Book Read Free

The Fear Factor

Page 22

by Abigail Marsh


  The second key mammalian function that oxytocin performs is enabling nursing. It’s not involved in milk production itself, but in making sure the milk can be drunk. Milk that is inside a breast will more or less stay there until a baby latches onto the nipple and starts to suck. That strange drawing sensation is transmitted from the nipple up to the hypothalamus, where it prompts a few small clusters of cells to start churning out oxytocin and relaying it to the nearby pituitary gland. From there, the pituitary releases the oxytocin into the bloodstream, where it filters down to myoepithelial cells in the breast, alerting them to release the milk they contain into the nipple. And voilà, out it flows. This neat little process is called the milk ejection reflex, and it has been keeping baby mammals alive for many millions of years.

  Of course, again, the physiological ability to lactate is not all that neat in isolation. Producing milk is only useful if it is accompanied by all the behavioral, cognitive, and emotional changes that allow the young to access it and benefit from it. These changes include the desire to spend lots of time in close proximity to the offspring, a lack of fear of them, and various nurturing and protective behaviors that keep them fed, clean, and safe. This seems like an enormous number of new adaptations for mammalian mothers to have acquired in a short period of time, and indeed it was. But incredibly, they are all supported by oxytocin, the same chemical that gets the babies born and the milk flowing. If milk and maternal care are the defining characteristics of mammals—and they are—then you could say that we mammals owe everything we are to this one little clump of nine amino acids. It gives me goose bumps when I really stop to contemplate it.

  Oxytocin’s importance was initially discovered in studies of rats. You might remember that Wilsoncroft’s wonderfully maternal subjects were all first-time mothers. That is an important detail, as rats who have never had babies before act quite differently toward pups. They are worse than ewes, believe it or not. Virgin female rats find the smell and cries of rat pups highly upsetting and generally go out of their way to avoid them. If forced to remain in close proximity to pups, they sometimes attack or even cannibalize them. “Ew,” you can almost hear them saying to themselves. “I really, really, really don’t like rat pups.”

  But one thing can turn these callous, cannibalizing monsters into attentive mothers willing to spend hours tirelessly rescuing rat pups from a dish, and it can do so almost instantaneously. That thing is oxytocin.

  In the days and hours before a rat gives birth for the first time, neurons that produce oxytocin begin to multiply in her hypothalamus. Receptors for the oxytocin molecule also proliferate throughout her brain, sprouting up where none had been before in regions like the olfactory nucleus, which moderates responses to smells; the hypothalamus; the stria terminalis, a ribbon of fibers connecting the hypothalamus and amygdala; and the amygdala itself. These changes seem to set the stage for caring behavior to flourish.

  Up through the 1970s, many efforts had been made to identify the neurotransmitters responsible for caring maternal behavior, and they had largely failed. Estrogen, progesterone, prolactin—all are hormones involved in female reproduction and seemed likely candidates for motherliness, but when they were injected into the brains of virgin rats, their aversion to pups remained unchanged. But when Cort Pedersen and his colleagues injected oxytocin into female rats’ brains, their responses to pups were transformed in minutes.

  At the start of one experiment, the researchers divided up more than 200 virgin female rats into groups. One group was randomly selected to have an inert saline solution injected into the fluid-filled cavities inside their brains. Then they were placed in the center of a cage that also held three squirming pups arrayed in a triangle, three inches apart from each other. Most of the rats did what virgin female rats usually do, which was to ignore the pups. Fewer than one in five showed signs of maternal behavior—although those that did showed the full suite within the hour, including picking up the pups and putting them into a little pile together, licking them clean, building a nest for them out of any materials handy, crouching over them, and retrieving back to the makeshift nest any pups who wriggled free.

  For comparison, the researchers tried injecting a number of other chemicals into other groups of rats, including vasotocin, vasopressin, and estrogen. Still no change. But when they injected oxytocin into the rats’ brains, the change was fast and profound. Now nearly three-quarters of the rats started to care assiduously for the strange pups—a 400 percent change over baseline. This showed that oxytocin can singlehandedly promote not just mothering but allomothering, as all the pups were strangers to the female. But again, note that even the powerful oxytocin didn’t affect all the rats equally. About 20 percent of the rats given oxytocin still failed to show any maternal behavior, and another 7 percent killed at least one pup—as did about 7 percent of the rats in every group.

  A tidal wave of research has since fleshed out the critical roles that oxytocin plays in mothering and allomothering. Oxytocin has been shown to induce maternal (and allomaternal) care in a wide variety of species, including rats, mice, rhesus monkeys, meerkats, goats, and sheep—and thus we can reasonably assume that it similarly affects all the other mammals that haven’t yet been tested. Female rats whose brains produce more oxytocin tend to be better mothers. Less devoted mothers get better, though, if you give them more oxytocin. In contrast, they do much worse if you chemically block oxytocin receptors in their brain, which eliminates maternal care almost completely. Oxytocin can induce sheep to begin tenderly caring for unfamiliar lambs, something they would normally never do, thirty seconds after it is injected into their brains. This is not terribly useful information for a rancher, as most sheep ranches are not equipped to perform brain surgery on their livestock. But the knowledge of oxytocin’s importance has given ranchers another tool for inducing ewes to accept orphaned lambs, which is—and I know this is gross, but it works—to poke either their hand or a special balloon up into the ewe’s birth canal and massage her cervix. Fun, huh? But this stimulation triggers a wave of oxytocin production, much the same way a baby suckling does, and causes the motherly lightbulb in even a sheep’s fairly dim brain to flicker to life.

  Oxytocin acts throughout the brain, and the specific locales in which it acts vary somewhat from species to species. But across species, the amygdala is a central locus of oxytocin’s effects. Oxytocin seems to act in the amygdala to reduce any aversion to the unfamiliar smell, sight, or sounds of an infant, preventing avoidance or aggression and opening the door—or switching the track, if you prefer—for care. Very recent research suggests that this is true in people just as it is in other mammals.

  Studying oxytocin’s effect on people is more challenging than studying it in rats or sheep. Researchers inject oxytocin directly into these animals’ brains because oxytocin is a very big molecule—so big that it is thought to have trouble crossing the blood-brain barrier if it’s injected into the bloodstream or swallowed. Researchers usually avoid injecting hormones directly into living human brains, so efforts to investigate oxytocin’s effects on humans were stymied for a while. But a simple method discovered in the 1990s solved the problem: squirting oxytocin up the nose as a nasal spray. There it gets absorbed through the thin, porous skin lining the sinuses and into the brain itself.

  When I began my postdoctoral work at NIMH, I was dying to set up an intranasal oxytocin study to test how oxytocin affects care-based responses in humans. I started getting the paperwork drawn up in 2004, before almost any human oxytocin research had been published. Unfortunately, the internecine bureaucracy of the NIH prevented me from actually running the study until 2006, by which point a wave of human oxytocin research had gotten under way, accompanied by huge swells of hype. Early studies found that oxytocin increased the amount of money people offered strangers in economic games, or the amount of time they spent focusing on their eyes. Oxytocin was quickly dubbed the “cuddle hormone” and the “love hormone.” News articles suggeste
d that car dealerships should pipe it through their HVAC systems to increase sales (really). Psychiatrists speculated that it might cure autism (it doesn’t, sadly). The research that generated all this hyperbole ended up getting a fairly bad rap, understandably, and later studies have questioned whether some of these early findings were even true. In truth, oxytocin is not a panacea for making all social interactions more cuddly and lovey. Why should it be? Its essential purpose is supporting the care of vulnerable offspring. (In some species, it has also since been exapted to serve related functions, like pair bonding and social recognition.) Thus, sometimes it will promote sweetness and cuddling and other times it will promote wariness and aggression against intruders—these are all forms of maternal care supported by oxytocin.

  In an effort to explore oxytocin’s care-related functions in humans, my research assistant Henry Yu and I spent two years asking people to squirt doses of oxytocin or a saline placebo up their noses in the Clinical Center of the NIH, just upstairs from where I was scanning the brains of psychopathic adolescents on the weekends. None of our oxytocin subjects tried to cuddle with us, but their behavior did change in ways that were consistent with increased parental care. In one study we found that a few squirts of oxytocin up each nostril increased subjects’ preference for infants’ faces, but decreased their preference for unfamiliar adults’ faces. This is just what you would expect if oxytocin influences not just parental care but alloparental care: increased preference for unfamiliar babies, whose faces carry key infantile stimuli, and wariness of unfamiliar adults who might do a baby harm. Think of sweet Mimi the chihuahua fiercely defending her pups from her owner once her squirrel babies got her oxytocin production humming, or the lioness who attacked other adult lions when they threatened her baby baboon.

  We also found—in keeping with the fact that there is always variability—that the degree to which oxytocin increased subjects’ preference for babies’ faces depended on variation in a gene called OXTR, which affects the activity of oxytocin receptors in the brain. We found that people who carried the “A” version of a particular segment of the OXTR gene preferred babies’ faces no matter what we gave them, but that people who carried only the “G” version of this polymorphism preferred infants’ faces only after a hit of oxytocin. These findings remind me of Pedersen’s rat studies, which found that a few rats are maternal even without extra oxytocin, but that most need an oxytocin boost to become fully maternal. Finally, we found that oxytocin also increased subjects’ ability to recognize happy facial expressions, although only happy expressions that were fairly subtle and hard to recognize. This was an interesting result, but several other researchers have since reported results that I find much more interesting: oxytocin has even stronger effects on increasing accuracy for recognizing fear. One study found that oxytocin improved people’s ability to recognize fear (and only fear) by about 7 percent. Two more studies, conducted in Israel by Meytal Fischer-Shofty and Simone Shamay-Tsoory and their colleagues, found selective fear-recognition improvements of 13 and 20 percent, respectively.

  That oxytocin strongly increases sensitivity to fearful faces—the same faces that psychopathic people fail to recognize, that highly altruistic people recognize with exquisite sensitivity, and that evoke approach and caring in people who see them—is, to me, remarkably clear evidence that oxytocin underlies the power of these vulnerable, infantile expressions to cause those who see them to care.

  Now, for this to be true, oxytocin would need to be able to accomplish two things simultaneously. It would need to promote a strong empathic response to distress cues like fearful expressions to enable these expressions to be interpreted. But it also would need to inhibit the urge to avoid or escape in favor of approaching and caring for the fearful. As it happens, the findings reported in a 2016 study of rats strongly suggest that oxytocin can accomplish exactly this delicate balance. When researchers gave oxytocin to rats who were under threat, the rats showed all the usual physiological signs of fear—like elevated heart rates—that are part of an empathic fear response. But the rats didn’t show any of the avoidance or freezing that normally accompanies fear. This striking pairing—intact fear physiology but not fear behavior, which would enable an animal to feel scared but not act scared—was mediated by oxytocin’s effects in two separate groups of cells within the central nucleus of the amygdala. These findings help to explain what might otherwise be a puzzling phenomenon in rats: anxious rats make much better mothers, including being braver when defending their pups from harm. Their courage seems to result from their unusually vigorous amygdala response coupled with a rush of oxytocin in this structure when their pups are in danger. I felt a thrill of joy when I first encountered these findings, which represent an essential piece of the strange puzzle of parental care, and of altruism more broadly.

  So, although no current technology can directly test this hypothesis in humans, here’s what I think is going on. Once a signal arrives in a human amygdala (the basolateral region, specifically) that somebody is frightened, two things happen. First, the basolateral nucleus responds vigorously, reflecting the importance of what it has detected. It then transmits the signal it receives to the central nucleus, which gins up an empathic response. For example, it tells the hypothalamus to increase physiological fear responding—pounding heart and sweaty palms and spiking blood pressure. Simultaneously, the vulnerable, infantile qualities of the expression are also being processed, setting off a wave of increased oxytocin production in the hypothalamus. When the oxytocin reaches the central nucleus of the amygdala, it triggers a response in the huge number of oxytocin-sensitive neurons that populate the lateral part of this nucleus. Neurons here suppress fear-relevant activity in other regions of the amygdala in response. These neurons may signal other cells in the central amygdala to inhibit what would otherwise have been an avoidant fear response in favor of a caring approach. These behaviors are regulated through the amygdala’s connections with various other structures in the parental care network, like the striatum and the periaqueductal gray, both of which are densely packed with oxytocin receptors.

  The sum total of all this activity is a signal from the parental care system to the rest of the brain that there’s a sweet, juicy baby out there in the world who’s in trouble and needs your help, so don’t be a nervous Nellie—go out and get it!

  I’d bet money on this system being fouled up in psychopaths (as well as other systems, I should add—psychopathy almost certainly represents not one single dysfunction but a constellation of them). Dysfunction throughout the amygdala prevents people who are psychopathic from registering others’ fear strongly in the first place, and even when they do, their oxytocin system probably isn’t set up to generate the urge to care, perhaps owing to observed abnormalities in their OXTR gene or other yet-to-be-discovered causes. In altruists, on the other hand, it’s likely that both of these systems are exquisitely sensitive. We know already that altruists’ amygdalas are highly sensitive to signs of others’ distress. Although so far we have data only about how they respond to fearful facial expressions, I would venture that a much wider array of other signs that a person is vulnerable and distressed—screams, frightened body language, crying, or other forms of helplessness or suffering conveyed verbally—might affect them similarly. Many of the altruistic kidney donors I’ve worked with say that they first felt the spontaneous urge to donate a kidney when they saw, heard, or read a news story about someone suffering from kidney disease. One read a harrowing Reddit post in which a stranger described what life in kidney failure is like and decided to donate that day. Several have been moved to donate after watching someone close to them suffer on dialysis. Harold says that one of the motivations for his donation was seeing an obituary for a child who had died of blood cancer because no bone marrow donor could be found. Lenny Skutnik was moved to dive into the icy Potomac after hearing the chilling scream of a drowning woman. I’ve always wondered whether my own rescuer caught a glimpse of my frightened
face through the windshield. I’ll probably never know. Altruists uniformly seem to get a blast of empathic distress in response to these cues, but perhaps owing to highly responsive oxytocin-producing cells in their hypothalamus, or an unusually high density of oxytocin receptors in certain regions of the amygdala, they don’t respond by avoiding or escaping what they’ve seen. They dive right in with care.

  If all this is true (and again, I’m betting it is), it would finally explain the fact that altruists are both sensitive to fear and brave in the face of others’ distress, and that their bravery is instinctive and intuitive. A complicated chain of events within some of the deepest recesses of their brains cause them to act on some of humankind’s most primitive, atavistic urges—urges that trace back to our earliest mammalian ancestors, whose babies needed food and care and would never have survived unless their very babyishness had prompted an unconditional desire to provide care and protection.

  A high degree of variation in the specificity and sensitivity of these systems results in a small fraction of the population being so acutely sensitive to others’ distress and vulnerability that they will respond with the same conviction and certainty when saving the life of a stranger as an ordinary person might feel only when the life of their own child (or mother) is at stake.

  7

  CAN WE BE BETTER?

  AS SHOULD BE clear by this point, an overwhelming body of scientific data supports the conclusion that human beings are in no way fundamentally selfish or callous. All the neural and cognitive tools needed to experience genuine concern for others’ welfare and the desire to help those in distress are part of our birthright as mammals who bear altricial, helpless infants in need of care and protection from both their parents and adults other than their parents. These tools include the ability to detect others’ distress, the tendency to feel concern when we do, and the desire to help those in distress, even individuals who are unrelated to us. Naturally, people vary considerably in terms of both concern for others and desire to help them, but it is the rare person who is totally blind to others’ distress and completely without concern for their welfare. That gives most of us plenty of innate capacity to build on.

 

‹ Prev