The Anatomy of Violence

by Adrian Raine

  One would think that today mothers fully understand and recognize that smoking is bad for their unborn child. Yet the unfortunate reality is that in the United States about a quarter of all pregnant mothers smoke, while in the United Kingdom a quarter of smokers who become pregnant continue to do so during pregnancy.96 Smoking remains a likely contributor to the violent offending in the offspring of these mothers.

  ALCOHOL CONSUMPTION DURING PREGNANCY

  In 1992, the double killer Robert Alton Harris was gassed to death in San Quentin prison, the first execution in California in twenty-five years. The terrible nature of his crime, in fact, likely contributed to the new wave of executions in the state. The murders occurred in 1978. Harris and his brother were looking to steal a getaway car for a planned bank holdup when they spotted a couple of teenagers in a green Ford eating Jack in the Box burgers. At gunpoint the boys were forced to drive to a wooded area near a lake under the promise that they would not be harmed. Once there, Harris shot and killed both boys. And it’s at this point that, for a jury listening to testimony, it becomes more psychologically gruesome.

  Just as Harris was about to execute the terrified boys, one of them—sixteen-year-old Michael Baker—pleaded for his life. According to a witness who shared a cell with Harris after his arrest, Harris boasted that he told the poor boy, “Quit crying and die like a man.” In those moments of impending doom, the petrified boy started to pray to God. Harris’s response? “God can’t help you now, boy; you’re going to die.”97 After the executions, and just as callously, Harris calmly finished off the rest of the murdered boys’ half-eaten hamburgers, and flicked pieces of the homicidal gore from the barrel of his gun.98 The heartlessness and clear lack of conscience, combined with the fact that Harris had just been released from prison for another murder, made him a clear candidate for the gas chamber.

  What is also true of Harris is that he was born with fetal alcohol syndrome. If smoking is problematic during pregnancy, you can imagine what the negative effects of consuming significant amounts of alcohol may be. Again, the model here is that alcohol consumed by a woman during pregnancy is a significant source of damage to the fetal brain, and this brain impairment predisposes her offspring to violence. There are four features of fetal alcohol syndrome as first established by the pediatrician Kenneth Jones in 1973:99 exposure to alcohol during pregnancy, craniofacial abnormalities, growth retardation, and central nervous system (CNS) dysfunction as evidenced by learning disabilities or low IQ. The craniofacial abnormalities in fetal alcohol syndrome sufferers can be striking. The middle part of the face is relatively flat, the upper lip is quite thin, and the eyes tend to be widely spaced. The uncanny result of this is that two unrelated babies in a hospital can look alike if they both have fetal alcohol syndrome. The rate of this syndrome is about 3 babies in every 1,000.100 More common, however, is the condition of “fetal alcohol effects”—in which just some of the symptoms described above are present—with a base rate of approximately 1 percent.

  Matching the striking nature of fetal alcohol syndrome is the relationship it bears to crime and delinquency. Perhaps the most comprehensive study conducted to date is that of Ann Streissguth and her colleagues at the University of Washington in Seattle.101 Although fetal alcohol syndrome is relatively rare, Streissguth was able to obtain an incredible 473 cases of either fetal alcohol syndrome or fetal alcohol effects from the Pacific Northwest, and assessed outcomes for antisocial behavior at age fourteen. A full 61 percent of the sample evidenced juvenile delinquency. Sixty percent were expelled or suspended from school. Forty-five percent showed some form of inappropriate sexual behavior, such as incest, sex with animals, or masturbation in public. More than half of the boys and 33 percent of the girls went on to be arrested or convicted for their offending.

  Streissguth’s work started off with fetal alcohol syndrome and looked at outcome for antisocial behavior. Another way one could look at it is to start off with a population of antisocial individuals and examine rates of fetal alcohol syndrome and fetal alcohol effects. This is exactly what Diane Fast and her colleagues did, and they found rates of 1 percent for fetal alcohol syndrome—more than three times the expected base rate—and a full 22 percent for fetal alcohol effects.102 There is little doubt that the mother’s intake of alcohol during pregnancy can raise the odds of problematic behavior.

  As with smoking during pregnancy one can argue that there is a third factor that underlies the fetal alcohol syndrome–antisocial relationship. Again, a genetically informative adoption study came to the fore in ruling this factor out. Remi Cadoret at the University of Iowa studied the adopted-away offspring of mothers who drank alcohol during pregnancy and found that they too showed higher rates of conduct disorder and adult antisocial behavior compared with adopted children whose biological mothers did not drink during pregnancy. Because the children were adopted away from their biological mothers after birth, their antisocial behavior cannot be attributed to the fact that mothers who drink may be poor caregivers during their child’s development. It does look as if the exposure to alcohol during pregnancy is causally related to crime outcome.

  Figure 6.2 Brain of a normal six-week-old baby (left) and brain of a baby with fetal alcohol syndrome

  The mechanism of action? The brain again has to be the number one suspect. Alcohol exposure ravages the brain during fetal development. The atrophy in brain tissue is striking and is widespread (Figure 6.2). Particularly affected is the corpus callosum, the band of white nerve fibers that connects the two hemispheres and allows for effective communication.103 Poorer executive functions are also an almost inevitable consequence of fetal alcohol syndrome.104 Experiments on animals have demonstrated that during the latter half of pregnancy, when the brain is rapidly developing, alcohol exposure results in a loss of neurons. It also affects glutamatergic neurotransmitter functioning, which in turn reduces hippocampal plasticity and the ability to learn.105 Just as we saw with prenatal smoking exposure, those born with fetal alcohol syndrome show widespread structural and functional impairments when given brain scans in late childhood.106

  Can pregnant women get away with just one alcoholic drink a week? It seems not. As with smoking, there is a dose-response relationship such that with increasing degrees of alcohol consumption, aggressive behavior and other externalizing behavior problems show a steady increase. A study of African-American mothers demonstrated that having just one alcoholic drink a week during pregnancy was enough to raise the odds of aggression and delinquency in the children.107 Indeed, this study documented that drinking any alcohol at all during pregnancy tripled the odds that the child would have clinically significant delinquency. At a causal level, animal studies yet again show dose-response relationships between increasing amounts of alcohol exposure and increasing degrees of structural brain impairments.108 It would be unwise for anyone who is pregnant to ignore the potential effects of consuming alcohol during pregnancy.

  So is there such a thing as a natural-born killer? If by this question we mean, “Is there an unalterable destiny for violence?,” the answer is no. But we have seen here that there are multiple health-related factors that occur right at birth and even before birth that are architects in shaping the landscape of violence. Birth complications, disruption to the developing brain of the fetus, exposure to smoking and alcohol, and testosterone exposure are significant elements in the genesis of violence. Yet these marks of Cain, while biologically based, are essentially environmental processes—not genetic. We get back to the crucial point that biological and social processes are inextricably mixed, and that a true appreciation of the biology of violence needs to take this mix into full consideration.

  What is certainly true is that in casting the lots that determine who will become an offender, for some the dice are loaded very early in life. Yet these very early health processes are just the beginning in a mixture of influences that can become toxic. We’ll see in the next chapter that health risk factors continue throughout de
velopment to create the deadly cocktail. As the example of Peter Sutcliffe shows us, there is more to homicide than birth complications—biologically based mental illness can be a critical factor in shaping a criminal career, as it did with him and with many others.

  7.

  A RECIPE FOR VIOLENCE

  Malnutrition, Metals, and Mental Health

  Amsterdam was a bad place to be in the winter of 1944–1945, especially if you were a pregnant mother. It was the beginning of the Dutch Hunger Winter. The Allied invasion of Normandy the previous June eventually gave relief to the Dutch, but in its immediate aftermath it brought misery. The Allies had been blocked at the Rhine and could not free much of the Netherlands from German occupation. In September the exiled Dutch government in London ordered railway workers in the Netherlands to go on strike in order to aid the Allies. They duly followed instructions, but the result was disastrous. German administrators retaliated with a food blockade, cutting off the western Netherlands from its food supplies.1

  It went from bad to worse. First, winter came very early that year and was unrepentantly harsh. Canals froze over. Food could not be transported. Retreating German troops destroyed bridges and docks, making transportation even more difficult. Second, much of the arable land had been ravaged by warfare and was barren—unable to provide sustenance for the Dutch citizenry. Two more painful blows to aching, empty stomachs.

  People began to starve. In November, city residents were rationed only 1,000 kilocalories of food a day. By February 1945 conditions deteriorated further with diets dropping to 580 kilocalories a piece.2 Ten thousand died of malnutrition, particularly in the cities, which were cut off from the countryside and food. Many thousands more are thought to have died of complications as a result of the famine.3 And for the remaining millions, life was wretched and depressing. Relief came only with liberation, in May 1945—ending a bitter eight months for the Dutch people.

  This seems a peculiar starting point for a window into antisocial personality, but the seeds of violence were being sown in that harsh winter, concealed in out-of-sight little victims—the unborn babies of starved, pregnant women. We know this because in 1963, when the male babies who were in utero during the famine turned eighteen, they underwent compulsory military service, and at that time they were subjected to a psychiatric examination that included formal assessment of antisocial personality disorder.4 Data collected from these examinations became the foundations of a unique epidemiological study on the effects of prenatal malnutrition on later behavior.

  In this breakthrough research study, Richard Neugebauer and colleagues from the New York State Psychiatric Institute conducted detailed analyses on these data. They divided the enormous sample of 100,543 men into those who were exposed to the famine—especially in the large cities in the west, including Amsterdam, Rotterdam, Leiden, Utrecht, and the Hague—and those in the north and south who were not exposed to the famine.5 The key result? Those exposed to the famine were two and a half times more likely to develop antisocial personality disorder in adulthood than those not exposed to the famine. The effects being especially true if the food shortage occurred during the first or second trimester of pregnancy. These findings were the first to demonstrate that poor nutrition during pregnancy predisposes to antisocial behavior in the offspring.

  This chapter on nutrition, toxins, and mental health is yet another that highlights the importance of the environment in causing the brain impairments that can contribute to crime. From the gut to the teeth to the hair and back to the brain, this particular close-up of the anatomy of violence shows that human and animal studies are building a persuasive picture of how a lack of iron, zinc, protein, riboflavin, and omega-3 in our diets may dump some of us into the violence trash bin. It’s a question of how both too little and too much food is bad for us. We’ll also see that these dietary deficiencies can be compounded by an overexposure to heavy metals in the environment, including lead and manganese. Finally, we’ll round off this physical-health perspective with a mental-health perspective, showing how major mental illness, with its base in biology, also contributes to violence.

  My own research into nutritional deficiencies and violence was inspired during a visit with Danny Pine, a brilliant and energetic researcher at Columbia University who had been working on heart rate and cognitive functioning in conduct-disordered children. We were walking to a meeting with Neugebauer, and Danny, with his sparkling glasses and wild beard, which had a life of its own, was talking a mile a minute, as he often does. “And Adrian, you just have to meet Richard. What a story from Holland—World War II, starvation, crime. It’s really something, you’re gonna love this.” And then he added with a mysterious twinkle and a wry smile, “Don’t forget to ask him about the tulip bulbs.”

  Tulips? What’s that all about? That song “When it’s spring again, I’ll bring again tulips from Amsterdam” flashed through my mind—but how does that fit into an academic meeting on violence? That was as much as Danny left me with before I met Richard Neugebauer and heard firsthand the astonishing story of the Dutch Winter Famine—and the tulip bulbs. Apparently, in the final months of the food blockade the starving Dutch began to eat tulip bulbs. These are toxic, and as we shall see later in this chapter, toxins have been associated with offending. Richard acknowledged that other issues remained unresolved. Only male adults had been studied. What about females? Could this malnutrition story apply to aggressive and antisocial behavior in children? And do social factors like poverty play a hidden role?

  These were the issues that percolated through my mind and ultimately stimulated us to look into nutrition in our Mauritius study. When our subjects were three years old, 1,559 of them came to the laboratory with their mothers to be examined by a pediatrician. We looked for five internal and external signs of malnutrition. First, they had their blood analyzed in a lab to assess hemoglobin levels. This gave us a handle on iron deficiency. Second, we had pediatricians conduct a physical examination of each child to look for four other external signs of malnutrition. Do you ever remember as a kid having cracks at the corners of your mouth? I seemed to have them on and off, and I’d poke them with my tongue when they felt hard and dry in order to soften them up. This is angular stomatitis, caused by a riboflavin deficiency, specifically a deficit in vitamin B2, but it can also reflect niacin deficiency.6

  Then the pediatrician would take a good look at the child’s hair. What color was it? In Mauritius, almost all of the children have black hair because they are of Indian, African, or Chinese extraction. But some kids had an orange tinge to their hair. It wasn’t some kind of funky look that their parents gave them to make them look cute and artsy—it was kwashiorkor. This is African dialect referring to “red hair.” It’s a sign of zinc, copper, and protein malnutrition that causes dyspigmentation of the hair—essentially a loss of the natural black color.7 The pediatrician also looked to see if the hair was sparse and thin, a sign of zinc, iron, and protein deficiency.8 Then, after these two careful looks, the pediatrician would grab a piece of hair and give it a tug. If it came out easily, it was a sign of protein energy malnutrition.9 There we have the five strands—all clinical indicators of malnutrition.

  At this point, Jianghong Liu, who at that time was a research fellow at the University of Southern California, entered the picture. She was the driving force behind the results I’ll discuss here. If a child had any one of these significant indicators, she assigned them to the malnourished group. Those who lacked malnutrition were the normal controls. She assessed the kids again at ages eight, eleven, and seventeen, the ages at which we had obtained teacher and parent ratings of their aggressive, antisocial, and hyperactive behavior. The results are shown in Figure 7.1. As you can see, at every single age the malnourished kids had higher scores on all dimensions of what we call “externalizing behavior”—aggression, delinquency, and hyperactivity.10

  Hold on a second. Aren’t kids with poor nutrition more likely to have parents with low levels of e
ducation and income? And aren’t low levels of education and income social risk factors for childhood behavior problems? Maybe poor nutrition itself makes no active contribution to aggression, but is linked to social deprivation, which causes aggression. Point taken. So Jianghong Liu controlled for poverty and twelve other social factors that could be driving the increase in aggressive behavior in the malnourished kids. The result? The malnutrition-aggression link was obstinate—it just would not budge. And it did not matter whether you were Creole or Indian, a boy or a girl. Poor nutrition does not respect race or gender when it comes to raising the risk of aggression. Furthermore, we also saw a dose-response relationship at age seventeen. If you look at Figure 7.2 you’ll see that the more signs of malnutrition the child had, the higher the score for conduct disorder. This result really reinforces the link between malnutrition and conduct disorder.

  Figure 7.1 Scores on externalizing behavior problems in malnourished and control groups across three time periods