The term “paranoia,” derived from the Greek &lduo;para” (beside) and “nous” (mind), was coined as a descriptor of psychopathology by Heinroth in 1818.¹ By the end of the 19th century, 50% to 80% of patients in asylums in German-speaking coun­tries had received a diagnosis of paranoia.¹ Beginning in 1899, Kraepelin’s efforts to define paranoia more precisely resulted in a decrease in diagnoses of paranoia in favor of dementia praecox and, later, schizophrenia.^1,2 This narrowing of the definition of paranoia is reflected in current nosology and practice. In DSM-IV-TR, the prevalence of delusional disorder is estimated at 0.03% of the general population and accounts for 1% to 2% of psychiatric admissions. The prevalence of paranoid personality disorder is 0.5% to 2.5%; this condition accounts for 10% to 30% of psychiatric admissions.³

The association of paranoia with violence was most rigorously theorized in the French psychiatric literature by Jacques Lacan.⁴ In the context of ongoing efforts by psychiatrists to distinguish paranoia from schizophrenia, Lacan outlined a review of the literature on paranoia, parsing earlier theories into 2 main groups:

• The first argued that paranoia is largely psychogenic and integrated into, or continuous with, the individual’s preexisting personality and life story.
• The second argued for an organic basis for the disorder, citing the experience of strangeness at the time of onset, taking the form of an abrupt eruption of symptoms that the individual experiences as alien and meaningless.

Lacan synthesized these views by conceptualizing the latter eruption of symptoms as the sudden appearance of an affective state with an organic basis that was initially devoid of meaning. The symptom in a second step, became necessarily situated within the person’s life story and premorbid personality. Because of their role in the onset of psychotic illness, these early symptoms marked the spot within the psyche where the person was most vulnerable to future decompensation, including passage to the act of violence, with implications for treatment and prevention.⁵

The assessment of violence risk: risk status vs risk state
Contemporary violence risk assessment in both criminal and psychiatric populations can emphasize either the prediction of an individual’s level of future risk or the management of risk.⁶ The latter takes a dynamic approach, which focuses on variables that are susceptible to change with treatment.⁷ In this context, for the assessment of violence risk, it is more relevant to know whether a person is paranoid than whether he or she is a paranoiac, and to know what makes him less paranoid.

The 3 main types of contemporary risk assessment integrate information about paranoid states in distinct ways:

Clinical judgment. The most common approach-clinical judgment in usual practice settings-;takes paranoia into account in the diagnostic process, and not only in schizophre­nia, delusional disorder, or paranoid personality disorder. Many other psychiatric disorders are associated with paranoid delusions and even more with paranoid ideation. In his 2007 review, Freeman⁸ noted that in addition to the nearly 50% of patients with schizophrenia who have persecutory delusions, 44% with psychotic unipolar depression, 31% with de­mentia, and 28% of patients who have had a manic episode also have persecutory delusional thoughts.

Freeman⁸ also reported that in studies of a nonclinical population from which persons with psychotic disorders were excluded, at least 10% to 15% of individuals regularly experienced paranoid thoughts, and these were increased in the presence of anxiety or depression. Whether paranoid ideation is related to a risk of harm to others is then evaluated on a case-by-case basis in the context of other factors, including ones that may be highly specific to the individual.

Actuarial risk assessment. A second type of risk assessment, actuarial risk assessment, has not consistently integrated an assessment of paranoid ideas into its approach. This may be because of a lack of reliable constructs for measuring nondelusional paranoia in the types of large studies on which actuarial instruments are usually based. For instance, the MacArthur Violence Risk Assessment Study demonstrated that fear and anger are 2 of the strongest predictors of violence, which suggests that perceived threat or a feeling of being wronged by another are at issue in the individual’s violent behavior.⁹

But an analysis of delusions in the MacArthur study showed that the presence of delusions alone did not predict future violence in the 10-week period following discharge from a hospital¹⁰ The study did show an increased level of violence among men with delusions when gender was taken into account.¹¹ In their discussion, Appelbaum and colleagues¹⁰ propose that suspiciousness may be a more relevant predictor of violence than delusions.

The literature on the relationship between paranoid delusions and violence nonetheless indicates that under certain conditions paranoid delusions are associated with an increased risk of violence. Threats made by eroto­manic stalkers or unusually persistent litigants with delusional vindications should be taken seriously because they may lead to violence.^12,13 Persecutory delusions in patients with schizophrenia correlate with higher rates of violence than those with other types of delusions.^14,15

Threat/control-override (TCO) de­lusions, which can be conceptualized as a form of paranoia, involve the belief that one is being threatened or controlled by forces outside oneself. These delusions have been generally found to be associated with increased violence.¹¹

Structured clinical judgment. This type of risk assessment involves psychometrically validated categories that are broad and are rated on a clinical basis, using objective features of the case as well as subjective features gleaned from the clinical interview.

The most robust of this type of rating scales, the Historical, Clinical, Risk Management–20 (HCR-20), includes the possibility of accounting for paranoid ideas under the variables “major mental disorder” and “active psychotic symptoms.”¹⁶ However, there is no specific item that focuses on persecutory or paranoid ideas per se.

Although there have been tremendous advances in our understanding of risk factors, recent trends in violence risk-assessment research emphasize the need for an explanatory theory of the choice to act at a particular moment and the evolution of dynamic risk over time.¹⁷ These elements are difficult if not impossible to measure at the time that violence is occurring. Strategies for approaching an explanatory model include observation of the temporal convergence of risk factors and hypotheses regarding the possible existence of a final common pathway for known risk factors.

Susceptibility to transient paranoid ideation would appear to be a good candidate for such a convergent pathway for variables as diverse as previous violence, substance abuse, personality disorder, and exposure to environmental stressors. Each of these may contribute to the experience of perceived threat and feelings of fear or anger, while examining paranoid delusions per se as a risk factor for violence may have limited utility. The study of paranoid ideation as a dynamic process that varies over time under specific conditions may allow us to develop a picture of what happens immediately before an act of violence.