Psychiatric disorder and offending Research findings

The relationship between psychiatric disorder and violence has attracted great interest because of the political sensitivity of the issue and the public policy implications. Teasing out the nature of the relationship is necessary to inform the task of identifying and managing risk in psychiatric practice (see Ch§pter..11.4.3). However, the association between psychiatric disorder and other types of offending is equally important and difficult to establish. We have referred above to the methodological hurdles in establishing the relationship between psychiatric disorder and offending. More reliable findings emanate from research based on the following:

• large and comprehensive samples

• reliable psychiatric case registers to identify admissions

• research diagnostic criteria for defining psychiatric disorder

• measurement of non-recorded as well as recorded crime

• lengthy periods of follow-up or data collection.

Some of these difficulties have been addressed in recent research based on population surveys, birth cohorts, and the longitudinal careers of psychiatric offenders; no research is likely to address all the methodological problems. Some examine only violent offending, while others take a wider perspective.

Population surveys

The Epidemiologic Catchment Area Survey by Swanson et al.(58) in the United States collected data on adults with psychiatric disorder living in the community and included questions on violent behaviour. A total of 10 000 respondents living in Baltimore, Raleigh-Durham, and Los Angeles were interviewed for the presence of certain psychiatric disorders (major mental illness, substance abuse, and some neurotic disorders) in the last 12 months. Personality disorder and mental retardation were not examined. Data on self-reported violence in the last 12 months was collected. Of those who reported some violence, 55 per cent had a psychiatric disorder compared with 20 per cent of the non-violent responders. The most common diagnosis in violent respondents was substance misuse (10 times more common) followed by affective and schizophrenic disorders (four times more common), with only a slight increase in neurotic disorders. However, of greater significance seems to be the prevalence of violence when there were multiple psychiatric disorders. The incidence of violent subjects with one, two, or three and more diagnoses was 7 per cent, 18 per cent, and 22 per cent respectively.

The problem of multiple diagnosis was addressed by Stueve and Link(59> in a community-based study in Israel of 2678 young adults. For those with a psychiatric disorder a five-category hierarchical system was applied: psychotic disorder, bipolar disorder, non-psychotic depressive disorder, anxiety disorder, and phobias. Separate assessments were made of lifetime presence of a substance abuse disorder and of antisocial personality disorder. Violence was assessed by self-report. The researchers found a significant association between fighting and weapon-use and psychotic disorder and bipolar disorder, while controlling for substance abuse disorder, antisocial personality disorder, and demographic factors. Non-psychotic depression, anxiety, and phobic disorders showed no association.

Examination of non-violent as well as violent crime in a New York population sample by Link et al.(69 showed those who had ever had psychiatric contact differed from those who had never had contact only in respect of violent crimes; non-violent crime was equally prevalent in the two groups.

Birth cohorts

Studies of birth cohorts provide data based on records rather than interview, but make possible lengthy follow-up of large and complete samples. Hodgins (61> examined the mental hospital and criminal records for the subsequent 30 years of all 15 117 persons born in Stockholm in 1953 and residing there in 1963. Diagnostic categories were major mental disorder, intellectual handicap, substance abuse, and other disorder; thus personality disorder was not separately identified. The sample did not include never-admitted patients. Men with a major mental disorder were 2.5 times more likely to have committed any offence and four times more likely to have committed a violent offence. In women the equivalent increased rates were five times and 27 times respectively. In those with intellectual handicap, violent offending was five times more likely in men, and 25 times more likely in women, than in those with no psychiatric admissions.

Broadly similar findings were reported by Tiihonen et al.(62.) in a cohort of all 12 058 births in 1966 in Northern Finland followed up for 26 years. Men with an admission for a major mental disorder (including alcohol-related disorder but excluding personality disorder and mental retardation) were up to 40 times more likely than controls to have a criminal conviction by the age of 26. These were mostly minor offences by men with alcohol-related disorders. But those men with schizophrenia or a psychotic mood disorder were up to nine times more likely than controls to have committed a crime of violence.

The largest birth cohort study is by Hodgins et al.{63) in which all Danish citizens born in the 4 years between 1944 and 1947 were followed up until aged 43. This was a linkage study of psychiatric admissions and criminal records. Subjects with at least one admission were assigned in a hierarchy to one of seven diagnostic categories: major mental disorder, mental retardation, organic disorder, antisocial personality disorder, drug use disorder, alcohol use disorder, and other mental disorder.

The relative risk estimates described by Hodgins et al.(63> (I.abie.3) are for each diagnostic group compared with the never-admitted subjects. The increased risk for committing any type of crime in all diagnostic groups was substantial—particularly in men with alcohol or drug misuse, antisocial personality disorder, and mental retardation. For crimes of violence, the risk in females was greater than in males because of the lower base rate for violent crime in women.

11 Q












H *







1É F





DMLlH'IcdfHil a. L

Table 3 Danish 1944-1947 birth cohort study: relative risk estimates of crime between 1978 and 1990

Longitudinal studies

These studies, which aim to follow the criminal careers of former psychiatric patients, can produce conflicting results that probably depend on the accuracy with which subsequent offending is identified and recorded. Two Swedish studies demonstrate the point. Lindqvist and Allebeck (64) reported a 17-year follow-up of 644 patients with schizophrenia discharged from mental hospitals in Stockholm during 1971. Although minor crimes of violence were four times higher in patients than controls, there was almost no major violent crime. Belfrage(65) suggests that peculiarities of the Swedish Police Register account for permanent loss of data on older offenders who survive lengthy periods without offending. He found 11 cases of murder, manslaughter, or attempted murder in 1056 patients with psychotic disorders discharged in 1986 and followed up for 10 years.

Identifying violent behaviour that did not result in criminal charges, and the use of a control group, were important methodological features in the MacArthur Violence Risk Assessment Study, a meticulous follow-up study by Steadman et a/.(66) of 1136 patients discharged from acute inpatient units in Pittsburgh, Kansas City, and Worcester, Massachusetts. Patients and a collateral informant were interviewed by researchers every 10 weeks for 1 year after discharge. A carefully matched community sample, with informants, were also interviewed, though only once. Results showed that discharged patients who did not abuse alcohol or drugs were no more violent than their community neighbours. However, a higher proportion of patients than controls abused alcohol or drugs. This led to increased violence in the patients during the first 20 weeks after discharge; thereafter there were no differences in the rates of violence between patients (whether or not abusing substances) and controls—86 per cent of violence by discharged patients was directed at family members or friends. Link and Stueve (67> commenting on the study, point to the remission of symptoms that is likely in the first 12 months after inpatient treatment as a significant factor in the low level of violence.


The literature concerning psychiatric disorder and offending confirm that this a complex area with, as yet, few incontrovertible facts established about the nature of the relationship, let alone the possible reasons for it. The bulk of the evidence to date does, however, point to certain conclusions.

• People with serious psychiatric disorders are more likely than members of the general public to acquire convictions for violent and other crimes (by factors of approximately 8 and 30 respectively).

• This increased likelihood is altered in strength by local factors such as the crime rate and sociodemographic variables.

• Antisocial personality disorder and substance misuse disorders have greater associations with offending than does major mental illness.

• A combination of psychiatric disorders (particularly when one is substance misuse disorder) may be more relevant than any single category of major mental illness.

• Most offending by those with major mental illness is minor in nature; violence when it occurs is likely to be targeted at a family member. Are psychiatric factors causal?

The relationship between psychiatric disorder and offending is unlikely to be one of exclusive causality. We have noted that an offence or crime is a particular form of behaviour that conflicts with the laws of society at a particular time. Like other behaviours or deeds or acts it is dependent on a complex interplay of forces. Determining the cause of any act may be as much a matter of philosophy or sociology as one of behavioural science or psychiatry. Any answer will depend largely on the perspective of the assessor and is untestable by scientific means, even though it may be 'proved' to the satisfaction of a criminal court.

Some would argue that any criminal act is the product of an abnormal mind and that therapy requires understanding the psychodynamic meaning of the act for that individual's mind. Exploration of early experiences, object relations, unconscious mechanisms, fantasies, and conscious thoughts provides the key to such understanding. The psychodynamic understanding of offending has become unfashionable, yet unconscious mechanisms can be identified even in those crimes that seem to have been driven by psychotic mental disorder. To that extent psychodynamic understanding of a criminal act should contribute to treatment in most mentally disordered offenders, though in the United Kingdom would more often than not be unavailable. For a full description of forensic psychotherapy see Cordess and Cox.(68)

In practice, when we seek to identify the role of psychiatric disorder in any criminal act, we are faced with a complex task that conflicts with our instinctive wish to have simple explanations for complex events. All crimes require a perpetrator, a victim, and a set of circumstances; crimes committed by the mentally ill are no exception. The most simple assessment of what caused a particular crime to happen will therefore include assessment of the following:

• the perpetrator (mental and physical health, personality)

• the victim (relationship with perpetrator, contributory factors)

• the situation (location, whether alcohol or drugs present).

Psychiatric assessment of offenders is concerned with the mental condition of the offender and in particular the relevance of the mental condition in negotiating the legal hurdles described in Chapter J 1.J . Assessment of the psychiatric condition of an offender is described further in Chapter 1...1..6. It requires the same process as any other psychiatric evaluation; history-taking and mental state examination are the cornerstones, supplemented by appropriate special tests and information from other sources. In practice, the great majority of offenders who have a mental disorder have evidence of that disorder before the index offence. Indeed, a mentally disordered offender who has had no previous contact with medical or psychiatric services is unusual. In most cases the offence is the latest occasion on which an already present psychiatric disorder manifested itself. Cases in which, but for the offence, there would have been no contact with medical or psychiatric services are bound to raise suspicion that the significance of any disorder is being exaggerated, possibly in an attempt to meet the particular criteria of the legal process.

Degrees of causality

The nature of the relationship between the offence and the disorder requires consideration. In practice, the strength of this relationship can vary on a hypothetical scale of 0 to 10. Sometimes the presence of a psychiatric disorder may have a minimal relationship to the commission of the offence (1 to 2 on the hypothetical scale). Conversely, in other cases the disorder may provide a complete explanation such that without the illness there would have been no offence (8 to 10). Between these extremes lie the majority of cases where psychiatric disorder operates (3 to 7) with other factors, such as situational events, relationship problems, or intoxication, in contributing to the offence.

Social and other interactions with psychiatric disorder

A systematic and hierarchical approach has been suggested by Hiday (69> in respect of mental illness and violent crime, but her thesis has relevance for other types of offence. She suggests a series of mechanisms, some direct and some indirect, by which mental illness may be related to offending.

Direct: severe mental illness leads to criminal behaviour

In this, the most traditional mechanism, various symptoms of illness including, but not exclusively, delusions and hallucinations cause the offending behaviour. Direct: severe mental illness plus threat-control override

Only those symptoms that produce a sense of personal threat within the mentally ill person are significant because these are the symptoms that override normal self-control, so-called 'threat-control override' (see below and Link and Stueve (79).

Indirect: symptoms frighten others and cause tense situations

Here threat-control override symptoms cause others to be frightened, thus generating situations of tension. Attempts to challenge the symptoms or to ask the person to desist in his frightening behaviour or to persuade him to comply with treatment aggravate the tension. This mechanism may be recognized in violence by psychiatric patients in hospital settings, but it is also a factor in those community settings where violence is commonly used in disputes.

Breaking Bulimia

Breaking Bulimia

We have all been there: turning to the refrigerator if feeling lonely or bored or indulging in seconds or thirds if strained. But if you suffer from bulimia, the from time to time urge to overeat is more like an obsession.

Get My Free Ebook

Post a comment