خود دلسوزی در افسردگی: ارتباط با نشانه های افسردگی، نشخوار فکری و اجتناب در بیماران سرپایی افسرده

Abstract Self-compassion involves being kind to oneself when challenged with personal weaknesses or hardship and has been claimed to be associated with resilience in various areas. So far, there are only a handful of studies that investigate self-compassion and its relation to clinical depression. Therefore, the principal goals of the present study were (a) to compare self-compassion in clinically depressed patients and never-depressed subjects, (b) to investigate self-compassion and its relation to cognitive-behavioral avoidance and rumination in depressed outpatients, and (c) to investigate rumination and avoidance as mediators of the relationship between self-compassion and depressive symptoms. One hundred and forty-two depressed outpatients and 120 never-depressed individuals from a community sample completed a self-report measure of self-compassion along with other measures. Results indicate that depressed patients showed lower levels of self-compassion than never-depressed individuals, even when controlled for depressive symptoms. In depressed outpatients, self-compassion was negatively related to depressive symptoms, symptom-focused rumination, as well as cognitive and behavioral avoidance. Additionally, symptom-focused rumination and cognitive and behavioral avoidance mediated the relationship between self-compassion and depressive symptoms. These findings extend previous research on self-compassion, its relation to depression, as well as processes mediating this relationship, and highlight the importance of self-compassion in clinically depressed patients. Since depressed patients seem to have difficulties adopting a self-compassionate attitude, psychotherapists are well advised to explore and address how depressed patients treat themselves.

Results Preliminary Analyses The DS and the NDS differed significantly regarding depressive symptoms (DS: M = 25.61, SD = 9.04; NDS: M = 3.95, SD = 3.88; t[198, 93] = 25.86, p < .001). Further, the samples differed significantly regarding age, (t[258, 84] = 7.05, p < .001) and marital status (χ2[2] = 10.99, p < .01), and there was an insignificant trend with regard to education (χ2[3] = 9.31, p = .051). Samples did not differ regarding gender distribution (χ2[1] = 0.28, p = .87). Due to the results of these analyses, we decided to include age, marital status, and education as covariates in subsequent analyses comparing the two samples. Self-Compassion in Clinically Depressed Patients and Never-Depressed Subjects As Table 1 indicates, clinically depressed patients were significantly less self-compassionate than never-depressed subjects. Furthermore, the two samples differed regarding all subscales of self-compassion.1 In order to investigate whether levels of depressive symptoms may account for this difference, we reconducted the analyses of depressed and never-depressed subjects regarding levels of self-compassion with depressive symptoms as an additional covariate. Results indicated that the two samples still differed significantly regarding the self-compassion total score, as well as the self-kindness, self-judgment, isolation, and overidentification subscale. Table 1. Self-Compassion in Never-Depressed Subjects and Clinically Depressed Patients Never-Depressed SubjectsM (SD) Depressed PatientsM (SD) F Partial η2 F Partial η2 (Controlling for depressive symptoms) SCS 3.31 (.36) 2.75 (.41) 126.70** .34 12.39** .05 SCS-sk 3.31 (.78) 2.47 (.78) 59.49** .19 4.82* .02 SCS-ch 2.97 (.76) 2.55 (.71) 15.00** .06 0.67 .00 SCS-m 3.32 (.70) 2.82 (.70) 37.23** .13 0.84 .00 SCS-sj 2.34 (.70) 3.55 (.62) 230.65** .48 30.73** .11 SCS-i 1.93 (.84) 3.40 (.81) 220.52** .47 26.76** .10 SCS-oi 2.51 (.80) 3.62 (.66) 146.21** .37 12.40** .05 Notes. Due to missing values in covariates, sample size was n = 119 for the never-depressed sample and n = 134 for the depressed sample. SCS = Self-Compassion total score, -sk = Self-Kindness subscale, -ch = Common Humanity subscale, -m = Mindfulness subscale, -sj = Self-Judgment subscale, -i = Isolation subscale, -oi = Over-Identification subscale. Age, marital status and education were entered as covariates in all analyses. * p < .05. ** p < .01. Table options Self-Compassion, Its Components, and Depressive Symptoms Next, we performed correlational analyses for self-compassion as measured by the SCS, its subscales, and depressive symptoms in both samples. As can be seen in Table 2, self-compassion was significantly negatively correlated with depressive symptoms in the DS (r = -.23, p < .01) as well as in the NDS (r = -.37, p < .01). These correlation coefficients did not differ significantly from each other (Zdiff = 1.23, p = .22). All subscales of the SCS correlated in the expected direction with the total self-compassion score in the depressed sample as well as in the never-depressed sample. Furthermore, correlational analyses indicated that in both samples all subscales except common humanity correlated with depressive symptoms in the expected direction. Table 2. Intercorrelations of Self-Compassion, Its Components, and Depressive Symptoms in Depressed Outpatients and Never-Depressed Subjects 1 2 3 4 5 6 7 8 1. SCS - .79** .65** .68** -.43** -.52** -.22* -.37** 2. SCS-sk .87** - .46** .44** -.22* -.23* -.11 -.19* 3. SCS-ch .74** .61** - .50** .14 .02 .04 -.11 4. SCS-m .79** .68** .55** - -.17 -.25** -.38** -.31** 5. SCS-sj -.70** -.61** -.28** -.37** - .61** .61** .37** 6. SCS-i -.58** -.37** -.20** -.39** .52** - .62** .48** 7. SCS-oi -.35** -.37** -.21** -.36** .44** .68** - .40** 8. BDI-II -.23** -.21** -.15 -.18* .23** .22** .24** - Notes. SCS = Self-Compassion total score, -sk = Self-Kindness subscale, -ch = Common Humanity subscale, -m = Mindfulness subscale, -sj = Self-Judgment subscale, -i = Isolation subscale, -oi = Over-Identification subscale, BDI-II = Beck Depression Inventory II. Correlation coefficients for the depressed sample (n = 142) are shown in the lower-left part of the table and for the never-depressed sample (n = 120) in the upper-right part of the table. * p < .05, ** p < .01. Table options Self-Compassion, Rumination, and Avoidance in Depressed Patients Table 3 gives an overview of how self-compassion is related to the different facets of rumination and avoidance in the DS. As expected, self-compassion was significantly negatively associated with symptom-focused rumination, but not with self-focused rumination. Furthermore, self-compassion was significantly negatively associated with all avoidance scales except for cognitive social avoidance. Table 3. Descriptives and Intercorrelations of Self-Compassion, Rumination, Avoidance and Depressive Symptoms in Depressed Outpatients 1 2 3 4 5 6 7 8 M (SD) 1. SCS - 2.76 (.41) 2. RSQ-SYM -.19* - 2.70 (.54) 3. RSQ-SELF -.16 .52** - 2.42 (.49) 4. CBAS-T -.30** .38** .25** - 2.86 (.63) 5. CBAS-BS -.27** .25** .28** .84** - 2.70 (.84) 6. CBAS-BN -.32** .39** .35** .84** .62** - 3.12 (.71) 7. CBAS-CS -.16 .22** .01 .80** .57** .52** - 2.68 (.77) 8. CBAS-CN -.23** .38** .20* .80** .51** .63** .53** - 2.95 (.73) 9. BDI-II -.23** .40** .17* .36** .37** .28** .20** .34** 25.61 (9.04) Notes. n = 140-142. SCS = Self-Compassion total score, RSQ-SYM = Symptom-focused rumination, RSQ-SELF = Self-focused rumination, CBAS-T = Cognitive-Behavioral Avoidance total score, -BS = CBAS, Behavioral Social subscale, -BN = CBAS, Behavioral Nonsocial subscale, -CS = CBAS, Cognitive Social subscale, -CN = CBAS, Cognitive Nonsocial subscale, BDI-II = Beck Depression Inventory-II. * p < .05. ** p < .01. Table options Mediation Analyses: Rumination and Avoidance as Mediators of the Relation Between Self-Compassion and Depressive Symptoms The results of the mediation analyses are presented in Table 4. The indirect effect of self-compassion on depressive symptoms through symptom-focused rumination was ab = -.06 (SE = .03) with a 95% bootstrap confidence interval (CI95%) of -.12 to -.002. Because zero is not in the 95% confidence interval for symptom-focused rumination, it can be concluded that this indirect effect is significantly different from zero at p < .05, and that symptom-focused rumination mediates the relationship between self-compassion and depressive symptoms. With respect to self-focused rumination, this was not the case (ab = -.02, SE = .02, CI95% = -.07 to .002). Table 4. Summary of Mediational Analyses (1000 Bootstrap Samples) Independent variable (IV) Mediating variable (M) Dependent variable (DV) Effect of IV on M (a) Effect of M on DV (b) Direct effect (c’) Indirect effect (ab) Total effect (c) 1. SCS RSQ-SYM BDI-II -.22* .28** -.12 -.07a -.18** 2. SCS RSQ-SELF BDI-II -.19 .10 -.17* -.02 -.18** 3. SCS CBAS-T BDI-II -.34** .24** -.11 -.08a -.19** 4. SCS CBAS-BS BDI-II -.31** .24** -.12 -.07a -.19** 5. SCS CBAS-BN BDI-II -.36** .16** -.13 -.06a -.19** 6. SCS CBAS-CS BDI-II -.18 .12* -.17* -.02 -.19** 7. SCS CBAS-CN BDI-II -.27** .22** -.13 -.06a -.19** Notes. n = 140-142. SCS = Self-Compassion total score, RSQ-SYM = Symptom-focused rumination, RSQ-SELF = Self-focused rumination, CBAS-T = Cognitive-Behavioral Avoidance total score, -BS = CBAS, Behavioral Social subscale, -BN = CBAS, Behavioral Nonsocial subscale, -CS = CBAS, Cognitive Social subscale, -CN = CBAS, Cognitive Nonsocial subscale, BDI-II = Beck Depression Inventory-II. a Significant point estimate (p < .05). * p < .05. ** p < .01. Table options Next, we tested whether total avoidance and the subscales of avoidance mediate the relationship between self-compassion and depressive symptoms. Results of these analyses suggest that total avoidance completely mediated the relation between self-compassion and depression. The indirect effect of self-compassion on depressive symptoms through avoidance was ab = -0.08 (SE = .03) with a 95% bootstrap confidence interval of -.17 to -.02. Subsequent analyses for the subscales of avoidance showed that all subscales of the CBAS (CBAS-BN: ab = -.06, SE = .03, CI95% = -.13 to -.01; CBAS-BS: ab = -.07, SE = .03, CI95% = -.15 to -.02; CBAS-CN: ab = -.06, SE = .03, CI95% = -.13 to -.01), except for cognitive social avoidance (ab = -.02, SE = .02, CI95% = -.08 to .002), mediated the relation between self-compassion and depressive symptoms. In a last step, we tested the hypothesis that the relation between self-compassion and depressive symptoms is mediated through a latent variable “avoidant functioning”, which was composed of the avoidance total score and symptom-focused rumination. A model of the association between participants’ self-compassion and their current depressive symptoms was specified. Low levels of self-compassion were significantly associated with increased depressive symptoms (β = -.23, z = -2.91, p < .01). Subsequently, a model including the direct association between self-compassion and current depressive symptoms as well as the indirect association of self-compassion through its association with avoidant functioning (defined as a latent factor and including avoidance total score and symptom-focused rumination) on depressive symptoms was specified and estimated. This model was found to fit the observed data adequately: χ2(1) = 1.53, p = .22, CFI = 0.99, TLI = 0.95, RMSEA = .06, SRMR = .02. As shown in Fig. 1, low levels of self-compassion were significantly associated with high levels of avoidant functioning (β = -.39, z = -3.03, p < .01). Avoidant functioning, in turn, was significantly associated with high levels of depressive symptoms (β = .63, z = 3.49, p < .01). Furthermore, the significant association between participants’ self-compassion and their depressive symptoms decreased when avoidant functioning was included in the model (β = .03, z = .23, p = .82). Finally, a bootstrapping procedure with 1000 resamples indicated that the indirect effect of self-compassion on depressive symptoms through avoidant functioning was significant (p = .04). Direct and indirect standardized association between self-compassion, avoidant ... Figure 1. Direct and indirect standardized association between self-compassion, avoidant functioning and levels of depressive symptoms in depressed patients (n = 140). Rectangles indicate measured variables and the circle represents a latent construct. * p < .05. ** p < .01.