Defining guilt in depression: a comparison of subjects with major depression, chronic medical illness and healthy controls

Background: Although guilt is a widely accepted feature of depression, there is limited and inconsistent data defining the nature of this symptom. The purpose of the current study was to examine the specificity and nature of guilt in subjects with major depression as compared to patients with another chronic medical illness and healthy controls. Methods: Outpatients with current major depressive episode (MDE n = 34), past-MDE (n = 22), chronic cardiac illness (n = 20) and healthy controls (n = 59) were administered the following measures: The Guilt Inventory (GI), State Shame and Guilt Scale (SSGS), 17-item Hamilton Rating Scale for Depression (Ham-D) and the Structured Clinical Interview for DSM-IV. Results: Overall multivariate analysis of covariance comparing mean scores for the six guilt subscales [state-guilt, trait-guilt, moral standards (from the GI); state-guilt, -pride, and -shame (from the SSGS)] across the four groups was significant (F = 9.1, df = 6:121, p < 0.0001). Post-hoc analysis revealed the following differences (each at least p < 0.01): for state-guilt (GI), current-MDE > past-MDE > cardiac = healthy controls for trait-guilt (GI), current-MDE = past-MDE > cardiac = healthy controls; for state-shame, -guilt and -pride (SSGS), current-MDE > past-MDE, past-MDE = cardiac, past-MDE > healthy, cardiac = healthy controls. Among depressed patients, there was significant correlation between Ham-D score and all guilt sub-scales (p < 0.01), except moral standards. Limitations: The cardiac group may have less illness burden than currently depressed. Conclusions: State expression of guilt, shame and low pride distinguish acutely depressed from all other groups, and are highly influenced by severity of depression. Trait-guilt does not differentiate acute from past depressed. Data suggests guilt may represent both an enduring and fluctuating feature of depressive illness over its longitudinal course. (C) 2002 Elsevier Science BV All rights reserved.