Abstract
This study used relational turbulence theory to examine (a) how the relational impact of miscarriage corresponded with bereaved mothers’ grief responses and (b) the association between a partner’s supportive conversational behaviors and relationship qualities. The sample included 193 women who had experienced a miscarriage within the previous year. As predicted, relational uncertainty was positively associated with women’s negative appraisals of miscarriage; facilitation from a partner was positively associated with women’s negative emotions; bereaved mothers’ negative appraisals and negative emotional responses covaried; and a partner’s supportive conversational behaviors were positively associated with facilitation from a partner. Contrary to our predictions, interference from a partner was not associated with more intense negative emotions, and a partner’s supportive conversational behaviors were positively associated with relational uncertainty and interference from a partner. The discussion highlights the relational impact of miscarriage on bereaved mothers’ experiences of grief.
Miscarriage, the spontaneous ending of a pregnancy before 20 weeks’ gestation, occurs in approximately 20% of all pregnancies in the U.S. (American Pregnancy Association, 2015). Following a miscarriage, bereaved mothers frequently experience intense grief symptoms (Athey & Spielvogel, 2000). Miscarriage can also negatively influence the couple’s relationship, such that partners often face difficulties with expressing emotions, decreased sexual intimacy, and a decreased sense of togetherness (Salakari et al., 2014). Indeed, the relationship issues that accompany grief can make it challenging for bereaved parents to cope with miscarriage (McGreal et al., 1997). Notably, miscarriage is a devastating experience for both the women and their partner (Holman & Horstman, 2019); however, women are especially likely to report depressive symptoms and grief after a miscarriage (Brier, 2008). Thus, in this study, we focus on the relationship between partners to understand bereaved mothers’ experiences of grief after miscarriage.
Compared to other types of bereavement, individuals coping with miscarriage tend to receive less emotional support from family and friends because miscarriage is not generally talked about (Bute & Brann, 2015). As a result, partners often turn to each other for support, such that they may collectively seek to make sense of their loss (Frost et al., 2007; Horstman & Holman, 2018). At the same time, some women report that miscarriage causes tension in their marital relationship (McGreal et al., 1997). Partners may have different expectations about how to react to the miscarriage, how to behave, what is an appropriate length of time to grieve, and how to disclose their loss to social network members (Bute et al., 2019), which can lead partners to feel disconnected from each other and find it hard to grieve together.
Relational turbulence theory (Solomon et al., 2016) is a theoretical perspective that can provide insight into how a woman’s miscarriage experience is affected by her romantic relationship. In articulating relational turbulence theory, Solomon and colleagues (2016) argued that changes in the relationship environment can amplify relationship qualities that intensify specific experiences. Miscarriage sparks a transitional period for couples because it marks a shift in course for partners anticipating the birth of a child (Brier, 2008), and it is often associated with relational decline (McGreal et al., 1997). Because miscarriage may set in motion relational turbulence processes, we examine women’s experiences of grief following miscarriage through the lens of relational turbulence theory.
The specific aims of this article are twofold. First, we test whether relationship qualities, including relational uncertainty and partner influence, correspond with women’s cognitive and emotional grief responses. Studies have shown that relational issues following miscarriage can exacerbate women’s grief (Brier, 2008; McGreal et al., 1997), and thus, we examine the association between relationship qualities and women’s grief responses to miscarriage. Second, we engage theorizing about conversationally induced reappraisal (Burleson & Goldsmith, 1998) to consider how a partners’ supportive conversational behaviors are associated with relationship qualities following a miscarriage. We evaluate partners’ conversational behaviors because engaging in supportive conversations can help couples improve relational outcomes during stressful transitions (Weigel et al., 2011). We begin by describing the effects of relationship qualities on women’s cognitions and emotions comprising grief responses to miscarriage. Then, we discuss how a partner’s conversational behaviors correspond with women’s perceptions of relationship qualities. Finally, we report and discuss a study that tests our logic.
Relational turbulence theory
Building upon the relational turbulence model (Solomon & Knobloch, 2004), relational turbulence theory was developed to explain people’s cognitive, emotional, and communicative reactions to relationship experiences. The theory emphasizes cognitive appraisals and emotions as indices of the intensity of experiences in close relationships (Solomon et al., 2016). Cognitive appraisals encompass women’s assessment or interpretation of their miscarriage, which influences the extent to which they experience the miscarriage as stressful. Cognitive reactions that follow a miscarriage may include preoccupation with thoughts of the deceased baby, longing for the baby, and negative self-evaluations (Brier, 2008). In terms of emotional responses to miscarriage, bereaved mothers may also experience profound sadness, anger, guilt, and despair (McGreal et al., 1997).
Relational turbulence theory argues that cognitive appraisals and intensified emotions in the face of a relational experience covary (Solomon et al., 2016). In particular, cognitive appraisals about the relational circumstance inform the activation of an emotional response. At the same time, intense emotional responses to specific events shape cognitive perceptions of the situation. Thus, following a miscarriage, bereaved mothers’ negative cognitive appraisals of the experience should be positively correlated with their negative emotional responses. Correspondingly, we propose the following hypothesis:
Considering bereaved mothers’ cognitive and emotional reactions to miscarriage provides insight into women’s experiences of miscarriage. Next, we review how relationship qualities emphasized by relational turbulence theory are associated with bereaved mothers’ cognitive appraisals and emotions about miscarriage.
Relationship qualities and the experience of miscarriage
Relational turbulence theory posits two relationship qualities that shape people’s reactions to specific episodes: relational uncertainty and qualities of interdependence (Solomon et al., 2016). The theory highlights those two relationship qualities as factors that respectively amplify cognitive biases and the intensity of emotional responses to specific experiences.
Relational turbulence theory suggests that people experience more relational uncertainty when there is a change in the relationship environment. Relational uncertainty is defined as the degree of confidence that individuals have in the nature of involvement in a relationship (Knobloch & Solomon, 2002). Relational uncertainty encompasses three interrelated sources of ambiguity: (a) self uncertainty refers to questions people have about their own involvement in the relationship, (b) partner uncertainty entails the questions individuals experience about their partner’s participation in a relationship, and (c) relationship uncertainty includes questions about the status of the relationship itself. Women undergoing fertility issues, such as miscarriage, may experience uncertainty about their relationship (Steuber & Solomon, 2008). For example, miscarriage involves the loss of identity as a pregnant couple, the potential for becoming a parent, and the future as a family (Bute & Brann, 2015; Horstman & Holman, 2018), which may spark women’s doubts about the partnership.
Solomon and colleagues (2016) proposed that relational uncertainty biases cognitive appraisals of specific relational episodes. In particular, people who experience high levels of relational uncertainty lack information to guide the sense-making process, which can lead to biased interpretations of the situation. For example, studies have shown that people perceive hurtful episodes to be more severe (Theiss et al., 2009) and unexpected events to be more distressing (Knobloch & Solomon, 2002) when they have doubts about the relationship. In the context of miscarriage, bereaved mothers’ questions about the nature of their marriage may distort their appraisals of the impact of miscarriage, such that greater relational uncertainty may be associated with more negative thoughts about miscarriage. Thus, we propose the following hypothesis:
Relational turbulence theory denotes the characteristics of interdependence as another relationship quality that influences people’s reactions to specific episodes. Interdependence reflects the dyadic coordination of partners’ capacity to influence, disrupt, and assist each other’s daily activities (Berscheid, 1983). The theory argues that a partner’s influence can be experienced as either disruptive or facilitative (Solomon et al., 2016). Interference from a partner is the extent to which a partner disrupts the other’s ability to accomplish a goal. Facilitation from a partner occurs when a partner’s interruption makes achieving goals or performing activities easier. Miscarriage has the potential to alter the coordinated routines that partners previously enjoyed. For example, following the loss of a child, bereaved parents report that they struggle with having mundane conversations, performing everyday routine, being sexually intimate, and maintaining social relationships (Salakari et al., 2014).
In a departure from previous work on relational turbulence (e.g., Solomon & Knobloch, 2004), relational turbulence theory proposes that both interference and facilitation from a partner amplify both positive and negative emotional reactivity (Solomon et al., 2016). According to the theory, exposure to recent and frequent interruptions to goal-directed behavior, whether helpful or disruptive, tunes emotional reflexes such that emotions evoked by subsequent experiences are stronger than they would otherwise be. By this logic, any change in interdependence or involvement, to the extent that it alters previously nonconscious behavioral routines, also intensifies emotional experiences. These emotional experiences will be more intensely positive when emotional arousal is contextualized by positive cues and more intensively negative when circumstances evoke negative emotions. For example, receiving a welcomed gift from a partner that might generally elicit positive emotions will result in especially strong positive feelings when the gift follows recent and frequent goal interference or facilitation.
In the context of miscarriage, where negative emotions, such as fear, anger, disappointment, guilt, and sadness, are commonly experienced by bereaved mothers (McGreal et al., 1997), the theory would suggest that experiences of interference and facilitation from a partner would correspond with stronger negative emotional experiences. Theiss and Solomon (2006) found that individuals grappling with interference from a partner experience more negative emotions such as anger, sadness, fear, and jealousy. Solomon and Brisini (2018) showed that, for married men, facilitation from a partner was associated with more anger, sadness, and fear. Thus, we predict that interference and facilitation from a partner are associated with intensified, negative emotional responses to miscarriage. Specifically:
Relational turbulence theory specifies that communication between partners can shape relational uncertainty and qualities of interdependence (Solomon et al., 2016). Next, we consider how a partner’s supportive conversational behaviors may be linked to bereaved mothers’ perceptions of relationship qualities following miscarriage.
Partners’ supportive conversational behaviors and relationship qualities
Relational turbulence theory argues that partners may use communication to mitigate relational uncertainty and enhance interdependence (Solomon et al., 2016), but the theory does not specify the types of partner communication that alleviate negative transition experiences (Brisini & Solomon, 2019). In the face of miscarriage, a partner’s supportive conversational behaviors that facilitate talk about the miscarriage may potentially help improve relational outcomes. For example, Horstman and Holman (2018) found that perceptions of spousal communicated perspective-taking following a miscarriage are positively associated with relational well-being. Notably, supportive communication between partners following a miscarriage can be complicated by the fact that both partners have experienced the loss and are grieving. For example, Stroebe et al. (2013) found that following the death of a child, bereaved parents often try to protect each other from further pain and suffering by avoiding talking about the death and remaining strong in each other’s presence; however, their data suggested that holding back from communicating about the loss had a detrimental effect on both partners’ well-being. Although bereaved fathers are a key source of support for their partner following a miscarriage, men also struggle to make sense of the loss (Horstman et al., 2019), and thus, husbands’ ability to provide support in the aftermath of miscarriage may not be representative of their support style during a time when they are not also personally grieving.
To illuminate the association between partners’ supportive conversational behaviors and relational outcomes following a miscarriage, we turned to research on conversationally induced reappraisal (Burleson & Goldsmith, 1998). According to this perspective, engaging in supportive conversations can help distressed people reframe a stressor, which can improve their negative emotional states and instill hope in difficult situations. Burleson and Goldsmith (1998) proposed three conversational conditions that are likely to change people’s appraisals of a stressful situation: creating and sustaining a supportive conversational environment, encouraging disclosure of thoughts and feelings about the upsetting experience, and offering assistance to shift negative appraisals to adaptive ones. One limitation associated with work on conversationally induced reappraisal is that it does not advance specific types of supportive behaviors that meet these optimal conversational conditions. A consideration of research on supportive communication and conversational moves (e.g., Jones & Wirtz, 2006; Kellas & Trees, 2006) suggests four features of partners’ supportive conversational behaviors that are likely to generate adaptive evaluations of stress, as well as positive evaluations of relational outcomes following a miscarriage.
One feature of a partner’s supportive conversational behaviors that encourages bereaved mothers’ disclosure of the miscarriage experience is eliciting thoughts and emotions. A partner’s supportive conversational behaviors, such as problem inquiry and analysis, can generate extended talk about the difficult event and negative emotions (Burleson & Goldsmith, 1998). Focused analysis of a distressed individual’s problem, such as asking for specific clarification of some detail or asking nonspecific questions (e.g., “How does that make you feel?”), allows that person to articulate and elaborate explanations for acts, beliefs, and emotions, which is likely to promote a new way of viewing things or a new way of coping (Feng, 2009). Elliott and colleagues (1985) found that problem exploration in a psychotherapy setting allows clients to realize something new about themselves, to gain cognitive insight, and to see new connections with others.
Second, both nonverbal and verbal listening behaviors function to demonstrate attention, convey responsiveness, and encourage continued expression of thoughts and feelings (Bodie et al., 2015). Nonverbal effective listening typically involves head nods, eye contact, and forward body lean that reflect the degree of closeness with others (Jones & Bodie, 2014). Effective listeners can also engage in a range of verbal behaviors, such as paraphrasing, reflecting feelings, assumption checking, and asking questions to demonstrate responsiveness. These verbal and nonverbal listening behaviors are beneficial because they serve to coordinate interaction and generate rapport, as well as bringing people closer in the face of stress (Bodie et al., 2015).
Another aspect of a partner’s supportive conversational behaviors that may be associated with positive relational outcomes is displaying care. Research suggests that the expressions of affection, care, warmth, and positive regard in times of stress help to create a supportive conversational environment (Goldsmith, 1994) and promote relational health (Pauley et al., 2015). During a supportive interaction, a partner’s invalidation of the other person’s self-identity, such as minimizing or criticizing the other person’s perspectives, may lead to face threat and negative source evaluation (Burleson & Goldsmith, 1998). In contrast, facework that conveys to distressed people that their feelings are recognized and appreciated, that they are valued as persons, and that they are accepted by others is likely to lead to positive relational outcomes (Goldsmith, 1992).
Finally, a partner’s ability to empathize and engage in perspective-taking is likely to overcome bereaved mothers’ reluctance to openly discuss their negative feelings. Wondra and Ellsworth (2015) argued that empathy or emotion sharing leads to understanding and legitimizing others’ emotions. When emotionally empathizing with another person, one does not merely recognize the other’s emotions and also shares these emotions. Through a partner’s confirmation and validation, bereaved mothers may come to realize that they are not alone and that their grief reactions and thoughts are shared by their partner (Dyregrov & Dyregrov, 2008).
Taken together, a supportive partner who is attentive and responsive during a conversation about miscarriage can elicit more detailed disclosures from a bereaved mother (Miller et al., 1983), and that disclosure about the loss can help enhance interpersonal relationships (Altman & Taylor, 1973; Horstman & Holman, 2018). In other words, a partner’s supportive conversational behaviors have the potential to alleviate the deleterious effects of miscarriage on relational uncertainty, as well as help partners reestablish integrated routines. Correspondingly, we propose the following hypothesis:
Method
We tested our hypotheses using self-report data from women who had experienced a miscarriage within the previous 12 months and who were married or in a committed relationship. An online questionnaire was used to elicit the participants’ current relationship and miscarriage experience.
Participants
The sample included 193 women recruited from a variety of sources. 1 Most of the participants (66%) were recruited from the Infertility Family Research Registry, which consists of a pool of volunteers who have faced different infertility issues, including miscarriage. We solicited some participants (27%) from local communities in Omaha and Denver, as well as national organizations that provide perinatal bereavement care (e.g., Star Legacy Foundation). Approximately 6% of our participants were recruited from online support groups on Facebook and Reddit. A small number of participants (less than 1%) were drawn from Amazon’s Mechanical Turk website. 2
Respondents were an average age of 30.54 years (SD = 3.44, range: 19–42). Most participants (97.4%, n = 188) were married, and those marriages were an average of 4.87 years in length (SD = 2.27); participants who were in a domestic partnership (2.6%, n = 5) were together for an average of 2.40 years (SD = 0.55). All the respondents were in a cross-sex relationship, and their male partner was an average age of 32.52 years (SD = 3.39, range: 21–45). The majority of the participants (95.3%) were White. The average time since the miscarriage was 5.95 months (SD = 2.33), and respondents were pregnant for an average of 5.98 weeks (SD = 6.51) when miscarriage occurred. Most of the participants (90.2%) did not know the cause of their miscarriage. The number of children in the participants’ household ranged from 0 to 4 (mode =1, M = 1.03, SD = 0.67), and the majority of the participants (n = 157, 81.3%) reported that they had at least one child, whereas 36 (18.7%) subjects did not have any children.
Procedures
Participants completed an online questionnaire via Qualtrics. After providing informed consent, participants responded to scales concerning their relationship qualities, miscarriage experience, and partners’ supportive conversational behaviors. The survey took approximately 20–40 min to complete, after which participants were thanked for their participation and received small compensation.
Measures
Relational uncertainty
Solomon and Brisini’s (2017) revised measure of relational uncertainty was used to measure self, partner, and relationship uncertainty, and responses were recorded on a 6-point Likert-type scale (1 = strongly disagree, 6 = strongly agree). Participants were asked to think about their relationship with their spouse or partner in the past month. Self uncertainty was measured with 6 items (e.g., “I sometimes wonder whether or not I want my marriage/partnership to work out in the long run”). Partner uncertainty was measured with 6 items (e.g., “I sometimes wonder how much my spouse/partner likes me as a person”). Relationship uncertainty was measured with 6 items (e.g., “I sometimes wonder whether or not my spouse/partner and I will stay together”).
Qualities of interdependence
Participants’ perceptions of characteristics of interdependence were measured using Knobloch and Solomon’s (2004) scale. Interference from a partner was operationalized using 5 items (e.g., “My spouse/partner disrupts my daily routine”). Facilitation from a partner was operationalized using 5 items (e.g., “My spouse/partner helps me in my efforts to make plans”). The two measures solicited responses using a 6-point Likert-type scale (1 = strongly disagree, 6 = strongly agree).
Negative appraisals of miscarriage
The Revised Impact of Miscarriage Scale (Huffman et al., 2014) was used to assess participants’ cognitive appraisals of miscarriage. Participants were asked to evaluate their miscarriage and indicate the degree to which they agree or disagree with 18 statements (e.g., “Through miscarriage, I feel I lost a part of myself”). Responses were solicited using a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree).
Negative emotions about miscarriage
Respondents were asked to think about what types of emotions they felt in the past month as a result of the miscarriage and to rate the degree to which they experienced 21 possible negative emotions (e.g., anxious, sad, worried, angry, frustrated; Dillard et al., 1996; Fowler & Afifi, 2011). Responses were solicited using a 5-point frequency scale (0 = not at all, 4 = very much).
Partners’ supportive conversational behaviors
Respondents were asked to reflect on the conversations they had with their partner about the miscarriage in the past month and evaluate their partner’s conversational behaviors. Items measuring the four features of partners’ supportive conversational behaviors were adapted from existing measures and were presented in random order within the questionnaire. We included all four features of partners’ supportive conversational behaviors in the measure, including elicitation of thoughts and feelings, effective listening, displaying care, and demonstrating empathy, because they are all implied by theorizing about conversationally induced reappraisal and are theoretically meaningful to the context of miscarriage. All responses were solicited using a 7-point scale (1 = strongly disagree, 7 = strongly agree). Elicitation of thoughts and feelings was operationalized using a 6-item scale drawn from Hill and Kellems’ (2002) and Weber and Patterson’s (1996) studies. Effective listening was assessed using Bodie’s (2011) Active-Empathic Listening Scale (11 items). Displaying care was measured using Rogers’ (2007) 5-item scale of unconditional positive regard. Finally, empathy was assessed using Barrett-Lennard Relational Inventory (16 items; Barrett-Lennard, 2015). 3
Relationship satisfaction
Participants’ perceptions of the overall relationship quality were measured using the Marital Quality Index (Norton, 1983). The scale included 5 items (e.g., “We have a good marriage/partnership”) using a 5-point scale (1 = strongly disagree, 5 = strongly agree).
Results
As a first step, we undertook preliminary analyses to identify items comprising our final measures and to assess sources of variance in the data related to contextual variables. Then, we conducted structural equation modeling (SEM) analysis to test all of our hypotheses.
Preliminary analyses
To evaluate the measures included in the study, we conducted exploratory factor analyses (EFAs) using SPSS Version 25 and confirmatory factor analyses (CFAs) using AMOS Version 25. First, we assessed the dimensionality of the two new scales, including the 21-item measure of negative emotions about miscarriage and the 38-item measure of partners’ supportive conversation behaviors. Convergence was achieved using principle axis factoring with direct oblimin oblique rotation. For the Negative Emotions Scale, we removed 6 cross-loading items including anxious, insecure, hurt, envious, irritated, and discouraged. EFA results suggested two factors for negative emotions: unhappy emotions (sad, angry, distressed, upset, scared, nervous, pessimistic, guilty, and unhappy) and concerned emotions (worried, frustrated, fearful, concerned, and jealous).
For partners’ supportive conversational behaviors, we removed 9 cross-loading items (1 in elicitation of thoughts and feelings, 3 in effective listening, 2 in displaying care, and 3 in empathy). EFA results suggested two factors: one contained all the reverse-coded items for empathy and the other one contained the rest of the items for partners’ supportive behaviors. Because one of the factors consisted only of reverse-coded items for empathy, which is not theoretically meaningful, we only used the remaining items to operationalize partners’ supportive conversational behaviors. Thus, only one factor was retained for supportive behaviors.
Second, we examined the measurement properties of all the scales included in the model using CFA. We modeled negative emotions as a second-order construct. We loaded all the items for self uncertainty, partner uncertainty, and relationship uncertainty onto one latent construct because those three relational uncertainty measures were highly correlated with each other in this study (r’s ranged from .84 to .89). 4 All items for each latent construct were included in the measurement model, and we allowed the latent constructs to covary. We consulted modification indices and removed items with the highest score one at a time until a model was identified that met these criteria for a good fitting factor structure: χ2/df < 3.00, comparative fit index (CFI) > .90, and root mean square error of approximation (RMSEA) < .08 (Browne & Cudek, 1993; Kline, 1998). Specifically, 6 items were dropped from the relational uncertainty scale: 3 in self uncertainty, 2 in partner uncertainty, and 1 in relationship uncertainty. Four items were eliminated from the negative cognitive appraisals of miscarriage (e.g., “The miscarriage is a nightmare”). The subscale of concerned emotions, which included 5 items, was removed from the negative emotions scale. The resulting measurement model showed acceptable fit (χ2/df = 1.29, CFI = .91, RMSEA = .04). Table 1 presents descriptive statistics and reliability estimates for these measures.
Descriptive statistics, reliability for substantive variables, and correlations among the variables.
Note. N = 193 individuals. Whether having other children (0 = did not have other children, 1 = did have other children) and knowing the cause of the miscarriage (0 = did not know, 1 = did know) were dichotomous variables.
*p < .05; **p < .01.
Next, we evaluated demographic and contextual variables that might influence women’s perceptions of relationship qualities and their miscarriage experience. Correlations among the variables revealed significant associations between the variables of interest and having other children or not, number of children, time since the miscarriage, length of pregnancy, knowing the cause of the miscarriage, and relationship satisfaction (see Table 1). Consequently, we controlled for these variables in our substantive analyses.
Tests of hypotheses
We used SEM (AMOS Version 25) to test our hypotheses (see Figure 1). All items for each latent construct were included in the SEM. To control for having other children or not, number of children, time since the miscarriage, length of pregnancy, knowing the cause of the miscarriage, and relationship satisfaction, we entered those covariates in the model and constructed paths from those covariates to endogenous variables with which they had significant associations (e.g., paths from the number of children to relational uncertainty and interference from a partner). The fit statistics were χ2/df = 1.39, CFI = .88, RMSEA =.05. Next, we removed all the nonsignificant paths linking the covariates and endogenous variables. Guided by modification indices, we also covaried the disturbance terms for relational uncertainty and interference from a partner; this addition is consistent with the correlation between these variables observed in prior research (e.g., Theiss & Knobloch, 2014). The resulting model had adequate fit to the data (χ2/df = 1.33, CFI = .90, RMSEA =.04).
Final structural model. Note. Model presents standardized regression weights. Indicators of latent variables, covariates, covariances between exogenous variables, and error terms were included in the analysis and are omitted from the figure for the sake of parsimony. The dashed double-headed arrow was added to the structural model guided by modification indices. **p < .01.
Consistent with H1, negative cognitive appraisals of miscarriage were positively correlated with negative emotions about miscarriage (r = .84, p < .01). As predicted by H2, relational uncertainty was positively associated with women’s negative appraisals of miscarriage (β = .24, p < .01). H3 addressed the association between interruptions in the form of either interference or facilitation from a partner and negative emotions about miscarriage. As predicted, results indicated a positive association between facilitation from a partner and negative emotions (β = .62, p < .01). The association between interference and negative emotions was not statistically significant (β = .11, p = .12). Thus, H3 was partially supported.
H4 predicted that partners’ supportive conversational behaviors were negatively associated with relational uncertainty and interference from a partner and positively associated with facilitation from a partner. Contrary to our hypothesis, results indicated that partners’ supportive conversational behaviors were positively associated with relational uncertainty (β = .62, p < .01) and interference from a partner (β = .61, p < .01). Consistent with H4c, we found that partner supportive communication was positively associated with facilitation from a partner (β = .52, p < .01).
Discussion
The goal of this article was to draw upon relational turbulence theory to provide insight into the experiences of bereaved mothers coping with miscarriage. We used an online questionnaire to assess women’s perceptions of relationship qualities, grief responses, and partners’ supportive conversational behaviors. To conclude this article, we consider the implications of our results as well as strengths and limitations of this study.
Theoretical implications
We used relational turbulence theory to understand miscarriage as a nonnormative experience in romantic relationships, which offers insights into the theory’s prediction about more intense relational episodes. Contrary to prior research that has consistently demonstrated low levels of relational uncertainty among people coping with significant transitions in marriage (e.g., Knobloch et al., 2013), we found that relational uncertainty seems to be present or even high after a miscarriage. In fact, we observed that the means of relational uncertainty measures ranged from 4.26 to 4.29 on 6-point scales. This suggests that miscarriage may prompt bereaved mothers to raise questions about the state of their marital relationship and that relational uncertainty can be a useful indicator of assessing relational well-being in the face of transitional events.
Similar to prior research (e.g., Solomon & Brisini, 2018), we found distinctive effects of interference and facilitation from a partner on negative emotions about miscarriage. Specifically, we did not observe a significant association between interference from a partner and negative emotions, whereas facilitation from a partner was positively associated with negative emotions. In keeping with relational turbulence theory, we found that interruptions from a partner, particularly helpful interruptions, are likely to intensify negative emotional reactions to miscarriage. Although speculative, we wonder if positive interruptions from a partner following a miscarriage, such as helping to schedule doctor’s appointments, manage finances, and readjust plans related to the birth of an anticipated child, may make the adverse impact of miscarriage more salient to bereaved mothers. As a result, these helpful interruptions in daily routines inadvertently amplify women’s negative emotions about miscarriage. Although experiences of facilitation can promote positive emotions and thereby offset the arousal of negative emotions (Berscheid, 1983), future research should continue to explore the pattern of facilitation from a partner and emotional responsiveness.
The high correlation (r = .77, p < .01) between facilitation from a partner and partners’ supportive conversational behaviors raises questions about the extent to which facilitation overlaps with perceived supportiveness. We wonder if facilitation from a partner may function as a form of invisible support following a miscarriage. Invisible support involves subtle or skilled provision of support without being recognized by the support receiver, which has shown to elicit positive psychological and relational outcomes (e.g., Bolger & Amarel, 2007). Following a miscarriage, bereaved mothers may be shielded from exposure to household stressors by a partner who works quietly behind the scenes (Conway & Russell, 2000). As Goldsmith (2004) suggested, when partners engage in routine talk, such as making plans, recapping the day’s events, and discussing ordinary events, such relational maintenance behaviors can enhance people’s global evaluations of their partner and the relationship. Thus, bereaved mothers are likely to perceive their partner’s conversational behaviors as supportive, helpful, and understanding when a partner routinely helps them accomplish day-to-day activities.
Finally, following Brisini and Solomon’s (2019) investigation, this study expands the conception of communication beyond the dimensions of valence and engagement emphasized in relational turbulence theory. We proposed four features of partners’ supportive conversational behaviors, including elicitation of thoughts and feelings, effective listening, displaying care, and demonstrating empathy, guided by research on conversationally induced reappraisal (Burleson & Goldsmith, 1998), and we assessed how they are related to each of the relational quality. The unexpected associations between partners’ supportive conversational behaviors and relational uncertainty and interference from a partner suggest that the unintended relational consequences of supportive communication may occur in the face of nonnormative relational stressors.
Previous research suggests that women coping with infertility often desire more emotional support than they receive from their spouse, and such a support gap corresponds with lower support quality (High & Steuber, 2014). Interestingly, our samples reported that their partner was relatively supportive. We wonder if such perceptions of partners’ supportive conversational behaviors may be based on what women expect to receive from their partner, perhaps accounting for their partner’s grief, rather than what they actually desire. Thus, even if women perceive that their partner attempted to be understanding, a partner’s supportive behaviors may be still inadequate to help address women’s concerns about their relationship.
Another explanation for our unexpected results is that women frequently experience a sense of guilt after miscarriage because they feel responsible for the loss of an unborn baby (McGreal et al., 1997). For women whose partner is comforting and supportive in the aftermath of miscarriage, they may feel a greater sense of guilt and disappointment for shattering their partner’s hopes of becoming a parent (Abbound & Liamputtong, 2003). Consequently, women may become particularly attuned to the negative relationship issues that arise following a miscarriage. The positive association between partners’ supportive conversational behaviors and relational uncertainty and interference from a partner is somewhat surprising, and thus, additional research is needed to examine the mechanism through which partner communication affects specific relationship quality during transitional events.
Practical implications
Our study underscores the relational impact of miscarriage that has been documented in prior research (e.g., Bute & Brann, 2015; Horstman & Holman, 2018). The fact that we observed relatively high levels of relational uncertainty and interference from a partner in our samples suggests that miscarriage undermines bereaved mothers’ relational well-being. These observations lead us to wonder if professionals can help women recover from their loss by focusing on the marital relationship, in addition to the miscarriage. Indeed, miscarriage does not only affect women’s psychological and physiological well-being, but it can also pose challenges to the quality of marital relationships (Horstman & Holman, 2018). Thus, by addressing the salient relational consequences of miscarriage, practitioners and counselors may help bereaved mothers cope with miscarriage.
The positive correlation between women’s negative cognitive and emotional responses to miscarriage points to a potential intervention that targets bereaved mothers’ meaning making of the loss. Appraisal-based theories of emotion suggest that changing or modifying one’s interpretation of a stressful event can alter its emotional impact (Lazarus & Folkman, 1984). Correspondingly, emotional regulation through cognitive reappraisal of miscarriage is a viable way of reducing negative emotional responses to this loss (Burleson & Goldsmith, 1998). For example, Holman and Horstman (2019) found that women's perceived stress differed based on the story they told about their miscarriage experience, such that redemptive stories that not only recognized the negative impact of miscarriage, but also reflected an optimistic perspective to their loss, resulted in better psychological well-being. Thus, clinicians can help bereaved mothers assess and reevaluate the impact of miscarriage on their well-being to improve their negative emotions.
Finally, partners should be mindful of the unintended consequences of enacting supportive behaviors during conversations about miscarriage. Supportive conversational behaviors may have negative effects on relational uncertainty and perceptions of interference of daily activities. At the same time, enacting supportive behaviors may contribute to the facilitation of day-to-day activities. Because couples are likely to share their pregnancy with their social network members past the first trimester (Bute & Brann, 2015), we wonder if women who are going through a late pregnancy loss are more likely to receive support from other sources to help process their grief. Indeed, in the face of reproductive challenges, people desire emotional, informational, and tangible support from multiple sources outside the marriage (High & Steuber, 2014). Although speculative, we wonder if receiving support from other social network members, such as those who have also had a pregnancy loss, may be particularly helpful for women who are coping with miscarriage.
Strengths and limitations
This study sheds light on the relational aspects of miscarriage for bereaved mothers. Previous research on miscarriage has largely focused on the detrimental effects of the loss on bereaved individuals’ emotional, psychological, and physiological health (e.g., Athey & Spielvogel, 2000). We provided an account of the relational implications of miscarriage that allows partners and counselors to have a more complete understanding of this type of loss. By recruiting women who have recently experienced a miscarriage, we were able to assess bereaved mothers’ current relationship states and salient grief experiences.
Our claims, however, are tempered by the limitations of this study. First, the cross-sectional data do not permit conclusions about relationship changes that unfold as partners cope with miscarriage. Future research should consider incorporating a longitudinal design that can assess causal effects and document whether the relational consequences of miscarriage dissipate or increase over time. Second, the majority of our participants self-identified as White/Caucasian, which is a limitation to assessing how culture may shape the experience of miscarriage and loss. Future research should explore and clarify the distinct experiences of miscarriage among various ethnic and cultural groups; doing so has value because there are racial differences in rates of miscarriage (Mukherjee et al., 2013) and cultural differences in bereavement about miscarriage (Chalmers & Meyer, 1992).
Third, we did not mandate that the participants’ current partner be the father of the miscarried baby because the current partner is the relevant source of support and focus of perceived relationship qualities. The demographics of our sample, including length of the marriage (M = 4.87 years, SD = 2.27 years) and domestic partnership (M = 2.40 years, SD = 0.55 years) and time since the miscarriage (M = 5.95 months, SD = 2.33), suggest that it is likely that the current partner is also a parent of the miscarried baby. Because we cannot verify this, some participants may be reporting on a different relational partner. This is a limitation because women in these circumstances may be more willing to seek support outside the romantic relationship to manage their grief.
Fourth, for women who are dealing with chronic fertility issues, their miscarriage experience could be different than those who are facing miscarriage for the first time. Women who have never had a successful birth may have a stronger sense of injustice than those who have children (Simmons et al., 2006). Moreover, for couples coping with chronic conception issues, their processing of the experience as part of a collective journey may be different than those who experience miscarriage as a singular event. For example, fertility issues present ongoing challenges to the marital relationship by forcing partners to revisit their family plans, to navigate the constant emotional highs and lows, and to cope with the uncertainty of their fate as biological parents (High & Steuber, 2014). Men and women also differ in how they respond to infertility, such that men are less likely to talk about fertility issues than women, which may breed women’s doubts about their partner’s commitment to the relationship. Future research should consider how the number of miscarriages influences women’s perceptions of the relationship, grief, and partner supportive communication.
Finally, we note the indistinctiveness of facets of relational uncertainty observed in this study. Substantial evidence suggests that the three measures of relational uncertainty are distinct constructs (e.g., Knobloch et al., 2007; Solomon & Brisini, 2017). In addition, self uncertainty and partner uncertainty share variance with relationship uncertainty, but the correlation between them decreases when the variance they share with relationship uncertainty is partialed (Knobloch & Solomon, 1999). The fact that all three facets of relational uncertainty are relatively high in our samples implies that miscarriage can severely affect a relationship. In particular, we wonder whether self uncertainty, partner uncertainty, and relationship uncertainty feed each other more in this context compared to people experiencing normative life stressors.
Conclusion
This study identified the effects of relational consequences of miscarriage on bereaved mothers’ experiences of grief. Our findings suggest that relationship issues that accompany miscarriage, specifically relational uncertainty and facilitation from a partner, respectively, intensify women’s negative cognitive and emotional responses to miscarriage. Importantly, our results suggested that a partner’s supportive conversational behaviors may correspond with more negative evaluations of relational uncertainty and interference from a partner, but a more positive evaluation of facilitation from a partner. Given the pervasive relational impact of miscarriage on bereaved mothers’ well-being, scholars should continue to advance research that aims to understand how partners manage to cope with this devastating loss.
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Open research statement
As part of IARR’s encouragement of open research practices, the authors have provided the following information: This research was not pre-registered. The data and materials used in the research are available. The data and materials can be obtained by emailing: