The Extreme Warrior gene: a reality check

by Alondra Oubré

MAOA — Genetic culprit for violence?

Theories about inborn race-based aggression, violence, and criminality are back in the news [1]. In the ongoing search for genes underlying social behavior, none has sparked more curiosity, if not controversy, than the gene that codes for monoamine oxidase A — MAOA [2, 3, 4]. Nicknamed the “warrior gene,” a variant of the MAOA drew international attention nearly a decade ago when geneticist Rod Lea reported that it was more common in Maori — the indigenous Polynesians of New Zealand — than in whites [2]. According to one journalist, Lea suggested this gene might be the source of poor health and increased rates of violent crime in Maori [5]. The media frenzy over “bad genes causing bad behavior” didn’t stop there. A rare, seemingly even more detrimental version — the “extreme warrior gene” — has since stirred debate because it occurs more frequently in African Americans than in whites [6, 7].

MAOA — an enzyme that degrades neurotransmitters such as serotonin and dopamine in the brain — is coded for by the MAOA gene [8, 9, 10]. Neurotransmitters play a pivotal role in mood, arousal, and emotions, even affecting impulse control. Since the 1990s scientists have identified several versions of the MAOA, which are usually categorized as low-activity or high-activity variants. MAOA genes are classified based on how many times a short sequence — a functional strip of DNA — repeats itself within a variable region of the gene [8]. The most common variant, MAOA-4R, has four repeats and is associated with high-activity breakdown of neurotransmitters [8]. Alternate forms of the MAOA, including the 2-repeat (2R) and 3-repeat (3R) versions, contain fewer repeat sequences.

The 2R and 3R variants are often lumped together in studies of the low-activity MAOA gene. (Although the 5R version has a large number of repeats, it too is less active than the 4R version.) The two classes of MAOA versions correlate with different behavioral tendencies. Low-activity variants are thought to lead to reduced levels of MAOA in the brain, possibly shifting mood by changing serotonin levels [8].

Over the past 12 years research on MAOA genes has examined how low-activity gene variants interact with environmental factors to influence violence and other antisocial behaviors [11]. In 2002, Avshalom Caspi, then at King’s College in London, and his colleagues published their landmark study [12]. Caspi’s team reported that adults with the low-expression MAOA who were mistreated as children were more prone to developing antisocial problems later in life. But maltreated children with the high-activity variant were less likely to engage in delinquent or criminal activities. It seems low-activity MAOA variants make people more responsive to abuse [12]. Up to this point, all of the studies on the MAOA gene had been conducted in Caucasians.

That changed when researchers started investigating this gene in the Maori of New Zealand. Historically, warfare was a central part of traditional Maori culture because, after all, these South Pacific islanders had to compete vigorously for limited natural resources. Today, some Maori are integrated into New Zealand society. Yet, overall they still lag behind other ethnic groups in their country in income, education, and health, and crime rates are higher. For many experts, this ethnic gap is the result of numerous environmental causes, including poverty [13]. In 2006, Lea reported that MAOA-3R — one of the low-activity risky variants — was more common in Maori males than in white males [2]. According to Lea, the 3R version was associated with a lineup of undesirable personality traits: risk-taking, violence, aggression, gambling, addiction and criminal behavior. Suddenly, it seemed genetics could possibly explain the Maori/white ethnic divide in achievement and social outcomes [2].

MAOA-3R — the “original warrior gene” — was the first gene linked with antisocial characteristics. But Maori were not the only ethnic group with a high frequency of this variant. It turned out that while 3R was found in 56% of Maori males, it occurred in 58% of African American males and 34% of European males [2]. Misinterpreted by the media, the 3R variant quickly became a lead character in a pop science narrative intended to explain why certain racial groups appear to have increased tendencies toward violence. When a disproportionately high number of males of an ethnic group carries a less common gene linked with aggressive behaviors, the discussion about that gene immediately takes on racial overtones [3, 14]. (Interestingly, the press ignored studies indicating that the 3R variant occurred in 61% of Taiwanese males [15] and 56% of Chinese males [16]).

Research on the MAOA gene 

Over the last few years, multiple studies have replicated the original findings of Caspi’s team. The evidence as a whole continues to show that the interaction between low-activity MAOA variants and early exposure to abuse increases the risk for antisocial behavior in men throughout their lifetime [11, 17]. Offending, conduct problems, and hostility have been observed in males carrying low-expression versions of the MAOA gene [6, 7, 18].

Kevin Beaver of Florida State University is a researcher in biosocial criminology — a field that explores the role of both genes and environment in criminal and other antisocial behaviors. One of Beaver’s studies has linked low-activity MAOA variants with increased likelihood of males joining a gang and using a weapon in a fight [18]. Most of the early investigations comparing low- and high-expression MAOA genes probed only the moderately risky 3R version. A few looked at a combination of 3R and 2R. However, the effects of these two variants on social behaviors were not teased apart in most of the initial studies [6, 7, 8].

In 2008, University of North Carolina sociologist Guang Guo and his colleagues found that antisocial behaviors in male youth were associated with three genes — low-activity MAOA variants and two dopamine-related genes [19]. But it was 2R — the “extreme warrior gene” — that captivated researchers searching for a still illusive genetic basis of criminal predispositions. Guo’s team analyzed data on male youth from Add Health — a national sample of adolescents in grades 7-12. Their findings showed that the rare variant, 2R, was correlated with higher levels of self-reported serious and violent delinquency. The association was also observed in females, but it was too weak to merit further study [19].

More recently, Beaver’s team has focused only on the 2R variant rather than the low-expression variants combined [6, 7]. He and his colleagues have discovered that African American males carrying 2R were more likely to be involved in extreme violence — shooting and stabbing — than African American men with other MAOA variants [6]. The relationship between the rare MAOA version and antisocial behaviors has raised eyebrows because, quite simply, this gene is not distributed equally across ethnic groups. In the Add Health database, 5.5% of African American men, 0.9% of Caucasian men, and 0.00067% of Asian men have 2R. (No information is currently available on the frequency of 2R in males of African black descent outside the United States.) Since the rare MAOA variant is virtually non-existent in whites, all of the males in Beaver’s study were black Americans [6].

Beaver’s sample of 133 African American men from the Add Health database included 6% that carried 2R. Overall, 5.6% of the men in the sample reported shooting or stabbing someone at some point in their lifetime. The association between 2R and committing a shooting or stabbing crime was statistically significant. Based on Beaver’s evidence, 2R appears to increase the risk of shooting or stabbing a victim during adolescence or adulthood [6]. For some commentators in the public arena, MAOA-2R has become a symbol of a new era in behavioral genetics research — an era that has reintroduced race into the nature versus nurture debate over the source of ethnic behavioral differences [1].

In a recent interview I asked Kevin Beaver if he had found any correlation between males in his study who carried 2R and socioeconomic status — SES. After all, a sample of African American young men is likely to disproportionately come from lower SES backgrounds. Beaver noted that the Add Health survey had deliberately over-sampled African Americans from the middle and upper middle classes to compensate for this imbalance. “No one knows how the over-sampling — the relatively larger number of middle to upper income participants — translates into the frequencies of MAOA-2R in the sample,” Beaver said. “The small number of 2R subjects, however, makes it difficult to examine the link between SES and the 2R variant” [20].

Beaver’s studies have shown that the 2R variant has a robust association with violent behaviors, arrest, and incarceration [6, 7]. His research is applauded by supporters of behavioral genetics, but it has also drawn criticism. It focuses on an antisocial-linked gene that reportedly occurs more frequently in African American men than in males of other ethnic groups. This has led some popular writers to speculate that MAOA-2R might account for — or at least play a significant role in — the relatively higher rates of violent crime in African Americans. Not everyone agrees [21].

Part of the skepticism surrounding Beaver’s studies may lie in popular misinterpretations of his research. As Beaver explains, “It is probably correct to assume that social behaviors are due to gene-environment interaction. But statistical models are quantifying variance — that is, they are looking at differences between persons. Why an individual turns out a certain way might be due to gene-environment interaction. But person-to-person differences do not always result from gene-environment interaction. The reason that people vary in criminal propensities could be due to only genetics, only environments, or either of these free from gene-environment interaction.”

Beaver’s findings may shed light on whether a single gene might underlie individual differences in criminal tendencies. So far, his investigations have targeted only African American males because too few whites carry the rare MAOA variant to include them. The rates of 2R are more than five times higher in African American males than in American white males, at least in the Add Health sample [6]. Beaver claims that 2R alone may be strong enough to account for a significant amount of violent behavior in African American men. But he doesn’t think this rare gene version explains all of the variation between men who have and don’t have severe antisocial traits. As he puts it, “Even if MAOA-2R is causally linked with antisocial behaviors, it is not common enough in African Americans to solely account for crime rates in blacks” [20].

Like many other genetic studies in criminology, Beaver’s research on MAOA-2R explores the heritability of specific antisocial behaviors — in this case, shooting and stabbing. Heritability — not to be confused with heredity — refers to the proportion of variance in a trait within a population due to genetic variation [22]. A heritability estimate does not pertain to the amount of genetic influence on a particular trait in a particular person. Each estimate is valid only for a single population at a specific point in time. Heritability estimates can change, depending upon the strength or weakness of environmental factors, which along with various genes, shape social behaviors [22].

Although genes affect individual differences in behavior, the effect of each individual gene is usually small. The genetic underpinnings of a specific social behavior typically involve multiple genes that have a cumulative influence [8]. It is not clear if MAOA-2R is an exception. The more common low-activity variant, 3R, interacts with adverse social effects such as childhood maltreatment [11, 12]. But other possible environmental factors, which conceivably could interact with the 2R, may not have been explored in-depth as yet. One such environmental influence that has recently received attention is parents and caregivers’ punitive discipline — spanking and yelling — of a young child [23]. Punitive practices are not necessarily abuse. But in families that traditionally use harsh discipline with their children, corporeal punishment or even loud verbal chastising can at times turn into maltreatment.

Daniel Choe, a developmental psychologist, and his colleagues at the University of Pittsburgh investigated the effects of punitive discipline on antisocial behavior in young white and African American men [23]. The researchers examined 189 young, low-income white and African American males with both low- and high-expression MAOA genes. As the researchers predicted, punitive discipline was associated with increased antisocial behavior only in men with the low-activity 3R variant. This pattern held for both white and black males. There was no relationship between harsh punishment and antisocial behavior in men carrying 4R, the high-activity version of MAOA [23].

Importantly, the effects on behavior depended on the age at which the children were punished [23]. Kids who had been disciplined at 1.5, 2, and 5 years were more likely to develop antisocial behaviors when they were older — between 15 and 20 years old. Specific antisocial behaviors, including violent attitudes and juvenile arrests, were more likely to occur at a specific age and to be linked with the age when the boys were abused [23].

Choe’s study is the first to demonstrate that ethnic minority children— African Americans, not just Caucasians — with a low-expression MAOA gene variant who face harsh discipline have an increased risk for antisocial behavior [23]. Choe’s team published the effects of just the 3R variant, excluding five African American participants in their study carrying the 2R version. Curious about possibly different effects of 2R, they then reanalyzed the data to include the five black males with 2R. The findings remained the same. Combining the boys with 2R — the highest risk variant — and those with the less severe risky 3R did not change the differences the researchers found between the 3R and 4R variants. The five males with 2R comprised a very small sample, but the fact that both low-activity MAOA variants, 2R and 3R, interacted with an environmental factor — punitive discipline — at specific ages, or developmental milestones, is noteworthy. It suggests the effects of MAOA-2R on antisocial behaviors are partly mediated by non-genetic factors [24].

Choe stresses that genetic influences on social behaviors such as juvenile delinquency cannot be fully understood outside the context of social circumstances. He is referring not only to parenting styles, but also to the in utero environment of the unborn fetus. As he explains, compared to the white youth in his study, the African Americans were more likely to grow up in poorer, urban, dangerous neighborhoods. A high percentage of these youth are being raised by single mothers, and they grow up without the attention found in most middle-class homes. Choe acknowledges the role of genes in behavior, but he clearly thinks that environmental factors contribute substantially to ethnic differences in antisocial behaviors. As he points out, the white kids in the sample were also poor, but they lived in low-income suburban communities, not in densely concentrated inner cities. The suburbs pose less of a risk than urban communities for group delinquent behavior [24].

Most experts agree that social behaviors stem from complex interactions between genes and environment [11]. Does MAOA-R2 go against the grain? Is it unaffected or only minimally affected by social experience and other elements of the environment? According to Beaver, MAOA-2R might act independently of environmental influence, but its effects might be masked by MAOA-3R. If the 2R version raises the risk of criminal behavior regardless of environmental influence, then perhaps it is indeed the source of a strong genetic propensity toward violence. If so, then violent tendencies associated with 2R — the “extreme warrior gene” — are not likely to be easily curtailed.

Yet many scientists think that behavioral traits are determined not only by the interplay between genes and environment. Antisocial behaviors also may be molded by the interaction of multiple genes — not simply a single gene [10, 19, 25]. When asked if he plans to examine the effects of 2R combined with genes other than MAOA, Beaver said he does not. As he explains, “the frequency of the 2R variant is too low to analyze. In the future we’ll need extremely large samples to have enough males with 2R to study.”

Epigenetics and MAOA in the brain 

Epigenetics is revolutionizing how scientists think about genetics. Epigenetics refers to external changes to DNA that turn genes “on” or “off” without altering the DNA sequence [26]. Gene expression — the manifestation of genetic potential — is modified in epigenetic processes, even though the gene itself stays intact. The field of epigenetics is largely theoretical, at least insofar as humans are concerned. But growing evidence suggests that epigenetic changes can, in some cases, be passed on from parents to children. They are handed down not as inherited traits, but as non-hereditary modifications transmitted to offspring along with genes from their parents [26].

Various environmental factors are thought to influence epigenetic processes. Could epigenetics modify behavioral traits by acting on MAOA gene activity? Scientists are just beginning to understand the effects of MAOA variants on the brain. The low-expression MAOA-3R variant has been linked with a heightened response from the amygdala, a structure in the brain that regulates emotion [27]. 3R is also associated with decreased activity in prefrontal regions of the brain that protect against anxiety [27].

Elena Shumay of the Brookhaven National Laboratory and her team conducted a study to determine how MAOA variants affect brain levels of the MAOA enzyme in healthy men [28]. Using PET imaging scans, these researchers found no correlation between MAOA brain levels and MAOA gene variants. Shumay and her colleagues reasoned that MAOA levels must be regulated by the same region of the MAOA gene where the 2R, 3R, 4R, or other repeat sequence are located. The evidence supported their prediction: it appears that MAOA expression associated with MAOA brain levels is under the control of epigenetic mechanisms [28].

In other words, epigenetics may influence whether a tendency toward higher or lower MAOA genetic activity actually manifests itself. The amount of genetic activity, in turn, determines whether there is a larger or smaller quantity of the MAOA enzyme in the brain, which is needed to break down certain neurotransmitters [28]. The findings of Shumay’s team are preliminary, however. Their data do not prove that antisocial behaviors are not influenced by the low-activity 2R and 3R variants of the MAOA [8]. Nonetheless, their results suggest that MAOA brain levels, which affect mood, are at least partially regulated by non-genetic factors — i.e., epigenetically.

Genes, environment and plasticity

There are limits to studying the role of a single gene in antisocial behavior outside of its environmental context. Even when a gene correlates closely with violence or criminal acts, it does not mean that the gene itself codes for aggressive tendencies. According to Kevin Beaver and University of California at Davis’ Jay Belsky, plasticity genes seem to affect how much or how little male youth are influenced by their parents. Beaver and Belsky claim these genes appear to increase susceptibility to environmental effects, “for better and for worse” [29]. Supportive and unsupportive parents are more likely to have a positive or negative impact, respectively, on their children if their kids carry plasticity genes [29].

Yet plasticity genes appear to have a cumulative effect. Determining the influence of each separate gene on a behavior can be difficult. The combined genetic effects may vary, depending upon the individual. For a gene to have a plasticity effect on a behavior, it has to interact with an environmental factor [29, 30]. Are we then back to the notion that gene-environment interactions ultimately determine social behaviors?

MAOA is one of several candidate genes for plasticity that appear to mediate a person’s susceptibility to his or her environment [8]. Complex interactions between genes — and between genes and environmental factors — may explain why males with multiple plasticity genes are at heightened risk for developing aggressive behaviors if, at a young age, they have traumatic experiences with their caregivers. MAOA variants are not necessarily directly associated with brain changes that could lead to violence. But two or three plasticity genes working in tandem might increase a young male’s risk of sensitivity to early terrifying encounters with parental figures [29]. As Choe’s findings demonstrate, the timing of stressful life events may influence whether or not a genetic proclivity for antisocial behavior manifests itself [23].

MAOA research — The future 

In matters as sensitive as race, genes, and behaviors — especially antisocial behaviors reported in African American males — the conventional wisdom is to balance the search for behavior-linked genes with a probe of environmental influences. Many experts doubt that violent behaviors are conditioned exclusively by genetics without any influence from social circumstances [11, 31]. Unless scientists have ruled out all the subtle and nuanced (or even plainly obvious) adverse social and ecological factors that can affect gene expression, they might miss profound interactions between MAOA-2R and the environment. As Choe and his colleagues point out, “multiple genes of small effects are likely to interact with multiple environments to lead to many outcomes” [23].

The recent work of both Kevin Beaver and Daniel Choe highlights just how complicated research on behavior-linked genes — particularly MAOA-2R — can be. The findings of a study may depend partly on whether scientists are looking for genetic effects, environmental effects, various combinations of gene-environment interactions, or genetic variance (heritability) between individuals — not to mention epigenetic complications. If researchers are focusing only on a genetic influence on adversity, they might miss environmental contributions. Conversely, by honing in on a single gene, investigators might discover a genetic trait that helps to differentiate males who do and don’t develop antisocial behaviors. With improved understanding of how violence-linked genes are expressed, it may someday be feasible to develop safe, noninvasive, and ethical psychosocial interventions to reduce offending and potential crime in males carrying high-risk genes linked with antisocial proclivities.

The jury is still out on whether 2R, the rare MAOA gene, acts independently of the environment (and independently of other genes) to shape antisocial personality traits.  While experts continue to unravel complex interactions between genes, epigenetics, and environment, it may be best for scientists and society alike to take a prudent position on this issue. We forsake our scientific heritage if, at this point in time, we leap to conclusions about what MAOA-2R means — or doesn’t mean — for antisocial tendencies in males of any ethnic or racial group.

_____

My thanks to Kevin Beaver and Daniel Choe for their input while writing this article.

Alondra Oubré is a science and medical writer who works primarily for the medical device, pharmaceutical and biotechnology industries. She holds a doctorate in medical anthropology, and is the author of various publications on human biodiversity, the ethnic achievement divide, health disparities, and plant drug research. She has published a two-volume collection entitled Race, Genes and Ability: Rethinking Ethnic Differences.

[1] Wade N. A Troublesome Inheritance: Genes, Race and Human History. New York: Penguin Press. 2014.

[2] Lea R, Chambers G. Monoamine oxidase, addiction, and the “warrior” gene hypothesis. N Z Med J. 2007. 120 (1250) PMID: 17339897.

[3] Merriman T, Cameron V. Risk-taking: behind the warrior gene story. N Z Med J. 2007 Mar 2;120(1250):U2440.

[4] Perbal L. The ‘warrior gene’ and the Mãori people: the responsibility of the geneticists. Bioethics. 2013 Sep;27(7):382-7. doi:10.1111/j.1467-8519.2012.01970.x.

[5] Stokes J. Scientist defends ‘warrior’  gene. The New Zealand Herald. Mar 5, 2007.

[6] Beaver K, Barnes J, Boutwell B. The 2-repeat allele of the MAOA gene confers an increased risk for shooting and stabbing behaviors. Psychiatr Q. 2013a. Dec 11.

[7] Beaver K, Wright, J, Boutwell B, Barnes J, DeLisi M, Vaughn M. Exploring the association between the 2-repeat allele of the MAOA gene promoter polymorphism and psychopathic personality traits, arrests, incarceration, and lifetime antisocial behavior. Personality and Individual Differences. 2013b. 54(2):164-168.

[8] Buckholtz JW, Meyer-Lindenberg A. MAOA and the neurogenetic architecture of human aggression. Trends Neurosci. 2008 Mar;31(3):120-9. doi: 10.1016/j.tins.2007.12.006.

[9] Dorfman H, Meyer-Lindenberg A, Buckholtz JW. Neurobiological mechanisms for impulsive-aggression: The role of MAOA. Curr Top Behav Neurosci. 2014 Jan 28.

[10] Pavlov KA, Chistiakov DA, Chekhonin VP. Genetic determinants of aggression and impulsivity in humans. J Appl Genet. 2012 Feb;53(1):61-82. doi:10.1007/s13353-011-0069-6.

[11] Byrd AL, Manuck SB. MAOA, childhood maltreatment, and antisocial behavior: meta-analysis of a gene-environment interaction. Biol Psychiatry. 2014 Jan 1;75(1):9-17. doi:10.1016/j.biopsych.2013.05.004.

[12] Caspi A, McClay J, Moffitt TE, Mill J, Martin J, Craig IW, Taylor A, Poulton R. Role of genotype in the cycle of violence in maltreated children. Science. 2002. Aug 2;297(5582):851-4.

[13] O’Sullivan J, Dana T. Redefining Maori economic development. International Journal of Social Economics. 2008. 35(5):364-379.

[14] Crampton P, Parkin C. Warrior genes and risk-taking science. N Z Med J. 2007. Mar 2;120(1250):U2439.

[15] Lu RB, Lee JF, Ko HC, Lin WW, Chen K, & Shih JC. No association of the MAOA gene with alcoholism among Han Chinese males in Taiwan. Progress in Neuro-psychopharmacology & Biological Psychiatry. 2002. 26 (3), 457-61 PMID: 11999895

[16] Lung FW, Tzeng DS, Huang MF, Lee MB. Association of the MAOA promoter uVNTR polymorphism with suicide attempts in patients with major depressive disorder. BMC Med Genet. 2011. 24;12:74. doi:10.1186/1471-2350-12-74.

[17] Fergusson DM, Boden JM, Horwood LJ, Miller AL, Kennedy MA. MAOA, abuse exposure and antisocial behaviour: 30-year longitudinal study. Br J Psychiatry. 2011. 198(6):457-63. doi:10.1192/bjp.bp.110.086991.

[18] Beaver K, DeLisi M, Vaughn M, Barnes JC. Monoamine oxidase A genotype is associated with gang membership and weapon use. Compr Psychiatry. 2010. 51(2):130-4. doi:10.1016/j.comppsych.2009.03.010.

[19] Guo G, Ou X, Roettger M, Shih J. The VNTR 2-repeat in MAOA and delinquent behavior in adolescence and young adulthood: Associations and MAOA promoter activity. European Journal of Human Genetics. 2008. 16(5):626-634

[20] Beaver K. Personal communication. May 2014.

[21] Allen A. Charging Into the Minefield of Genes and Racial Difference. Nicholas Wade’s ‘A Troublesome Inheritance.’ Book Review. New York Times. May 15, 2014.

[22] Visscher PM, Hill WG, Wray NR. Heritability in the genomics era–concepts and misconceptions. Nat Rev Genet. 2008 Apr;9(4):255-66. doi:10.1038/nrg2322.

[23] Choe D, Shaw D, Hyde L, Forbes E. Interactions between monoamine oxidase A and punitive discipline in African American and Caucasian men’s antisocial behavior. Clinical Psychological Science. 2014. March 14. doi:10.1177/2167702613518046

[24] Choe D. Personal communication. May 2014.

[25] Simons RL, Beach SR, Barr AB. Differential susceptibility to context: A promising model of the interplay of genes and the social environment. Adv Group Process. 2012;29. doi:10.1108/S0882-6145(2012)0000029008.

[26] Rothstein MA, Cai Y, Marchant GE. The ghost in our genes: legal and ethical implications of epigenetics. Health Matrix Clevel. 2009 Winter;19(1):1-62.

[27] Zhong S, Israel S, Hong X, Ebstein R, Chew S. Monoamine oxidase A gene (MAOA) associated with attitude towards longshot risks. PLoS ONE. 2009. 4(12):e8516 doi:10.1371/journal.pone.0008516

[28] Shumay E, Logan J, Volkow ND, Fowler JS. Evidence that the methylation state of the monoamine oxidase A (MAOA) gene predicts brain activity of MAO A enzyme in healthy men. Epigenetics. 2012 Oct;7(10):1151-60. doi:10.4161/epi.21976.

[29] Belsky J, Beaver KM. Cumulative-genetic plasticity, parenting and adolescent self-regulation. J Child Psychol Psychiatry. 2011 May;52(5):619-26. doi:10.1111/j.1469-7610.2010.02327.x.

[30] Belsky J, Jonassaint C, Pluess M, Stanton M, Brummett B, Williams R. Vulnerability genes or plasticity genes? Mol Psychiatry. 2009 Aug;14(8):746-54. doi:10.1038/mp.2009.44.

[31] Weder N, Yang BZ, Douglas-Palumberi H, Massey J, Krystal JH, Gelernter J, Kaufman J. MAOA genotype, maltreatment, and aggressive behavior: the changing impact of genotype at varying levels of trauma. Biol Psychiatry. 2009 Mar 1;65(5):417-24. doi:10.1016/j.biopsych.2008.09.013.