Criminal offending, aggression and violence
How is criminal offending, aggression and violence related to schizophrenia?
Criminal offending covers a wide range of behaviours from destructive acts, stealing, sexual assaults, to physical assaults causing injury or death. The majority of people with schizophrenia will never commit a crime, however the few who do may help perpetuate a negative public stereotype that schizophrenia is associated with violent behaviour. It is difficult to determine whether any criminal acts committed by people with schizophrenia are a consequence of the illness or are traits of the particular individual. This is confounded by the fact that people with schizophrenia may be at high risk of exposure to the social factors that contribute to criminal offending, such as social disadvantage and substance abuse.
What is the evidence for criminal offending, aggression and violence?
Any criminal offending
Moderate to low quality evidence suggests arrest rates in people with schizophrenia or bipolar disorder are around 40%, which is similar to arrest rates in people with other mental disorders. There was a small increased risk of repeat offending in people with schizophrenia compared to people with depression, or compared to the general population. There were no differences in repeat offending rates when comparing people with schizophrenia to people with substance use disorders, mental retardation, or learning disabilities.
Moderate to low quality evidence suggests the prevalence of any aggression in people during a first episode of psychosis is around 31%, and the prevalence of serious aggression is around 16%.There was a small, decreased risk of aggression with better cognitive functioning. Moderate quality evidence suggests a large increased risk of inpatient aggression with a history of previous inpatient admissions, a small to medium-sized increased risk of inpatient aggression with a history of illicit substance abuse or involuntary admissions, and a small increased risk of inpatient aggression in males, people with schizophrenia, inpatients with a history of self-destructive behavior, and inpatients who are not married. Moderate to low quality evidence suggests a small increased risk of inpatient aggression in those with a history of violence, and in younger patients.
Moderate quality evidence suggests a small increased risk of violence in people with schizophrenia compared to the general population. The risk was lower in people with schizophrenia than in people with personality disorders. The factors associated with a large increased risk of violence were; previous violent victimisation, high verbal aggression, polysubstance use, non-adherence to psychological therapies, and previous hospital admissions. The factors associated with a medium-sized increased risk of violence were; homelessness, childhood maltreatment, aggression, hostility, any substance misuse, poor impulse control, psychopathy, antisocial personality disorder, a history of conviction, imprisonment, assault or involuntary hospital admission, and a lack of insight. The factors associated with a small increased risk of violence were; parental criminal involvement, parental alcohol misuse, previous suicide attempts, higher symptom scores, excitement and angry affect scores, non-white ethnicity, low socio-economic status, non-adherence to antipsychotic medication, and a history of self-destructive behaviour. For people with first-episode psychosis, involuntary treatment, hostility, having a forensic history, manic symptoms, illegal drug use, being male, being younger, and having a longer duration of untreated psychosis were associated with a medium-sized increase risk of violence.
Moderate to high quality evidence suggests a large increased risk of homicide in people with first-episode psychosis prior to treatment compared to after treatment. Prior to treatment, the rate of homicide in first-episode patients is around 0.16%, and after treatment it is around 0.01%. The proportion of stranger homicide by people with psychotic disorders is significantly lower than the proportion of other homicides.
Green - Topic summary is available.
Orange - Topic summary is being compiled.
Red - Topic summary has no current systematic review available.