherhood responsibility can result in new mothers experiencing disenfranchised grief.

Disenfranchised grief generally refers to situations of loss that are not socially recognized or acknowledged as valid, or the loss itself is not socially supported or is stigmatized in some way (Robson & Walter, 2012; Doka, 1989, cited in Winokuer & Harris, 2012). In the case of new mothers, they are not viewed by society as valid grievers as it is counter-intuitive to associate grief with the birth of a new born baby (Nicolson, 1999). This, in conjunction with continual and pervasive messages of motherhood being the prototypical happy time in a woman’s life, appears to make it difficult for women to feel that they are allowed to acknowledge what they have lost, along with what they have gained. Furthermore, in the context of attachment theory, when a mother’s inner working model of her child does not match the actual experience itself, the inconsistency creates immense distress for the mother. This disparity was operationalized and tested via the residualized discrepancy scores in this study. The modification of this inner working model occurs slowly through adaptation to the new environment—in this case a baby and/or support system not behaving as the mother initially expected—which eventually results in a reduction of grief intensity and a re-worked and modified perception of the new reality (Simon, 2013).

An examination of individual survey items revealed some notable findings. The first of these was that 2.3% of mothers expected to experience disturbing feelings toward their baby but 7.3% of women actually experienced these disturbing feelings in the postnatal period. This is an extremely important finding for it suggests that a significant number of women feel overwhelmed to the point of experiencing feelings toward their baby that are of great concern. In a similar pairing of questions, it was found that, while 0% of participants expected to have thoughts that they may do something extreme to quieten or harm their baby, 2.8% of the respondents did have such thoughts. As in the above survey items, this is also a tremendously important finding and is evidence of the high risk of possible infanticide in almost 3% of new mothers. Both of these results signify the immense importance of normalizing the mood and emotional lability involved in becoming a new mother; further, that we as a society, rather than criticizing and anthologizing the experience of childbirth, should be having conversations about the difficulties many women experience following the birth of a child. Additionally, this study uncovered some interesting results in the statistics regarding the diagnosis of PND and whether women personally felt that they had depression. As mentioned, a diagnosis of PND ranges from 3% to 20% depending on the criteria and methodology used; and the results of this study are in agreement with the higher end of this range, with 21% of the respondents having had a PND diagnosis. When designing this study, the authors considered it important to also explore how women themselves felt about their experience of depression post birth, and they found, quite surprisingly, that a total of 39.8% of women reported feeling that they had experienced PND. This result, which is almost twice the highest reported estimate of diagnosed PND, suggests not only that PND may be significantly underreported but that current diagnostic and assessment tools may not be adequately identifying women who are suffering from PND. This is an important area for future research to continue to explore for these reasons. This finding also provides evidence for the supposition that women withdraw and conceal their true feelings when they are feeling low after having a baby.

The findings of this study in the context of past research have significant implications concerning the education of women before they have a child and highlight the importance of modifying the (currently) deeply ingrained and unrealistic conceptual framework to which society holds all new mothers to account. This perception of women as “super mums” is not merely unattainable but rather it creates an environment that promotes the development of depression in new mothers, given that women feel strongly obliged to isolate themselves and conceal their true feelings when they are experiencing difficulties and/or depression post birth. These behaviors only succeed in further feeding the symptoms of depression by avoiding the issues at hand. As such, it is strongly recommended that health professionals (such as GPs and midwives) provide early education about PND to all women preparing for motherhood—and this should not be restricted to first-time mothers. Additionally, an education program focusing on compromised infant, support, and self-expectations during the first year post birth should be created and incorporated into existing antenatal classes. Current antenatal classes focus mainly on the birthing process; however, as the results of this study reveal, it is the transition that occurs once the mother is at home that is pivotal in the development of depression, anxiety, stress, and low-self-esteem. This type of education program should perhaps emphasize the potential realities of having a child (such as a difficult and painful birthing experience or a baby with a more difficult temperament) but, most importantly, it should normalize the ambivalence and doubt that a mother may experience post birth, and stress the importance of reaching out for help and talking to others if she experiences even the smallest difficulty during the transition to becoming a new mother. Furthermore, symptoms of PND should be thoroughly explored to assist mothers and family members in their decision about when to seek professional help. This education program could also highlight how current social norms for new mothers are essentially unachievable, and that women are often condemned if they do not follow these norms after childbirth. It should also be made clear that, if a healthy shift to new motherhood is to transpire, it is the rule rather than the exception that this transition may be accompanied by some degree of grief and loss and changes in mood.

There are limitations in this study. Firstly, as is the case with the majority of studies that utilize self-report measures, it is susceptible to social desirability biases (Burn, 2008). The unexpected finding in regard to the mean infant discrepancy scores (the comparison of expectation and experience scores described in Table 2), as compared to the support and self-discrepancy scores for Baby 1 and Baby 2, may be the result of the respondents feeling obliged to choose the most socially acceptable responses regarding their children, resulting in scores that are not reflective of actual experience. Another limitation related to the instrument is the lack of restriction on the timeframe since birth. The length of time between having each child and filling out the questionnaire may have meant that respondents had difficulty remembering precisely how they felt during the 6 months after they had their babies; consequently, their responses may have been less accurate than if they had filled out the questionnaire sooner. Secondly, it is important to note that all independent variables (infant, support, and self-residualized discrepancy scores) were significantly associated, revealing small to moderate correlations. Such relationships can make it difficult to accurately partial out covariance; however, assumption checks did suggest that multicollinearity was not an issue, and none of the correlations were above .63. Furthermore, the correlational design of this study makes it difficult to be confident of cause and effect; it is therefore important to note that interpretation of such is based on theory only. Lastly, several of the expectations measures displayed lower reliability, therefore the results gleaned from these scales should be interpreted with caution. Future studies in the field would benefit from the development and psychometric testing of a valid measurement instrument of motherhood expectations and experiences (particularly regarding self-expectations and experiences) targeted at women whose last child was born at least two years prior.

In conclusion, the results of this study support the prediction that when a mother’s prenatal expectations regarding her infant, support network, and sense of self as a mother are compromised, this leads to lower levels of self-esteem and higher levels of depression, anxiety, and stress. In addition, how the mother feels about herself was found to significantly mediate both infant and support residualized discrepancy scores with respect to all the outcome variables, particularly for Baby 1 and Baby 2, which suggests that how a mother views herself is one of the salient factors in predicting postnatal emotional states. This study provides empirical support for the hypothesis that compromised prenatal expectations play a critical role in a woman’s transition to motherhood, and this effect is not restricted to first-time mothers. When considered in conjunction with previous qualitative studies, these findings highlight the importance of reflecting on society’s ideology of motherhood and the negative impact that this unrealistic vision can have on new mothers. In addition, the finding that mothers’ self-reported rates of PND are double those of reported diagnosed PND rates not only provides important evidence concerning the lack of efficacy of current assessment and diagnostic procedures but also highlights that women may feel the need to conceal symptoms of low mood when around others. This research has revealed the importance of educating women about the impact of current conceptual frameworks regarding happy and energetic new mothers (Nicolson, 1999), and how this image significantly (but unconsciously) has a heavy impact on their decision-making and behavior post birth. These findings also emphasize the importance of normalizing negative feelings associated with the transition to motherhood, along with the importance of talking about these feelings and reaching out for help if it is needed.

 

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Australian Institute of Health and Welfare. (2012). Experience of perinatal depression: Data from the 2010 Australian national infant feeding survey. (Cat. No. PHE 161). Canberra, Australia: Author.

Bates, J. E., Freeland, C., & Lounsbury, M. L. (1979). Infant characteristics questionnaire-6 month form. Retrieved from http://www.psychology.org.nz

Bowlby, J. (1980). Attachment and loss: Loss sadness and depression (Vol. III). New York, NY: Basic Books.

Burn, S. M. (2008). A situational model of sexual assault prevention through bystander intervention. Sex Roles, 60, 779–792. doi:10.1007/s11199-008-9581-5

Buultjens, M., & Liamputtong, P. (2007). When giving life starts to take the life out of you: Women’s experiences of depression after childbirth. Midwifery, 23, 77–91. doi:10.1016/j.midw.2006.04.002

Confidential Enquiries into Maternal Deaths (2001). Why mothers die 1997–1999: The fifth report of the Confidential Enquiries into Maternal Deaths. London, United Kingdom: RCOG Press.

Confidential Enquiry into Maternal and Child Health (2004). Why mothers die 2000–2002: The sixth report of the Confidential Enquiries into Maternal Deaths in the United Kingdom. London, United Kingdom: RCOG Press.

Choi, P., Henshaw, C., Baker, S., & Tree, J. (2005). Supermum, superwife, supereverything: Performing femininity in the transition to motherhood. Journal of Reproductive and Infant Psychology, 23, 167–180. doi:10.1080/02646830500129487

Delmore-Ko, P., Pancer, S. M., Hunsberger, B., & Pratt, M. (2000). Becoming a parent: The relation between prenatal expectations and postnatal experience. Journal of Family Psychology, 14, 625–640. doi:10.1037//0893-3200.14.4.625

Flykt, M., Lindblom, J., Punamäki, R. L., Poikkeus, P., Repokari, L., Unkila-Kallio, L., . . . Tulppala, M. (2009). Prenatal expectations in transition to parenthood: Former infertility and family dynamic considerations. Journal of Family Psychology, 23, 779–789. doi:10.1037/a0016468

Flykt, M., Palosaari, E., Lindblom, J., Vänskä, M., Poikkeus, P., Repokari, L., . . . Punamäki, R. L. (2014). What explains violated expectations of parent-child relationship in transition to parenthood? Journal of Family Psychology, 28, 148–159. doi:10.1037/a0036050

Franck, E., & De Raedt, R. (2007). Self-esteem reconsidered: Unstable self-esteem outperforms level of self-esteem as vulnerability marker for depression. Behaviour Research and Therapy, 45, 1531–1541. doi:10.1016/j.brat.2007.01.003

Gloster, A. T., Rhoades, H. M., Novy, D., Klotsche, J., Senior, A., Kunik, M., . . . Stanley, M. A. (2008). Psychometric properties of the Depression Anxiety and Stress Scale-21 in older primary care patients. Journal of Affective Disorders, 110, 248–259. doi:10.1016/j.jad.2008.01.023

Goodman, J. H. (2004). Paternal postpartum depression, its relationship to maternal postpartum depression, and implications for family health. Journal of Advanced Nursing, 45, 26–35. doi:10.1046/j.1365-2648.2003.02857.x

Goodman, J. H. (2009). Women’s attitudes, preferences, and perceived barriers to treatment for perinatal depression. Birth, 36, 60–69. doi:10.1111/j.1523-536X.2008.00296.x

Harwood, K. (2004). Women’s parenting expectations and their influence on adjustment to parenthood (Unpublished doctoral dissertation). University of Western Australia, Crawley, Australia.

Harwood, K., McLean, N., & Durkin, K. (2007). First-time mothers’ expectations of parenthood: What happens when optimistic expectations are not matched by later experiences? Developmental Psychology, 43, 1–12. doi:10.1037/0012-1649.43.1.1

Henry, J. D., & Crawford, J. R. (2005). The short-form version of the Depression Anxiety Stress Scales (DASS-21): Construct validity and normative data in a large non-clinical sample. The British Journal of Clinical Psychology, 44(Pt 2), 227–239.

doi:10.1348/014466505X29657

Hunter, J. (2007). Bereavement: An incomplete rite of passage. Omega, 56, 153–173. doi:10.2190/OM.56.2.b

Lawrence, E., Nylen, K., & Cobb, R. J. (2007). Prenatal expectations and marital satisfaction over the transition to parenthood. Journal of Family Psychology, 21, 155–164. doi:10.1037/0893-3200.21.2.155

Lovibond, S. H., & Lovibond, P. F. (1995) Manual for the depression anxiety stress scales (2nd ed.). Sydney, Australia: Psychology Foundation.

Mauthner, N. S. (2002). The darkest days of my life: Stories of postpartum depression. Cambridge, MA: Harvard University Press.

National Collaborating Centre for Mental Health. (2007). Antenatal and postnatal mental health: The NICE guideline on clinical management and service guidance. Leicester, United Kingdom: The British Psychological Society.

National Health and Medical Research Council. (2000). Postnatal depression: Not just the baby blues. Retrieved from https://www.nhmrc.gov.au/_files_nhmrc/publications/attachments/wh30.pdf

Nicolson, P. (1999). Loss, happiness and postpartum depression: The ultimate paradox. Canadian Psychology, 40, 162–178. doi:10.1037/h0086834

Oates, M. (2003). Suicide: The leading cause of maternal death. British Journal of Psychiatry, 183, 279–281. doi:10.1192/02-488

Ogrodniczuk, J. S. (2004). Increasing a partner’s understanding of motherhood significantly reduces postnatal distress and depression in first time mothers with low self esteem. Evidence Based Mental Health, 7, 116. doi:10.1136/ebmh.7.4.116

Pancer, S. M., Pratt, M., Hunsberger, B., & Gallant, M. (2000). Thinking ahead: Complexity of expectations and the transition to parenthood. Journal of Personality, 68(2), 253–280.

Parent–Infant Research Institute. (2009). Postnatal depression: Evidence relating to infant cognitive emotional development. Retrieved from http://www.fremantlemedicarelocal.com.au/wp-content/uploads/2012/05/Postnatal-Depression-evidence-relating-to-infant-cognitive-emotional-devt-for-health-professionals.pdf

Parent–Infant Research Institute. (2010). Perinatal depression and anxiety: Evidence relating to infant cognitive and emotional development. Retrieved from http://resources.beyondblue.org.au/prism/file?token=BL/0470

Read, D., Crockett, J., & Mason, R. (2012). It was a horrible shock: The experience of motherhood and women’s family size preferences. Women’s Studies International Forum, 35, 12–21. doi:10.1016/j.wsif.2011.10.001

Robson, P., & Walter, T. (2012). Hierarchies of loss: A critique of disenfranchised grief. Omega, 66, 97–119. doi:10.2190/OM.66.2.a

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Rucker, D. D., Preacher, K. J., Tormala, Z. L., & Petty, R. E. (2011). Mediation analysis in social psychology: Current practices and new recommendations. Social and Personality Psychology Compass, 5, 359–371. doi:10.1111/j.17519004.2011.00355.x

Sayil, M., Güre, A., & Uçanok, Z. (2006). First time mothers’ anxiety and depressive symptoms across the transition to motherhood: Associations with maternal and environmental characteristics. Women & Health, 44, 61–77. doi:10.1300/J013v44n03_04

Shear, K., & Shair, H. (2005). Attachment, loss, and complicated grief. Developmental Psychobiology, 47, 253–267. doi:10.1002/dev.20091

Simon, N. M. (2013). Treating complicated grief. JAMA, 310, 416–423. doi:10.1001/jama.2013.8614

Sinclair, S. J., Blais, M. A., Gansler, D. A., Sandberg, E., Bistis, K., & LoCicero, A. (2010). Psychometric properties of the Rosenberg Self-Esteem Scale: Overall and across demographic groups living within the United States. Evaluation & the Health Professions, 33, 56–80. doi:10.1177/0163278709356187

Staneva, A., & Wittkowski, A. (2013). Exploring beliefs and expectations about motherhood in Bulgarian mothers: A qualitative study. Midwifery, 29, 260–267. doi:10.1016/j.midw.2012.01.008

Whitton, A., Warner, R., & Appleby, L. (1996). The pathway to care in post-natal depression: Women’s attitudes to post-natal depression and its treatment. The British Journal of General Practice, 46, 427–428.

Winokuer, H. R., & Harris, D. L. (2012). Principles and practice of grief counseling. New York, NY: Springer Publishing Company.

Women’s Health Queensland Wide. (2009). Antenatal and postnatal depression. Retrieved from http://www.womhealth.org.au/pregnancy-and-parenting/antenatal-and-postnatal-depression-booklet

Wylie, L., Hollins Martin, C. J., Marland, G., Martin, C. R., & Rankin, J. (2011). The enigma of post-natal depression: An update. Journal of Psychiatric and Mental Health Nursing, 18, 48–58. doi:10.1111/j.1365-2850.2010.01626.x

 

Tables

Table 1

Descriptive Statistics for Nominal Variables

Nominal variables
Variable n %
Number of babies
1 87 49.4
2 60 34.1
3 21 11.9
Age
18–21 2 1.1
22–25 11 6.3
26–30 37 21.0
31–40 82 46.6
41–50 30 17.0
51–60 10 5.7
61-70 2 1.1
Missing data 2 1.1
Marital status
Single, never married 11 6.3
Married 123 77.0
Separated 4 2.3
Divorced 6 3.4
De facto 29 16.7
Widowed 1 .6
Missing data 2 1.1
Education level
University degree 108 61.3
TAFE/trade certificate/diploma 33 18.8
Year 12 21 11.9
Year 10 10 5.8
Did not finish school 1 .6
Missing data 3 1.7
Diagnosis of PND
Yes 37 21.0
No 139 79.0
Mother personally felt she had PND
Yes 70 39.8
No 106 60.2

 

 

Table 2

Descriptive Statistics and t-Test Results for Continuous Variables

Continuous variables
Prenatal Expectations Postnatal Experience
Variable M SD M SD Min Max t(175)
1st baby: infant 13.50 2.98 13.84 3.53 6 24 -.96
1st baby: support 19.90 3.53 17.81 4.21 7 28 6.76***
1st baby: self 32.11 3.48 28.81 5.31 10 40 7.16***
1st baby: depression 4.43 4.72 0 21
1st baby: anxiety 3.08 3.31 0 21
1st baby: stress 7.45 4.67 0 21
1st baby: self-esteem 18.31 7.02 0 30
2nd baby: infant 14.70 3.51 16.43 3.71 6 24 3.72***
2nd baby: support 18.65 3.58 18.35 4.37 7 28 .77
2nd baby: self 32.18 3.53 32.20 4.92 10 40 -.05
2nd baby: depression 4.38 5.54 0 21
2nd baby: anxiety 2.63 4.20 0 21
2nd baby: stress 6.45 5.26 0 21
2nd baby: self-esteem 19.94 7.56 0 30
3rd baby: infant 15.86 3.60 15.59 3.46 6 24 .42
3rd baby: support 19.34 3.75 18.62 4.16 7 28 .18
3rd baby: self 33.28 3.34 32.62 6.04 10 40 .77
3rd baby: depression 4.10 5.29 0 21
3rd baby: anxiety 2.45 3.87 0 21
3rd baby: stress 6.79 5.41 0 21
3rd baby: self-esteem 21.17 7.86 0 30

Note. Due to multiple t-tests a more stringent significance level of .0167 was calculated using a Bonferroni correction.

PND: Postnatal Depression

*** p <.001

 

 

Table 3

Zero-Order Correlations for Independent and Dependent Variables (DVs) Used in the Hierarchical Regression Analyses

 

Variable 1 2 3 4 5 6 7
Baby 1:
1.          Infant Expectations .28** .63** -.40** -.29** -.47** .55**
2.          Support Expectations .43** -.45** -.29** -.36** .44**
3.          Self Expectations -.68** -.57** -.71** .73**
4.          Depression (DV) .71** .78** -.79**
5.          Anxiety (DV) .74** -.64**
6.          Stress (DV) -.73**
7.          Self-esteem (DV)
Baby 2:
1.          Infant Expectations .33** .58** -.36** -.28** -.38** .45**
2.          Support Expectations .47** -.26* -.10 -.17 .33**
3.          Self Expectations -.49** -.43** -.49** .56**
4.          Depression .80** .87** -.81**
5.          Anxiety .82** -.71**
6.          Stress -.73**
7.          Self-esteem
Baby 3:
1.          Infant Expectations .72** .41* -.38* -.28 -.41* .31
2.          Support Expectations .51** -.48** -.27 -.36 .36
3.          Self Expectations -.65** -.50** -.49 .60**
4.          Depression .76** .84** -.79**
5.          Anxiety .86** -.68**
6.          Stress -.79**
7.          Self-esteem

* p < .05. ** <.01.

Table 4

Multiple Regression Analysis Predicting Depression With First Baby

Model R2 Variables
(Residualized Scores)
β B SE B t
1 .28 Infant -.30 -1.40 .32 -4.40***
Support -.37 -1.74 .32 -5.48***
2 .50 Infant .06 .27 .33 .81
Support -.19 -.91 .28 -3.22**
Self -.64 -3.01 .35 -8.59***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

** p <.01. *** p < .001

 

Table 5

Multiple Regression Analysis Predicting Anxiety With First Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .13 Infant Expectation -.23 -.75 .25 -3.05**
Support Expectation -.22 -.73 .25 -2.99**
2 .33 Infant .12 .38 .27 1.44
Support Expectation -.05 -.17 .23 -.74
Self Expectation -.62 -2.05 .28 -7.21***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 6

Multiple Regression Analysis Predicting Stress With First Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .28 Infant Expectation -.40 -1.88 .32 -5.96***
Support Expectation -.24 -1.13 .32 -3.58***
2 .50 Infant .04 .19 .33 .58
Support Expectation -.06 -.29 .28 -1.03
Self Expectation -.65 -3.06 .35 -8.85***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 7

Multiple Regression Analysis Predicting Self-Esteem With First Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .39 Infant Expectation .47 3.28 .44 7.53***
Support Expectation .30 2.14 .44 4.90***
2 .56 Infant .15 1.08 .46 2.36
Support Expectation .15 1.04 .40 2.64**
Self Expectation .56 3.97 .49 8.13***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 8

Multiple Regression Analysis Predicting Depression With Second Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .15 Infant Expectation -.31 -1.72 .59 -2.93**
Support Expectation -.15 -.86 .59 -1.46
2 .25 Infant .12 .64 .65 .99
Support Expectation -.03 -.17 .60 -.28
Self Expectation -.41 -2.26 .69 -3.27**

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 9

Multiple Regression Analysis Predicting Anxiety With Second Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .08 Infant Expectation -.28 -1.16 .47 -2.50
Support Expectation -.00 -.01 .47 -.03
2 .20 Infant -.05 -.22 .51 -.43
Support Expectation .14 .59 .47 1.27
Self Expectation -.47 -1.97 .54 -3.66***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 10

Multiple Regression Analysis Predicting Stress With Second Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .15 Infant Expectation -.36 -1.93 .56 -3.44**
Support Expectation -.04 -.23 .56 -.41
2 .51 Infant -.15 -.81 .61 -1.32
Support Expectation .09 .49 .56 .87
Self Expectation -.44 -2.34 .65 -3.60**

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 11

Multiple Regression Analysis Predicting Self-Esteem With Second Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .24 Infant Expectation .38 2.91 .76 3.84***
Support Expectation .20 1.55 .76 2.05
2 .34 Infant .18 1.37 .83 1.66
Support Expectation .08 .57 .76 .75
Self Expectation .42 3.21 .88 3.65***

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 12

Multiple Regression Analysis Predicting Depression With Third Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .23 Infant Expectation -.09 -.46 1.33 -.35
Support Expectation -.42 -2.24 1.33 -1.68
2 .45 Infant .04 .22 1.15 -.19
Support Expectation -.17 -.92 1.22 -.75
Self Expectation -.55 -2.94 .93 -3.16**

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

 

Table 13

Multiple Regression Analysis Predicting Anxiety With Third Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .09 Infant Expectation -.18 -.72 1.06 -.67
Support Expectation -.14 -.56 1.06 -.52
2 .26 Infant -.14 -.56 .98 -.57
Support Expectation .07 .29 1.04 .28
Self Expectation -.48 -1.87 .79 -2.37

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

 

Table 14

Multiple Regression Analysis Predicting Stress With Third Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .18 Infant Expectation -.31 -1.71 1.41 -1.21
Support Expectation -.14 -.77 1.41 -.55
2 .30 Infant -.28 -1.53 1.34 -1.14
Support Expectation .04 .21 1.42 .15
Self Expectation -.40 -2.19 1.08 -2.03

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

Table 15

Multiple Regression Analysis Predicting Self-Esteem With Third Baby

Model R2 Variables
(Residualized Score)
β B SE B t
1 .13 Infant Expectation 1.00 .80 2.10 .38
Support Expectation .29 2.31 2.10 1.10
2 .36 Infant .05 .43 1.85 .23
Support Expectation .04 .31 1.96 .16
Self Expectation .55 4.44 1.49 2.98**

Note. Due to undertaking four multiple regressions per baby, a more stringent significance level of .013 was calculated using a Bonferroni correction.

* p < .05. ** p <.01. *** p < .001

 

 

Figure 1 (see downloadable PDF for this figure). Standardized regression coefficients for the relationship between infant discrepancy scores, support discrepancy scores and depression (DV), anxiety (DV), stress (DV) and self-esteem (DV) for Baby 1, as mediated by self-discrepancy scores.
Note. Infant and support are correlated as per the zero-order correlation table (see Table 3).

** p <.01. *** p < .001

 

Figure 2 (see downloadable PDF for this figure). Standardized regression coefficients for the relationship between infant discrepancy scores, support discrepancy scores and depression (DV), anxiety (DV), stress (DV) and self-esteem (DV) for Baby 2, as mediated by self-discrepancy scores.
Note. Infant and support are correlated as per the zero-order correlation table (see Table 3).

** p <.01. *** p < .001

Figure 3 (see downloadable PDF for this figure). No significant relationships were found between infant discrepancy scores, support discrepancy scores and depression (DV), anxiety (DV), stress (DV) and self-esteem (DV) for Baby 3, as mediated by self-discrepancy scores.
Note. Infant and support are correlated as per the zero-order correlation table (see Table 3).

** p <.01. *** p < .001

 

Author Note:

Kathryn Lazarus, School of Psychology, the University of Queensland; Pieter Rossouw, School of Psychology and School of Nursing, Midwifery and Social Work, the University of Queensland.

Correspondence concerning the article should be addressed to Kathryn Lazarus, School of Psychology, the University of Queensland, St Lucia, QLD 4072.

Email: kathryn1@onthenet.com.au

 

support
Need Help?
Support Ticket