Dissociation-Induced Shame in Those with a Dissociative Disorder: Assessing the Impact of Relationship context using Vignettes

ABSTRACT

Some evidence in non-clinical groups suggests that the relationship context in which dissociation is experienced might moderate its association with shame. The current study used vignettes detailing either dissociative symptoms or the expression of sadness occurring in three different relationship contexts: with a friend, an acquaintance, or when alone. Ratings of emotional (e.g. shame, anxiety) and behavioral (e.g. leave, talk) reactions were made on single-item measures, and shame feelings were further assessed with the State Shame Scale. Participants were in treatment for either dissociative identity disorder (n = 31) or other specified dissociative disorder (n = 3; N = 34). Feelings of shame were elevated in the acquaintance condition compared to when with a close friend or alone regardless of whether dissociation or sadness was experienced. In the acquaintance context, participants exposed to dissociation or sadness reported feeling annoyed at themselves, having a greater desire to leave, and a lesser desire to talk compared to when these experiences happened with a close friend or alone. Results suggest those with a dissociative disorder appraise themselves as more vulnerable to shame if experiencing dissociation or sadness when with an acquaintance, potentially because the risk of not being understood and rejected is heightened.

KEYWORDS:

• Dissociation
• shame
• dissociative disorders

Theoretical and empirical work suggests a relationship between experiences of dissociation and feelings of shame (e.g., Benau, 2020, 2022; Platt & Freyd, 2015; Rudy et al., 2022; Thomson & Jaque, 2013). However, it is less well known, especially in those with dissociative disorders (DDs), to what degree this association is influenced by factors such as the relationship context within which shame or dissociation occurs (e.g., with a friend, acquaintance, or when alone). Given the relationship context of therapy, this becomes an important question to address. Studies examining these issues have been largely limited to non-clinical participants. This study examined whether relationship context influences the connection between dissociative experiences and reactive shame feelings in those with DDs using experiential vignettes.

Clinical observations have long connected shame and dissociation, typically noting that shame may underpin dissociative experiences and disorders (Benau, 2022; Bromberg, 2001; Kluft, 2007; Lightstone, 2004; Nathanson, 1992). As M. Lewis (1992) notes, “shame and dissociative processes have some systematic relation” (p. 172). Empirically, Irwin (1998) found that shame predicted dissociation in college students (see also Sutin & Stockdale, 2011), while Thomson and Jaque (2013) found a similar result in a sample of athletes and dancers, with other work reporting an association in psychiatric groups (e.g., Franzoni et al., 2013). Studies in traumatized samples have consistently found a correlation between shame and dissociation (e.g., Dorahy et al., 2013; Dorahy, 2010; Oh et al., 2016; Rudy et al., 2022; Talbot et al., 2004). Recent experimental work has shown dissociative-type experiences elevating as shame feelings increase, and conversely rises in shame correspond to increases in dissociative experiences (Dorahy et al., 2017, 2021). As an example of dissociation directly impacting on shame, Platt et al. (2017) used an induction task to elevate dissociation-type experiences in college students and found an increase in immediate shame and fear reactions.

Efforts to understand the association between shame and dissociation have queried whether the context in which one occurs influences the strength of relationship with the other. For example, does experiencing a dissociative episode in the presence of a new therapist lead to more or less shame compared to having the same experience in an established therapeutic relationship? Using vignettes to manipulate relationship context, McKeogh et al. (2018) randomly assigned non-clinical participants to scenarios in which dissociation (i.e., amnesia, flashback, or depersonalization) occurred while in the presence of a close friend, an acquaintance, or when alone. Each vignette was written in the second person (e.g., “you are meeting up with a very old friend today … ”), to increase the self-referential nature of the task. Upon exposure to the vignettes, participants completed measures of state shame, related behaviors, and other emotions. The single-item rating of shame and the state shame psychometric tool both showed higher shame ratings in the dissociation compared to sadness condition when with a close friend, but no differences across conditions when alone or with an acquaintance. These findings suggest that the association between dissociation and shame may be influenced by relationship context, especially ones characterized by more intimacy. Interestingly, no matter whether participants were exposed to the dissociation or sadness condition, they had a greater desire to talk about their experience with a close friend compared to an acquaintance or when alone, and less desire to leave their close friend following the experience than when in the other two relationship contexts. This suggests that while dissociation-induced shame may elevate in an intimate context, that context also offers a means of managing the experience (McKeogh et al., 2018).

In another study, participants were exposed to vignettes where dissociation or sadness occurred while alone, with a close friend or with a doctor (Dorahy et al., in press, Study 2). These latter two scenarios were relationally very familiar to nearly all participants, making it easier to identify with all scenarios. Participants who were exposed to dissociation showed higher scores on the single-item shame measure compared to sadness while with a close friend or doctor. However, scores between conditions did not differ when alone. Thus, dissociation-induced shame seemed to be more related to intimate, or at least interpersonal, contexts. For the state shame measure, scores did not differ across conditions in the specific relationship contexts, but overall, with dissociation and sadness grouped together, higher scores were evident when with a doctor compared to when alone. This again raises the possibility of relationship context being important to the production of shame after dissociation is experienced. Regarding behavioral responses, there was a desire to talk with one’s doctor about dissociating in their presence.

While such work has provided a starting point for exploring the role of relationship context in moderating the shame and dissociation linkage, it has been difficult understanding the clinical implications of such studies as they have only used samples from the general population. In therapeutic contexts, different interventions would apply if dissociative experiences were likely to induce shame than if another response was more likely. Dissociative identity disorder (DID) is the most severe dissociative disorder, with the personality of the individual divided into organized self-states that typically are not well co-ordinated to allow smooth adaptation to the environment, and which by definition are not integrated to allow a more unitary experience of self (e.g., Kluft, 2013; Van der Hart et al., 2006). DID has been associated with high levels of shame and dissociation (Dorahy et al., 2015; Kluft, 2007; Laddis & Dell, 2012) with these experiences at least in part attributed to high level of early relational trauma that characterizes the majority of DID cases (e.g., Chefetz, 2015; D. O. Lewis et al., 1997; Dorahy et al., 2016; Middleton & Butler, 1998). Thus, shame and dissociation in complex DDs are to a greater or lesser extent related to the interpersonal context of maltreatment, but it is not known whether relationship context might influence the likelihood of dissociation producing shame feelings.

This study drew on the methodology originally utilized by McKeogh et al. (2018) to assess if relationship context influenced the association between dissociation and shame in those with DDs. Vignettes describing experiences of dissociation or sadness when with a close friend, acquaintance, or alone were utilized. In line with previous work (e.g., McKeogh et al., 2018), acute experiences of shame and other emotions (to examine the unique effect of shame) were assessed, along with shame-related behaviors. Given dissociation-induced shame generally appears to be more evident in interpersonal and potentially intimate relationships, it was hypothesized that participants with a major DD would show a greater connection between dissociation and shame when dissociation compared to sadness occurs with a close friend than when with an acquaintance or when alone. In addition, and consistent with earlier findings, it was expected that participants were more likely to want to stay connected with a close friend by talking to them about the experience. Thus, while the friend context was predicted to elevate dissociation-related shame, it was also expected to provide some solution to it.

Method

Participants

Fifty-seven participants in treatment for a DD opened the survey. Twenty did not start or complete it, while three failed the validity tests. The remaining 34 were used for this study and contained two men, 30 women, and two non-binary people with a mean age of 46.71 years (SD = 10.05). Thirty-one participants (91.2%) reported being in treatment for DID, and three (8.8%) indicated being treated for Other Specified Dissociative Disorder. Comorbid diagnoses included personality disorder, PTSD, depression, and anxiety disorders. Twenty-five (73.5%) reported being single, while being in a relationship, married, or separated had three (8.8%) in each group. Regarding ethnicity, the majority identified as white European (n = 26, 76.5%), with a smaller number identifying as Asian (n = 2, 5.9%), nonwhite European (n = 1, 2.9%) or “other” (n = 5, 14.7%). For education, the majority had a post-high-school qualification (e.g., course, diploma, degree; n = 27, 79.4%).

Measures

A demographic questionnaire assessed age, gender, relationship status, ethnicity, and highest qualification, as well as current diagnoses. To assess the current level of trait dissociation and trait shame, the Dissociative Experiences Scale (DES; Carlson & Putnam, 1993) and the Experience of Shame Scale (ESS; Andrews et al., 2002) were utilized. The DES contains 28 self-report items assessed on an 11-point scale from 0% (never) to 100% (always). In the current study, the DES had a Cronbach’s alpha of r = .94. The ESS uses a 4-point scale ranging from 1 (“not at all”) to 4 (“very much”) to assess 25 shame-related items. A validity control question was included near the middle of the scale that read, “If you have read this question, please indicate not at all” (item 17). The ESS had a Cronbach’s alpha in the current study of r = .95.

Experimental stimuli

Vignettes

The vignettes utilized in this study were adapted from those used by McKeogh et al. (2018). They found that in some contexts, the amnesia vignette produced a little less state shame than the flashback vignette and a little more state shame than the depersonalization vignette. Because all people with DID experience amnesia, and to reduce the likelihood that the potential shame produced by the vignette would lead to dissociative episodes that may impact on self-reporting internal experiences, the amnesia vignette was used in this study as a proxy for dissociation experience. It was compared with the “sob” vignette as a proxy for sadness. The amnesia (dissociation) and sob (sadness) vignettes were given in the context of having the experience with a close friend (briefly defined as a long term relationship where there is safety and support), with an acquaintance (briefly defined as someone not known well and not seen in a long while) or when alone to manipulate relational context/intimacy. Thus, there were two independent variables: experience (dissociation, sadness) and relationship context (friend, alone, acquaintance). For the dissociation condition, the amnesia experience had the person feeling disoriented and experiencing a feeling of drifting off such that when they “returned,” they were not aware of what had been happening and why they were sitting in a different position in the chair. For the sadness condition, the sobbing experience had tears rolling down the person’s face while remembering a recent occurrence. Regarding the relationship context, the alone scenario involved being alone at home reading a book, the close friend scenario entailed talking to a close friend about a recent experience, while the acquaintance scenario involved discussing the years gone by since last seeing an old acquaintance. The Experience independent variable was manipulated between-subjects, while relationship context was manipulated within-subjects. The relationship context was randomized so it was presented in different orders. This led to six different vignette sets (e.g., dissociation: close friend, acquaintance, alone; dissociation: alone, close friend, acquaintance; sadness: acquaintance, alone, close friend) where participants were randomized to complete one (see https://osf.io/s8du7/).

To heighten their self-referential nature, vignettes were written in second-person narratives (e.g., “you become upset about something that happened to you recently”). Additionally, in all vignettes, specific phrases and words were bolded to emphasize the dissociation or sadness experience being presented.

Measures of state shame

To assess whether the vignettes evoked an emotional response, namely shame, three separate measures were used. The first assessed seven different emotions using individual questions (e.g., “Would you feel:” Angry? Ashamed? Sad? Disgust? Anxious? Guilty? Relaxed/Calm?). Ratings were made on a 5-point scale from 0 (not at all) to 4 (extremely).

The second assessment was the state shame subscale from the State Shame and Guilt Scale (SSGS, Marshall et al., 1994), which contains 5-items rated from “not at all” (1) to “very strongly” (5). The wording on the original shame subscale was slightly altered by taking the word “I” off the beginning of each sentence (“I feel worthless, powerless”) and creating a question targeted at the vignette for each item. For example, “would you feel worthless, powerless?” The SSGS has sound psychometric properties (Marshall et al., 1994).

The final measure was developed by McKeogh et al. (2018) and drew partly on the Compass of Shame model (Nathanson, 1992). It assessed eight behavioral responses to shame (e.g., withdraw, attack self and attack other). For both the close friend and acquaintance contexts, the eight responses were as follows: talk to your close friend/acquaintance about it, quickly leave the room to get away from your close friend/acquaintance, hide your head in your hands or divert your gaze from your close friend/acquaintance, get annoyed with yourself for having this experience in the presence of your close friend/acquaintance, distract attention away from what happened and talk to your close friend/acquaintance about something else, get frustrated at your close friend/acquaintance for being with you when you had this experience, wished you had had this experience alone rather than in the presence of your close friend/acquaintance, and sit calmly with your close friend/acquaintance. The alone context contained the same eight behavioral responses, but they were designed to reflect having the experience when alone (i.e., talk to someone about it, quickly leave the room where it occurred, hide your head in your hands, get annoyed with yourself for having this experience, distract attention away from what happened and think about something else, feel frustrated and think of someone who recently annoyed you and turn your frustration toward them, feel relieved you had this experience while alone rather than while you were with someone else, and sit calmly on your own). Ratings were made on a 5-point scale from 0 (never) to 4 (definitely) in terms of the likelihood of respondents engaging in each behavioral responses (for all state measures see https://osf.io/s8du7/).

Procedure

Participants were recruited via an invitation letter sent to them by their treating physician or nurse specialist engaged in their care. This letter contained the link for the survey which was administered via Qualtrics software. Upon clicking the link, participants read the information and consent form, completed the demographic questions, and were randomly administered the DCI and ESS. Participants were then randomly placed in one of two conditions (dissociation or sadness) where each had three randomized presentations of the relationship contexts. They were asked to imagine themselves in each scenario, as if it was their own experience. Participants then responded to shame measures randomly presented. Upon completion, participants were provided debriefing information, a list of support services, and offered the opportunity to go into a draw for one of three $50 shopping vouchers. The study was approved by the relevant human ethics committee.

Design and analysis

Mixed two-way multivariate analysis of variance (MANOVA) was used on the single-item emotion measures and the behavioral responses. A mixed two-way ANOVA was used on the state shame scale. In all cases, the independent variables were Experience (dissociation, sadness; between-subjects) and Relationship context (close friend, acquaintance, alone; within-subjects). One participant was an extreme outlier when rating the single-item emotion measure “sad” and one was an extreme outlier for the behavioral response, “annoyed at self.” These were winsorized. All variables were within skewness and kurtosis normality tolerances (≤2; George & Mallery, 2010), unless noted.

The interaction effect for the two shame measures (single item and State Shame Scale) and the interaction effect for the behavioral items were examined to assess hypotheses 1 and 2, respectively. The Wilks’ Lambda statistic was interpreted for MANOVA results, while Gabriel’s was used for post-hoc assessments as sample sizes were unequal (Field, 2013). Data analysis used Statistical Package for Social Sciences (IBM SPSS Statistics 28). All data are available at https://osf.io/s8du7/.

Results

Due to the randomization process and the nature of those excluded from the final data analysis, there were 20 participants in the dissociation experience and 14 participants in the sadness experience. There were no differences across experiences for age, F(1, 32) = .64, p = .431, η²_p = .02, and gender, χ²(2) = 1.52, p = .467, V = .21. Table 1 shows descriptive statistics. Due to the low cell counts, analyses across experience were not conducted for ethnicity, relationship status, and education. These were generally similar for dissociation and sadness experiences.

Table 1. Descriptive statistics for demographic variables across experience (dissociation vs sadness).

		Dissociation n = 20	Sadness n = 14
Age M (SD)		45.55 (9.85)	48.35 (10.45)
Gender n (%) Male Female Non-binary		2 (10%) 17 (85%) 1 (5%)	0 (0%) 13 (93%) 1 (7%)
Ethnicity n (%)
White European Other European Asian Other		13 (65%) 1 (5%) 1 (5%) 5 (25%)	13 (93%) 1 (7%) 0 (0%) 0 (0%)
Relationship status n (%) Single In relationship Married Separated		15 (75%) 3 (15%) 1 (5%) 1 (5%)	10 (71%) 0 (0%) 2 (14%) 2 (14%)
Education n (%) No high school High school Course/diploma Degree Masters PhD Other		3 (15%) 2 (10%) 3 (15%) 7 (35%) 3 (15%) 1 (5%) 1 (5%)	1 (7%) 1 (7%) 3 (21%) 6 (43%) 3 (21%) 0 (0%) 0 (0%)

Trait dissociation and trait shame (Table 2)

For trait dissociation (DES), a between-subject ANOVA showed no main effect for Experience, F(1, 30) = .65, p = .43, η²_p = .02, Relationship Context randomizations, F(2, 30) = 1.17, p = .33, η²_p = .07, nor an interaction, F(2, 30) = .78, p = .47, η²_p = .05. Similarly, for trait shame (ESS) there was no Experience, F(1, 30) = .53, p = .47, η²_p = .02, or Relationship Context, F(2, 30) = .59, p = .56, η²_p = .04, main effects, nor an interaction, F(2, 30) = .53, p = .60, η²_p = .03. Thus, trait dissociation and shame were similar across conditions.

Table 2. Means and standard deviations for experience (dissociation, sadness) by relationship context randomizations for the DES and ESS.

		Close friend, alone, acquaintance M (SD)	Alone, acquaintance, close friend M (SD)	Acquaintance, close friend, alone M (SD)
DES Dissociation Sadness		56.58 (26.15) 52.14 (24.14)	78.39 (2.77) 59.92 (17.08)	54.38 (18.74) 59.40 (5.77)
ESS Dissociation Sadness		83.14 (21.16) 79.20 (18.40)	96.05 (3.54) 83.40 (5.55)	80.45 (18.88) 83.67 (11.06)

Single-item emotions

For the single-item emotion ratings, “relaxed/calm” was considerably non-normally distributed and given it was not central to our hypotheses, it was dropped from further analyses. A 2 (Experience) × 3 (Relationship Context) mixed MANOVA on anger, shame, sad, disgust, anxious, and guilty ratings (see Table 3) produced no multivariate effect for Experience, Λ = .90, F(6, 27) = .49, p = .812, η²_p = .09. However, the multivariate effect for Relationship Context, Λ = .22 F(12, 21) = 6.40, p < .001, η²_p = .79, and the multivariate interaction effect, Λ = .40, F(12, 21) = 2.66, p = .024, η²_p = .60, were significant. Regarding Relationship Context, there were significant univariate effects for shame, F(2, 64) = 16.89, p < .001, η²_p = .35, disgust, F(2, 64) = 3.43, p = .038, η²_p = .10, anxiety, F(2, 64) = 4.19, p = .019, η²_p = .12, and guilt, F(2, 64) = 13.57, p < .001, η²_p = .30. Participants reported more shame following either dissociation or sadness conditions in the acquaintance context compared to when alone (p < .001) or with a close friend (p = .001). They trended toward having more shame when with a close friend than when these experiences happened alone (p = .058). For disgust, ratings were higher in the acquaintance context compared to when alone (p = .02), and trended toward higher ratings when with an acquaintance than with a close friend (p = .054). There was no difference between ratings for close friend and alone (p = .64). Participants reported more anxiety following exposure to dissociation or sadness vignettes with an acquaintance than with a close friend (p = .011) or when alone (p = .045), but the latter two did not differ (p = 1.00). For guilt ratings, participants showed higher scores in the acquaintance and close friend contexts compared to when alone (p < .001; p = .003, respectively). The two conditions involving another person did not differ (p = .102).

Table 3. Means and standard deviations for relationship context (close friend, acquaintance, alone) by experience (dissociation, sadness) on seven single-item emotion ratings.

		Close friend M (SD)	Acquaintance M (SD)	Alone M (SD)	Total M (SD)
Anger Dissociation Sadness		2.55 (1.64) 2.71 (1.49)	2.60 (1.53) 3.50 (1.34)	2.55 (1.57) 2.57 (1.70)	2.56 (1.58) 2.93 (1.51)
Shame Dissociation Sadness		3.15 (1.42) 3.79 (1.53)	3.95 (1.36) 4.57 (.51)	2.95 (1.39) 2.93 (1.54)	3.35 (1.39) 3.76 (1.19)
Sad Dissociation Sadness		3.30 (1.30) 3.71 (.99)	3.70 (1.22) 3.14 (1.09)	3.25 (1.41) 4.14 (.95)	3.42 (1.31) 3.59 (1.17)
Disgust Dissociation Sadness		2.85 (1.60) 3.29 (1.82)	3.20 (1.64) 4.00 (1.18)	2.60 (1.64) 3.36 (1.60)	2.88 (1.63) 3.55 (1.53)
Anxious Dissociation Sadness		3.70 (1.17) 3.86 (1.17)	4.00 (1.30) 4.50 (.65)	4.10 (1.21) 3.29 (1.44)	3.93 (1.23) 3.88 (1.08)
Guilty Dissociation Sadness		3.10 (1.52) 4.07 (1.33)	3.65 (1.50) 4.07 (1.44)	2.70 (1.69) 3.14 (1.66)	3.15 (1.57) 3.76 (1.48)

In terms of the significant multivariate interaction, univariate effects were evident for sad, F(2, 64) = 8.34, p < .001, η²_p = .21, and anxious F(2, 64) = 5.41, p = .007, η²_p = .15, feelings. Simple effects analyses showed a trend toward more sad feelings for the sadness versus dissociation experience when alone (p = .081). Neither the close friend (p = .434) nor acquaintance (p = .147) contexts showed differences across sadness and dissociation experiences for sad feelings. Regarding the interaction for anxious feelings, there was a trend toward more anxiety in the dissociation compared to sadness vignette when alone (p = .083). No difference in anxiety was evident across dissociation and shame conditions when with a close friend (p = .703) or an acquaintance (p = .194).

State shame

For state shame, the ANOVA showed no main effect for Experience, F(1, 32) = 1.53, p = .226, η²_p = .05, nor an interaction between Experience and Relationship Context, F(2, 64) = .32, p = .731, η²_p = .01 (see Table 4). However, there was a significant main effect for Relationship Context, F(2, 64) = 13.19, p < .001, η²_p = .29, with greater shame in the acquaintance compared to close friend (p < .001) or alone (p < .001) contexts. Shame ratings did not differ across close friend and alone contexts (p = .338).

Table 4. Means and standard deviations for relationship context (close friend, acquaintance, alone) by experience (dissociation, sadness) for state shame scale scores.

	Close friend M (SD)	Acquaintance M (SD)	Alone M (SD)	Total M (SD)
Experience Dissociation Sadness	3.06 (1.45) 3.34 (1.44)	3.89 (1.35) 4.47 (.57)	3.16 (1.45) 3.69 (.73)	3.37 (1.42) 3.83 (.91)

Behavioral responses

In terms of the eight behavioral responses, the two-way MANOVA across Experience and Relationship Context (see Table 5) showed no multivariate effect for Experience, Λ = .71, F(8, 25) = 1.27, p = .301, η²_p = .29, nor the interaction, Λ = .35, F(16, 17) = 1.98, p = .086, η²_p = .65. The Relationship Context multivariate effect was significant, Λ = .25, F(16, 17) = 3.62, p = .010, η²_p = .75, with univariate effects for talk, F(2, 64) = 10.73, p < .001, η²_p = .25, leave, F(2, 64) = 10.01, p < .001, η²_p = .24, annoyed at self, F(2, 64) = 6.17, p = .004, η²_p = .16, alone, F(2, 64) = 4.82, p = .011, η²_p = .13, and Calm, F(2, 64) = 3.77, p = .028, η²_p = .11. Participants wanted to talk more following either dissociation or sadness conditions in the close friend and alone contexts compared to the acquaintance context (p < .001; p = .003). There was no difference between close friend and alone contexts for wanting to talk (p = .14). For the desire to leave, ratings were higher in the acquaintance context compared to when alone (p = .002), or with a close friend (p < .001). There was no difference between ratings for close friend and alone (p = .79). Participants reported more annoyance at self if dissociation or sad feelings occurred with an acquaintance than when alone (p = .002) or with a close friend (p = .004), while the latter two did not differ (p = .810). For wishing the experience had occurred while alone, scores were higher in the acquaintance context compared to close friend (p = .003) or alone (p = .015) contexts. Finally, in terms of the desire to sit calmly in the context following the dissociation and sadness vignettes, scores were higher with a close friend than when with an acquaintance (p = .026) and trended in that direction when alone compared to when with an acquaintance (p = .054). There was no difference between when with a close friend or when alone (p = .49).

Table 5. Means and standard deviations for relationship context (close friend, acquaintance, alone) by experience (dissociation, sadness) for the eight behavioral responses.

		Close friend M (SD)	Acquaintance M (SD)	Alone M (SD)	Total M (SD)
Talk Dissociation Sadness		2.95 (1.31) 3.07 (1.64)	1.80 (1.06) 2.00 (1.18)	2.80 (1.15) 2.29 (1.20)	2.52 (1.17) 2.45 (1.34)
Leave Dissociation Sadness		2.30 (1.21) 2.93 (1.54)	3.10 (1.33) 3.79 (1.37)	2.75 (1.37) 2.43 (1.22)	2.72 (1.30) 3.05 (1.38)
Hide Dissociation Sadness		3.45 (1.32) 3.57 (1.50)	3.20 (1.28) 3.57 (1.40)	2.80 (1.36) 3.36 (1.08)	3.15 (1.32) 3.50 (1.33)
Annoyed at self Dissociation Sadness		3.75 (1.33) 3.64 (1.33)	4.10 (1.33) 4.64 (.84)	3.50 (1.43) 4.14 (.95)	3.78 (1.36) 4.14 (1.04)
Distract Dissociation Sadness		3.40 (1.47) 3.71 (1.27)	3.85 (1.14) 4.07 (1.14)	3.20 (1.24) 3.71 (1.33)	3.48 (1.28) 3.83 (1.25)
Annoyed at other Dissociation Sadness		1.60 (1.04) 2.07 (1.44)	1.45 (.68) 2.14 (1.09)	1.75 (1.12) 2.36 (1.39)	1.60 (.95) 2.19 (1.31)
Wished to experience alone Dissociation Sadness Sit Dissociation Sadness		3.85 (1.35) 3.64 (1.44) 2.85 (1.04) 2.29 (.83)	4.50 (.89) 4.57 (.94) 2.60 (1.05) 1.71 (.73)	4.00 (1.33) 3.71 (1.59) 2.85 (1.04) 2.64 (1.65)	4.12 (1.19) 3.97 (1.32) 2.77 (1.04) 2.21 (1.07)

Discussion

The hypothesis that the close friend context would show heightened shame if dissociation compared to sadness occurred was not supported. Rather, the dissociative disorder sample showed no difference in shame feelings across dissociation and sadness experiences, with both elevating shame when with an acquaintance compared to being alone or with a close friend. These findings were evident for both single-item shame and state shame scale measures. Compared to being alone or with a close friend, being with an acquaintance and experiencing dissociation or sadness also elevated disgust and anxiety. Guilt feelings were elevated for dissociation and sadness conditions when with an acquaintance or close friend compared to when alone. For the single-item measures, there was an interaction between experience and relationship context for sadness and anxiety, but further analyses were not robust and so cannot be easily interpreted. The second hypothesis that the close friend context would be associated with the desire to talk about the experience was partially supported, with higher scores in that context and when alone, compared to when with an acquaintance. However, again this effect was not isolated to the dissociation condition but also occurred with sadness. There was a greater desire to want to be alone and leave the acquaintance context, as well as feeling annoyed at oneself for having those experiences in this context, compared to when with a close friend or alone. Being motivated to remain calmly present was heightened in the close friend compared to the acquaintance context.

Both dissociation and sadness conditions showed the capacity to heighten feelings of shame in those with a dissociative disorder. In the sample, mean scores on single item and state shame measures for both experiences, and across all contexts, were well above scores that would indicate no reported acute shame feelings. Thus, for this dissociative disorder sample, putting themselves in the vignette and imagining experiencing dissociation or sadness elevated feelings of shame. This finding is consistent with previous work suggesting that dissociation had a natural propensity to activate shame feelings (Dorahy et al., 2021), and this may be especially the case for individual with a history of interpersonal trauma from a trusted other (Platt et al., 2017), which characterizes the history of most people with a major dissociative disorder (e.g., Middleton & Butler, 1998; Ross et al., 1989). Seemingly, for those with a dissociative disorder, experiences of sadness also have a propensity to activate shame.

A study in the general population found that dissociation more than sadness activated shame (McKeogh et al., 2018). Yet, for those with DID or related disorders, the dissociation and sadness conditions equally heightened shame, and this was particularly evident when with an acquaintance, a relationship that did not have the same level of intimacy as being with a close friend.

Dorahy et al. (2021) found that feeling flawed and exposed were dominant appraisals for experiencing shame after dissociative experiences in college students (Study 1) and those in treatment for the adult outcomes of child sexual abuse (Study 2). Work using a design comparable to the current study, found that ideas around losing control, being excluded or being judged were dominant appraisals provided for shame when exposed to dissociation and sadness vignettes in the general population (Dorahy et al., in press). One possible interpretation for heightened shame following the dissociation and sadness vignettes in the acquaintance context of the current dissociative disorder sample is that such experiences in the presence of a person less emotionally familiar activate a sense of being flawed, exposed, judged, and not in control of internal experiences. Dissociative disorders have been described as conditions of secrecy (Kluft, 1991; Loewenstein et al., 2017), with at least some dissociative symptoms (e.g., amnesia, dissociative identities) keeping experiences hidden from the person themselves and potentially from others (e.g., they do not consciously disclose information not known to them). Dissociative symptoms, therefore, can operate to assist an individual experience themselves as operating in an apparently normal manner, despite significant psychological disruption (Van der Hart et al., 2006). This dissociative cloak requires that not only the content of psychological experience is kept invisible to self and others but also the mechanism assisting it (i.e., dissociation), so control and poise is maintained. Thus, dissociative symptoms are often kept secret or very reluctantly disclosed in DDs, especially during assessment and at the start of therapy, for shame-based reasons (Loewenstein et al., 2017). The occurrence of dissociative and emotional (e.g., sadness) experiences in the presence of an acquaintance may be interpreted as exposing the flaw of losing control of psychological functioning in the presence of a somewhat unfamiliar and potentially less accepting (i.e., rejecting) other, which in turn heightens reactive shame feelings in those with DDs.

The elevation of anxiety and disgust in the acquaintance condition could be understood in light of this interpretation. Feelings of anxiety following the dissociation or sadness conditions for an acquaintance may relate to fear of possible rejection and the shame it brings, and disgust (seemingly at oneself) may arise from not maintaining the wished-for psychological poise in such a situation. The disgust interpretation is limited by a lack of specificity of the emotion in the current research, as it was not explicitly assessed whether disgust was toward oneself, another, or the situation. Future research should directly address the factors underpinning shame following dissociative and emotional experiences in those with clinical dissociative difficulties.

Findings in the general population using the same study design have generally, though not invariably, suggested that dissociation- or sadness-induced shame is more evident in intimate (e.g., close friend, therapist) versus less intimate (e.g., acquaintance, alone) contexts. McKeogh et al. (2018) argue that more intimate relationships bring with them a greater risk of emotionally meaningful rejection, a key driver of shame (Gilbert, 1997; Nathanson, 1992). As such, dissociative and sadness manifestations when with a close friend may be seen in the general population as producing a greater risk of shame-related rejection. The dissociative disorder sample in the current study showed heightened dissociation- and sadness-induced shame when in the less intimate context of being with an acquaintance. Given their history of dissociative symptoms and emotional dysregulation, perhaps they felt that dissociation and sadness were more acceptable to a close friend and they were less likely to risk rejection by exposing a flaw or losing control, than when with an acquaintance. This finding is consistent with what has been observed clinically, that risk of rejection and shame is especially heightened in therapeutic work with DD clients before the patient experiences the therapist having a greater acceptance of them (Chu, 2011; Steele et al., 2017).

The notion that those with DDs feel less fearful of shame-related rejection when with a close friend than with an acquaintance was evident in their behavioral responses. Not only did they feel more shame following the dissociation or sadness conditions when with an acquaintance compared to a close friend, but they also had less desire to talk or sit with an acquaintance and a greater desire to leave them, while feeling more annoyed at themselves for having such experiences. Thus, the acquaintance context evoked the most shame when exposed to the dissociation and sadness conditions and a greater desire to leave that context, not talk about what happened, and feel annoyed at oneself. Unlike with those in the general population where a close friend context prompted more shame but also provided a sense of support, with participants wanting to stay and talking about the experience (McKeogh et al., 2018), the dissociative disorder sample in this study did not feel like the context they felt the most shame in was also the context that could provide relief from shame. Instead, and consistent with the central dynamics of shame management, the context promoting shame after dissociation and sadness experiences, was the context prompting most avoidance behaviors (Nathanson, 1992).

Limitations and conclusion

One major limitation of the current study is its small sample size. While this is a common issue with studies relying on clinical samples, the use of a small sample in a between-subjects design requires exercising some caution, especially interpreting non-significant results which may be the result of limited power rather than true null findings. Future research should therefore build on this novel work and utilize larger sample sizes to replicate and expand on these findings.

Second, the study utilized participants with a major dissociative disorder, the large majority with DID. Therefore, the findings speak directly to the responses of individual with such presentations. However, the participants were initially selected by their treating clinician and then self-selected themselves into the study. Consequently, they were not a representative dissociative disorder sample, and efforts were not made to assess factors like time and progress in treatment which ideally reduce schematized shame beliefs about self and may influence results.

In addition, like with this methodological approach generally, the current design asked participants to imagine themselves in particular relational contexts having specific experiences, rather than testing the variables of interest directly (e.g., inducing dissociation in different relationship contexts). Thus, the degree to which findings and conclusions drawn from them translate into psychological functioning in real-life setting for those with DDs is not clear. However, such findings do provide directions and hypotheses for future work, and clinically could prime clinicians to the possibility that clients with DDs may experience shame when dissociating or expressing sadness in their presence, especially early in therapy (i.e., when the therapist is more like an acquaintance). Finally, a manipulation assessment was not made following the dissociation induction to assess for heightened dissociative experiences.

Using vignettes describing experiences of dissociation and sadness when alone, with a close friend or with an acquaintance, those with a DD reported heightened shame in the presence of an acquaintance, someone they had the least emotional intimacy with. They also had a greater desire to withdraw from this context and not to share their experience with the acquaintance. In this context, there is a risk of rejection before the person is known to the other, and having dissociative and sadness experiences in the presence of someone who is perceived not to understand may drive an increase in shame in those with a dissociative disorder.

Acknowledgment

The authors acknowledge Kate McKeogh for assistance with the design of the study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported that there is no funding associated with the work featured in this article.