By Dr. E. Fuller Torrey
In response to the recent spate of mental illness-related tragedies, there have been occasional claims that the antipsychotic drugs used to treat schizophrenia and bipolar disorder cause homicidal behavior. Such allegations have been heard mostly on social media, on call-in radio shows, and even in letters to the editor. For example, a Jan. 8 letter in the “Toronto Star” claims that “psychotropic drugs have horrendous side effects such as homicide and/or suicide. The perpetrators of almost every mass shooting were on psychotropic drugs.”
Some of these ideas come from previous allegations concerning antidepressant drugs, especially following the Columbine High School shootings in which one of the shooters was alleged to have been prescribed an antidepressant. I am not aware of any definite resolution of this question, but studies I have seen would suggest that antidepressants decrease, rather than increase, suicidal and/or homicidal ideation, eg. Gibbons et al, Suicidal thoughts and behavior with antidepressant treatment, Archives of General Psychiatry, 2012 Jun;69(6):580-7.
Antipsychotics are of course a different class of drugs from antidepressants. In searching PubMed and the psychiatric literature, I can find no published evidence whatsoever that suggests that antipsychotic drugs increase homicidal ideation. On the contrary, those studies which have addressed this issue have concluded that homicides by individuals with severe mental illnesses are almost always committed by individuals who have never been treated or are not taking their antipsychotic medication.
For example, a meta-analysis of 10 studies of homicides and psychotic illness reported that 39 percent of such homicides occur in individuals with psychoses before they have ever been treated. The homicide rate in such individuals who had never been treated was 22 times higher than the rate in individuals who had been treated (Nielssen and Large, Rates of homicide during the first episode of psychoses and after treatment, Schizophrenia Bulletin 2010;36(4):702-12.)
Many other studies have noted an association between violent behavior by individuals with severe mental illnesses and the fact that they were not being treated at the time they committed the violent act. Several such studies are summarized on the Backgrounder on the TAC website (“Violent behavior: One of the consequences of failing to treat individuals with severe mental illnesses”) e.g. Elbogen et al 2006; Swanson et al. 2002; Swanson et al. 1997; Smith et al. 1989. Looking at the question in another way, among all the studies which have examined violent behavior in individuals with severe mental illnesses, the study which reported the lowest rate of violence was the CATIE study in which all the patients were taking antipsychotic drugs.
Anecdotally there is abundant evidence to support the idea that homicides committed by individuals with severe mental illnesses are in most instances associated with the failure to treat (Jared Loughner being exhibit A) or the failure of the individuals to take his/her antipsychotic medication. There are literally hundreds of such examples on the Preventable Tragedies Database on the TAC website.
In conclusion, there is no evidence to support the allegation that homicides committed by individuals with severe mental illnesses are caused by the antipsychotic medication they are taking. Individuals making such allegations should be asked to cite the evidence for their allegation. On the contrary, studies strongly suggest that