2

Framing the Paradigm

The opening panels of the workshop set the stage for the subsequent discussions. The keynote address provided an overview of the evidence for and the policy involving mental illness and its relationship to violence. Speakers presented on operational definitions, ecological frameworks, cultural context, risk and protective factors, and neurobiology. They noted the common misperception that mental illness plays a greater role in the risk of violence than it actually does. Although, under certain circumstances, persons with mental illness are indeed at a greater risk of violence to others and, in general, are at greater risk for suicide.

VIOLENCE AND MENTAL ILLNESS:
WHAT DO WE KNOW? WHAT DO WE NEED? WHAT CAN WE DO?¹

Mental health and violence are often addressed in a manner that adds to the confusion rather than the clarity, stated Tom Insel in his keynote address, and it is important to disseminate accurate and evidence-based information about the relationship between the two. Currently, he remarked, there is tremendous focus from the public on mass violence (e.g., shootings) and linking it to mental illness—a situation that requires untangling. Insel evoked President Barack Obama’s suggested action plan to address gun violence and mental health services, noting that these issues go beyond the

___________________

¹ This section summarizes information presented by Tom Insel, National Institute of Mental Health.

domain of criminal justice and point to social inequities. Public health tools are essential in reducing violence, Insel asserted.

At the highest levels of the U.S. government, there is both a desire to address the recent shootings in schools and public places and a hesitancy to directly address gun violence. This desire has translated into transforming mental health care to reduce additional violence. Furthermore, Insel remarked that the framing of mental health and violence on the same axis—though done with good intentions—has resulted in more misconceptions. To highlight this point, he shared the following data:

• Untreated active psychosis, whether because of mental illness or drug use, is associated with irrational behavior, which could include violence. Notably, 38 to 48 percent of the homicides and suicides associated with people who are diagnosed with schizophrenia or bipolar disorder occur at the beginning of illness, often before treatment and sometimes before a diagnosis (Nielssen et al., 2012; Short et al., 2013). That risk of violence is most likely to be directed toward family and friends.
• People with treated mental illness are at no higher risk for committing violence than the general population and are at higher risk for being the victims of violence. Scandinavian studies have indicated that treatment of mental illness can reduce violence risk 15-fold (Nielssen and Large, 2010).
• Violence associated with a diagnosed serious mental illness is more likely to be self-directed than directed at others, even if one includes family and friends. Ninety percent of the approximately 38,000 suicides each year in the United States involve mental illness, while less than 5 percent of the approximately 14,000 homicides each year involve mental illness (CDC, 2005). Insel calculated this to be a 50-fold differential. Occasionally, as was the case in the school shooting in Newtown, Connecticut, an event might include homicide before suicide.
• The risk of suicide is greater in people with mental illness than in the general population—almost half to three-quarters of the population-wide suicide risk can be explained by mental illness (Hawton and Heeringen, 2009). The lifetime risk of suicide for men with mental illness is 4.3 percent, and for women, it is 2.1 percent, whereas it is 0.7 percent and 0.2 percent for men and women without mental illness, respectively.
• Misinformation exists around a supposed increase in homicides in the United States when in fact it has decreased from 9.8 per 100,000 people in 1990 to 4.8 per 100,000 in 2010. On the other hand, suicide rates have remained relatively stable over the past

• 20 years. Other causes of mortality, including road traffic fatalities, have also dropped in that same time frame.

• The United States is disproportionately represented in firearm deaths among high-income countries, accounting for almost 80 percent of the total. For people 15 to 24 years old (i.e., the peak period for developing psychotic illness), the risk of homicide involving a firearm is 42 times higher and suicide involving a firearm is 8.8 times higher, relative to other countries (Richardson and Hemenway, 2011).

Insel referenced the popular perception that although the number of homicides has decreased, mass shootings have increased. He questioned this conclusion, as these are relatively rare events; while there is potential for this trend, the evidence is not clear. These tragic events capture national attention even though they are a small part of overall risk.

Because violence is a relatively rare event (even if it does not seem rare), it is difficult to predict at the individual level which prevention efforts will be the most effective. From a public health perspective, means restriction can work to reduce both homicide and suicide, he asserted. The knowledge on how to do so exists but can be a difficult sell in the policy arena. To this end, Insel emphasized that to the extent that mental illness is a risk factor for violence, treatment can help reduce it.

Treatment, however, is usually targeted to specific populations. Insel described the findings of a mapping exercise used to determine the domains in which suicides occur. Some populations, such as military personnel, do not constitute a large proportion of the overall group of suicides. Others, such as emergency department workers and health care providers, are a much larger percentage. Deaths from gun violence compose the largest portion. To reduce suicides, Insel proposed four measures for continued exploration, refinement, and improvement:

1. Predictors of risk and resilience, though he acknowledged this would be a difficult path
2. Surveillance, as data on suicides lag by almost 3 years
3. Tools for prevention and treatment
4. Evidenced-based policies for limiting access to means

An important measure that could be undertaken now is to better assist adolescents who are on the pathway to psychosis. In the United States, the duration of untreated psychosis is about 110 weeks—more than 2 years. One program the National Institute of Mental Health (NIMH) supports to address this issue is called RAISE (Recovery After an Initial Schizophrenia Episode), which ensures that following a diagnosis, a person receives a

package of treatment in a family-centered approach with a goal of improving function in addition to reducing symptoms. This program is currently being implemented in a few states, with data collection efforts under way to determine its effectiveness.

A second measure Insel recommended is to move upstream in terms of interventions. Early detection and early intervention have reduced mortality in other areas and have the potential to reduce violence, as well. Rather than focusing on the “21-year-old who’s been psychotic for 2 years and now gets a label in a treatment program, [focus should be placed on] the 15-year-old who’s at highest risk, and figuring out what could be provided to that 15-year-old so that at 19 he or she doesn’t have a psychotic illness,” Insel explained. He expressed that treatment is not only medication, but a whole series of interventions that build resilience and executive function, provide family psychoeducation and peer support, and improve other skills.

Insel closed with a few summary remarks:

• Most people with mental illness are not violent, and most acts of violence are not committed by people with mental illness.
• Some people with mental illness are a danger to themselves and others.
• Fear of those with mental illness confounds the assessment of risk (i.e., people with mental illness are more likely to be victims than perpetrators).
• Early detection and early treatment can reduce risk.

In the question-and-answer session following Insel’s presentation, additional salient points were raised. In terms of means restriction, participants discussed that the people who attempt suicide are often somewhat ambivalent about their success—restricting one’s means does not often result in the substitution of another means. Additionally, Insel emphasized that school-based interventions should not be focused on addressing mental illness, which often has not developed by adolescence, but rather on reframing the issue as improving adolescents’ school performance and relationship skills. The majority of those who demonstrate “precursors” do not develop mental illness, he noted.

OPERATIONAL DEFINITIONS FOR THE WORKSHOP²

Several workshop participants and speakers mentioned that confusion and miscommunication abound in the field of mental health, in no small

___________________

² This section summarizes information presented by Vickie Mays, University of California, Los Angeles.

part due to the ambiguity around terms. Vickie Mays noted that even the term “mental health” is confusing; some equate it with mental illness, while others place it on the side of well-being. To create a foundation for workshop discussions, Mays presented a series of operational definitions for common terminology in the field.

Mental health is defined as “a state of well-being in which the individual realizes his or her own abilities, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to his or her community” (WHO, 2001).

On mental illness, Mays acknowledged there is no one encompassing definition because perspectives, such as health care assessment and justice, often have different aims. However, despite this difficulty, stakeholders share a common goal of developing research that produces better predictors, interventions, and treatments.

On the other hand, severe mental illness (SMI) has a greater consensus in definition and comprises several disorders including bipolar disorder, depression, obsessive compulsive disorder (OCD), panic disorder, posttraumatic stress disorder (PTSD), and schizophrenia. SMI is disruptive, not only for individuals, but also for families, communities, and sometimes in the broader system. On the positive side, there are treatments, including not just medication but also therapies. One important policy direction is ensuring these treatments reach the people who need them. “In terms of serious mental illness, we need to remember that recovery is possible when we can get these treatments to people in an effective manner and in a timely manner,” she stated.

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), defines a mental disorder as “a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning” (APA, 2013). While Mays acknowledges some level of controversy around the DSM-5 system, she highlighted the key elements of the definition of a syndrome that is characterized by “clinically significant disturbance” in the areas of cognition, emotional regulation, and behavior (APA, 2013).

Violence, as defined by the World Health Organization (WHO), is the “intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment, or deprivation” (WHO, 1996). Mays emphasized the broad definition in thinking about interventions.

Conduct disorder refers to a group of behavioral and emotional problems that usually begin during childhood or teenage years. Children with the disorder have “long-term, continual patterns of behavior” that tend

to impact others or go against what is deemed typical by society for their age group (WHO, 1996). Mays explained that this particular concept was important because of the age of occurrence and because it is associated with more severe behavior later in life. She cautioned that some symptoms of conduct behavior, such as rule breaking, are also indicative of other disorders such as attention deficit hyperactivity disorder (ADHD) or abuse victimization, and that physical, family, school, and social factors could also explain the behavior. Moreover, several factors could contribute to the development of conduct disorder, including brain damage, child abuse, genetic vulnerability, school failure, and traumatic life experiences.

Alcohol use disorder is a pattern of consumption that results in clinically significant impairment or distress with a cluster of behavioral or physical symptoms, which can include withdrawal, craving, and tolerance. In the DSM-5, alcohol use disorder can be mild, moderate, or severe (APA, 2013).

Similarly, substance use disorder is marked by a constellation of cognitive, behavioral, and physiological symptoms and a continued use of a substance despite the harm it causes.

Perpetrator refers to the person who has actually committed the violent act or is responsible for it occurring. However, Mays stated that sometimes perpetrators have been or are currently victims themselves. Additionally, perpetrators are often addressed within the legal context, which only defines the person by his or her action at the time and does not address either the likelihood of reoccurrence or the measures that might prevent reoccurrence.

Victim refers to the person who has been “directly and proximately harmed as a result of the commission of an offense for which restitution may be ordered” (USSC, 2011). This is a legal definition, Mays noted. In terms of victims, stakeholders should examine the range of interventions needed to make a person whole and functional again.

Participants further discussed these definitions, noting that the variety of stakeholders with their different perspectives also have varying definitions. As science in the field advances and the biological basis of mental illness is further illuminated, these definitions can be refined. Depending on the aims and the principles of a particular discipline, however, definitions can be narrowed or expanded. Mays explained that these definitions are evolving in the context of a changing health care field, in which both services and actors are continuously being defined.

ECOLOGICAL FRAMEWORK

The ecological framework session included an overview and discussions of risk and protective factors and intervention points related to mental health and violence at the individual, relationship, community, and societal levels.

Models for Suicide Prevention and Treatment³

The focus of the workshop spans a range of problems that “encompass individuals’ unique life circumstances, communities, societies, and the globe all into one thoughtful discussion,” Eric Caine observed. This is chaotic, he noted, but holds possibility for intervention. On a national level, it is possible to prevent suicide. Notably, homicides and road traffic fatalities have decreased in part because of interventions when the outcomes were predictable at the population level, if not at the individual level. Although violence prevention has traditionally fallen under the purview of the criminal justice system, the responses have usually been reactionary. In the case of suicide, Caine explained, a reactionary response is one that is too late. He also observed that violence has a major global dimension: suicide, self-harm, interpersonal violence, war, and conflict are all very different around the world. However, looking at the burden of disease, suicide far outweighs the combined international impact of war and interpersonal violence.

In the United States, suicides not only carry a larger burden than homicides, but they are also substantially underestimated, particularly self-poisoning. Deliberate self-poisoning is not the same as unintentional poisoning, but occasionally “unintentional” means unable to determine intent and not specifically accidental. When the number of suicides is combined with the number of self-poisonings of uncertain intent, fatalities due to self-injury climb to approximately 50,000 per year. Caine also stated that firearms account for half of suicides, mostly due to their lethality (CDC, 2005). Because most people who attempt suicide and fail do not make another attempt, restricting lethal means could potentially reduce suicides, he argued.

The increase in suicides over the past decade has mostly been seen in the middle years of life in both men and women—not in youth. While the bulk of prevention resources is directed at youth, the suicide rate is highest among middle-aged white men, who compose a relatively small part of the population. The middle years are the most common time for women to commit suicide, as well. Caine further noted that suicides are not equally distributed across the United States. In ways not yet fully understood, suicide is driven by community and by location. In fact, suicides by poisoning and by firearm overlap significantly. There is also geographical variation in the suicide rate: it is higher in rural areas than in urban areas, whereas the opposite is true of homicides.

Across the United States, there is no uniform picture of self-directed violence or interpersonal violence. However, there are common

___________________

³ This section summarizes information presented by Eric Caine, University of Rochester Medical Center.

community-level factors that may lead to violent injuries. He noted that, although people often choose to focus on one type of violence or another, prevention should be addressed earlier in the trajectory.

Caine shared an ecological model from WHO, which explores the overlap among different levels in determining risk and protective factors of violence. There are risks at the individual, interpersonal, community, and societal levels, all of which offer potential points of intervention. In fact, some interventions occur at multiple levels or require cooperation with other interventions, such as youth and family prevention programs. Caine cautioned that exploring all levels and factors at one time can be overwhelming.

Risk is a cumulative “unfolding phenomenon,” he explained. All risks do not happen at once. This yields several points of potential intervention along a developmental pathway. He shared a model that depicted risks as a mountain range: the peaks are violent outcomes; the bases are larger social, economic, and family factors; and in the middle are individual and situational factors. In terms of interventions, the base includes those that are universal, while clinical ones are at the top; in the middle are selected and indicated interventions. Addressing the bottom and the middle would be necessary to make a difference, he stated.

Insel shared a second framework developed from the Haddon Matrix, which examines injury through the lens of multiple layers (i.e., individual, agent, and environment) and along a continuum of the event (i.e., pre-event, during, and post-event). This model, adapted for suicide, provides a starting point for the integration of neurobiological research and social research, as well as policy analysis.

He closed by sharing a third model focused on treatment, called the health impact pyramid. The pyramid shows that selected clinical interventions sit at the top and have a smaller population-level impact. Interventions at the bottom of the pyramid, which are more universal, have a greater population-level impact.

Culture, Mental Health, and Violence⁴

Cultural meaning is essential in considering the intersection of mental health and violence, and what it means for a person to accept treatment or medication, explained Janis Jenkins. A cultural lens highlights multiple perspectives and subjective experiences in a global, comparative context. From her research, which is team based and employs mixed methods including clinical diagnostic criteria, statistical analysis, and ethnography,

___________________

⁴ This section summarizes information presented by Janis Jenkins, University of California, San Diego.

she observed that culture overlaps sectors of the ecological framework and is central to mental health and illness. She described several fundamental aspects of mental illness that are shaped by culture:

• Risk and vulnerability factors (e.g., gender inequity)
• Symptom content, form, and constellation
• Clinical diagnostic process
• Illness experience: identification, explanatory model, and meaning
• Kin emotional response and bonds and attachment
• Community social support stigma
• Service use and preferred treatment modalities
• Resources for resilience and recovery
• Course and outcome

In terms of symptoms, Jenkins advised further attention to not only whether the symptom is present, but how it presents, as well as how symptoms aggregate. Furthermore, culture influences what is considered typical and atypical, how explanatory models are developed, how those with mental illness are viewed, how the illness course proceeds, and what outcomes develop. She further explained that whether mental illness is perceived as a personality defect or a legitimate illness affects the experience of illness as well as the recovery.

What is involved in culture and conceptualizing it? She noted that perception of the self is highly variable. In some places, the self is more individualistic or self-centered, while in others the self is more socio-centric. This orientation of the self influences assumptions about the world and one’s place in it, particularly in terms of labeling or self-identification. Emotions are also products of cultural systems and cultural rules.

In a cross-cultural survey of four African societies, perceptions of psychosis were examined (Edgerton, 1966). When queried on the behaviors associated with psychosis, participants responded in a similar manner: they were not tolerant of, and were concerned about, murder, assault, and disruptive behaviors. “These data show both a kind of universality to the conceptions of psychosis as well as some cultural specificity,” she stated.

In terms of violence, Jenkins shared ways in which culture can broaden the understanding of violence. Many developing countries have high rates of homicide, particularly Belize, Côte d’Ivoire, El Salvador, Guatemala, Honduras, Jamaica, Malawi, and Zambia; and among developed countries, the United States outstrips its peers. Violence is not uniform; rather, cultural, socioeconomic, and political factors influence violence and instability in these arenas can particularly disadvantage individuals with increased vulnerability to violence. Subjective dimensions and structural arrangements of violence, she postulated, could offer greater insight. These would include

conceptualizing violence as ordinary, contextually specific, lived experience, or an organized set of ideologies and practice. Structural violence, exerted systematically, causes harm or violence and disadvantages groups of people (Farmer, 2004). Examples include ethnocentrism, poverty, racism, and sexism, all of which constrain individual agency.

Jenkins further described research she had undertaken, investigating “the nexus among the role of the state in constructing a political ethos, the personal emotions of those who dwell in that ethos, and the mental health consequences for refugees or displaced persons” (Jenkins, 1991). Researchers have examined the experience of refugees fleeing political violence and identified terms to describe that experience, including “calor” or an intense feeling of heat that was presenting in the emergency room. They found in their study population that people had symptoms of PTSD and depression, but were also working hard to raise money to send home to their families. This raised the concept of “engaged depression” and the importance of resilience. She and her colleagues proposed that “calor” was a cultural manifestation of the political ethos of fear and violence inflicted by the state, presenting in a physical aspect. They used this concept to understand the connection among “symptom, emotion, culture, bodily experience, [and] political ethos.”

In her concluding remarks, Jenkins described research into youth violence and mental health issues, showing links with structural issues such as poverty. She also noted a lack of gender difference in both perpetration and victimization. The ethnographic analysis also suggested that youth who are preoccupied with issues of violence have comorbid mental health conditions, as well.

Discussion

Following the presentations, speakers delved into the concepts and themes they raised. They spoke of the failure of detecting mental illness related to violence before the occurrence of such violence, particularly suicide. Caine remarked that instead of focusing on individual risk, going “upstream” at the population level means examining life circumstances in the community and the family. The need to address the “bottom of the pyramid” is felt around the world, and there is “tremendous commonality around community engagement,” he continued. However, the focus is too often on suicide or homicide as an individual problem, partly because of the stigma around mental health. Caine also emphasized the importance of assessing the continuum of the problem rather than the event itself, which would include considering morbidity and disability when assessing the burden of violence, and not just at mortality. Jenkins added that addressing

these societal issues would mean rethinking the concept of resource scarcity and instead generating political will to build the needed capacity.

RELATIONSHIP BETWEEN MENTAL ILLNESS AND VIOLENCE⁵

Mark Rosenberg presented information prepared by Paul Appelbaum on the evidence base for the relationship between mental illness and violence. Rosenberg began with the four conclusions of the presentation:

• The public perceives a strong relationship between mental health and violence.
• Although the rates of violence are increased in patients with serious mental illness, the relative risk is moderate and well below those public popular perceptions.
• Mental illness accounts for only a small proportion of the overall violence risk.
• Beyond substance abuse, it is the case that hostility, suspiciousness, agitation, and psychotic experiences may further increase the risk of violent behavior; however, violence is varied and multicausal, which has implications for both prediction and treatment.

Rosenberg addressed definitional issues around the concept of risk, including perceived risk, measured risk, absolute risk versus relative risk, and population-attributable risk. The intersection of mental health and violence is highly dependent on how these terms are defined. Additionally, he asked, what is the mechanism for increased risk? Mental health includes not only the biological basis of mental illness, but also the orientation of the individual within the family, society, and greater cultural context. Violence can be defined as physical, psychological, or emotional, Rosenberg explained, and it can result in death or injury, both physical and psychological. Threat could also be considered a form of violence.

The definition of mental illness is inconsistent in the literature. Some studies look at schizophrenia only, and others look at serious mental illnesses. Some look at Axis I disorders, while others look at personality disorders, or a combination of the two. Substance use disorders can be included as well. Because of this lack of agreement, comparative analyses should be performed carefully to ensure comparisons are equal. Appelbaum’s intention was to focus on studies that examine violence toward others and their relationship with Axis I disorders and substance abuse, because methods and definitions are more refined in this area.

___________________

⁵ This section summarizes information presented by Mark Rosenberg, Task Force for Global Health, with information prepared by Paul Appelbaum, Columbia University.

In looking at the perceived risk of violence by people with mental illness, Rosenberg noted that the public sees a strong relationship. The General Social Survey (GSS) provides some illustrative data. For the GSS, people were queried about the risk of violence in a specific situation. The situation involved a hypothetical person named John whose mental health is described as deteriorating over a period of a few months, until he became housebound, neglected his hygiene, and began to hear voices. Participants in the survey responded that if John had schizophrenia, it was 61 percent likely that he would commit violence; if major depression, then 34 percent; and if drug dependence, then 87 percent. Rosenberg also cited a public opinion survey in which 46 percent of participants thought people with a serious mental illness are “by far more dangerous” than the general population; and one-third thought locating a group home for people with mental illness in a residential neighborhood endangered local residents (Smith et al., 2013).

Though these data indicate that perceived risk is vastly exaggerated, evidence does suggest that there is an incremental risk associated with mental illness. The Epidemiologic Catchment Area Surveys looked at violence in the year prior in a sample representative of the general population. Researchers defined violence as hitting or throwing things at a partner or spouse, hitting a child and causing injury, using a weapon in a fight, and fighting while drinking. They looked at the distribution of violent risk and found that the percentage of violent people in the group with no diagnosed disorder was 2 percent. By specific diagnostic groups, there was an increasingly elevated risk of violent behavior: 2.37 percent in those with anxiety disorders, 8.36 percent in those with schizophrenia, and up to 21.3 percent in those with substance use disorder (NIMH, 1991).

Rosenberg stated that Appelbaum examined a second dataset, the National Epidemiologic Survey on Alcohol and Related Conditions, which assessed violence in the past year and mental illness. For those with mental illness, the relative risk of violence was 2.0, serious mental illness was 3.5, and substance use disorder was 3.3; for serious mental illness plus substance use disorder, it was 11.5 (NIAAA, 2005). Notably, the highest risk was found with the combination of mental illness and substance use disorder. At the same time, Appelbaum cautions that relative risk is dependent on the comparison group, and the general population might not be the best control.

In another study, the MacArthur Violence Risk Assessment Study, 1,000 people were followed for 1 year after discharge and interviewed every 10 weeks. They were compared with people in their own neighborhoods. In the first 10 weeks after discharge, people with mental illness did have an elevated risk compared with the community, which was higher when substance use disorders were included. Additionally, when using the

appropriate control groups, the relative risk of violence for discharged individuals within the study community decreased slightly compared with results of relative risk in the general population. When controlling for substance abuse, this relative risk almost disappeared. It is the combination of substance abuse and mental illness that saw the highest increase.

Rosenberg also raised the issue of population-attributable risk. What proportion of violence in the population as a whole is due to mental illness? In other words, if mental illness were reduced, how much reduction would be seen in violence? A study in the United Kingdom found that the population-attributable risk for any personality disorder was 37 percent, for hazardous drinking it was 50 percent, and for antisocial personality disorder, 24 percent. Similar studies have found lower relative risk; mental illness is less important of a risk factor. Some studies suggest that substance use accounts for an increased risk, and others have found several risk factors related to violence, few of which are also related to mental illness.

Rosenberg closed by reiterating the four conclusions: the public perceives a strong association between mental illness and violence; rates of violence are increased but only moderately; only a small proportion of violence is attributable to mental health; and violence is variable with multiple causes and implications for treatment.

During the ensuing discussion, speakers raised additional issues around definitions. One issue is that there are no diagnostic categories for someone who is hostile all the time—there is a marked difference in the way angry affect is treated versus other types of affect. Another issue raised was measuring the adverse impact of exposure to violence on the mental health of children; violence prevention could be framed as mental health promotion in this respect.

NEUROCOGNITIVE MECHANISMS OF VIOLENT BEHAVIOR⁶

James Blair spoke about the neurocognitive systems that mediate or increase risk of interpersonal violence. He distinguished between two forms of interpersonal violence: reactive violence is frustration based or threat based, while instrumental violence is used to achieve a goal. Several mental health conditions increase the risk of reactive aggression, such as anxiety, borderline personality disorder, childhood bipolar disorder, depression, intermittent explosive disorder, and psychopathy. Whereas only one mental illness increases the risk of instrumental aggression—psychopathy (i.e., callous and unemotional [CU] traits). However, both types of aggression are normative behaviors; reactive aggression is the ultimate response to a

___________________

⁶ This section summarizes information presented by James Blair, National Institute of Mental Health.

threat, and in some circumstances, instrumental aggression might be the appropriate decision to make.

Blair explained the brain mechanism responsible for reactive aggression, a threat-response circuitry that includes the amygdala hypothalamus and extends into the periaqueductal gray. This neurocognitive system generates the response to a threat: in the distance, it might cause a person to freeze; closer, it might cause flight; and in very close proximity, it might result in fighting. This process is somewhat regulated by various frontal systems, as well. It is also highly responsive to the amount of stimulation—from low, freeze, to high, fight. He suggested that this should mean that individuals who are at heightened risk for reactive aggression should also have a heightened responsiveness of this circuitry. In fact, this is the case in brain scans of people with PTSD and other disorders known to increase risk of reactive aggression. Trauma and neglect also increase the responsiveness of this threat circuitry, and problems of emotional regulation block the ability to reduce the responsiveness.

In terms of instrumental aggression, Blair pointed to a dysfunction in empathic responsiveness that increases risk. This dysfunction manifests clinically in CU traits, such as low pro-social emotions, including a lack of remorse or guilt, lack of empathy, or lack of attachment to other individuals. This brain circuitry includes the amygdala and the ventral medial prefrontal cortex, as well. The amygdala is responsible for basic socialization, such as learning how others react to one’s actions. Depending on those reactions, one might choose to repeat or avoid that particular action in the future. However, if there is dysfunction in this circuitry, then there is an increase in CU traits and an inability to respond to the distress or pain of other individuals. Blair indicated that this inability to respond is not general, as there is no problem with recognizing anger or disgust.

A third brain mechanism described by Blair is a set of systems responsible for reward- and punishment-based decision making, which are not specifically related to CU traits. Problems in this circuitry tend to be prevalent across conduct disorders and to some extent in substance abuse populations. It also involves the amygdala, the ventral prefrontal cortex, and the caudate. Researchers have hypothesized, based on data from rat studies, that in the face of stimulus, a person might expect a positive outcome (i.e., reward) or a negative outcome (i.e., punishment). Normally, once a person determines which response will generate which outcome, he or she will continue the behavior that earns the reward. They learn to anticipate, or predict, which stimulus generates the reward and adapt their behavior accordingly when that feedback changes. However, in people with disruptive behavior or conduct disorder, this process does not occur. Additionally, those with conduct disorder show problems in the representation

of value in the ventral medial prefrontal cortex. These issues are also seen in people with substance use disorders, ADHD, and externalizing disorders.

In summary, Blair noted that the three neurocognitive systems he discussed might have a relationship with certain disorders, but are not disorder specific. The acute threat response, if overly responsive, is more likely to have an episode of reactive aggression. If an individual has empathic problems, then he or she will not be as responsive to the distress or pain of others and is less likely to be inhibited in causing harm. And those with problems in the reward-and-punishment circuitry have issues with externalizing disorders.

He speculated that there are additional factors that affect brain processes, such as poverty, which modulates decision making, and impoverished diet, which affects the development of brain structures such as the amygdala. Genetics, too, might play a role in increased responsiveness in the acute threat circuitry, and possibly other systems. And finally, he mentioned the role of alcohol, which in healthy individuals reduces response to distress of others and affects reward–punishment decision making.

In the discussion following the presentation, Blair addressed a question regarding how suicide plays out in these neurocognitive systems by noting that it is difficult to determine because the brain architecture explored does not generate self-harm behavior in mammalian species. Impulsivity plays a role, but the process is not necessarily within one of the three systems he described. While reactive and instrumental aggression are different processes, what is reactive is subjective and lies within the perception of the perpetrator and his or her social milieu. He also discussed the implication of this research on treatment; presumably, treatments that teach pro-social behavior should recalibrate these systems in individuals with mental health conditions. This seems to be true in several cases, but he noted that conduct disorders, for example, might also require pharmacology.

REFERENCES

NIAAA (National Institute on Alcohol Abuse and Alcoholism). 2005. National epidemiologic survey on alcohol and related conditions. Rockville, MD: U.S. Department of Health and Human Services, National Institute on Alcohol Abuse and Alcoholism.

Nielssen, O. B., and M. M. Large. 2010. Rates of homicide during the first episode of psychosis and after treatment: A systematic review and meta-analysis. Schizophrenia Bulletin 36(4):702–712.

Nielssen, O. B., G. S. Malhi, P. D. McGorry, and M. M. Large. 2012. Overview of violence to self and others during the first episode of psychosis. Journal of Clinical Psychiatry 73(5):e580–e587.

NIMH (National Institute of Mental Health). 1991. Epidemiologic catchment area survey of mental disorders, wave I (household), 1980–1985. Rockville, MD: U.S. Department of Health and Human Services, National Institute of Mental Health.

Richardson, E. G., and D. Hemenway. 2011. Homicide, suicide, and unintentional firearm fatality: Comparing the United States with other high-income countries. Journal of Trauma and Acute Care Surgery 70(1):238–243.

Short, T., S. Thomas, P. Mullen, and J. R. Ogloff. 2013. Comparing violence in schizophrenia patients with and without comorbid substance-use disorders to community controls. Acta Psychiatrica Scandinavica 128(4):306–313.

Smith, T. W., P. V. Marsden, and M. Hout. 2013. General social survey, 1972–2012. Chicago, IL: National Opinion Research Center.

USSC (U.S. Sentencing Commission). 2011. Guidelines manual. Washington, DC: U.S. Sentencing Commission.

WHO (World Health Organization). 1996. Violence: A public health priority. Geneva, Switzerland: World Health Organization.

WHO. 2001. Strengthening mental health promotion. Fact sheet no. 220. Geneva, Switzerland: World Health Organization.