Objectives
Gilbert (Compassion: Conceptualisations, Research, and Use in Psychotherapy. London: Routledge, 2005) theorized that self‐critical individuals have more severe psychopathology due in part to their elevated feelings of shame. We sought to test this model in a sample of eating disorder sufferers.
Method
Seventy‐four patients admitted to a specialized day or inpatient eating disorders treatment programme completed the Forms of Self‐Criticism and Self‐Reassurance Scale, Rosenberg Self‐Esteem Inventory, Experience of Shame Scale, Beck Depression Inventory, Positive and Negative Affect Schedule, and Eating Disorder Examination Questionnaire.
Results
We tested our mediational model with Preacher and Hayes' (Behavior Research Methods, 40, 879, 2008) bootstrapping approach entering self‐criticism as a predictor, self‐esteem as a covariate, and shame, negative affect, positive affect, and depressive symptoms as simultaneous mediators. Applying a 95% confidence interval, the total indirect effect of self‐criticism on eating disorder pathology was significantly different from zero suggesting that its influence occurred through the proposed set of mediators. Specific indirect effects revealed that shame was the only mediator to contribute significantly to the model.
Conclusions
Results support the theory that among eating disorder patients, higher self‐criticism is associated with elevated eating disorder pathology through feelings of shame. Interventions that target the shame of self‐critical patients might therefore facilitate their recovery.
Practitioner Points
- Assessing and intervening with the feelings of shame experienced by high self‐critical patients might be particularly important in helping them overcome their eating disorder.
- Working with patients to understand and alter the links between their self‐criticism, shame, and eating disorder behaviours might be a fruitful therapeutic avenue.
- Case conceptualizations in the eating disorders and perhaps other forms of psychopathology might benefit from integrating theoretical models that integrate a focus on self‐criticism and shame.
Limitations
- This is a cross‐sectional study, meaning that we cannot make conclusions about the directionality of the relationships observed. It will be important to examine the temporal relationships between self‐criticism, shame, and eating disorder symptoms in future research.
- We rely on correlational data and so can only make predictive inferences based on our results. An experimental study in which self‐criticism and/or shame are manipulated will be an important next step to allow for causal inferences.