Postnatal Depression

Abstract

Although breastfeeding has multiple benefits for baby and mother, including maternal mental well-being, many mothers terminate breastfeeding earlier than they desire. We examined two key factors in breastfeeding duration and maternal mental health––breastfeeding efficacy and family–work conflict. Specifically, we examined the moderating role of family–work conflict in the process of breastfeeding efficacy as a predictor of maternal depression by way of duration. In a sample of 61 first-time mothers, we found that breastfeeding duration mediated the relation between prenatal breastfeeding efficacy and depression at 9 months postpartum for working mothers who experienced low levels of family-to-work conflict. That is, for mothers with low family-to-work conflict, higher expected breastfeeding efficacy during pregnancy predicted a longer duration of breastfeeding, which in turn was associated with lower depression at 9 months postpartum. However, for working mothers with high family-to-work conflict, breastfeeding duration did not emerge as an indirect effect on the relation between efficacy and depression. These findings have important implications for a healthy family–work balance to help new mothers adjust when they return to the workforce and as they transition to parenthood.

Keywords

parental status breastfeeding family–work conflict self-efficacy depression

For many first-time mothers, breastfeeding is perceived as an integral and key component in caring for their new infant. Most individuals perceive breastfeeding positively (Avery & Magnus, 2011), as an array of evidence suggests benefits for infants’ overall physical (e.g., Davis, 2001; Van Odjik et al., 2003) and mental health (e.g., Oddy et al., 2010). Additionally, when mothers breastfeed, they experience stabilized hormones, greater postpartum weight loss, a closer connection to their child (e.g., Allen & Hector, 2005; Blincoe, 2005), and a reduction in their own levels of stress (Mezzacappa & Katkin, 2002), depression, and anxiety (Hahn-Holbrook, Haselton, Dunkel-Schetter, & Glynn, 2013; Rondó & Souza, 2007). Despite these numerous positive influences, breastfeeding is a challenge for many mothers, due to pain and difficulty with the baby latching on properly, as well as the significant time commitment (Leeming, Williamson, Lyttle, & Johnson, 2013). Furthermore, many first-time mothers return to work and need to adapt their breastfeeding (or pumping) to their work hours. As a result, breastfeeding duration is often curtailed earlier than mothers desire or anticipate (e.g., Cardenas & Major, 2005), or breastfeeding does not occur at all (e.g., Strong, 2013).

Given the large number of mothers in the paid workforce and dual-earner families in the United States (e.g., Bianchi, Sayer, Milkie, & Robinson, 2012; Del Mar Alonso-Almeida, 2014), family–work balance can have important implications for breastfeeding practices. Moreover, issues with family–work balance and breastfeeding can be critical for maternal well-being. As women transition into their new role as mothers and return to their role as employees, family–work conflict may occur and, despite its many potential benefits, breastfeeding can be a part of this conflict. And, breastfeeding could interfere with work responsibilities specifically for women, because as new fathers, men do not have to ask for accommodations such as a private location to nurse or pump or specific breaks to do so. As a result, breastfeeding could limit gender equality in the workplace and lead to feelings of distress for new mothers (McKinley & Hyde, 2004). Moreover, family–work conflict may be related to working mothers’ decisions to stop breastfeeding (Cardenas & Major, 2005). To further understand the role of family–work conflict in breastfeeding, we sought to expand on prior research by examining whether breastfeeding, as a potentially protective role in mothers’ mental health, is conditional on levels of family–work conflict among first-time mothers who are working full time. More specifically, we focused on one factor that is particularly important in predicting breastfeeding duration–breastfeeding efficacy, or the confidence mothers have in their ability to breastfeed (e.g., Dennis & McQueen, 2007).

Breastfeeding Duration and Mental Health

Breastfeeding is associated with lower levels of psychological distress, negative mood, anxiety, and depressive symptoms (e.g., Groer, Davis, & Hemphill, 2001; Mezzacappa, 2004; Pippins, Brawarsky, Jackson, Fuentes-Afflick, & Haas, 2006). The relation between breastfeeding and reduced psychological distress is often attributed to oxytocin, a stress-reducing hormone that is released during the act of breastfeeding (e.g., Heinrichs, Neumann, & Ehlhert, 2002). Thus, there are acute mental health benefits of breastfeeding. For instance, negative mood decreased immediately after a breastfeeding session (Mezzacappa & Katkin, 2002), and breastfeeding can also have long-term benefits––when mothers continued to breastfeed for a longer duration in the postpartum period, they reported lower depressive symptoms (Dias & Figuerido, 2015; Hahn-Holbrook et al., 2013).

Despite the benefits of long-term breastfeeding and recommendations from the World Health Organization (2009) and the American Academy of Pediatrics (2012) to breastfeed exclusively for at least the first 6 months of a child’s life, only about 20% of new mothers breastfeed for the suggested time frame (Blyth et al., 2002). Failing to meet breastfeeding goals, specifically duration, is linked with poorer maternal psychological outcomes (Rondó & Souza, 2007). Considering early breastfeeding cessation occurs for a majority of mothers (Centers for Disease Control and Disease Prevention, Division of Nutrition, Physical Activity, and Obesity, 2014), researchers have looked into the role of efficacy in predicting breastfeeding duration. As expected, the more efficacious a mother reported feeling about her ability to breastfeed her child (both prenatally and postpartum), the longer she engaged in breastfeeding (Blyth et al., 2002; O’Brien, Buikstra, & Hegney, 2008; Semenic, Carmen, & Gottlieb, 2008; Ystrom, Niegel, Klepp, & Vollrath, 2008). We reasoned that it is important to examine how breastfeeding efficacy in primiparas (or first time) mothers is related to breastfeeding outcomes, as prior breastfeeding experiences can confound measures of efficacy (Blyth et al., 2002; Dennis & Faux, 1999) and duration (Bai, Fong, & Tarrant, 2015; Kronborg & Vaeth, 2004). Given these considerations, the current study focused on the role of breastfeeding efficacy exclusively among first-time mothers.

Breastfeeding Efficacy and Mental Health

In addition to breastfeeding duration, breastfeeding efficacy is associated with depression—the more efficacious a mother felt about her breastfeeding ability, the fewer depressive symptoms she experienced in the postpartum period (Dai & Dennis, 2003; Dennis & McQueen, 2007, 2009; Haga et al., 2011; Ystrom et al., 2008). One substantial limitation in these prior studies is that breastfeeding efficacy was measured only at postpartum, despite likely differences between postpartum and prenatal breastfeeding efficacy. Prenatal efficacy refers to mothers who have yet to meet expectations about breastfeeding because they have not breastfed. Prenatal efficacy is important to examine because violations of expectations in breastfeeding have predicted mental health (e.g., Redshaw & Henderson, 2012). However, current and ongoing challenges and difficulties in attempting to breastfeed, as well as general parenting, are likely to cloud postpartum breastfeeding efficacy. In other words, mothers may adjust their postpartum efficacy based on current experiences, in addition to their prenatal breastfeeding expectations. Thus, it is important to distinguish between prenatal and postpartum breastfeeding efficacy, because efficacy can change after the birth of the baby. For new mothers, examining prenatal efficacy can affirm the importance of preparing for breastfeeding in the anticipation of potential challenges that may arise after the birth. Medical professionals and practitioners can use the information to encourage expectant mothers to sustain breastfeeding. Distinguishing between prenatal and postpartum breastfeeding efficacy could then provide insight as to when breastfeeding interventions are best implemented. To our knowledge, no one has examined whether prenatal or postpartum breastfeeding efficacy is more influential on maternal mental health outcomes. In the current study, we compared the relation between prenatal expected breastfeeding efficacy (during the third trimester of pregnancy) and 1 month postpartum breastfeeding efficacy on 9 months postpartum maternal mental health (Aim 1). We also tested a mediational model that examined the indirect effect of duration on the breastfeeding efficacy and maternal mental health relation (Aim 2).

We proposed that breastfeeding duration may be the underlying mechanism for the association between breastfeeding efficacy and postnatal depression for two reasons. First, new mothers may base their expectations and beliefs about their ability to breastfeed on societal pressure and peer norms (Sheehan, Schmied, & Barclay, 2013). Medical practitioners and general society perceive breastfeeding as the superior form of infant feeding and new mothers may feel stigmatized for bottle-feeding (Wirihana & Barnard, 2012). As a result, mothers may feel pressure not only to breastfeed in general but also to breastfeed for a particular duration of time. Second, the duration of breastfeeding may be an indicator of whether mothers feel successful in not only meeting their breastfeeding goals but also in their expectations or beliefs about themselves as a competent parent (i.e., parenting efficacy; McCarter-Spaulding & Kearney, 2001). If new mothers feel as though they failed in their breastfeeding goals, specifically to breastfeed for a certain amount of time, they may be at risk for greater distress because of perceptions of failure as a parent.

Breastfeeding and Family–Work Conflict

Understanding the issues involved in breastfeeding is especially important for mothers today in the United States, as early return to work makes breastfeeding (especially for the recommended duration) more difficult (Lindberg, 1996; Ogbuanu, Glover, Probst, Liu, & Hussey, 2011). For working mothers, both structural factors (e.g., maternity leave length and flexibility in the workplace) and personal attitudes (e.g., enjoyment of breastfeeding) may predict breastfeeding duration (McKinley & Hyde, 2004). Although general stress responses are diminished while mothers breastfeed (e.g., Groer et al., 2001; Mezzacappa & Katkin, 2002), it is unknown if breastfeeding benefits still occur when working mothers experience conflict between their family and work lives.

People experience family–work conflict when they attempt to simultaneously meet demands in both the home domain and work domain; the direction of the interference occurs both ways, such that family demands can conflict or interfere with work (family-to-work conflict [FWC]) and work demands can interfere with home (work-to-family conflict [WFC]; Edwards & Rothbard, 2000; Frone, Russell, & Cooper, 1992). As new mothers transition back into the workforce and adjust to the responsibilities of their job, they must also maintain their responsibilities related to motherhood, including childcare tasks such as breastfeeding. As a result, family–work conflict is likely to occur among working mothers (Cardenas & Major, 2005). A recent meta-analysis on gender differences in home and work conflict indicated that feelings of domain conflict were more prevalent among women only when they had children (Byron, 2005), and childcare responsibilities were negatively associated with mothers’ feelings of family–work balance (Buffardi, Smith, O’Brien, & Erdwins, 1999). In addition, family–work conflict can have negative implications for breastfeeding adherence and depression for new mothers (Goodman & Crouter, 2009).

We proposed that family–work conflict would complicate the relation between breastfeeding efficacy, duration, and maternal mental health (Aim 3). Specifically, we predicted a moderated mediation model (see Figure 1), such that for mothers with low levels of family–work conflict, breastfeeding efficacy would be indirectly related to their mental health through breastfeeding duration. However, we predict that for mothers with high levels of family–work conflict, breastfeeding efficacy would not be indirectly related to their mental health through duration. Given that domain conflict is bidirectional (e.g., Frone et al., 1992), it is plausible one (or both) direction of family–work conflict would be important for this moderated mediation model of breastfeeding. However, due to the lack of theory and research on this topic, we did not make specific predictions about whether FWC or WFC is more important than the other; rather, these analyses were exploratory.

Figure 1.
Proposed conceptual moderated mediational model. FWC = family-to-work conflict; WFC = work-to-family conflict. FWC and WFC were entered as separate moderators.

The Current Study

In the current study, we proposed that family–work conflict would act as a moderator in the mediational process of breastfeeding efficacy–duration–mental health. Specifically, we predicted breastfeeding duration would only mediate the indirect effect of breastfeeding efficacy on mental health when family–work conflict is low in employed new mothers. To test study predictions, we utilized a sample of 61 employed, first-time mothers. To test whether prenatal or postnatal breastfeeding efficacy is directly related to postnatal depression, we utilized data collected at pregnancy and 1 month postpartum because breastfeeding challenges are likely to occur within the first few weeks of birth. Moreover, we focused on mothers’ family–work conflict at 9 months postpartum because this is the time frame when a majority of employed mothers are likely to return to work, there is greater variability in breastfeeding practices, and mothers are adjusted to the challenges of first-time parenthood—capturing a time when mothers’ mental health is most stable.

Method

Participants

Participants were first-time expectant mothers (N = 61). Mothers were either married or cohabiting (98.4% were married), and on average had been married/cohabiting for about 3 years (M = 3.52, SD = 1.98, range = 1–11). The average age for mothers at baseline was 28 years (SD = 3.60, range = 19–41). The sample was largely homogenous in terms of most demographic characteristics: Mothers were mostly White (90.2%), had at least an undergraduate college degree (85.2%), and middle- to upper income (78.8% of households made more than US$60,000 a year). For full details of participant characteristics, please view Table 1.

Table 1.
Participant Characteristics.

M SD

Age 28.51 3.60

Years in relationship 3.52 1.98

Frequency %

Baby’s sex^a

Male 26 42.6

Female 31 50.8

Relationship status

Married 60 98.4

Cohabiting 1 1.6

Race

White 55 90.2

Black 0 0

Hispanic/Latino 2 3.3

Asian 2 3.3

Other 2 3.3

Education

High school diploma 1 1.6

Some college 8 13.1

College degree 33 54.1

Graduate degree 19 31.1

Total household income

<US$20,000 1 1.6

US$20,001–US$40,000 6 9.8

US$40,001–US$60,000 6 9.8

US$60,001–US$80,000 18 29.5

US$80,001–US$100,000 13 21.3

US$100,001–US$120,000 8 13.1

>US$120,000 9 14.8

Note. Participants (N = 61).

^aFour participants in the current study did not report the sex of their baby.

Procedure

The participants in the current study were from a larger, yearlong study that originally recruited 104 first-time expectant couples (see Biehle & Mickelson, 2011, for additional information). Participant eligibility requirement included fluency in English and employment at the time of the first interview. Recruitment was conducted at local birthing classes (47.6%) and through online message boards (52.4%). Participants were interviewed at four time points. The interviews occurred during the final trimester of pregnancy, 1 month postpartum, 4 months postpartum, and 9 months postpartum. After obtaining informed consent, participants first completed online questionnaires and then within 24 hr they completed a second portion of the interview over the phone with trained interviewers. The combined online and phone interview took each participant approximately 1 hr to complete; participants were compensated US$25 each for their time.

The original study included interview questions that assessed several structural and psychosocial factors related to the transition to parenthood, such as mental health (e.g., postnatal depression, positive affect, anxiety); physical health related to the pregnancy and birth, concerns about parenting, pregnancy, and the birth; measures related to interpersonal relationships (e.g., relationship satisfaction, social support). In the current study, participants (N = 61) provided responses to their breastfeeding efficacy, breastfeeding duration, family–work conflict, and depression across all four waves (i.e., pregnancy, 1 month postpartum, 4 months postpartum, and 9 months postpartum).

Participant eligibility requirements for the current study included intention to breastfeed during pregnancy and full- or part-time employment at 9 months postpartum. After accounting for attrition between the first and last waves of data collection (n = 21) and excluding mothers who were not planning on breastfeeding at the time of pregnancy or were unemployed at 9 months postpartum (n = 22), the final sample size for the study was N = 61 (for access to raw data of the sample of current study, please contact the first author via e-mail). For attrition analyses to assess any differences between the current study sample and the remaining participants from the original study sample (n = 43) at the first wave of data collection, please refer to the preliminary analyses presented in the Results section.

Measures

Sociodemographics

We assessed a number of sociodemographic characteristics potentially related to the major study variables, including age, race/ethnicity, marital status, education, and income, as well as baby’s sex and baby’s health. The current sample ranged from 18 to 52 years old. Race was dichotomized as White or non-White due to our largely homogenous sample. Marital status (cohabiting [i.e., not married] vs. married) was also dichotomized. Education consisted of five categories: some high school, high school, some college, college, or an advanced degree. Income was a measure of the total household income at the time of the first interview, with seven categories: below US$20,000, US$20,001–US$40,000, US$40,001–US$60,000, US$60,001–US$80,000, US$80,001–US$100,000, US$100,001–US$120,000, and more than US$120,000. Baby’s sex was coded as male (0 = 42.6%) or female (1 = 50.8%). Baby’s health problems asked whether the baby had health problems, at each wave (1 = yes, 0 = no). For analyses, we calculated whether the mother ever said yes at any of the three postpartum waves to create a dichotomous variable of baby health problems (1 = yes, 0 = no).

Breastfeeding efficacy

We used the Breastfeeding Self-Efficacy Scale (Dennis, 2003). During pregnancy, we asked participants 5 items about the extent to which they feel confident they will be able to breastfeed their child. At 1 month postpartum, we rephrased the stem question to assess their current breastfeeding efficacy. We asked participants how confident they were in their ability to “determine that baby is getting enough milk,” “breastfeed your baby without using formula as a supplement,” “ensure that your baby is properly latched on for the whole feeding,” “manage the breastfeeding situation to your satisfaction,” and “deal with the fact that breastfeeding can be time-consuming.” Potential responses ranged from 1 = not at all confident to 5 = very confident. A sum score was calculated for the 5 items, with higher scores indicating higher efficacy for breastfeeding. Reliability for the measure was adequate (pregnancy: α = .80; 1 month postpartum: α = .76).

Breastfeeding duration

We measured time duration by calculating the number of months mothers reported to have breastfed. To do this, data about when mothers ceased breastfeeding were collected at 9 months postpartum. If mothers reported that they had ceased breastfeeding, they also indicated at which time point postpartum they stopped (e.g., 6 months). If mothers reported to still be breastfeeding at 9 months postpartum, we coded their score as 9. If participants did not breastfeed at all, we coded their scores as 0. Thus, breastfeeding duration in months ranged from 0 to 9.

A modified version of a single-item measure from Rice, Frone, and McFarlin (1992) and Voyandoff (1988) was used to assess FWC (e.g., “How often do you feel the demands of your family interfere with your work or your job?”) and WFC (e.g., “How often do the demands of work interfere with your family life?”). Responses for FWC and WFC ranged from 1 = never to 4 = very often, with higher scores reflecting higher levels of domain conflict.

Maternal mental health

We assessed maternal mental health using a measure of depressive symptomatology, the Center for Epidemiologic—Depression Inventory (CES-D), which is a well-validated and reliable measure (Radloff, 1977). Participants answered 20 questions assessing their mood over the past week. Example items included “How often have you felt depressed in the past 7 days?” and “How often did you feel that your life was hopeless over the past 7 days?” Responses ranged from 0 = none/rarely (<1 day) to 3 = most (5–7 days), and a total depression score was created by summing scores from the individual items. The scale demonstrated adequate internal consistency for mothers at pregnancy (α = .87) and 9 months postpartum (α = .88). The sample was not highly depressed at either wave, as participants reported only moderate levels of depression (see Table 2 for means). Traditionally, researchers use a score of 16 or higher on the CES-D as a marker for individuals who may suffer from a clinical level of depression (McDowell, 2006).

Table 2.
Descriptives and Bivariate Correlations of Major Study Variables.

Variables Mean SD Range 1 2 3 4 5 6 7

1. Breastfeeding efficacy (prenatal) 16.92 3.48 9–25 —

2. Breastfeeding efficacy (1 MPP) 18.02 4.41 10–25 0.29* —

3. Breastfeeding duration 6.12 3.56 0–9 0.33* 0.54*** —

4. Depression (prenatal) 10.20 6.36 1–35 −0.24 −0.30* 0.02 —

5. Depression (9 MPP) 9.72 7.14 0–31 −0.37 −0.13 −0.12 0.33* —

6. Work-to-family conflict (9 MPP) 1.57 0.50 1–2 −0.07 −0.11 0.11 0.05 0.23 —

7. Family-to-work conflict (9 MPP) 1.95 0.75 1–4 0.00 −0.21 −0.08 0.30* 0.12 0.21 —

Note. Participants (N = 61). MPP = months postpartum.

^†p < .10. p < .05. p ≤ .01. p ≤ .001.

Results

Analytic Strategy

We tested our research aims in three steps. First, we compared the predictive power of prenatal breastfeeding efficacy versus 1 month postpartum breastfeeding efficacy using hierarchical linear regression (Aim 1). To test the mediational model that breastfeeding duration explains the association between breastfeeding efficacy and depression (Aim 2), we examined the significance of the indirect effect with bootstrapped confidence intervals using an SPSS macro, MEDIATE, developed by Preacher and Hayes (2004). Previous researchers recommend examining the significance tests of the indirect effect through bootstrapping because it produces a test that does not depend on large samples and does not assume a normal distribution of the data (e.g., Preacher & Hayes, 2004; Shrout & Bolger, 2002). Bootstrapped mediation analyses used 5,000 replications.

Finally, we tested the moderated mediation model whether the indirect effect of breastfeeding duration on breastfeeding efficacy and depression is conditional on levels of family and work conflict (Aim 3). To test this conditional indirect effect, we used the MODMED macro for SPSS (developed by Preacher, Rucker, & Hayes, 2007). The current study tested the proposed moderated mediational model in which the moderator (FWC and WFC) affects the b path (Preacher et al., 2007)—that is, the association between breastfeeding duration and depression. The MODMED macro also implements bootstrapping to detect indirect effects. As above, we bootstrapped analyses with 5,000 replications. We used bias corrected and accelerated (B_ca) bootstrapped confidence intervals to detect indirect effects in a moderated mediation analysis (Preacher et al., 2007), as it is a more accurate use of bootstrapping because it adjusts for biased and skewed estimations (e.g., Efron & Tibshirani, 1993; Preacher & Hayes, 2004). Note that we examined FWC and WFC, each as a single moderator in a separate model to avoid power and collinearity issues, given the small sample size. Prior to conducting the moderation analyses, we centered all predictor variables in the model for ease of interpretation (Aiken & West, 1991).

Preliminary Analyses

Prior to conducting the main study analyses, we tested for differences in sociodemographic and major study variables between the current study sample (N = 61) and the remaining participants from the 104 expectant mothers in the first wave of data collection (n = 43). With respect to sociodemographic variables, participants in the current study were more likely to be married, χ²(1) = 9.18, p = .002, 57.7%, and have a higher level of education, b = 0.56, SE = .16, OR = 2.68, p = .002. Using a multivariate analysis of variance, we found no significant differences between the current study sample and the remaining sample from the initial data collection in any of the major study variables from the first wave of data collection (prenatal breastfeeding efficacy, depression, FWC, WFC). We also tested for any potential sociodemographic covariates among major study variables, but none emerged as significant predictors.

Table 2 provides means, standard deviations, and bivariate correlations of major study variables. Prenatal breastfeeding efficacy was positively related to breastfeeding duration, r(56) = .33, p = .01, and negatively related to depression at 9 months postpartum, r(59) = −.37, p = .003. Breastfeeding efficacy at 1 month postpartum was also related to breastfeeding duration, r(44) = .54, p < .001, but not to 9 months postpartum depression, r(45) = −.13, p = .38. Furthermore, breastfeeding duration was not related to depression at 9 months postpartum. The lack of significance between breastfeeding duration and depression at 9 months suggests it might be moderated by a third factor (i.e., family–work conflict). Moreover, breastfeeding duration was unrelated to either FWC or WFC, meeting the assumptions that the predictor and moderator variables are orthogonal and that FWC and WFC could serve as potential moderators of the path from duration to depression.

Prenatal Versus Postnatal Breastfeeding Efficacy on Depression

Our first aim of our study was to examine whether prenatal or postpartum breastfeeding efficacy had a stronger association with maternal depression. To test this aim, we conducted hierarchical linear regressions. First, we examined the prenatal breastfeeding efficacy, while controlling for depression at pregnancy as predictors and together these variables accounted for 20% of the variance in depression at 9 months postpartum, R² = .20, F(2, 58) = 7.22, p = .002. We found the higher mothers’ prenatal breastfeeding efficacy, the lower their depression was at 9 months postpartum, b = −0.63, SE = .25, p = .013. Next, we examined whether 1 month postpartum breastfeeding efficacy was related to depression at 9 months postpartum (controlling for depression during pregnancy) in a separate model. Breastfeeding efficacy at 1 month postpartum and depression during pregnancy accounted for 18% of the variance in depression at 9 months postpartum, R² = .18, F(2, 44) = 4.78, p = .013. However, breastfeeding efficacy at 1 month postpartum did not significantly predict depression at 9 months postpartum. Finally, to examine the unique associations of prenatal versus postnatal efficacy and depression, we entered both variables in the same model. Prenatal breastfeeding efficacy, breastfeeding efficacy, and depression at 1 month postpartum accounted for a significant amount of variance in depression at 9 months postpartum, R² = .23, F(3, 43) = 4.36, p = .009. Although only marginally significant, only prenatal breastfeeding efficacy, b = −0.50, SE = .28, p = .09, was negatively related to depression at 9 months postpartum, such that greater prenatal breastfeeding efficacy predicted lower depression. Breastfeeding efficacy at 1 month postpartum was not significantly related to depression at 9 months postpartum, b = 0.07, SE = .23, p = .75.

Breastfeeding Duration as a Mediator

Next, we tested whether breastfeeding duration was an explanatory factor (or mediator) in the relation between breastfeeding efficacy and depression at 9 months postpartum. Given our results above, we only examined prenatal breastfeeding efficacy as a predictor in the mediational model. Table 3 displays the results of the initial mediation model. Together, the variables in the proposed mediation model accounted for 20% of the total variance in depression at 9 months postpartum, R² = .20, F(3, 54) = 4.61, p = .002. The direct effect (c path) between breastfeeding efficacy and depression was significant, b = −0.58, SE = .28, p = .04, such that, at pregnancy, the more efficacious mothers expected to be in their ability to breastfeed, the fewer depressive symptoms they reported at 9 months postpartum (controlling for depression at pregnancy). In accordance with Aim 2, prenatal breastfeeding efficacy was also positively related to breastfeeding duration, a path; b = 0.37, SE = .14, p ≤ .008, indicating that feeling more efficacious at pregnancy about their ability to breastfeed was related to a longer duration of breastfeeding later in the postpartum period (refer to Table 3). However, the direct effect from breastfeeding duration to depression at 9 months postpartum (b path) was nonsignificant, indicating a longer duration of breastfeeding was not related to lower depression at 9 months postpartum. Additionally, the indirect effect of breastfeeding duration on the relation between breastfeeding efficacy at pregnancy and depression at 9 months postpartum was nonsignificant, b = −0.03, SE = .10, 95% CI [−0.24, 0.17], p = .79. Thus, contrary to our prediction for Aim 2, length of breastfeeding duration did not mediate the relation between prenatal breastfeeding efficacy and depression at 9 months postpartum for new mothers.

Table 3.
Regression Results for Initial Mediation Model.

Direct Effects

b SE t p

Breastfeeding efficacy on depression (c path) –0.58 0.28 –2.09 .04

Breastfeeding efficacy on duration (a path) 0.37 0.14 2.75 .01

Duration on depression (b path) –0.07 0.26 –0.26 .79

Bootstrap results for indirect effect

Effect SE LL 95% CI UL 95% CI

Indirect effect of duration on efficacy and depression (c’ path) –0.10 0.10 –0.24 0.17

Note. Participants (N = 61). LL = lower limit; UL = upper limit.

Test of Moderated Mediation

Our final research aim (Aim 3) was to examine whether family–work conflict moderates the overall mediating process of breastfeeding efficacy on depression by way of breastfeeding duration. Using the MODMED macro, we separately tested FWC and WFC as moderators of the mediation between prenatal breastfeeding efficacy, duration, and 9 months postpartum depression (controlling for depression at pregnancy). According to the MODMED analysis, FWC moderated the indirect effect of duration on the association between breastfeeding efficacy and depression, b = 0.92, SE = .30, p < .001 (see Table 4). We decomposed the interaction and examined the conditional indirect effect of prenatal breastfeeding efficacy on 9 months postpartum depression via breastfeeding duration at three values of the centered variable, FWC: the mean (0), 1 SD above the mean (0.75), and 1 SD below the mean (−0.75). A decomposition of the interaction using normal theory tests revealed that only low levels (−1 SD of the mean) of FWC significantly moderated the process of prenatal efficacy on depression through breastfeeding duration, b = −0.44, SE = .22, p = .047. Bootstrap B_ca confidence intervals confirmed these results, B_ca 95% CI [−1.31, −0.01]. This finding suggests that breastfeeding duration only mediates the association between prenatal breastfeeding efficacy and maternal depression at 9 months postpartum, if mothers had low levels of FWC. For mothers with low FWC, greater prenatal breastfeeding efficacy predicted a longer duration of breastfeeding, b = 0.43, SE = .14, p = .003, and a longer duration of breastfeeding resulted in lower depression at 9 months postpartum, b = −1.02, SE = .38, p = .009 (see Figure 2). Separate analyses with WFC revealed that WFC was not a significant moderator in the mediational model of prenatal breastfeeding efficacy, duration, and 9 months postpartum depression.

Figure 2.
Influence of low family-to-work conflict on indirect effect of breastfeeding duration on breastfeeding efficacy and depression. FWC = family-to-work conflict; **p < .01. p < .05.

Table 4.
Regression Results of Moderated Mediation Model.

Variables b SE t p

Breastfeeding Duration (Direct Effects)

Constant −0.04 0.44 −0.08 .93

Efficacy at pregnancy 0.43 0.14 3.12 .00

Depression at pregnancy (control) 0.06 0.07 0.84 .40

Depression at 9MPP (direct effects)

Constant 10.05 0.81 12.43 .00

Efficacy at pregnancy −0.26 0.28 −0.94 .35

Duration −0.36 0.26 −1.39 .17

FWC at 9 MPP 0.83 1.18 0.70 .49

Duration × FWC 0.92 0.30 3.06 .00

Depression at pregnancy (control) 0.33 0.14 2.40 .02

Conditional effects of FWC on indirect effect of duration on efficacy and depression

Indirect Effect SE z p

−0.75 (−1 SD) −0.44 0.22 −1.97 <.05

0 (Mean) −0.14 0.13 −1.43 .25

0.75 (+1 SD) 0.15 0.14 1.06 .29

Note*. Participants (N = 61). All predictor variables were centered prior to analyses. MPP = months postpartum; FWC = family-to-work conflict.

As a further exploration, we predicted that WFC was not a significant moderator in the indirect effect of duration on the efficacy and depression relation because new mothers might experience greater FWC than WFC at 9 months postpartum. As a post hoc analysis, we conducted a paired samples t-test. As we predicted, at 9 months postpartum, new mothers in our sample had a higher amount of FWC, M = 1.95, SD = .75, than WFC, M = 1.57, SD = .50, t(59) = 3.69, p < .001.

Discussion

New mothers often feel encouraged to breastfeed in order to reap its many benefits for their babies and themselves. However, balancing family and work demands can produce challenges for breastfeeding and may limit the benefits of breastfeeding for working mothers. The current study was the first to examine the role of family–work conflict on breastfeeding and maternal depression in first-time mothers who are employed. We specifically focused on whether prenatal versus postpartum breastfeeding efficacy was more important for maternal postpartum depression as well as whether breastfeeding duration mediated this association. Finally, we tested whether working mothers’ levels of family–work conflict moderated the mediational model. Our results indicated that prenatal breastfeeding efficacy, as opposed to postpartum breastfeeding efficacy, had a stronger association with maternal depression at 9 months postpartum. Moreover, length of breastfeeding duration mediated this relation, but only when mothers reported experiencing low levels of FWC. Below we discuss and interpret these results in light of previous findings.

Prenatal Versus Postnatal Breastfeeding Efficacy on Depression

A majority of prior work has focused solely on postpartum breastfeeding efficacy (e.g., Dai & Dennis, 2003; Dennis & McQueen, 2007, 2009; Haga et al., 2011). We found that prenatal breastfeeding efficacy, rather than 1 month postpartum efficacy, was related to less depression at 9 months postpartum for new mothers. Prenatal breastfeeding efficacy and 1 month postpartum efficacy were only moderately correlated, suggesting there is something unique about the association between prenatal breastfeeding efficacy and 9 months postpartum maternal depression. Whereas 1 month postpartum efficacy is more directly tied to a mother’s actual experiences with breastfeeding, prenatal breastfeeding efficacy may be tapping into a general sense of self-efficacy. In other words, prenatal breastfeeding efficacy may be capturing a trait-like view of a mother’s belief in her general competence or abilities. Another possible explanation is that expected breastfeeding efficacy (controlling for postpartum efficacy) is tapping into met versus unmet expectations. During pregnancy, primiparas mothers have yet to experience breastfeeding, but at 1 month postpartum, mothers are experiencing the actual challenges of breastfeeding, including whether or not they have met their expectations for breastfeeding (Kearney, Cronenwett, & Barrett, 1990). The disappointment or frustration often associated with violated expectations (Miceli & Castelfranchi, 2002) is likely to have a stronger association with depression than actual levels of efficacy alone. Thus, these frustrations may confound efficacy at postpartum; at pregnancy, on the other hand, the measure of efficacy is reflecting a mother’s anticipated beliefs, without the influences of the ongoing challenges. While the current study did not examine the role of violated expectations of breastfeeding, future research should attempt to better understand what exactly it is about breastfeeding efficacy that is influencing maternal depression—that is, trait self-efficacy or violated expectations.

Role of Family and Work Conflict

Our third research aim focused on whether family–work conflict moderated the relation between breastfeeding efficacy, duration, and maternal depression. Although our findings did not support our initial mediational model for working first-time mothers, we found support for our prediction that the relation between breastfeeding efficacy, duration, and maternal depression was conditional on levels of family–work conflict. Specifically, the process and experience of breastfeeding was only beneficial for mothers with low levels of FWC. This finding is in accordance with prior research, suggesting that breastfeeding may not offer mental health benefits for women experiencing chronic, psychosocial stressors (Heinrichs et al., 2002).

Of note, we found that only FWC, not WFC, acted as a moderator in our mediational model. This differential finding that FWC but not WFC moderated the mediation model supports the conceptual distinction between these constructs (Greenhaus & Beutell, 1985). It further suggests that the predictive power of the direction of conflict (WFC vs. FWC) may depend on the importance or salience of a given role for new mothers. Cinamon and Rich (2002) found women tended to endorse the family role more than the work role. However, McKinley and Hyde (2004) found that neither work nor family role salience was a significant predictor of breastfeeding duration for employed mothers—albeit their measures of role salience were assessed prenatally rather than postnatally. Our finding that FWC, rather than WFC, was related to the association between breastfeeding efficacy, duration, and depression is also in line with Cinamon and Rich’s (2002) finding that new mothers may place greater importance on their family roles, particularly after the birth of the baby. In the current study, it is possible that FWC is associated with the breastfeeding process because first-time mothers endorsed higher importance in their family role—as suggested in the post hoc analyses showing new mothers in our sample reported higher FWC than WFC.

Although WFC was not related to the breastfeeding process and postpartum depressive symptoms, it is possible that later on in the postpartum period (i.e., 1 year and beyond), mothers may experience fluctuations in levels of FWC and WFC as family and work demands change, based on the child’s growth and independence (Erickson, Martinengo, & Hill, 2010). In addition, how FWC and WFC each interact with breastfeeding duration may also offer interesting insights for mothers who decide to breastfeed their child for longer than the recommended period of time—as extended breastfeeding could elicit feelings of perceived stigma (e.g., Kendall-Tackett & Sugarman, 1995). Researchers should examine the effects of both directions of domain conflict on breastfeeding and maternal mental health beyond the first year of the transition to parenthood.

Theoretical Implications

The current study has made important contributions to both role strain (e.g., Greenhaus & Beutell, 1985) and role enhancement theory (Marks, 1977). Role strain theory argues that engagement in multiple roles (i.e., being a mother and an employee) can leave one feeling distressed because of the inability or difficulty in meeting the demands of each role with a limited amount of resources, time, and energy (Greenhaus & Beutell, 1985). Our results suggest that for some new mothers, breastfeeding may be perceived as a stressor that contributes to family–work conflict. Specifically, for those with high levels of family–work conflict, breastfeeding while working may be challenging because of the lack of support at work for a place to pump or store milk, or limited time to take breaks in a work schedule to pump milk. The challenges of maintaining breastfeeding during return to employment may lead new mothers to experience a strain between their roles as a parent and worker, resulting in potential distress.

However, for other mothers, breastfeeding may tap into feelings of fulfillment or accomplishment as a parent (Ryan, Todres, & Alexander, 2011) and continuation of breastfeeding during return to employment may especially be perceived as an achievement. Thus, breastfeeding may contribute to family–work facilitation, rather than conflict. Family–work facilitation supports the notion of role enhancement, which suggests that multiple roles and demands can be beneficial because of additional resources for support, transmission of positive mood and skills, and opportunities (Greenhaus & Powell, 2006). Thus, positive aspects of their multiple roles could be related to new mothers’ low FWC in their breastfeeding experiences. Specifically, for working mothers, structural support at work (e.g., a comfortable place and scheduled breaks to pump, available storage space for expressed breast milk) combined with support at home (e.g., encouragement from spouse or family to continue breastfeeding) can all contribute to feeling efficacious about breastfeeding, resulting in a longer duration of breastfeeding and reduced levels of postnatal depression.

Practice Implications

The current study findings have important implications for health practitioners and employers and organizations. Although health practitioners often encourage breastfeeding, most interventions are aimed at new mothers during pregnancy or immediately after hospital discharge (Hannula, Kaunonen, & Tarkka, 2008). Based on the findings from the current study, we suggest that in addition to the first few weeks following the baby’s birth, breastfeeding interventions may be most useful in the prenatal period, given that prenatal breastfeeding efficacy was related to a longer duration and lower postnatal depression for mothers with low family–work conflict. Moreover, because mothers with low family–work conflict experienced a longer duration of breastfeeding and lower depression, prenatal breastfeeding interventions should aim to prepare expectant mothers to manage breastfeeding when returning to work. Specifically, breastfeeding interventions should educate expectant mothers on how to effectively balance their new family demands when returning to their work demands.

In addition, organizations and employers can utilize the study findings. Because study results suggest that new mothers with low family–work conflict benefit most from breastfeeding, organizations can assist in the well-being of their employees by encouraging policies that promote a healthy sense of family–work balance and encourage breastfeeding continuation at work. Policies can include having a generous family leave, allowing new parents adequate time to adjust to their new family roles, and providing support for breastfeeding at work (time and space to pump, space to store milk, etc.). By supporting new mothers in their adjustment to parenthood and continuation of breastfeeding, organizations are likely to benefit as well. Indeed, organizational support for breastfeeding has been associated with less job turnover and reduced health-care costs (Cardenas & Major, 2005).

Limitations and Future Directions

While the findings from the current study offer valuable and interesting insight on the process of breastfeeding and its associations with working mothers’ mental health, the study is not without limitations. First, the small sample size (N = 61) is of concern. Although we utilized bootstrapping to test for moderated mediation, as this test does not depend on large sample sizes (Preacher & Hayes, 2004), we advise future researchers to replicate these findings among a larger a sample for increased power. Moreover, increasing the sample size could also address the issue of the level of significance for our moderated meditation model. Although we found that duration significantly mediated the relation between prenatal breastfeeding efficacy and postnatal depression, the effect is not robust and should be interpreted with caution; the confidence intervals from the B_ca bootstrapped analyses indicate the results bordered on nonsignificance. An increased sample size may provide a more accurate depiction of the proposed model.

Second, the sample for the current study was homogenous, with a majority of mothers being White, highly educated, and upper middle class in income. Furthermore, our preliminary analyses indicated that mothers who were more educated and with a higher income breastfed for a longer period of time—which is in line with the social gradient of breastfeeding. The higher the socioeconomic status (SES) of mothers, the more likely they are to initiate and sustain breastfeeding (Heck, Braveman, Cubbin, Chavez, & Kiely, 2006). Because of the high SES of the study sample, it is unknown if the results would be generalizable to a more diverse population. Specifically, would these results replicate among lower income mothers, as there are distinct differences in breastfeeding and employment conditions based on SES. Lower income mothers experience more family–work conflict (Griggs, Casper, & Eby, 2012), and breastfeeding may be more challenging to them because of stress from their relatively greater financial dependency. Moreover, SES can also determine employment status for new mothers, such as the need or desire to work full-time versus part-time, which could play a role in new mothers’ breastfeeding experiences. For instance, mothers of a higher SES may have more flexibility to work part-time, resulting in additional opportunities to continue breastfeeding (Mandal, Roe, & Fein, 2010). We recommend that future research examine the role of different forms of employment and how it may relate to new mothers’ breastfeeding experiences. In addition to SES, there are cultural differences in breastfeeding practices—women of color are more likely to breastfeed than their White counterparts (Celi, Rich-Edwards, Richardson, Kleinman, & Gillman, 2005). Given that mothers of racial and ethnic minority groups are more likely to breastfeed and are disproportionately lower SES, future research is needed to understand the role of family–work conflict on breastfeeding efficacy, duration, and maternal mental health among a more diverse sample.

Third, the current study did not assess the role of structural factors such as maternity leave or organizational work support. Prior research has shown that a flexible maternity leave and supportive work policies (e.g., having a place and time to pump or store breast milk; McKinley & Hyde, 2004) affect feelings of family–work conflict (Bernas & Major, 2000) and postnatal depression (Klein, Hyde, Essex, & Clark, 1998). We recommend that researchers examine family-related leave policies and organizational work support as potential factors influencing breastfeeding efficacy and duration, WFC, and maternal depression.

Finally, our single-item indicators of FWC and WFC may be limiting. Although prior researchers have utilized single-item measures of FWC and WFC (e.g., Rice, Frone, & McFarlin, 1992; Voyandoff, 1988) and have produced valid results, others have argued that it can lack reliability (Netemeyer, Boles, & McMurrian, 1996). Thus, future research should replicate the findings of the current study using other measures of FWC and WFC, such as the Survey of Work–Home Interaction (Geurts et al., 2005) or the 5-item scale developed by Netemeyer, Boles, and McMurrian (1996).

Conclusions

The current study was the first to examine the conditional mechanism behind the association between prenatal breastfeeding efficacy and depression for first-time mothers as they returned to work. Findings from the current study revealed that FWC plays an important role in the process of breastfeeding and mental health outcomes during the early transition to parenthood. While there are a number of interventions designed to promote breastfeeding, these interventions should also include ways to help first-time mothers balance their home and work lives, as they continue to engage in breastfeeding while returning to the paid workforce. For new mothers who are employed, effectively balancing their home and work demands may help them to experience the full benefits of breastfeeding. From a feminist perspective, although breastfeeding can be beneficial for both mother and baby, it might also threaten the success of mothers in the paid workforce, as they adjust to working in the face of first-time parenting (McKinley & Hyde, 2004). While it is only biologically feasible for women to breastfeed, greater effort to understand how breastfeeding and simultaneously balancing work and home demands is related to maternal well-being can help to achieve greater gender equality in the workplace, and a healthier adjustment during the transition to parenthood.

	M	SD
Age	28.51	3.60
Years in relationship	3.52	1.98
	Frequency	%
Baby’s sex^a
Male	26	42.6
Female	31	50.8
Relationship status
Married	60	98.4
Cohabiting	1	1.6
Race
White	55	90.2
Black	0	0
Hispanic/Latino	2	3.3
Asian	2	3.3
Other	2	3.3
Education
High school diploma	1	1.6
Some college	8	13.1
College degree	33	54.1
Graduate degree	19	31.1
Total household income
<US$20,000	1	1.6
US$20,001–US$40,000	6	9.8
US$40,001–US$60,000	6	9.8
US$60,001–US$80,000	18	29.5
US$80,001–US$100,000	13	21.3
US$100,001–US$120,000	8	13.1
>US$120,000	9	14.8

Variables	Mean	SD	Range	1	2	3	4	5	6	7
1. Breastfeeding efficacy (prenatal)	16.92	3.48	9–25	—
2. Breastfeeding efficacy (1 MPP)	18.02	4.41	10–25	0.29*	—
3. Breastfeeding duration	6.12	3.56	0–9	0.33*	0.54***	—
4. Depression (prenatal)	10.20	6.36	1–35	−0.24	−0.30*	0.02	—
5. Depression (9 MPP)	9.72	7.14	0–31	−0.37**	−0.13	−0.12	0.33***	—
6. Work-to-family conflict (9 MPP)	1.57	0.50	1–2	−0.07	−0.11	0.11	0.05	0.23	—
7. Family-to-work conflict (9 MPP)	1.95	0.75	1–4	0.00	−0.21	−0.08	0.30*	0.12	0.21	—

Direct Effects
Breastfeeding efficacy on depression (c path)	–0.58	0.28	–2.09	.04
Breastfeeding efficacy on duration (a path)	0.37	0.14	2.75	.01
Duration on depression (b path)	–0.07	0.26	–0.26	.79
Bootstrap results for indirect effect
	Effect	SE	LL 95% CI	UL 95% CI
Indirect effect of duration on efficacy and depression (c’ path)	–0.10	0.10	–0.24	0.17

Variables	b	SE	t	p
Constant	−0.04	0.44	−0.08	.93
Efficacy at pregnancy	0.43	0.14	3.12	.00
Depression at pregnancy (control)	0.06	0.07	0.84	.40
	Depression at 9MPP (direct effects)
Constant	10.05	0.81	12.43	.00
Efficacy at pregnancy	−0.26	0.28	−0.94	.35
Duration	−0.36	0.26	−1.39	.17
FWC at 9 MPP	0.83	1.18	0.70	.49
Duration × FWC	0.92	0.30	3.06	.00
Depression at pregnancy (control)	0.33	0.14	2.40	.02
	Conditional effects of FWC on indirect effect of duration on efficacy and depression
	Indirect Effect	SE	z	p
−0.75 (−1 SD)	−0.44	0.22	−1.97	<.05
0 (Mean)	−0.14	0.13	−1.43	.25
0.75 (+1 SD)	0.15	0.14	1.06	.29

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was supported by a grant to Kristin D. Mickelson from the Ohio Board of Regents.

References

Aiken

L. S.

West

S. G.

(1991). Multiple regressions: Testing and interpreting interactions. Thousand Oaks, CA: Sage.

Allen

Hector

(2005). Benefits of breastfeeding. NSW Public Health Bulletin, 16, 42–46. doi:10.1071/NB05011

American Academy of Pediatrics. (2012). Breastfeeding and the use of human milk. Pediatrics, 129, 827–841. doi:10.1542/peds.2011-3552

Avery

A. B.

Magnus

J. H.

(2011). Expectant fathers’ and mothers’ perceptions of breastfeeding and formula feeding: A focus group study in three US cities. Journal of Human Lactation, 27, 147–154. doi:10.1177/0890334410395753

Bai

D. L.

Fong

D. Y. T.

Tarrant

(2015). Previous breastfeeding experience and duration of any and exclusive breastfeeding among multiparous mothers. Birth, 42, 70–77. doi:10.1111/birt.12152

Bernas

K. H.

Major

D. A.

(2000). Contributors to stress resistance: Testing a model of women’s family-work conflict. Psychology of Women Quarterly, 24, 170–178. doi:10.1111/j.1471-6402.2000.tb00198.x

Bianchi

S. M.

Sayer

L. C.

Milkie

M. A.

Robinson

J. P.

(2012). Who did, does, or will do it, and how much does it matter? Social Forces, 91, 55–63. doi:10.1093/sf/sos120

Biehle

S. N.

Mickelson

K. D.

(2011). Worries in expectant parents: Its relation with perinatal well-being and relationship satisfaction. Personal Relationships, 18, 697–713. doi:10.1111/j.1475-6811.2010.01335.x

Blincoe

A. J.

(2005). The health benefits of breastfeeding for mothers. British Journal of Midwifery, 13, 398–401. doi:10.12968/bjom.2005.13.6.18361

10.

Blyth

Creedy

D. K.

Dennis

C. L.

Moyle

Pratt

De Vries

S. M.

(2002). Effect of maternal confidence on breastfeeding duration: An application of breastfeeding self-efficacy theory. Birth, 29, 278–284. doi:10.1046/j.1523-536X.2002.00202.x

11.

Buffardi

L. C.

Smith

J. L.

O’Brien

A. S.

Erdwins

C. J.

(1999). The impact of dependent-care responsibility and gender on work attitudes. Journal of Occupational Health Psychology, 4, 356–367. doi:10.1037/1076-8998.4.4.356

12.

Byron

(2005). A meta-analytic review of work-family conflict and its antecedents. Journal of Vocational Behavior, 67, 169–198. doi:10.1016/j.jvb.2004.08.009

13.

Cardenas

R. A.

Major

D. A.

(2005). Combining employment and breastfeeding: Utilizing a work-family conflict framework to understand obstacles and solutions. Journal of Business and Psychology, 20, 31–51. doi:10.1007/s10869-005-6982-0

14.

Celi

A. C.

Rich-Edwards

J. W.

Richardson

M. K.

Kleinman

K. P.

Gillman

M. W.

(2005). Immigration, race/ethnicity, and social and economic factors of breastfeeding initiation. Archives of Pediatrics and Adolescent Medicine, 159, 255–260. doi:10.1001/archpedi.159.3.255

15.

Centers for Disease Control and Disease Prevention, Division of Nutrition, Physical Activity, and Obesity. (2014, 7). Breastfeeding report card. Retrieved from http://www.cdc.gov/breastfeeding/pdf/2014breastfeedingreportcard.pdf

16.

Cinamon

R. G.

Rich

(2002). Gender differences in the importance of work and family roles: Implications for work-family conflict. Sex Roles, 47, 531–541. doi:10.1023/A:1022021804846

17.

Dai

Dennis

(2003). Translation and validation of the breastfeeding self-efficacy scale into Chinese. The American College of Nurse-Midwives, 48, 350–356. doi:10.1053/S1526-9523(03)00283-6

18.

Davis

M. K.

(2001). Breastfeeding and chronic disease in childhood and adolescence. Pediatric Clinics of North America, 48, 125–141. doi:10.1016/S0031-3955(05)70289-3

19.

Del Mar Alonso-Almeida

(2014). Women (and mothers) in the workforce: Worldwide factors. Women’s Studies International Forum, 44, 164–171. doi:10.1016/j.wsif.2014.01.010

20.

Dennis

C. L.

(2003). The breastfeeding self-efficacy scale: Psychometric assessment of the short form. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 32, 734–744. doi:10.1177/0884217503258459

21.

Dennis

C. L.

Faux

(1999). Development and psychometric testing of the breastfeeding self-efficacy scale. Research in Nursing & Health, 22, 399–409. doi:10.1002/(SICI)1098-240X(199910)

22.

Dennis

C. L.

McQueen

(2007). Does maternal postpartum depressive symptomatology influence infant feeding outcomes? Acta Paediatrica, 96, 590–594. doi:10.1111/j.1651-2227.2007.00184.x

23.

Dennis

C. L.

McQueen

(2009). The relationship between infant-feeding outcomes and postpartum depression: A qualitative systematic review. American Academy of Pediatrics, 123, 734–751. doi:10.1542/peds.2008-1629

24.

Dias

C. C.

Figueiredo

(2015). Breastfeeding and depression: A systematic review of the literature. Journal of Affective Disorders, 171, 142–154. doi:10.1016/j.jad.2014.09.022

25.

Edwards

J. R.

Rothbard

N. P.

(2000). Mechanisms linking work and family: Clarifying the relationship between work and family constructs. Academy of Management Review, 25, 178–199. doi:10.5465/AMR.2000.2791609

26.

Efron

Tibshirani

R. J.

(1993). An introduction to the Bootstrap. Boca Raton, FL: Chapman & Hall/CRC.

27.

Erickson

J. J.

Martinengo

Hill

E. J.

(2010). Putting work and family experiences in context: Differences by family life stage. Human Relations, 63, 977–979. doi:10.1177/0018726709353138

28.

Frone

M. R.

Russell

Cooper

M. L.

(1992). Antecedents and outcomes for work-family conflict: Testing a model of the work-family interface. Journal of Applied Psychology, 77, 65–78. doi:10.1037/0021-9010.77.1.65

29.

Geurts

S. E.

Taris

T. T.

Kompier

M. J.

Dikkers

J. S. E.

van Hooff

M. M.

Kinnunen

U. M.

(2005). Work-home interaction from a work psychological perspective: Development and validation of a new questionnaire, the SWING. Work & Stress, 19, 319–339. doi:10.1080/02678370500410208

30.

Goodman

W. B.

Crouter

A. C.

(2009). Longitudinal associations between maternal work stress, negative work-family spillover, and depressive symptoms. Family Relations, 58, 245–258. doi:10.1111/j.1741-3729.2009.00550.x

31.

Greenhaus

J. H.

Beutell

N. J.

(1985). Sources of conflict between work and family roles. The Academy of Management Review, 10, 76–88. doi:10.5465/AMR.1985.4277352

32.

Greenhaus

J. H.

Powell

G. N.

(2006). When work and family are allies: A theory of work-family enrichment. Academy of Management Review, 31, 72–92. doi:10.5465/AMR.2006.19379625

33.

Griggs

T. L.

Casper

W. J.

Eby

L. T.

(2012). Work, family and community support as predictors of work-family conflict: A study of low-income workers. Journal of Vocational Behavior, 82, 59–68. doi:10.1016/j.jvb.2012.11.006

34.

Groer

M. W.

Davis

M. W.

Hemphill

J. H.

(2001). Postpartum stress: Current concepts and the possible protective role of breastfeeding. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 31, 411–417. doi:10.1111/j.1552-6909.2002.tb00063.x

35.

Haga

S. M.

Ulleberg

Slinning

Kraft

Steen

T. B.

Staff

(2011). A longitudinal study of postpartum depressive symptoms: Multilevel growth curve analyses of emotion regulation strategies, breastfeeding self-efficacy, and social support. Archives of Women’s Mental Health, 15, 175–184. doi:10.1007/s00737-012-0274-2

36.

Hahn-Holbrook

Haselton

M. G.

Dunkel Schetter

Glynn

L. M.

(2013). Does breastfeeding offer protection against maternal depressive symptomology? Archives of Women’s Mental Health, 16, 411–422. doi:10.1007/s00737-013-0348-9

37.

Hannula

Kaunonen

Tarkka

M. T.

(2008). A systemic review of professional support interventions for breastfeeding. Journal of Clinical Nursing, 17, 1132–1143. doi:10.1111/j.1365-2702.2007.02239.x

38.

Heck

K. E.

Braveman

Cubbin

Chavez

G. F.

Kiely

J. L.

(2006). Socioeconomic status and breastfeeding initiation among California mothers. Public Health Reports, 121, 51–59.

39.

Heinrichs

Neumann

Ehlhert

(2002). Lactation and stress: Protective effects of breast-feeding in humans. Stress, 5, 195–203. doi:10.1080/1025389021000010530

40.

Kearney

M. H.

Cronenwett

L. R.

Barrett

J. A.

(1990). Breastfeeding problems in the first week postpartum. Nursing Research, 39, 90–95. doi:10.1097/00006199-199003000-0000510.1111/j.1552-6909.1988.tb00519.x

41.

Kendall-Tackett

K. A.

Sugarman

(1995). The social consequences of long-term breastfeeding. Journal of Human Lactation, 11, 179–183. doi:10.1177/089033449501100316

42.

Klein

M. H.

Hyde

J. S.

Essex

M. J.

Clark

(1998). Maternity leave, role quality, work involvement and mental health one year after delivery. Psychology of Women Quarterly, 22, 239–266. doi:10.1111/j.1471-6402.1998.tb00153.x

43.

Kronborg

Vaeth

(2004). The influence of psychosocial factors on the duration of breastfeeding. Scandinavian Journal of Public Health, 32, 210–216. doi:10.1080/14034940310019218

44.

Leeming

Williamson

Lyttle

Johnson

(2013). Socially sensitive information: Exploring the social context of breastfeeding. Psychology and Health, 28, 450–468. doi:10.1080/08870446.2012.737465

45.

Lindberg

L. D.

(1996). Women’s decisions about breastfeeding and maternal employment. Journal of Marriage and the Family, 58, 239–251. doi:10.2307/353392

46.

Mandal

Roe

B. E.

Fein

S. B.

(2010). The differential effects of full-time and part-time work status on breastfeeding. Health Policy, 97, 79–86. doi:10.1016/j.healthpol.2010.03.006

47.

Marks

S. R.

(1977). Multiple roles and role strain: Some notes on human energy, time and commitment. American Sociological Review, 42, 921–936.

48.

McCarter-Spaulding

D. E.

Kearney

M. H.

(2001). Parenting self-efficacy and perception of insufficient breast milk. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 30, 515–522. doi:10.1111/j.1552-6909.2001.tb01571.x

49.

McDowell

(2006). Measuring health: A guide to rating scales and questionnaires (3rd ed.). Oxford, England: Oxford University Press.

50.

McKinley

N. M.

Hyde

J. S.

(2004). Personal attitudes or structural factors? A contextual analysis of breastfeeding duration. Psychology of Women Quarterly, 28, 388–399. doi:10.1111/j.1471-6402.2004.00156.x

51.

Mezzacappa

E. S.

(2004). Breastfeeding and maternal stress response and health. Nutrition Reviews, 62, 261–268. doi:10.1111/j.1753-4887.2004.tb00050.x

52.

Mezzacappa

E. S.

Katkin

E. S.

(2002). Breast-feeding is associated with reduced perceived stress and negative mood in mothers. Health Psychology, 21, 187–193. doi:10.1037/0278-6133.21.2.187

53.

Miceli

Castelfranchi

(2002). The mind and the future. The (negative) power of expectations. Theory & Psychology, 12, 335–336. doi:10.1177/0959354302012003015

54.

Netemeyer

R. G.

Boles

McMurrian

R. C.

(1996). Development and validation of work-family conflict and family-work conflict scales. Journal of Applied Psychology, 81, 400–410. doi:10.1037/0021-9010.81.4.400

55.

O’Brien

Buikstra

Hegney

(2008). The influence of psychological factors on breastfeeding duration. Journal of Advanced Nursing, 63, 397–408. doi:10.1111/j.1365-2648.2008.04722.x

56.

Oddy

W. H.

Kendall

G. E.

Jacoby

Robinson

de Klerk

N. H.

… Stanley

F. J.

(2010). The long-term effects of breastfeeding on child and adolescent mental health: A pregnancy cohort study followed for 14 years. The Journal of Pediatrics, 156, 568–574. doi:10.1016/j.jpeds.2009.10.020

57.

Ogbuanu

Glover

Probst

Liu

Hussey

(2011). The effect of maternity leave length and time of return to work on breastfeeding. Pediatrics, 127, e1414–e1427. doi:10.1542/peds.2010-0459

58.

Pippins

J. R.

Brawarsky

Jackson

R. A.

Fuentes-Afflick

Haas

J. S.

(2006). Association of breastfeeding with maternal depressive symptoms. Journal of Women’s Health, 15, 754–762. doi:10.1089/jwh.2006.15.754

59.

Preacher

K. R.

Hayes

A. F.

(2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, and Computers, 36, 717–731. doi:10.3758/BF03206553

60.

Preacher

K. J.

Rucker

D. D.

Hayes

A. F.

(2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185–227. doi:10.1080/00273170701341316

61.

Radloff

L. S.

(1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. doi:10.1177/014662167700100306

62.

Redshaw

Henderson

(2012). Learning the hard way: Expectations and experiences of infant feeding support. Birth, 39, 21–29. doi:10.1111/j.1523-536X.2011.00509.x

63.

Rice

R. W.

Frone

M. R.

McFarlin

D. B.

(1992). Work-nonwork conflict and the perceived quality of life. Journal of Organizational Behavior, 13, 155–168. doi:10.1002/job.4030130205

64.

Rondó

P. H. C.

Souza

M. R.

(2007). Maternal distress and intended breastfeeding duration. Journal of Psychosomatic Obstetrics & Gynecology, 28, 55–60. doi:10.1080/01674820600999811

65.

Ryan

Todres

Alexander

(2011). Calling, permission, and fulfillment: The interembodied experience of breastfeeding. Qualitative Health Researcher, 21, 731–742. doi:10.1177/1049732310392591

66.

Semenic

Carmen

Gottlieb

(2008). Predictors of the duration of exclusive breastfeeding among first-time mothers. Research in Nursing & Health, 31, 428–441. doi:10.1002/nur.20275

67.

Sheehan

Schmied

Barclay

(2013). Exploring the process of women’s infant feeding decisions in the early postbirth period. Qualitative Health Research, 23, 989–998. doi:10.1177/1049732313490075

68.

Shrout

P. E.

Bolger

N. B.

(2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7, 422–445. doi:10.1037/1082-989X.7.4.422

69.

Strong

(2013). Barriers to breastfeeding during the neonatal period. Journal of Neonatal Nursing, 19, 134–138. doi:10.1016/j.jnn.2013.04.005

70.

Van Odjik

Kull

Borres

M. P.

Brandtzaeg

Edberg

Hanson

L. A.

… Wille

(2003). Breastfeeding and allergic disease: A multidisciplinary review of the literature (1966-2001) on the mode of early feeding in infancy and its impact on later atopic manifestations, Allergy, 58, 833–843. doi:10.1034/j.1398-9995.2003.00264.x

71.

Voyandoff

(1988). Work role characteristics, family structure demands, and work/family conflict. Journal of Marriage and Family, 50, 749–761. doi:10.2307/352644

72.

Wirihana

L. A.

Bernard

(2012). Women’s perceptions of their healthcare experience when they choose not to breastfeed. Women and Birth, 25, 135–141. doi:10.1016/j.wombi.2011.08.005

73.

World Health Organization. (2009). Infant and young child feeding: Model chapter for textbooks for medical students and allied health professionals. Geneva, Switzerland: WHO Press.

74.

Ystrom

Niegel

Klepp

Vollrath

M. E.

(2008). The impact of maternal negative affectivity and general self-efficacy on breastfeeding: The Norwegian mother and child cohort study. The Journal of Pediatrics, 152, 68–72. doi:10.1016/j.jpeds.2007.06.005

Variables	b	SE	t	p
Variables	Breastfeeding Duration (Direct Effects)
Constant	−0.04	0.44	−0.08	.93
Efficacy at pregnancy	0.43	0.14	3.12	.00
Depression at pregnancy (control)	0.06	0.07	0.84	.40
	Depression at 9MPP (direct effects)
Constant	10.05	0.81	12.43	.00
Efficacy at pregnancy	−0.26	0.28	−0.94	.35
Duration	−0.36	0.26	−1.39	.17
FWC at 9 MPP	0.83	1.18	0.70	.49
Duration × FWC	0.92	0.30	3.06	.00
Depression at pregnancy (control)	0.33	0.14	2.40	.02
	Conditional effects of FWC on indirect effect of duration on efficacy and depression
	Indirect Effect	SE	z	p
−0.75 (−1 SD)	−0.44	0.22	−1.97	<.05
0 (Mean)	−0.14	0.13	−1.43	.25
0.75 (+1 SD)	0.15	0.14	1.06	.29