نقص های تنظیم احساسات در اختلالات غذایی: یک نشانگر آسیب شناسی خوردن و یا آسیب شناسی روانی عمومی؟

Abstract Preliminary evidence indicates that individuals with eating disorders (ED) show emotion regulation (ER) difficulties. However, it is yet unclear whether different types of ED differ in their ER profile and whether certain ER difficulties are specific for ED or rather a transdiagnostic factor. Twenty women with anorexia nervosa (AN), 18 with bulimia nervosa (BN), 25 with binge eating disorder (BED), 15 with borderline personality disorder (BPD), 16 with major depressive disorder (MDD) and 42 female healthy controls (HC) were administered the Emotion Regulation Questionnaire, the Inventory of Cognitive Affect Regulation Strategies, the Difficulties in Emotion Regulation Scale and the Affect Intensity Measure. The ED groups reported significantly higher levels of emotion intensity, lower acceptance of emotions, less emotional awareness and clarity, more self-reported ER problems as well as decreased use of functional and increased use of dysfunctional emotion regulation strategies when compared to HC. No significant differences between the ED groups emerged for most ER variables. However, there were indications that the BED group may show a slightly more adaptive pattern of ER than the two other ED groups. As a whole, all clinical groups performed very similar on most ER variables and reported more difficulties regulating their emotions than HC. The findings suggest that ER difficulties are not linked to a particular diagnostic category. Instead, ER difficulties appear to be a transdiagnostic risk and/or maintenance factor rather than being disorder-specific.

. Introduction With respect to eating disorders (ED), a great amount of theoretical and empirical attention so far has been devoted to cognitive biases and their relevance for the maintenance of pathological eating behavior (Lee and Shafran, 2004). In recent years, however, there has been a proliferation of research focusing on the dispositional use of emotion regulation (ER) strategies in individuals with ED. This research interest was triggered by empirical findings showing that negative mood is a reliable antecedent of binge eating in binge eating disorder (BED; Wegner et al., 2002, Chua et al., 2004, Hilbert et al., 2007 and Stein et al., 2007) and binge/purging behavior in bulimia nervosa (BN; Stice, 2001, Waters et al., 2001, Smyth et al., 2007, Crosby et al., 2009 and Smyth et al., 2009). A key hypothesis in this context is that individuals with ED display deficits in ER and lack the skills required to adaptively and effectively cope with negative affective states. Binge attacks and/or purge behavior are then seen as attempts to cope with negative affect by providing short term comfort or distraction (Smyth et al., 2007 and Wild et al., 2007). Although most theoretical accounts on ER deficits in ED have focused on BN or BED, some theorists have suggested problems with ER to be present in other ED, too. For example, in their transdiagnostic model of ED, Fairburn, Cooper and Shafran (2003) suggest mood intolerance to be a process involved in the maintenance of ED in general. More specifically, excessive exercising in anorexia nervosa (AN) has been suggested to serve as an ER strategy and thereby replace the binge/purge cycle and binge behavior typical for BN and BED respectively (Penas-Lledo et al., 2002). Preliminary evidence for the assumed lack of effective ER skills in ED comes from self-report studies comparing eating disordered individuals with healthy controls (HC). Several of these studies focused on the concept of alexithymia, which refers to the difficulty in identifying and describing feelings (Taylor et al., 1992, Bagby et al., 1994 and Taylor et al., 1996). Using the Toronto Alexithymia Scale (TAS; Taylor et al., 1988), increased difficulties with respect to emotional awareness have been reported both for AN and BN (Bydlowski et al., 2005 and Carano et al., 2006) and BED (Svaldi et al., 2010) when compared to HC. Furthermore, compared to HC and girls with unipolar depression, young women with BN show inferior interoceptive awareness and greater reluctance to express emotions (Sim and Zeman, 2004). In another study (Sim and Zeman, 2005), negative affect, poor awareness of emotions and nonconstructive coping with negative emotions were shown to partially mediate the link between body dissatisfaction and bulimic symptoms in a sample of early adolescent girls. Corstorphine et al. (2007) found women with AN, BN and ED not otherwise specified (EDNOS) to self-report greater avoidance of emotions and lower capacity to accept and manage emotions on the Distress Tolerance Scale (Corstorphine et al., 2007), an instrument which assesses adaptive and maladaptive means of coping with affect. In BED, the frequency of binge eating in a large sample of college students could be predicted by the total score, as well as by all the subscales of the Difficulties in Emotion Regulation Scale (DERS; Gratz and Roemer, 2004) including non-acceptance of emotions, difficulties with engaging in goal-directed activities, impulsiveness, lack of awareness of emotions and non-availability of effective ER strategies (Whiteside et al., 2007). Similarly, Harrison et al. (2009) found women with AN to have significantly more difficulties in all the subscales of the DERS compared to HC. In a recent meta-analysis on the dispositional use of ER strategies (Aldao et al., 2010), rumination and suppression were positively and problem solving was negatively associated with ED symptoms. While effect sizes for rumination and suppression were medium, effect sizes for reappraisal were only small and non-significant. Hence, some ER strategies may be more important than others. In sum, there is preliminary evidence supporting the hypothesis that individuals with ED show ER difficulties. However, a number of limitations regarding past research in this area are noteworthy. First, ER is a broad concept that has been defined as “the extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially their intensive and temporal features, to accomplish one's goals” (Thompson, 1994, pp. 27–28). Conceptualized in this way, ER encompasses a number of different processes involving not only “the modulation of emotional arousal, but also the awareness, understanding, and acceptance of emotions, and the ability to act in desired ways regardless of emotional state” (Gratz and Roemer, 2004, p. 41). It is yet unclear, which of these processes are involved in the maintenance of ED. Past studies have tended to focus on only a few variables at a time. In order to get a more complete picture of ER difficulties in ED, it appears important to assess ER in a more comprehensive way. Second, it is as yet unclear whether different types of eating disorders differ in their ER profile. In an earlier study using the Negative Mood Regulation Scale (Catanzaro and Mearns, 1990), no differences between different ED were found with regard to individuals' sense of competence in using various strategies of negative mood regulation (Gilboa-Schechtman et al., 2006). Although participants with AN reported lower levels of emotional awareness than individuals with BN, this difference disappeared when controlling for levels of depression and anxiety. Similarly, in another study TAS differences between AN and BN disappeared when taking depression into account as a confounding variable (Corcos et al., 2000). In a third study, no differences between AN and BN were found on the TAS, but AN females reported significantly lower levels of emotional awareness by means of the Level of Emotional Awareness Scale (LEAS; Lane et al., 1990) than participants with BN (Bydlowski et al., 2005). The current study aimed to systematically investigate ER difficulties across different ED using a large range of ER variables. Most studies to date have compared eating disordered participants to HC only. It therefore remains unclear in which ways ER may be specific for eating pathology or rather a transdiagnostic maintenance factor. In fact, ER difficulties have been found across a large range of emotional disorders (Campbell-Sills and Barlow, 2006 and Kring, 2008) and feature prominently in current theoretical models of psychopathology (Barlow et al., 2004 and Mennin and Fresco, 2010). In the previously mentioned meta-analysis on the dispositional use of ER (Aldao et al., 2010), maladaptive strategies such as rumination, avoidance and suppression were associated with higher levels of depression, anxiety, substance abuse and eating pathology. Likewise, more adaptive strategies as acceptance, reappraisal and problem solving were associated with lower psychopathology. However, in contrast to adaptive ER strategies, maladaptive strategies were more strongly related to psychopathology. Furthermore, the relationship between certain ER strategies and psychopathology differed depending on the type of disorder examined. For example, effect sizes for rumination were large for anxiety and depression, but only medium for eating disturbances and substance abuse. Effect sizes for avoidance were large for depression, medium to large for anxiety, and medium for eating and substance problems. Finally, effect sizes for reappraisal were small to medium for depression and anxiety, but only small for eating pathology and substance abuse. A subsequently conducted study (Aldao and Nolen-Hoeksema, 2012) found the relationship between adaptive ER strategies and symptoms of depression, anxiety and alcohol problems to be moderated by levels of maladaptive strategies. Specifically, adaptive strategies were negatively associated with psychopathology symptoms only at high levels of maladaptive strategies. Moreover, maladaptive, but not adaptive ER strategies were prospectively associated with symptoms of psychopathology. Based on the limitations of earlier research, the aims of the current study were threefold. It was firstly aimed to replicate earlier findings showing that ED are related to ER difficulties. In line with existing conceptualizations (Gratz and Roemer, 2004 and Berking et al., 2008), a range of different characteristics of ER were examined. It was hypothesized that compared to healthy controls, participants with ED would (1) report a higher intensity of emotions, (2) be less acceptant of negative emotions, (3) report less awareness, clarity and understanding of their emotions, (4) report more ER problems, (5) report using less functional ER strategies, and (6) report using more dysfunctional ER strategies. The study further aimed to investigate whether different types of eating disorders (AN, BN and BED) differ regarding their ER profile. For most ER variables, the three groups of ED were expected to show elevated levels when compared to HC, but not to significantly differ from each other. However, based on preliminary earlier findings described above individuals with AN were hypothesized to show lower levels of emotional awareness, but also lower levels of impulse control difficulties than those with BN or BED. Finally, we aimed to investigate whether ER difficulties are specific of ED or rather a general distress deficiency also common to other psychological disorders. In order to test this important issue, we compared our ED groups to a group of women with major depressive disorder (MDD) and a group of women with borderline personality disorder (BPD). There is extensive evidence that BPD is related to ER difficulties (Levine et al., 1997 and Ebner-Priemer et al., 2007) and deficient ER is a key feature of theoretical models of the disorder (Linehan, 1993 and Gratz et al., 2009). Similarly, there is emerging evidence linking depression to difficulties in ER (Ehring et al., 2010 and Joormann, 2010). In line with a transdiagnostic view of ER, we expected all patient groups to differ from HC on all ER variables with only few differences between the different diagnostic categories. However, we hypothesized that BPD, BN and BED would show higher levels of impulse control difficulties than all other groups. In addition, individuals with MDD were expected to report the lowest levels of using functional cognitive ER strategies.

Results 3.1. Group differences in demographic variables and symptom severities Demographic variables and symptom severities split by diagnostic groups are shown in Table 1. Results of one-way ANOVAs and subsequent follow-up tests showed significant differences in age (BED = MDD > BPD > BN = AN > HC). In addition, groups differed regarding their educational level, whereby more participants in the HC, AN and BN groups belonged to the high education group than those from the BED, MDD and BPD conditions. As to be expected, the groups also differed regarding their body mass index (BMI = kg/m2) and symptom severities. The BMI was lowest in the AN, and highest in the BED group. In addition, AN and BN participants received the highest scores and HC the lowest scores on the EDE-Q restraint and eating concern subscales. With respect to EDE-Q weight and shape concern, the ED groups and the BPD group had significantly higher scores than the MDD and HC groups. When looking at symptom levels of depression, MDD and BPD participants showed the highest scores and HC the lowest. Table 1. Sample characteristics. Controls (N = 42) AN (N = 20) BN (N = 18) BED (N = 25) MDD (N = 16) BPD (N = 15) ANOVA F(5, 130) or χ2(10) ηp2 Age1 (years): M (S.D.) 27.76 (6.34)a 22.85 (4.38)b 25.89 (7.84)ab 43.46 (11.95)c 46.38 (7.29)c 35.13 (7.17)d 32.21⁎⁎⁎ 0.55 Education2: N (%) Low education 0 (0%) 1 (5%) 2 (11%) 8 (32%) 4 (25%) 4 (27%) 49.04⁎⁎⁎ Middle school 2 (5%) 7 (35%) 3 (17%) 7 (28%) 7 (44%) 6 (40%) High (A-level and/or university) 40 (95%) 12 (60%) 13 (72%) 10 (40%) 5 (31%) 5 (33%) BMI1 (kg/m2) M (S.D.) 21.36 (2.51)a 16.28 (1.82)b 22.25 (2.77)a 37.55 (6.72)c 30.26 (9.29)d 27.25 (8.95)d 44.14⁎⁎⁎ 0.63 BDI1 2.62 (2.85)a 23.79 (12.02)b 17.94 (8.61)c 15.19 (8.26)c 30.94 (11.01)d 31.21(9.46)d 45.20⁎⁎⁎ 0.64 EDE-Q global score1 0.52 (0.59)a 3.83 (1.53)bd 4.15 (1.04)b 4.18 (1.09)b 1.61 (1.23)c 3.30 (1.52)d 55.81⁎⁎⁎ 0.68 EDE-Q restraint scale1 0.48 (0.89)a 3.74 (2.04)b 3.48 (1.64)bc 2.87 (1.39)ce 1.11 (1.22)ad 2.11 (1.77)de 21.98⁎⁎⁎ 0.46 EDE-Q shape concerns1 0.71 (0.76)a 4.25 (1.48)b 4.73 (0.94)bc 5.15 (1.39)c 2.31 (1.82)d 4.22 (1.61)b 55.27⁎⁎⁎ 0.68 EDE-Q weight concerns1 0.55 (0.76)a 3.76 (1.89)b 4.11 (1.49)b 4.47 (1.37)b 1.76 (1.51)c 3.60 (1.71)b 37.51⁎⁎⁎ 0.59 EDE-Q eating concerns1 0.11 (0.21)a 3.16 (1.54)b 3.81 (1.30)b 3.27 (1.61)b 0.55 (0.63)ac 2.31 (1.68)d 49.94⁎⁎⁎ 0.63 AN = anorexia nervosa; BN = bulimia nervosa; BED = binge eating disorder; MDD = major depressive disorder; BPD = borderline personality disorder; BMI = body mass index (kg/m2); BDI = Beck Depression Inventory; EDE-Q = Eating disorder Examination Questionnaire. a,b,c,d,eDifferent superscripts denote significant differences between groups. ⁎⁎⁎ P < 0.001. 1 Means (standard deviations). 2 Frequencies (percentages). Table options 3.2. Group differences in emotion regulation Results of the MANOVAs showed significant main effects of diagnostic group for all six groups of variables (see Table 2 for all MANOVA and ANOVA results). Table 2. Means (M), standard deviations (S.D.) and group differences in emotion regulation. Controls (N = 42) AN (N = 20) BN (N = 18) BED (N = 25) MDD (N = 16) BPD (N = 15) (M)ANOVA M (S.D.) M (S.D.) M (S.D.) M (S.D.) M (S.D.) M (S.D.) F(15, 384) or F(5, 128) ηp2 Intensity of emotions 4.36⁎⁎⁎ 0.15 AIMintensity of positive emotions 21.59 (6.01) 18.55 (9.79) 23.44 (9.32) 21.44 (9.06) 16.40 (10.95) 20.19 (11.28) 1.43 0.05 AIMintensity of negative emotions 19.09 (5.80)a 23.47 (5.97)b 22.69 (6.22)ab 25.84 (8.99)b 24.40 (9.19)b 32.23 (7.84)c 7.69⁎⁎⁎ 0.23 AIMSerenity 20.26 (5.84)a 14.95 (7.87)b 15.22 (6.86)b 16.60 (7.34)b 16.07 (7.91)b 9.31 (4.87)c 5.93⁎⁎⁎ 0.19 Acceptance of emotions 4.01⁎⁎⁎ 0.13 DERSnon-acceptance 11.00 (4.08)a 20.25 (7.61)bc 19.17 (7.66)bc 16.96 (6.60)b 18.50 (6.19)bc 21.33 (7.02)c 10.72⁎⁎⁎ 0.29 ICARUSacceptance of feelings 5.93 (1.69)a 3.65 (1.63)bc 3.94 (1.83)bc 4.56 (1.61)b 3.75 (1.91)bc 3.20 (1.15)c 10.12⁎⁎⁎ 0.28 ICARUSacceptance of the situation 8.60 (2.12)a 6.50 (2.35)b 7.39 (2.45)ab 7.96 (2.19)ac 6.75 (2.59)bc 6.20 (2.18)b 4.25⁎⁎ 0.14 Clarity, consciousness and understanding of emotions 4.18⁎⁎⁎ 0.14 DERSlack of emotional awareness 14.76 (5.86)a 19.30 (4.37)b 20.22 (4.88)b 19.80 (4.78)b 21.66 (12.66)b 21.00 (4.47)b 4.56⁎⁎ 0.15 DERSlack of emotional clarity 8.88 (2.79)a 15.20 (5.83)bcd 17.94 (7.35)cd 13.12 (4.97)b 14.94 (5.88)bc 18.53 (4.41)d 13.74⁎⁎⁎ 0.35 ICARUSanalysis of feelings and situation 16.42 (3.33)a 13.90 (4.22)b 14.22 (3.96)ab 14.32 (4.51)b 12.88 (5.08)b 16.42 (3.33)ab 2.54⁎ 0.09 Self-reported ER problems 9.19⁎⁎⁎ 0.26 DERSdifficulties engaging in goal directed behavior 11.10 (4.35)a 19.70 (8.16)b 17.67 (5.71)b 16.56 (4.60)b 18.75 (5.17)b 19.87 (3.36)b 11.75⁎⁎⁎ 0.31 DERSimpulse control difficulties 8.31 (2.75)a 17.31 (6.56)b 19.43 (5.68)bd 16.56 (5.49)b 13.13 (5.54)c 21.28 (4.36)d 25.20⁎⁎⁎ 0.49 DERSlimited access to strategies 13.10 (4.67)a 27.45 (8.96)b 26.63 (6.09)b 22.28 (7.89)c 26.94 (6.25)b 30.73 (3.71)b 28.52⁎⁎⁎ 0.52 Controls (N = 43) AN (N = 20) BN (N = 19) BED (N = 29) MDD (N = 16) BPD (N = 15) (M)ANOVA M (S.D.) M (S.D.) M (S.D.) M (S.D.) M (S.D.) M (S.D.) F(15, 384) or F(5, 128) ηp2 Functional ER strategies 2.51⁎⁎⁎ 0.09 ICARUSpositive thoughts 16.02 (3.24)a 13.70 (3.47)b 15.22 (4.22)ab 13.72 (3.52)b 10.13 (4.32)c 12.93 (4.61)b 6.49⁎⁎⁎ 0.20 ICARUSreframing and growth 16.14 (3.10)a 11.75 (3.92)b 14.33 (3.22)ac 13.88 (3.66)c 10.63 (3.50)b 12.47 (2.95)bc 8.87⁎⁎⁎ 0.25 ICARUSdownward comparison & reality testing 9.19 (2.77) 8.45 (3.10) 7.94 (2.82) 8.16 (2.01) 7.38 (3.32) 8.11 (3.50) 1.21 0.05 ICARUSmindful observation 7.71 (2.28)a 5.35 (2.21)b 6.56 (2.38)ab 5.84 (1.95)b 5.31 (2.09)b 5.13 (2.09)b 6.41⁎⁎⁎ 0.20 ERQreappraisal 4.54 (1.07)a 3.64 (0.95)b 3.73 (1.12)b 3.89 (1.51)b 3.32 (1.74)b 3.23 (1.74)b 3.90⁎⁎ 0.13 Dysfunctional ER strategies 4.64⁎⁎⁎ 0.18 ICARUSself criticism/self-balme 11.81 (3.85)a 18.20 (4.12)b 18.11 (3.79)b 17.68 (3.70)b 17.38 (4.53)b 21.39 (1.95)c 19.99⁎⁎⁎ 0.44 ICARUSthought suppression/mental distraction 15.13 (3.35) 16.85 (5.10) 17.17 (4.76) 17.00 (4.92) 16.75 (4.42) 15.38 (4.59) 1.08 0.04 ICARUSblaming others 5.10 (1.92) 4.65 (2.46) 4.78 (2.49) 5.04 (1.88) 4.94 (2.02) 6.80 (3.05) 2.00 0.07 ICARUSthoughs of suicide 2.10 (0.37)a 5.25 (2.49)b 3.56 (2.09)c 2.40 (1.29)a 5.31 (2.70)b 5.60 (2.38)b 17.81⁎⁎⁎ 0.41 ICARUSfutile planning 6.36 (2.86)a 7.20 (3.64)ab 8.61 (2.53)b 7.44 (2.95)ab 6.25 (2.52)a 8.73 (2.69)b 2.76⁎ 0.10 ERQsuppression 2.23 (0.98)a 4.20 (1.59)b 3.89 (1.17)b 3.54 (1.54)b 3.77 (0.98)b 3.75 (1.89)b 7.00⁎⁎⁎ 0.21 a,b,c,dDifferent superscripts denote significant differences between groups. Note. ER = emotion regulation; ICARUS = Inventory of Cognitive Affect Regulation Strategies; ERQ = Emotion Regulation Questionnaire; DERS = Difficulties in Emotion Regulation Scale; AIM = Affect Intensity Measure; AN = anorexia nervosa; BN = bulimia nervosa; BED = binge eating disorder; MDD = major depressive disorder; BPD = borderline personality disorder. ⁎ P < 0.05. ⁎⁎ P < 0.01. ⁎⁎⁎ P < 0.001. Table options 3.2.1. Intensity of emotions Follow-up one-way ANOVAs showed significant main effects of group for the AIM subscales intensity of negative emotions and serenity, but not intensity of positive emotions. All disorder groups showed significantly lower scores for serenity than HC, and all groups apart from BN showed significantly higher scores for negative emotions than HC. The only difference within disorder groups was that participants with BPD showed significantly more extreme scores than all other disorder groups. Apart from a large effect size between BN and MDD participants (d = 0.69), and a medium effect size between AN and BN participants (d = 0.51) and BED and MDD participants (d = 0.50) with regard to the intensity of positive emotions, all non-significant group differences had small to moderate effect sizes (all ds < 0.34). With regard to the intensity of negative emotions, all non-significant group differences had small to moderate effect sizes (all ds < 0.41). Regarding serenity, all non-significant group differences had small effect sizes (all ds < 0.22). 3.2.2. Acceptance of emotions Participants from all five disorder groups reported significantly less emotional acceptance than HC. In addition, the AN, MDD, and BPD groups also showed lower scores regarding situational acceptance. The only difference between disorders was between BPD and BED groups with the former reporting less acceptance than the latter for all three subsumed subscales. With regard to non-acceptance of emotions, all non-significant group differences had small to moderate effect sizes (all ds < 0.43). With regard to acceptance of feelings, there was a large effect size for differences between AN and BED participants (d = 0.56), while all other non-significant group differences had small to moderate effect sizes (all ds < 0.48). Regarding the acceptance of the situation, there was a large effect size for differences between AN and BED participants (d = 0.64) and a medium effect size for differences between BN and BPD participants (d = 0.51) and BED and MDD participants (d = 0.50). All other effect sizes for non-significant group differences were small to moderate (all ds < 0.37). 3.2.3. Clarity, consciousness, and understanding of emotions Participants with disorders reported less emotional awareness and clarity. Similarly, the AN, BED and MDD groups reported analyzing their feelings less than HC on the ICARUS. The disordered groups did not differ from each other regarding their levels of emotional awareness or the use of analyzing their feelings as an ER strategy. However, the BPD group reported significantly lower levels of emotional clarity than the BED and MDD groups, and participants with BN showed lower scores than participants with BED. Regarding the lack of emotional awareness, a large effect size was found for differences between AN and BPD participants (d = 0.84), while all other effect sizes were small to moderate (all ds < 0.26). For lack of emotional clarity, large effect sizes for non-significant group differences were found between AN and BPD participants (d = 0.64), while all other effect sizes were small to moderate (all ds < 0.45). With regard to analysis of feelings and situation, there was further a large effect size for non-significant group differences between AN and BPD participants (d = 0.66) and MDD and BPD participants (d = 0.82), as well as a moderate effect size for differences between BED and BPD participants (d = 0.53), while all other effect sizes for non-significant group differences were small to moderate (all ds < 0.29). 3.2.4. Self-reported ER problems All groups with disorders reported significantly more ER problems than HC. No differences between the disorder groups emerged for the DERS subscale difficulties engaging in goal-directed behavior. As expected, participants with BPD showed significantly higher levels of impulse control difficulties than all other groups except the BN group. In addition, all ED groups showed significantly higher scores on this variable than participants with MDD. However, no significant group difference regarding impulse control difficulties were found between the different types of ED. Although differing from HC on this variable, participants with BED showed lower scores on the DERS subscale limited access to strategies than all other disorder groups. Regarding difficulties engaging in goal directed behavior, there was a large effect size for differences between BED and BPD participants (d = 0.82), while all other effect sizes for non-significant group differences were small to moderate (all ds < 0.47). Regarding impulse control difficulties, there was a moderate effect size for differences between BN and BED participants (d = 0.51), while all other non-significant group differences had small to moderate effect sizes (all ds < 0.37). Finally, with regard to limited access to strategies, there was a large effect size for differences between BN and BPD participants (d = 0.81), while all other effect sizes for non-significant group differences were small to moderate (all ds < 0.48). 3.2.5. Functional ER strategies No group differences were found regarding the use of downward comparison and reality testing. However, HC reported using reappraisal as a functional ER strategy significantly more often than all other disorder groups with no difference between disorders. In addition, all disorder groups except BN showed significantly lower scores on the scales positive thoughts, reframing and growth and mindful observation. The hypothesis that participants with MDD would specifically differ from the other groups was only supported for the use of positive thoughts to regulate emotions. With regard to positive thoughts, there were large effect sizes for non-significant group differences between AN and BN participants (d = 0.93) and BN and BPD participants (d = 0.52), while effect sizes for all other non-significant group differences were small to moderate (all ds < 0.39). With regard to reframing and growth, effect sizes were large for non-significant group differences between BN and BPD participants (d = 0.60) and MDD and BPD participants (d = 0.57), while all other effect sizes for non-significant group differences were small to moderate (all ds < 0.42). With regard to downward comparison and reality testing, all effect sizes for non-significant group differences were small to moderate (all ds < 0.33). For mindful observation, effect sizes were large for non-significant group differences between AN and BN participants (d = 0.53), BN and MDD participants (d = 0.56) and BN and BPD participants (d = 0.64). All other effect sizes were small to moderate (all ds < 0.35). Regarding cognitive reappraisal, all effect sizes for non-significant group differences were small to moderate (all ds < 0.41). 3.2.6. Dysfunctional ER strategies No group differences were found regarding the use of thought suppression and blaming others. However, all disorder groups showed significantly heightened levels of self-criticism and emotion suppression when compared to controls. Suicidal thoughts were elevated in AN, MDD and BPD groups, whereas participants with BN and BPD showed the highest scores regarding futile planning. Regarding self-criticism and self-blame, all effect sizes for non-significant group differences were small (all ds < 0.19). Regarding thought suppression and mental distraction all effect sizes for non-significant group differences were small to moderate (all ds < 0.38). For the scale blaming others, large effect sizes were found for non-significant group differences between AN and BPD participants (d = 0.78), BN and BPD participants (d = 0.73), BED and BPD participants (d = 0.69) and MDD and BPD participants (d = 0.72). All other effect sizes were small (all ds < 0.18). For thoughts of suicide, effect sizes for non-significant group differences were small (all ds < 0.14). With regard to futile planning and suppression, effect sizes for non-significant group differences were small to moderate (all ds < 0.48). 3.3. Correlations Correlations between the severity of ED symptoms (EDE-Q) as well as depressive symptoms (BDI) and the different ER variables are shown in Table 3. Significant correlations between EDE-Q scores and ER difficulties emerged for all ER variables, apart from the intensity of positive emotions, the functional strategy down comparison and reality testing and the dysfunctional strategy blaming others. Similarly, the levels of depression were significantly correlated with all indices of ER, apart from the dysfunctional strategies thought suppression/mental distraction, blaming others, and futile planning. Table 3. Correlations between emotion regulation strategies, eating pathology and depression. EDE-Q global score EDE-Q restraint scale EDE-Q eating concerns EDE-Q weight concerns EDE-Q shape concerns BDI Intensity of emotions AIMintensity of positive emotions − 0.07 − 0.08 − 0.01 − 0.07 − 0.08 − 0.21⁎ AIMintensity of negative emotions 0.39⁎⁎ 0.22⁎ 0.35⁎⁎ 0.43⁎⁎ 0.41⁎⁎ 0.43⁎⁎ AIMserenity − 0.24⁎⁎ − 0.11 − 0.24⁎⁎ − 0.24⁎⁎ − 0.25⁎⁎ − 0.41⁎⁎ Acceptance of emotions DERSnon-acceptance 0.46⁎⁎ 0.32⁎⁎ 0.45⁎⁎ 0.45⁎⁎ 0.47⁎⁎ 0.60⁎⁎ ICARUSacceptance of feelings − 0.43⁎⁎ − 0.36⁎⁎ − 0.33⁎⁎ − 0.42⁎⁎ − 0.43⁎⁎ − 0.56⁎⁎ ICARUSacceptance of the situation − 0.25⁎⁎ − 0.20⁎ − 0.20⁎ − 0.27⁎⁎ − 0.26⁎⁎ − 0.41⁎⁎ Clarity, consciousness and understanding of emotions DERSlack of emotional awareness 0.33⁎⁎ 0.24⁎⁎ 0.28⁎⁎ 0.32⁎⁎ 0.35⁎⁎ 0.43⁎⁎ DERSlack of emotional clarity 0.44⁎⁎ 0.34⁎⁎ 0.42⁎⁎ 0.42⁎⁎ 0.45⁎⁎ 0.52⁎⁎ ICARUSanalysis of feelings and situation − 0.21⁎ − 0.13 − 0.17 − 0.22⁎ − 0.23⁎⁎ − 0.27⁎⁎ Self-reported ER problems DERSdifficulties engaging in goal directed behavior 0.46⁎⁎ 0.30⁎⁎ 0.44⁎⁎ 0.46⁎⁎ 0.47⁎⁎ 0.60⁎⁎ DERSimpulse control difficulties 0.57⁎⁎ 0.43⁎⁎ 0.54⁎⁎ 0.56⁎⁎ 0.57⁎⁎ 0.60⁎⁎ DERSlimited access to strategies 0.51⁎⁎ 0.37⁎⁎ 0.46⁎⁎ 0.50⁎⁎ 0.51⁎⁎ 0.78⁎⁎ Functional ER strategies ICARUSpositive thoughts − 0.19⁎ − 0.13 − 0.09 − 0.22⁎ − 0.22⁎ − 0.44⁎⁎ ICARUSreframing and growth − 0.26⁎⁎ − 0.19⁎ − 0.21⁎ − 0.23⁎⁎ − 0.29⁎⁎ − 0.51⁎⁎ ICARUSdownward comparison & reality testing − 0.10 0.01 − 0.07 − 0.13 − 0.14 − 0.23⁎⁎ ICARUSmindful observation − 0.35⁎⁎ − 0.26⁎⁎ − 0.28⁎⁎ − 0.36⁎⁎ − 0.36⁎⁎ − 0.46⁎⁎ ERQreappraisal − 0.24⁎⁎ − 0.14 − 0.23⁎⁎ − 0.24⁎⁎ − 0.27⁎⁎ − 0.43⁎⁎ Dysfunctional ER strategies ICARUSself criticism/self-balme 0.55⁎⁎ 0.44⁎⁎ 0.48⁎⁎ 0.55⁎⁎ 0.54⁎⁎ 0.62⁎⁎ ICARUSthought suppression/mental distraction 0.21⁎ 0.17⁎ 0.18⁎ 0.16 0.22⁎ 0.07 ICARUSblaming others − 0.003 − 0.10 − 0.03 0.06 0.02 − 0.01 ICARUSthoughs of suicide 0.34⁎⁎ 0.30⁎⁎ 0.23⁎⁎ 0.33⁎⁎ 0.34⁎⁎ 0.71⁎⁎ ICARUSfutile planning 0.22⁎⁎ 0.20⁎ 0.21⁎ 0.20⁎ 0.21⁎ 0.14 ERQsuppression 0.47⁎⁎ 0.40⁎⁎ 0.38⁎⁎ 0.44⁎⁎ 0.48⁎⁎ 0.48⁎⁎ Severity of depression BDI 0.49⁎⁎ 0.38⁎⁎ 0.41⁎⁎ 0.47⁎⁎ 0.50⁎⁎ 1 Note. ICARUS = Inventory of Cognitive Affect Regulation Strategies; ERQ = Emotion Regulation Questionnaire; DERS = Difficulties in Emotion Regulation Scale; AIM = Affect Intensity Measure; AN = anorexia nervosa; BN = bulimia nervosa; BED = binge eating disorder; MDD = major depressive disorder; BPD = borderline personality disorder; BDI = Beck Depression Inventory; EDE-Q = Eating disorder Examination Questionnaire. ⁎ P < 0.05. ⁎⁎ P < 0.001 Table options Statistically controlling for the intensity of negative emotions (AIM negative emotions subscale) did not change the results.