Monoamine and neuroendocrine gene-sets associate with frustration-based aggression in a gender-specific manner
Introduction
Aggression has been defined as any behavior directed toward the goal of causing harm or injury to others (Baron and Richardson, 1994). From an evolutionary perspective, aggressive behaviors can be adaptive and have an important role in survival and competition for resources (Georgiev et al., 2013). In modern societies, aggression often is maladaptive and associated with negative consequences, causing psychological and somatic burden to victims as well as to aggressive individuals themselves (Fergusson et al., 2005, Reef et al., 2010). Aggression poses a substantial financial burden on society, for example caused by increased legal costs and work absence (WHO, 2007). A better understanding of the subtypes and etiology of aggression is needed to facilitate prevention and to improve treatment options (Fergusson et al., 2005). Given that about half of the variance in aggressive behaviors may be explained by genetic influences (Tuvblad and Baker, 2011, Veroude et al., 2016), studying the molecular genetics underlying these behaviors can provide important mechanistic insights. Research into aggression etiology is, however, complicated by several factors, including considerable phenotypic as well as genetic heterogeneity and the existence of sex differences in aggressive behaviors (Baker et al., 2008, Georgiev et al., 2013).
Heterogeneity in the etiology of aggression may be parsed by considering subtypes. Different classification systems have been proposed; one based on biological hypotheses is the distinction of proactive and reactive aggression (Dodge and Coie, 1987). Proactive aggression, also referred to as instrumental aggression, is goal-oriented, organized behavior often associated with low autonomic arousal and affect. Reactive aggression on the other hand, is also known as impulsive or affective aggression, and occurs in response to provocation or a negative emotional state (Raine et al., 2006, Stanford et al., 2003). Importantly, the subtypes have been associated with distinct behavioral, neurocognitive, and neural characteristics. For example, proactive aggression has been related to psychopathic traits and delinquent behavior (Cima and Raine, 2009, Cima et al., 2013), while the reactive subtype of aggression has been associated with impulsivity, anxiety, and hostile interpretation bias (Brugman et al., 2015, Bubier and Drabick, 2009). Twin studies showed slightly higher heritability estimates for proactive than reactive aggression (Baker et al., 2008, Brendgen et al., 2005, Tuvblad et al., 2009). The two aggression subtypes may have partially distinct genetic contributions. Serotonergic and dopaminergic neurotransmission may regulate both reactive and proactive aggression, whereas endocrine signaling seems to be more involved in the regulation of reactive aggression, e.g. through modulation of impulsivity and the stress response (Waltes et al., 2015). Recently, a further subdivision of reactive aggression has been proposed based on an exploratory factor analysis of the Reactive Proactive Questionnaire (RPQ). This analysis was conducted in a sample of adolescents (71.6% male), who were referred to clinical services for externalizing behavior problems (Smeets et al., 2016). Besides a proactive factor, reactive aggression was further subdivided into a subtype associated with external provocation or threat and another one associated with internal frustration. Improved fit indices for this three factor model compared to the original two-factor model were also reported based on an adult, males-only sample recruited partly in forensic psychiatric in- and outpatient clinics and partly from the general population. (Brugman et al., 2016). The reactive subtypes differed in their associated behavioral correlates, which suggests that the three-factor model may further reduce phenotypic heterogeneity and facilitate the search for genes involved in the etiology of aggression.
The most convincing observation supporting the existence of sex differences in aggression is the difference in crime rate statistics between males and females. Females are vastly less likely to commit serious offenses than males, and males are more likely to display antisocial behavior than females (Stephenson et al., 2014). Males are also overrepresented in aggression-related disorders such as conduct disorder (CD), where the gender ratio is approximately 2.5 (Hill, 2002). Importantly, sex differences are also found in the type of aggressive behavior displayed (Collett et al., 2003). The clear gender-specificity of aggression is thought to have evolved by sexual selection, and to reflect differences in optimal strategies in the competition for resources for males and females (Georgiev et al., 2013). Sex differences in heritability estimates have been observed in some but not all of the aggression twin studies conducted to date, with higher heritability estimates for boys than girls, when self-report measures were assessed (Baker et al., 2008, Wang et al., 2013). Incorporation of sex in aggression studies may be essential to identify the underlying biological mechanisms of aggressive behaviors.
The biological systems most investigated in the context of aggression phenotypes (as well as related traits such as mood disturbances and impulsivity) are the monoaminergic neurotransmitter systems related to serotonin and dopamine and the neuroendocrine system. Multiple reviews to date discuss these systems in the context of aggression and list the candidate genes that have been investigated for association with aggressive behaviors (Pavlov et al., 2012, Veroude et al., 2016, Waltes et al., 2015).
The serotonergic system is hypothesized to play a key role in aggression due to its influence on functions including social cognition, emotional regulation, and cognitive control (Lesch et al., 2012). Both human and animal studies link genes within these systems to aggressive behavior. For example, the serotonin transporter gene (SLC6A4) is one of the most investigated candidate genes for aggression. Variation in the serotonin receptor 2B gene (5-HT2B) has been associated with violent impulsivity in a Finnish population, and 5-HT2B and 5-HT1B knockout studies in mice implicate these genes in aggression and/or impulsivity (Bevilacqua et al., 2010, Nautiyal et al., 2015). While candidate genetic association studies have often produced equivocal results, investigations measuring levels of the serotonin metabolite 5-HIAA in cerebrospinal fluid, e.g. (Brown et al., 1979, Coccaro and Lee, 2010), or manipulating central serotonin function through tryptophan depletion/loading, e.g. (Bjork, 2000), have revealed a highly significant relationship between serotonin availability and aggression (Rosell and Siever, 2015). Dopamine is relevant for understanding aggression because of its effects on reward, motivated behavior, and decision making (Costa et al., 2012). While studies of dopamine manipulation have mostly been conducted in animals, the involvement of dopamine in aggression is also evidenced by the fact that in humans, D2-receptor antagonists have been used effectively to treat aggressive behavior (Nelson and Trainor, 2007). Additional evidence linking the serotonergic and dopaminergic neurotransmitter systems comes from genetic association studies of the MAOA gene. This X-linked gene encodes the enzyme monoamine oxidase A, which breaks down both serotonin and dopamine, and has been robustly associated with aggression, especially in the context of stress and maltreatment (Brunner et al., 1993; Caspi et al., 2002; Byrd and Manuck, 2014). The third system implicated in aggression is the neuroendocrine system, including both stress-related hypothalamic-pituitary-adrenal (HPA) axis signaling and sex-hormone-related hypothalamo-pituitary-gonadal (HPG) axis signaling. As early life stress is known to increase risk for the development of mood and aggression-related disorders (Agid et al., 1999, Éthier et al., 2004, Fonagy, 2006, Heim et al., 2001), the neuroendocrine stress response with its genetic components is a major candidate system for the development of aggressive behaviors. The relation of the HPA axis to aggression has been well established, especially through animal studies (Veenema, 2009). Also in humans, cortisol levels have been related to aggression repeatedly (Alink et al., 2012, Loney et al., 2006, Popma et al., 2007, Shirtcliff et al., 2005, van Bokhoven et al., 2004). The HPG axis involves signaling between hypothalamus, pituitary, and the gonadal glands, which produce estrogen and testosterone. Testosterone levels have been related to human aggression (Book et al., 2001, Brown et al., 2008, Chichinadze et al., 2010, Yu and Shi, 2009) and it has been hypothesized that especially the interplay between cortisol and sex steroids is important in determining aggression liability (Pavlov et al., 2012, Terburg et al., 2009).
Extensive reviews of aggression candidate gene studies have recently been published (Fernandez-Castillo and Cormand, 2016, Pavlov et al., 2012, Veroude et al., 2016, Waltes et al., 2015). Although a moderate number of studies has been conducted, a meta-analysis of individual candidate variants did not reveal any significant associations with aggressive behavior (Vassos et al., 2014). One reason for this may be the complex genetic background of aggression in most people. While a few monogenic aggression disorders caused by rare genetic variations with a high effect size exist (Brunner et al., 1993; Zhang-James et al., 2016), aggression in the population has a complex and polygenic genetic background, which can be aggravated by environmental factors (Veroude et al., 2016).
In the current study, we assessed the genetic mechanisms underlying aggression subtypes in the general population. Firstly, we aimed to verify the existence of three aggression subtypes in adult males and females from the general population based on the RPQ. Second, we aimed to assess the association of common genetic variants in the three biological systems with most evidence for a role in aggression, i.e. the serotonergic system, the dopaminergic system, and the neuroendocrine system with the different subtypes. We aimed to maximize power for finding genetic associations by (1) parsing phenotypic heterogeneity through differentiating between subtypes, (2) by assessing males and females separately, and (3) by combining genetic variants in a gene-set analysis (Bralten et al., 2011, Bralten et al., 2013, Naaijen et al., 2017).
Section snippets
Sample
The investigated sample consisted of participants of the Brain Imaging Genetics (BIG) study conducted at the Donders Institute for Brain, Cognition and Behaviour (Franke et al., 2010). The BIG study consists of self-reported healthy adults, who participated in smaller-scale imaging studies at the institute and gave consent to be included in the BIG study. Saliva samples for genetic testing were collected, and an internet-based test-battery of questionnaires was applied. The Reactive Proactive
Results
The general characteristics of our sample of 661 participants and the genotyped sample of n = 395 are shown in Table 2. The tree factor model of the RPQ, consisting of a proactive factor, a reactive factor due to internal frustration, and a reactive factor due to external provocation or threat, showed a good model fit in the healthy adults (RMSEA 90% CI: .041–.051, RMSEA: .046, CFI: .915, TLI: .905), Cronbach's alpha = 0.687 (proactive), 0.663 (reactive internal frustration), 0.684 (reactive
Discussion
In the current study, we investigated genetic mechanisms underlying aggression subtypes in the healthy population. Factor analysis confirmed that three correlated but separate dimensions of aggression can be distinguished in healthy adults, using the self-report scale RPQ (‘proactive aggression’, ‘reactive aggression due to internal frustration’, and ‘reactive aggression due to external provocation’). Aggregated analysis of common variants within monoaminergic and neuroendocrine systems
Role of funding source
The authors would like to acknowledge grants supporting their work from the Netherlands Organization for Scientific Research (NWO), i.e. the NWO Brain & Cognition Excellence Program (grant 433-09-229) and the Vici Innovation Program (grant 016–130-669 to BF). Additional support is received from the European Community's Seventh Framework Programme (FP7/2007 – 2013) under grant agreements n° 602805 (Aggressotype), n° 603016 (MATRICS), n° 602450 (IMAGEMEND), and n° 278948 (TACTICS), and from the
Contributors
Marjolein M.J. van Donkelaar, Martine Hoogman, Janita Bralten, Jan Buitelaar and Barbara Franke designed the study. Marjolein van Donkelaar managed literature searches, analyses and wrote the first draft of the manuscript. Elena Shumskaya was responsible for data management, data preparation and data quality control. Marjolein van Donkelaar, Janita Bralten and Martine Hoogman where involved in the statistical analysis. All authors contributed to and have approved the final manuscript.
Conflicts of interest
Barbara Franke discloses having received educational speaking fees from Merz and Shire. Jan K Buitelaar has been in the past 3 years a consultant to / member of advisory board of / and/or speaker for Janssen Cilag BV, Eli Lilly, Lundbeck, Shire, Roche, Novartis, and Servier. He is not an employee of any of these companies, and not a stock shareholder of any of these companies. He has no other financial or material support, including expert testimony, patents, royalties. None of the other
Acknowledgements
The authors would like to thank the Genetics of Multifactorial Diseases group of the Department of Human Genetics, Radboud university medical center, for their valuable contributions to genetic data collection, preparation and management.
References (83)
Differential behavioral effects of plasma tryptophan depletion and loading in aggressive and nonaggressive men
Neuropsychopharmacology
(2000)- et al.
The relationship between testosterone and aggression: a meta-analysis
Aggress. Violent Behav.
(2001) - et al.
Aggression in humans correlates with cerebrospinal fluid amine metabolites
Psychiatry Res.
(1979) - et al.
Salivary cortisol, dehydroepiandrosterone, and testosterone interrelationships in healthy young males: a pilot study with implications for studies of aggressive behavior
Psychiatry Res.
(2008) - et al.
Co-occurring anxiety and disruptive behavior disorders: the roles of anxious symptoms, reactive aggression, and shared risk processes
Clin. Psychol. Rev.
(2009) Androgen production in women
Fertil. Steril.
(2002)- et al.
MAOA, childhood maltreatment, and antisocial behavior: meta-analysis of a gene-environment interaction
Biol. Psychiatry
(2014) - et al.
Distinct characteristics of psychopathy relate to different subtypes of aggression
Pers. Individ. Differ.
(2009) - et al.
Ten-year review of rating scales. VI: scales assessing externalizing behaviors
J. Am. Acad. Child Adolesc. Psychiatry
(2003) - et al.
A longitudinal study of the effects of chronic maltreatment on children's behavioral and emotional problems
Child Abus. Negl.
(2004)