Printer Friendly

Condemning self, condemning other: blame and mental health in women suffering stillbirth.

Every year around the globe there are more than two million stillbirths, yet stillbirth is generally treated as a non-event, considered less impactful than the death of a life-born child. In up to 60 percent of third-trimester stillbirths, the causes of death were attributed to maternal conditions or were "undetermined." As a result, mothers blame themselves or specific others. This analysis set out to determine how the attitudes of 2,232 bereaved mothers predict their mental health outcomes measuring depressive and anxious symptoms with the Hopkins Symptom Checklist (HSCL). Of the women sampled, 24.6% reported blaming themselves, and 42.3% reported elevated HSCL mean scores. Serf-blame in particular is correlated with symptoms of anxiety and depression. Multivariate analyses predicting elevated HSCL scores demonstrated the importance of time after death, level of education, and reported abuse during pregnancy in the models, as did serf-blame and blaming others. Controlling for other demographic and pregnancy-related variables, self-blame was the strongest predictor of poor mental health outcomes. Implications for mental health counselors are discussed.


Somehow I feel I've failed as a woman. I just didn't get it quite right. Most women come home with a baby after nine months of pregnancy. I came home with a tabletop full of drugs.

Kara Jones, Flash of Life


The effects of child death are traumatic and profound because the nature of the relationship is so intense and enduring (Boyle, Vance, Najman, & Thearle, 1996; DeFrain, Martens, Stork, & Stork, 1990; Vance & Najman, 1995). The effects of traumatic grief may include emotional, physiological,

social, and psychological symptoms ranging from alterations in appetite and sleep patterns to pining and social withdrawal to guilt and shame (Bright, 1991; Condon, 2010; Lovell, 1983; Peppers & Knapp, 1980). Parents also report feeling anger, rage, and inadequacy (Hunfield, Wladimiroff, Verhage, & Passchier, 1995) and often struggle with a profound sense of purposelessness (Boyle et al., 1996). For grieving parents, psychologically this tragedy represents not only the child's death but also, in an existential sense, the death of self as well (Rando, 1985). In the case of stillbirth, just as parents are expecting to welcome a new baby with all attention turned toward the pregnant woman, the baby dies.

Yet stillbirth, the death of a baby after 20 completed gestation weeks but before birth, has long been marginalized as the "invisible death" (DeFrain, Martens, Stork, & Stork, 1986), in part because of its unjustifiable association with abortion and pregnancy loss (Cacciatore, 2008). Concerned researchers and clinicians have asserted that the stillbirth is systemically avoided as a public health issue by governments and the public health system, reproductive rights foes and activists, and even the medical establishment (Cacciatore, 2010; Froen et al., 2011; Kelley, 2011). While most infant and child deaths are socially recognized as traumatic and worthy of mourning, stillbirth is generally treated as a non-event (Cacciatore, 2010; DeFrain et al., 1986; Kirkley-Best & Kellner, 1982; Radestad, Steineck, Nordin, & Sjogren, 1996) that is not as weighty as the death of a live-born child (Froen et al., 2011).


In up to 60% of third-trimester stillbirths, the causes of death were directly attributed to either maternal conditions, such as genetic blood disorders or preeclampsia, or were undetermined even after a thorough postmortem evaluation (Lawn et al., 2011). As a result, mothers often take responsibility for the death (Dunn, Goldenbach, Lasker, & Toedter, 1991), attributing the cause to their "body's failure" in "killing" the baby (Cacciatore, 2010, p. 140). In a global study on stillbirth, Froen et al. (2011) found that healthcare providers in non-Western countries often make erroneous assumptions about the cause, with 29-43% attributing the baby's death to the mother's previous sins, lifestyle or diet, or superstition. Another 20-46% considered the woman to be a failure as a mother and cited the baby's death as "her own fault" (Froen et al., 2011, p. 1356).

Janoff-Bulman (1979) identified both behavioral and characterological self-blame: Behavioral self-blame is more easily controlled or modified; characterological is nonmodifiable and is a greater risk to self-esteem. DeFrain et al. (1990) found that nearly all mothers of stillborn babies report intense behavioral and characterological self-blame following the baby's death:

"It was my fault. I wasn't sure what happened, I just knew it had to be my fault ... I felt guilt over the death because the baby died inside of me. I didn't know of anything I did, but felt I must have made some mistake" ... and (they feel) they helped to kill the baby. (DeFrain et al., 1990, pp. 88-89)

The same study revealed that nearly 30% of bereaved mothers seriously considered suicide after their babies were stillborn, 13% admitted to using substances to cope with the loss, and 62% wanted to "go to sleep and wake up after the pain was gone" (p. 90). Still, few quantitative studies have explored the relationship between self-blame and psychological outcomes after the death of a baby to stillbirth.

Affect Theory and the Crucible of Self-Blame
 According to Kaufman (1996), The essence of the self-blame identity
 script is the repeated accusation of the self for real or imagined
 mishaps ... and humiliation of the self with accusations of fault.
 There is no way to maintain dignity and self-respect: A self-blame
 script remorselessly calls these into question. (p. 103)

The tenets of attribution theory assert that the process of causal determination is facilitative. That is, exogenous blame has been deemed a defensive attribution and self-protective mediator (Shaver, 1970), moderating feelings of helplessness and vulnerability (Taylor, 1983; Wong & Weiner, 1981). Conversely, there is little dispute that self-blame is a danger to mental health, rendering an individual more vulnerable to guilt and shame (Cacciatore, 2010; Tennen & Affleck, 1990). Weinberg (1994) found that dual blame, that is blaming self and other, is more common after traumatic and unnatural deaths, and those who did not report self-blame were better able to cope than those who blamed themselves. Both endogenous (blaming self) and exogenous (blaming others) blame have been recognized as predictors of poor adjustment in bereavement (Field & Bonnano, 2001; Weinberg, 1994). Self-blame, common among bereaved parents, has been shown to predict depressive symptoms (Downey, Silver, & Wortman, 1990) and may give rise to feelings of guilt and shame. Yet differentiating the two affective states may be difficult. Schmader & Lickel (2006) found that guilt and shame conflate, becoming harder to distinguish, when a person engages in self-blame.

According to the Tomkins affect theory (1962, 1963), the underpinnings of optimal mental health are influenced most by nine affective states: joy and excitement are positive, surprise is neutral, and negative are anger or rage, disgust, distress, fear or terror, and shame or humiliation. The expression of each, particularly the negative affects, becomes important for both its recognition and processing. Those specific emotional states, according to Kaufman (1996), are so deeply disturbing to the psyche that "distinct syndromes of shame arise" (p. 114), distorting esteem. Such self-scorning psychopathology includes narcissistic, compulsive, and addictive disorders incited by internalized self-hatred, self-blame, and shame. In particular, psychoanalysts recognize the relationship between these negative emotional states and such psychiatric sequelae as depression and anxiety (Jacoby, 1996; Kaufman, 1996).

Because so little is understood about how self-blame affects maternal mental health after stillbirth, the aim of this analysis was to determine how women's attitudes toward their pregnancy, especially the degrees of wantedness and self-blaming scripts, may have influenced their psychological outcomes. This analysis tested the association between psychological outcomes in terms of maternal demographics, characteristics of the pregnancy, and participants' attribution of blame. It also tested the inter-correlation of different attributions of blame and different psychological outcomes. We predicted that attitudes toward pregnancy and self-blame would both worsen two negative states in bereaved mothers--anxiety and depression--as measured using the 25-item Hopkins Symptom Checklist (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974) to estimate outcomes.



The Maternal Observations and Memories of Stillbirth (MOMStudy) and the website were established to study women's experiences during and after a stillbirth. Participants were recruited into the study by identifying organizations online, informative web pages, and online forums that offer free information on pregnancy and childbirth, including stillbirth. Organizations were sought through 12 Internet search engines and directories using the terms "stillbirth," "pregnancy loss," "fetal loss," and "fetal death." All inflections of these terms were searched where possible. From the first 100 matches in English from each search engine, 749 web pages were identified. Excluded were web pages that only sporadically presented information, such as online scientific journals, dictionaries or databases, and sites for specific scientific, professional, or government use (e.g., American College of Obstetricians and Gynecologists and U.S. National Center for Health Statistics). Ultimately 104 sites or organizations were asked to extend an invitation to women affected by stillbirth to respond to a questionnaire through their websites, mailing lists, newsletters, forums, and other means using ready-made invitations. Of these, 37 accepted.

Data were collected over an 18-month period. Participation and response to the online questionnaire were monitored by an Internet protocol address using The Webalizer[C], freeware from The questionnaire was designed to be interactive in terms of participant responses in order to minimize the number of non-applicable questions. It incorporated topics related to the experience of stillbirth, and repeated questions were used to confirm accuracy. For example, questions about the participant's current age were later compared to others asking her age when experiencing stillbirth and number of years since stillbirth. Cases with inconsistencies beyond obvious, unintended errors were excluded from the sample. For this reason, one case with extreme values was excluded. The study protocol was approved by the Regional Ethics Committee for Medical Research of Southern Norway.

While the questionnaire was online, 3,519 women reported experiencing a stillbirth, and 2,900 completed the entire questionnaire and confirmed their submission, a completion rate of 82.4%. Reported were the stillbirths of 2,769 singletons, 119 twins, and 12 triplets. The 2,332 women who reported a singleton stillbirth after 20 weeks of gestation and were certain about whether they were pregnant were participants in the current study. These were mainly from the United States (71.2%), United Kingdom (10.4%), Australia (8.3%), and Canada (4.4%). Among the babies who died, 52.2% were boys and 47.8% were girls.


The Hopkins Symptom Checklist (HSCL; Derogatis et al., 1974) was used to measure anxiety and depression in the MOMStudy. This 25-item measure has been in use for over 50 years to measure mental distress. The HSCL has been used with samples of adults in medical trials, program evaluations, family practice, and family planning services (Hesbacher, Rickels, Morris, Newman, & Rosenfeld, 1980; Olsen, Mortensen, & Bech, 2006; Parloff, Kelman, & Frank, 1954). Several studies of diverse cultural groups, including samples of pregnant women, have used the HSCL, which produced reliable and valid mental distress scores (Kaaya et al., 2002).

The HSCL includes a global score of mental distress and two subscales. The 10-item anxiety subscale and the 15-item depression subscale each have responses coded 1-4 (1 = "Not at all"; 2 = "A little"; 3 = "Quite a bit"; 4 = "Extremely") with higher scores indicating more mental distress. Anxiety and depression scores, the dependent variables in the current study, are the respondents' mean scores on the HSCL's 10 anxiety and 15 depression items. Any mean symptom scores for anxiety or depression symptoms greater than 1.75 were coded as clinically elevated anxiety or depression and used as dichotomous dependent variables. Several studies have used this cut-off score as indicating mental health concerns (Roberts, Damundu, Lomoro, & Sondorp, 2009; Sandanger et al., 1998; Sandanger et al., 1999; Veijola et al., 2003), and it has been supported in previous research (Mollica, Massagli, & Silove, 2004; Nettelbladt, Hansson, Stefansson, Borgquist, & Nordstrom, 1993). In this study the global scale ([alpha] = .93), the anxiety subscale ([alpha] = .84), and the depression subscale ([alpha] = .91) each demonstrated good to excellent reliability.

Self-blame. Respondents in the MOMStudy were asked, first, if they felt blame toward someone for the baby's death and then who they thought was to blame. The options for the placement of blame were prenatal care provider, labor unit provider, the father of the child, myself, others, God, no blame, and do not know. In the current study, blame placed on either the prenatal care provider or the labor unit provider was coded as blaming a health care provider. Blaming the father, others, or God were coded as blaming others.

Control variables. The current study incorporates several control variables that account for differences among respondents that may be associated with the anxiety or depression scores they reported. Years since the stillbirth and the respondent's age at the time of the stillbirth were both recorded as continuous measures. Respondents were asked about their education (coded as less than college = 0, at least some college = 1) and urbanieity (rural = 0, urban = 1, large urban = 2). The ethnicity reported was coded as a binary variable (white = 0, nonwhite = 1). Responses to the question whether respondents had been victims of physical or psychological abuse during the pregnancy were coded as binary variables (no = 0, yes = 1). Whether the participant reported wanting the pregnancy that had resulted in the stillbirth and whether she reported being pregnant when the questionnaire was completed were coded as binary variables (for both, no = 0, yes = 1). Other demographics, such as residence in the United States, marital status, gestational age at the time of stillbirth, living arrangements during pregnancy, whether the stillbirth occurred during the third trimester, and whether the mother attended a prenatal care program or ultrasound screenings were also collected (for all: no = 0, yes = 1).

Data Analysis

The data were analyzed using SPSS 18.0. First, bivariate logistic regression models were conducted for each of the dependent and control variables with each of the outcome variables (the HSCL global, anxiety, and depression mean scores). Logistic regression predicts the likelihood of group membership in a dichotomous outcome and was selected to predict the likelihood of high scores on the three HSCL scales. All variables within the bivariate logistic regression models were then introduced into a stepwise multivariate logistic regression model for each of the three HSCL outcome scores. During this process, nonsignificant variables (p > 0.5) were removed and the next variable was added to the model in order to reach a final model for each of the three HSCL scores. An exploratory process using step-wise multiple regression was used to assess the strength of the predictors and to develop models predicting HSCL scores. Because the sample was obtained through nonrandom procedures, all p-values obtained through parametric statistics should be interpreted with caution.


The 2,332 women who reported a singleton stillbirth after 20 weeks of gestation were on average 28 years old when the baby died and participated in the study within about three years (Table 1). Most of the women were from urban or large urban settings. At the time of the death, the vast majority of the women reported that they were married or cohabiting with a partner. The majority reported at least a college education, were primarily white, and were not pregnant at the time of the study. Most (82.1%) also reported wanting the pregnancy, attending prenatal care, receiving ultrasound screenings, and having a third-trimester stillbirth; the average gestational age at the time of the baby's death was 34 weeks.

Self-Blame and Blame Variables

Of the total sample, 42.1% of the women reported blaming at least one source for the death; 25% blamed medical providers, 11.1% blamed others (father, others, God), and 24.6% blamed themselves. More than 27% of respondents blamed one source, 11% at least two sources, and 4% all three possible sources.

Blaming one of the sources was associated with blaming other sources as found in a series of Chi-square analyses: Blaming self was significantly associated with blaming health care providers ([chi square] = 114.78, df = 1, p < .001) and blaming others ([chi square] = 291.97, df = 1, p < .001). Blaming others was significantly associated with blaming health care providers ([chi square] = 38.13, df = 1, p < .001). Those women who reported wanting the pregnancy were more likely to blame themselves for the loss of the child (Wald = 4.358, df = 1, p = .037)--wanting the pregnancy was associated with a 31% increase in the likelihood of also reporting self-blame.

Depression and Anxiety Variables

As measured by the HSCL total score, 42.3% of the sample had elevated mental distress scores. Mean scores on the depression subscale indicated that a majority of the sample had a clinically elevated score (Table 2). Respondents reported clinically elevated scores (> 1.75 mean score) on each subscale of the HSCL, with 41.3% reporting clinically elevated anxiety and 61.7% reporting depressive symptoms. All three HSCL scores demonstrated significant inter-correlations (Table 2).

Bivariate Models and Variables

The association between the predictors and reporting clinically elevated scores on the HSCL and each subscale was tested first by logistic regression. Each predictor was introduced individually (Table 3). Maternal age at the time of stillbirth and years elapsed since the stillbirth were both significantly associated with elevated total, depression, and anxiety scores. Each additional year of age at the time of the stillbirth was associated with a 2-4% drop in the likelihood of reporting elevated scores. Similarly, each additional year since the stillbirth was associated with about a 6-8% drop in the likelihood of reporting elevated scores.

Less education was significantly associated with elevated scores; those reporting at least a college education were 40-47% less likely to have elevated scores. Those with less education were also more likely to have higher scores across the HSCL and subscales. Ethnicity was not related to elevated scores. Urbanicity was significantly associated with the likelihood of elevated total and anxiety scores. Residence in an urban environment was associated with a 20-21% lower likelihood of elevated scores than residence in a rural area.

Variables surrounding pregnancy and experiences during pregnancy were significantly associated with elevated scores. Women reporting that they wanted the pregnancy that resulted in a stillbirth were significantly more likely to have elevated scores on the HSCL and subscales. Wanting the pregnancy was associated with a 43% increased likelihood of an elevated total score, a 39% increase in the likelihood of reporting depression, and a 35% increase in the likelihood of reporting anxiety. Women who reported experiencing domestic abuse during their pregnancies were more than twice as likely to have elevated HCSL total and depression scores, and abuse was also associated with a 73% increase in the likelihood of reporting anxiety. Being pregnant during participation in the study was associated only with elevated depression scores; these women were 1.5 times more likely to report an elevated score.

Women blaming health care providers for the stillbirth were about 1.5 times more likely to report higher scores on the HSCL total and subscale scores. Women blaming others were about 3.2 times more likely to have an elevated total score, 3.5 times more likely to report depressive symptoms, and 2.3 times more likely to report anxiety. Finally, blaming themselves increased the likelihood of elevated HSGL total scores by 3 times, elevated depression scores by 3.9 times, and anxiety scores by 2.2 times.

Multivariate Models

A step-wise process of entering each variable into a model was used to produce final multivariate models predicting HSCL total and subscale scores. In each case, the variables were entered starting with maternal age and ending with three blame variables as the predictors of interest (Table 4). A single variable was entered into the model, its ability to predict the outcomes was tested, then it was either retained or dropped from the model. This process was repeated for each variable predicting each of the three HSCL outcome scores.

The three resulting models significantly predicted the HSCL total and subscales scores. The step-wise introduction of variables in each model resulted in the same model predicting the HSCL total score ([chi square] = 254.86, df = 8, p < .001) and HSCL anxiety score ([chi square] = 165.81, df = 8, p < .001). The model predicting anxiety scores explained approximately 10% of the variance in the outcome (Nagelkerke [R.sup.2] = .097) where these variables predicted approximately 15% of the variance in the HSCL total scores.

The model predicting HSCL depression scores differed from the other two models. The step-wise process resulted in six variables significantly predicting likelihood of elevated depression ([chi square] = 255.96, df = 6, p < .001) and explained about 15% of the variance in the outcome (Nagelkerke [R.sup.2] = .147).

The final multivariate models produced are shown in Table 3. Ethnicity and whether the pregnancy was wanted were not significant predictors when added to each multivariate model after controlling for the significant predictors added in prior iterations. Whether the participant was pregnant while participating was significant only in the model predicting the depression score. Maternal age at the time of the stillbirth and blaming health care providers for the loss were found not to be significant predictors in the depression model but were significant for the total and anxiety scores. For each model, women who reported blaming themselves for the stillbirth were 2.7 times more likely to have higher depression scores, 1.8 times as likely to report anxiety, and almost 2.5 times as likely to have an elevated total score.


Trait guilt--an individual's personality threshold and the regulation of emotional and behavioral responses to guilt--is an enduring and dynamic aspect of long-term major mood disorders (Ghatavi, Nicolson, MacDonald, Osher, & Levitt, 2002). According to Kaufman (1996), shame, a form of inner torment, specifically is "central to conscience, indignity, identity, and disturbances in self-functioning" (p. 5), posing a major threat to self-esteem and having a binding effect on the development and maintenance of the self (Tomkins, 1963). Whereas guilt is about real or perceived transgressions, shame is about self-blame and failure that can lead to self-contempt, self-anger, increased distress, worthlessness, powerlessness, and fear and anxiety (Kaufman, 1996). Barr (2004) posited that being prone to guilt and shame is responsible for a substantial variance in maternal grief after the death of a baby, and this state may be predicted best by self-esteem (Jacoby, 1996). Mothers of stillborn infants appear to be particularly vulnerable to these negative affect states.

While age, education, and years since stillbirth were shown to have significant relationships to anxiety and depressive symptoms, self-blame is highly significant and positively related to both. This suggests that self-blame has a central role in the manifestation of mental distress among women whose baby was stillborn, whether or not the baby was planned or wanted. When controlling for other demographic and pregnancy-related variables, self-blame was the strongest predictor of the HSGL scores these women reported.

Psychological and physical abuse during pregnancy also predicted elevated scores in the model; however, even when controlling for abuse, blaming self and blaming others remained significant in the three multivariate models. The contribution of abuse in an intimate partner relationship and its relationship to blaming self and others are complex and may partially mediate those variables.

The results of the logistic regression models (see Table 3) also indicate that age, education, and wantedness in addition to self-blame all have direct effects on anxiety and depression in the bivariate models. Younger, less educated women who wanted the pregnancy were less likely to report elevated scores. These results were not consistent when added to the multivariate models. Age was predictive of elevated anxiety and total scores, yet not elevated depression scores. Wanting the pregnancy was not predictive when added to the multivariate models, and ethnicity was not a significant variable in any of the analyses. Further study is needed with a more diverse sample to explore differences in blame and psychological outcomes. Blame is a deeply personal and social experience that manifests differently in different ethnic and cultural groups.

It is necessary to pay attention to the differences between the models for anxiety and depression. Though the step-wise processes predicting elevated anxiety and total HSGL scores produced similar models, in the step-wise multivariate process, modeling variables differed. While maternal age and blaming health care providers were significant predictors of depression in the bivariate models, that did not hold true when controlling for other variables. Being pregnant during the study remained a significant predictor in the multivariate model even after introducing the controls. For these women, and possibly other women who have experienced the death of a child in this way, blaming may contribute differently to the development of anxiety and depression. No conclusion about possible differences in this process can be drawn from this inquiry; further research into blaming scripts could be essential to forming and guiding evidence-based practice.


While having the advantage of a very large number of participants, as was needed to include significant confounders in the analyses, this study is cross-sectional and correlational. Bias may have been introduced not only by dependence on informed consent from participants but also by a skewed recruitment of participants in comparison to the total population. Participants represent younger and higher socioeconomic groups (Internet access), and our adjustments for age and education will not fully correct for this fact. A large proportion of the women were recruited while actively searching on the Internet for support or information about stillbirths. Participant self-selection into the study may be correlated with characterologieal traits that may influence the woman's psychological response to having a stillborn baby. Methodologically, the selection effect may thus compromise the generalizability of the outcomes. However, it is also possible that mental health providers are more likely to encounter help-seeking clients like the women who participated in this research. Finally, the fact that the study is quantitative limits the richness of the data when the topic is so nuanced and complex, failing to sufficiently capture the subjective experience.


The experience of stillbirth has only recently been recognized as a serious public health problem (Cacciatore, 2010; Froen et al., 2011). It is possible that the long history of societal mispereeptions about stillbirth have exacerbated private feelings of inferiority, impotence, failure, and scripts of self-blame, exposing bereaved mothers to the risk of maladaptive adjustment and psychopathology. The outcome of such nonrecognition and devaluing may well be profound feelings of guilt and shame (Cacciatore, 2010; Jacoby, 1996; Kirkley-Best & Kellner, 1982). Guilt and shame, closely associated with self-blame (Schmader & Lickel, 2006), incite feelings of self-reproach (Kaufman, 1996), and chronic guilt contributes most to the variance in grief responses (Barr & Cacciatore, 2007). At an already vulnerable time in the aftermath of a baby's death, these emotional states can pose great psychological risk to grieving mothers and their families. In addition, just under 11% (n = 255) of the mothers in this sample reported abuse during the pregnancy. It is important for counselors to recognize the complex, interrelated variables associated with blame under these circumstances.

In affect theory, the discrete categories of emotional bases are the wellspring of optimal mental health, emphasizing the positive states of affect while noting a particular imperative to approach, explore, and minimize the negative (Tomkins, 1962, 1963). Bereaved mothers who were avoidant and who deflected their emotional states were inclined to protracted mental health problems (LaRoche et al., 1982). Thus, it is critical for medical and mental health providers to (a) be aware of the role of self-blame, shame, and guilt on negative psychological outcomes for mothers grieving a stillbirth; (b) be willing to approach and tolerate enduring discussions of negative affective states when the mother is ready; (c) be mindful of how they are communicating, choosing words carefully and avoiding any shaming language; (d) provide a safe environment in which shame and guilt may be expressed and received nonjudgmentally; (e) be prepared to offer creative interventions that aid and approach, rather than impede and avoid, adjustment; and (f) be aware of the increased mental health risk during a subsequent pregnancy.

A willingness to approach painful emotional states has been associated with reconciliation, while avoidance tends toward inaction (Schmader & Lickel, 2006). Kurtz (2007) emphasized the importance of recognizing shame and then transforming it into positive shame in order that clients might "live constructively within their essential limitation" (p. 22). In order to do this, however, a person must come from behind the mask of shame.

Middelton-Moz (1990) referred to shame as the master of disguise, provoking an insidious cycle of reshaming. Thus, providers will need to watch for its markings through communication (nonverbal as well as verbal), family history, and careful assessment. Acknowledging the existence of shame in the therapeutic relationship, to both the self and a safe other, is the first step toward healing it. The process of unmasking shame, though painful, is necessary; because it is a subjective emotion, it must be treated subjectively (Kurtz, 2007).

Kurtz (2007) proposed principles for healing shame: (a) acknowledge shame and become conscious of its presence; (b) help clients to recognize the need for others; (c) help clients feel safe enough to discuss shame with transparency and without reshaming; and (d) find meaning and purposefulness without the intense and enduring need to focus on the self. Kaufman (1996) recommended a shame profile that explores affect-shame binds, the internalized linkages that trigger shame when various emotional states are experienced, connecting the person's history of shaming, even back to childhood, and reliving governing scenes. These become repaired by paying attention and increasing awareness of "clusters of shame nuclei" (p. 27), patterns of the internal linkages between emotions and shame. Counseling can provide a space where mindfulness or awareness activities, such as introspective journaling, can occur. This process leads to an evolving language that sufficiently demarcates emotional states, drives, needs, purpose, and function and unearths internalized patterns of self-sabotage.

Of particular import is the commonality of enduring distress in this sample, suggesting that perhaps strong emotional reactions including shame and guilt, though difficult for the individual, may not be abnormal. Mindfulness-based therapies may help to heal shame because they underscore the need for acceptance, attunement, and trust in the therapeutic relationship and may facilitate a feeling of safety for the client (Cacciatore & Flint, 2012). These types of interventions tend to diverge from the medicalized model of grief in accepting that some degree of negative affect and expression may be a normal part of adjusting to the traumatic death of a loved one rather than representing psychopathology. This type of approach is more acceptance-based and egalitarian. Kaufman (1996) acknowledged the therapeutic benefits of shared power in healing shame. Mindfulness-based approaches are also more integrative and thus therapy can become a catalyst for the creation of new scripts and scenes, an integral force in healing shame.

Once clients come out from behind the mask of shame, as Kurtz (2007) suggested, they will be able to take more positive action. For example, a mother who is experiencing deep shame and guilt over the death of her baby may want to compose a letter of apology to the baby who died. In this letter, she can state and restate her transgressions, become increasingly conscious of the internal scripts, and eventually ask for her child's forgiveness. After several hours of contemplation, she can then write another letter as if she were hearing her child's voice of forgiveness rather than condemnation.

Reconciliatory actions, such as volunteering, ritual or ceremony, random acts of kindness, and the example of an apology letter may mediate self-blame, guilt, and shame, redirecting them into more useful states of mind, such as self-forgiveness, self-compassion, meaning-making, and self-affirmation (Cacciatore & Flint, 2012). Self-affirmation translates to improvements in esteem, respect, and pride in the self. These creative strategies, which require conscious and intentional living, may be a means through which clients become more aware of and channel or absolve negative self-scripts. They may also provide some relief from the "governing scenes of negative affect" that underlie "addictive, compulsive, and narcissistic" behaviors and significant issues of self-esteem and self-worth by reclaiming parts of the self that were formerly disowned (Kaufman, 1996, p. 110).

The process of mourning a dead baby may be more enduring than some clinicians might realize. Women suffer these losses as a threat to the self. They may wonder about their role in the baby's death and be plagued with self-doubt. They may wonder about their worthiness as mothers to surviving or subsequent children and about their relationship to partners (Cacciatore, 2010). During the process of mourning, a mother not only grapples with the physical absence of the baby but also with its social nonexistence, inciting an existential crisis that questions her identity as a mother and incites shame and guilt over her failure to give birth to a healthy, living baby (Cacciatore, 2008). The movement to understand and destigmatize stillbirth, as well as toward a model of dignified and respectful psychosocial care, is most certainly a worthwhile endeavor that may improve outcomes for mothers, children, and families. It also holds far broader implications for communities and society. Indeed, a world where women can transition from the condemning to the compassionate self would undoubtedly be an improvement for many.

Acknowledgement: The study received financial support from the Norwegian Society for Unexpected Infant Death, Oslo, Norway. The authors declare no competing interests


Barr, P. (2004). Guilt- and shame-proneness and the grief of perinatal bereavement. Psychology Psychotherapy: Theory, Research & Practice, 77, 493-510.

Barr, P., & Cacciatore, J. (2007). Problematic emotions and maternal grief. Omega Journal of Death and Dying, 56, 329-346.

Boyle, F. M., Vance, J. C., Najman, J. M., & Thearle, M. J. (1996). The mental health impact of stillbirth, neonatal death or SIDS: Prevalence and patterns of distress among mothers. Social Science & Medicine, 43, 1273-1282.

Bright, D. A. (1991). Stillbirth. Journal of Family Practice, 32, 245-253.

Cacciatore, J. (2008). Stillbirth: A sociopolitical issue. Affilia, 23, 378-387.

Cacciatore, J. (2010). Unique stories of women and their families after the death of a baby. Journal of Healthcare Social Work, 49, 134-148.

Cacciatore, J., & Flint, M. (2012). ATTEND: A Mindfulness-based bereavement care model. Death Studies, 36, 61-82.

Condon, I. (2010). Women's mental health: A wish-list for the DSM V. Archives of Women's Mental Health, 13, 5-10.

DeFrain, J., Martens, L., Stork, J., & Stork, W. (1990). The psychological effects of a stillbirth on surviving family members. Omega, 22, 81-108.

DeFrain, J., Martens, L., Stork, J., & Stork, W. (1986). Stillborn: The invisible death. Lexington, MA: Lexington Books.

Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins Symptom Checklist (HSCL): A self-report inventory. Behavioral Science, 19, 1-15.

Downey, G., Silver, R. C., & Wortman, C. (1990). Reconsidering the attribution adjustment relation following a major negative life event: Coping with the loss of a child. Journal of Personality and Social Psychology, 59, 925-940.

Dunn, D. S., Goldenbach, K., Lasker, J., & Toedter, L. J. (1991). Explaining pregnancy loss: Parents' and physicians' attributions. Omega, 23, 13-23.

Field, N. P., & Bonanno, G.A. (2001). The role of blame in adaptation in the first five years following the death of a spouse. American Behavioral Scientist, 44, 764-781.

Froen, E, Cacciatore, J., McClure, E., Kuti, O., Hakeem Jokhio, A., Islam, M., & Shiffman, J. (2011). Still births: Why they matter. The Lancet, 377, 1353-1366.

Ghatavi, K., Nicolson, R., MacDonald, C., Osher, S., & Levitt, A. (2002). Defining guilt in depression: A comparison of subjects with major depression, chronic medical illness and healthy controls. Journal of Affective Disorders, 68, 307-315.

Hesbacher, E T., Rickels, K., Morris, R. J., Newman, H., & Rosenfeld, H. (1980). Psychiatric illness in family practice. Journal of Clinical Psychiatry, 41, 6-10.

Hunfield, J., Wladimiroff, J., Verhage, F., & Passchier, J. (1995). Previous stress and acute psychological defense as predictors of perinatal grief. Social Science and Medicine, 40, 829-835.

Jacoby, M. (1996). Shame and the origins of self-esteem. New York, NY: Routledge.

Janoff-Bulman, R. (1979). Characterological versus behavioral self-blame. Inquiries into depression and rape. Journal of Personality and Social Psychology, 37, 1798-1809.

Jones, K. (2002). Flash of life (2nd ed.). Vashon, WA: Kota Press. Kaaya, S. E, Fawzi, M. C. S., Mbwambo, J. K., Lee, B., Msamanga, G. I., & Fawzi, W. (2002). Validity of the Hopkins Symptom Checklist-25 amongst HIV-positive pregnant women in Tanzania. Acta Psychiatrica Scandinavica, 106, 9-19.

Kaufman, G. (1996). The psychology of shame (2nd ed.). New York, NY: Springer Publications.

Kelley, M. (2011). Counting stillbirths: Women's health and reproductive rights. The Lancet, 377, 1313-1314.

Kirkley-Best, E., & Kellner, K R. (1982). The forgotten grief: A review of the psychology of stillbirth. American Journal of Orthopsyehiatry, 52, 420-429.

Kurtz, E. (2007). Shame and guilt (2nd ed.). New York, NY: iUniverse.

LaRoche, C., Lalinec-Michaud, M., Engelsmann, F., Fuller, N., Copp, M., McQuade-Soldatos, P., Azima, R. (1984). Grief reactions to perinatal death--a follow up study. Canadian Journal of Psychiatry, 29, 14-19.

Lawn, J., Blencowe, H., Pattinson, R., Cousens, S., Kumar, R., Ibiebele, I., ... Stanton, C. (2011). Stillbirths: Where? When? Why? How to make the data count? The Lancet, 377, 1368-1389.

Lovel], A. (1983). Some questions of identity: Late miscarriage, stillbirth, and perinatal loss. Social Science and Medicine, 17, 755-761.

Middelton-Moz, J. (1990). Shame and guilt: Masters of disguise. Deerfield, FL: Healthcommunications, Inc.

Mollica, R. M. L., Massagli, M., & Silove, D. (2004). Measuring trauma, measuring torture. Cambridge, MA: Harvard University.

Nettelbladt, P., Hansson, L., Stefansson, C. G., Borgqnist, L., & Nordstrom, G. (1993). Test characteristics of the Hopkins Symptom Check List-25 (HSCL-25) in Sweden, using the Present State Examination (PSE-9) as a caseness criterion. Social Psychiatry and Psychiatric Epidemiology, 28, 130-133.

Olsen, L. R., Mortensen, E. L., & Beth, E (2006). Mental distress in the Danish general population. Acta Psychiatrica Scandinavica, 113,477-484.

Parloff, M. B., Kelman, H. C., & Frank, J. D. (1954). Comfort, effectiveness, and self-awareness as criteria for improvement in psychotherapy. American Journal of Psychiatry, 3, 343-351.

Peppers, L. G., & Knapp, R. J. (1980). Motherhood and mourning. New York, NY: Praeger.

Radestad, I., Steineck, G., Nordin, C., & Sjogren, B. (1996). Psychological complications after stillbirth--influence of memories and immediate management: Population based study. British Medical Journal, 312, 1505-1508.

Rando, T. A. (1985). Bereaved parents: Particular difficulties, unique factors, and treatment issues. Social Work, 30, 19-23.

Roberts, B., Damnndu, E. Y., Lomoro, O., & Sondorp, E. (2009). Post-conflict mental health needs: A cross sectional survey of trauma, depression and associated factors in Juba, Southern Sudan. BMC Psychiatry, 9, 1-10.

Sandanger, I., Mourn, T., Ingebrigtsen, G., Dalgard, O. S., Sorensen, T., & Bruusgaard, D. (1998). Concordance between symptom screening and diagnostic procedure: The Hopkins Symptom Checklist-25 and the Composite International Diagnostic Interview II. Social Psychiatry and Psychiatric Epidemiology, 33, 345-354.

Sandanger, I., Moum, T., Ingebrigtsen, G., Sorensen, T., Dalgard, O. S., & Bruusgaard, D. (1999). The meaning and significance of caseness: The Hopkins Symptom Checklist-25 and the Composite International Diagnostic Interview II. Social Psychiatry and Psychiatric Epidemiology, 34, 53-79.

Schmader, T., & Lickel, B. (2006). The approach and avoidance function of guilt and shame emotions. Motivation and Emotion, 30, 43-56.

Shaver, K. G. (1970). Defensive attribution: The effects of severity and relevance on assigned responsibility for an accident. Journal of Personality and Social Psychology, 14, 100-113.

Taylor, S. (1983). Adjustment to life-threatening events. American Psychologist, 38, 1161-1172.

Tennen, H., & Affleck, G. (1990). Blaming of others for threatening events. Psychological Bulletin, 108, 209-231.

Tomkins, S. (1962). Affect imagery consciousness: The positive affects (Vol.1). New York, NY: Springer.

Tomkins, S. (1963). Affect imagery consciousness: The negative affects (Vol. 2). New York, NY:


Vance, J. C., & Najman, J. M. (1995). Psychological changes in parents eight months after the loss of an infant from stillbirth. Pediatrics, 96, 933-938.

Veijola, J., Jokelainen, J., Laksy, K., Kantojarvi, L., Kokkonen, P., Jarvelin, M. R., & Joukamaa, M. (2003). The Hopkins Symptom Checklist-25 in screening DSM-III-R axis-I disorders. Nordic Journal of Psychiatry, 5;7, 119-123.

Weinberg, N. (1994). Self-blame, other blame, and desire for revenge: Factors in recovery from bereavement. Death Studies, 18, 583-593.

Wong, P., & Weiner, B. (1981). When people ask why questions and the heuristics of attributional search. Journal of Personality and Social Psychology, 40, 650-663.

Joanne Cacciatore is affiliated with Arizona State University, J. Frederik Froen with the Norwegian Institute of Public Health, and Michael Killian with Florida State University. Correspondence about this article should be directed to Joanne Cacciatore, PhD, LMSW, FT, Arizona State University School of Social Work, 411 N. Central Rd., Phoenix, AZ 85004. Email: or jcaccia@me.corn,
Table 1. Demographics and Sample Characteristics

Variable Mean (SD) N (percentage)

Maternal age when stillbirth 28.7 (5.6) --
Maternal age at participation 31.7 (6.6) --
Years since experience of 2.6 (4.6) --
Urbanicity -- --
 Large Urban -- 684 (29.4%)
 Urban -- 1,130 (48.5%)
 Rural -- 516 (22.1%)
Participants from United States -- 1,660 (71.6%)
Marital status
 Single -- 167 (7.2%)
 Cohabitant -- 12.3 (12.4%)
 Married -- 1,865 (80.1%)
 Divorced -- 8 (0.3%)
Level of education (at least -- 1,778 (76.3%)
Primary race White -- 2,082 (90.1%)
Pregnant while participating -- 286 (12.3%)
Wanted pregnancy -- 1,914 (82.1%)
Gestational age at stillbirth in 33.9 (6.3) --
Third trimester stillbirth -- 1,853 (79.5%)
Attended a prenatal care program -- 1,660 (71.2%)
Attended ultrasound screenings -- 2,183 (93.6%)
Experience of physical or -- 255 (10.9%)
 psychological abuse during
Blame medical providers for -- 583 (25.0%)
Blame myself for stillbirth -- 574 (24.6%)
Blame others for stillbirth -- 258 (11.1%)
HSCL total score -- --
 Mean score 1.91 (0.52) --
 Participants with clinically -- 986 (42.3%)
 elevated score
HSCL depression score -- --
 Mean score 2.03 (0.60) --
 Participants with clinically -- 1,420 (61.7%)
 elevated score
HSCL anxiety score -- --
 Mean score 1.74 (0.49) --
 Participants with clinically -- 962 (41.3%)
 elevated score

Table 2. HSCL Descriptive Statistics and Correlations

 Intercorrelations and Descriptives

Variables HSCL Total Depression HCSL Anxiety

HSCL total -- r = .963 ** r = .868 **
HSCL depression r = .963 ** -- r = .703 **
HCSL anxiety r = .868 ** r = .703 ** --
Mean score (SD) 1.91 (.52) 2.03 (.60) 1.74 (.49)
Range 1.00-3.68 1.00-3.93 1.00-3.70
Percentage of 55.8% 60.9% 41.3%
 participants reporting
 a mean score higher
 than 1.75

* p<0.05 ** p<0.01

Table 3. Bivariate Models

 Predicting clinical elevation of a mean score
 of 1.75 or greater (reference category of
 not clinically elevated score)

Variables HSCL Total HSCL Depression
(reference category) OR, 95% CI OR, 95% CI

Maternal age at .969 *** (.95,.98) .982 *** (.97-99)
Years since .933 *** (.91-95) .922 *** (.90, .942)
Education .525 *** (.43-64) .605 *** (.49, .74)
 vs. not)
Urbanicity (urban .795 * (.65, .97) .826 (.67, 1.01)
 vs. rural)
Ethnicity (white vs. 1.057 (.80, 1.40) 1.06 (.80, 1.41)
Pregnant while 1.20 (.93, 1.54) 1.58 *** (1.23, 2.03)
 (yes vs. no)
Pregnancy wanted 1.43 ** (1.12, 1.82) 1.39 ** (1.09, 1.78)
 (yes vs. no)
Abuse during 2.24 *** (1.67, 3.0) 2.04 *** (1.5 1, 2.75)
 (yes vs. no)
Blame health care 1.58 *** (1.31, 1.93) 1.43 *** (1.17, 1.74)
 system (yes vs. no)
Blame myself 3.02 *** (2.44, 3.74) 3.87 *** (2.62, 4. 13)
 (yes vs. no)
Blame others 3.20 *** (2.33, 4.38) 3.53 *** (2.51, 4.97)
 (yes vs. no)

 Predicting clinical
 elevation of a mean score
 of 1.75 or greater
 (reference category of
 not clinically elevated score)

Variables HCSL Anxiety
(reference category) OR, 95% CI

Maternal age at .962 *** (.95-98)
Years since .940 *** (.92, .96)
Education .634 *** (.52, .77)
 vs. not)
Urbanicity (urban .788 * (.65, .96)
 vs. rural)
Ethnicity (white vs. 1.02 (.78, 1.35)
Pregnant while .802 (.63, 1.03)
 (yes vs. no)
Pregnancy wanted 1.35 * (1.07, 1.7 1)
 (yes vs. no)
Abuse during 1.73 *** (1.33, 2.25)
 (yes vs. no)
Blame health care 1.47 *** (1.21, 1.77)
 system (yes vs. no)
Blame myself 2.20 *** (1.82, 2.67)
 (yes vs. no)
Blame others 2.30 *** (1.76, 3.00)
 (yes vs. no)

* p<0.05 ** p<0.01 *** p<0.001

Table 4. Multivariate Models

 Predicting clinical elevation of
 a mean score of 1.75 or greater
 (reference category of not
 clinically elevated score)

Variables HSCL Total HSCL Depression
(reference category) OR, 95% CI OR, 95% CI

Nagelkerke [R.sup.2] .146 .147
 for model
Model [chi square] 254.86 ***, df=8 255.96 ***, df=6

Maternal age at .979 * (.96-99) --
Years since .914 *** (.89, .94) .908 *** (.89, .93)
Education .554 *** (.44, .70) .599 *** (.48,.75)
 vs. not)
Urbanicity (urban .761 * (.61,.95) --
 vs. rural)
Ethnicity (White vs. -- --
Pregnant while -- 1.71 *** (1.31, 2.23)
 (yes vs. no)
Pregnancy wanted -- --
 (yes vs. no)
Abuse during 1.89 *** (1.38, 2.60) 1.83 *** (1.33, 2.53)
 (yes vs. no)
Blame health care 1.30 * (1.05, 1.62) --
 (yes vs. no)
Blame myself 2.46 *** (1.94,3.11) 2.73 *** (2.13, 3.49)
 (yes vs. no)
Blame others (yes 1.78 ** (1.26,2.52) 2.01 *** (1.39, 2.92)
 vs. no)

 Predicting clinical elevation of
 a mean score of 1.75 or greater
 (reference category of not
 clinically elevated score)

Variables HCSL Anxiety
(reference category) OR, 95% CI

Nagelkerke [R.sup.2] .097
 for model
Model [chi square] 165.81 ***, df=8

Maternal age at .969 *** (.95, .99)
Years since .923 *** (.90, .946)
Education .698 ** (.56,.86)
 vs. not)
Urbanicity (urban .771 * (.63, .95)
 vs. rural)
Ethnicity (White vs. --
Pregnant while --
 (yes vs. no)
Pregnancy wanted --
 (yes vs. no)
Abuse during 1.40 * (1.05, 1.86)
 (yes vs. no)
Blame health care 1.24 * (1.0 1, 1.53)
 (yes vs. no)
Blame myself 1.85 *** (1.49, 2.30)
 (yes vs. no)
Blame others (yes 1.46 * (1.08, 1.96)
 vs. no)

* p<0.05 ** p<0.01 *** p<0.001
COPYRIGHT 2013 American Mental Health Counselors Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:RESEARCH
Author:Cacciatore, Joanne; Froen, J. Frederik; Killian, Michael
Publication:Journal of Mental Health Counseling
Date:Oct 1, 2013
Previous Article:Recovering identity: a qualitative investigation of a survivor of dissociative identity disorder.
Next Article:Scope of practice impact on employability in New York State: director and counselor views.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |