DISCUSSION
We performed a forensic psychiatric assessment of a sample of 28 patients that were institutionalized in a forensic psychiatric facility (custodial hospital) in Rio de Janeiro, Brazil, for offenses involving violent behavior. Many of these women had little education and were single (57%), were not working (32%), and had low family income at the time of the crime. The main psychiatric diagnosis was schizophrenia (n=13; 46%). Our findings corroborate the results reported in the literature, which has identified schizophrenia as the main mental disorder associated with aggressive behavior in both developed17-19 and developing countries20,21.
Few studies in Latin America have addressed aggressive behavior in individuals with schizophrenia. One study22 included 253 stable outpatients. Prevalence of aggressive behavior in the sample was 3.5% in Chile, 14.6% in Peru, and 55.4% in Bolivia. Aggressive behavior was associated with the severity of psychotic symptoms, low family income, young age at onset of the illness, and more hospitalizations. Most of these variables were present in our sample.
Analysis of the means and standard deviations for positive and negative symptom scores in our sample using the Positive and Negative Syndrome Scale (PANSS)16 showed lower scores for positive symptoms (12 ± 8.92) than for negative symptoms (38.5 ± 10.91), which may be related to the evolution of this severe mental disorder, with predominantly negative symptoms 23,24.
A history of previous violence has been considered consistently as predictive of subsequent violence in various patient populations25-27. Six patients (21%) in our sample had a history of aggressive behavior, two of whom had been involved in three criminal cases (with diagnoses of schizoaffective disorder and bipolar disorder, respectively) and one patient had been involved in five criminal cases (her diagnosis was schizophrenia).
The relationship between severe mental disorders and crime is more complex than simple causality. Factors such as age, gender, socioeconomic status, and prior crime are important, as are other potentially treatable factors such as substance abuse, personality disorders, and regular use of medications. Comorbidity with substance abuse increases the risk of aggressive behavior in individuals with severe mental disorders 28-33. According to Jaffeet al .34, individuals with substance use disorders and comorbidity with mental disorders show lower success with addiction treatment and more involvement in the criminal justice system. It is highly important to track substance use disorders in order to prevent aggressive behavior in these individuals.
On this issue, 10 patients (35.7%) in our sample reported alcohol and/or psychoactive substance use on the day of the offense. Curiously, all three patients with involvement in more than one criminal case reported such use. The association between mental disorders and violence can definitely be affected by various factors, including comorbidity with substance use, negative life events, and low social support. Public health strategies to reduce violence in individuals with and without mental disorders should focus on substance abuse prevention.
Fourteen patients (50%) in the total sample (n=28) presented psychotic symptoms at the time of the initial forensic psychiatric assessment resulting in involuntary hospitalization. Auditory hallucinations (n=8 cases) and persecutory delusions (n=6 cases) were the most common psychotic symptoms. These data illustrate the relevance of psychotic symptoms prior to manifestation of the violent behavior. Another key finding was that only four patients (14.3%) in the sample were in psychiatric treatment or in use of psychiatric medication on the days preceding the offense. We believe that the lack of treatment may have contributed to perpetration of the violent behavior.
Homicide and unusual forms of aggression often result from symptoms leading persons with mental disorders to believe they are in danger. A study by Taylor35 found a strong association between psychotic symptoms and recent aggressive behavior, since 93% of the sample displayed psychotic symptoms when they committed these offenses, and 47% were “definitely” or “probably” motivated by these symptoms. Other studies have found an association between auditory hallucinations and persecutory delusions and motivation to commit homicide36-40. More severe paranoia is associated with greater aggressiveness, even when controlling for such factors as impulsiveness, command hallucinations, treatment with antipsychotics, substance abuse, age, and gender41. Many homicide cases definitely involve individuals with personality disorders rather than other forms of mental illness42.
In this study’s sample, violent behavior was largely targeted to family members. Of the 30 victims of violent behavior, 14 (47%) were the patients’ family members, including a grandmother, mother, brother, husbands, and children. One patient had tried to drown her three children, who fortunately managed to survive. The most frequent means for committing the violent behavior was cold steel weapons (n=9, or 32% of the sample). A systematic review by Minero et al. 43 also found that cold steel weapons (knives, scissors, etc.) were the most frequent means used by individuals with psychotic disorders to commit homicide.
Severe mental illness has been associated with certain cases of familicidal behavior, with an emphasis on psychotic motivations. Other purportedly associated factors are: loss of family control; revenge for loss of the female partner; fear of abandonment; narcissistic rage; financial difficulties; altruistic urges to defend the family from real or imaginary catastrophes; and instrumental violence44-45. One patient in our sample poisoned and killed her two-year child as revenge for her male partner having left her for another woman. We recently published case reports of parricide46 and fratricide47 by individuals with psychotic disorders.
Various studies have provided evidence that the risk of aggressive behavior can increase in patients with affective disorders48-51. However, estimates of the risk of violence associated with bipolar disorder have not differed statistically from those associated with psychotic depression in many studies, often limited by the sample sizes. Manic symptoms such as impulsiveness, grandiose delusions, impaired judgment, and psychosis can also contribute to aggressive behavior in these patients52. Our study found only two cases of bipolar disorder. We believe that this finding may reflect the relatively small sample size.
A wide variety of mental disorders have been described in association with matricide, including schizophrenia53,54, depression53,55 personality disorders53,55, and alcohol and psychoactive substance abuse55,56. Our sample included one case of matricide and another of assault with a deadly weapon against the mother. The first patient had bipolar disorder and the second schizophrenia. We published a case report on the patient with bipolar disorder who killed her mother57. It should be noted that many cases of matricide are not associated with mental disorders58.
Five patients in the sample (18%) had committed or attempted filicide. Three of the five patients presented diagnoses of schizophrenia. The other two patients had diagnoses of mental retardation and borderline personality disorder (the latter being considered partially responsible to stand trial, thus her confinement to a forensic facility). Two of these five patients had a history of alcohol abuse. The fact that there was only one case of personality disorder may have been due to the fact the it was a sample of patients with more severe mental disorders, under more rigorous security. However, studies have found aggressive behavior against intimate partners59,60 and children61 in individuals with borderline personality disorder. Another study found that filicidal mothers, compared to fathers, were more subject to compulsory psychiatric hospitalization than to incarceration62. We published a report of two cases of filicide and attempted filicide, both with diagnosis of schizophrenia63.
Various studies have found a persistent pattern of interrupted contact with mental health services, while in others, the aggressive behavior appears to occur right after the onset of the mental disorder, before the offender has made contact with these services64. In our study, 21 of 28 patients (75%) had a history of psychiatric treatment. However, only four of the 28 patients (14.3%) were in psychiatric treatment shortly before the crime, and the same percentage were on psychoactive medication. The findings illustrate the fact that these patients with aggressive behavior were not in regular psychiatric treatment before they displayed this behavior. According to a cohort study65, persons with schizophrenia show lower risk of committing violent crimes when they are on antipsychotic medication, compared to periods when they fail to receive such treatment.
Psychiatric patients with a criminal history would benefit from treatment programs which nevertheless frequently exclude them. The presence of criminal history should be an indicator of increased need for integrative approaches, as opposed to exclusion from treatment. Mental health services should strive to prevent patients’ loss to follow-up and nonadherence to treatment, which often precede the aggressive behavior committed by persons with severe mental disorders. It is also essential for society and government authorities to mitigate barriers to psychiatric and psychosocial treatment.
Although the statistical and empirical evidences point to a direct relationship between serious mental disorders and aggressive behavior, this certainly represents a small proportion of the violence occurring in the community. In countries like Brazil with high rates of violence, and where violence and crime show a strong association with precarious socioeconomic conditions, the percentage of violent crimes such as homicides associated with mental disorders may be even smaller. The aim of studies on the association between violence and mental disorders is not to stigmatize, but to better understand the factors contributing to this association, as well as to propose mental health policies and therapeutic interventions for patients with mental disorders and aggressive behavior.
Current mental health policies and clinical practices have failed to recognize that aggressive behavior and victimization are problems for many patients with severe mental disorders. Thus, treatment services have not provided sufficient services to treat these problems, increasing the number of patients transferred to forensic services.
The media coverage on violence committed by individuals with serious mental disorders increases the alarm and stigma towards these individuals. They rarely commit violence, and they are actually the victims66,67 more often than the perpetrators of violence. Some studies have found that patients with mental illness were more prone to dying from homicide than persons in the general population68,69.
One limitation to our study was the small sample size (n=28). Our sample is definitely not adequate for obtaining epidemiological data on a potential association between mental illness and aggressive behavior in women. Our sample consisted of court referrals on aggressive behavior in women in a forensic facility and is thus not representative of the overall population of women with aggressive behavior. However, the sample included all the female patients that were committed to psychiatric hospitalization for violent behavior in the state Rio de Janeiro at the time of the study.
Another limitation was the use of retrospective data, although all the participating patients were interviewed during the study itself. Further studies are definitely necessary to explore the risk of violence in women with mental disorders from various populations and to assess the benefits of the therapeutic intervention in these factors, in reducing the risk of violence.
Although our study cannot be considered representative of all the women that committed homicide, we believe it can contribute to understanding the relationship between homicide and mental disorders in women. The study of motivating factors for violent behavior can provide knowledge for establishing therapeutic interventions in women with mental disorders that present risk of violent behaviors.