Abstract
Joyce and associates report an association between wrist-cutting and bipolar disorder and provide a welcome warning against premature exclusion of the diagnosis of major mental illness when patients present with this sort of self-harm [1]. Our first concern is with the use of the term ‘self-mutilation’ to denote minor self-inflicted injuries in this and some other papers. Minor self-mutilation is sometimes used to refer to self-inflicted injuries that occur without the intention of suicide and do not cause a lasting disability [2]. However, ‘minor self-mutilation’ is a non sequitur because the Oxford English Dictionary (OED) emphasizes the seriousness of mutilation and implies that the term should not be used unless the injuries are severe. The OED defines ‘mutilate’ as ‘to deprive of a limb or organ, destroy the use of a limb or organ or render imperfect by an act of excision or some act of destruction’ [3]. The definition of ‘mutilate’ in the Macquarie Dictionary is similar to the OED, and includes to ‘injure, disfigure, or make imperfect by removing or irreparably damaging body parts’ [4]. The Webster's Online Dictionary settles for ‘injury that deprives you of a limb or other important body part’ [5]. Self-cutting can sometimes result in permanent loss of function, but even this type of injury might not always to be considered to be primarily a mutilation if the intent was suicide. In most instances, self-cutting does not result in loss of function or lasting disability and therefore cannot be properly regarded as self-mutilation using dictionary definitions.
Our second concern is that in contrast to the findings of Joyce, we believe self-mutilation is most strongly associated with schizophrenia. We conducted a systematic review of published cases of self-mutilation in which there was total removal of an eye or testicle, or the amputation of the penis or a limb [6,7]. Less serious injuries and suicide attempts by self-stabbing were excluded. Of 143 cases associated with a psychotic illness, 119 patients had a schizophrenia-spectrum psychosis, 12 cases were attributed to drug-induced or organic psychosis, and 12 had bipolar disorder or psychotic depression. Over half of the patients had never been treated and could be regarded as being in first episode psychosis. Guided by Austin Bradford Hill's criteria for causation in epidemiology we concluded that there was good evidence for a strong and causal association between schizophrenia-spectrum psychosis and self-mutilation, especially prior to initial antipsychotic treatment. In contrast to schizophrenia, the association between affective psychosis such as bipolar disorder and self-mutilation is weak.