The mediating role of self-compassion in the relationship between internalized stigma and psychological resilience in bipolar disorder
Abstract
Introduction
Internalized stigma is known to be high in bipolar disorder (BD). Concepts such as self-compassion and psychological resilience have recently begun to be studied as protective factors for BD. The aim of the current study was to examine the relationships between internalized stigma, self-compassion and resilience among individuals with BD.
Method
One hundred and thirty-two male and female (18-65 years of age) participants with a DSM 5 diagnosis of BD (BD- I & BD- II) were included. The remission criteria (YMRS< 5 and HDRS< 7) was evaluated using clinician-administered measures and all participants were found to be remitted. Correlation and mediation analyses were performed. Participants completed the Internalized Stigma in Mental Illness Scale (ISMI), the Self-Compassion Scale (SCS) and the Resilience Scale for Adults (RSA).
Results
Significant correlations were found between internalized stigma, sub-dimensions of self-compassion (self-kindness, self-judgement, common humanity, isolation, mindfulness, and over-identification), and resilience in the expected directions like negative correlations between internalized stigma and positive dimensions of self-compassion (self-kindness, common humanity and mindfulness). Self-judgement and self-kindness mediated the relationship between internalized stigma and psychological resilience.
Conclusions
The findings of the study shed light on which dimensions of self-compassion might be more beneficial to work with in order to increase resilience when working with internalized stigma in BD. This strengths-based investigation would be valuable to enrich psycho-social interventions for the prevention of relapse in BD.
INTRODUCTION
Bipolar disorder (BD) is a life-long, chronic and often debilitating condition characterized with (hypo)manic and depressive episodes with/without mixed symptoms. These episodes may recur without a regular pattern and a normal mood (euthymia) can be seen between these episodes (Goodwin & Jamison, 2007). BD is accepted as the sixth leading cause of disability worldwide (Ferrari et al., 2016).
Since patients with BD have almost non-symptomatic periods, the difference between normal, euthymic and symptomatic periods becomes blurred. This causes individuals with this disorder to perceive any mood swings as a danger sign of stigmatization. Internalized stigma is about having the inner subjective counterpart of negative stereotypes and stigmatizing attitudes imposed by society related to receiving a psychiatric diagnosis, such as inadequacy or dangerousness, even if they are not exposed to explicit discrimination or any negative behaviour (Yanos et al., 2015). As a result of internalized stigma, the person withdraws from society with negative feelings like worthlessness and shame (Corrigan, 1998), which decreases their self-esteem and lowers their levels of functioning (Hawke et al., 2013). Many studies have revealed that internalized stigma is quite high in patients with BD (Boyd-Ritsher & Phelan, 2004; Grover et al., 2020).
Resilience and mental illness have recently been explored (Davydov et al., 2010). It is the human's dynamic capacity to withstand or recover from significant life challenges that might threaten its development, viability or stability (Masten, 2011). Despite of the scarcity of the research on the role of resilience in BD, the existing research indicates that resilience is lower in BD when compared with healthy controls (Choi et al., 2015; Hofer et al., 2017; Lee et al., 2017; Mizuno et al., 2016; Post et al., 2018).
Since internalized stigma is frequently observed in mental illnesses like BD, several researchers state that the concept of self-compassion might be important for patients to behave themselves understandingly and kindly (Brohan et al., 2011; Edwards, 2022; Heath et al., 2018; Mateer et al., 2023; Wong et al., 2019). Self-compassion involves being open to one's own suffering, being able to contact this suffering, and easing it without avoidance and a desire for self-healing with kindness. Self-compassionate individuals are more understanding and kind to themselves when they experience failure, inadequacy, or misfortune (Neff, 2003; Neff, 2022a, 2022b). It has been found that individuals with BD had lower levels of self-compassion than healthy controls and that self-compassion could act as a buffer against depression, anxiety, and difficulties in emotion regulation in BD (Døssing et al., 2015; Fletcher et al., 2019). Also, lower scores on ‘self-kindness’ subscale and higher scores on ‘self-judgement’ subscale of SCS were found (Døssing et al., 2015). Ranjbar et al. (2020) found that dimensions of internalized stigma were correlated with self-compassion among BD patients. Heath et al. (2018) and Wong et al. (2019) stated that self-compassion might help people with BD experiencing internalized stigma to accept and see themselves and situations in a more balanced way.
Since internalized stigma makes a person more critical and judgemental about oneself, a negative association between self-stigma and resilience was found in BD (Post et al., 2021). However, self-compassion as a positive human strength augments feelings related to inter-connectedness, stimulates kindness and provides more balanced and broader awareness. Thus, it can promote resilience in BD (Keyes, 2005; Neff et al., 2007).
The studies mentioned above reveal that individuals with BD suffer from internalized stigma and have lower self-compassion and resilience levels than healthy individuals (e.g. Fletcher et al., 2019; Grover et al., 2020; Mizuno et al., 2016). Regarding BD, the relationship between internalized stigma and resilience (Ellison et al., 2013; Post et al., 2021), the relationship between internalized stigma and self-compassion (Heath et al., 2018; Ranjbar et al., 2020; Wong et al., 2019) and the relationship between self-compassion and resilience (Keyes, 2005; Neff et al., 2007) were presented by different researchers. However, no research has been found in the literature regarding BD examining the relationships between these three concepts. To fill this gap, the primary objective of this study was to look into the relationships between internalized stigma, self-compassion and resilience in BD. Then, the impact of internalized stigma and self-compassion on resilience was examined. Lastly, the mediating effects of sub-dimensions of self-compassion between internalized stigma and psychological resilience were examined. It was thought that taking into account protective factors such as self-compassion and resilience to deal with the perception related with internalized stigma might be beneficial in enriching and strengthening the content of psychosocial interventions (individual or group psychotherapies, psychoeducation, etc.) for bipolar disorder.
MATERIALS AND METHODS
Participants and procedures
For this study, 143 male and female patients (18–65 years of age) who were diagnosed with bipolar disorder (BB-I & BB-II) according to the Diagnostic and Statistical Manual of Mental Diseases (DSM-5) were recruited from the University of Health Sciences Bakirkoy Mazhar Osman Mental and Neurological Diseases Education and Research Hospital Mood Disorder Center and Psychiatry Outpatient Unit. Other eligibility criteria included being on the maintenance treatment, being in remission, being literate in Turkish and agreeing to participate in the study. Exclusion criteria were set as having any comorbid personality disorders or psychiatric/neurological disorders and not having any physical disability that would prevent filling the research battery. Clinical diagnoses were made by specialist psychiatrists working in the University of Health Sciences Bakirkoy Mazhar Osman Mental and Neurological Diseases Education and Research Hospital Mood Disorder Center and Psychiatry Outpatient Unit, and the participants included in the study were patients who had been followed up for at least a year in this hospital. Diagnoses of personality disorders were evaluated by reviewing the information obtained in the patient files and the interview notes in these files. One patient was excluded because of substance use disorder diagnosis, and 10 patients were excluded because they filled in insufficiently, so the study was continued with 132 patients.
The study was approved by the ethics committee of the University of Health Sciences Bakirkoy Mazhar Osman Mental and Neurological Diseases Research and Training Hospital with the decision number 379. First, informed consent form that gave information about the purpose of the study was obtained. Then, the Young Mania Rating Scale (YMRS) and the Hamilton Depression Rating Scale (HDRS) were applied to evaluate the remission criteria. This criteria were set as YMRS<5 and HDRS<7. After that, all participants who met the inclusion and exclusion criteria were asked to fill in the Sociodemographic and Clinical Data Form, the Internalized Stigma in Mental Illness Scale (ISMI), the Self-Compassion Scale (SCS) and the Resilience Scale for Adults (RSA). It took, on average, 20–30 min to complete the scales.
Measures
Sociodemographic and clinical data form
It is a detailed interview form prepared by the researchers in accordance with the aim of the research, evaluating the sociodemographic characteristics of the participants, the clinical diagnosis process, the course of the disorder, physical examination of the patient and past suicide attempts.
Internalized Stigma of Mental Illness Scale
Internalized Stigma of Mental Illness (ISMI) is a 4-point Likert scale developed by Ritsher and Phelan (2004). It consists of 29 items and has five subscales: “alienation” “stereotype endorsement”, “perceived discrimination”, “social withdrawal” and “stigma resistance”. Larger scores indicate greater self-stigma. The Turkish adaptation was made by Ersoy and Varan (2007). In the original study, the Cronbach alpha coefficient was found as .85, in the Turkish adaptation study as .93, and in the present study as .89.
The Resilience Scale for Adults
Resilience Scale for Adults (RSFA), which was developed by Friborg et al. (2003), is a self-report scale that aims to measure the level of psychological resilience. It consists of 33 items and a total of six sub-dimensions: “perception of self” (6 items), “perception of future” (4 items), “structured style” (4 items), “social competence” (6 items), “family cohesion” (6 items) and “social resources” (7 items). The Turkish adaptation was made by Basim and Cetin (2011). The 6-factor structure was confirmed for the Turkish version of RSA. Alpha coefficients of these six sub-dimensions were found to range from .66 to .81 and the Cronbach alpha coefficient of the total scale was .86. In the present study, the Cronbach alpha coefficient of the total scale was calculated as .92.
Self-Compassion Scale
Self-Compassion Scale (SCS) is a 26-item and 5-Likert (1 = almost never, 5 = almost always) self-report scale comprising six sub-scales, which measure dimensions of global self-compassion: “self-kindness” (being understanding and kind towards oneself in difficult times), “self-judgement” (while suffering, punishing oneself with criticism), “common humanity” (seeing difficulties as part of a common experience), “isolation” (feeling isolated by emotional pain and distress), “mindfulness” (being mindful of the present moment by fully accepting feelings, thoughts, and bodily sensations), and “over-identification” (having inclination to over-identify with negative experiences). Higher scores indicate greater self-compassion. In the original study (Neff, 2003), the internal consistency was found as .92. In the Turkish adaptation study (Akin et al., 2007), the Cronbach alpha internal coefficient was calculated as .94. In terms of subscales of SCS, the Cronbach alpha values was ranged from .87 to .94. In the present study, the Cronbach alpha value was calculated as .68. In terms of subscales, the alpha values were varied from .75 to .86.
The Young Mania Rating Scale (YMRS)
It is a clinician-administered scale to measure the severity and change of manic symptoms (Young et al., 1978). Seven items of the scale, which consists of 11 items in total, are of 5-point Likert type, and the other four items are of 9-point Likert type. While having well-established reliability and validity (Young et al., 1978), higher scores point out more severe mania. In the Turkish validity and reliability study, the inter-rater reliability was found as 0.79; the correlations between the items with statistical significance and the total score was .40–.84. The agreement among the scale items of two researchers was 63.3%–95%, the kappa values was found between 0.114 and 0.849 (Karadağ et al., 2001).
The Hamilton Depression Rating Scale
Hamilton Depression Rating Scale (HDRS) is a clinician-administered scale to evaluate the severity of depressive symptoms. Originally, developed as 17-items, later four items were added. A total score is obtained for the sum of the ratings of each item made, and the increase in the score indicates the severity of depressive symptoms. In Turkish validity and reliability study (Akdemir et al., 1996), the test–retest reliability of the scale was found to be .85; two-half reliability was .76; the Cronbach Alpha internal consistency coefficient was .75; and the reliability coefficients were found between .87 and .97.
Statistical analyses
Statistical analyses were performed using IBM SPSS Statistics Version 25.0. First, to see whether the scale scores showed a normal distribution, skewness and kurtosis values were examined. As a result of the examination, it was observed that the skewness and kurtosis values were between [−1.5, +1.5] (for ISMI, skewness = .38, kurtosis = −.26; for RSFA, skewness = −.52, kurtosis = −.47; for SCS, skewness = −.21, kurtosis = .19). So, it was accepted that the data had a normal distribution (Tabachnick & Fidell, 2013). Then, descriptive statistics (mean, standard deviations, ranges, Cronbach's alpha) were reported for each of the measures. Second, Pearson product moment correlations (r) were computed to indicate bivariate correlations between internalized stigma, self-compassion and psychological resilience. Then, the possible confounding influence of demographic variables were analysed using a series of multivariate regression analyses. After that, mediation analyses were undertaken using structural equation modelling (SEM) using SPSS Amos 24.0. SEM represents a range of cause-effect relationships among variables in testable composite models (Shipley, 2002). In this study, the mediator effect of self-compassion was explored by testing the relationship between internalized stigma (IV) and psychological resilience (DV) based on Baron and Kenny's (1986) method for mediation. Therefore, at first the relationships between internalized stigma and psychological resilience were examined without including mediator variables (sub-dimensions of self-compassion) with path analysis. In the next step, the model of mediating variables was tested and mediation was determined. In the last stage, the theoretically constructed model was tested. The significance values of the estimation (β), standard error, critical ratio, standardized estimation and model paths of the tested model were presented.
Power analysis
G-Power 3.1.9.7 was performed by assuming a Type-I error of 0.05 and an effect size of 0.80. When the correlation coefficient is calculated as a result of the power analysis performed, ρ ≥ 0.3 is obtained and it is seen that the sample size should be at least 84. Accordingly, the sample size of 132 patients with BD allows for the detection of the relationships between variables by Pearson correlation analysis. This is a medium effect correlation according to Cohen's classification (Cohen, 1992). In addition, using the online calculator (https://www.analyticscalculators.com/calculator.aspx) for the SEM analysis which includes two latent variables and six observed variables, the expected effect size was calculated as 0.5 and the statistical power as .80, and the p value as .05, then the minimum sample size required for this study was set as 23, so a sample size of 132 people was thought to be sufficient to determine the statistical significance of a single predictor variable.
RESULTS
Mean scores, standard deviations, score ranges in the sample are shown in Table 1. All measures represented adequate variability and internal consistency.
| Mean (SD) | Range | |
|---|---|---|
| HDRS | 2.01 (1.57) | 0–7 |
| YDRS | .72 (.69) | 0–2 |
| ISMI | 2.13 (.41) | 1.50–3.77 |
| RSFA | 3.62 (.83) | 1.52–5 |
| SCS | 2.65 (.42) | 1.52–5 |
| SK | 2.97 (1.0) | 1–5 |
| SJ | 2.24 (.99) | 1–4.8 |
| CH | 2.79 (.96) | 1–5 |
| ISL | 2.39 (1.06) | 1–4.5 |
| OI | 2.51 (1.04) | 1–5 |
| MF | 3.01 (1.03) | 1–5 |
- Abbreviations: CH, Common humanity; ISL, Isolation; ISMI, Internalized Stigma of Mental Illness; MF, Mindfulness; OI, Over-identification; RSFA, Resilience Scale for Adults; SCS, Self-compassion Scale; SJ, Self-judgement; SK, Self-kindness.
Sociodemographic and clinical characteristics
Table 2 summarizes sociodemographic and clinical data. In the study, 59.8% (n = 79) were women. The majority of the sample consisted of people between the ages of 36–50 (53%) and their education level was college and higher (31.8%). More than half (n = 77) were married. Also, 54.5% (n = 72) of them were living with their spouse/husband or child, 38.6% (n = 51) were living with their family, and 38.6% (n = 51) were working. The majority of participants (n = 116; 87.9%) had a duration of illness of 5 years or more. 68.9% (n = 91) had no history of suicide attempt. While 31 participants (23.5%) had never been hospitalized, 43.9% (n = 58) had more than one hospitalization.
| n | % | |
|---|---|---|
| Sex | ||
| Male | 53 | 40.2 |
| Female | 79 | 59.8 |
| Age | ||
| 18–35 | 36 | 27.3 |
| 36–50 | 70 | 53.0 |
| 51–65 | 26 | 19.7 |
| Educational level | ||
| Literate | 2 | 1.5 |
| Primary (5 years) | 30 | 22.7 |
| Secondary (8 years) | 18 | 13.6 |
| High school | 40 | 30.3 |
| College and above | 42 | 31.8 |
| Working status | ||
| Working | 51 | 38.6 |
| Not working | 81 | 61.4 |
| Marital status | ||
| Married | 77 | 58.3 |
| Single | 42 | 31.8 |
| Divorced/widow | 13 | 9.8 |
| With whom is living | ||
| Alone | 9 | 6.8 |
| Partner or child | 72 | 54.5 |
| Family | 51 | 38.6 |
| Insight | ||
| Present | 108 | 81.8 |
| None | 24 | 18.2 |
| Family psychiatric history | ||
| Present | 78 | 59.1 |
| None | 54 | 40.9 |
| Suicide attempt | ||
| Yes | 41 | 31.1 |
| No | 91 | 68.9 |
| Duration of illness | ||
| 0–2 years | 6 | 4.5 |
| 2–5 years | 10 | 7.6 |
| 5 years and more | 116 | 87.9 |
| Number of hospitalization | ||
| None | 31 | 23.5 |
| Once | 43 | 32.6 |
| More than once | 58 | 43.9 |
Bivariate correlations
According to Baron and Kenny (1986), in order to test the mediation model, significant correlations should be found between (1) the predictor (internalized stigma), and the mediator (self-compassion), (2) the mediator (self-compassion) and the outcome (psychological resilience), and (3) the predictor (internalized stigma) and the outcome (psychological resilience). As depicted in Table 3, significant correlations were found among study measures (all p's < .01) in predicted directions. In the first step, internalized stigma was positively correlated with “self-judgement”, “isolation” and “over-identification” (respectively r = .64; p < .01, r = .61; p < .01, r = .58; p < .01). It was also found to be negatively correlated with “self-kindness”, “common humanity” and “mindfullness” (respectively, r = −.51, p < .01, r = −.41, p < .01, r = −.51, p < .01). In the second step, it was determined that there was a negative relationship with “self-judgement”, “isolation” and “over-identification” (respectively, r = −.67; p < .01, r = −.62; p < .01, r = −.63, p < .01) and psychological resilience. On the other hand, “self-kindness, “common humanity” and “mindfullness” subdimensions of SCS were positively correlated with psychological resilience (r = .69; p < .01, r = .48; p < .01, r = .61, p < .01, respectively). In the last step, there was a negative relationship between psychological resilience and internalized stigma, r = −.62; p < .01.
| ISMI | RSFA | SK | SJ | CH | ISL | MF | OI | |
|---|---|---|---|---|---|---|---|---|
| ISMI | 1 | −.62** | −.51** | .64** | −.41** | .61** | −.51** | .58** |
| RSFA | 1 | .69** | −.67** | .48** | −.62** | .61** | −.63** | |
| SK | 1 | −.63** | .67** | −.55** | .82** | −.61** | ||
| SJ | 1 | −.32** | .78** | −.55** | .81** | |||
| CH | 1 | −.29** | .68** | −.31** | ||||
| ISL | 1 | −.47** | .72** | |||||
| MF | 1 | −.53** | ||||||
| OI | 1 |
- Abbreviations: CH, Common humanity; ISL, Isolation; ISMI, Internalized Stigma for Mental Illness; MF, Mindfulness; OI, Over-identification; RSFA, Resilience Scale for Adults; SJ, Self-judgement; SK, Self-kindness.
- ** p < .01.
Mediation analyses
The possible confounding influence of demographic variables (IVs) on key measures (DVs: ISMI, RSFA and SK; SJ, CH, ISL, MFNS, OI) were tested using a series of multivariate regression analyses. Categorical variables were turned into dummy variables. No significant relationships were found between demographic variables and key measures, so no covariate was entered in mediation analyses.
Determining the effect of internalized stigma on psychological resilience was found to be significant at the p < .001 level of the way tested in the established model. It was observed that internalized stigma had a negative effect on resilience (β = −.62). That is, one unit change in the level of internalized stigma caused −0.62-unit change in resilience. Accordingly, it was seen that the first hypothesis proposed by Baron and Kenny (1986) was fulfilled.
In the second step, a new model was established in which the sub-dimensions of self-compassion were included. The goodness of fit values of this analysed model were as such: ×2/df = 26.15, RMSEA = .44, NFI = .51, CFI = .51, GFR = .54. Not only the goodness of fit index values did not exhibit acceptable fit but also not all paths in the research model were found to be significant. For this reason, the recommended covariances within the framework of the modification indices were established and the relationships of the variables were examined, and non-significant paths were removed from the model, and the analysis was performed again. As a result of this analysis, the final version of the model with meaningful paths is shown in Figure 1.
After the improvements made in the model shown in Figure 1, the goodness-of-fit values of the research model, which includes meaningful paths, are as follows: ×2/df = 1.85, RMSEA = .08, NFI = .97, CFI = .99, GFR = .96. (See Table 4).
| ×2/df | RMSEA | NFI | CFI | GFI | |
|---|---|---|---|---|---|
| Research model | 1.85 | .08 | .97 | .99 | .96 |
It has been observed that the model's goodness-of-fit values have an acceptable level of fit. The regression weights of the significant paths in the model are shown in Table 5.
| Estimate (β) | S.E. (Standard error) | C.R. (Critical rate) | Standardized estimates | p | |
|---|---|---|---|---|---|
| MF ← ISMI | −1.265 | .186 | −6.805 | −.511 | .000*** |
| SJ ← ISMI | 1.548 | .161 | 9.596 | .642 | .000*** |
| SK ← ISMI | −.301 | .137 | −2.202 | −.124 | .028* |
| SK ← MF | .746 | .055 | 13.543 | .761 | .000*** |
| ISO ← ISMI | .474 | .178 | 2.656 | .185 | .008** |
| OI ← ISMI | 1.460 | .178 | 8.200 | .582 | .000*** |
| CH ← ISMI | −.943 | .190 | −4.972 | −.398 | .000*** |
| ISO ← SJ | .698 | .074 | 9.430 | .413 | .000*** |
| RSFA ← SK | .338 | .055 | 6.155 | −.250 | .000*** |
| RSFA ← ISMI | −.497 | .151 | −3.297 | −.258 | .000*** |
| RSFA ← SJ | −.213 | .062 | −3.425 | .658 | .000*** |
- Abbreviations: CH, Common humanity; ISMI, Turkish version of the Internalized Stigma of Mental Illness Scale; ISO, Isolation; MF, Mindfulness; OI, Over-identification; RSFA, Resilience Scale for Adults; SJ, Self-judgement; SK, Self-kindness.
- ***p < .001; **p < .01; *p < .05.
According to the results of the analysis of the final model, which was formed to examine whether the second and third assumptions proposed by Baron and Kenny (1986) were realized, the level of internalized stigma had a negative influence on the sub-dimensions of “common humanity”, “self-kindness” and “mindfullness” (respectively, β = −.40; p < .001, β = −.12; p < .05, β = −.51; p < .001), it was seen that “over-identification”, “isolation” and “self-judgement” dimensions were positively affected (respectively, β = .58; p < .001, β = .19; p < .01, β = .64; p < .001). In addition, it was seen that only the “self-kindness” and “self-judgement” had an effect on psychological resilience (respectively β = .66; p < .001, β = −.25; p < .001). For this reason, since the variables “common humanity”, “mindfullness”, “over-identification” and “isolation” did not meet the assumption that the mediator variable should have a predictive effect on the dependent variable, which is required for the mediator relationship, these variables were not included as mediator variables in the final model.
With the inclusion of the sub-dimensions of self-compassion in the model, the negative effect of internalized stigma on psychological resilience decreased in the first model (β = −.26; p < .001). Thus, it was seen that all the assumptions required for the mediation expressed by Baron and Kenny (1986) were met. Finally, it was examined whether “self-judgement” and “self-compassion” significantly mediated the relationship between internalized stigma and psychological resilience. Accordingly, the indirect effect between internalized stigma and resilience was obtained in the range of −.49 to −.26 (p < .05). Since this range did not include a 0 value, the indirect effect was found to be statistically significant. In this case, it was seen that “self-judgement” and “self-kindness” played mediator roles in the effect of internalized stigma on psychological resilience.
DISCUSSION
This is the first study to examine the relationship between internalized stigma, self-compassion and psychological resilience in euthymic bipolar patients. The results of this study reflect important implications for enriching psychosocial interventions for bipolar disorder.
Bivariate associations
The present study showed that there were positive correlations with negative dimensions of self-compassion (self-judgement, isolation and over-identification) and internalized stigma, also negative correlations with positive dimensions of self-compassion (self-kindness, common humanity and mindfulness). Neither Døssing et al. (2015) and Ranjbar et al. (2020) did not find any correlation between internalized stigma and self-compassion, Ranjbar et al. (2020) observed that dimensions of internalized stigma were correlated with self-compassion. So, their perspective was quite different from ours that we chose to look for the correlations with dimensions of self-compassion. However, our findings were in line with Heath et al. (2018) and Wong et al. (2019) who mentioned that, thanks to self-compassion, individuals who experience internalized stigma can approach themselves and situations with a more balanced perspective and accept their identities as they are, together with their positive and negative characteristics and outcomes. This may also be interpreted that negative experience like having internalized stigma may be modified through having more compassionate perspective.
Moreover, positive dimensions of self-compassion (self-kindness, common humanity, and mindfulness) were positively correlated with psychological resilience, and negative dimensions (self-judgement, isolation, and over-identification) had negative correlations with resilience. These findings imply that self-compassion could be a resilience mechanism (Trompetter et al., 2017). Self-compassion as a positive human strength, evokes kindness, provides more broadened and balanced awareness, and reinforces feelings related with inter-connectedness, so it can contribute to resilience (Keyes, 2005; Neff et al., 2007). Interestingly, correlation levels were quite similar for both positive and negative dimensions. This could be interpreted with the ‘bipolar continuum’ of self-compassion, that is, the opposite ends of the bipolar continuum [from compassionate self-responding (CS) (self-kindness, common humanity, mindfulness) to uncompassionate self-responding (UCS) (self-judgement, isolation, over-identification) had the same strength of associations (Neff, 2022a, 2022b).
Lastly, a negative relationship between internalized stigma and psychological resilience was found. Only one study (Post et al., 2021) was able to be reached to explore the relationship between resilience and stigma, particularly self-stigma and stigma resistance. A negative correlation between resilience and the self-stigma, also a positive correlation with the stigma resistance were obtained. Our finding was interpreted as being in line with the finding of negative relationship with resilience and self-stigma. Another study (Unal et al., 2022) indicated a negative correlation between internalized stigma and self-efficacy, self-esteem and social support which are related with psychological resilience. In this sense, resilience could be seen as a protective mechanism for bipolar disorder.
Our main aim for this research was to examine the mediating effects of the sub-dimensions of self-compassion between internalized stigma and psychological resilience. Self-kindness and self-judgement were found to play mediator roles in this relationship. This finding is meaningful that internalized stigma is about exerting negative stereotypes and criticizing attitudes towards oneself for having BD (Ellison et al., 2013), which leads to being more judgemental and critical about oneself. In line with this, Post et al. (2021) showed that there was a negative relationship between self-stigma and resilience. However, in our model self-kindness had more powerful impact on the relationship with internalized stigma and resilience. In other words, to alleviate the negative impact of internalized stigma on resilience, self-kindness may be boosted, aside from dealing with self-judgements related with internalized stigma. As Neff (2022a, 2022b) stated, self-kindness is more than ceasing self-criticism, which is quite common in clinical population (Shapero et al., 2015). Lowens (2010) pointed out that individuals with BD tend to use self-attacking strategy to motivate their goal-directed behaviours. This is a common strategy when they feel powerless and desperate on account of dependence on medication, when they fear of potential future relapses, when they recover not as fast as they would like to see, all are related with internalized stigma. However, self-kindness is about taking an active interest in our distress. In our context, self-kindness implies being emotionally available when life gets hard due to stigmatization. So, responding to ourselves with warmth makes us feel supported, encouraged and validated (Neff, 2022a, 2022b). By this way, self-compassion can help a person with BD use healthier ways to reach their goals and establish more realistic expectations (Yang et al., 2018).
Clinical implications
This research is grounded in the positive clinical psychology perspective (Wood & Johnson, 2016). Regarding this perspective, the focus of the psychological treatment process should be to increase the positive characteristics as well as to reduce the negative ones (Wood & Tarrier, 2010). Within the positive clinical psychology, risk (internalized stigma) can be located at one end of the continuum, so resilience at the other end (Johnson et al., 2011). Positive characteristics can act as a buffer against the effects of negativities that cause distress and stress; also, positive traits (like being self-compassionate) can be strengthened to increase psychological resilience (Wood & Tarrier, 2010). Psychological resilience can be seen as a developmental process that includes the skills to use internal and external resources by adapting to difficulties in a positive way rather than being well in the face of difficulties (Yates et al., 2003). Thus, adversity (so, internalized stigma) can be seen as a factor for enhancing psychological resilience (Luthar & Cicchetti, 2000). According to Mooney and Padesky (2000), focusing on building resilience in psychotherapy is more effective than focusing on behaviours and thoughts that interfere with resilience. Therefore, depending on Bonanno's (2004) conceptualisation, to improve the psychological resilience of people with BD experiencing internalized stigma may be helped by enhancing their social (social support, communication skills), emotional (hope, self-compassion, optimism, self-respect, self-efficiency) and cognitive (cognitive flexibility, emotion regulation, problem-solving) skills. In our study, we just focused on self-compassion, which is about being kind and understanding towards oneself in hard times, seeing difficulties as part of a common experience, not tending to over-identification with negative experiences, and not feeling isolated by distress and emotional pain (Gilbert, 2009; Neff, 2003). In the future, both research can be done to explore the relationships between these variables, and also psychosocial interventions could focus on the mentioned variables to dampen the effect of internalized stigma and to boost psychological resilience in bipolar disorder.
While integrating compassion into psychosocial treatment for BD, Lowens (2010) stated that self-concept is excessively evaluated in both negative and positive sense on the perceived social comparison between superiority and inferiority. In line with this, Power (2005) and Potter (2013) mentioned about over-attachment to self-concept in BD being contradictory and fragmented. In their psychological model for BD, Mansell et al. (2007) pointed out that multiple, and excessive self-directed appraisals make management of behaviours harder, so mood states become escalated. To regulate these self-directed appraisals, nonattachment to self might be cultivated through dampening the over-active self-related appraisals and/or managing dysregulated self-perceptions through decreasing the inclination to engage in maladaptive coping behaviours. So, these biased appraisals might be buffered by self-compassion (Yang et al., 2018).
Limitations and future suggestions
The present study has a cross-sectional design and relatively small sample size. So, the generalizability of the results is limited and longitudinal investigation might be suggested for causal explanations (Kazdin, 2007). Patients with acute symptoms or patients at different stages of illness might experience internalized stigma, resilience, or self-compassion differently. Data were collected only from the outpatient clinic of the University of Health Sciences Bakirkoy Dr. Mazhar Osman Mental and Neurological Diseases Education and Research Hospital. So, larger and more representative samples of BD patients might further confirm the obtained findings. Unfortunately, part of the data collection process of this study coincided with the initial phase of the Covid-19 process. This made data collection harder, also it is very possible that this situation affected the psychological processes of the participants. For this reason, anxiety levels could also be evaluated. Data about the clinical characteristics of the patients were created with the information obtained from them, their medical files, and their relatives. Obtaining information retrospectively might limit the reliability of this information. In addition, the use of a self-report scale can be considered as another limitation. Since this study did not have any control group or could not allow for comparison between BB-I and BB-II patients, future studies could make comparisons with either different sub-types of BD or other psychiatric diagnoses like schizophrenia or MDB. In addition, only one mediator was examined; multiple mediation models, including insight, emotion regulation, mindfulness, or any other related variables, would allow determining their relative effects on the relationship between internalized stigma and resilience, which might empower psychosocial interventions for BD. Finally, mixed-method research might also be preferred, thus, a qualitative part could be added to explore how internalized stigma, self-compassion, and resilience are made sense in the participants' experiences.
CONCLUSION
The current study indicated that decreased level of internalized stigma was associated with improving resilience. In this relationship, self-compassion, particularly self-kindness and self-judgement, could be a beneficial target in psychosocial interventions for enhancing resilience and decreasing the effect of internalized stigma. Research findings might make a contribution to the growing research on positive psychology concepts in understanding bipolar disorder. This study was a strengths-based investigation in BD to target to alleviate the effect of internalized stigma. Also, the findings of the study shed light on which dimensions of self-compassion (self-kindness and self-judgement) might be useful to consider in psychotherapeutic studies for bipolar disorder, and would be valuable to enhance psycho-social interventions for prevention of relapse in BD.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
AUTHOR CONTRIBUTIONS
Zeynep Mackali: Conceptualization; methodology; writing – original draft; project administration; supervision. Saadet Cetinkaya: Writing – original draft; investigation; formal analysis; resources. Nur Ay: Methodology; formal analysis.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
Full text views and downloads on Wiley Online Library.
` document.querySelector(".tooltip__body a")?.addEventListener("mousedown",(e)=>{ window.location.href = e.target.href } )}