Razi University

   Abstract    Eating disorders, characterized by abnormal or disturbed eating behaviors, are currently recognized as one of the significant psychological disorders, and their prevalence has shown an increasing trend in recent decades. The present study aimed to investigate the mediating role of self-compassion in the relationship between attachment styles and eating disorder symptoms among adult women in Kermanshah. This study employed a descriptive-correlational design, and the data were analyzed using structural equation modeling. The study population consisted of all women aged 18 to 40 years in Kermanshah, from whom 366 participants were selected through convenience sampling. The research instruments included the Eating Attitudes Test (Garner & Garfinkel, 1982), the Revised Adult Attachment Scale (Collins & Read, 1996), and the Short Form of the Self-Compassion Questionnaire (Raes et al., 2011). Data were analyzed using    version 26 and AMOS version 24.    Findings: The results of the structural model indicated that secure attachment had a positive and significant direct relationship with self-compassion (? = 0.402, p < 0.001), whereas anxious attachment had a negative and significant direct relationship with self-compassion (? = ?0.521, p < 0.001). The direct effect of secure attachment on eating disorder symptoms was negative but non-significant (? = ?0.065, p = 0.642), while the direct effect of anxious attachment on eating disorder symptoms was positive but non-significant (? = 0.057, p = 0.617). In contrast, self-compassion had a negative and significant direct effect on eating disorder symptoms (? = ?0.205, p = 0.035).  Furthermore, the indirect effect of secure attachment on eating disorder symptoms through self-compassion was significant (? = ?0.182, p = 0.031; 95% bootstrap confidence interval: ?0.202 to ?0.015). Similarly, the indirect effect of anxious attachment on eating disorder symptoms through self-compassion was significant (? = 0.208, p = 0.018; 95% bootstrap confidence interval: 0.025 to 0.239). Avoidant attachment was not entered into the final structural model because its relationships with the main study variables were not significant. The structural model demonstrated an acceptable fit to the data (?²/df = 2.14, RMSEA = 0.07, CFI = 0.90, TLI = 0.90).    Conclusion: The findings indicated that self-compassion had a significant negative direct relationship with eating disorder symptoms and also played a significant mediating role in the relationship between secure attachment and anxious attachment and eating disorder symptoms. In contrast, the direct paths from secure attachment and anxious attachment to eating disorder symptoms were not significant. Therefore, self-compassion may be considered in understanding the association between attachment styles and eating disorder symptoms.    Keywords: Eating disorder symptoms, attachment styles, self-compassion, adult women, Kermanshah.