Pregnancy Intendedness, Maternal-Fetal Bonding, and Postnatal Maternal-Infant Bonding

Abstract

This study models associations between pregnancy intendedness and prenatal maternal-fetal bonding with postnatal maternal-infant bonding. Unintended pregnancies are associated with disruptions in maternal-infant bonding, which has long-term adverse implications for maternal and child well-being. Given the high proportion of births that are unintended in the U.S., identifying protective factors is critical. Pregnant women (ages 16-38) were recruited from two prenatal clinics in a metropolitan city in the South-Central U.S. at their first prenatal visit and followed throughout pregnancy and post-birth. Multiple regression analyses examined associations, mediation, and moderation. Results indicate that the more unintended/unwanted women reported their pregnancies to be, the lower they reported their maternal-infant bonding post-birth, and higher maternal-fetal bonding scores are associated with higher postnatal maternal-infant bonding. An interaction revealed that a higher level of prenatal bonding is protective for postnatal bonding among those with unintended/unwanted pregnancies. Because prenatal bonding can be enhanced through intervention, it is a promising target to reduce the risks associated with unintended pregnancy.

1. INTRODUCTION

Nearly half (45%) of all pregnancies in the United States are classified as unintended (e.g., reported as occurring sooner than desired or as unwanted) (Finer & Zola, 2016). Unintended births are associated with numerous adverse consequences including poorer maternal and child health and development (Brown & Eisenberg, 1995; Cheng et al., 2009; Shah et al., 2011); these outcomes are due, in part, to disruptions in mother-infant bonding (Kinsey & Hupcey, 2013) which are strongly correlated with maternal caregiving (Coyl et al., 2002). It is therefore critical to explore modifiable factors that can buffer the adverse outcomes associated with unintended pregnancy.

Maternal-fetal bonding (MFB), referring to the affectionate feelings a mother develops for her baby during pregnancy (Brandon et al., 2009; Cranley, 1981; Müller & Mercer, 1993), is one such factor. MFB, sometimes called prenatal attachment, predicts the postnatal mother-child relationship and the quality of maternal care provided (Siddiqui & Hagglof, 2000). In particular, MFB serves as a direct indicator for maternal sensitivity and postnatal mother-infant bonding (Maas et al., 2013; 2016; Rossen et al., 2017). Although women with unintended pregnancies tend to report less MFB initially (Damato, 2004; Rubertsson et al., 2014), the effect size is small (Yarcheski et al., 2009). Life course characteristics (e.g., maternal age, relationship status, maternal employment, and number of children) have previously been correlated with MFB (Petri et al., 2018; Rossen et al., 2016). Moreover, MFB is modifiable and can be enhanced through interventions (de Jong-Pliej et al., 2013; Nishikawa & Sakakibara, 2013; Sandbrook & Adamson-Macedo, 2004; Shreffler et al., 2019). While evidence for associations between MFB and postnatal mother-infant bonding has been established in the literature, studies have not explored whether the impact of an unintended pregnancy on postnatal maternal-infant bonding is mediated or modified by MFB.

In this study, we use data collected from participants at two reproductive health clinics in a South-Central metropolitan city. We examine the associations between pregnancy intendedness as reported at the first prenatal appointment, MFB measured during the 2^nd trimester, and postnatal mother-infant bonding measured approximately one-month post-birth and examine the interaction between intentions and MFB for maternal-infant bonding. We use a multidimensional measure of pregnancy intendedness designed to capture the wantedness and timing regarding a current pregnancy (Santelli et al., 2009).

1.1. Prenatal Bonding in the Context of Unintended Pregnancies

Deeply rooted in Bowlby’s (1982; 1988) attachment theory, MFB includes a mother’s perceptions, emotions, and behaviors relative to her baby and the internalized representation of the baby that takes place during pregnancy (Brandon et al., 2009). Rubin (1975) proposed the construct of early bonding that occurs during pregnancy to include communication with the baby, fantasy about the baby, mental images of the baby, and planning for a future with the baby. These “internalized representations” (Bretheron & Munholland, 2008) provide information about the meaning of the unborn children for their parents and have been found to contribute to how a mother will interact with her newborn in the first several months postpartum (Tambelli et al., 2014; Viaux-Savelon et al., 2012) and first year of her infant’s life (Dayton et al., 2010; Maas et al., 2012). Studies have shown that these prenatal representations, or how the unborn child is viewed or held in the mind of the mother, can influence a number of notable outcomes for both mother and child, including developmental progression of the parent-infant attachment relationship (Raval et al., 2001; Theran et al., 2005), the child’s neurological development (Fonagy et al., 2004), the affective regulation style employed by the parent (i.e., balanced, disengaged, or distorted; Zeanah et al., 1993), and the parent’s openness and responsiveness to their child’s expressions and needs (van IJzendoorn, 1995; Zeanah et al., 1994).

Leifer (1977) noted that women begin an affectionate relationship with their baby during the first trimester of pregnancy, but that feelings are abstract and fantasy-like until the pregnancy progresses to a point where the baby becomes more concrete through movement and perceived personality characteristics. At this point, mothers are likely to actively engage with their babies through both verbal and tactile interaction, which includes talking to the baby, calling the baby by name, pushing the baby to move, and watching the baby move through belly movements. Supporting this theory, MFB scores have been found to increase as pregnancy progresses, with a particularly large increase around 20 weeks gestation when quickening occurs and baby’s movements can be felt (Laxton-Kane & Slade, 2002; Tichelman et al., 2019; Vedova et al., 2008).

Feelings of prenatal bonding may begin even before gestation, as the cognitive and emotional abilities to conceptualize a child are possible before women become pregnant (Doan & Zimerman, 2003). Pregnancies that are intended are associated with higher MFB scores (Damato, 2004; Rubertsson et al., 2014), particularly when both partners view the pregnancy favorably (Siddiqui, Hagglof, & Eisemann, 1999). Yet additional life course characteristics are also associated with MFB; younger maternal age, being a first-time mother, having children previously, and being employed full-time are independently correlated with higher prenatal bonding scores (Malm et al., 2016; Rossen et al., 2016). It is possible, therefore, that MFB may serve as a mediator of the association between pregnancy intentions and postnatal bonding or as a moderator. As MFB has gained increased recognition for playing a critical role in optimal parenting and child outcomes (Salehi & Kohan, 2017) it has been recently examined in a number of studies, acting as the mediator between factors such as maternal mental health, dyadic adjustment, and postnatal maternal-infant bonding (Brandão et al., 2019; Matthies et al., 2020; Punamäki et al., 2017). Additionally, MFB has been found to moderate the pathway connecting prenatal depression and anxiety to social support (Hopkins et al., 2018), as well as changing the manner in which personality influences maternal postpartum depression (Priel & Besser, 1999).

Despite its increased use in mediation and moderation analyses, MFB has yet to be examined as a mediator or moderator in the pathway connecting pregnancy intendedness and postnatal maternal-infant bonding. While mediation and moderation represent different conceptual processes, these previous findings provide support for considering both statistical methods in an attempt to study the interconnection of these factors, as well as contribute to increased understanding of causation and model development, which are crucial for developing effective, efficient, and relevant intervention and support programs (Walsh et al., 2013), particularly for mothers who may experience unintended pregnancy.

1.2. Prenatal Maternal-Fetal Bonding and Postnatal Mother-Infant Bonding

The early mother-infant relationship is critical to a child’s future (Ainsworth, 1979), as it profoundly affects later social, cognitive, and behavioral development along with the physical health of the child (Mantymaa, 2006). Brockington (2004) asserts that the mother-infant relationship is the most essential and significant process in the postpartum period. This relationship is comprised of a complex interplay between maternal caregiving behaviors, maternal emotional regulation, and maternal cognitive appraisal in concert with signaling and care-eliciting behaviors on the part of the infant (Galbally et al., 2011). Evolutionarily, the function of maternal bonding is to secure nurturing and protection, and ultimately, survival of the child (Bowlby, 1982; Siegel, 2020).

Attachment theory contends that sensitive interactions between a mother and her infant provide a basis for secure attachment and development of a positive internal working model that allows for the development of future relationships, expectations of others, as well as the development of one’s self-perception (Bowlby, 1982; Bretherton, 1985; Bretherton, & Munholland, 2008). Once formed, this internal model is thought to remain relatively stable throughout the lifespan (Bar-Haim et al., 2000; Bowlby, 1982), though additional factors such as experiences, temperament, and epigenetic characteristics also play a critical role in development (Seigel, 2020).

MFB is highly correlated with postnatal maternal-infant bonding (Dubber et al., 2015; Müller, 1996; Siddiqui & Hagglof, 2000; Tichelman et al., 2019). Indeed, Müller (1996) found that 17% of postnatal bonding is explained by prenatal bonding. Salisbury, Law, LaGasse, and Lester (2003) suggest that the behavioral acts during pregnancy are critical in the development of the mother-infant relationship; these acts demonstrate care and commitment to the baby, and include aspects of nurturance, comforting, and physical preparation. These prenatal behaviors have been referred to as the earliest form of parenting (Condon & Corkindale, 1997; Glover & Capron, 2017). The implications of prenatal bonding for the postpartum mother-infant relationship are therefore substantial. Prenatal bonding has been linked to postpartum maternal sensitivity (Shin, Park, & Kim, 2006), postpartum maternal involvement (Foley & Hughes, 2018; Siddiqui & Hagglof, 2000), and mother-infant attachment scores within the first postpartum year (Shin et al., 2006). Greater levels of MFB are also associated with secure mother-infant attachment styles and earlier achievement of developmental milestones for infants (Alhusen et al., 2012), with positive effects spanning into adulthood (Waters et al., 2003).

1.3. Current Study

Lasting effects on the physical health of a child in addition to issues with social, cognitive, and behavioral development have been shown to manifest as a result of deficits in postnatal mother-infant bonding (Condon, 1993; Doan & Zimmerman, 2003; Mantymaa, 2006). As such, it is important to identify potential factors that strengthen the maternal-infant relationship or buffer the impact of risk factors for deficits in maternal-infant bonding, such as unintended pregnancy (Damato, 2004). Prenatal bonding is a promising target. Prior research indicates that maternal prenatal bonding can be enhanced through interventions with items such as abdominal palpitation, listening to the heartbeat, and viewing ultrasound scans (Cuijlits et al., 2019; Nishikawa & Sakakibara, 2013; Sandbrook & Adamson-Macedo, 2004; Shreffler et al., 2019). The goal of this study is therefore to examine whether prenatal bonding can buffer the impact of unintended pregnancy on postnatal maternal-infant bonding. The study draws from a diverse sample of 124 pregnant women who were recruited during their first prenatal visit and followed after birth. Although the sample is small, the longitudinal data allow us to prospectively examine prenatal factors that are significantly associated with subsequent maternal-infant bonding.

2. METHOD

2.1. Sample

The study was reviewed and approved by the Oklahoma State University Institutional Review Board before data collection began. Between October, 2017 and May, 2018, women (aged 16-38 years, mean = 25; SD = 5.5) attending their first scheduled prenatal appointment at either of the two area prenatal clinics serving predominately low-income patients in a South Central urban city were invited to participate in the study. Written informed consent was obtained from all participants (N=177). Informed consent was also obtained from a parent or legal guardian for participants younger than 18. Surveys were conducted in English. Of the 177 women who participated in the first survey assessment, 124 participated in multiple waves, including the survey conducted approximately one month postpartum. The sample for this study includes the 124 women who participated through the one-month postpartum survey.

2.2. Measures

The first survey assessed factors that shaped women’s childbearing attitudes, intentions, and behaviors. The survey included various measures addressing women’s fertility intentions, reproductive histories, and sociodemographic information, as well as various factors that could increase risk for adverse birth outcomes, such as prenatal care, partner relationship, health insurance, etc. More than 70% of participants were in their first trimester at the time of the first assessment. An additional survey conducted during the latter part of the second pregnancy trimester assessed MFB. The one-month post-birth survey inquired about the birth experience and maternal-infant bonding.

2.2.1. Pregnancy Intendedness

Per recommendations by Santelli and colleagues (2003; 2009), we assessed pregnancy intendedness with three questions regarding different aspects of intentions; whether the pregnancy was viewed as “unplanned,” whether the pregnancy was considered mistimed, and whether the respondent indicated not wanting to have a baby. To assess whether the pregnancy was unplanned, participants were asked, “Was this pregnancy unplanned?” In order to assess whether a pregnancy was mistimed, women were asked, “Would you say that you became pregnant too soon, about the right time, later than you wanted to, didn’t care about the timing, or don’t know/not sure?” To assess unwantedness, participants were asked, “Right before you became pregnant, did you want to have a baby at any time in the future?” Pregnancy intendedness was coded such that “1” indicated that participants reported their pregnancies as planned, occurring on time or later (or that timing did not matter), and that they wanted to have a baby in the future. Responses were coded as a “2” if they reported their pregnancies as unplanned or occurring “too soon,” but that they wanted to have a baby in the future. Responses were coded as a “3” if participants responded that they did not want to have a baby in the future. Coding pregnancy intendedness as a single continuous variable in this way addresses previous concerns that an intentions measure needs to incorporate both desires for a baby as well as planning (Stanford et al., 2000) and should represent a continuum of responses as opposed to a dichotomous intended/unintended measure (Bachrach & Newcomer, 1999), with mistimed or ambivalence as a middle category. The coding used in this study is similar to that recommended by Kost and Lindberg (2015), though we were unable to parse out the extent of mistiming due to the smaller sample size in this study.

2.2.2. Maternal-Fetal Bonding

The Prenatal Attachment Inventory (PAI) (Müller & Mercer, 1993; Siddiqui et al., 1999) was utilized to measure maternal-fetal bonding (MFB) levels in the second trimester survey of this study. The PAI is based on principles of attachment theory and is a self-report scale consisting of 21 Likert-type items ranging from 1 (almost never) to 4 (almost always). The scale was designed to measure how often a mother has affectionate thoughts or behaviors directed towards her baby. Total possible scores range from 21 to 84, with higher scores representing higher levels of prenatal bonding. Our sample PAI scores also ranged from 21 to 84. Reliability analysis with the current sample indicates a Cronbach’s alpha = .93. The construct validity of this measure has been assessed using confirmatory factor analysis, which indicated the PAI is a valid instrument to assess levels of prenatal bonding in expectant mothers (Gau & Lee, 2003).

2.2.3. Postnatal Mother-Infant Bonding

The Mother-to-Infant Bonding Scale (MIBS) (Bienfait et al., 2011: Taylor, Atkins, Kumar, Adams, & Glover, 2005) was utilized to measure postnatal bonding levels and is the only variable in the current study drawn from the one-month post-birth survey. The MIBS is a self-rated questionnaire consisting of 8 feeling descriptors (loving, resentful, neutral/felt nothing, joyful, dislike, protective, disappointed, and aggressive) assessed on a 4 point Likert scale ranging from very much (0) to not at all (3). When the feeling descriptor represents a negative emotional response, the scoring is reversed. Total possible scores on the MIBS range from 0-24, with high scores indicating a higher level of bonding. The current study range was from 13-24, with a Cronbach’s alpha = .71. The MIBS has been found to satisfactorily detect problematic mother-infant bonding (Bienfait at al., 2011; van Bussel, Spitz, & Demyttenaere, 2010; Wittkowski, Wieck, & Mann, 2007) and the potential rejection of the infant by the mother (van Bussel et al., 2010; Wittkowski et al., 2007).

2.2.4. Sociodemographic Control Variables

Several control variables previously associated with postnatal mother-infant bonding were included in the current study, including maternal age, employment status, marital status, and number of children (Petri et al., 2018; Rossen et al., 2016). Age is included as a continuous variable. Employment status was coded as a dichotomous variable, with participants working either full- or part-time coded as a “1” and those not working as a “0.” Those who indicated they lived in a marriage or cohabiting union were coded as “1,” with those not in a union coded as “0.” Number of children is included as a continuous variable ranging from 0 to 5 in our sample. Education in years is included as a continuous variable ranging from 8 to 19 in our sample. Race/ethnicity was measured using the standard Census questions about race and Hispanic ethnicity and coded into dummy variables with non-Hispanic White as the reference category and Black, Hispanic, American Indian, and “Other” included in the analysis.

2.3. Data Analysis

Descriptive statistics including means and standard deviations summarized the study variables. Ordinary least squares regression analyses were used to examine associations with postnatal mother-infant bonding. The first model includes pregnancy intendedness only, the second model adds MFB, model 3 includes the life course control variables of maternal age, employment status, union status, number of children, educational attainment, and race/ethnicity. The interaction term was added in the final model.

3. RESULTS

3.1. Descriptive Statistics

Table 1 presents the descriptive statistics and Pearson correlation coefficients of the study variables. On average, maternal-infant (i.e., postnatal) bonding scores were quite high, with the mean (M=23.11; SD=1.97) near the upper range of the scale. Maternal-infant bonding was significantly negatively correlated with pregnancy intendedness (r=−.32; p<.01) and number of children (r=−.33; p<.001) indicating that postnatal bonding scores tended to be lower among pregnancies that were more unwanted and multiparous. Scores were lower among Black women (r=−.23; p<.01). Maternal-infant bonding was also significantly positively correlated with maternal-fetal bonding (r=.39; p<.001), suggesting that women with higher levels of bonding prenatally also reported higher levels of bonding after childbirth.

Table 1.

Descriptive Statistics and Correlations of Study Variables (N=124).

		(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)
(1)	Maternal-Infant Bonding	--
(2)	Pregnancy Intendedness	−.32 **	--
(3)	Prenatal Maternal-Fetal Bonding	.39 ***	−.13	--
(4)	Maternal Age	−.03	−.03	−.11	--
(5)	Employment	.09	−.13	.00	.09	--
(6)	Union (Marriage/Cohabiting)	.06	−.23 **	−.05	−.05	.09	--
(7)	Number of Children	−.33 ***	.19 *	−.10	.54 ***	−.16 *	−.08	--
(8)	Education in Years	.08	−.16	−.03	37 ***	.22 **	−.01	−.06	--
(9)	White	.15	−.16 *	.08	.21 **	.10	.18 *	.06	.12	--
(10)	Black	−.23 *	.08	−.04	−.002	−.12	−.35 ***	.07	.07	−.51 ***	--
(11)	Hispanic	−.06	.09	.03	−.22 **	.10	.07	−.06	−.01	−.31 ***	−.25 **	--
(12)	American Indian	.11	.05	−.07	−.09	−.08	.09	−.10	−.27 **	−.38 ***	−.20 ***	−.17 *	--
(13)	Other race/ethnicity	.04	.04	.08	.04	.09	.07	.05	.05	−.06	−.05	−.03	−.04	--
M		23.11	1.71	63.66	25.16	.41	.58	1.23	12.79	.40	.28	.14	.18	.01
SD		1.97	.60	13.08	5.54	.49	.50	1.28	1.95	.49	.45	.34	.38	.08
Range		13-24	1-3	21-84	16-38	0-1	0-1	0-5	8-19	0-1	0-1	0-1	0-1	0-1

Note: Range refers to sample range.

^*p<.05

^**p<.01

^***p<.001.

The mean score of pregnancy intendedness was 1.71 (SD=.60), indicating that more pregnancies were reported as intended or unplanned/mistimed as opposed to unwanted. We did not find a significant correlation between pregnancy intendedness and MFB, indicating that a mediation model is unwarranted. We found significant correlations between pregnancy intendedness and being in a marital or cohabiting union (r=−.23; p<.01), number of children (.19; p<.05), and being White (r=−.16; p<.01). MFB scores in the study spanned the possible range from 21-84 (M=63.66; SD=13.08). The average age was 25 (M=25.16; SD=5.54), and 41% of the study participants reported part- or full-time employment. Approximately 58% of participants were living in a union, with one child on average (M=1.23; SD=1.28). The average level of education in the sample was a high school diploma (M=12.79; 1.95). Approximately 40% of the sample reported race/ethnicity as White; 28% Black, 14% Hispanic; 18% American Indian, and 1% “Other.”

3.2. Postnatal Maternal-Infant Bonding

Table 2 presents the multivariate analyses (OLS regression) examining postnatal maternal-infant bonding in a series of four regression models. Model 1 displays coefficients for pregnancy intendedness before other variables were entered. Pregnancy intendedness was significantly associated with postnatal maternal-infant bonding scores (b=−.97, p<.01), indicating that pregnancies reported as more unwanted were associated with lower postnatal bonding. Model 2 included MFB, which was significantly associated with higher postnatal bonding (b=.05, p<.001). Pregnancy intendedness remained significantly associated with postnatal bonding (b=−.90, p<.01). Model 3 included the control variables. The associations between intendedness, MFB, and postnatal bonding remained significant; additionally, number of children (b=−.38; p<.05) and Black race (b=−1.07; p<.05) were significantly associated with postnatal bonding. An interaction term between pregnancy intendedness and MFB was included in the final model, which was significantly associated with postnatal bonding (b=.09, p<.001).

Table 2.

Linear Regression Predicting Maternal-Infant Bonding at One Month Post-Birth (N=124).

	Model 1		Model 2		Model 3		Model 4
Variables	b	SE	b	SE	b	SE	b	SE
Pregnancy Intendedness	−.97 **	.31	−.90 **	.29	−.69 *	.33	−6.18 ***	1.42
Prenatal Maternal-Fetal Bonding			.05 ***	.01	.05 ***	.01	−.11 *	.04
Maternal Age					.02	.05	.03	.05
Employment					.36	.35	.40	.32
Union (Marriage/Cohabiting)					−.30	.40	−.05	.37
Number of Children					−.30	.19	−.34	.18
Education in Years					−.02	.10	−.03	.09
White (reference)
Black					−1.07 *	.44	−1.05 *	.40
Hispanic					−.21	.62	−.15	.57
American Indian					−.11	.51	−.47	.48
Other race/ethnicity					.12	1.69	−.37	1.56
Intentions X MFB							.09 ***	.02
Constant	24.85 ***	.54	21.44 ***	1.04	21.64 ***	1.95	31.35 ***	3.04

Note:

^*p<.05

^**p<.01

^***p<.001.

To ease readability of the interaction term, a figure is presented depicting the interaction between pregnancy intendedness and MFB (1 standard deviation below and above the mean) on postnatal maternal-infant bonding. The graph highlights that postnatal bonding was stable and higher among women with an intended pregnancy, regardless of level of MFB. For pregnancies that were unintended and unwanted, however, MFB was protective for postnatal bonding.

Figure 1.

Interaction between pregnancy intention and prenatal maternal-fetal bonding predicting postnatal mother-infant bonding.

4. DISCUSSION

The primary goal of this study was to examine whether MFB—a modifiable factor—buffers the association between pregnancy intendedness and postnatal bonding. Our regression results indicate that mothers with unintended/unwanted pregnancies reported lower levels of bonding with their newborn infants. Additionally, MFB exerted an independent and positive effect on postnatal bonding, suggesting that prenatal bonding contributes to or predicts postnatal bonding regardless of whether or not a pregnancy was intended. These findings are aligned with prior research; systematic reviews also highlight the associations between unintended pregnancy and maternal-infant bonding (Kinsey & Hupcey, 2013) and prenatal bonding and maternal-infant bonding (Misri & Kendrick, 2008). Although pregnancy intendedness and MFB were independently related to postnatal bonding, we did not find a significant association connecting pregnancy intendedness to MFB. Consistent with prior research, we additionally found that having more children was associated with lower levels of postnatal bonding following the target pregnancy (Tichelman et al., 2019).

We also found a significant interaction between pregnancy intendedness and prenatal bonding for post-birth maternal-infant bonding. Among women with pregnancies that were unintended or unwanted, higher levels of bonding during pregnancy served as a protective factor for postnatal bonding. The lowest levels of postnatal bonding were reported by women who said their pregnancies were unwanted and reported lower levels of bonding during pregnancy. Yet we did not see lower levels of postnatal bonding among those with unintended or unwanted pregnancies when prenatal bonding levels are higher. Taken together, these findings suggest that a window of opportunity may exist where attachment-based intervention during unwanted pregnancies may contribute to enhanced outcomes related to postnatal maternal-infant bonding and infant mental health, especially for those with lower levels of prenatal attachment.

This study contributes to the literature on pregnancy intendedness and postnatal bonding in several key ways. First, recent studies that examined the association between pregnancy intendedness and postnatal mother-child relationships often use a dichotomous measure of unplanned/planned or unintended/intended despite acknowledging the multidimensional nature of pregnancy intendedness (e.g., Boden et al., 2015; Claridge et al., 2017; Nelson & O’Brien, 2011). Our measure also captured pregnancies reported as unwanted—which tend to be associated with worse outcomes than those that are reported as mistimed. This enabled us to ascertain that higher levels of prenatal bonding is protective for postnatal bonding among those with unwanted as well as mistimed pregnancies.

Second, we were able to take advantage of a prospective, longitudinal clinic-based sample that asked pregnant women about the intendedness of their pregnancies in the first trimester, level of MFB during the second trimester, and level of postnatal mother-infant bonding approximately one month after birth. This design enables us to overcome some problems with recall bias as opposed to designs that retrospectively ask women about their intentions and feelings during pregnancy.

We believe that these findings contribute to the theoretical understanding of prenatal and postnatal bonding. Although women with unintended or unwanted pregnancies report lower postnatal bonding, and MFB is also associated with levels of postnatal bonding, there does not appear to be a mediation effect between pregnancy intentions, MFB, and postnatal bonding. Some women with unwanted pregnancies reported high levels of MFB, and this turned out to be protective for postnatal mother-infant bonding. Considering these findings, future studies might look deeper into specific factors or conditions that contribute to varying levels of MFB for mothers with unwanted pregnancies. It is possible that women’s values about motherhood, religion, or culture, for example, impact MFB regardless of pregnancy intendedness or wantedness.

These findings are important as they confirm previous research while addressing some limitations present within other studies, such as the use of a dichotomous pregnancy intention variable and lack of prospective longitudinal data. There are several limitations to the current study findings, however. Self-reported intentions during pregnancy may not be accurately reflective of pre-pregnancy intentions, as perceptions of intentions may change during or after pregnancy (Santelli et al., 2003). However, because women are less likely to admit a pregnancy was unintended during pregnancy than before (David, 2006), this may lead to more conservative estimates of the association between unintended pregnancy and maternal-infant bonding. Furthermore, a self-assessment data gathering technique was employed for this study; therefore, recall and selective bias must be considered along with the impact of social desirability reflected in reported answers on pregnancy intendedness, prenatal bonding, and postnatal bonding. Observational research techniques such as video-recording would strengthen the reliability of the postnatal maternal-infant bonding assessment. In addition, video recordings could be reviewed with the participants to gain greater understanding and context related to bonding-based decisions and interactions. Finally, our finding that Black women reported significantly lower levels of postnatal mother-infant bonding was unexpected and might be an artifact associated with the questions in the bonding measure. Similarly, research examining why Black women reported lower importance of motherhood scores than White women (McQuillan et al., 2008) found that the discrepancy was explained by question wording (Tichenor et al., 2017); Black women were more likely to report valuing raising children than biologically giving birth to them. Ensuring that bonding measures are culturally sensitive and appropriate is a critical future direction for this line of research.

There is, therefore, a need for future research to continue exploring these relationships. Research on the associations between prenatal bonding and postnatal mother-infant bonding may benefit from a representative sample as well as a longitudinal design that follows women during each trimester of pregnancy and well into the postpartum period. Prenatal bonding is likely multi-dimensional and contextual (Doan & Zimmerman, 2003) and culturally relevant (Alhusen, 2008; Cannella, 2005), yet little research has been done to explore these aspects of the development of MFB, which may explain inconsistencies and gaps in the literature. As such, qualitative inquiry that examines narratives of women’s pregnancies and the way they speak about their unborn children may provide greater insight into how bonding during pregnancy develops and is associated with postnatal maternal-infant bonding and may also provide contextual aspects to examine that have yet to be linked to this construct. The identification of simplistic methods to diagnose prenatal bonding disturbances as well as MFB-promoting techniques would also be useful, as it is important for clinicians to be able to easily identify and treat mothers who score low on MFB assessments (Pisoni et al., 2014). Moreover, a number of evidence-based interventions have been established to both assess and educate mothers regarding their relationships with their unborn babies and increase prenatal bonding (Morris et al., 2020). For example, the Nurturing Parenting Programs have been developed specifically to reduce intergenerational transmission of maltreatment and increase parent-child attachment by teaching parents about the importance of prenatal bonding and neonatal attachment (Maher et al., 2011). Additionally, the Parents as Teachers Program emphasizes developmental-centered parenting, focusing on prenatal and postnatal interaction, attachment, and positive parent-child relationships (Zigler et al., 2008).

While effective practices and interventions have been developed, this may not be an easy undertaking as women may find it difficult to express negative feelings about pregnancy and/or motherhood and therefore nurturing and trusted relationships with medical professionals are necessary to normalize and offer judgment-free assistance. Expectant mothers may benefit from referrals to counseling or therapeutic services as well as through direct promotion of methods known to promote MFB during routine obstetric appointments including ultrasound imagining, listening to fetal heart tones, drawing attention to baby through abdominal palpation maneuvers, and kick counting. Expectant mothers may also benefit from direct conversations regarding pregnancy and baby-related topics to promote MFB or to identify possible attachment issues following birth (Pallant et al., 2015; Rubertsson et al., 2015). Reliance on postnatal assessment and screening does a disservice to expectant mothers who could benefit from prenatal intervention (Pisoni et al., 2014). Although the focus on reducing the proportion of pregnancies that are unintended is a worthwhile goal, nearly half of all pregnancies in the U.S. are unintended (Finer & Zolna, 2016)—a rate that has changed little for decades (Mosher et al., 2012). Because many women opt to continue their unintended pregnancies, we argue that focusing on prenatal bonding as a dynamic characteristic capable of growth may help to ameliorate some of the risks of unintended pregnancy for postnatal maternal-infant bonding, potentially promoting improved maternal and child health outcomes (Alhusen et al., 2012; Cannella, 2005), postpartum maternal sensitivity (Shin et al., 2006) and maternal involvement (Foley & Hughes, 2017; Siddiqui & Hagglof, 2000).

Key findings and implications:

1. Women with unintended pregnancies report lower levels of maternal-infant bonding post-birth.

2. Prenatal maternal-fetal bonding moderates the association between unintended pregnancies and postnatal maternal-infant bonding.

3. Because prenatal bonding can be enhanced through intervention, it is a promising target to reduce the risks associated with unintended pregnancy.

Statement of relevance to infant and early childhood mental health:

Our results indicate that maternal-fetal bonding buffers the association between unintended pregnancy and maternal-infant bonding. Reducing the risks of unintended pregnancy for postnatal bonding through increased prenatal bonding has the potential to enhance maternal and child health outcomes, postpartum maternal sensitivity, and maternal involvement.