My Age of Anxiety: Fear, Hope, Dread, and the Search for Peace of Mind

by Scott Stossel

  Those infants or toddlers identified as inhibited are more likely to become shy, nervous adolescents—and then shy, nervous adults—than their peers. They are much more likely to develop clinical anxiety or depression as teens or adults than their less physiologically reactive peers. Even those high-reactive babies who don’t grow up to get officially diagnosed with anxiety disorders tend to remain more nervous, on average, than their peers.

  In believing that temperament is innate and largely fixed from birth, Kagan falls squarely in the intellectual tradition that stretches back to Hippocrates, who in the fourth century B.C. argued that personality and mental health derived from the relative balance of the four humors in the body: blood, phlegm, black bile, and yellow bile. According to Hippocrates, as noted in chapter 1, a person’s relative humoral balance accounted for his temperament: whereas someone with relatively more blood might have a lively or “sanguine” temperament and be given to hot-blooded explosions of temper, someone with relatively more black bile might have a melancholic temperament. Hippocrates’s theory of humoral balance directly anticipates, metaphorically anyway, the serotonin hypothesis of depression and other modern theories of the relation between chemical imbalances in the brain and mental health. For instance, Hippocrates attributed what we would by the mid-twentieth century be calling neurotic depression—and what we would today call generalized anxiety disorder (DSM code 300.02)—to an excess of black bile (melain chole). As Hippocrates described it, this condition was characterized by both physical symptoms (“pains at the abdomen, breathlessness … frequent burps”) and emotional ones (“anxiety, restlessness, dread … fear, sadness, fretfulness” often accompanied and usually caused by “meditations and worries exaggerated in fancy”).

  Hippocrates may have had the wrong explanatory metaphor, but modern science is proving him to have been basically right about temperament’s fixity and biological basis. Kagan, now in his eighties, is semiretired, but four major longitudinal studies started by him or by a former protégé, Nathan Fox at the University of Maryland, are still under way. All four are reaching conclusions that support Kagan’s long-held theory that the anxious temperament is an innate, genetically determined phenomenon that characterizes a relatively fixed percentage of the population.† His studies have repeatedly found that those 15 to 20 percent of infants who react strongly to strangers or novel situations are much more likely to grow up to develop anxiety disorders than their less physiologically reactive peers. If you are born highly reactive and inhibited, you tend to stay highly reactive and inhibited. In decades of longitudinal studies, only rarely has Kagan seen anyone move from one temperamental category to another.

  All of which would seem to complicate, if not undermine, what I said about attachment theory in the previous chapter. Indeed, Kagan believes John Bowlby, Mary Ainsworth, and their colleagues were largely wrong about how anxiety gets transmitted from one generation to the next. In Kagan’s view, insecure attachment style per se does not produce an anxious child. Rather—and I’m oversimplifying a little—genes produce a mother with an anxious temperament; that temperament, in turn, leads her to demonstrate an attachment style that psychologists observe to be insecure. The mother then transmits this anxiety to her children—not primarily, as Bowlby and Ainsworth would have it, through her nervous parenting style (though, to be sure, that may intensify the transmission), but rather by the passing on of her anxious genes. Which, if true, would make it harder to break the transmission of anxiety from generation to generation through changes in parenting behavior—and which might explain why, despite our best efforts, my wife and I have been unable to prevent our children from developing signs of incipient anxiety disorders.

  John Bowlby cited animal studies to buttress his theory of attachment. But Jerome Kagan can also cite animal studies to rebut Bowlby and support his own theory of temperament. In the 1960s, researchers at the Maudsley Hospital in London bred what became known as the Maudsley strain of reactive rats, which responded to stress with pronounced anxious behavior. These breeding experiments were performed without the benefit of modern genomics. Researchers simply observed rat behavior, noted the “emotionally reactive” ones (mainly by measuring their defecation rates when placed in open spaces), and mated them with one another—and in so doing managed to produce this highly anxious strain. (By the same selective breeding technique, they also produced a strain of nonreactive rats, which responded less fearfully than average to open spaces and other stressors.) This seemed to be evidence of a potent hereditary component to anxiety in rat populations.

  Modern experimental techniques have advanced beyond selective breeding. Scientists are now able to chemically switch different mouse genes on or off, allowing researchers to observe how the genes affect behavior. By deactivating certain genes, researchers have created mice that, for instance, no longer experience anxiety, and in fact cannot recognize real danger, because their amygdalae have stopped working. Researchers, churning out hundreds of studies of this sort a year, have so far identified at least seventeen genes that seem to affect various parts of the fear neurocircuitry in mice.

  For instance, Eric Kandel, a Nobel Prize–winning neuroscientist at Columbia, has discovered one gene (known as Grp) that seems to encode a mouse’s ability to acquire new phobias through fear conditioning and another gene (known as stathmin) that regulates innate levels of physiological anxiety. Mice whose Grp gene has been switched off cannot learn to associate a neutral tone with an electric shock the way normal mice do. Mice whose stathmin gene has been switched off become daredevils: instead of instinctively cowering on the edge of open spaces like normal mice, they venture boldly into exposed areas.

  Evolution has conserved many genes, so humans and rodents share many of the same ones. Consider RGS2, a gene that in both mice and humans seems to regulate the expression of a protein that modulates serotonin and norepinephrine receptors in the brain. After it was noted that mice without the RGS2 gene displayed markedly anxious behavior and “elevated sympathetic tone” (meaning their bodies were on constant low-grade fight-or-flight alert), a series of studies on humans by Jordan Smoller and his team at Harvard Medical School found a relation between certain variations of the RGS2 gene and human shyness. In one study of children from 119 families, the kids who displayed characteristics of a “behaviorally inhibited” temperament tended to have the same variant of the RGS2 gene. Another study, of 744 college students, found that students with the “shy” variant of the gene were more likely to describe themselves as introverted. A third study revealed how the gene exerts its effect on the brain: when fifty-five young adults were placed in a brain scanner and shown pictures of angry or fearful faces, those with the relevant RGS2 variation were more likely to show increased “neuronal firing” in the amygdala and the insula, a part of the cortex associated not only with limbic system expressions of fear but also with “interoceptive awareness,” that explicit consciousness of inner bodily functions that can give rise to “anxiety sensitivity.” A fourth study, of 607 people who lived through the severe Florida hurricane season of 2004, found that those who had the relevant variant of the RGS2 gene were more likely to have developed an anxiety disorder in the aftermath of the hurricanes.

  None of these studies prove that simply having a certain variant of the RGS2 gene causes anxiety disorder. But they do suggest that the RGS2 gene affects the functioning of fear systems in the insula and the amygdala—and that individuals who have the “shy” variant of the gene are more likely to have hyperactive amygdalae and to experience higher levels of autonomic arousal in social situations and therefore to be shy or introverted. (Shyness and introversion are both predisposing factors in the anxiety disorders.)

  Lauren McGrath, a researcher in the Psychiatric and Neurodevel-opmental Genetics Unit at Massachusetts General Hospital, studied 134 babies over nearly twenty years. When the babies were four months old, McGrath’s team divided them into groups of (using Kagan’s terms) “high-reactives” a
nd “low-reactives.” At four months, the high-reactives cried and moved more in response to the movement of a mobile than the low-reactives did; at fourteen and twenty-one months, those same high-reactives still tended to demonstrate fearful reactions in response to novel situations. Eighteen years later, McGrath’s team tracked down the original study subjects and looked at the structure and reactivity of their amygdalae. Sure enough, the babies identified as high-reactives at four months had larger, more hyperactive amygdalae at age eighteen than the low-reactives did—yet more evidence that the amygdala’s response to novelty is a strong predictor of temperamental anxiety level. In a final wrinkle, McGrath’s team, using new genetic coding techniques, discovered that high amygdala reactivity at age eighteen was highly correlated with a particular variation on a specific gene known as RTN4. McGrath and her colleagues hypothesize that the RTN4 gene helps determine how hyperreactive your amygdala will be, which in turn helps determine whether your temperament will be high-reactive or low-reactive—which in turn helps determine your vulnerability to clinical anxiety.

  This alphabet soup of genetic studies—hundreds if not thousands of which are being conducted at any given time—can seem inanely reductionist. A few years ago, I read a New York Times article about studies attributing a correlation between certain variants of two human genes—AVPR1a and SLC6A4—and a “talent for creative dance performance.”‡ The good news, I suppose, is that if this changes how we think about character and fate, it might also change how we think about courage and cowardice, shame and disease, stigma and mental illness. If extreme anxiety is owed to genetic anomalies, should it be any more shameful than multiple sclerosis or cystic fibrosis or black hair, all diseases or traits encoded by the genes?

  Fifty years ago, we could plausibly blame the behavior of our mothers for all manner of neuroses and unhappiness and bad behavior. Today, we can perhaps still blame our mothers—but it may be more plausible to blame the genes they conferred upon us than the behaviors they displayed or the emotional wounds they inflicted.

  For that which is but a flea-biting to one, causeth insufferable torment to another.

  —ROBERT BURTON, The Anatomy of Melancholy (1621)

  A number of private companies will, in exchange for a drop of your saliva and a hefty fee, sequence part of your genome in order to provide information about your relative risk factor for various diseases. A few years ago, I paid a few hundred dollars to a company called 23andMe, and so I now know that my genes have left me with, all things being equal, a modestly higher than average likelihood of getting gallstones, a modestly lower than average likelihood of developing type 2 diabetes or skin cancer, and a roughly average likelihood of suffering a heart attack or developing prostate cancer. I also learned that I am, according to my genotype, a “fast caffeine metabolizer,” that I am at “typical” risk for heroin addiction and alcohol abuse, and that I have fast-twitch sprinter’s muscles. (I also learned that I have “wet” earwax.)

  I was hoping to find out which variants I have of two particular genes, each of which has at different times been dubbed the “Woody Allen gene.” The first gene, known as COMT, is found on chromosome 22 and encodes the production of an enzyme (catechol-O-methyltransferase) that breaks down dopamine in the prefrontal cortex of the brain. The second gene, SLC6A4, also known as the SERT gene, is found on chromosome 17 and encodes for how efficiently serotonin gets ferried across the synapses of your neurons.

  The COMT gene has three variants.§ One (known as val/val) encodes for a high level of the enzyme that breaks down dopamine very effectively; the others (val/met and met/met) encode for lower levels that break down less dopamine and leave more of it in the synapses. Recent studies have found that people with the met/met version tend to have a harder time regulating their emotional arousal. The excess levels of dopamine, researchers speculate, are linked to “negative emotionality” and to an “inflexible attentional focus” that leaves people unable to tear themselves away from obsessional preoccupation with frightening stimuli—traits that are, in turn, linked to depression, neuroticism, and, especially, anxiety. People with the met/met variant show an inability to relax after exposure to apparently threatening stimuli, even when those stimuli are revealed to be not dangerous after all. In contrast, the val/val variant was associated with less intense experiencing of negative emotions, a less reactive startle reflex, and less behavioral inhibition.‖

  David Goldman, the chief of human neurogenetics at the National Institutes of Health, has labeled COMT the “worrier-warrior gene.” Those who possess the val/val version, according to Goldman, are “warriors”: under stressful conditions, this gene variant gives them a beneficial increase in the extracellular brain level of dopamine, which presumably makes them less anxious and less susceptible to pain and allows them to focus better. The extra dopamine also gives them “better working memory” while under stress. I would imagine that, for example, the NFL quarterback Tom Brady—who is legendary for his ability to make quick, smart decisions while under tremendous pressure (throwing the ball accurately to the correct receiver even as thousands of pounds of linebacker are bearing down on him at high speed and millions of people are watching and judging him)—has the warrior variant. But there are situations in which the worrier version, which 25 percent of the world population has, confers evolutionary advantages. Studies have shown that carriers of the met/met version perform better on cognitive tasks requiring memory and attention when not under severe stress; this suggests the worriers may be better at evaluating complex environments and therefore at avoiding danger. Each version confers a different adaptive strategy: those with the worrier variant are good at staying out of danger; those with the warrior variant act effectively once they’re in danger.a

  The SERT gene also has three variants: short/short, short/long, and long/long (these get abbreviated to s/s, s/l, and l/l). Starting in the mid-1990s, many studies have found that people with one or more short SERT allele (that is, those with the s/s or s/l variant of the gene) tend to process serotonin less efficiently than those with only long alleles—and that when carriers of short-allele polymorphisms are shown fear-producing images, they display more amygdala activity than carriers of the l/l pairing. This correlation between a specific gene and activity in the amygdala, researchers hypothesize, helps account for the higher rates of anxiety disorder and depression that other studies were finding in people with the s/s version.

  In the absence of life stress, people with the s/s and s/l genotypes weren’t any likelier to become depressed than people with the l/l genotypes. When stressful situations arose, however—whether in the form of financial, employment, health, or relationship problems—those individuals with short versions were more likely to become depressed or suicidal. Put the other way around, those people with the l/l variant seemed partly insulated against depression and anxiety even when under stress.b

  Kerry Ressler, a psychiatrist at Emory University, has produced similar findings about other genes. Ressler has discovered that whereas some genotypes seem to confer increased vulnerability to certain forms of anxiety disorder, other genotypes seem to confer almost complete resistance to them. For instance, a gene called CRHR1 encodes the structure of brain receptors for corticotropin-releasing hormone (CRH), which is released during activation of the fight-or-flight response or during times of prolonged stress. To oversimplify a little, there are three variants of this gene: C/C, C/T, and T/T (the letters refer to the sequence of proteins that encode the amino acids that make up your DNA). Looking at a group of five hundred people from inner-city Atlanta who had suffered high rates of poverty, trauma, and child abuse, Ressler found that the variant of the CRHR1 gene you inherited strongly predicted the likelihood of your developing depression as an adult if you were abused as a child. One homozygous version of the gene (C/C) was associated with child abuse victims being very likely to develop clinical depression in adulthood; the heterozygous version of the gene (C/T) was associated with a moderate like
lihood of depression in adulthood; and, most fascinatingly, the other homozygous version of the gene (T/T) was not associated at all with depression in adulthood—the T/T version of the gene seems to confer on these child abuse victims almost complete immunity to depression. Child abuse seemed to have had no long-term psychological effect at all on those with this version of the gene.

  Ressler has also discovered similar findings in studies on the gene responsible for coding the feedback sensitivity of glucocorticoid receptors. Variations in this gene, known as FKBP5, seem to have a powerful effect on the susceptibility of children to post-traumatic stress disorder. Whereas one variant of the FKBP5 gene seems to be associated with high rates of PTSD, another variant seems to confer strong resistance: kids with the G/G variant developed PTSD at only around a third the rate of kids with other variants.

  Research like this suggests that your susceptibility to nervous breakdown is strongly determined by your genes. Certain genotypes make you especially vulnerable to psychological breakdown when subjected to stress or trauma; other genotypes make you naturally resilient. No single gene, or even set of genes, programs you to be anxious per se. But certain gene combinations program you to have either a high or a low level of hypothalamic-pituitary-adrenal activity: if you were born with a sensitive autonomic nervous system and then get exposed to stress in early childhood, your HPA system gets sensitized even further, so it’s always hyperactive later in life, producing an excessively twitchy amygdala—which in turn primes you to develop depression or anxiety disorders. If, however, you were born with genes encoding low baseline levels of HPA activity, you will tend to have a high level of immunity to the effects of even severe stress.