A range of findings indicate that aggression and criminal behaviour are to some extent genetically inherited (Cadoret et al. 1995). To date, no genome-wide association studies have been published on aggression and violence. The candidate genes being studied in this context include those associated with the personality trait “novelty seeking,” impulsivity or hostility, and also ADHD. Serotonin has been postulated as the major neurotransmitter in the regulation of aggression (Siever 2008), but enhancement of central dopaminergic or noradrenergic function is also of relevance (Comai et al. 2012). Since the serotonin system plays an important role in anxiety, impulsivity, and aggression on the genetic level, polymorphic genetic variants of the serotonergic system have been studied as possible genetic vulnerability markers for aggression (Beitchman et al. 2004). Among these variants are functional polymorphisms in monoamine oxidase A (MAO A) and the serotonin transporter (5-HTT) (Brunner et al. 1993; Fresan et al. 2007), for a review of the genetic determinants of aggression, see (Pavlov et al. 2012).

Catechol-O-methyltransferase (COMT) is one of the key enzymes in this area and involved in the degradation of dopamine, especially in the prefrontal cortex (Shehzad et al. 2012). In contrast to enzymes coded for by other candidate genes (e.g., monoamine oxidase), COMT metabolizes dopamine, noradrenalin, and adrenalin, but not serotonin. COMT has a functional polymorphism resulting from a single amino acid exchange in which valine (Val) is replaced by methionine (Met). The polymorphism is relatively common and was first reported at the beginning of the 1980s. Val/Met and Met/Met carriers have four- to fivefold lower COMT activity than Val/Val homozygotes (Lachman et al. 1996; Strous et al. 1997, 2003).

A recent meta-analysis of the Val158Met polymorphism and violent behaviour in schizophrenia for individuals with Met/Met genotype when compared to Val/Val genotype carriers suggests a trend association of Met/Met homozygosity and violence in men (Singh et al. 2012), also confirmed by another meta-analysis (Bhatka et al. 2012). In addition, a recent study in children suggests an association of the COMT gene with high aggression (Hirata et al. 2012).

A dysfunction of certain neural circuits responsible for emotional and impulse control and self-regulation is believed to be part of the neurobiological basis of aggression. Key structures include the anterior cingulate cortex, the limbic structures, the amygdala, and the prefrontal cortex.

There are a number of theories on the brain structures, abnormalities, and neuromodulators underlying aggression. On the cortical level, lesions in certain areas, a decreased cortical volume – possibly linked to developmental disorders – and an orbitofrontal/cingulated cortex processing inefficiency have been discussed, together with a reduced serotonergic and enhanced dopamine/norepinephrine activity (Siever 2008). On the limbic level, a hyperactivity of the amygdala or other limbic structures, a reduced amygdalar volume, the role of hypersensitivity and kindling phenomena, a reduced GABA, and enhanced glutamate and acetylcholine activity have been proposed (Siever 2008). A prefrontal dysfunction has especially been linked to impulsive aggression (Brower and Price 2001). Hyperactivity of the limbic system in response to negative or provocative stimuli was shown in particular in antisocial and borderline patients (Herpetz et al. 2001; Donegan et al. 2003; Schmahl et al. 2003, 2004; Minzenberg et al. 2007) and is discussed also as an important mechanism underlying impulsive aggression (Palijan et al. 2010).

(For a review, see Wahlund and Kristiansson 2009; Dolan 2010; Anderson and Kiehl 2012).

Structural imaging has shown reductions in prefrontal grey matter (Raine et al. 2000; Narayan et al. 2007; Müller et al. 2008; Glenn et al. 2010; Yang et al. 2005, 2009; de Oliveira-Souza et al. 2008; Tiihonen et al. 2008), although some results are conflicting (Laakso et al. 2002; Dolan et al. 2002; Barkataki et al. 2006). In addition, volume reductions in the left frontal cortex and right anterior cingulated cortex and the temporolimbic cortex, including the hippocampus, were found (Glenn et al. 2010; Yang et al. 2009; de Oliveira-Souza et al. 2008; Dolan et al. 2002; Barkataki et al. 2006; Hazlett et al. 2005; Raine et al. 2004; Matsuo et al. 2008; Antonuci et al. 2006). Tiihonen et al. (2008) studied a group of persistently violent offenders and found larger white matter volumes in various brain regions and reduced grey matter volume in the postcentral gyri, frontopolar cortex, and orbitofrontal cortex. The authors discussed whether the larger volumes in posterior brain areas may reflect atypical neurodevelopmental processes that underlie early-onset persistent antisocial and aggressive behaviour.

An excellent study on this topic was recently published by Schiffer et al. (2011) who examined patients from forensic settings and controlled for substance use disorders. Compared to nonoffenders, violent offenders presented with a larger grey matter volume in the amygdala bilaterally, the left nucleus accumbens, and the right caudate head and less grey matter volume in the left insula. Men with substance use disorders exhibited a smaller grey matter volume in the orbitofrontal cortex, ventromedial prefrontal cortex, and premotor cortex than patients without substance use. Regression analysis showed that the alterations in grey matter volume that distinguished the violent offenders from nonoffenders were associated with psychopathy scores and scores for lifelong aggressive behaviour. The grey matter volumes of the orbitofrontal and prefrontal cortex that distinguished the men with substance use from those without were correlated with scores for response inhibition. The authors concluded that a greater grey matter volume in the mesolimbic reward system may be associated with violent behaviour and that reduced grey matter volumes in the prefrontal cortex, orbitofrontal cortex, and premotor area characterize men with substance use disorders.

Data on functional neuroimaging are largely restricted to PET studies, mostly using FDG-PET, in part with challenge tests, and very few SPECT data are available. In the field of aggression and violence, PET and SPECT are only used for research and scientific purposes, not for clinical diagnosis or psychiatric or forensic assessment.

In general, far fewer data are available from functional imaging studies than from structural imaging studies. Early studies using FDG-PET revealed decreased glucose metabolism in the temporal and frontal cortices of patients with violent behaviour (Volkow et al. 1995). A resting FDG-PET study in borderline patients revealed an inverse correlation between lifetime aggression and metabolism in the orbitofrontal cortex (Goyer et al. 1994). Later on these findings were replicated in murderers (Raine et al. 1994, 1997, 1998). In an FDG-PET study, George et al. (2004) reported a decreased glucose metabolism in the right hypothalamus and reduced relationships between cortical and subcortical brain structures in perpetrators of domestic violence. Another PET study in a small sample of borderline patients revealed an inverse relationship between a history of impulsive aggressive behaviour and glucose metabolism in the prefrontal cortex, Brodmann’s area 46 (Goyer et al. 1994). A challenge study with mCPP in patients with a history of physical aggression showed decrements in the lateral, medial, and frontal cortices at baseline (New et al. 2007). Another PET study on laboratory-induced aggression in patients from the same group with borderline personality disorder (New et al. 2009) showed an increased relative glucose metabolism in the orbital frontal cortex and amygdala in borderline patients with aggression, compared to controls. In contrast, an increased glucose metabolism was found in the anterior, medial, and dorsolateral prefrontal regions during provocation in healthy controls compared to borderline patients. The authors concluded that aggressive patients showed increased glucose metabolism in “emotional” brain areas and not in those areas associated with cognitive control of aggression.

Other studies used laboratory-induced aggression or challenge tests as a model to understand the anatomy of aggression. An experimental imaging study in healthy volunteers imagining aggressive behaviour showed blood flow reductions in the orbital frontal cortex (Pietrini et al. 2000). Another study on laboratory-induced aggression in patients with borderline personality disorder showed diminished responses to provocation in the medial frontal cortex and the anterior frontal cortex but greater responses in the orbitofrontal cortex compared to controls (New et al. 2006). Previously, two functional imaging studies showed an association of anger induction and activation of the orbitofrontal cortex in healthy adults (Dougherty et al. 1999; Kimbrell et al. 1999). Taken together, these neuroimaging findings give further evidence for the contribution of the ventromedial cortex, limbic system, amygdala, and thalamus to impulsivity control and aggression, which was previously shown in individual’s brain damage in these regions (Berlin et al. 2004; Grafman et al. 1996; Bechara et al. 1999).

A number of other neuroimaging studies were performed not on aggression per se but on psychopathy, a related phenotype (for a review see Anderson and Kiehl (2012) and Yang and Raine 2009), mostly challenge tasks using psychological tests to study emotional responses to different scenarios. In sum, these studies also suggest that the hypoactive prefrontal cortex is involved in mediating the clinical features of psychopathy.

Very few imaging studies have been performed on neurotransmitter function in aggressive individuals, some in healthy individuals.

MAO A is an enzyme involved in the release and degradation of dopamine and serotonin. MAO A genotype may be associated with aggression, as shown in some studies (Meyer-Lindenberg et al. 2006; Kim-Cohen et al. 2006; Newman et al. 2005). Caspi et al. (2002) demonstrated that low MAO genotype predicts high trait aggression in men with childhood trauma and MAO A genotype is one of the possible genetic determinants of aggression and impulsivity (Pavlov et al. 2012). A PET study measuring trait aggression using clogyline labelled with carbon 11 as the radioligand in healthy males (N = 27) found an inverse correlation between the measured amount of aggression and the multidimensional personality questionnaire (Alia-Klein et al. 2008). These data are not derived from patients with aggression or from forensic samples but are nevertheless interesting and suggest that lower MAO A activity in cortical and subcortical brain regions may predict aggressive or antisocial behaviour.

The relationship between MAO A binding and maladaptive personality traits has also been studied by Soliman et al. (2011). The group studied healthy nonsmokers using [¹¹C] harmine PET in prefrontal regions. In addition, personality traits were measured. Prefrontal MAO A binding correlated negatively with anger-hostility and positively with deliberation. In a two-factor regression model, these factors explained 35% of variance in prefrontal MAO A binding.

Other PET studies focused on serotonergic transmission or a serotonergic deficiency and its relationship with aggression. 5-HT (1a) and 5-HT (1B) receptors are very probably linked to aggression (De Boer and Koolhaas 2005). Correspondingly, Witte et al. (2009) studied 33 healthy volunteers using the radioligand (carbonyl-(11)C)WAY-100635 to quantify 5-HT (1A) binding potentials in the prefrontal cortex, limbic areas, and midbrain. In addition, testosterone and other hormone levels were measured. Statistical analysis revealed higher 5-HT (1A) receptor binding in subjects exhibiting higher aggression scores in prefrontal and anterior cingulated cortices. The authors discussed a reduced downstream control due to higher activity of frontal 5-HT (1A) receptors in more aggressive subjects, presumably modulated by sex hormones.

Since there is robust evidence from postmortem, in vivo imaging and genetic studies that the serotonin transmitter system is involved in the regulation of impulsive aggression PET studies have addressed this system. Da Cunha-Bang et al. (2013) studied trait aggression and impulsivity in healthy individuals. Trait aggression and impulsivity were assessed with the Buss-Perry Aggression Questionnaire and the Barratt Impulsiveness Scale. The 5T2A receptor binding was measured in a PET study. 94 individuals were included. Contrary to the hypothesis, results revealed no significant associations with 5-HT2AR and the psychopathology scales.

An interesting study has been published by Booij et al. (2010). The group studied the brain serotonin synthesis in a 21-year longitudinal study in individuals with a childhood-limited physical aggression (N = 8) and with normal level of aggression (N = 18) in a PET study using the tracer [¹¹C]methyl-l-tryptophan (¹¹C-AMT). Individuals with a history of aggression had significantly lower trapping of ^11C-AMT bilaterally in the orbitofrontal cortex and self-reported impulsiveness. Despite this, in adulthood there were no group differences in plasma tryptophan levels, genotyping, aggression, emotional intelligence, working memory, computerized measures of impulsivity, psychosocial functioning/adjustment, and personal and family history of mood and substance use disorders. The authors concluded that these results force a re-examination of the low 5-HT hypothesis as central in the biology of violence.