Aggression Research

Some individuals are abnormally aggressive. That is, they respond with disproportionate retaliation relative to provocation and can escalate their behavior to violate the safety and well-being of themselves and others. Such behavior is repeated over the individual’s lifespan, resulting in severe consequences for the aggressor, ranging from job loss and divorce to severe injury and jail time. The loss of occupational and social functioning, and damage to victims, calls for the systematic study of individuals who express repeated aggression. A phenotype is a readily observable trait (e.g., brown eyes). The phenotype of aggressive behavior is also observable, though not as readily; still, teachers can point to the kids who initiated many physical fights this school year and friends know of their aggressive buddy who flew into a rage at the bar. Through interviews and tasks, we, too, can identify the aggressive phenotype and our aim is to study the multiple factors that are related to this particular behavior.

Developing an Inhibitory Paradigm: Just Say No?

Studies have shown that people vary in their ability to say "no" to things they are compelled to do even if these things are damaging to them. We set out to study how the word "no" is perceived by people and processed in their brains. The words "no" and “yes” are involved in learning to prohibit and encourage behavior, respectively. We hypothesized that these are emotional words, and as such would activate neural circuits involved with emotional control. We used fMRI to record brain activity while people were listening to these words. "No" and “yes” were associated with opposite brain-behavior responses.  While "no" was perceived as a negative word and produced a slowing of behavior, it also evoked a negative signal in the orbitofrontal cortex. By contrast, “yes” was perceived as a positive word, produced faster response times, and evoked a positive signal in the same frontal region. We also found that the more people reported controlling their anger, the more their orbitofrontal region was responsive to "no" (Alia-Klein et al., 2007).

Gene-Brain-Behavior Studies

Aggressive behavior results when the monoamine-oxidase-A (MAOA) gene is deleted. This relationship was initially found in one human family having a very rare gene mutation. It was later validated in gene knockout studies of mice in which this gene was experimentally deleted.

Monoamine oxidase A is an enzyme important for brain chemistry because it breaks down the monoamine neurotransmitters norepinephrine, serotonin, and dopamine. Through this mechanism, MAOA plays an important role during brain development and in brain functioning. The ability to map activity of the MAOA enzyme in vivo allowed us to ask whether MAOA brain activity is related to aggression. Brain MAOA activity was measured with PET using clorgyline labeled with carbon 11. We asked the individuals who were scanned to also complete a personality questionnaire where some of the questions were designed to capture aggressive personality traits. Results pointed to a relationship between aggressive personality and the activity of brain MAOA. People who reported more aggression in the questionnaire also had lower levels of the MAOA enzyme throughout their brain (Alia-Klein et al., 2008).

The genotype of genetic risk carriers of the gene allele for low MAOA activity may have an increased probability of expressing their anger, potentially through differences in brain functioning during emotional contexts. Emphatically expressed, the emotional word “no” described above (Alia-Klein et al., 2007) can come to inhibit behavior through associations of hearing “no” with stopping behavior during early childhood. We assessed brain activity during the emotional context established while people heard “no” among other words. Results showed that carriers of the low MAOA genotype had reduced frontal cortex brain activation to “no” compared with carriers of the high MAOA variant. Furthermore, and only for carriers of the low-MAOA genotype, the more brain regions such as the amygdala (associated with emotional response) and thalamus (associated with regulating arousal) activated in response to “no,” the more individuals reported expressing their anger in situations of negative evaluation and rejection (Alia-Klein et al., 2009).

Cocaine addiction is also known to be related to brain changes and difficulties with self-control. We therefore asked if there are differences in the MAOA genotype and in gray-matter volume in the brains of cocaine-addicted individuals as compared to healthy controls. We found that individuals with cocaine addiction showed reductions in gray-matter volume in the frontal cortex and temporal cortex and in the hippocampus as compared to controls. We found reductions in the orbitofrontal cortex (a region in the frontal cortex involved in evaluation and self-control) only in cocaine-addicted individuals who carry the low allele of the MAOA genotype. Furthermore, gray-matter reductions were related to the number of years cocaine and alcohol were used: the more years of use, the more reductions were observed in the dorsolateral prefrontal cortex (part of the frontal cortex involved in sustaining attention) and in the hippocampus (a region important for learning and memory) (Alia-Klein et al., 2011). Subsequent study of the MAOA genotype in cocaine addiction has revealed that carriers of the low allele are also more reactive to threatening and other emotionally unpleasant stimuli (pictures) (Moeller et al., 2014).

Other genotypes beyond MAOA also influence relevant phenotypes in addiction. For example, we have examined a gene involved with functioning of the dopamine transporter (DAT1). We found that carriers of a DAT1 9R-allele were more reactive to drug-related cues than carriers of two DAT1 10R alleles, especially when these 9R allele carriers were also in withdrawal from cocaine (Moeller et al., 2013).

Illness-to-Recovery: Insight

In problem behaviors, insight or awareness of the behaviors is a very important requirement for changing the behaviors. Some individuals’ awareness and appreciation of the severity of their illness and its impact on their disadvantageous behaviors (such as taking drugs or attacking others) are compromised. When insight is impaired there is less perceived need for treatment.

The vast majority of people who suffer from severe illness are neither aggressive nor violent. However, violence towards others by a minority of individuals is a significant public health concern and could be preventable. The severity of violence may be influenced by the extent to which individuals are aware they have a condition that influences their thinking and their behaviors.

We wanted to find out whether poor insight (in schizophrenia) is related to violent behavior and to problems with treatment adherence. We assessed a group of psychotic patients detained at a forensic unit in New York City through interviews, supplementing this data with information from hospital and official records, family members, and the treating clinician. Severity of violent behavior was related to poor adherence to medications and to poor insight and drug use (Alia-Klein et al., 2007).

We are now conducting similar studies on insight into severity of illness/self-awareness and association with treatment outcome in drug-addicted individuals (see reviews by Goldstein et al., 2009; Moeller and Goldstein, 2014). Our work has shown that cocaine-addicted individuals were less able than healthy controls to report correctly on their drug-related choices, a deficit associated with more severe drug taking (Moeller et al., 2010). We have since demonstrated that this unawareness of choice (Moeller et al., 2014), as well as more general metacognition (that is, thinking about thinking) deficits (Moeller et al., 2016), are associated with impaired function and structure of the rostral anterior cingulate cortex, a region that is involved with processing information relevant to oneself.