HPQ0010.1177/1359105317702212Journal of Health PsychologyPalmeira et al. 702212research-article2017

Article

Journal of Health Psychology 1 –10 The role of self- in eating © The Author(s) 2017 Reprints and permissions: psychopathology in overweight sagepub.co.uk/journalsPermissions.nav DOI:https://doi.org/10.1177/1359105317702212 10.1177/1359105317702212 and obesity: Can self-compassion journals.sagepub.com/home/hpq be useful?

Lara Palmeira1, José Pinto-Gouveia1 and Marina Cunha1,2

Abstract This study explores the relationship between self-disgust and eating psychopathology and whether self- compassion plays a mediator role on this relationship. Participants were 203 adults, from both genders,

with overweight and obesity (MBMI = 31.17, standard deviation = 5.43). Women reported higher levels of self-disgust and eating psychopathology and lower levels of self-compassion than men. Path analysis results suggested that the effect of self-disgust on eating psychopathology occurred partially through one’s inability to be self-compassionate. The results highlight the damaging role of self-disgust on eating psychopathology and the importance of developing a more compassionate attitude towards the self to promote healthy eating behaviors.

Keywords eating psychopathology, overweight and obesity, path analysis, self-compassion, self-disgust

Introduction Evolutionarily, disgust refers to feelings of revul- social deviance and immoral behavior, interper- sion (Rozin et al., 1999) and is an aversive response sonal relationships (being disgusted by other’s rooted in the threat-protection system that signals behaviors or presentations), and self-judgments possible contamination or attacks, aiming to avoid (regarding one’s body, emotions, thoughts, behav- diseases and protect us from ingesting harmful or iors, or even the self; Chapman and Anderson, poisoned substances (Gilbert, 2015; Oaten et al., 2009). It motivates avoidance-based responses, such as moving away, getting rid of, or eradicating 1CINEICC, Universidade de Coimbra, Portugal 2 what is dangerous or disgusting (Rozin et al., Miguel Torga Superior Institute (ISMT), Portugal 1999). Nevertheless, disgust evolved to be a multi- Corresponding author: faceted and heterogeneous emotional construct Lara Palmeira, CINEICC, Faculdade de Psicologia e shaped through sociocultural learning, which can Ciências da Educação, Universidade de Coimbra, Rua do Colégio Novo, Apartado 6153, 3001-802 be prompted by different stimuli (Power and Coimbra, Portugal. Dalgleish, 2008; Rozin et al., 1999), including Email: [email protected] 2 Journal of Health Psychology 00(0)

2012; Gilbert, 2015; Ille et al., 2014; Overton et al., 2012). Nonetheless, individuals can still engage 2008), and lead to distinct responses. Although the in avoidance behaviors, such as avoiding touch- disgust response serves an important and adaptive ing or looking at themselves or being looked by function in the external world, feelings of disgust others, masking the disgusting aspects of them- may be generalized and directed toward the self selves or even trying to distract themselves when some aspects are seen as toxic, repugnant, from the object of disgust (Espeset et al., 2012; and dangerous (Gilbert, 2015; Overton et al., 2008; Powell et al., 2013). These avoidance-based Power and Dalgleish, 2008). strategies tend to yield paradoxical effects and Self-disgust or self-loathing is considered to contribute to increase or maintain the disgust arise from the self-directed generalization of response (Powell et al., 2015). Given that self- the basic disgust response and relates to feel- disgust is particularly enduring and that avoid- ings of aversion, deep grief, or even repugnance ance-based strategies seem to produce unwanted toward the one’s self, reflecting a noxious, effects, individuals with self-disgust may harsh, and embodied feeling state (Roberts and become somehow trapped with the desire to get Goldenberg, 2007). It is considered a stable and rid of those parts considered disgusting (Powell dysfunctional psychological phenomenon that et al., 2013). encloses two interrelated domains of the self Until now, research on self-disgust is still (physical and behavioral) and that is intrinsi- scant and its relationship with psychopathol- cally dependent on one’s social environment ogy needs to be further explored. Nonetheless, (Powell et al., 2013). Even though self-disgust recently, in a longitudinal study, self-disgust origins are yet to be fully understood, it is likely has been found to be involved in the genesis of that it arises from social learning experiences depressive experience, as an important predic- such as disgust-related criticism, internalization tor of depressive symptoms (Powell et al., of other’s disgust reactions, and negative social 2013). Self-disgust has also been related to comparisons, where the individuals learn what unhealthy eating behaviors (e.g. restrictive eat- attributes others consider physically or socially ing, purging, and vomiting), body dissatisfac- repulsive (Powell et al., 2015). Moreover, self- tion, urge to lose weight, and being often disgust is intrinsically linked to other threaten- directed at undesirable and unattractive body based emotions, such as , , and shame features (Espeset et al., 2012; Powell et al., that when directed toward one’s self may be 2014; Shanmugarajah et al., 2012). Moreover, extremely pathogenic (Gilbert, 2015; Powell evidence suggests that individuals with eating et al., 2015). It is unclear whether self-disgust disorders may use unhealthy eating behaviors represents a severe form of shame or whether to avoid or regulate negative affect and diffi- shame and self-disgust are variants of the dis- cult emotions, such as shame, anger, and sad- gust response (Roberts and Goldenberg, 2007). ness (Espeset et al., 2012; Fox et al., 2015; Nevertheless, several authors have been claim- Power & Dalgleish, 2008). In a study compar- ing that self-disgust reflects an important and ing patients with different psychological prob- distinct construct that deserves more empirical lems and healthy controls, Ille et al. (2014) attention (Powell et al., 2015; Roberts and found that those with psychological problems Goldenberg, 2007). showed higher levels of self-disgust. Among According to Gilbert (2005), individuals use those with mental problems, those with border- identical psychological mechanisms to deal line personality and eating disorders presented with both external and internal attacks and the highest self-disgust levels. Furthermore, threats. Although disgust tends to elicit the urge evidence suggests that self-disgust plays a cru- to get rid of, avoid, or reject what is considered cial role in eating disorders’ maintenance, elic- disgusting, this can be particularly difficult iting food, eating, and body-related stimuli when the stimulus that elicits the disgust feeling avoidance in women (Espeset et al., 2012; Fox is a part (or the whole) of self (Espeset et al., and Power, 2009). However, as far as we know, Palmeira et al. 3 the role of self-disgust in individuals with This study aims to explore the associations overweight and obesity has never been studied. between self-disgust, self-compassion, and eat- Nevertheless, unveiling the role of self-disgust ing psychopathological symptoms in individu- in this population may be especially important als with overweight and obesity. We expect that given that in modern society, physical appear- self-disgust would be positively related to eat- ance (particularly for women) has become a ing psychopathology and negatively related to major source of social acceptance (Gilbert self-compassion. Gender differences concern- et al., 1995) and is something that cannot be ing all studies’ variables were also explored easily hidden from others. Thus, presenting a since literature has been highlighting that body that is different from the one that is women are more prone to present eating psy- socially valued and that others may view as chopathology than men (e.g. Buchanan et al., disgusting can be threatening and become 2013). Finally, this study also investigated internalized, leading to feelings of shame and whether the ability to be self-compassionate self-disgust. In turn, this increases one’s vul- mediated the relationship between self-disgust nerability to engage in disordered eating pat- feelings and eating psychopathological symp- terns (Fox et al., 2015). Gilbert (2015) proposes toms in individuals with overweight and obe- that self-disgust encloses the desire to avoid sity, while controlling gender and BMI. the object of disgust, to become an acceptable and valued self within the social context. Methods On the other hand, research has been consist- ently highlighting the importance of developing Participants self-compassion, that is, having a warm, kind, and accepting relationship with one’s self as a The sample comprised 203 individuals (50.2% protective emotional regulation process in men- males and 49.8% females) with overweight and tal health and well-being (MacBeth and obesity (MBMI = 31.17; SDBMI = 5.43) seeking nutri- Gumley, 2012; Neff, 2003). However, individu- tional treatment for weight loss in several public als who struggle with eating and weight prob- and private health care units in Portugal. Inclusion lems seem to present difficulties in being criteria were being an adult (age > 18 years old) 2 self-compassionate toward themselves, espe- and having a BMI ⩾25 kg/m . Participants pre- cially when facing setbacks or failures (Adams sented a mean age of 40.08 years (standard devia- and Leary, 2007; Gilbert et al., 2014). tion (SD) = 11.74 years), with a mean of 12.67 Evidence suggests that self-compassion may (SD = 3.74) years of education. No gender differ- be linked to eating pathology in multiple ways ences were found for age, t(201) = −1.228, p = .221, (Braun et al., 2016). Self-compassion has been years of education, t(201) = 1.463, p = .145, and BMI found to mediate the relationship between body t(201) = −0.451, p = .653. mass index (BMI), shame, body dissatisfaction, Concerning marital status, 55.1 percent of the body image–related unfavorable social com- participants were married or living together, parisons, eating psychopathological symptoms, 36 percent were single, 8.4 percent were divorced, and quality of life in females from clinical and and 0.5 percent were widowed. The majority non-clinical samples (e.g. Duarte et al., 2015; (68.5%) came from low to medium socio-eco- Ferreira et al., 2013). A large German study nomic status. (N = 1158), with individuals with overweight and obesity, found that self-compassion medi- Procedures ated the relationship between weight self- stigma and global health (Hilbert et al., 2015). Before data collection, ethical approval was Taken together, these studies emphasize self- obtained from all institutions that participated in compassion abilities as key resources for the the study. Participants were invited to the study adoption of healthy behaviors. by their nutritionist on the day of their nutritional 4 Journal of Health Psychology 00(0) appointment. In the first page of the protocol, the vs. isolation, and mindfulness vs. over-identifi- study’s goals as well as the voluntary and confi- cation) All items are rated on a 5-point Likert dential nature of the data were stated. Participants scale (1 = almost never to 5 = almost always). were required to sign an informed consent before The total score reflects the mean of all items, completing the self-reported questionnaires. The with higher values indicating more self-com- study’s protocol took approximately 15 minutes passion. Both the original (α = .92) and the to be completed. Portuguese version (α = .94) showed very good internal consistency (Castilho et al., 2015; Neff, Measures 2003). Eating Disorder Examination Questionnaire Demographic data. Participants reported their (EDE-Q; Fairburn and Beglin, 1994; Machado age, educational level, current height, and et al., 2014) is a well-known 36-item self-report weight. Participants were asked to report the instrument that measures eating psychopatho- weight of their current or previous nutritional logical symptoms using a 6-point rating scale. appointment. Then BMI (weight/height2) was EDE-Q has been consistently considered a calculated. reliable measure of eating psychopathology Multidimensional Self-Disgust Scale (Fairburn, 2008). The Portuguese version also (MSDS; Carreiras, 2014) is a self-report meas- presented very good internal consistency (α = ure with 33 items that assesses self-disgust con- .94; Machado et al., 2014). In this study, we cerning physical, behavioral, and functioning only used the EDE-Q global score as we were aspects. It encloses four subscales: defensive interested in capturing eating psychopathologi- activation (physiological component; “I have cal symptoms’ severity. the feeling my body contracts”), cognitive- emotional (cognitive and emotional compo- Data analysis nent; “I hate/despise that part of me”), avoidance (behavioral component; “I disguise/dissimulate IBM SPSS Statistics 20 and AMOS software those aspects of me that I disgust”), and exclu- were used to perform all data analysis. Gender sion (behaviors used to eliminate and exclude variable was dummy coded as 0 = female and disgusting characteristics of the self; “I feel like 1 = male. Preliminary data analyses were exe- cutting, burning, or excluding that part of cuted to explore the adequacy of the data. myself”). Participants rate the frequency they Pearson’s correlation coefficients were calcu- experience for each item on a 5-point Likert lated to explore the associations between BMI, scale (0 = never and 4 = always). In the original self-disgust (cognitive-emotional subscale), study, conducted in a Portuguese sample, all self-compassion, and eating psychopathologi- subscales showed good internal consistency cal symptoms. Then independent t-tests were (α = .95 for defensive activation, α = .97 for cog- performed to explore gender differences in all nitive-emotional subscale, α = .77 for exclusion, studies’ variables (Field, 2013). Effect sizes and α = .84 for avoidance), good convergent, were calculated with Cohen’s d. Following and divergent validities (Carreiras, 2014). For Cohen’s guidelines (1988 cited in Tabachnick this study, only the cognitive-emotional sub- and Fidell, 2007), values <.4 were considered scale was used to assess participant’s disgust small effects, from .5 to .7 medium effects, and thoughts and feelings toward the self. >.8 large effect sizes. Self-Compassion Scale (SCS; Castilho et al., Finally, to explore the mediator effect of self- 2015; Neff, 2003) is a 26-item self-report meas- compassion on the relationship between self-dis- ure that assesses the tendency to be compas- gust and eating psychopathological symptoms, sionate toward the self when facing setbacks. while controlling for BMI and gender, a path The SCS encloses six subscales that measure analysis was used. Path analysis allows the simul- three components of self-compassion (self- taneous examination of structural relationships as kindness vs. self-judgment, common humanity well as the examination of direct and indirect Palmeira et al. 5

Table 1. Means (M), SDs, Alpha coefficients, and Pearson’s moment correlation for all studies’ variables (N = 203).

Measures M SD α 1 2 3 4 1. Gender – – – – 2. BMI 31.17 5.43 – .03 – 3. Self-disgust 8.04 7.44 .94 −.18** .27*** – 4. Self-compassion 3.25 0.60 .92 −.28*** −.11 −.59*** – 5. EDE-Q 1.77 1.25 .94 .21** .37*** .65*** −.48***

SD: standard deviation; BMI: body mass index; EDE-Q: Eating Disorder Examination Questionnaire. **p < .01; ***p < .001. paths (e.g. Schumacker and Lomax, 2004). self-disgust and EDE-Q and a low and positive Maximum likelihood method was used since it correlation with self-compassion. BMI showed allows for the estimation of all path coefficients positive and low associations with self-disgust and computes fit statistics. Model fit was assessed and EDE-Q. BMI was not significantly related using several goodness-of-fit measures and rec- to self-compassion. Self-disgust was nega- ommended cut-points: chi-square (χ2), normed tively and moderately related to self-compas- chi-square (χ2/degree of freedom (df)), compara- sion and positively and moderately associated tive fit index (CFI ⩾ .95, desirable; Hu and with EDE-Q. Finally, self-compassion was Bentler, 1998), goodness-of-fit index (GFI ⩾ .95, negatively and moderately correlated with desirable; Jöreskog and Sörbom, 1996), and root EDE-Q. mean square error of approximation (RMSEA ⩽ .08, acceptable fit; Kline, 2005) with a 95 per- Gender differences cent confidence interval (CI). The mediation effect was examined using bootstrap (2000 resa- To explore differences regarding BMI, self- mples) with 95 percent bias-corrected CI. The disgust, self-compassion, and eating psycho- effect is statistically significant at p <.05 if zero is pathological symptoms between females and not included in the interval between the lower and males, independent samples t-test were per- the upper bounds (Kline, 2005). formed. Means, SDs, t-test differences, and Cohen’s d for all variables for each gender are displayed in Table 2. As can be seen, signifi- Results cant differences were found for all studies’ Preliminary data analyses variables, except for BMI (t(201) = −0.451, p = .653; d = .06). Overall, females reported All variables showed acceptable skewness and higher levels of self-disgust (t(201) = 2.625, kurtosis values below the recommend cut-points p = .009; d = .37) and eating psychopathological (SK < |3| and Ku < |8–10|). Also, multicollinearity symptoms (t(201) = 4.116, p ⩽ .001; d = .58). was not identified as all variables had variance On the contrary, men presented higher levels of inflation factor (VIF) <5. Finally, Mahalanobis self-compassion abilities (t(201) = −3.076, p = distance statistic (D2) did not detect the presence .002; d = .43). All differences represent small of any outliers (Kline, 2005). effect sizes, with the exception for eating psy- chopathological symptoms, where the effect Correlation analysis size was medium. Pearson’s correlation analyses are displayed in Path analysis Table 1. The results reveal that gender and BMI were not significantly correlated. Gender also Path analysis was conducted to test the mediational showed negative and low associations with role of self-compassion on the relationship between 6 Journal of Health Psychology 00(0)

Table 2. Means (M), SDs, t-test differences by gender for all variables, and Cohen’s d effect size (N = 203).

Female (n = 101) Male (n = 102) t (df) P Cohen’s d M SD M SD BMI 31.00 5.42 31.34 5.46 −0.451 (201) .653 .06 Self-disgust 9.40 8.40 6.70 6.56 2.625 (201) .009 .37 Self-compassion 3.12 0.61 3.38 0.56 −3.076 (201) .002 .43 EDE-Q 2.12 1.23 1.42 1.19 4.116 (201) <.001 .58 df: degree of freedom; SD: standard deviation; BMI: body mass index; EDE-Q: Eating Disorder Examination Questionnaire.

Figure 1. Graphic representation of the mediation model (N = 203). **p ⩽ .01; ***p ⩽ .001; standardized path coefficients are presented. self-disgust and eating psychopathological symp- based on 95% CI: .007 to .159, p = .033). The toms, while controlling for the effect of BMI and total effect (which represents the sum of the gender. The model tested contained 15 parameters. standardized direct and indirect effects) of self- As fully saturated models always have a perfect disgust on EDE-Q was β = .551 based on 95 per- model fit, model fit indices were neither examined cent CI (.438 to .640, p = .001). Moreover, there nor reported. All path coefficients were statistically was a significant direct effect of self-disgust on significant with the exception of the direct path self-compassion (β = −.586, based on 95% CI: from BMI → self-compassion (b = .005; SE = .006; −.666 to −.498, p = .001) and a significant direct Z = 0.718; p = .473) and the direct path from gen- effect of self-compassion on EDE-Q (β = −.146, der → self-compassion (b = .129; SE = .068; Z = based on 95% CI: −.272 to −.012, p = .033). 1.888; p = .059) that were non-significant and were Also, BMI (β = .237, based on 95% CI: .113 to progressively removed. The final model (Figure 1) .338, p = .001) and gender (β = −.172, based on presented a very good model fit: χ2(2) = 4.049, 95% CI: −.267 to −.068, p = .001) both pre- p = .473; χ2/df = 2.025; CFI = .992; GFI = .992; sented significant direct effects on EDE-Q. The RMSEA = .071 (CI = .000 to .172); p = .266. final model accounted for 34 percent of self- Self-disgust was significantly correlated compassion and 51 percent of EDE-Q. with both BMI (r = .27, p ⩽ .001) and gender (r = −.18, p = −011). Concerning the meditation Discussion analysis, the results showed that self-disgust had a significant direct (β = .465 based on 95% Self-disgust has been described as an enduring CI: .326 to .590, p = .001) and an indirect effect feeling of aversion, deep grief, or repugnance on EDE-Q through self-compassion (β = .085 toward some parts of the self (e.g. physical, Palmeira et al. 7 psychological, or behavioral) that stimulates and obesity. It seems that individuals who avoidance-based responses (Roberts and experience more self-disgust-related thoughts Goldenberg, 2007). Especially when directed at and emotions present more eating psychopath- relatively stable attributes of the self (e.g. body ological symptoms. This is in line with the weight), self-disgust seems to be particularly existent studies with patients with anorexia and maladaptive and associated with disordered eat- bulimia nervosa that suggest that self-disgust is ing (Espeset et al., 2012; Fox and Power, 2009). closely linked to body-image dissatisfaction, However, research on self-disgust is still in its avoidance, tendency to restrict food intake, and early stages. As far as we know, this is the first purging behaviors (Espeset et al., 2012; Powell study that explored self-disgust in a sample of et al., 2014). Furthermore, these findings give individuals with overweight and obesity seek- empirical support to Gilbert’s (2015) idea that ing treatment for weight loss. trying to avoid or get rid of what is considered The results revealed that self-disgust was disgusting enfolds the wish to become accepted positively and moderately associated with eat- and valued in the eyes of others. Given the fact ing psychopathological symptoms, whereas that one’s weight and physical appearance is BMI only revealed positive but low associa- easily judged by others and that, especially for tions with self-disgust and eating psychopatho- women, physical appearance is a crucial ele- logical symptoms. This result mirrors the results ment for social acceptance, being considered found in previous studies with non-clinical and overweight may present a threat to one’s social eating disorders samples which showed that acceptance and status (Gilbert et al., 1995; self-disgust was related to restrictive eating, Simpson et al., 2010). In fact, Gilbert (2015) purging, body dissatisfaction, and urge to lose argued that if the self, or parts of the self, are weight (Espeset et al., 2012; Powell et al., seen as disgusting by others (e.g. weight or 2013). Contrarily, and as expected, the ability to physical appearance), this view can become have a warm and caring relationship with one- built in the self-system. Thus, self-disgust feel- self was negatively associated with self-disgust ings may lead people to increase their weight and eating psychopathological symptoms and control strategies and become more preoccu- was not significantly related to BMI. pied with their weight, eating, and body image Consistent with previous findings, women as a way to be accepted and valued in their presented higher levels of self-disgust and eat- social context. ing psychopathological symptoms and fewer Consistent with previous research (Buchanan self-compassion abilities than men. All gender et al., 2013; Ferreira et al., 2014), having a differences reflected small effects, except the higher BMI and being female (although weakly) medium effect size found for eating psychopath- were directly related to higher levels of eating ological symptoms. In fact, the literature sug- psychopathological symptoms. The fact that in gests that women are more self-critic and have our male sample the EDE-Q mean score was fewer self-compassion skills than men (Neff, below the cut-off for eating psychopathology 2003; Yarnell et al., 2015) and are more vulner- may also account for the results found. able to eating psychopathology (Buchanan Moreover, in our model, the relationship et al., 2013). Likewise, although less studied, between self-disgust and eating psychopatho- women present higher levels of self-disgust than logical symptoms was partially mediated by men (e.g., Ille et al., 2014). These findings sug- individual’s inability to adopt a compassionate gest that women are more prone to develop a attitude toward themselves. Overall, the model negative and harmful internal relationship and tested accounted for 34 percent of self-compas- may require differentiated interventions. sion and 51 percent of eating psychopathologi- In addition, this study highlights the role of cal symptoms. This result points out that the self-disgust on eating psychopathological effect of feeling disgust toward the self on dis- symptoms in people living with overweight ordered eating symptomatology seems to occur 8 Journal of Health Psychology 00(0) partially through the difficulty in accepting and internal experiences, instead of being judgmental having a warm and kind attitude toward oneself and becoming overidentified with them. In turn, and one’s imperfections. This is noteworthy as this may help decrease their eating psychopatho- once acquired, self-disgust is considered hard to logical symptoms. unlearn and that avoidance-based strategies tend to be ineffective (Powell et al., 2015). Acknowledgements To sum up, this study highlights the harmful The authors would like to thank the nutritionists who role of self-disgust on eating psychopathologi- collaborated with the data collection and all partici- cal symptoms in people living with overweight pants who took part in the study. and obesity. It also points out that being self- compassionate when facing failures or errors Declaration of conflicting interests can be a useful resource in self-disgust and eat- The author(s) declared no potential conflicts of inter- ing psychopathology. Still, research on self- est with respect to the research, authorship, and/or disgust is recent and more studies are needed to publication of this article. better understand its origins and impact on the lives of people living with overweight and obe- Funding sity. Future studies could continue to explore The author(s) disclosed receipt of the following the impact of self-disgust particularly on binge financial support for the research, authorship, and/or eating and quality of life. publication of this article: This research was sup- This study encloses some limitations. This is a ported by the first author’s PhD grant (SFRH/ cross-sectional study, which precludes conclusions BD/84452/2012), sponsored by FCT (Portuguese regarding causality. Clearly, longitudinal studies Foundation for Science and Technology). are needed to determine the directionality of the associations found. Likewise, we relied on partici- References pants’ self-reports (including height and weight), Adams CE and Leary MR (2007) Promoting self- which may be biased and influenced the results. compassionate attitudes toward eating among Additionally, we were unable to assess participant’s restrictive and guilty eaters. Journal of Social medical and psychiatric history or the existence of and Clinical Psychology 26: 1120–1144. eating disorders. Future studies should also take Braun TD, Park CL and Gorin A (2016) Self- these aspects into account as they may influence the compassion, body image, and disordered eat- results found. Finally, the model tested is limited as ing: A review of the literature. Body Image 17: 117–131. it is likely that other variables (e.g. body dissatisfac- Buchanan NT, Bluestein BM, Nappa AC, et al. tion, self-criticism) and other emotional regulation (2013) Exploring gender differences in body processes (e.g. decentering, experiential avoidance, image, eating pathology, and sexual harass- cognitive fusion) may be involved in the relation- ment. Body Image 10: 352–360 ship between self-disgust and eating psychopatho- Carreiras D (2014) The toxicity of the self: logical symptoms. However, we intentionally Developing a new measure and testing a com- restrained this model to specifically explore the role prehensive model of the nature of self-disgust. of self-disgust and self-compassion abilities. Unpublished Master Dissertation, University of In conclusion, this study offers new insights Coimbra. Available at: https://estudogeral.sib. for future research on self-disgust and stresses uc.pt/handle/10316/27653 the importance of assessing and targeting self- Castilho P, Pinto-Gouveia J and Duarte J (2015) Evaluating the multifactor structure of the disgust—an often neglected emotion—in adults long and short versions of the self-compassion (especially women) with overweight and obesity scale in a clinical sample. Journal of Clinical seeking treatment. Furthermore, our findings Psychology 71(9): 856–870. suggest the importance of fostering self-compas- Chapman HA and Anderson AK (2012) sion skills, so individuals may develop a more Understanding disgust. Annals of the New York detached and accepting relationship with their Academy of Sciences 1251: 62–76. Palmeira et al. 9

Duarte C, Ferreira C, Trindade IA, et al. (2015) Body Gilbert P, Price J and Allan S (1995) Social compari- image and college women’s quality-of-life: The son, social attractiveness and evolution: How importance of being self-compassion. Journal might they be related? New Ideas in Psychology of Health Psychology 20(6): 754–764. 13: 149–165. Espeset EMS, Gulliksen KS, Nordbø RHS, et al. Hilbert A, Braehler E, Schmidt R, et al. (2015) Self- (2012) The link between negative emotions compassion as a resource in the self-stigma and eating disorder behaviour in patients with process of overweight and obese individuals. anorexia nervosa. European Eating Disorders Obesity Facts 8: 293–301. Review 20: 451–460. Hu L and Bentler PM (1998) Fit indices in covariance Fairburn CG (2008) Cognitive Behavior Therapy and structure modeling: Sensitivity to underparam- Eating Disorders. New York: Guilford Press. eterized model misspecification. Psychological Fairburn CG and Beglin S (1994) Assessment of Methods 3: 424–453. eating disorders: Interview or self-report ques- Ille R, Schögglb H, Kapfhammer H, et al. (2014) tionnaire? International Journal of Eating Self-disgust in mental disorders—Symptom- Disorders 16(4): 363–370. related or disorder-specific? Comprehensive Ferreira C, Palmeira L and Trindade IA (2014) Psychiatry 55: 938–943. Turning eating psychopathology risk fac- Jöreskog KG and Sörbom D (1996) LISREL 8 tors into action: The pervasive effect of body User’s Reference Guide. Chicago, IL: Scientific image-related cognitive fusion. Appetite 80(1): Software International. 137–142. Kline RB (2005) Principles and Practice of Ferreira C, Pinto-Gouveia J and Duarte C (2013) Structural Equation Modeling (2nd edn). New Self-compassion in the face of shame and body York: Guilford. image dissatisfaction: Implications for eating MacBeth A and Gumley A (2012) Exploring com- disorders. Eating Behaviors 14: 207–210. passion: A meta-analysis of the association Field A (2013) Discovering Statistics Using IBM between self-compassion and psychopathol- SPSS Statistics (4th edn). Los Angeles, CA: ogy. Clinical Psychology Review 32(6): 545– SAGE. 552. Fox JR and Power MJ (2009) Eating disorders and Machado P, Martins C, Vaz A, et al. (2014) multi-level models of emotion: An integrated Eating Disorder Examination Questionnaire: model. Clinical Psychology & Psychotherapy Psychometric properties and norms for the 16: 240–267. Portuguese population. European Eating Fox JR, Grange N and Power MJ (2015) Self- Disorders Review 22(6): 448–453. disgust in eating disorders: A review of the lit- Neff K (2003) The development and validation of erature and clinical implications. In: Powell PA, a scale to measure self-compassion. Self and Overton PG and Simpson J (eds) The Revolting Identity 2(3): 223–250. Self: Perspectives on the Psychological, Social, Oaten M, Stevenson RJ and Case TI (2009) Disgust as and Clinical Implications of Self-Directed a disease-avoidance mechanism. Psychological Disgust, pp. 167–186. London: Karnac Books Bulletin 135: 303–321. Gilbert J, Stubbs J, Gale C, et al. (2014) A qualitative Overton PG, Markland FE, Taggart HS, et al. (2008) study of the understanding and use of “compas- Self-disgust mediates the relationship between sion focused coping strategies” in people who dysfunctional cognitions and depressive symp- suffer from serious weight difficulties. Journal tomatology. Emotion 8(3): 379–385. of Compassionate Health Care 1: 9. Powell PA, Overton PG and Simpson J (2014) The Gilbert P (2005) Compassion: Conceptualizations, revolting self: An interpretative phenomenolog- Research and Use in Psychotherapy. London: ical analysis of the experience of self-disgust in Routledge. females with depressive symptoms. Journal of Gilbert P (2015) Self-disgust, self-hatred and Clinical Psychology 70(6): 562–578. compassion focused therapy. In: Powell PA, Powell PA, Simpson J and Overton PG (2013) When Overton PG and Simpson J (eds) The Revolting disgust leads to dysphoria: A three-wave longi- Self: Perspectives on the Psychological, Social, tudinal study assessing the temporal relationship and Clinical Implications of Self-Directed between self-disgust and depressive symptoms. Disgust, pp. 233–254. London: Karnac Books. Cognition & Emotion 27(5): 900–913. 10 Journal of Health Psychology 00(0)

Powell PA, Simpson J and Overton PG (2015) An divinity). Journal of Personality and Social introduction to the revolting self: Self-disgust Psychology 76: 574–586. as an emotion schema. In: Powell PA, Overton Schumacker RE and Lomax RG (2004) A Beginner’s PG and Simpson J (eds) The Revolting Self: Guide to Structural Equation Modeling Perspectives on the Psychological, Social, and (2nd edn). Mahwah, NJ: Lawrence Erlbaum Clinical Implications of Self-Directed Disgust, Associates. pp. 1–24 London: Karnac Books. Shanmugarajah K, Gaind S, Clarke A, et al. (2012) Power MJ and Dalgleish T (2008) Cognition and The role of disgust emotions in the observer Emotion: From Order to Disorder (2nd edn). response to facial disfigurement. Body Image 9: Hove: Psychology Press. 455–461. Roberts TA and Goldenberg JL (2007) Wrestling Simpson J, Hillman R, Crawford T, et al. (2010) with nature: An existential perspective on the Self-esteem and self-disgust both mediate the body and gender in self-conscious emotions. relationship between dysfunctional cognitions In: Tracy JL, Robins RW and Tangney JP (eds) and depressive symptoms. Motivation and The Self-Conscious Emotions: Theory and Emotion 34: 399–406. Research. New York: Guildford, pp. 389–406. Tabachnick B and Fidell L (2007) Using Multivariate Rozin P, Lowery L, Imada S, et al. (1999) The CAD Statistics. New York: Pearson Education Inc. triad hypothesis: A mapping between three Yarnell LM, Stafford RE, Neff KD, et al. (2015) moral emotions (, anger, disgust) and Meta-analysis of gender differences in self- three moral codes (community, autonomy, compassion. Self and Identity 14(5): 499–520.