Clinical and Research News
Psychiatrists Disagree About Crime-Schizophrenia Link
Psychiatric News
Volume 39 Number 10 page 34-35

People with schizophrenia appear to have a higher rate of criminal and violent behavior than people in the general population.

That finding does not appear to be explained solely by the presence of active symptoms, the concurrence of substance abuse, or the effects of deinstitutionalization, according to a report in the April American Journal of Psychiatry.

Rather, the pattern of offending among patients with schizophrenia reflects a range of factors that appear to be operative before, during, and after periods of illness, said Paul E. Mullen, M.B.B.S., D.Sc., of the Victoria Institute of Forensic Mental Health, Victoria, Australia, and colleagues.

The controversial finding, however, is disputed by at least one expert who reviewed the report for Psychiatric News.

Darrel Regier, M.D., M.P.H., director of the APA Office of Research, commented that the study’s methods for controlling all potentially confounding factors that might influence criminal offending among people with schizophrenia are insufficient to support the conclusions.

Study author Mullen, however, insists the study undermines theories that claim a single cause for the association between criminal offending and schizophrenia, and suggests that the disorder may be one that "makes individuals more vulnerable to the criminogenic influences which determine crime in general."

Mullen said the higher rate of criminal offending should be taken into consideration by treating psychiatrists.

"The study suggests that many of the factors that produce offending in the general population are important in producing offending in schizophrenia," he said. "These include disturbed backgrounds, poor social conditions, unemployment, and substance abuse, among others. But those with schizophrenia may be more vulnerable to these influences. It argues for a much more holistic approach to the treatment of schizophrenia, taking far greater note of the psychological and social problems of these patients if we are to do anything to reduce offending in our patients."


In the study, the criminal records of 2,861 patients in five different cohorts based on year of first admission for schizophrenia—1975, 1980, 1985, 1990, 1995—were compared with those of an equal number of comparison subjects matched for age, gender, and neighborhood of residence.

The patient cohorts were drawn from the Australian state of Victoria’s psychiatric case registers. The comparison subjects were drawn from a database generated from criminal-record searches performed as part of the state’s process for selecting citizens for jury duty.

(Potential jurors are randomly selected from the state’s electoral rolls, and their criminal history is checked before they are placed on jury lists.)

Compared with the control subjects, the patients with schizophrenia accumulated a greater total number of criminal convictions (8,791 versus 1,119) and were significantly more likely to have been convicted (21.6 percent versus 7.8 percent). They were also more likely to have been convicted of an offense involving violence (8.2 percent versus 1.8 percent).

Mullen and colleagues also found that the proportion of patients who had a conviction increased from 14.8 percent of the 1975 first-admission cohort to 25 percent of the 1995 first-admission cohort.

But, importantly, a proportionately similar increase occurred among the comparison subjects. The proportionately similar increase in offending occurred despite a dramatic increase in substance abuse among patients with schizophrenia (8.3 percent in the 1975 cohort to 26.6 percent in the 1995 cohort).


The relative constancy in rates of offending over time is crucial, Mullen told Psychiatric News. It suggests that patients with schizophrenia who offend may have much more in common with offenders among the general population than generally thought, and that the association between schizophrenia and offending cannot be easily explained by such mediating factors as the effect of deinstitutionalization, the presence of active symptoms, or the concurrence of substance abuse.

For instance, he said, if deinstitutionalization accounted for offending among patients, the rate of increase in offending over the 25-year period (a period marked by deinstitutionalization) would have been greater than the rate of increase among the general population. Since the increase was proportionately similar, then offending among schizophrenia patients cannot be attributed solely to the effects of deinstitutionalization.

The same is true of the effect of substance abuse on offending. Substance abuse expanded dramatically during the 25-year period, even while the rate of increase in offending among patients with schizophrenia and the rate of increase among the general population was proportionately similar.

"The study does not demonstrate that there is no effect of deinstitutionalization or no effect of alcohol and drug abuse," Mullen told Psychiatric News. "What it demonstrates is that these effects, if they exist, are certainly not the whole story explaining the correlation between having schizophrenia and acting or at least being convicted of a criminal offense."

The same is true of the effect of active symptoms, Mullen said.

"If the active symptoms of the illness were the major determinant of offending, then the increase in offending in schizophrenia would occur after the emergence of those active symptoms," he said. "The time at which a patient is first diagnosed with an illness does not necessarily correspond to the time at which active symptoms emerge. But whenever those symptoms begin, there should be a skewed distribution in those with schizophrenia toward a later onset of offending than the general population of controls, and a continuation of higher rates to a later age."


The study received mixed reviews from experts who read the report for Psychiatric News.

Edward Mulvey, Ph.D., called the study a "very impressive effort" to examine the issue of offending among patients with schizophrenia.

"I have never been convinced that active symptomatology is the only component of the relationship between schizophrenia and violence," he told Psychiatric News. "The idea that people with schizophrenia are at increased risk for involvement in violence has been out there for a while and is not disputed. What is disputed is the reason for that relationship. The idea that active symptomatology is the only reasonable explanation is probably simplistic, and this study demonstrates that."

Mulvey is a professor of psychiatry and director of law and psychiatry research at the Western Psychiatric Institute of the University of Pittsburgh.

But APA’s Regier said the study provides no detailed analysis of level of treatment and clinical outcome that might have influenced the propensity to criminal offense.

Neither has the influence of socioeconomic background been sufficiently explored to determine its effect. Regier drew attention to the fact that none of the patients in the study was treated in the private sector.

"In the U.S. Epidemiologic Catchment Area study, most patients with schizophrenia and schizoaffective disorder were actually treated in the private sector," Regier told Psychiatric News. "If private sector patients were added, I have little doubt that this would diminish the magnitude of the difference [in criminal offending] between patients and controls. The authors haven’t adequately described the socioeconomic characteristics of patients and controls, an essential component of any case-control study of schizophrenia."

Regier is also executive director of the American Psychiatric Institute for Research and Education.

Regier said the fact that patients and controls were drawn from the same neighborhood does not necessarily mean they are from the same socioeconomic background.

He also drew attention to the fact that the control subjects were drawn from jury rolls. "The fact that the controls are drawn from jury rolls implies that none of the controls is homeless or keeping a low profile because of criminal behavior," he said. "It would be interesting to know how many of the patients were actually on jury rolls as a basis of comparison."

Mulvey, too, said that "an inherent problem with a study like this is that there is really no controlling for the types of social contextual factors and opportunities for violence that exist over the lifetime of an individual with schizophrenia."

He added, "We know from other research that having schizophrenia produces a social slide downward, and living in a tough neighborhood gets you in more fights. We can’t ever adequately control for something like that."


However controversial, the phenomenon of criminal offending among patients with schizophrenia is a matter of clinical and public policy concern—a point on which Mullen, Mulvey, and Regier agreed.

"I think that there is little doubt that patients with schizophrenia who develop command delusions or hallucinations are a danger to themselves and others if they are not medicated, and particularly if they turn to alcohol and other drugs to medicate themselves," Regier said. "Hence, I think there is clear risk associated with schizophrenia in a deinstitutionalized context where inadequate services are provided to maintain such patients in a functional state in the community.

"The message for psychiatrists and policymakers is that failure to provide adequate treatment resources for patients with schizophrenia, particularly in the public sector, can result in an increased social burden of criminal offending," he added. "This is particularly true if patients with dual diagnoses of mental and addictive disorders are left untreated."

The study, "Criminal Offending in Schizophrenia Over a 25-Year Period Marked by Deinstitutionalization and Increasing Prevalence of Comorbid Substance Use Disorders," is posted online at ajp.psychiatryonline.org/cgi/content/full/161/4/716?.

Interactive Graphics


Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Related Articles