Self-mutilation is one of the most severe methods of acting out internal distress upon oneself. Various researchers have found that it is related to histories of childhood abuse, parental abandonment/neglect, and later symptomatology such as borderline personality disorder, dissociative disorders, and posttraumatic stress disorder. A further link between self-mutilation and insecure attachment has been hypothesized but never empirically evaluated. The purpose of this study was to assess the direct and indirect relationships of these factors to self-mutilating (SM) behaviors.

To assess these relationships, 78 survivors of childhood sexual abuse who were receiving treatment at an outpatient facility devoted to the treatment of sexual abuse were assessed. Because of the nature of the outpatient facility, survivors had experienced more severe abuse and harsher childhood environments, and were particularly symptomatic. Sexual abuse most often included penetration, lasted for an extended time, and was one among other types of traumatic childhood experiences. Not surprisingly, then, half or more of respondents exhibited moderate to severe dissociative, borderline, or posttraumatic stress symptoms. Forty-nine percent had a history of self-mutilation, with the most frequent types being cutting and hitting oneself (each of which were endorsed by half of those who self-mutilated). The most endorsed reasons for engaging in these behaviors was to reduce tension or, secondarily, to feel something.

Using a path analysis, the relationship between attachment and self-mutilation (SM) behaviors was found to be direct, with dismissing attachment being the best predictor. Both intrusive posttraumatic stress symptoms and dissociation had moderate to strong direct relationships with SM behaviors and also acted as mediators for other childhood abuse/neglect and symptoms of borderline personality disorder. This latter finding is particularly important because of the historical inference that individuals with borderline personality disorder may use self mutilation as a method of manipulation. Instead, this analysis suggests that individuals in a more dissociated state are more likely to self mutilate, and that it is the comorbidity of borderline personality disorder with dissociation that accounts for the relationship of borderline personality disorder with self-mutilation. The final important and intriguing finding is that intrusive posttraumatic stress symptoms were positively related to SM behaviors, whereas avoidant posttraumatic symptoms were negatively associated. In comparing symptoms between SM and nonSM respondents in this and another study, it was determined that SM respondents scored higher on virtually all symptoms except avoidance. SM respondents were instead less likely, or perhaps, less capable of engaging in avoidant behaviors. These intriguing findings are explained within a theoretical and neurophysiological framework.