A2 -CT

Sapolsky, R. M. (2004). Why zebras don't get ulcers: The, acclaimed guide to stress, stress- related diseases and coping (3rd ed.). St. Martin's Griffin. ISBN: 9780805073690. Chapter 3

Stroke, Heart Attacks, and Voodoo Death

It’s one of those unexpected emergencies: you’re walking down the street, on your way to meet a friend for dinner. You’re already thinking about what you’d like to eat, savoring your hunger. Come around the corner and—oh no, a lion! As we now know, activities throughout your body shift immediately to meet the crisis: your digestive tract shuts down and your breathing rate skyrockets. Secretion of sex hormones is inhibited, while epinephrine, norepinephrine, and glucocorticoids pour into the bloodstream. And if your legs are going to save you, one of the most important additional things that better be going on is an increase in your cardiovascular output, in order to deliver oxygen and energy to those exercising muscles. The Cardiovascular Stress-Response Activating your cardiovascular system is relatively easy, so long as you have a sympathetic nervous system plus some glucocorticoids and don’t bother with too many details. The first thing you do is shift your heart into higher gear, get it to beat faster. This is accomplished by turning down parasympathetic tone, and in turn activating the sympathetic nervous system. Glucocorticoids add to this as well, both by activating neurons in the brain stem that stimulate sympathetic arousal, and by enhancing the effects of epinephrine and norepinephrine on heart muscle. You also want to increase the force with which your heart beats. This involves a trick with the veins that return blood to your heart. Your sympathetic nervous system causes them to constrict, to get more rigid. And that causes the returning blood to blast through those veins with more force. Blood returns to your heart with more force, slamming into your heart walls, distending them more than usual…and those heart walls, like a stretched rubber band, snap back with more force. So your heart rate and blood pressure have gone up. The next task is to distribute the blood prudently throughout that sprinting body of yours. Arteries are relaxed—dilated—that lead to your muscles, increasing blood flow and energy delivery there. At the same time, there is a dramatic decrease in blood flow to nonessential parts of your body, like your digestive tract and skin (you also shift the pattern of blood flow to your brain, something that will be discussed in chapter 10). The decrease in blood flow to the gut was first noted in 1833, in an extended study of a Native American who had a tube placed in his abdomen after a gunshot wound there. When the man sat quietly, his gut tissues were bright pink, well supplied with blood. Whenever he became anxious or angry, the gut mucosa would blanch, because of decreased blood flow. (Pure speculation, perhaps, but one suspects that his transients of anxiety and anger might have been related to those white folks sitting around experimenting on him, instead of doing something useful, like sewing him up.) There’s one final cardiovascular trick in response to stress, involving the kidneys. As that zebra with its belly ripped open, you’ve lost a lot of blood. And you’re going to need that blood to

deliver energy to your exercising muscles. Your body needs to conserve water. If blood volume goes down because of dehydration or hemorrhage, it doesn’t matter what your heart and veins are doing; your ability to deliver glucose and oxygen to your muscles will be impaired. What’s the most likely place to be losing water? Urine formation, and the source of the water in urine is the bloodstream. Thus, you decrease blood flow to your kidneys and, in addition, your brain sends a message to the kidneys: stop the process, reabsorb the water into the circulatory system. This is accomplished by the hormone vasopressin (known as antidiuretic hormone for its ability to block diuresis, or urine formation), as well as a host of related hormones that regulate water balance. A question no doubt at the forefront of every reader’s mind at this point: if one of the features of the cardiovascular stress-response is to conserve water in the circulation, and this is accomplished by inhibition of urine formation in the kidneys, why is it that when we are really terrified, we wet our pants? I congratulate the reader for homing in on one of the remaining unanswered questions of modern science. In trying to answer it, we run into a larger one. Why do we have bladders? They are dandy if you are a hamster or a dog, because species like those fill their bladders up until they are just about to burst and then run around their territories, demarcating the boundaries—odoriferous little “keep out” signs to the neighbors.* A bladder is logical for scent-marking species, but I presume that you don’t do that sort of thing.* For humans, it is a mystery, just a boring storage site. The kidneys, now those are something else. Kidneys are reabsorptive, bidirectional organs, which means you can spend your whole afternoon happily putting water in from the circulation and getting some back and regulating the whole thing with a collection of hormones. But once the urine leaves the kidneys and heads south to the bladder, you can kiss that stuff good-bye; the bladder is unidirectional. When it comes to a stressful emergency, a bladder means a lot of sloshy dead weight to carry in your sprint across the savanna. The answer is obvious: empty that bladder.*

“So! Planning on roaming the neighborhood with some of your buddies today?”

Everything is great now—you have kept your blood volume up, it is roaring through the body with more force and speed, delivered where it is most needed. This is just what you want when running away from a lion. Interestingly, Marvin Brown of the University of California at San Diego and Laurel Fisher of the University of Arizona have shown that a different picture emerges

when one is being vigilant—a gazelle crouching in the grass, absolutely quiet, as a lion passes nearby. The sight of a lion is obviously a stressor, but of a subtle sort; while having to remain as still as possible, you must also be prepared, physiologically, for a wild sprint across the grasslands with the briefest of warnings. During such vigilance, heart rate and blood flow tend to slow down, and vascular resistance throughout the body increases, including in the muscles. Another example of the complicating point brought up at the end of chapter 2 about stress signatures—you don’t turn on the identical stress-response for every type of stressor. Finally, the stressor is over, the lion pursues some other pedestrian, you can return to your dinner plans. The various hormones of the stress-response turn off, your parasympathetic nervous system begins to slow down your heart via something called the vagus nerve, and your body calms down. Chronic Stress and Cardiovascular Disease So you’ve done all the right things during your lion encounter. But if you put your heart, blood vessels, and kidneys to work in this way every time someone irritates you, you increase your risk of heart disease. Never is the maladaptiveness of the stress-response during psychological stress clearer than in the case of the cardiovascular system. You sprint through the restaurant district terrified, and you alter cardiovascular functions to divert more blood flow to your thigh muscles. In such cases, there’s a wonderful match between blood flow and metabolic demand. In contrast, if you sit and think about a major deadline looming next week, driving yourself into a hyperventilating panic, you still alter cardiovascular function to divert more blood flow to your limb muscles. Crazy. And, potentially, eventually damaging. How does stress-induced elevation of blood pressure during chronic psychological stress wind up causing cardiovascular disease, the number one killer in the United States and the developed world? Basically, your heart is just a dumb, simple mechanical pump, and your blood vessels are nothing more exciting than hoses. The cardiovascular stress-response essentially consists of making them work harder for a while, and if you do that on a regular basis, they will wear out, just like any pump or hose you’d buy at Sears. The first step in the road to stress-related disease is developing hypertension, chronically elevated blood pressure.* This one seems obvious: if stress causes your blood pressure to go up, then chronic stress causes your blood pressure to go up chronically. Task accomplished, you’ve got hypertension. It’s a bit messier because a vicious cycle emerges at this point. The little blood vessels distributed throughout your body have the task of regulating blood flow to the local neighborhoods as a means of ensuring adequate local levels of oxygen and nutrients. If you chronically raise your blood pressure—chronically increase the force with which blood is coursing through those small vessels—those vessels have to work harder to regulate the blood flow. Think of the ease it takes to control a garden hose spritzing water versus a firehose with a hydrant’s worth of force gushing through it. The latter takes more muscle. And that’s precisely what happens at these small vessels. They build a thicker muscle layer around them, to better control the increased force of

blood flow. But as a result of these thicker muscles, these vessels now have become more rigid, more resistant to the force of blood flow. Which tends to increase blood pressure. Which tends to further increase vascular resistance. Which tends… So you’ve gotten yourself chronically high blood pressure. This isn’t great for your heart. Blood is now returning to your heart with more force and, as mentioned, this makes for a greater impact upon the heart muscle wall that encounters that tsunami. Over time, that wall will thicken with more muscle. This is termed “left ventricular hypertrophy,” which means increasing the mass of the left ventricle, the part of the heart in question. Your heart is now lopsided, in a sense, being overdeveloped in one quadrant. This increases the risk of developing an irregular heartbeat. And more bad news: in addition, this thickened wall of ventricular heart muscle may now require more blood than the coronary arteries can supply. It turns out that after controlling for age, having left ventricular hypertrophy is the single best predictor of cardiac risk. The hypertension isn’t good for your blood vessels, either. A general feature of the circulatory system is that, at various points, large blood vessels (your descending aorta, for example) branch into smaller vessels, then into even smaller ones, and so on, down to tiny beds of thousands of capillaries. This process of splitting into smaller and smaller units is called bifurcation. (As a measure of how extraordinarily efficient this repeated bifurcation is in the circulatory system, no cell in your body is more than five cells away from a blood vessel—yet the circulatory system takes up only 3 percent of body mass.) One feature of systems that branch in this way is that the points of bifurcation are particularly vulnerable to injury. The branch points in the vessel wall where bifurcation occurs bear the brunt of the fluid pressure slamming into them. Thus, a simple rule: when you increase the force with which the fluid is moving through the system, turbulence increases and those outposts of wall are more likely to get damaged. With the chronic increase in blood pressure that accompanies repeated stress, damage begins to occur at branch points in arteries throughout the body. The smooth inner lining of the vessel begins to tear or form little craters of damage. Once this layer is damaged, you get an inflammatory response—cells of the immune system that mediate inflammation aggregate at the injured site. Moreover, cells full of fatty nutrients, called foam cells, begin to form there, too. In addition, during stress the sympathetic nervous system makes your blood more viscous. Specifically, epinephrine makes circulating platelets (a type of blood cell that promotes clotting) more likely to clump together, and these clumped platelets can get gummed up in these aggregates as well. As we’ll see in the next chapter, during stress you’re mobilizing energy into the bloodstream, including fat, glucose, and the “bad” type of cholesterol, and these can also add to the aggregate. All sorts of fibrous gunk builds up there, too. You’ve now made yourself an atherosclerotic plaque. Therefore, stress can promote plaque formation by increasing the odds of blood vessels being damaged and inflamed, and by increasing the likelihood that circulating crud (platelets, fat, cholesterol, and so on) sticks to those inflamed injury sites. For years, clinicians have tried to get a sense of someone’s risk of cardiovascular disease by measuring how much of one particular type of crud there is in the bloodstream. This is, of course, cholesterol, leading to such a skittishness about cholesterol that the egg industry has to urge us to give their cholesterol-filled products a break. High levels of cholesterol, particularly of “bad” cholesterol, certainly increase

the risk for cardiovascular disease. But they’re not a great predictor; a surprising number of folks can tolerate high levels of bad cholesterol without cardiovascular consequences, and only about half of heart attack victims have elevated cholesterol levels. In the last few years, it is becoming clear that the amount of damaged, inflamed blood vessels is a better predictor of cardiovascular trouble than is the amount of circulating crud. This makes sense, in that you can eat eleventy eggs a day and have no worries in the atherosclerosis realm if there are no damaged vessels for crud to stick to; conversely, plaques can be forming even amid “healthy” levels of cholesterol, if there is enough vascular damage.

A healthy blood vessel (left), and one with an atherosclerotic plaque (right).

How can you measure the amount of inflammatory damage? A great marker is turning out to be something called C-reactive protein (CRP). It is made in the liver and is secreted in response to a signal indicating an injury. It migrates to the damaged vessel where it helps amplify the cascade of inflammation that is developing. Among other things, it helps trap bad cholesterol in the inflamed aggregate. CRP is turning out to be a much better predictor of cardiovascular disease risk than cholesterol, even years in advance of disease onset. As a result, CRP has suddenly become quite trendy in medicine, and is fast becoming a standard endpoint to measure in general blood work on patients. Thus, chronic stress can cause hypertension and atherosclerosis—the accumulation of these plaques. One of the clearest demonstrations of this, with great application to our own lives, is to be found in the work of the physiologist Jay Kaplan at Bowman Grey Medical School. Kaplan built on the landmark work of an earlier physiologist, James Henry (who was mentioned in the previous chapter), who showed that purely social stress caused both hypertension and atherosclerosis in mice. Kaplan and colleagues have shown a similar phenomenon in primates, bringing the story much closer home to us humans. Establish male monkeys in a social group, and over the course of days to months they’ll figure out where they stand with respect to one another. Once a stable dominance hierarchy has emerged, the last place you want to be is on the bottom: not only are you subject to the most physical stressors but, as will be reviewed in chapter 13 on psychological stress, to the most psychological stressors as well. Such subordinate males show a lot of the physiological indices of chronically turning on their stress-responses. And often these animals wind up with atherosclerotic plaques—their arteries are all clogged up. As evidence that the atherosclerosis arises from the overactive sympathetic nervous system component of the stress-response, if Kaplan gave the monkeys at risk drugs that prevent sympathetic activity (beta-blockers), they didn’t form plaques.

Kaplan showed that another group of animals is also at risk. Suppose you keep the dominance system unstable by shifting the monkeys into new groups every month, so that all the animals are perpetually in the tense, uncertain stage of figuring out where they stand with respect to everyone else. Under those circumstances, it is generally the animals precariously holding on to their places at the top of the shifting dominance hierarchy who do the most fighting and show the most behavioral and hormonal indices of stress. And, as it turns out, they have tons of atherosclerosis; some of the monkeys even have heart attacks (abrupt blockages of one or more of the coronary arteries). In general, the monkeys under the most social stress were most at risk for plaque formation. Kaplan showed that this can even occur with a low-fat diet, which makes sense, since, as will be described in the next chapter, a lot of the fat that forms plaques is being mobilized from stores in the body, rather than coming from the cheeseburger the monkey ate just before the tense conference. But if you couple the social stress with a high-fat diet, the effects synergize, and plaque formation goes through the roof. So stress can increase the risk of atherosclerosis. Form enough atherosclerotic plaques to seriously obstruct flow to the lower half of the body and you get claudication, which means that your legs and chest hurt like hell for lack of oxygen and glucose whenever you walk; you are then a candidate for bypass surgery. If the same thing happens to the arteries going to your heart, you can get coronary heart disease, myocardial ischemia, all sorts of horrible things. But we’re not done. Once you’ve formed those plaques, continued stress can get you in trouble another way. Again, increase stress and increase blood pressure, and, as the blood moves with enough force, increase the chances of tearing that plaque loose, rupturing it. So maybe you’ve had a plaque form in a huge aqueduct of a blood vessel, with the plaque being way too small to cause any trouble. But tear it loose now, form what is called a thrombus, and that mobile hairball can now lodge in a much smaller blood vessel, clogging it completely. Clog up a coronary artery and you’ve got a myocardial infarct, a heart attack (and this thrombus route accounts for the vast majority of heart attacks). Clog up a blood vessel in the brain and you have a brain infarct (a stroke). But there’s more bad news. If chronic stress has made a mess of your blood vessels, each individual new stressor is even more damaging, for an additional insidious reason. This has to do with myocardial ischemia, a condition that arises when the arteries feeding your heart have become sufficiently clogged that your heart itself is partially deprived of blood flow and thus of oxygen and glucose.* Suppose something acutely stressful is happening, and your cardiovascular system is in great shape. You get excited, the sympathetic nervous system kicks into action. Your heart speeds up in a strong, coordinated fashion, and its contractive force increases. As a result of working harder, the heart muscle consumes more energy and oxygen and, conveniently, the arteries going to your heart dilate in order to deliver more nutrients and oxygen to the muscle. Everything is fine.

But if you encounter an acute stressor with a heart that has been suffering from chronic myocardial ischemia, you’re in trouble. The coronary arteries, instead of vasodilating in response to the sympathetic nervous system, vasoconstrict. This is very different from the scenario described at the beginning of the chapter, where you are constricting some big blood vessels that deliver blood to unessential parts of your body. Instead, these are the small vessels diverting blood right to your heart. Just when your heart needs more oxygen and glucose delivered through these already clogged vessels, acute stress shuts them down even more, producing a shortage of nutrients for the heart, myocardial ischemia. This is exactly the opposite of what you need. Your chest is going to hurt like crazy—angina pectoris. And it turns out that it takes only brief periods of hypertension to cause this vasoconstrictive problem. Therefore, chronic myocardial ischemia from atherosclerosis sets you up for, at the least, terrible chest pain whenever anything physically stressful occurs. This is the perfect demonstration of how stress is extremely effective at worsening a pre-existing problem.

A necrotic heart.

When cardiology techniques improved in the 1970s, cardiologists were surprised to discover that we are even more vulnerable to trouble in this realm than had been guessed. With the old techniques, you would take someone with myocardial ischemia and wire him (men are more prone to heart disease than women) up to some massive ECG machine (same as EKG), focus a huge X-ray camera on his chest, and then send him running on a treadmill until he was ready to collapse. Just as one would expect, blood flow to the heart would decrease and his chest would hurt. Some engineers invented a miniature ECG machine that can be strapped on while you go about your daily business, and ambulatory electrocardiography was invented. Everyone got a rude surprise. There were little ischemic crises occurring all over the place in people at risk. Most ischemic episodes turned out to be “silent”—they didn’t give a warning signal of pain. Moreover, all sorts of psychological stressors could trigger them, like public speaking, pressured interviews, exams. According to the old dogma, if you had heart disease, you had better worry when you were undergoing physical stress and getting chest pains. Now it appears that, for someone at risk, trouble is occurring under all sorts of circumstances of psychological stress in everyday life, and you may not even know it. Once the cardiovascular system is damaged, it appears to be immensely sensitive to acute stressors, whether physical or psychological. One last bit of bad news. We’ve been focusing on the stress-related consequences of activating the cardiovascular system too often. What about turning it off at the end of each psychological stressor? As noted earlier, your heart slows down as a result of activation of the vagus nerve by the parasympathetic nervous system. Back to the autonomic nervous system never letting you put

your foot on the gas and brake at the same time—by definition, if you are turning on the sympathetic nervous system all the time, you’re chronically shutting off the parasympathetic. And this makes it harder to slow things down, even during those rare moments when you’re not feeling stressed about something. How can you diagnose a vagus nerve that’s not doing its part to calm down the cardiovascular system at the end of a stressor? A clinician could put someone through a stressor, say, run the person on a treadmill, and then monitor the speed of recovery afterward. It turns out that there is a subtler but easier way of detecting a problem. Whenever you inhale, you turn on the sympathetic nervous system slightly, minutely speeding up your heart. And when you exhale, the parasympathetic half turns on, activating your vagus nerve in order to slow things down (this is why many forms of meditation are built around extended exhalations). Therefore, the length of time between heartbeats tends to be shorter when you’re inhaling than exhaling. But what if chronic stress has blunted the ability of your parasympathetic nervous system to kick the vagus nerve into action? When you exhale, your heart won’t slow down, won’t increase the time intervals between beats. Cardiologists use sensitive monitors to measure interbeat intervals. Large amounts of variability (that is to say, short interbeat intervals during inhalation, long during exhalation) mean you have strong parasympathetic tone counteracting your sympathetic tone, a good thing. Minimal variability means a parasympathetic component that has trouble putting its foot on the brake. This is the marker of someone who not only turns on the cardiovascular stress-response too often but, by now, has trouble turning it off. Sudden Cardiac Death The preceding sections demonstrate how chronic stress will gradually damage the cardiovascular system, with each succeeding stressor making the system even more vulnerable. But one of the most striking and best-known features of heart disease is how often that cardiac catastrophe hits during a stressor. A man gets shocking news: his wife has died; he’s lost his job; a child long thought to be dead appears at the door; he wins the lottery. The man weeps, rants, exults, staggers about gasping and hyperventilating with the force of the news. Soon afterward, he suddenly grasps at his chest and falls over dead from sudden cardiac arrest. A strong, adverse emotion like anger doubles the risk of a heart attack during the subsequent two hours. For example, during the O. J. Simpson trial, Bill Hodgman, one of the prosecutors, got chest pains around the twentieth time he jumped up to object to something Johnnie Cochran was saying, and collapsed afterward (he survived). This sort of cardiac vulnerability to strong emotions has led Las Vegas casinos to keep defibrillators handy. It also is thought to have a lot to do with why exposure to New York City is a risk factor for a fatal heart attack.* The phenomenon is quite well documented. In one study, a physician collected newspaper clippings on sudden cardiac death in 170 individuals. He identified a number of events that seemed to be associated with such deaths: the collapse, death, or threat of loss of someone close; acute grief; loss of status or self-esteem; mourning, on an anniversary; personal danger; threat of an injury, or recovery from such a threat; triumph or extreme joy. Other studies have shown the same. During the 1991 Persian Gulf war fewer deaths in Israel were due to SCUD missile damage than to sudden cardiac death among frightened elderly people. During the 1994 L. A. earthquake, there was similarly a big jump in heart attacks.*

The actual causes are obviously tough to study (since you can’t predict what’s going to happen, and you can’t interview the people afterward to find out what they were feeling), but the general consensus among cardiologists is that sudden cardiac death is simply an extreme version of acute stress causing ventricular arrhythmia or, even worse, ventricular fibrillation plus ischemia in the heart.* As you would guess, it involves the sympathetic nervous system, and it is more likely to happen in damaged heart tissue than in healthy tissue. People can suffer sudden cardiac death without a history of heart disease and despite increased blood flow in the coronary vessels; autopsies have generally shown, however, that these people had a fair amount of atherosclerosis. Mysterious cases still occur, however, of seemingly healthy thirty-year-olds, victims of sudden cardiac death, who show little evidence of atherosclerosis on autopsy. Fibrillation seems to be the critical event in sudden cardiac death, as judged by animal studies (in which, for example, ten hours of stress for a rat makes its heart more vulnerable to fibrillation for days afterward). As one cause, the muscle of a diseased heart becomes more electrically excitable, making it prone to fibrillation. In addition, activation of stimulatory inputs to the heart becomes disorganized during a massive stressor. The sympathetic nervous system sends two symmetrical nervous projections to the heart; it is theorized that during extreme emotional arousal, the two inputs are activated to such an extent that they become uncoordinated—major fibrillation, clutch your chest, keel over. Fatal Pleasures Embedded in the list of categories of precipitants of sudden cardiac death is a particularly interesting one: triumph or extreme joy. Consider the scenario of the man dying in the aftermath of the news of his winning the lottery, or the proverbial “at least he died happy” instance of someone dying during sex. (When these circumstances apparently claimed the life of an ex-vice president a few decades back, the medical minutiae of the incident received especially careful examination because he was not with his wife at the time.) The possibility of being killed by pleasure seems crazy. Isn’t stress-related disease supposed to arise from stress? How can joyful experiences kill you in the same way that sudden grief does? Clearly, because they share some similar traits. Extreme anger and extreme joy have different effects on reproductive physiology, on growth, most probably on the immune system as well; but with regard to the cardiovascular system, they have fairly similar effects. Once again, we deal with the central concept of stress physiology in explaining similar responses to being too hot or too cold, a prey or a predator: some parts of our body, including the heart, do not care in which direction we are knocked out of allostatic balance, but rather simply how much. Thus wailing and pounding the walls in grief or leaping about and shouting in ecstasy can place similarly large demands on a diseased heart. Put another way, your sympathetic nervous system probably has roughly the same effect on your coronary arteries whether you are in the middle of a murderous rage or a thrilling orgasm. Diametrically opposite emotions then can have surprisingly similar physiological underpinnings (reminding one of the oft-quoted statement by Elie Wiesel, the Nobel laureate writer and Holocaust survivor: “The opposite of love is not hate. The opposite of love is indifference.”). When it comes to the cardiovascular system, rage and ecstasy, grief and triumph all represent challenges to allostatic equilibrium.

Women and Heart Disease Despite the fact that men have heart attacks at a higher rate than women, heart disease is nonetheless the leading cause of death among women in the United States—500,000 a year (as compared to 40,000 deaths a year for breast cancer). And the rate is rising among women while cardiovascular death rates in men have been declining for decades. Moreover, for the same severity of heart attack, women are twice as likely as men to be left disabled. What are these changes about? The increased rate of being disabled by a heart attack seems to be an epidemiological fluke. Women are still less subject to heart attacks than are men, with the onset of vulnerability delayed about a decade in women, relative to men. Therefore, if a man and woman both have heart attacks of the same severity, the woman is statistically likely to be ten years older than the man. And because of this, she is statistically less likely to bounce back afterward. But what about the increasing incidence of heart disease in women? Various factors are likely to be contributing to it. Obesity is skyrocketing in this country, more so in women, and this increases the risk of heart disease (as discussed in the next chapter). Moreover, though smoking rates are declining in the country, they are declining more slowly among women than men. Naturally, stress seems to have something to do with it as well. Kaplan and Carol Shively have studied female monkeys in dominance hierarchies and observe that animals chronically stuck in subordinate positions have twice the atherosclerosis as dominant females, even when on a low- fat diet. Findings with a similar theme of social subordination emerge among humans. This period of increasing rates of cardiovascular disease in women corresponds to a time when increasing percentages of women are working outside the home. Could the stressfulness of the latter have something to do with the former? Careful studies have shown that working outside the home does not increase the risk of cardiovascular disease for a woman. Unless she is doing clerical work. Or has an unsupportive boss. Go figure. And just to show what a myth it is that women working outside the home causes a shift toward men shouldering more of the burden of work at home, the other predictor of cardiovascular disease for women working outside the home is having kids back home. So why does stress increase the risk of cardiovascular disease in female primates, human or otherwise? The answer is all the usual suspects—too much sympathetic nervous system arousal, too much secretion of glucocorticoids. But another factor is relevant, one that is wildly controversial, namely estrogen. At the time of the previous edition of this book, estrogen was boring news. People had known for decades that estrogen protects against cardiovascular disease (as well as stroke, osteoporosis, and possibly Alzheimer’s disease), mostly thanks to estrogen working as an antioxidant, getting rid of damaging oxygen radicals. This explained why women didn’t start to get significant amounts of heart disease until after estrogen levels dropped with menopause. This was widely known and was one of the rationales for post-menopausal estrogen replacement therapy.

The importance of estrogen in protecting against cardiovascular disease came not just from statistics with human populations, but from careful experimental studies as well. As will be discussed in chapter 7, stress causes a decline in estrogen levels, and Kaplan’s low-ranking female monkeys had estrogen levels as low as you would find in a monkey that had had her ovaries removed. In contrast, subject a female to years of subordinance but treat her with estrogen, raising her levels to those seen in dominant animals, and the atherosclerosis risk disappears. And remove the ovaries of a high-ranking female, and she was no longer protected from atherosclerosis. Studies like these seemed definitive. Then in 2002 came a landmark paper, based on the Women’s Health Initiative, a study of thousands of women. The goal had been to assess the effects of eight years of post-menopausal replacement therapy with estrogen plus progestin. The expectation was that this was going to be the gold-standard demonstration of the protective effects of such therapy against cardiovascular disease, stroke, and osteoporosis. And five years into it, the codes as to who was getting hormone and who placebo were cracked, and the ethics panel overseeing the mammoth project brought it to a halt. Because the benefits of estrogen plus progestin were so clear that it was unethical to give half the women placebo? No—because estrogen plus progestin was so clearly increasing the risk of heart disease and stroke (while still protecting against osteoporosis) that it was unethical to continue the study. This was a bombshell. Front-page news everywhere. Similar trials were halted in Europe. Pharmaceutical stocks plummeted. And zillions of perimenopausal women wondered what they were supposed to do about estrogen replacement therapy. Why such contradictory findings, with years of clinical statistics and careful laboratory studies on one side, and this huge and excellent study on the other? As one important factor, studies like those of Kaplan’s involved estrogen, while this clinical trial was about estrogen plus progestin. This could well make a big difference. Then, as an example of the nit-picking that scientists love and which drives everyone else mad, the doses of hormones used probably made a difference, as did the type of estrogen (estradiol versus estriol versus estrone, and synthetic versus natural hormone). Finally, and this is an important point, the laboratory studies suggest that estrogen protects against the formation of atherosclerosis, rather than reverses atherosclerosis that is already there. This is quite relevant because, given our Western diets, people are probably just starting to form atherosclerotic plaques in their thirties, not in their post-menopausal fifties or sixties. The jury is still out on this one. And though it may not turn out that post-menopausal estrogen protects against cardiovascular disease, it seems plausible that estrogen secreted by women themselves at much younger ages does. And stress, by suppressing such estrogen levels, could be contributing to cardiovascular disease through that route. Voodoo Death The time has come to examine a subject far too rarely discussed in our public schools. Well- documented examples of voodoo death have emerged from all sorts of traditional non- westernized cultures. Someone eats a forbidden food, insults the chief, sleeps with someone he

or she shouldn’t have, does something unacceptably violent or blasphemous. The outraged village calls in a shaman who waves some ritualistic gewgaw at the transgressor, makes a voodoo doll, or in some other way puts a hex on the person. Convincingly soon, the hexed one drops dead. The Harvard team of ethnobotanist Wade Davis and cardiologist Regis DeSilva reviewed the subject.*Davis and DeSilva object to the use of the term voodoo death, since it reeks of Western condescension toward non-Western societies—grass skirts, bones in the nose, and all that. Instead, they prefer the term psychophysiological death, noting that in many cases even that term is probably a misnomer. In some instances, the shaman may spot people who are already very sick and, by claiming to have hexed them, gain brownie points when the person kicks off. Or the shaman may simply poison them and gain kudos for his cursing powers. In the confound (that is, the source of confusion) that I found most amusing, the shaman visibly puts a curse on someone, and the community says, in effect, “Voodoo cursing works; this person is a goner, so don’t waste good food and water on him.” The individual, denied food and water, starves to death; another voodoo curse come true, the shaman’s fees go up. Nevertheless, instances of psychophysiological death do occur, and they have been the focus of interest of some great physiologists in this century. In a great face-off, Walter Cannon (the man who came up with the fight-or-flight concept) and Curt Richter (a grand old man of psychosomatic medicine) differed in their postulated mechanisms of psychophysiological death. Cannon thought it was due to overactivity of the sympathetic nervous system; in that scheme, the person becomes so nervous at being cursed that the sympathetic system kicks into gear and vasoconstricts blood vessels to the point of rupturing them, causing a fatal drop in blood pressure. Richter thought death was due to too much parasympathetic activity. In this surprising formulation, the individual, realizing the gravity of the curse, gives up on some level. The vagus nerve becomes very active, slowing the heart down to the point of stopping—death due to what he termed a “vagal storm.” Both Cannon and Richter kept their theories unsullied by never examining anyone who had died of psychophysiological death, voodoo or otherwise. It turns out that Cannon was probably right. Hearts almost never stop outright in a vagal storm. Instead, Davis and DeSilva suggest that these cases are simply dramatic versions of sudden cardiac death, with too much sympathetic tone driving the heart into ischemia and fibrillation. All very interesting, in that it explains why psychophysiological death might occur in individuals who already have some degree of cardiac damage. But a puzzling feature about psychophysiological death in traditional societies is that it can also occur in young people who are extremely unlikely to have any latent cardiac disease. This mystery remains unexplained, perhaps implying more silent cardiac risk lurking within us than we ever would have guessed, perhaps testifying to the power of cultural belief. As Davis and DeSilva note, if faith can heal, faith can also kill. Personality and Cardiac Disease: A Brief Introduction

Two people go through the same stressful social situation. Only one gets hypertensive. Two people go through a decade’s worth of life’s ups and downs. Only one gets cardiovascular disease. These individual differences could be due to one person already having a damaged cardiovascular system—for example, decreased coronary blood flow. They could also be due to genetic factors that influence the mechanics of the system—the elasticity of blood vessels, the numbers of norepinephrine receptors, and so on. They could be the result of differences in how many risk factors each individual experiences—does the person smoke, eat a diet teeming with saturated fats? (Interestingly, individual differences in these risk factors explain less than half the variability in patterns of heart disease.) Faced with similar stressors, whether large or small, two people may also differ in their risk for cardiovascular disease as a function of their personalities. In chapters 14 and 15 I will review some of these—how the risk of cardiovascular disease is increased by hostility, a Type-A personality, and by clinical depression. The bad news is that these personality risk factors are substantial in their impact. But the good news is that something can often be done about them. This discussion has served as the first example of the style of analysis that will dominate the coming chapters. In the face of a short-term physical emergency, the cardiovascular stress- response is vital. In the face of chronic stress, those same changes are terrible news. These adverse effects are particularly deleterious when they interact with the adverse consequences of too much of a metabolic stress-response, the subject of the next chapter. Chapter 4

Stress, Metabolism, and Liquidating Your Assets

So you’re sprinting down the street with the lion after you. Things looked grim for a moment there, but—your good luck—your cardiovascular system kicked into gear, and now it is delivering oxygen and energy to your exercising muscles. But what energy? There’s not enough time to consume a candy bar and derive its benefits as you sprint along; there’s not even enough time to digest food already in the gut. Your body must get energy from its places of storage, like fat or liver or non-exercising muscle. To understand how you mobilize energy in this circumstance, and how that mobilization can make you sick at times, we need to learn how the body stores energy in the first place. Putting Energy in the Bank

The basic process of digestion consists of breaking down chunks of animals and vegetables so that they can then be transformed into chunks of human. We can’t make use of the chunks exactly as they are; we can’t, for example, make our leg muscles stronger by grafting on the piece of chicken muscle we ate. Instead, complex food matter is broken down into its simplest parts (molecules): amino acids (the building blocks of protein), simple sugars like glucose (the building blocks of more complex sugars and of starches [carbohydrates]), and free fatty acids and glycerol (the constituents of fat). This is accomplished in the gastrointestinal tract by enzymes, chemicals that can degrade more complex molecules. The simple building blocks thus produced are absorbed into the bloodstream for delivery to whichever cells in the body need them. Once you’ve done that, the cells have the ability to use those building blocks to construct the proteins, fats, and carbohydrates needed to stay in business. And just as important, those simple building blocks (especially the fatty acids and sugars) can also be burned by the body to provide the energy to do all that construction and to operate those new structures afterward. It’s Thanksgiving, and you’ve eaten with porcine abandon. Your bloodstream is teeming with amino acids, fatty acids, glucose. It’s far more than you need to power you over to the couch in a postprandial daze. What does your body do with the excess? This is crucial to understand because, basically, the process gets reversed when you’re later sprinting for your life. To answer this question, it’s time we talked finances, the works—savings accounts, change for a dollar, stocks and bonds, negative amortization of interest rates, shaking coins out of piggy banks—because the process of transporting energy through the body bears some striking similarities to the movement of money. It is rare today for the grotesquely wealthy to walk around with their fortunes in their pockets, or to hoard their wealth as cash stuffed inside mattresses. Instead, surplus wealth is stored elsewhere, in forms more complex than cash: mutual funds, tax-free government bonds, Swiss bank accounts. In the same way, surplus energy is not kept in the body’s form of cash—circulating amino acids, glucose, and fatty acids—but stored in more complex forms. Enzymes in fat cells can combine fatty acids and glycerol to form triglycerides (table). Accumulate enough of these in the fat cells and you grow plump. Meanwhile, your cells can stick series of glucose molecules together. These long chains, sometimes thousands of glucose molecules long, are called glycogen. Most glycogen formation occurs in your muscles and liver. Similarly, enzymes in cells throughout the body can combine long strings of amino acids, forming them into proteins. The hormone that stimulates the transport and storage of these building blocks into target cells is insulin. Insulin is this optimistic hormone that plans for your metabolic future. Eat a huge meal and insulin pours out of the pancreas into the bloodstream, stimulating the transport of fatty acids into fat cells, stimulating glycogen and protein synthesis. It’s insulin that’s filling out the deposit slips at your fat banks. We even secrete insulin when we are about to fill our bloodstream with all those nutritive building blocks: if you eat dinner each day at six o’clock, by five forty-five you’re already secreting insulin in anticipation of the rising glucose levels in your bloodstream. Logically, it is the parasympathetic nervous system that stimulates the anticipatory secretion, and this ability to secrete insulin in preparation for the glucose levels that are about to rise is a great example of the anticipatory quality of allostatic balance.

Emptying the Bank Account: Energy Mobilization During a Stressor This grand strategy of breaking your food down into its simplest parts and reconverting it into complex storage forms is precisely what your body should do when you’ve eaten plenty. And it is precisely what your body should not do in the face of an immediate physical emergency. Then, you want to stop energy storage. Turn up the activity of the sympathetic nervous system, turn down the parasympathetic, and down goes insulin secretion: step one in meeting an emergency accomplished. The body makes sure that energy storage is stopped in a second way as well. With the onset of the stressful emergency, you secrete glucocorticoids, which block the transport of nutrients into fat cells. This counteracts the effects of any insulin still floating around. So you’ve made sure you don’t do anything as irrational as store away new energy at this time. But in addition, you want your body to gain access to the energy already stored. You want to dip into your bank account, liquidate some of your assets, turn stored nutrients into your body’s equivalent of cash to get you through this crisis. Your body reverses all of the storage steps through the release of the stress hormones glucocorticoids, glucagon, epinephrine, and norepinephrine. These cause triglycerides to be broken down in the fat cells and, as a result, free fatty acids and glycerol pour into the circulatory system. The same hormones trigger the degradation of glycogen to glucose in cells throughout the body, and the glucose is then flushed into the bloodstream. These hormones also cause protein in non-exercising muscle to be converted back to individual amino acids. The stored nutrients have now been converted into simpler forms. Your body makes another simplifying move. Amino acids are not a very good source of energy, but glucose is. Your body shunts the circulating amino acids to the liver, where they are converted to glucose. The liver can also generate new glucose, a process called gluconeogenesis, and this glucose is now readily available for energy during the disaster. As a result of these processes, lots of energy is available to your leg muscles. There’s a burst of activity; you leave the lion in the dust and arrive at the restaurant only a smidgen late for your five forty-five anticipatory insulin secretion.

The scenario I’ve been outlining is basically a strategy to shunt energy from storage sites like fat to muscle during an emergency. But it doesn’t make adaptive sense to automatically fuel, say, your arm muscles while you’re running away from a predator if you happen to be an upright human. It turns out that the body has solved this problem. Glucocorticoids and the other hormones of the stress-response also act to block energy uptake into muscles and into fat tissue. Somehow the individual muscles that are exercising during the emergency have a means to override this blockade and to grab all the nutrients floating around in the circulation. The net result is that you shunt energy from fat and from non-exercising muscle to the exercising ones. And what if you can’t mobilize energy during a crisis? This is what occurs in Addison’s disease, where people cannot secrete adequate amounts of glucocorticoids, or in Shy-Drager syndrome, where it is epinephrine and norepinephrine that are inadequate, having an inability to mobilize the body during energetic demands. Obviously, the lion is more likely to feast. And in a more subtle scenario, if you live in a westernized society and tend to have a somewhat underactive stress-response? Just as obviously, you’ll have trouble mobilizing energy in response to the demands of daily life. And that is precisely what is seen in individuals with chronic fatigue syndrome, which is characterized by, among other things, too low levels of glucocorticoids in the bloodstream. So Why Do We Get Sick? You most definitely want to have a metabolic stress-response if you’re evading a lion, and even if you are doing anything as taxing as walking up a flight of stairs (or even getting up in the morning, the time of day when our glucocorticoid levels normally peak). But what about the more typical scenario for us, one of turning on the stress-response too often, for months on end? We get into metabolic trouble for many of the same reasons that constantly running to the bank and drawing on your account is a foolish way to handle your finances. On the most basic level, it’s inefficient. Another financial metaphor helps. Suppose you have some extra money and decide to put it away for a while in a high-interest account. If you agree not to touch the money for a certain period (six months, two years, whatever), the bank agrees to give you a higher-than-normal rate of interest. And, typically, if you request the money earlier, you will pay a penalty for the early withdrawal. Suppose, then, that you happily deposit your money on these terms. The next day you develop the financial jitters, withdraw your money, and pay the penalty. The day after, you change your mind again, put the money back in, and sign a new agreement, only to change your mind again that afternoon, withdraw the money, and pay another penalty. Soon you’ve squandered half your money on penalties. In the same way, every time you store energy away from the circulation and then return it, you lose a fair chunk of the potential energy. It takes energy to shuttle those nutrients in and out of the bloodstream, to power the enzymes that glue them together (into proteins, triglycerides, and glycogen) and the other enzymes that then break them apart, to fuel the liver during that gluconeogenesis trick. In effect, you are penalized if you activate the stress-response too often: you wind up expending so much energy that, as a first consequence, you tire more readily—just plain old everyday fatigue.

As a second consequence, your muscles can waste away, although this rarely happens to a significant degree. Muscle is chock-full of proteins. If you are stressed chronically, constantly triggering the breakdown of proteins, your muscles never get the chance to rebuild. While they atrophy ever so slightly each time your body activates this component of the stress-response, it requires a really extraordinary amount of stress for this to happen to a serious extent. As we will see in later chapters, sometimes clinicians give patients massive doses of synthetic glucocorticoids. In this scenario, significant amounts of myopathy—atrophy of muscle—can occur, of a type similar to that seen in people who are bedridden for long periods. Finally, another problem with constantly mobilizing the metabolic stress-response was hinted at in the last chapter. You don’t want to have tons of fat and glucose perpetually circulating in your bloodstream because, as we saw, that increases the chances of the stuff glomming on to some damaged blood vessel and worsening atherosclerosis. Cholesterol also plays into this. As is well understood, there is “bad” cholesterol, also known as low-density lipoprotein-associated cholesterol (LDL) and “good” cholesterol (high-density lipoprotein-associated cholesterol, HDL). LDL-cholesterol is the type that gets added to an atherosclerotic plaque, whereas HDL- cholesterol is cholesterol that has been removed from plaques and is on its way to be degraded in the liver. As a result of this distinction, your total level of cholesterol in the bloodstream is not actually a meaningful number. You want to know how much of each type you have, and lots of LDL and minimal HDL are independently bad news. We saw in the last chapter that the amount of vascular inflammation, as measured by CRP levels, is the best predictor out there of cardiovascular disease risk. Nonetheless, you don’t want to have tons of LDL-cholesterol floating around and not enough HDL to counteract it. And during stress, you increase LDL- cholesterol levels and decrease HDL.* Therefore, if you are stressed too often, the metabolic features of the stress-response can increase your risks of cardiovascular disease. This becomes particularly relevant with diabetes. Juvenile Diabetes There are multiple forms of diabetes, and two are relevant to this chapter. The first is known as juvenile diabetes (or type 1, insulin-dependent diabetes). For reasons that are just being sorted out, in some people the immune system decides that the cells in the pancreas that secrete insulin are, in fact, foreign invaders and attacks them (such “autoimmune” diseases will be discussed in chapter 8). This destroys those cells, leaving the person with little ability to secrete insulin. For equally mysterious reasons, this tends to hit people relatively early in life (hence the “juvenile” part of the name) although, to add to the mystery, in recent decades, the rate at which adults, even middle-aged adults, are getting diagnosed with juvenile diabetes is climbing. Because the person can no longer secrete adequate amounts of insulin (if any), there is little ability to promote the uptake of glucose (and, indirectly, fatty acids) into target cells. Cells starve—big trouble, not enough energy, organs don’t function right. In addition, there’s now all that glucose and fatty acid circulating in the bloodstream—oleaginous hoodlums with no place to go, and soon there’s atherosclerotic trouble there as well. The circulating stuff gums up the blood vessels in the kidneys, causing them to fail. The same can occur in the eyes, causing blindness.

Blood vessels elsewhere in the body are clogged, causing little strokes in those tissues and, often, chronic pain. With enough glucose in the circulation, it begins to stick to proteins, begins to Velcro proteins together that have no business being connected, knocking them out of business. None of this good. And what is the best way to manage insulin-dependent diabetes? As we all know, by accommodating that dependency with insulin injections. If you’re diabetic, you never want your insulin levels to get too low—cells are deprived of energy, circulating glucose levels get too high. But you don’t want to take too much insulin. For complex reasons, this deprives the brain of energy, potentially putting you into shock or a coma and damaging neurons. The better the metabolic control in a diabetic, the fewer the complications and the longer the life expectancy. Thus, there’s a major task for this type of diabetic to keep things just right, to keep food intake and insulin dosages balanced with respect to activity, fatigue, and so on. And this is an area where there has been extraordinary technological progress enabling diabetics to monitor blood glucose levels minute by minute and make minuscule changes in insulin dosages accordingly. How does chronic stress affect this process? First, the hormones of the stress-response cause even more glucose and fatty acids to be mobilized into the bloodstream. For a juvenile diabetic, this increases the likelihood of the now-familiar pathologies of glucose and fatty acids gumming up in the wrong places. Another, more subtle problem occurs with chronic stress as well. When something stressful happens, you don’t just block insulin secretion. Basically, the brain doesn’t quite trust the pancreas not to keep secreting a little insulin, so a second step occurs. As noted earlier, during stress, glucocorticoids act on fat cells throughout the body to make them less sensitive to insulin, just in case there’s some still floating around. Fat cells then release some newly discovered hormones that get other tissues, like muscle and liver, to stop responding to insulin as well. Stress promotes insulin resistance. (And when people get into this diabetic state because they are taking large amounts of synthetic glucocorticoids [to control any of a variety of diseases that will be discussed later in the book] they have succumbed to “steroid diabetes.”) Why is this stress-induced insulin resistance bad for someone with juvenile diabetes? They have everything nice and balanced, with a healthy diet, a good sensitivity to their body’s signals as to when a little insulin needs to be injected, and so on. But throw in some chronic stress, and suddenly insulin doesn’t work quite as well, causing people to feel terrible until they figure out that they need to inject more of the stuff…which can make cells even more resistant to insulin, spiraling the insulin requirements upward…until the period of stress is over with, at which point it’s not clear when to start getting the insulin dose down…because different parts of the body regain their insulin sensitivity at different rates…. The perfectly balanced system is completely upended. Stress, including psychological stress, can wreak havoc with metabolic control in a juvenile diabetic. In one demonstration of this, diabetics were exposed to an experimental stressor (speaking in public) and their glucocorticoid secretion was monitored. Those who tended to have the largest stress-response under those circumstances were the ones least likely to have their

diabetes well controlled. Moreover, in related studies, those who had the strongest emotional reactions to an experimental stressor tended to have the highest blood glucose levels. Stress may sneak in another way. Some careful studies have shown higher rates of major stressors suffered by people during the three years before the onset of their juvenile diabetes than would be expected by chance. Does this mean that stress can make the immune system more likely to attack the pancreas? There is a little bit of evidence for this, which will be discussed in chapter 8 on immunity. A more likely explanation is built around the fact that once the immune system begins to attack the pancreas (that is, once the diabetes has started), it takes a while before the symptoms become apparent. By having all the adverse effects just talked about, stress can speed up the whole process, making the person notice sooner that he or she is just not feeling right. Thus, frequent stress and/or big stress-responses might increase the odds of getting juvenile diabetes, accelerate the development of the diabetes, and, once it is established, cause major complications in this life-shortening disease.* Therefore, this is a population in which successful stress management is critical. Adult-Onset Diabetes In adult-onset diabetes (type 2, non-insulin-dependent diabetes), the trouble is not too little insulin, but the failure of the cells to respond to insulin. Another name for the disorder is thus insulin-resistant diabetes. The problem here arises with the tendency of many people to put on weight as they age. (However, if people do not put on weight as they age, they show no increased risk of this disease. This is the case among people in non-westernized populations. The disease is not, therefore, a normal feature of aging; instead, it is a disease of inactivity and fat surplus, conditions that just happen to be more common with age in some societies.) With enough fat stored away, the fat cells essentially get full; once you are an adolescent, the number of fat cells you have is fixed, so if you put on weight, the individual fat cells are distended. Yet another heavy meal, a burst of insulin trying to promote more fat storage by the fat cells, and the fat cells refuse—“Tough luck, I don’t care if you are insulin; we’re completely full.” No room at the inn. The fat cells become less responsive to insulin trying to promote more fat storage, and less glucose is taken up by these cells.* The overstuffed fat cells even release hormones that trigger other fat cells and muscle into becoming insulin resistant. Do the cells now starve? Of course not, the abundant amounts of fat stored in them was the source of the trouble in the first place. The body gets into trouble because of all that circulating glucose and fatty acids, damaging blood vessels. Same old problem. And if the adult-onset diabetes goes on for a while, an additional, miserable development can occur. Your body has become insulin-resistant. Your pancreas responds by secreting even more insulin than usual. You’re still resistant. So the pancreas secretes even more. Back and forth, your pancreas pumping out ever higher levels of insulin, trying to be heard. Eventually, this burns out the insulin-secreting cells in the pancreas, actually destroying them. So you finally get your adult- onset diabetes under control, thanks to losing weight and exercising, and you discover you’ve now got juvenile diabetes, thanks to that damage to your pancreas.

Photomicrograph of bloated fat cells.

How does chronic stress affect adult-onset diabetes? Once again, constantly mobilizing glucose and fatty acids into the bloodstream adds to the atherosclerotic glomming. And there’s that problem of the stress-response involving your fat cells being instructed to become less responsive to insulin. Suppose that you’re in your sixties, overweight, and just on the edge of insulin resistance. Along comes a period of chronic stress with those stress hormones repeatedly telling your cells what a great idea it is to be insulin-resistant. Enough of this and you pass the threshold for becoming overtly diabetic. Why is any of this worth paying attention to? Because there is a worldwide epidemic of adult- onset diabetes going on, especially in our country. As of 1990, about 15 percent of Americans over age sixty-five had adult-onset diabetes. That was considered a health disaster then. As of a decade later, there’s been a 33 percent increase above that, and among middle-aged adults as well. And this disease of aging is suddenly hitting far younger people as well—in the last decade, there’s been a 70 percent increase in its incidence among thirty-year-olds. In addition, something like 20 million Americans are “pre-diabetic”—barreling toward a formal diagnosis. Adult-onset diabetes has even become more prevalent among kids than juvenile diabetes, which is pretty horrifying. Moreover, as people in the developing world are first being exposed to westernized diets, not only do they develop diabetes, they develop it at a faster rate than do westerners, for reasons that are probably both cultural and genetic. This once nonexistent disease afflicts an estimated 300 million people worldwide and killed 200,000 Americans last year. What’s this about? It’s obvious. Despite the impression that everyone spends their days eating low-fat/carb/cholesterol/cardboard diets and power walking uphill while loudly reciting the writings of Atkins or Ornish, with each passing year, we are eating more food—more junk food—and exercising less. Twenty percent of Americans are now technically “obese” (versus 12 percent in 1990), and 54 percent are “overweight” (versus 44 percent then). To paraphrase the allostasis theorist Joseph Eyer, prosperity has become a cause of death.* Metabolic Syndrome/Syndrome X

In the well-entrenched tradition of medical compartmentalizing, there’s a whole set of things that can go wrong in you that would get you sent to a cardiologist, whereas a bunch of different problems would get you turfed to an internal medicine doc who specializes in diabetes. With any luck, they’d even confer with each other now and then. What should be obvious over the last two chapters is that your metabolic and cardiovascular systems are intimately interconnected. “Metabolic syndrome” (also known as Syndrome X) is a new term recognizing this interconnection. It’s actually not so new, having been formalized in the late 1980s by Gerald Reaven of Stanford University. It’s just become tremendously trendy in the past few years (so trendy that it’s even been described in a population of wild baboons who forage through the desserts in a garbage dump at a tourist lodge in East Africa). Make a list of some of the things that can go wrong from the last two chapters: elevated insulin levels in the blood. Elevated glucose levels. Elevated systolic and diastolic blood pressure. Insulin resistance. Too much LDL-cholesterol. Too little HDL. Too much fat or cholesterol in the blood. Suffer from a subset of these, and you’ve got Metabolic syndrome (the formal diagnosis involves “one or more” from a list of some of these problems, and “two or more” from a list of the others).* The syndrome-ness is a way of stating that if you have a subset of those symptoms, you’re probably heading toward the rest of them, since they’re all one or two steps away from each other. Have elevated insulin levels, low HDL, and abdominal obesity and the chances are pretty good you’re going to get insulin resistance. Elevated LDL-cholesterol, high blood pressure, and insulin resistance, and you’re likely to be obese soon. Another bunch and they predict hypertension. Subsets of these clusters of traits not only predict each other, they collectively predict major disease outcomes, like heart attacks or stroke, and mortality rates. This was shown with particular subtlety in an impressive study carried out by a team headed by Teresa Seeman of UCLA. Medicine normally works in diagnostic categories: have glucose levels above X, and it’s official, you have hyperglycemia. Have blood pressure levels above Z, you’re hypertensive. But how about if your glucose levels, blood pressure, HDL-cholesterol, and so on, are all in the normal range, but all of them are getting near the edge of where you have to start worrying? In other words, no measure is abnormal, but there’s an abnormally large number of measures that are almost abnormal. Technically, nothing is wrong, amid it being obvious that things are not right. Take more than a thousand study subjects, all over age seventy, none of whom are certifiably sick—that is to say, where none of those measures are technically abnormal. Now, see how they’re doing on all those Metabolic syndrome measures. Throw in some other measures as well—including resting levels of glucocorticoids, epinephrine, norepinephrine. Combine the insights into these measures mathematically and, collectively, this information was significantly predictive of who was going to have heart disease, a decline in cognitive or physical functioning, and mortality, far more predictive than subsets of those variables alone. This is the essence of that “allostasis” concept, of keeping things in balance through interactions among different, far-flung systems in the body. This is also the essence of the wear-and-tear concept of allostatic “load,” a formal demonstration that even if there’s no single measure that’s certifiably wrong, if there are enough things that are not quite right, you’re in trouble. And, as the final, obvious point, this is also the essence of what stress does. No single disastrous effect, no

lone gunman. Instead, kicking and poking and impeding, here and there, make this a bit worse, that a bit less effective. Thus making it more likely for the roof to cave in at some point. Chapter 5

Ulcers, the Runs, and Hot Fudge Sundaes

Not having enough food or water definitely counts as a stressor. If you’re a human, having enough food and water for this meal, but not being sure where the next meal is coming from is a major stressor as well, one of the defining experiences of life outside the westernized world. And choosing not to eat to the point of starvation—anorexia—is a stressor as well (and one with an odd endocrine signature, harking back to chapter 2, in that glucocorticoids tend to be elevated while the sympathetic nervous system is unexpectedly inhibited). None of this is surprising. Nor is it surprising that stress changes eating patterns. This is well established. The question, of course, is in what way. Stress and Food Consumption From the previous chapter it’s perfectly obvious where we’re heading in terms of appetite. You’re the zebra running for your life, don’t think about lunch. That’s the reason why we lose our appetites when we’re stressed. Except for those of us who, when stressed, eat everything in sight, in a mindless mechanical way. And those who claim they’re not hungry, are too stressed to eat a thing, and just happen to nibble 3,000 calories’ worth of food a day. And those of us who really can’t eat a thing. Except for chocolate-chocolate chip hot fudge sundaes. With whipped cream and nuts. The official numbers are that stress makes about two-thirds of people hyperphagic (eating more) and the rest hypophagic.* Weirdly, when you stress lab rats, you get the same confusing picture, where some become hyperphagic, others hypophagic. So we can conclude with scientific certainty that stress can alter appetite. Which doesn’t teach us a whole lot, since it doesn’t tell us whether there’s an increase or decrease. It turns out that there are ways to explain why some of us become hyper- and others hypophagic during stress. To start, we extend the zebra scenario to the point of it surviving its encounter. During the stressor, appetite and energy storage were suppressed, and stored energy was mobilized. Thus, what’s the logic during the post-stress period? Obvious—recover from that, reverse those processes. Block the energy mobilization, store the nutrients in your bloodstream, and get more of them. Appetite goes up. This is accomplished through some endocrinology that is initially fairly confusing, but is actually really elegant. The confusing issue is that one of the critical hormones of the stress-response stimulates appetite, while another inhibits it. You might recall from earlier chapters that the hormone CRH is released by the hypothalamus and, by stimulating the pituitary to release ACTH, starts the cascade of events that culminates in adrenal release of glucocorticoids. Evolution has allowed the development of efficient use of the body’s chemical messengers, and CRH is no exception. It is also used in parts of the brain to regulate other features of the stress- response. It helps to turn on the sympathetic nervous system, and it plays a role in increasing vigilance and arousal during stress. It also suppresses appetite. (Unsuccessful dieters should be warned against running to the neighborhood pharmacist for a bottle of CRH. It will probably

help you lose weight, but you’ll feel awful—as if you were always in the middle of an anxiety- provoking emergency: your heart racing; feeling jumpy, hyposexual, irritable. Probably better to just opt for a few more sit-ups.) On the other side of the picture are glucocorticoids. In addition to the actions already outlined in response to stress, they appear to stimulate appetite. This is typically demonstrated in rats: glucocorticoids make these animals more willing to run mazes looking for food, more willing to press a lever for a food pellet, and so on. The hormone stimulates appetite in humans as well (although, to my knowledge, no one has stoked human volunteers on glucocorticoids and then quantified them scurrying up and down supermarket aisles). Scientists have a reasonably good idea where in the brain glucocorticoids stimulate appetite, which type of glucocorticoid receptors are involved, and so on.* What is really fascinating is that glucocorticoids don’t just stimulate appetite—they stimulate it preferentially for foods that are starchy, sugary, or full of fat—and we reach for the Oreos and not the celery sticks. Thus, we appear to have a problem here. CRH inhibits appetite, glucocorticoids do the opposite.* Yet they are both hormones secreted during stress. Timing turns out to be critical. When a stressful event occurs, there is a burst of CRH secretion within a few seconds. ACTH levels take about fifteen seconds to go up, while it takes many minutes for glucocorticoid levels to surge in the bloodstream (depending on the species). Thus, CRH is the fastest wave of the adrenal cascade, glucocorticoids the slowest. This difference in time course is also seen in the speed at which these hormones work on various parts of the body. CRH makes its effects felt within seconds, while glucocorticoids take minutes to hours to exert their actions. Finally, when the stressful event is over, it takes mere seconds for CRH to be cleared from the bloodstream, while it can take hours for glucocorticoids to be cleared. Therefore, if there are large amounts of CRH in your bloodstream, yet almost no glucocorticoids, it is a safe bet that you are in the first few minutes of a stressful event. Good time to turn off appetite, and the combination of high CRH and low glucocorticoids accomplishes that. Next, if there are large amounts of CRH and glucocorticoids in the bloodstream, you are probably in the middle of a sustained stressor. Also a good time to have appetite suppressed. You can pull this off only if the appetite-suppressing effects of CRH are stronger than the appetite- stimulating effects of glucocorticoids. And that’s exactly how it works. Finally, if there are substantial amounts of glucocorticoids in the circulation but little CRH, you have probably started the recovery period. That’s exactly when digestion starts up again and your body can begin to replenish those stores of energy consumed in that mad dash across the savanna. Appetite is stimulated. In chapter 4, we saw how glucocorticoids help to empty out the bank account of stored energy during a stressor. In this case, glucocorticoids would not so much serve as the mediator of the stress-response, but as the means of recovering from the stress- response. Things now begin to make sense when you consider both the duration of a stressor and the recovery period combined. Suppose that something truly stressful occurs, and a maximal signal to secrete CRH, ACTH, and glucocorticoids is initiated. If the stressor ends after, say, ten

minutes, there will cumulatively be perhaps a twelve-minute burst of CRH exposure (ten minutes during the stressor, plus the seconds it takes to clear the CRH afterward) and a two-hour burst of exposure to glucocorticoids (the roughly eight minutes of secretion during the stressor plus the much longer time to clear the glucocorticoids). So the period where glucocorticoid levels are high and those of CRH are low is much longer than the period of CRH levels being high. A situation that winds up stimulating appetite. In contrast, suppose the stressor lasts for days, nonstop. In other words, days of elevated CRH and glucocorticoids, followed by a few hours of high glucocorticoids and low CRH, as the system recovers. The sort of setting where the most likely outcome is suppression of appetite. The type of stressor is key to whether the net result is hyper- or hypophagia. Take some crazed, maze-running rat of a human. He sleeps through the alarm clock first thing in the morning, total panic. Calms down when it looks like the commute isn’t so bad today, maybe he won’t be late for work after all. Gets panicked all over again when the commute then turns awful. Calms down at work when it looks like the boss is away for the day and she didn’t notice he was late. Panics all over again when it becomes clear the boss is there and did notice. So it goes throughout the day. And how would that person describe his life? “I am like, SO stressed, like totally, nonstop stressed, 24/7.” But that’s not really like totally nonstop stressed. Take a whole body burn. That’s like totally nonstop stressed, 24/7. What this first person is actually experiencing is frequent intermittent stressors. And what’s going on hormonally in that scenario? Frequent bursts of CRH release throughout the day. As a result of the slow speed at which glucocorticoids are cleared from the circulation, elevated glucocorticoid levels are close to nonstop. Guess who’s going to be scarfing up Krispy Kremes all day at work? So a big reason why most of us become hyperphagic during stress is our westernized human capacity to have intermittent psychological stressors throughout the day. The type of stressor is a big factor. Another variable that helps predict hyperphagia or hypophagia during stress is how your body responds to a particular stressor. Put a bunch of subjects through the same experimental stressor (for example, a session on an exercise bicycle, a time-pressured set of math questions, or having to speak in public) and, not surprisingly, not everyone secretes the exact same amount of glucocorticoids. Furthermore, at the end of the stressor, everyone’s glucocorticoid levels don’t return to baseline at the same rate. The sources of these individual differences can be psychological—the experimental stressor may be an utter misery for one person and no big deal for another. Differences can also arise from physiology—one person’s liver may be pokier at breaking down glucocorticoids than the next person’s. Elissa Epel of UCSF has shown that the glucocorticoid hypersecreters are the ones most likely to be hyperphagic after stress. Moreover, when given an array of foods to choose from during the post-stress period, they also atypically crave sweets. This is an effect that is specific to stress. The people who secrete excess glucocorticoids during stress don’t eat any more than the other subjects in the absence of stress, and their resting, non-stressed levels of glucocorticoids aren’t any higher than the others.

What else separates the stress hyperphagics from the stress hypophagics? Some of it has to do with your attitude toward eating. Lots of people eat not just out of nutritional need, but out of emotional need as well. These folks tend both to be overweight and to be stress-eaters. In addition, there’s a fascinating literature concerning the majority of us, for whom eating is a regulated, disciplined task. At any given point, about two-thirds of us are “restrained” eaters. These are people who are actively trying to diet, who would agree with statements like, “In a typical meal, I’m conscious of trying to restrict the amount of food that I consume.” Mind you, these are not people who are necessarily overweight. Plenty of heavy people are not dieting, plenty of everyone else is at any point. Restrained eaters are actively restricting their food intake. What the studies consistently show is that during stress, people who are normally restrained eaters are more likely than others to become hyperphagic. This makes lots of sense. Things are a bit stressful—corporate thugs have looted your retirement savings, there’s anthrax in the mail, and you’ve realized that you hate how your hair looks. That’s exactly the time when most people decide that, as a coping device, as a means of being nice to themselves during a tough time, they need to ease up on something about which they’re normally pretty regimented. So if you normally force yourself to watch Masterpiece Theater instead of reality TV as some sort of gesture of self-improvement, on goes Survivor XII. And if it’s food intake that you’re normally regimented about, out come the fudge brownies.

Mark Daughhetee, The Sin of Gluttony, oil on silver print, 1985.

So we differ as to whether stress stimulates or inhibits our appetite, and this has something to do with the type and pattern of stressors, how reactive our glucocorticoid system is to stress, and whether eating is normally something that we keep a tight, superegoish lid on. It turns out that we also differ as to how readily we store food away after a stressor. And where in the body we store it. Apples and Pears

Glucocorticoids not only increase appetite but, as an additional means to recover from the stress- response, also increase the storage of that ingested food. Mobilize all that energy during that mad dash across the savanna, and you’re going to have to do a lot of energy storage during your recovery period. In order to have this effect, glucocorticoids trigger fat cells to make an enzyme that breaks down the circulating nutrients into their storage forms, ideal for storing them for next winter. It’s not just any fat cells that glucocorticoids stimulate. Time for one of the great dichotomies revered by fat cell aficionados: fat cells located in your abdominal area, around your belly, are known as “visceral” fat. Fill up those fat cells with fat, without depositing much fat elsewhere in your body, and you take on an “apple” shape. In contrast, fat cells around your rear end form “gluteal” fat. Fill those up preferentially with fat and you take on a “pear” shape, being round-bottomed. The formal way to quantify these different types of fat deposition is to measure the circumference of your waist (which tells you about the amount of abdominal fat) and the circumference of your hips (a measure of gluteal fat). Apples have waists that are bigger than hips, producing a “waist-hip ratio” (WHR) that is bigger than 1.0, while pears have hips that are bigger than waists, producing a WHR that is less than 1.0. It turns out that when glucocorticoids stimulate fat deposition, they do it preferentially in the abdomen, promoting apple-shaped obesity. This even occurs in monkeys. The pattern arises because abdominal fat cells are more sensitive to glucocorticoids than are gluteal fat cells; the former have more receptors that respond to glucocorticoids by activating those fat-storing enzymes. Furthermore, glucocorticoids only do this in the presence of high insulin levels. And once again, this makes sense. What does it mean if you have high glucocorticoid levels and low insulin levels in the bloodstream? As we know from chapter 4, you’re in the middle of a stressor. High glucocorticoids and high insulin? This happens during the recovery phase. Pack away those calories to recover from the grassland sprint. This stimulation of visceral fat deposition by glucocorticoids is not good news. This is because if you have to pack on some fat, you definitely want to become a pear, not an apple. As we saw in the chapter on metabolism, lots of fat is a predictor for Syndrome X. But it turns out that a large WHR is an even better predictor of trouble than being overweight is. Take some extremely applish people and some very peary ones. Match them for weight, and it’s the apples who are at risk for metabolic and cardiovascular disease. Among other reasons, this is probably because fat released from abdominal fat cells more readily finds its way to the liver (in contrast to fat from gluteal fat stores, which gets dispersed more equally throughout the body), where it is converted into glucose, setting you up for elevated blood sugar and insulin resistance. These findings lead to a simple prediction, namely that for the same stressor, if you tend to secrete more glucocorticoids than most, not only are you going to have a bigger appetite post- stressor, you’re going to go apple, preferentially socking away more of those calories in your abdominal fat cells. And that’s precisely what occurs. Epel has studied this in women and men across a range of ages, and she finds that a prolonged glucocorticoid response to novelty is a feature of applish people, not pears.

So with lots of stress, you get cravings for starchy comfort food and you pack it in the abdomen. One final distressing piece of information, based on some fascinating recent work by Mary Dallman from the University of California at San Francisco: consuming lots of those comfort foods and bulking up on abdominal fat are stress-reducers. They tend to decrease the size of the stress-response (both in terms of glucocorticoid secretion and sympathetic nervous system activity). Not only do the Oreos taste good, but by reducing the stress-response, they make you feel good as well. There seems to be a huge number of routes by which obesity can occur—too much or too little of this or that hormone; too much or too little sensitivity to this or that hormone.* But another route appears to involve being the sort of person who secretes too many glucocorticoids, either because of too many stressors, too many perceived stressors, or trouble turning off the stress- response. And thanks to that weird new regulatory loop discovered by Dallman, it appears as if abdominal fat is one route for trying to tone down that overactive stress-response. Bowel Movement and Bowel Movements Thanks to the preceding part of this chapter and to chapter 4, we’ve now sorted out how stress alters what you ingest, how it gets stored and mobilized. We have one last piece to fill in, which is getting food from your mouth to its digested form in your circulation. This is the purview of the gastrointestinal (GI) tract—your esophagus, stomach, small intestines and large intestines (also known as the colon or the bowel). When it comes to your GI tract, there’s no such thing as a free lunch. You’ve just finished some feast, eaten like a hog—slabs of turkey, somebody’s grandma’s famous mashed potatoes and gravy, a bare minimum of vegetables to give a semblance of healthiness, and—oh, why not— another drumstick and some corn on the cob, a slice or two of pie for dessert, ad nauseam. You expect your gut to magically convert all that into a filtrate of nutrients in your bloodstream? It takes energy, huge amounts of it. Muscular work. Your stomach not only breaks down food chemically, it does so mechanically as well. It undergoes systolic contractions: the muscle walls contract violently on one side of your stomach, and hunks of food are flung against the far wall, breaking them down in a cauldron of acids and enzymes. Your small intestines do a snake dance of peristalsis (directional contraction), contracting the muscular walls at the top end in order to squeeze the food downstream in time for the next stretch of muscle to contract. After that, your bowels do the same, and you’re destined for the bathroom soon. Circular muscles called sphincters located at the beginning and end of each organ open and close, serving as locks to make sure that things don’t move to the next level in the system until the previous stage of digestion is complete, a process no less complicated than shuttling ships through the locks of the Panama Canal. At your mouth, stomach, and small intestines, water has to be poured into the system to keep everything in solution, to make sure that the sweet potato pie, or what’s left of it, doesn’t turn into a dry plug. By this time, the action has moved to your large intestines, which have to extract the water and return it to your bloodstream so that you don’t inadvertently excrete all that fluid and desiccate like a prune. All this takes energy, and we haven’t even considered

jaw fatigue. All told, your run-of-the-mill mammals, including us, expend 10 to 20 percent of their energy on digestion. So back to our by-now-familiar drama on the savanna: if you are that zebra being pursued by a lion, you can’t waste energy on your stomach walls doing a rumba. There isn’t time to get any nutritional benefits from digestion. And if you are that lion running after a meal, you haven’t just staggered up from some all-you-can-eat buffet. Digestion is quickly shut down during stress. We all know the first step in that process. If you get nervous, you stop secreting saliva and your mouth gets dry. Your stomach grinds to a halt, contractions stop, enzymes and digestive acids are no longer secreted, your small intestines stop peristalsis, nothing is absorbed. The rest of your body even knows that the digestive tract has been shut down—as we saw two chapters ago, blood flow to your stomach and gut is decreased so that the blood-borne oxygen and glucose can be delivered elsewhere, where they’re needed. The parasympathetic nervous system, perfect for all that calm, vegetative physiology, normally mediates the actions of digestion. Along comes stress: turn off the parasympathetic, turn on the sympathetic, and forget about digestion.* End of stress; switch gears again, and the digestive process resumes. As usual, this all makes wonderful sense for the zebra or the lion. And as usual, it is in the face of chronic stress that diseases emerge instead. Bowels in an Uproar Regardless of how stressful that board meeting or examination is, we’re not likely to soil our pants. Nevertheless, we are all aware of the tendency of immensely terrified people—for example, soldiers amid horrifying battle—to defecate spontaneously. (This reaction is consistent enough that in many states, prisoners are clothed in diapers before an execution.) The logic as to why this occurs is similar to why we lose control of our bladders if we are very frightened, as described in chapter 3. Most of digestion is a strategy to get your mouth, stomach, bile ducts, and so forth to work together to break your food down into its constituent parts by the time it reaches the small intestines. The small intestines, in turn, are responsible for absorbing nutrients out of this mess and delivering them to the bloodstream. As is apparent to most of us, not much of what we eat is actually nutritious, and a large percentage of what we consume is left over after the small intestines pick through it. In the large intestines, the leftovers are converted to feces and eventually exit stage left. Yet again, you sprint across the veld. All that stuff sitting in your large intestines, from which the nutritive potential has already been absorbed, is just dead weight. You have the choice of sprinting for your life with or without a couple of pounds of excess baggage in your bowels. Empty them. The biology of this is quite well understood. The sympathetic nervous system is responsible. At the same time that it is sending a signal to your stomach to stop its contractions and to your small intestine to stop peristalsis, your sympathetic nervous system is actually stimulating muscular

movement in your large intestine. Inject into a rat’s brain the chemicals that turn on the sympathetic nervous system, and suddenly the small intestine stops contracting and the large intestine starts contracting like crazy. But why, to add insult to injury, is it so frequently diarrhea when you are truly frightened? Relatively large amounts of water are needed for digestion, to keep your food in solution as you break it down so that it will be easy to absorb into the circulation when digestion is done. As noted, a job of the large intestine is to get that water back, and that’s why your bowels have to be so long—the leftovers slowly inch their way through the large intestine, starting as a soupy gruel and ending up, ideally, as reasonably dry stool. Disaster strikes, run for your life, increase that large intestinal motility, and everything gets pushed through too fast for the water to be absorbed optimally. Diarrhea, simple as that. Stress and Functional Gastrointestinal Disorders Broadly, there are two types of gastrointestinal disorders. In the first, you feel terrible, something isn’t working right, and the doctors find something wrong. These are “organic” GI disorders. A gaping hole in the wall of your stomach, in other words, a peptic ulcer, counts as there being something demonstrably wrong. We’ll consider ulcers shortly. Out-of-control inflammation of tissue throughout your GI tract, which is what inflammatory bowel disease is, also counts as demonstrably wrong. This disorder will be briefly touched on in chapter 8. But suppose you feel terrible, something isn’t working right, and the docs can’t find a thing wrong. Congratulations, you now have a “functional” GI disorder. These are immensely sensitive to stress. And this is not just the touchy-feely psychologists saying this. Papers about stress and functional GI disorders are even published in tough-guy meat-and-potato scientific journals with names like Gut. The most common functional GI disorder, which will be considered here, is irritable bowel syndrome (IBS), which involves abdominal pain (particularly just after a meal) that is relieved by defecating and symptoms such as diarrhea or constipation, passage of mucus, bloating, and abdominal distention. Despite physicians checking you from every which end, they can’t find anything wrong, which qualifies IBS as a functional disorder. IBS is among the most common of stress-sensitive disorders. Personally, all the major rites of passage in my life have been marked by pretty impressive cases of the runs a few days before—my bar mitzvah, going away to college, my doctoral defense, proposing marriage, my wedding. (Finally, here’s that confessional tone obligatory to successful books these days. Now if I can only name some Hollywood starlet with whom I’ve taken diuretics, this may become a bestseller.) Carefully conducted studies show that major chronic stressors increase the risk of the first symptoms of IBS appearing, and worsen preexisting cases. This makes sense. As we saw, what stress does is increase the contractions in the colon, getting rid of that dead weight. And IBS— also known as “spastic colon”—involves the colon being too contractile, an excellent way of producing diarrhea. (It is not clear why lots of stress-induced contractions of the colon can lead to constipation. As a possible explanation, the stress-induced contractions in the colon are

directional, which is to say, they push the contents of the colon from the small intestinal end to the anus. And if they do that a lot, things get accelerated, resulting in diarrhea. However, in one plausible scenario, with long enough periods of stress, the contractions begin to get disorganized, lose their directionality, so that not much of anything moves toward the anus). So people with IBS are disproportionately likely to be experiencing a lot of stressors. But in addition, IBS can be a disorder of too much gastrointestinal sensitivity to stress. This can be shown in experimental situations, where a person with IBS is subjected to a controlled stressor (keeping her hand in ice water for a while, trying to make sense of two recorded conversations at once, participating in a pressured interview). Contractions in the colon increase in response to these stressors more in IBS patients than in control subjects. Another connection between stress and IBS concerns pain. As we’ll see in chapter 9, stress can blunt the sort of pain you feel in your skin and skeletal muscles while increasing the sensitivity of internal organs like the intestines to pain (something called “visceral” pain). And that is the profile seen in IBS patients—less sensitivity to skin (“cutaneous”) pain, and more visceral pain. Even more support for the stress/IBS link is that people with IBS don’t typically have hypercontractility of their bowels when they are asleep. Gut spasticity is not something that’s going on all the time—only when the person is awake, amid the opportunities to be stressed. What’s the physiology of this gut that is too contractile? As we saw earlier, the sympathetic nervous system is responsible for the increased large intestinal contractions during stress. And as would be expected, people with IBS have overactive sympathetic nervous systems (though it is less clear whether glucocorticoid levels are abnormal in IBS). And just to make the whole process worse, the pain of that gassy, distended, hypersensitive gut can stimulate sympathetic activation even further, making for a vicious circle. So ongoing stress can be closely associated with IBS. Interestingly, traumatic stress early in life (abuse, for example) greatly increases the risk of IBS in adulthood. This implies that childhood trauma can leave an echo of vulnerability, a large intestine that is hyperreactive to stress, long afterward. Animal studies have shown that this occurs. Despite these findings, there is a great deal of resistance to the link between stress and IBS (prompting some semi-irate letters to me from readers of earlier editions of this book). One reason for this is the linkage between IBS and certain personality types. In the cases of depression or anxiety, the connection is solid, but earlier linkages seem pretty suspect. These studies tended to focus on a lot of psychoanalytic gibberish (there, now I’ll get myself into trouble with that crowd)—some hoo-ha about the person being stuck in the anal stage of development, a regression to the period of toilet training where going to the bathroom gained great acclaim and, suddenly, diarrhea was a symbolic reach for parental approval. Or the approval of the doctor as a parental surrogate. Or something or other. I’m not sure how they factored in constipation, but I’m sure they did. Few gastroenterologists take these ideas seriously anymore. However, in less scientific circles, some still cling to these views. It is easy to see how someone suffering from IBS, who has just

managed to clear up the perception that they’re still having some potty-training issues, isn’t enthused about getting fingered for not dealing well with stress. Another reason why the stress/IBS connection is often viewed with suspicion is because there have been many studies that have failed to find a link. Why should this be? First, both the severity of IBS symptoms and the intensity of stressors that someone is experiencing tend to wax and wane over time, and detecting a link between two such fluctuating patterns takes some very fancy statistics. (Typically, a technique called time-series analysis, a subject four classes more advanced than the statistics that most biomedical scientists have sort of learned. When my wife had to do a time-series analysis as part of her doctoral research, it made me nervous just to have a textbook on the subject in the house.) Such waxing and waning of stress and of symptoms is particularly difficult to track because most studies are retrospective (they look at people who already have IBS and ask them to identify stressors in their past) rather than prospective (in which people who do not have a disease are followed to see if stress predicts who is going to get it). The problem here is that people are terribly inaccurate at recalling information about stressors and symptoms that are more than a few months old, a point we’re going to return to often in this book. Moreover, as was mentioned above, the sorts of stressors that can increase the risk of IBS can occur many years prior to the emergence of symptoms, making the link hard to detect even in prospective studies. Finally, “IBS” is probably a hodgepodge of diseases with multiple causes, and stress may be relevant to only some of them, and it takes some additional fancy statistics to detect those folks as a meaningful subset of the whole, instead of as just random noise in the data. At later junctures in this book, we will see other supposed links between stress and some disease, and be in the same quandary—there definitely is a link in some patients, or clinical impressions strongly support a stress-disease link, yet hard-nosed studies fail to show the same thing. As we will see repeatedly, the trouble is that the supposedly hard-nosed studies are often asking a fairly unsophisticated, straightforward question: does stress cause the disease in the majority of sufferers? The far more sophisticated questions to ask are whether stress worsens preexisting disease, whether patterns of symptoms and of stressors fluctuate in parallel over time, and whether these links occur only in a subset of vulnerable individuals. When asked in those ways, the stress-disease link becomes far more solid. Ulcers At last we arrive at the medical problem that started the stress concept on the road to fame and fortune. An ulcer is a hole in the wall of an organ, and ulcers originating in the stomach or in the organs immediately bordering it are termed peptic ulcers. The ones within the stomach are called gastric ulcers; those a bit higher up than the stomach are esophageal, and those at the border of the stomach and the intestine are duodenal (the most common of peptic ulcers).

Photomicrograph of a stomach ulcer.

As will be recalled, peptic ulcers were among the trio of symptoms Selye noted more than sixty years ago when he exposed his rats to nonspecific unpleasantness. Since then, stomach ulcers have emerged as the disorder most recognized by the lay public as a stress-related disease: in this view, you have upsetting thoughts for a long period of time and holes appear in the walls of your stomach. Most clinicians agree that there is a subtype of ulcers that forms relatively rapidly (sometimes over the course of days) in humans who are exposed to immensely stressful crises—hemorrhage, massive infection, trauma due to accident or surgery, burns over large parts of the body, and so on. Such “stress ulcers” can be life threatening in severe cases. But where a lot of contention has appeared has been with the issue of gradually emerging ulcers. This used to be a realm where people, including physicians, would immediately think stress. But a revolution has dramatically changed thinking about ulcers. That revolution came with the discovery in 1983 of a bacterium called Helicobacter pylori. This obscure microorganism was discovered by an obscure Australian pathologist named Robert Warren. He, in turn, interested an even more obscure younger colleague named Barry Marshall, who documented that this bacterium consistently turned up in biopsies of the stomachs of people with duodenal ulcers and stomach inflammation (gastritis). He theorized that it actually caused the inflammation and ulcers, announced this to the (gastroenterological) world at a conference, and was nearly laughed out of the room. Ulcers were caused by diet, genetics, stress—not bacteria. Everyone knew that. And besides, because the stomach is so incredibly acidic, owing to the hydrochloric acid in stomach juices, no bacteria could survive in there. People had known for years that the stomach was a sterile environment, and that any bacteria that might turn up were just due to contamination by some sloppy pathologist.

Marshall showed that the bacteria caused gastritis and ulcers in mice. That’s great, but mice work differently than humans, everyone said. So, in a heroic, soon-to-be-a-movie gesture, he swallowed some Helicobacter bilge and caused gastritis in himself. Still, they ignored Marshall. Eventually, some folks in the field got tired of hearing him go on about the damn bacteria at meetings, decided to do some experiments to prove him wrong, and found that he was absolutely right. Helicobacter pylori turns out to be able to live in the acidic stomach environment, protecting itself by having a structure that is particularly acid-resistant and by wrapping itself in a coat of protective bicarbonate. And this bacterium probably has a lot to do with 85 to 100 percent of ulcers in Western populations (as well as with stomach cancer). Nearly 100 percent of people in the developing world are infected with Helicobacter—it is probably the most common chronic bacterial infection in humans. The bacteria infect cells in the lining of the stomach, causing gastritis, which somehow compromises the ability of those cells lining the duodenum to defend themselves against stomach acids. Boom, under the right conditions, you’ve got a hole in that duodenal wall. Many of the details remain to be sorted out, but the greatest triumph for Marshall and Warren has been the demonstration that antimicrobial drugs, such as antibiotics, turn out to be the greatest things since sliced bread for dealing with duodenal ulcers—they are as good at getting rid of the ulcers as are antacids or antihistamine drugs (the main prior treatments) and, best of all, unlike the aftermath of other treatments, ulcers now stay away (or at least until the next Helicobacter infection). Once everybody in the field got used to the idea of Marshall and Warren being carried around on sedan chairs for their discovery, they embraced Helicobacter with a vengeance. It makes perfect sense, given the contemporary desire of medicine to move toward hard-nosed, reductive models of disease, rather than that wimpy psychosomatic stuff. The Center for Disease Control sent out educational pamphlets to every physician in America, advising them to try to disabuse their patients of the obsolete notion that stress has anything to do with peptic ulcers. Clinicians celebrated at never having again to sit down with their ulcer patients, make some serious eye contact, and ask them how their lives were going. In what one pair of investigators has termed the “Helicobacterization” of stress research on ulcers, the number of papers on stress as a component of the ulcer story has plummeted. Don’t bother with this psychological stuff when we finally have gotten some real science here, complete with a bacterium that’s got its own Latin name. The trouble is that one bacterium can’t be the whole story. For starters, up to 15 percent of duodenal ulcers form in people who aren’t infected with Helicobacter, or with any other known bacterium related to it. More damning, only about 10 percent of the people infected with the bacteria get ulcers. It’s got to be Helicobacter pylori plus something else. Sometimes, the something else is a lifestyle risk factor—alcohol, smoking, skipping breakfast habitually, taking a lot of nonsteroidal anti-inflammatory drugs like aspirin. Maybe the something else is a genetic tendency to secrete a lot of acid or to make only minimal amounts of mucus to protect stomach linings from the acid.

But one of the additional factors is stress. Study after study, even those carried out after the ascendancy of the bacteria, show that duodenal ulceration is more likely to occur in people who are anxious, depressed, or undergoing severe life stressors (imprisonment, war, natural disasters). An analysis of the entire literature shows that somewhere between 30 and 65 percent of peptic ulcers have psychosocial factors (i.e., stress) involved. The problem is that stress causes people to drink and smoke more. So maybe stress increases the risk of an ulcer merely by increasing the incidence of those lifestyle risk factors. But no—after you control for those variables, stress itself still causes a two- to threefold increase in the risk of an ulcer. Helicobacter is relevant to ulcers, but it is only in the context of its interactions with these other factors, including stress. You can show this statistically if you study a zillion ulcer patients. Then, do a fancy mathematical analysis that takes into account bacterial load, lifestyle risk factors, and stress (something aptly called a multivariate analysis). You’ll observe that ulcers can arise if you only have a little bit of one of the factors (bacterial load, stress, or lifestyle risks), so long as you have a lot of one or two of the others. As an example of that, if you expose lab rats to psychological stressors, they get ulcers—but not if they live in a germ-free environment that lacks Helicobacter. So how does stress exacerbate the process of ulcer formation? Some sixty years after Selye first noticed his rats’ ulcers, it is still not quite clear. There are some favorite scenarios, however. Acid Rebound To understand this mechanism, we have to grapple with the grim reality of what bizarre things we are willing to eat and expect our stomachs to digest. The only way that the stomach is going to be able to handle some of this stuff is if it has powerful degradative weapons. The contractions certainly help, but the main weapon is the hydrochloric acid that pours into your stomach from the cells lining it. Hydrochloric acid is immensely acidic; all well and good, but it raises the obvious question of why your stomach is not itself digested by the digestive acids. Eat somebody else’s stomach and your stomach disintegrates it. How do your own stomach walls remain unscathed? Basically, your stomach has to spend a fortune protecting itself. It builds many layers of stomach wall and coats them with thick, soothing mucus that buffers the acid. In addition, bicarbonate is secreted into the stomach to neutralize the acid. This is a wonderful solution, and you happily go about digestion. Along comes a stressful period that lasts months. Your body cuts down on its acid secretion— there are now frequent times when digestion is being inhibited. During this period, your stomach essentially decides to save itself some energy by cutting corners. It cuts back a bit on the constant thickening of the stomach walls, undersecretes mucus and bicarbonate, and pockets the difference. Why not? There isn’t much acid around during this stressful period anyway. End of stressful period; you decide to celebrate by eating a large chocolate cake inscribed for the occasion, stimulate your parasympathetic nervous system, start secreting hydrochloric acid, and…your defenses are down. The walls have thinned, there isn’t as thick a protective mucous layer as there used to be, the bicarbonate is overwhelmed. A couple of repeated cycles of stress and rebound with a bacterial infection that is already compromising the defenses and you’ve got an ulcer.

Suppose you are in the middle of a very stressful period, and you worry that you are at risk for an ulcer. What’s the solution? You could make sure that you remain under stress every second for the rest of your life. You definitely will avoid ulcers caused by hydrochloric acid secretion, but of course you’ll die for a zillion other reasons. The paradox is that, in this scenario, ulcers are not formed so much during the stressor as during the recovery. This idea predicts that several periods of transient stress should be more ulcerative than one long, continuous period, and animal experiments have generally shown this to be the case. Decreased Blood Flow As we know, in an emergency, you want to deliver as much blood as possible to the muscles that are exercising. In response to stress, your sympathetic nervous system diverts blood from the gut to more important places—remember the man with a gunshot wound in the stomach, whose guts would blanch from decreased blood flow every time he became angry or anxious. If your stressor is one that involves a dramatic decrease in blood flow to the gut (for example, following a hemorrhage), it begins to cause little infarcts—small strokes—in your stomach walls, because of lack of oxygen. You develop small lesions of necrotic (dead) tissue, which are the building blocks of ulcers. This condition probably arises for at least two reasons. First, with decreased blood flow, less of the acid that accumulates is being flushed away. The second reason involves another paradoxical piece of biology. We all obviously need oxygen and would turn an unsightly blue without it. However, running your cells on oxygen can sometimes produce an odd, dangerous class of compounds called oxygen radicals. Normally, another group of compounds (free radical quenchers, or scavengers) dispose of these villains. There is some evidence, however, that during periods of chronic stress, when blood flow (and thus oxygen delivery) to the gut decreases, your stomach stops making the scavengers that protect you from the oxygen radicals. Fine for the period of stress (since the oxygen radicals are also in shorter supply); it’s a clever way to save energy during a crisis. At the end of stress, however, when blood flow chock-full of oxygen resumes and the normal amount of oxygen radicals is generated, the stomach has its oxidative pants down. Without sufficient scavengers, the oxygen radicals start killing cells in the stomach walls; couple that with cells already in trouble thanks to bacterial infection and you’ve got an ulcer. Note how similar this scenario is to the acid-rebound mechanism: in both cases, the damage occurs not during the period of stress but in its aftermath, and not so much because stress increases the size of an insult (for example, the amount of acid secreted or the amount of oxygen radicals produced), but because, during the stressful emergency, the gut scrimps on defenses against such insults. Immune Suppression Helicobacter as a bacterium triggers your immune system into trying to defend against it.* As you will soon learn in sickening detail (chapter 8), chronic stress suppresses immunity, and in this scenario, lowered immune defenses equals more Helicobacters reproducing happily.

Insufficient Amounts of Prostaglandins In this scenario, micro-ulcers begin now and then in your gut, as part of the expected wear and tear on the system. Normally your body can repair the damage by secreting a class of chemicals called prostaglandins, thought to aid the healing process by increasing blood flow through the stomach walls. During stress, however, the synthesis of these prostaglandins is inhibited by the actions of glucocorticoids. In this scenario, stress does not so much cause ulcers to form as impair your body’s ability to catch them early and repair them. It is not yet established how often this is the route for ulcer formation during stress. (Aspirin also inhibits prostaglandin synthesis, which is why aspirin can aggravate a bleeding ulcer.) Stomach Contractions For unknown reasons, stress causes the stomach to initiate slow, rhythmic contractions (about one per minute); and for unknown reasons, these seem to add to ulcer risk. One idea is that during the contractions, blood flow to the stomach is disrupted, causing little bursts of ischemia; there’s not much evidence for this, however. Another idea is that the contractions mechanically damage the stomach walls. The jury is still out on that mechanism. Most of these mechanisms are pretty well documented routes by which ulcers can form; of those credible mechanisms, most can occur during at least certain types of stressors. More than one mechanism may occur simultaneously, and people seemingly differ as to how likely each mechanism is to occur in their gut during stress, and how likely it is to interact with bacterial infection. Additional mechanisms for stress’s role in ulcer formation will no doubt be discovered, but for the moment these should be quite sufficient to make anyone sick. Peptic ulcers are what the physician Susan Levenstein, the wittiest person on earth writing about gastroenterology, has termed “the very model of a modern etiology.”* Stress doesn’t cause peptic ulcers to form. But it makes the biological villains that do cause ulcers to form more likely to occur, or more virulent, or impairs your ability to defend yourself against those villains. This is the classic interaction between the organic (bacteria, viruses, toxins, mutations) and the psychogenic components of disease. Chapter 6

Dwarfism and the Importance of Mothers It still surprises me that organisms grow. Maybe I don’t believe in biology as much as I claim. Eating and digesting a meal seems very real. You put a massive amount of something or other in your mouth, and, as a result, all sorts of tangible things happen—your jaw gets tired, your stomach distends, eventually something comes out the other end. Growth seems pretty tangible, too. Long bones get longer, kids weigh more when you heft them. My difficulty is with the steps that connect digestion with growth. I know how it works; my university even allows me to teach impressionable students about it. But it just seems implausible. Someone ate a mountain of spaghetti, salad, garlic bread, and two slices of cake for dessert—and that has been transformed and is now partially inside this test tube of blood? And somehow it’s going to be reconstructed into bone? Just think, your femur is made up of tiny pieces of your mother’s chicken potpie that you ate throughout your youth. Ha! You see, you

don’t really believe in the process either. Maybe we’re too primitive to comprehend the transmogrification of material. How We Grow Nevertheless, growth does occur as a result of eating. And in a kid, it’s not a trivial process. The brain gets bigger, the shape of the head changes. Cells divide, grow in size, and synthesize new proteins. Long bones lengthen as cartilaginous cells at the ends of bones migrate into the shaft and solidify into bone. Baby fat melts away and is replaced by muscle. The larynx thickens and the voice deepens, hair grows in all sorts of unlikely places on the body, breasts develop, testes enlarge. From the standpoint of understanding the effects of stress on growth, the most important feature of the growth process is that, of course, growth doesn’t come cheap. Calcium must be obtained to build bones, amino acids are needed for all that protein synthesis, fatty acids build cell walls— and it’s glucose that pays for the building costs. Appetite soars, and nutrients pour in from the intestines. A large part of what various hormones do is to mobilize the energy and the material needed for all these civic expansion projects. Growth hormone dominates the process. Sometimes it works directly on cells in the body—for example, growth hormone helps to break down fat stores, flushing them into the circulation so they can be diverted to the growing cells. Alternatively, sometimes growth hormone must first trigger the release of another class of hormones called somatomedins, which actually do the job, such as promoting cell division. Thyroid hormone plays a role, promoting growth hormone release, making bones more responsive to somatomedins. Insulin does something similar as well. The reproductive hormones come into play around puberty. Estrogen promotes the growth of long bones, both by acting directly on bone and by increasing growth hormone secretion. Testosterone does similar things to long bones and, in addition, enhances muscle growth. Adolescents stop growing when the ends of the long bones meet and begin to fuse, but for complex reasons, testosterone, by accelerating the growth of the ends of long bones, can actually speed the cessation of growth. Thus, pubescent boys given testosterone will, paradoxically, wind up having their adult stature blunted a bit. Conversely, boys castrated before puberty grow to be quite tall, with lanky bodies and particularly long limbs. Opera history buffs will recognize this morphology, as castrati were famed for this body shape. Neurotic Parents: Beware! It is time to look at how stress disrupts normal development. As we’ll see, this not only involves impairing skeletal growth (that is, how tall you grow to be), but also how stress early in life can alter your vulnerability to disease throughout your lifetime. Now, before I launch into this, I have to issue a warning to anyone who is a parent, or who plans to be a parent, or who had parents. There’s nothing like parenthood to make you really neurotic, as you worry about the consequences of your every act, thought, or omission. I have young children, and here are some of the heinous things that my wife and I have done to irreparably harm them: there was the time we were desperate to placate them about something and allowed

them to eat some sugar-bomb breakfast cereal we’d normally ban; then there was the loud concert we went to when our firstborn was a third-trimester fetus, causing him to kick throughout, no doubt in pained protest; and there was the time we messed up with our otherwise ceaseless vigilance and allowed ten seconds of a violent cartoon to show on the television while we fumbled with the Kumbaya-esque video we were attempting to insert. You only want perfection for the ones you love beyond words, so you get nutsy at times. This section will make you nutsier. So keep this warning in mind, a point I will return to at the end. Prenatal Stress What is childhood about? It is a time when you make assessments about the nature of the world. For example, “If you let go of something, it falls down, not up.” Or, “If something is hidden underneath something else, it still exists.” Or, ideally, “Even if Mommy disappears for a while, she will come back because Mommy always comes back.” Often, these assessments shape your view of the world forever. For example, as will be discussed in chapter 14, if you lose a parent to death while you are a child, your risk of major depression has increased for the rest of your life. I will suggest that this arises from having learned at a premature age a deep emotional lesson about the nature of life, namely, that this is a world in which awful things can happen over which you have no control. It turns out that during development, beginning with fetal life, your body is also learning about the nature of the world and, metaphorically, making lifelong decisions about how to respond to the outside world. And if development involves certain types of stressors, some of these “decisions” cause a lifelong increase in the risk of certain diseases. Consider a female who is pregnant during a famine. She’s not getting enough calories, nor is her fetus. It turns out that during the latter part of pregnancy, a fetus is “learning” about how plentiful food is in that outside world, and a famine winds up “teaching” it that, jeez, there’s not a whole lot of food out there, better store every smidgen of it. Something about the metabolism of that fetus shifts permanently, a feature called metabolic “imprinting” or “programming.” Forever after, that fetus will be particularly good at storing the food it consumes, at retaining every grain of precious salt from the diet. Forever after, that fetus develops what has been termed a “thrifty” metabolism. And what are the consequences of that? Suddenly we find ourselves back in the middle of chapters 3 and 4. Everything else being equal throughout life, even late in life, that organism is more at risk for hypertension, obesity, adult-onset diabetes, and cardiovascular disease. Remarkably, things work precisely this way in rats, pigs, and sheep. And humans as well. The most dramatic and most cited example concerns the Dutch Hunger Winter at the end of World War II. The occupying Nazis were being pushed back on all fronts, the Dutch were trying to aid the Allies coming to liberate them, and, as punishment, the Nazis cut off all food transport. For a demarcated season, the Dutch starved. People consumed less than 1,000 calories a day, were

reduced to eating tulip bulbs, and 16,000 people starved to death. Fetuses, going about their lifelong metabolic programming, learned some severe lessons about food availability during that winter of starvation. The result is a cohort of people with thrifty metabolisms and increased risks of Metabolic syndrome a half-century later. Seemingly, different aspects of metabolism and physiology get programmed at different points of fetal development. If you were a first-trimester fetus during the famine, that programs you for a greater risk of heart disease, obesity, and an unhealthy cholesterol profile, whereas if you were a second- or third-trimester fetus, that programs you for a greater diabetes risk. The key to this phenomenon seems to be not only that you were undernourished as a fetus, but that after birth you had plenty of food and were able to recover from the deprivation quickly. Thus, from early in childhood, you not only were highly efficient at storing nutrients, but had access to plentiful nutrients.* So avoid starving a fetus while you’re pregnant. But this phenomenon also applies to less dramatic situations. Within the normal range of birth weights, the lower the weight of a baby (when adjusted for body length), the greater the risk of those Metabolic syndrome problems in adulthood. Even after you control for adult body weight, low birth weight still predicts an increased risk of diabetes and hypertension. These are big effects. When you compare those who were heaviest versus lightest at birth, you see an approximate eight-fold difference in the risk of pre-diabetes, and about an eighteen-fold difference in the risk of Metabolic syndrome. Among both men and women, compare those whose birth weights were in the lowest 25 percent versus those in the highest 25 percent, and the former have a 50 percent higher rate of death from heart disease. This relationship between fetal nutritional events and lifelong risks of metabolic and cardiovascular disease was first described by the epidemiologist David Barker of Southampton Hospital in England, and now goes by the name Fetal Origins of Adult Disease (FOAD). And we’re not done with this yet. Starvation is clearly a stressor, raising the question of whether the metabolic programming occurs because of the nutritional consequences of the shortage of calories, and /or because of the stressfulness of the shortage of calories. Asked another way, do non-nutritional stressors during pregnancy also induce FOAD-like effects? The answer is, yes. An extensive literature, stretching back decades, shows that stressing a female rat in any number of ways while she is pregnant will cause lifelong changes in the physiology of her offspring. Predictably, one set of changes involves glucocorticoid secretion. Once again, think of the fetal body “learning” about the outside world, this time along the lines of, “How stressful is it out there?” Fetuses can monitor signals of stress from the mother, insofar as glucocorticoids readily pass through to the fetal circulation, and lots of glucocorticoids “teach” the fetus that it is indeed a stressful world out there. The result? Be prepared for that stressful world: tend toward secreting excessive amounts of glucocorticoids. Prenatally stressed rats grow into adults with elevated glucocorticoid levels—depending on the study, elevated basal levels, a larger stress-response, and/or a sluggish recovery from the stress-response. The lifelong programming seems to be due

to a permanent decrease in the number of receptors for glucocorticoids in one part of the brain. The brain region is involved in turning off this stress-response by inhibiting CRH release. Fewer glucocorticoid receptors there mean less sensitivity to the hormone’s signal, which means less effective reining in of subsequent glucocorticoid secretion. The result is a lifelong tendency toward elevated levels. Is it the glucocorticoid secretion by the stressed pregnant female that gives rise to these permanent changes in the offspring? Seemingly yes—the effect can be replicated in a number of species, including nonhuman primates, by injecting the pregnant female with high glucocorticoid levels, instead of stressing her. A smaller but fairly solid literature shows that prenatal stress programs humans for higher glucocorticoid secretion in adulthood as well. In these studies, low birth weight (corrected for body length) is used as a surrogate marker for stressors during fetal life, and the lower the birth weight, the higher the basal glucocorticoid levels in adults ranging from age twenty to seventy; this relationship becomes even more pronounced when low birth weight is coupled with premature birth.* The excessive glucocorticoid exposure of a stressful fetal life seems to contribute to the lifelong increase in the risk of Metabolic syndrome as well. As evidence, if you expose a fetal rat, sheep, or nonhuman primate to lots of synthetic glucocorticoids during late gestational life (by injecting the mother with them), that fetus will be more at risk for the symptoms of Metabolic syndrome as an adult. How does this arise? A plausible sequence is that the prenatal exposure to high glucocorticoid levels leads to the elevated glucocorticoid levels in adulthood, which increases the risk of Metabolic syndrome. Those readers who have memorized the book so far will have no trouble recalling exactly how an excess of glucocorticoids in adulthood can increase the odds of obesity, insulin-resistant diabetes, and hypertension. Despite those potential links, the elevated glucocorticoid levels in adulthood are probably only one of the routes linking prenatal stress with the adult Metabolic syndrome. So now we have hypertension, diabetes, cardiovascular disease, obesity, and glucocorticoid excess in this picture. Let’s make it worse. How about the reproductive system? An extensive literature shows that if you stress pregnant rats, you “demasculinize” the male fetuses. They are less sexually active as adults, and have less developed genitals. As we will see in the next chapter, stress decreases testosterone secretion, and it seems to do so in male fetuses as well. Furthermore, glucocorticoids and testosterone have similar chemical structures (they are both “steroid” hormones), and a lot of glucocorticoids in a fetus can begin to gum up and block receptors for testosterone, making it impossible for the testosterone to have its effects. More FOADish problems. Seriously stress a pregnant rat and her offspring will grow up to be anxious. Now, how do you tell if a rat is anxious? You put it in a new (and thus, by definition, scary) environment; how long does it take for it to explore? Or take advantage of the fact that rats, being nocturnal, don’t like bright lights. Take a hungry rat and put some food in the middle of a brightly lit cage; how long until the rat goes for the food? How readily can the rat learn in a novel setting, or socially interact with new rats? How much does the rat defecate in a novel setting?* Prenatally stressed rats, as adults, freeze up when around bright lights, can’t learn in

novel settings, defecate like crazy. Sad. As we will see in chapter 15, anxiety revolves around a part of the brain called the amygdala, and prenatal stress programs the amygdala into a lifelong profile that has anxiety written all over it. The amygdala winds up with more receptors for (that is, more sensitivity to) glucocorticoids, more of a neurotransmitter that mediates anxiety, and fewer receptors for a brain chemical that reduces anxiety.* Does prenatal stress in humans make for anxious adults? It’s difficult to study this in humans, in that it is hard to find mothers who are anxious during pregnancy, or anxious while their child is growing up, but not both. So there’s not a huge amount of evidence for this happening in humans. Finally, chapter 10 will review how an excess of stress can have bad effects on the brain, particularly in the developing brain. Prenatally stressed rodents grow up to have fewer connections between the neurons in a key area of the brain involved in learning and memory, and have more impairments of memory in old age, while prenatally stressed nonhuman primates have memory problems and form fewer neurons as well. The human studies have been very hard to carry out for reasons similar to that of those examining whether prenatal stress increases the risk of anxiety. With that caveat, a number of studies have shown that such stress results in children born with a smaller head circumference (which certainly fits in with the picture of being underweight in general). However, it’s not clear whether head circumference at birth predicts how many academic degrees the kid is going to have after her name thirty years later. One final piece of the FOAD story is so intrinsically fascinating that it made me stop thinking like a worried parent for a few minutes and instead I just marveled at biology. Suppose you have a fetus exposed to lots of stress, say, malnutrition, and who thus programs a thrifty metabolism. Later, as an adult, she gets pregnant. She consumes normal amounts of food. Because she has that thrifty metabolism, is so good at storing away nutrients in case that fetal famine ever comes back again, her body grabs a disproportionate share of the nutrients in her bloodstream for herself. In other words, amid consuming an average amount of food, her fetus gets a less than average share of it, producing mild malnutrition. And thus programs a milder version of a thrifty metabolism. And when that fetus eventually becomes pregnant…. In other words, these FOADish tendencies can be transmitted across generations, without the benefit of genes. It’s not due to shared genes, but to shared environment, namely, the intimately shared blood supply during gestation. Amazing. This is precisely what is seen in the Dutch Hunger Winter population, in that their grandchildren are born with lower than expected birth weights. This is seen in other realms as well. Pick some rats at random and feed them on a diet that will make them become obese at the time of pregnancy. As a result, their offspring, despite being fed a normal diet, have an increased risk of obesity. As will their grandkids. Similarly, in humans, having insulin-resistant diabetes while pregnant increases the risk of the disorder in your offspring, after controlling for weight. Wait a second—going through a famine means less nutrients in the bloodstream, while having insulin-resistant diabetes means more. How can they produce the same thrifty metabolism in the fetus? Remember, you have elevated levels of glucose in the bloodstream in the case of diabetes because you can’t store the stuff. Recall a one-sentence factoid from chapter 4—when overstuffed fat cells begin to become insulin-resistant, they release hormones that urge other fat

cells and muscle to do the same. And those hormones get into the fetal circulation. So you have Mom, who is insulin-resistant because she has too much energy stored away, releasing hormones that make the normal-weight fetus bad at energy storage as well…and the fetus winds up underweight and with a thrifty metabolic view of the world. So expose a fetus to lots of glucocorticoids and you are increasing its risk for obesity, hypertension, cardiovascular disease, insulin-resistant diabetes, maybe reproductive impairments, maybe anxiety, and impaired brain development. And maybe even setting up that fetus’s eventual offspring for the same. Aren’t you sorry now that the two of you had that argument over whether to videotape the delivery? Now on to the next realm of worries. Postnatal Stress The obvious question to begin this section is, does postnatal stress have lifelong adverse effects on development as well? Of course it can. To begin, what’s the most stressful thing that could happen to an infant rat? Being deprived of its mother (while still receiving adequate nutrition). Work done by Paul Plotsky at Emory University shows that maternal deprivation causes similar consequences in a rat as prenatal stress: increased levels of glucocorticoids during stress and an impaired recovery at the end of stress. More anxiety, and the same sorts of changes in the amygdala as were seen in prenatally stressed adults. And impaired development of a part of the brain relevant to learning and memory. Separate an infant rhesus monkey from its mother and it grows up to have elevated glucocorticoid levels as well. How about something more subtle? What if your rat mom is around but is simply inattentive? Michael Meaney of McGill University has looked at the lifelong consequences for rats of having had a highly attentive or highly inattentive mother. What counts as attentiveness? Grooming and licking. Infants whose mothers groomed and licked the least produced kids who were milder versions of rats who were maternally deprived as infants, with elevated glucocorticoid levels.*

What are the consequences of childhood stress for disease vulnerability during adulthood in humans? This has been studied only minimally, which is not surprising, given how difficult such studies are. A number of studies, mentioned earlier, show that loss of a parent to death during childhood increases the lifelong risk of depression. Another, discussed in chapter 5, shows that early trauma increases the risk of irritable bowel syndrome in adulthood, and similar animal studies show that early stress produces large intestines that contract to abnormal extents in response to stress. Though the subject is still poorly studied, childhood stress may produce the building blocks for the sort of adult diseases we’ve been considering. For example, when you examine children who had been adopted more than a year before from Romanian orphanages, the longer the child spent in the orphanage, the higher the resting glucocorticoid levels.* Similarly, children who have been abused have elevated glucocorticoid levels, and decreased size and activity in the most highly evolved part of the brain, the frontal cortex. Skeletal Growth and Stress Dwarfism How about the effects of stress on how tall you grow (often referred to as skeletal growth)? Skeletal growth is great when you are a ten-year-old lying in bed at night with a full belly. However, it’s the usual scenario of it not making a whole lot of sense when you’re sprinting from a lion. If there is no time to derive any advantages from digesting your meal at that point, there certainly isn’t time to get any benefit from growth. To understand the process by which stress inhibits skeletal growth, it helps to begin with extreme cases. A child of, say, eight years is brought to a doctor because she has stopped growing. There are none of the typical problems—the kid is getting enough food, there is no apparent disease, she has no intestinal parasites that compete for nutrients. No one can identify an organic cause of her problem; yet she doesn’t grow. In many such cases, there turns out to be something dreadfully stressful in her life—emotional neglect or psychological abuse. In such circumstances, the syndrome is called stress dwarfism, or psychosocial or psychogenic dwarfism.* A question now typically comes to mind among people who are below average height. If you are short, yet didn’t have any obvious chronic diseases as a kid and can recall a dreadful period in your childhood, are you a victim of mild stress dwarfism? Suppose one of your parents had a job necessitating frequent moves, and every year or two throughout childhood you were uprooted, forced to leave your friends, moved off to a strange school. Is this the sort of situation associated with psychogenic dwarfism? Definitely not. How about something more severe? What about an acrimonious divorce? Stress dwarfism? Unlikely. The syndrome is extremely rare. These are the kids who are incessantly harassed and psychologically terrorized by the crazy stepfather. These are the kids who, when the police and the social workers break down the door, are discovered to have been locked in a closet for extended periods, fed a tray of food slipped under the door. These are the products of vast, grotesque psychopathology. And they appear in every endocrinology textbook, standing nude in front of a growth chart. Stunted little kids, years behind their expected height, years behind in

mental development, bruised, with distorted, flinching postures, haunted, slack facial expressions, eyes masked by the obligatory rectangles that accompany all naked people in medical texts. And all with stories to take your breath away and make you wonder at the potential sickness of the human mind. Invariably, on the same page in the text is a surprising second photo—the same child a few years later, after having been placed in a different environment (or, as one pediatric endocrinologist termed it, having undergone a “parentectomy”). No bruises, maybe a tentative smile. And a lot taller. So long as the stressor is removed before the child is far into puberty (when the ends of the long bones fuse together and growth ceases), there is the potential for some degree of “catch-up” growth (although shortness of stature and some degree of stunting of personality and intellect usually persist into adulthood). Despite the clinical rarity of stress dwarfism, instances pop up throughout history. One possible case arose during the thirteenth century as the result of an experiment by that noted endocrinologist, King Frederick II of Sicily. It seems that his court was engrossed in philosophic disputation over the natural language of humans. In an attempt to resolve the question, Frederick (who was apparently betting on Hebrew, Greek, or Latin) came up with a surprisingly sophisticated idea for an experiment. He commandeered a bunch of infants and had each one reared in a room of its own. Every day someone would bring the child food, fresh blankets, and clean clothes, all of the best quality. But they wouldn’t stay and play with the infant, or hold it— too much of a chance that the person would speak in the child’s presence. The infants would be reared without human language, and everyone would get to see what was actually the natural language of humans. Of course, the kids did not spontaneously burst out of the door one day reciting poetry in Italian or singing opera. The kids didn’t burst out of the door at all. None of them survived. The lesson is obvious to us now—optimal growth and development do not merely depend on being fed the right number of calories and being kept warm. Frederick “laboured in vain, for the children could not live without clappings of hands and gestures and gladness of countenance and blandishments,” reported the contemporary historian Salimbene. It seems quite plausible that these kids, all healthy and well fed, died of a nonorganic failure to thrive.*

A child suffering from stress dwarfism: changes in appearance during hospitalization (left to right).

Another study that winds up in half the textbooks makes the same point, if more subtly. The subjects of the “experiment” were children reared in two different orphanages in Germany after World War II. Both orphanages were run by the government; thus there were important controls in place—the kids in both had the same general diet, the same frequency of doctors’ visits, and so on. The main identifiable difference in their care was the two women who ran the orphanages. The scientists even checked them, and their description sounds like a parable. In one orphanage was Fräulein Grun, the warm, nurturing mother figure who played with the children, comforted them, and spent all day singing and laughing. In the other was Fräulein Schwarz, a woman who was clearly in the wrong profession. She discharged her professional obligations, but minimized her contact with the children; she frequently criticized and berated them, typically among their assembled peers. The growth rates at the two orphanages were entirely different. Fräulein Schwarz’s kids grew in height and weight at a slower pace than the kids in the other orphanage. Then, in an elaboration that couldn’t have been more useful if it had been planned by a scientist, Fräulein Grun moved on to greener pastures and, for some bureaucratic reason, Fräulein Schwarz was transferred to the other orphanage. Growth rates in her former orphanage promptly increased; those in her new one decreased.

Growth rates in the two German orphanages. During the first 26 weeks of the study,

growth rates in Orphanage A, under the administration of Fräulein Grun, were much greater than those in Orphanage B, with the stern Fräulein Schwarz. At 26 weeks (vertical line), Fräulein Grun left Orphanage A and was replaced by Fräulein Schwarz. The rate of growth in that orphanage promptly slowed; growth in Orphanage B, now minus the stern

Fräulein Schwarz, accelerated and soon surpassed that of Orphanage A. A fascinating elaboration emerges from the fact that Schwarz was not completely heartless, but had a

subset of children who were her favorites (Curve C), whom she had transferred with her. A final and truly disturbing example comes to mind. If you ever find yourself reading chapter after chapter about growth endocrinology (which I don’t recommend), you will note an occasional odd reference to Peter Pan—perhaps a quotation from the play, or a snide comment about Tinker Bell. I’d long noted the phenomenon and finally, in a chapter in one textbook, I found the explanation for it.

The chapter reviewed the regulation of growth in children and the capacity for severe psychological stress to trigger psychogenic dwarfism. It gave an example that occurred in a British Victorian family. A son, age thirteen, the beloved favorite of the mother, is killed in an accident. The mother, despairing and bereaved, takes to her bed in grief for years afterward, utterly ignoring her other, six-year-old son. Horrible scenes ensue. For example, the boy, on one occasion, enters her darkened room; the mother, in her delusional state, briefly believes it is the dead son—“David, is that you? Could that be you?”—before realizing: “Oh, it is only you.” Growing up, being “only you.” On the rare instances when the mother interacts with the younger son, she repeatedly expresses the same obsessive thought: the only solace she feels is that David died when he was still perfect, still a boy, never to be ruined by growing up and growing away from his mother. The younger boy, ignored (the stern, distant father seemed to have been irrelevant to the family dynamics), seizes upon this idea; by remaining a boy forever, by not growing up, he will at least have some chance of pleasing his mother, winning her love. Although there is no evidence of disease or malnutrition in his well-to-do family, he ceases growing. As an adult, he is just barely five feet in height, and his marriage is unconsummated. And then the chapter informs us that the boy became the author of the much-beloved children’s classic—Peter Pan. J. M. Barrie’s writings are filled with children who didn’t grow up, who were fortunate enough to die in childhood, who came back as ghosts to visit their mothers. The Mechanisms Underlying Stress Dwarfism Stress dwarfism involves extremely low growth hormone levels in the circulation. The sensitivity of growth hormone to psychological state has rarely been shown as clearly as in a paper that followed a single child with stress dwarfism. When brought to the hospital, he was assigned to a nurse who spent a great deal of time with him and to whom he became very attached. Row A in the table below shows his physiological profile upon entering the hospital: extremely low growth hormone levels and a low rate of growth. Row B shows his profile a few months later, while still in the hospital: growth hormone levels have more than doubled (without his having received any synthetic hormones), and the growth rate has more than tripled. The stress dwarfism is not a problem of insufficient food—the boy was eating more at the time he entered the hospital than a few months later, when his growth resumed.

A Demonstration of the Sensitivity of Growth to Emotional State

Source: From Saenger and colleagues, 1977. Growth hormone is measured in nanograms of the hormone per milliliter of blood following insulin stimulation; growth is expressed as centimeters per 20 days. Food intake is expressed in calories consumed per day. Row C profiles the period when the nurse went on a three-week vacation. Despite the same food intake, growth hormone levels and growth plummeted. Finally, Row D shows the boy’s profile after the nurse returned from vacation. This is extraordinary. To take a concrete, nuts and bolts feature of growth, the rate at which this child was depositing calcium in his long bones could be successfully predicted by his proximity to a loved one. You can’t ask for a clearer demonstration that what is going on in our heads influences every cell in our bodies. Why do growth hormone levels decline in these kids? Growth hormone is secreted by the pituitary gland, which in turn is regulated by the hypothalamus in the brain (see chapter 2). The hypothalamus controls the release of growth hormone through the secretion of a stimulatory hormone and an inhibitory one, and it looks as if stress dwarfism involves too much release of the inhibitory hormone. Stress-induced overactivity of the sympathetic nervous system may play some role in this. Furthermore, the body becomes less responsive to what little growth hormone is actually secreted. Therefore, even administering synthetic growth hormone doesn’t necessarily solve the growth problem. Some stress dwarfism kids have elevated glucocorticoid levels, and the hormone blunts growth hormone release as well as responsiveness of the body to growth hormone. Kids with stress dwarfism also have gastrointestinal problems, in that they’re impaired at absorbing nutrients from their intestines. This is probably because of the enhanced activity of their sympathetic nervous systems. As discussed in chapter 5, this will halt the release of various digestive enzymes, stop the muscular contractions of the stomach and intestinal walls, and block nutrient absorption. This tells us something about which stress hormones shut down growth. But what is it about being reared under pathological conditions that causes a failure of skeletal growth? Cynthia Kuhn and Saul Schanberg of Duke University and, in separate studies, Myron Hofer of the New York State Psychiatric Institute, have examined that question in infant rats separated from their mothers. Is it the smell of Mom that would normally stimulate growth? Is it something in her milk? Do the rats get chilly without her? Is it the rat lullabies that she sings? You can imagine the various ways scientists test for these possibilities—playing recordings of Mom’s vocalizations, pumping her odor into the cage, seeing what substitutes for the real thing. It turns out to be touch, and it has to be active touching. Separate a baby rat from its mother and its growth hormone levels plummet. Allow it contact with its mother while she is anesthetized, and growth hormone is still low. Mimic active licking by the mother by stroking the rat pup in the proper pattern, and growth normalizes. In a similar set of findings, other investigators have observed that handling neonatal rats causes them to grow faster and larger. The same seems to apply in humans, as demonstrated in a classic study. Tiffany Field of the University of Miami School of Medicine, along with Schanberg, Kuhn, and others, performed an incredibly simple experiment that was inspired both by the rat research and by the history of the

dismal mortality rates in orphanages and pediatric wards, as discussed earlier. Studying premature infants in neonatology wards, they noted that the premature kids, while pampered and fretted over and maintained in near-sterile conditions, were hardly ever touched. So Field and crew went in and started touching them: fifteen-minute periods, three times a day, stroking their bodies, moving their limbs. It worked wonders. The kids grew nearly 50 percent faster, were more active and alert, matured faster behaviorally, and were released from the hospital nearly a week earlier than the premature infants who weren’t touched. Months later, they were still doing better than infants who hadn’t been touched. If this were done in every neonatology ward, this would not only make for a lot more healthy infants, but would save approximately a billion dollars annually. It’s rare that the highest technology of medical instrumentation—MRI machines, artificial organs, pacemakers—has the potential for as much impact as this simple intervention.

Pigtailed macaque mother and infant.

Touch is one of the central experiences of an infant. We readily think of stressors as consisting of various unpleasant things that can be done to an organism. Sometimes a stressor can be the failure to provide something essential, and the absence of touch is seemingly one of the most marked developmental stressors that we can suffer. Stress and Growth Hormone Secretion in Humans The pattern of growth hormone secretion during stress differs in humans from rodents, and the implications can be fascinating. But the subject is a tough one, not meant for the fainthearted. So feel free to go to the bathroom now and come back at the next commercial break. When a rat is first stressed, growth hormone levels in the circulation decline almost immediately. If the stressor continues, growth hormone levels remain depressed. And as we have seen, in humans major and prolonged stressors cause a decrease in growth hormone levels as well. The weird thing is that during the period immediately following the onset of stress, growth hormone

levels actually go up in humans and some other species. In these species, in other words, short- term stress actually stimulates growth hormone secretion for a time. Why? As was mentioned, growth hormone has two classes of effects. In the first, it stimulates somatomedins to stimulate bone growth and cell division. This is the growing part of the story. But in addition, growth hormone works directly on fat cells, breaking down fat stores and flushing them into the circulation. This is the energy for the growth. In effect, growth hormone not only runs the construction site for the new building, but arranges financing for the work as well. Now that business about breaking down stored energy and flushing it into the circulation should sound familiar—that’s precisely what glucocorticoids, epinephrine, norepinephrine, and glucagon are doing during that sprint from the lion. So those direct growth hormone actions are similar to the energy mobilization that occurs during stress, while the somatomedin-mediated growth hormone actions are not what you want to be doing. During stress, therefore, it is adaptive to secrete growth hormone insofar as it helps to mobilize energy, but a bad move to secrete growth hormone insofar as it stimulates an expensive, long-term project like growth. As noted, during stress, somatomedin secretion is inhibited, as is the sensitivity of the body to that hormone. This is perfect—you secrete growth hormone during stress and still get its energy- mobilizing effects, while blocking its more explicit growth-promoting effects. To extend the metaphor used earlier, growth hormone has just taken out cash from the bank, aiming to fund the next six months of construction; instead, the cash is used to solve the body’s immediate emergency. Given this clever solution—spare the growth hormone, block the somatomedins—why should growth hormone levels decline at all during stress (whether immediately, as in the rat, or after a while, as in humans)? It is probably because the system does not work perfectly—somatomedin action is not completely shut down during stress. Therefore, the energy-mobilizing effects of growth hormone might still be used for growth. Perhaps the timing of the decline of growth hormone levels in each species is a compromise between the trait triggered by the hormone that is good news during stress and the trait that is undesirable. What impresses me is how careful and calculating the body has to be during stress in order to coordinate hormonal activities just right. It must perfectly balance the costs and benefits, knowing exactly when to stop secreting the hormone. If the body miscalculates in one direction and growth hormone secretion is blocked too early, there is relatively less mobilization of energy for dealing with the stressor. If it miscalculates in the other direction and growth hormone secretion goes on too long, stress may actually enhance growth. One oft-quoted study suggests that the second type of error occurs during some stressors. In the early 1960s, Thomas Landauer of Dartmouth and John Whiting of Harvard methodically studied the rites of passages found in various non-Western societies around the world; they wanted to know whether the stressfulness of the ritual was related to how tall the kids wound up being as adults. Landauer and Whiting classified cultures according to whether and when they subjected their children to physically stressful development rites. Stressful rites included piercing

the nose, lips, or ears; circumcision, inoculation, scarification, or cauterization; stretching or binding of limbs, or shaping the head; exposure to hot baths, fire, or intense sunlight; exposure to cold baths, snow, or cold air; emetics, irritants, and enemas; rubbing with sand, or scraping with a shell or other sharp object. (And you thought having to play the piano at age ten for your grandmother’s friends was a stressful rite of passage.) Reflecting the anthropological tunnel vision of the time, Landauer and Whiting only studied males. They examined eighty cultures around the world and carefully controlled for a potential problem with the data—they collected examples from cultures from the same gene pools, with and without those stressful rituals. For example, they compared the West African tribes of the Yoruba (stressful rituals) and Ashanti (nonstressful), and similarly matched Native American tribes. With this approach, they attempted to control for genetic contributions to stature (as well as nutrition, since related ethnic groups were likely to have similar diets) and to examine cultural differences instead. Given the effects of stress on growth, it was not surprising that among cultures where kids of ages six to fifteen went through stressful maturational rituals, growth was inhibited (relative to cultures without such rituals, the difference was about 1.5 inches). Surprisingly, going through such rituals at ages two to six had no effect on growth. And most surprising, in cultures in which those rituals took place with kids under two years of age, growth was stimulated—adults were about 2.5 inches taller than in cultures without stressful rituals. There are some possible confounds that could explain the results. One is fairly silly—maybe tall tribes like to put their young children through stressful rituals. One is more plausible—maybe putting very young children through these stressful rituals kills a certain percentage of them, and what doesn’t kill you makes you stronger and taller. Landauer and Whiting noted that possibility and could not rule it out. In addition, even though they attempted to pair similar groups, there may have been differences other than just the stressfulness of the rites of passage—perhaps in diet or child-rearing practices. Not surprisingly, no one has ever measured levels of growth hormone or somatomedins, in, say, Shilluk or Hausa kids while they are undergoing some grueling ritual, so there is no direct endocrine evidence that such stressors actually stimulate growth hormone secretion in a way that increases growth. Despite these problems, these cross- cultural studies have been interpreted by many biological anthropologists as evidence that some types of stressors in humans can actually stimulate growth, amid the broader literature showing the growth-suppressing effects of stress. Enough Already So there’s a whole bunch of ways that prenatal or early childhood stress can have bad and long- term consequences. This can be anxiety provoking; it gets me into a storm of parental agitation just to write about this. Let’s figure out what’s worrisome and what’s not. First, can fetal or childhood exposure to synthetic glucocorticoids have lifelong, adverse effects? Glucocorticoids (such as hydrocortisone) are prescribed in vast amounts, because of their

immunosuppressive or anti-inflammatory effects. During pregnancy, they are administered to women with certain endocrine disorders or who are at risk for delivering preterm. Heavy administration of them during pregnancy has been reported to result in children with smaller head circumferences, emotional and behavioral problems in childhood, and slowing of some developmental landmarks. Are these effects lifelong? No one knows. At this point, the experts have weighed in emphatically stating that a single round of glucocorticoids during either fetal or postnatal life has no adverse effects, though there is potential for problems with heavy use. But heavy doses of glucocorticoids are not administered unless there’s a serious illness going on, so the most prudent advice is to minimize their use clinically but to recognize that the alternative, the disease that prompted the treatment in the first place, is most probably worse. What about prenatal or postnatal stress? Does every little hiccup of stress leave an adverse scar forever after, unto multiple generations? Many times, some relationship in biology may apply to extreme situations—massive trauma, a whole winter’s famine, and so on—but not to more everyday ones. Unfortunately, even the normal range of birth weights predicts adult glucocorticoid levels and the risk of Metabolic syndrome. So these appear not to be phenomena only of the extremes. Next important question: How big are the effects? We’ve seen evidence that increasing amounts of fetal stress, over the normal range, predict increasing risk of Metabolic syndrome long afterward. That statement may be true and describes one of two very different scenarios. For example, it could be that the lowest levels of fetal stress result in a 1 percent risk of Metabolic syndrome, and each increase in stress exposure increases the risk until an exposure to a maximal fetal stress results in a 99 percent chance. Or the least fetal stress could result in a 1 percent risk, and each increase in stress exposure increases the risk until exposure to maximal fetal stress results in a 2 percent risk. In both cases, the endpoint is sensitive to small increments in the amount of stress, but the power of fetal stress to increase disease risk is vastly greater in the first scenario. As we will see in more detail in later chapters, early stress and trauma seem to have a tremendous power in increasing the risk of various psychiatric disorders many years later. Some critics of the FOAD literature seem to be of the opinion that it constitutes cool biology of the “Gee whiz, isn’t nature amazing” variety, but is not a major source of worry. However, the risks of some of these adult diseases vary manyfold as a function of birth weight—so these strike me as big effects. Next question: Regardless of how powerful these effects are, how inevitable are they? Lose it once in a crazed, sleepless moment at two in the morning and yell at your colicky infant and is that it, have you just guaranteed more clogging of her arteries in 2060? Not remotely. As discussed, stress dwarfism is reversible with a different environment. Studies have shown that the lifelong changes in glucocorticoid levels in prenatally stressed rats can be prevented with particular mothering styles postnatally. Much of preventative medicine is a demonstration that vast numbers of adverse health situations can be reversed—in fact, that is a premise of this book. The Cornell anthropologist Merideth Small has written a wonderfully un-neurotic book, Our Babies, Ourselves, which looks at child-rearing practices across the planet. In a particular culture, how often is a child typically held by parents, by non-parents? Do babies sleep alone

ever and, if so, starting at what age? What is the average length of time that a child cries in a particular culture before she is picked up and comforted? In measure after measure, westernized societies and, in particular, the United States, come out at the extreme in these cross-cultural measures, with our emphasis on individuality, independence, and self-reliance. This is our world of both parents working outside the home, of single-parent households, of day care and latchkey kids. There is little evidence that any of these childhood experiences leave indelible biological scars, in contrast to the results of horrific childhood trauma. But whatever style of child-rearing is practiced, it will have its consequences. Small makes a profound point. You begin by reading her book assuming it is going to be an assortment box of prescriptions, that at the end, you’ll emerge with a perfect combo for your kids, a mixture of the Kwakiutl Baby Diet, the Trobriand Sleeping Program, and the Ituri Pygmy Infant Aerobics Plan. But, Small emphasizes, there is no perfect, “natural” program. Societies raise their children so that they grow into adults who behave in a way valued by that society. As Harry Chapin sang in “Cat’s in the Cradle,” that ode to baby boomer remorse, “My boy was just like me.” Growth and Growth Hormone in Adults Personally I don’t grow much anymore, except wider. According to the textbooks, another half- dozen Groundhog Days or so and I’m going to start shrinking. Yet I, like other adults, still secrete growth hormone into my circulation (although much less frequently than when I was an adolescent). What good is it in an adult? Like the Red Queen in Alice in Wonderland, the bodies of adults have to work harder and harder just to keep standing in the same place. Once the growth period of youth is finished and the edifice is complete, the hormones of growth mostly work at rebuilding and remodeling—shoring up the sagging foundation, plastering the cracks that appear here and there. Much of this repair work takes place in bone. Most of us probably view our bones as pretty boring and phlegmatic—they just sit there, inert. In reality, they are dynamic outposts of activity. They are filled with blood vessels, with little fluid-filled canals, with all sorts of cell types that are actively growing and dividing. New bone is constantly being formed, in much the same way as in a teenager. Old bone is being broken down, disintegrated by ravenous enzymes (a process called resorption). New calcium is shuttled in from the bloodstream; old calcium is flushed away. Growth hormone, somatomedins, parathyroid hormone, and vitamin D stand around in hard hats, supervising the project. Why all the tumult? Some of this bustle is because bones serve as the Federal Reserve for the body’s calcium, constantly giving and collecting loans of calcium to and from other organs. And part is for the sake of bone itself, allowing it to gradually rebuild and change its shape in response to need. How else do cowboys’ bow-legged legs get bowed from too much time on a horse? The process has to be kept well balanced. If the bones sequester too much of the body’s calcium, much of the rest of the body shuts down; if the bones dump too much of their calcium into the bloodstream, they become fragile and prone to fracture, and that excess circulating calcium can start forming calcified kidney stones.

Predictably, the hormones of stress wreak havoc with the trafficking of calcium, biasing bone toward disintegration, rather than growth. The main culprits are glucocorticoids. They inhibit the growth of new bone by disrupting the division of the bone-precursor cells in the ends of bones. Furthermore, they reduce the calcium supply to bone. Glucocorticoids block the uptake of dietary calcium in the intestines (uptake normally stimulated by vitamin D), increase the excretion of calcium by the kidney, and accelerate the resorption of bone. If you secrete excessive amounts of glucocorticoids, this increases the risk that your bones will eventually give you problems. This is seen in people with Cushing’s syndrome (in which glucocorticoids are secreted at immensely high levels because of a tumor), and in people being treated with high doses of glucocorticoids to control some disease. In those cases, bone mass decreases markedly, and patients are at greater risk for osteoporosis (softening and weakening of bone).* Any situation that greatly elevates glucocorticoid concentrations in the bloodstream is a particular problem for older people, in whom bone resorption is already predominant (in contrast to adolescents, in whom bone growth predominates, or young adults, in which the two processes are balanced). This is especially a problem in older women. Tremendous attention is now being paid to the need for calcium supplements to prevent osteoporosis in postmenopausal women. Estrogen potently inhibits bone resorption, and as estrogen levels drop after menopause, the bones suddenly begin to degenerate.* A hefty regimen of glucocorticoids on top of that is the last thing you need. These findings suggest that chronic stress can increase the risk of osteoporosis and cause skeletal atrophy. Most clinicians would probably say that the glucocorticoid effects on bone are “pharmacological” rather than “physiological.” This means that normal (physiological) levels of glucocorticoids in the bloodstream, even those in response to normal stressful events, are not enough to damage bone. Instead, it takes pharmacological levels of the hormone (far higher than the body can normally generate), due to a tumor or to ingestion of prescription glucocorticoids, to cause these effects. However, work from Jay Kaplan’s group has shown that chronic social stress leads to loss of bone mass in female monkeys. A Final Word about the L-Word In looking at research on how stress and understimulation can disrupt growth and increase the risks of all sorts of diseases, a theme pops up repeatedly: an infant human or animal can be well fed, maintained at an adequate temperature, peered at nervously, and ministered to by the best of neonatologists, yet still not thrive. Something is still missing. Perhaps we can even risk scientific credibility and detachment and mention the word love here, because that most ephemeral of phenomena lurks between the lines of this chapter. Something roughly akin to love is needed for proper biological development, and its absence is among the most aching, distorting stressors that we can suffer. Scientists and physicians and other caregivers have often been dim at recognizing its importance in the mundane biological processes by which organs and tissues

grow and develop. For example, at the beginning of the twentieth century, the leading expert on child-rearing was a Dr. Luther Holt of Columbia University, who warned parents of the adverse effects of the “vicious practice” of using a cradle, picking up the child when it cried, or handling it too often. All the experts believed that affection not only wasn’t needed for development but was a squishy, messy thing that kept kids from becoming upright, independent citizens. Yet young organisms were able to teach about how these savants were wrong in a classic set of studies begun in the 1950s—studies that are, in my opinion, among the most haunting and troubling of all the pages of science. The work was carried out by the psychologist Harry Harlow of the University of Wisconsin, a renowned and controversial scientist. Psychology at that time was dominated by either Freudians or a rather extreme school of thought called behaviorism, in which behavior (of an animal or a human) was thought to operate according to rather simple rules: an organism does something more frequently because it has been rewarded for it in the past; an organism does something less frequently because it has failed to be rewarded, or has even been punished for that behavior. In this view, just a few basic things like hunger, pain, or sex lie at the basis of reinforcement. Look at the behaviors, view organisms as machines responding to stimuli, and develop a predictive mathematics built around the idea of rewards and punishments. Harlow helped to answer a seemingly obvious question in a non-obvious way. Why do infants become attached to their mothers? Because Mom supplies food. For behaviorists, this was obvious, as attachment was thought to arise solely from the positive reinforcement of food. For Freudians, it was also obvious—infants were thought to lack the “ego development” to form a relationship with any thing/one other than Mom’s breast. For physicians influenced by the likes of Holt, it was obvious and convenient—no need for mothers to visit hospitalized infants— anyone with a bottle would supply attachment needs. No need to worry about preemies kept antiseptically isolated in incubators—regular feeding suffices for human contact. No need for children in orphanages to be touched, held, noted as individuals. What’s love got to do with healthy development? Harlow smelled a rat. He raised infant rhesus monkeys without mothers. Instead, he gave them a choice of two types of artificial “surrogate” mothers. One pseudo-mother had a monkey head constructed of wood and a wire-mesh tube resembling a torso. In the middle of the torso was a bottle of milk. This surrogate mother gave nutrition. The other surrogate mother had a similar head and wire-mesh torso. But instead of containing a milk bottle, this one’s torso was wrapped in terry cloth. The behaviorists and the Freudians would be snuggling up to the milk-mom within seconds. But not the baby monkeys—they chose the terry-cloth mothers. Kids don’t love their mothers because Mom balances their nutritive intake, these results suggested. They love them because, usually, Mom loves them back, or at least is someone soft to cling to. “Man cannot live by milk alone. Love is an emotion that does not need to be bottle- or spoon-fed,” wrote Harlow.

Infant monkey and cloth mother, in a Harlow study.

Harlow and his work remain immensely controversial.* The controversy arises from the nature of his experiments and variations on them (for example, raising monkeys in complete social isolation, in which they never see another living animal). These were brutal studies, and they are often among the primary ones cited by those opposed to animal experimentation. Moreover, Harlow’s scientific writing displayed an appalling callousness to the suffering of these animals— I remember as a student being moved to tears of rage by the savage indifference of his writing. But at the same time, these studies have been extremely useful (although my feeling is that there should have been far fewer of them carried out). They have taught us the science of why we primates love individuals who treat us badly, why the mistreatment can at times increase the love. They have taught us about why being abused as a child increases the risk of your being an abusive adult. Other aspects of Harlow’s work have taught us how repeated separations of infants from their mothers can predispose those individuals to depression when they are adults. The irony is that it required Harlow’s pioneering work to demonstrate the unethical nature of that work. But wasn’t it obvious before? If you prick us, do we not bleed?; if you socially isolate us as infants, do we not suffer? Few in the know thought so. The main point of Harlow’s work wasn’t teaching what we might now wrongly assume to have been obvious then, namely that if you isolate an infant monkey, it is a massive stressor, and that she saddens and suffers for long after. It was to teach the utterly novel fact that if you do the same to a human infant, the same occurs. Chapter 7

Sex and Reproduction

Kidneys and pancreas and heart are important, but what we really want to know is why, when we are being stressed, our menstrual cycles become irregular, erections are more difficult to achieve, and we lose our interest in sex. As it turns out, there are an astonishing number of ways in which reproductive mechanisms may go awry when we are upset. Males: Testosterone and Loss of Erections It makes sense to start simple, so let’s initially consider the easier reproductive system, that of males. In the male, the brain releases the hormone LHRH (luteinizing hormone releasing hormone), which stimulates the pituitary to release LH (luteinizing hormone) and FSH (follicle- stimulating hormone).* LH, in turn, stimulates the testes to release testosterone. Since men don’t have follicles to be stimulated by follicle-stimulating hormone, FSH instead stimulates sperm production. This is the reproductive system of your basic off-the-rack male.

A simplified version of male reproductive endocrinology. The hypothalamus releases LHRH

into the private circulatory system that it shares with the anterior pituitary. LHRH triggers the release by the pituitary of LH and FSH, which work at the testes to cause testosterone

secretion and sperm production. With the onset of a stressor, the whole system is inhibited. LHRH concentrations decline, followed shortly thereafter by declines in LH and FSH, and then the testes close for lunch. The result is a decline in circulating testosterone levels. The most vivid demonstrations of this occur during physical stress. If a male goes through surgery, within seconds of the first slice of a scalpel through his skin, the reproductive axis begins to shut down. Injury, illness, starvation, surgery—all of these drive down testosterone levels. Anthropologists have even shown that in human societies in which there is constant energetic stress (for example, those of rural Nepalese

villagers), there are significantly lower testosterone levels than among sedentary Bostonian controls. But subtle psychological stressors are just as disruptive. Lower the dominance rank of a social primate and down go his testosterone levels. Put a person or a monkey through a stressful learning task and the same occurs. In a celebrated study several decades ago, U.S. Officer Candidate School trainees who underwent an enormous amount of physical and psychological stress were subjected to the further indignity of having to pee into Dixie cups so that military psychiatrists could measure their hormone levels. Lo and behold, testosterone levels were down; maybe not to the levels found in cherubic babies, but still it’s worth keeping in mind the next time you see some leatherneck at a bar bragging about his circulating androgen concentrations. Why do testosterone concentrations plunge with the onset of a stressor? For a variety of reasons. The first occurs at the brain. With the onset of stress, two important classes of hormones, the endorphins and enkephalins (mostly the former), act to block the release of LHRH from the hypothalamus. As will be discussed in chapter 9, endorphins play a role in blocking pain perception and are secreted in response to exercise (helping to account for the famed “runner’s high” or “endorphin high” that hits many hardy joggers around the 30-minute mark). If males secrete endorphins when they are experiencing runner’s high, and these compounds inhibit testosterone release, will exercise suppress male reproduction? Sometimes. Males who do extreme amounts of exercise, such as professional soccer players and runners who cover more than 40 or 50 miles a week, have less LHRH, LH, and testosterone in their circulation, smaller testes, less functional sperm. They also have higher levels of glucocorticoids in their bloodstreams, even in the absence of stress. (A similar decline in reproductive function is found in men who are addicted to opiate drugs.) To jump ahead to the female section, reproductive dysfunction is also seen in women athletes, and this is at least partially due to endorphin release as well. Up to half of competitive runners have menstrual irregularities, and highly athletic girls reach puberty later than usual. For example, in one study of fourteen-year-olds, approximately 95 percent of control subjects had started menstruating, whereas only 20 percent of gymnasts and 40 percent of runners had. This brings up a broader issue important to our era of lookin’ good. Obviously, if you don’t exercise at all, it is not good for you. Exercise improves your health. And a lot of exercise improves your health a lot. But that doesn’t mean that insanely large amounts of exercise are insanely good for your body. At some point, too much begins to damage various physiological systems. Everything in physiology follows the rule that too much can be as bad as too little. There are optimal points of allostatic balance. For example, while a moderate amount of exercise generally increases bone mass, thirty-year-old athletes who run 40 to 50 miles a week can wind up with decalcified bones, decreased bone mass, increased risk of stress fractures and scoliosis (sideways curvature of the spine)—their skeletons look like those of seventy-year-olds. To put exercise in perspective, imagine this: sit with a group of hunter-gatherers from the African grasslands and explain to them that in our world we have so much food and so much free time that some of us run 26 miles in a day, simply for the sheer pleasure of it. They are likely to say, “Are you crazy? That’s stressful.” Throughout hominid history, if you’re running 26 miles in a day, you’re either very intent on eating someone or someone’s very intent on eating you.

Thus, we have a first step. With the onset of stress, LHRH secretion declines. In addition, prolactin, another pituitary hormone that is released during major stressors, decreases the sensitivity of the pituitary to LHRH. A double whammy—less of the hormone dribbling out of the brain, and the pituitary no longer responding as effectively to it. Finally, glucocorticoids block the response of the testes to LH, just in case any of that hormone manages to reach them during the stressor (and serious athletes tend to have pretty dramatic elevations of glucocorticoids in their circulation, no doubt adding to the reproductive problems just discussed). A decline in testosterone secretion is only half the story of what goes wrong with male reproduction during stress. The other half concerns the nervous system and erections. Getting an erection to work properly is so incredibly complicated physiologically that if men ever actually had to understand it, none of us would be here. Fortunately, it runs automatically. In order for a male primate to have an erection, he has to divert a considerable amount of blood flow to his penis, engorging it.* This is accomplished by activating his parasympathetic nervous system. In other words, the guy has to be calm, vegetative, relaxed.

Overexercise can have a variety of deleterious effects. (Left) Max Ernst, Health Through

Sport, photographic enlargement of a photomontage mounted on wood, 1920; (right) Above the Clouds Midnight Passes, collage with fragments of photographs and pencil, 1920.

What happens next, if you are male? You are having a terrific time with someone. Maybe you are breathing faster, your heart rate has increased. Gradually, parts of your body are taking on a sympathetic tone—remember the four F’s of sympathetic function introduced in chapter 2. After awhile, most of your body is screaming sympathetic while, heroically, you are trying to hold on to parasympathetic tone in that one lone outpost as long as possible. Finally, when you can’t take it anymore, the parasympathetic shuts off at the penis, the sympathetic comes roaring on, and you ejaculate. (Incredibly complicated choreography between these two systems; don’t try this unsupervised.) This new understanding generates tricks that sexual therapists advise—if you are close to ejaculating and don’t want to yet, take a deep breath. Expanding the chest muscles briefly triggers a parasympathetic volley that defers the shift from parasympathetic to sympathetic.

What, then, changes during stress? One is that sufficient prior stress will damage and clog up your blood vessels—severe vascular disease can seriously impede blood flow. But what if you’re stressed in that immediate situation? Well, obviously, if you’re nervous or anxious, you’re not calm or vegetative. First, it becomes difficult to establish parasympathetic activity if you are nervous or anxious. You have trouble having an erection. Impotency. And if you already have the erection, you get in trouble as well. You’re rolling along, parasympathetic to your penis, having a wonderful time. Suddenly, you find yourself worrying about the strength of the dollar versus the euro and—shazaam—you switch from parasympathetic to sympathetic far faster than you wanted. Premature ejaculation. It is extremely common for problems with impotency and premature ejaculation to arise during stressful times. Furthermore, this can be compounded by the fact that erectile dysfunction is a major stressor on its own, getting men into this vicious performance anxiety cycle of fearing fear itself. A number of studies have shown that more than half the visits to doctors by males complaining of reproductive dysfunction turn out to be due to “psychogenic” impotency rather than organic impotency (there’s no disease there, just too much stress). How do you tell if it is organic or psychogenic impotency? This is actually diagnosed with surprising ease, because of a quirky thing about human males. As soon as they go to sleep and enter REM (rapid eye movement) dream sleep, they get erections. I’ve consulted with Earth’s penis experts, and no one is sure why this should occur, but that’s how it works.* So a man comes in complaining that he hasn’t been able to have an erection in six months. Is he just under stress? Does he have some neurological disease? Take a handy little penile cuff with an electronic pressure transducer attached to it. Have him put it on just before he goes to sleep. By the next morning you may have your answer—if this guy gets an erection when he goes into REM sleep, his problem is likely to be psychogenic.* Thus, stress will knock out erections quite readily. In general, the problems with erections are more disruptive than problems with testosterone secretion. Testosterone and sperm production have to shut down almost entirely to affect performance. A little testosterone and a couple of sperm wandering around and most males can muddle through. But no erection, and forget about it.* The erectile component is exquisitely sensitive to stress in an incredible array of species. Nonetheless, there are some circumstances where stress does not suppress the reproductive system in a male. Suppose you’re some big bull moose and it’s mating season. You’re spending all your time strutting your stuff and growing your antlers and snorting and having head-butting territorial disputes with the next guy and forgetting to eat right and not getting enough sleep and getting injured and worrying about the competition for some female moose’s favors.* Stressful. Wouldn’t it be pretty maladaptive if the male-male competitive behaviors needed to get the opportunity to mate were so stressful that when the opportunity came, you were sexually dysfunctional? Not a good Darwinian move. Or suppose that in your species, sex is this wildly metabolically demanding activity, involving hours, even days of copulation at the cost of resting or feeding (lions fall in this category, for example). High energetic demands plus little eating or sleeping equals stress. It would be disadvantageous if the stress of mating caused erectile dysfunction.

It turns out that in a lot of species, stressors associated with mating season competition or with mating itself not only don’t suppress the reproductive system, but can stimulate it a bit. In some species where this applies, the seeming stressor doesn’t cause secretion of stress hormones; in other cases, the stress hormones are secreted but the reproductive system becomes insensitive to them. And then there is one species which, regardless of whether it is mating season or not, breaks all the rules concerning the effects of stress on erectile function. It is time we had a little talk about hyenas. Our Friend, the Hyena The spotted hyena is a vastly unappreciated, misrepresented beast. I know this because over the years, in my work in East Africa, I have shared my campsite with the hyena biologist Laurence Frank of the University of California at Berkeley. For lack of distracting television, radio, or telephone, he has devoted his time with me to singing the hyena’s praises. They are wondrous animals who have gotten a bad rap from the press. We all know the scenario. It’s dawn on the savanna. Marlin Perkins of Mutual of Omaha’s Wild Kingdom is there filming lions eating something dead. We are delighted, craning to get a good view of the blood and guts. Suddenly, on the edge of our field of vision, we spot them—skulky, filthy, untrustworthy hyenas looking to dart in and steal some of the food. Scavengers! We are invited to heap our contempt on them (a surprising bias, given how few of the carnivorous among us ever wrestle down our meals with our canines). It wasn’t until the Pentagon purchased a new line of infrared night-viewing scopes and decided to unload its old ones on various zoologists that, suddenly, researchers could watch hyenas at night (important, given that hyenas mostly sleep during the day). Turns out that they are fabulous hunters. And you know what happens? Lions, who are not particularly effective hunters, because they are big and slow and conspicuous, spend most of their time keying in on hyenas and ripping off their kills. No wonder when it’s dawn on the savanna the hyenas on the periphery are looking cranky, with circles under their eyes. They stayed up all night hunting that thing, and who’s having breakfast now? Having established a thread of sympathy for these beasts, let me explain what is really strange about them. Among hyenas, females are socially dominant, which is fairly rare among mammals. They are more muscular and more aggressive, and have more of a male sex hormone (a close relative of testosterone called androstenedione) in their bloodstreams than males. It’s also almost impossible to tell the sex of a hyena by looking at its external genitals. More than two thousand years ago, Aristotle, for reasons obscure to even the most learned, dissected some dead hyenas, discussing them in his treatise Historia Animalium, VI, XXX. The conclusion among hyena savants at the time was that these animals were hermaphrodites— animals that possess all the machinery of both sexes. Hyenas are actually what gynecologists would call pseudohermaphrodites (they just look that way). The female has a fake scrotal sac made of compacted fat cells; she doesn’t really have a penis but, instead, an enlarged clitoris that can become erect. The same clitoris, I might add, with which she has sex and through which she

gives birth. It’s pretty wild. Laurence Frank, who is one of Earth’s experts on hyena genitals, will dart some animal and haul it, anesthetized, into camp. Excitement; we go to check it out, and maybe twenty minutes into examining it, he kind of thinks he knows what sex this particular one is. (Yes, the hyenas themselves know exactly who is which sex, most probably by smell.)

Behold, the female hyena.

Perhaps the most interesting thing about hyenas is that there is a fairly plausible theory as to why they evolved this way, a theory complicated enough for me mercifully to relegate it to the endnotes. For our purposes here, what is important is that hyenas have evolved not only genitals that look unique, but also unique ways to use these organs for social communication. This is where stress comes into play. Among many social mammals, males have erections during competitive situations as a sign of dominance. If you are having a dominance display with another male, you get an erection and wave it around in his face to show what a tough guy you are. Social primates do this all the time. However, among hyenas, an erection is a sign of social subordinance. When a male is menaced by a terrifying female, he gets an erection—“Look, I’m just some poor no-account male; don’t hit me, I was just leaving.” Low-ranking females do the same thing; if a low-ranking female is about to get trounced by a high-ranking one, she gets a conspicuous clitoral erection—“Look, I’m just like one of those males; don’t attack me; you know you’re dominant over me, why bother?” If you’re a hyena, you get an erection when you are stressed. Among male hyenas, the autonomic wiring has got to be completely reversed in order to account for the fact that stress causes erections. This hasn’t yet been demonstrated, but perhaps Berkeley scientists working on this, squandering tax dollars that could otherwise be going to Halliburton and Bechtel, will do it. Thus the hyena stands as the exception to the rule about erectile functions being adversely affected by stress, a broader demonstration of the importance of looking at a zoological oddity as a means of better seeing the context of our own normative physiology, and a friendly word of warning before you date a hyena. Females: Lengthened Cycles and Amenorrhea

We now turn to female reproduction. Its basic outline is similar to that of the male. LHRH is released by the brain, which releases LH and FSH from the pituitary. The latter stimulates the ovaries to release eggs; the former stimulates ovaries to synthesize estrogen. During the first half of the menstrual cycle, the “follicular” stage, levels of LHRH, LH, FSH, and estrogen build up, heading toward the climax of ovulation. This ushers in the second half of the cycle, the “luteal” phase. Progesterone, made in the corpus luteum of the ovary, now becomes the dominant hormone on the scene, stimulating the uterine walls to mature so that an egg, if fertilized just after ovulation, can implant there and develop into an embryo. Because the release of hormones has the fancy quality of fluctuating rhythmically over the menstrual cycle, the part of the hypothalamus that regulates the release of these hormones is generally more structurally complicated in females than in males.

A simplified version of female reproductive endocrinology. The hypothalamus releases LHRH into the private circulatory system that it shares with the anterior pituitary. LHRH triggers the

release by the pituitary of LH and FSH, which in turn bring about ovulation and hormone release from the ovaries.

The first way in which stress disrupts female reproduction concerns a surprising facet of the system. There is a small amount of male sex hormone in the bloodstream of females, even non- hyena females. In human beings, this doesn’t come from the ovaries (as in the hyenas), but from the adrenals. The amount of these “adrenal androgens” is only about 5 percent of that in males, but enough to cause trouble.* An enzyme in the fat cells of females usually eliminates these androgens by converting them to estrogens. Problem solved. But what if you are starving because the crops failed this year? Body weight drops, fat stores are depleted, and suddenly there isn’t enough fat around to convert all the androgen to estrogen. Less estrogen, therefore, is

produced. More important, androgen concentrations build up, which inhibits numerous steps in the reproductive system (it should be noted that this is but one of the mechanisms by which starvation inhibits reproduction). Reproduction is similarly inhibited if you starve voluntarily. One of the hallmarks of anorexia nervosa is disruption of reproduction in the (typically) young women who are starving themselves. There’s more to the reproduction cessation than just the weight loss, since cycling doesn’t necessarily resume in women when they regain the weight unless the initial psychological stressors have been sorted out. But the weight loss still plays a critical, initiating role. And loss of body fat leading to androgen buildup is one of the mechanisms by which reproduction is impaired in females who are extremely active physically. As noted above, this has been best documented in young girls who are serious dancers or runners, in whom puberty can be delayed for years, and in women who exercise enormous amounts, in whom cycles can become irregular or cease entirely. Overall, this is a logical mechanism. In the human, an average pregnancy costs approximately 50,000 calories, and nursing costs about a thousand calories a day; neither is something that should be gone into without a reasonable amount of fat tucked away. Stress also can inhibit reproduction in ways other than shrinkage of fat cells. Many of the same mechanisms apply as in the male. Endorphins and enkephalins will inhibit LHRH release (as discussed, this occurs in female athletes as readily as in males); prolactin and glucocorticoids will block pituitary sensitivity to LHRH; and glucocorticoids will also affect the ovaries, making them less responsive to LH. The net result is lowered secretion of LH, FSH, and estrogen, making the likelihood of ovulating decrease. As a result, the follicular stage is extended, making the entire cycle longer and less regular. At an extreme, the entire ovulatory machinery is not merely delayed, but shut down, a condition termed anovulatory amenorrhea. Stress can also cause other reproductive problems. Progesterone levels are often inhibited, which disrupts maturation of the uterine walls. The release of prolactin during stress adds to this effect, interfering with the activity of progesterone. Thus, even if there is still enough hormonal action during the follicular period to cause ovulation, and the egg has become fertilized, it is now much less likely to implant normally. The loss of estrogen with sustained stress has some consequences beyond the reproductive realm. For example, amid the controversies discussed in chapter 3 about whether estrogen protects against cardiovascular disease, it is quite clear that it protects against osteoporosis, and stress- induced declines in estrogen levels have bad effects on bone strength. Of all the hormones that inhibit the reproductive system during stress, prolactin is probably the most interesting. It is extremely powerful and versatile; if you don’t want to ovulate, this is the hormone to have lots of in your bloodstream. It not only plays a major role in the suppression of reproduction during stress and exercise, but it also is the main reason that breast feeding is such an effective form of contraception. Oh, you are shaking your head smugly at the ignorance of this author with that Y chromosome; that’s an old wives’ tale; nursing isn’t an effective contraceptive. On the contrary, nursing works

fabulously. It probably prevents more pregnancies than any other type of contraception. All you have to do is do it right. Breast feeding causes prolactin secretion. There is a reflex loop that goes straight from the nipples to the hypothalamus. If there is nipple stimulation for any reason (in males as well as females), the hypothalamus signals the pituitary to secrete prolactin. And as we now know, prolactin in sufficient quantities causes reproduction to cease. The problem with nursing as a contraceptive is how it is done in Western societies. During the six months or so that she breast-feeds, the average mother in the West allows perhaps half a dozen periods of nursing a day, each for 30 to 60 minutes. Each time she nurses, prolactin levels go up in the bloodstream within seconds, and at the end of the feeding, prolactin settles back to pre-nursing levels fairly quickly. This most likely produces a scalloping sort of pattern in prolactin release. This is not how most women on earth nurse. A prime example emerged a few years ago in a study of hunter-gatherer Bushmen in the Kalahari Desert of southern Africa (the folks depicted in the movie The Gods Must Be Crazy) Bushman males and females have plenty of intercourse, and no one uses contraceptives, but the women have a child only about every four years. Initially, this seemed easy to explain. Western scientists looked at this pattern and said, “They’re hunter-gatherers: life for them must be short, nasty, and brutish; they must all be starving.” Malnutrition induces cessation of ovulation.

A Kalahari Bushman mother with her child in a hip sling.

However, when anthropologists looked more closely, they found that the Bushmen were anything but suffering. If you are going to be nonwestemized, choose to be a hunter-gatherer over being a nomadic pastoralist or an agriculturist. The Bushmen hunt and gather only a few hours a day, and spend much of the rest of their time sitting around chewing the fat. Scientists have called them

the original affluent society. Out goes the idea that the four-year birth interval is due to malnutrition. Instead, the lengthy interval is probably due to their nursing pattern. This was discovered by a pair of scientists, Melvin Konner and Carol Worthman.* When a hunter-gatherer woman gives birth, she begins to breast-feed her child for a minute or two approximately every fifteen minutes. Around the clock. For the next three years. (Suddenly this doesn’t seem like such a hot idea after all, does it?) The young child is carried in a sling on the mother’s hip so he can nurse easily and frequently. At night, he sleeps near his mother and will nurse every so often without even waking her (as Konner and Worthman, no doubt with their infrared night-viewing goggles and stopwatches, scribble away on their clipboards at two in the morning). Once the kid can walk, he’ll come running in from play every hour or so to nurse for a minute. When you breast-feed in this way, the endocrine story is very different. At the first nursing period, prolactin levels rise. And with the frequency and timing of the thousands of subsequent nursings, prolactin stays high for years. Estrogen and progesterone levels are suppressed, and you don’t ovulate. This pattern has a fascinating implication. Consider the life history of a hunter-gatherer woman. She reaches puberty at about age thirteen or fourteen (a bit later than in our society). Soon she is pregnant. She nurses for three years, weans her child, has a few menstrual cycles, becomes pregnant again, and repeats the pattern until she reaches menopause. Think about it: over the course of her life span, she has perhaps two dozen periods. Contrast that with modern Western women, who typically experience hundreds of periods over their lifetime. Huge difference. The hunter-gatherer pattern, the one that has occurred throughout most of human history, is what you see in nonhuman primates. Perhaps some of the gynecological diseases that plague modern westernized women have something to do with this activation of a major piece of physiological machinery hundreds of times when it may have evolved to be used only twenty times; an example of this is probably endometriosis (having uterine lining thickening and sloughing off in places in the pelvis and abdominal wall where it doesn’t belong), which is more common among women with fewer pregnancies and who start at a later age.* Females: Disruption of Libido The preceding section describes how stress disrupts the nuts and bolts of female reproduction— uterine walls, eggs, ovarian hormones, and so on. But what about its effects upon sexual behavior? Just as stress does not do wonders for erections or for the desire of a male to do something with his erections, stress also disrupts female libido. This is a commonplace experience among women stressed by any number of circumstances, as well as among laboratory animals undergoing stress. It is relatively easy to document a loss of sexual desire among women when they are stressed— just hand out a questionnaire on the subject and hope it is answered honestly. But how is sexual drive studied in a laboratory animal? How can one possibly infer a libidinous itch on the part of a female rat, for example, as she gazes into the next cage at the male with the limpid eyes and cute incisors? The answer is surprisingly simple—how often would she be willing to press a lever in

order to gain access to that male? This is science’s quantitative way of measuring rodent desire (or, to use the jargon of the trade, “proceptivity”).* A similar experimental design can be used to measure proceptive behavior in primates. Proceptive and receptive behaviors fluctuate among female animals as a function of factors like the point in the reproductive cycle (both of these measures of sexual behavior generally peak around ovulation), the recency of sex, the time of year, or vagaries of the heart (who is the male in question). In general, stress suppresses both proceptive and receptive behaviors. This effect of stress is probably rooted in its suppression of the secretion of various sex hormones. Among rodents, both proceptive and receptive behaviors disappear when a female’s ovaries are removed, and the absence of estrogen after the ovariectomy is responsible; as evidence, injection of ovariectomized females with estrogen reinstates these sexual behaviors. Moreover, the peak in estrogen levels around ovulation explains why sexual behavior is almost entirely restricted to that period. A similar pattern holds in primates, but it is not as dramatic as in rodents. A decline in sexual behavior, although to a lesser extent, follows ovariectomy in a primate. For humans, estrogen plays a role in sexuality, but a still weaker one—social and interpersonal factors are far more important. Estrogen exerts these effects both in the brain and peripheral tissue. Genitals and other parts of the body contain ample amounts of estrogen receptors and are made more sensitive to tactile stimulation by the hormone. Within the brain, estrogen receptors occur in areas that play a role in sexual behavior; through one of the more poorly understood mechanisms of neuroendocrinology, when estrogen floods those parts of the brain, salacious thoughts follow. Surprisingly, adrenal androgens also play a role in proceptive and receptive behaviors; as evidence, sex drive goes down following removal of the adrenals and can be reinstated by administration of synthetic androgens. This appears to be more of a factor in primates and humans than in rodents. While the subject has not been studied in great detail, there are some reports that stress suppresses the levels of adrenal androgens in the bloodstream. And stress certainly suppresses estrogen secretion. As noted in chapter 3, Jay Kaplan has shown that the stressor of social subordinance in a monkey can suppress estrogen levels as effectively as removing her ovaries. Given these findings, it is relatively easy to see how stress disrupts sexual behavior in a female. Stress and the Success of High-Tech Fertilization In terms of psychological distress, few medical maladies match infertility—the strain placed on a relationship with a significant other, the disruption of daily activities and ability to concentrate at work, the estrangement from friends and family, and the rates of depression.* Thus, circumventing infertility with recent high-tech advances has been a wonderful medical advance. There is now a brave new world of assisted fertilization: artificial insemination; in vitro fertilization (IVF), in which sperm and egg meet in a petri dish, and fertilized eggs are then implanted in the woman; preimplantation screening, carried out when one of the couple has a serious genetic disorder; after eggs are fertilized, their DNA is analyzed, and only those eggs that

do not carry the genetic disorder are implanted. Donor eggs, donor sperm. Injection of an individual sperm into an egg, when the problem is an inability of the sperm to penetrate the egg’s membrane on its own. Some forms of infertility are solved with some relatively simple procedures, but others involve extraordinary, innovative technology. There are two problems with that technology, however. The first is that it is an astonishingly stressful experience for the individuals who go through it. Furthermore, it’s expensive as hell, and is often not paid for by insurance, especially when some of the fancier new experimental techniques are being tried. How many young couples can afford to spend ten to fifteen thousand dollars out of pocket each cycle they attempt to get pregnant? Next, most IVF clinics are located only near major medical centers, meaning that many participants have to spend weeks in a motel room in some strange city, far from friends and family. For some genetic screening techniques, only a handful of places in the world are available, thus adding a long waiting list to the other stress factors. But those stress-induced factors pale compared with the stress generated by the actual process. Weeks of numerous, painful daily shots with synthetic hormones and hormone suppressors that can do some pretty dramatic things to mood and mental state. Daily blood draws, daily sonograms, the constant emotional roller-coaster of whether the day’s news is good or bad: how many follicles, how big are they, what circulating hormone levels have been achieved? A surgical procedure and then the final wait to see whether you have to try the whole thing again. The second problem is that it rarely works. It is very hard to figure out how often natural attempts at fertilization actually succeed in humans. And it is hard to find out what the success rates are for the high-tech procedures, as clinics often fudge the numbers in their brochures— “We don’t like to publish our success rates, because we take on only the most difficult, challenging cases, and thus our numbers must superficially seem worse than those of other clinics that are wimps and take only the easy ones”—and thus, they say, it is hard to gauge just how bad the odds are for a couple with an infertility problem going this route. Nevertheless, going through one of those grueling IVF cycles has a pretty low chance of succeeding. All that has preceded in this chapter would suggest that the first problem, the stressfulness of IVF procedures, contributes to the second problem, the low success rate. A number of researchers have specifically examined whether women who are more stressed during IVF cycles are the ones less likely to have successful outcomes. And the answer is a resounding maybe. The majority of studies do show that the more stressed women (as determined by glucocorticoid levels, cardiovascular reactivity to an experimental stressor, or self-report on a questionnaire) are indeed less likely to have successful IVFs. Why, then, the ambiguity? For one thing, some of the studies were carried out many days or weeks into the long process, where women have already gotten plenty of feedback as to whether things are going well; in those cases, an emerging unsuccessful outcome might cause the elevated stress-response, rather than the other way around. Even in studies in which stress measures are taken at the beginning of the process, the number of previous cycles must be controlled for. In other words, a stressed woman may indeed be less likely to have a successful outcome, but both traits may be due to the fact that she is an especially poor candidate who has already gone through eight unsuccessful prior attempts and is a wreck.

In other words, more research is needed. If the correlation does turn out to be for real, one hopes that the outcome of that will be something more constructive than clinicians saying, “And try not to be stressed, because studies have shown it cuts down the chances IVF will succeed.” It would be kind of nice if progress in this area actually resulted in eliminating the stressor that initiated all these complexities in the first place, namely, the infertility. Miscarriage, Psychogenic Abortions, and Preterm Labor The link between stress and spontaneous abortion in humans prompted Hippocrates to caution pregnant women to avoid unnecessary emotional disturbances.* Since then, it is a thread that runs through some of our most florid and romantic interpretations of the biology of pregnancy. There’s Anne Boleyn attributing her miscarriage to the shock of seeing Jane Seymour sitting on King Henry’s lap, or Rosamond Vincy losing her baby when frightened by a horse in Middlemarch. In the 1990 movie Pacific Heights (which took the Reagan-Bush era to its logical extreme, encouraging us to root for the poor landlords being menaced by a predatory tenant), the homeowner, played by Melanie Griffith, has a miscarriage in response to psychological harassment by the Machiavellian renter. And in the less literary and more mundane realm of everyday life, the stress of a high-demand /low-control job increases the risk of miscarriage among women. Stress can cause miscarriages in other animals as well. This may occur, for example, when pregnant animals in the wild or in a corral have to be captured for some reason (a veterinary exam) or are stressed by being transported. Studies of social hierarchies among animals in the wild have revealed one instance in which stress-induced miscarriages often occur. In many social species, not all males do equivalent amounts of reproducing. Sometimes the group contains only a single male (typically called a “harem male”) who does all the mating; sometimes there are a number of males, but only one or a few dominant males reproduce.* Suppose the harem male is killed or driven out by an intruding male, or a new male migrates into the multi-male group and moves to the top of the dominance hierarchy. Typically, the now-dominant male goes about trying to increase his own reproductive success, at the expense of the prior male. What does the new guy do? In some species, males will systematically try to kill the infants in the group (a pattern called competitive infanticide and observed in a number of species, including lions and some monkeys), thus reducing the reproductive success of the preceding male. Following the killing, moreover, the female ceases to nurse and, as a result, is soon ovulating and ready for mating, to the convenient advantage of the newly resident male. Grim stuff, and a pretty strong demonstration of something well recognized by most evolutionists these days; contrary to what Marlin Perkins taught us, animals rarely behave “for the good of the species.” Instead, they typically act for the good of their own genetic legacy and that of their close relatives. Among some species—wild horses and baboons, for example—the male will also systematically harass any pregnant females to the point of miscarriage, by the same logic. This pattern is seen in a particularly subtle way among rodents. A group of females resides with a single harem male. If he is driven out by an intruder male who takes up residence, within days,

females who have recently become pregnant fail to implant the fertilized egg. Remarkably, this termination of pregnancy does not require physical harassment on the part of the male. It is his new, strange odor that causes the failed pregnancies by triggering a disruptive rise in prolactin levels. As proof of this, researchers can trigger this phenomenon (called the Bruce-Parkes effect) with merely the odor of a novel male. Why is it adaptive for females to terminate pregnancy just because a new male has arrived on the scene? If the female completes her pregnancy, the kids will promptly be killed by this new guy. So, making the best of a bad situation, evolution has sculpted this response to at least save the further calories that would be devoted to the futile pregnancy—terminate it and ovulate a few days later.* Despite the drama of the Bruce-Parkes effect, stress-induced miscarriages are relatively rare among animals, particularly among humans. It is not uncommon to decide retrospectively that when something bad happens (such as a miscarriage), there was significant stress beforehand. To add to the confusion, there is a tendency to attribute miscarriages to stressful events occurring a day or so preceding them. In actuality, most miscarriages involve the expelling of a dead fetus, which has typically died quite a while before. If there was a stressful cause, it is likely to have come days or even weeks before the miscarriage, not immediately preceding it. When a stress-induced miscarriage does occur, however, there is a fairly plausible explanation of how it happens. The delivery of blood to the fetus is exquisitely sensitive to blood flow in the mother, and anything that decreases uterine blood flow will be disruptive to the fetal blood supply. Moreover, fetal heart rate closely tracks that of the mother, and various psychological stimuli that stimulate or slow down the heart rate of the mother will cause a similar change a minute or so later in the fetus. This has been shown in a number of studies of both humans and primates. Trouble seems to occur during stress as a result of repeated powerful activation of the sympathetic nervous system, causing increased secretion of norepinephrine and epinephrine. Studies of a large number of different species show that these two hormones will decrease blood flow through the uterus—dramatically, in some cases. Exposing animals to something psychologically stressful (for example, a loud noise in the case of pregnant sheep, or the entrance of a strange person into the room in which a pregnant rhesus monkey is housed) will cause a similar reduction in blood flow, decreasing the delivery of oxygen (called hypoxia) to the fetus. This is certainly not a good thing, and this sort of prenatal stress returns us to all the issues of growth in chapter 6. The general assumption in the field is that it takes a number of these hypoxic episodes to cause asphyxiation. Thus, severe stress can increase the likelihood of miscarriage. Furthermore, if one is at a late stage in pregnancy, stress can increase the risk of preterm birth, an effect that is probably due to elevated glucocorticoids. Certainly not a good thing, given what we saw in the last chapter about the metabolic imprinting consequences of low birth weight. How Detrimental to Female Reproduction is Stress?

As we have seen, there is an extraordinary array of mechanisms by which reproduction can be disrupted in stressed females—fat depletion; secretion of endorphins, prolactin, and glucocorticoids acting on the brain, pituitary, and ovaries; lack of progesterone; excessive prolactin acting on the uterus. Moreover, possible blockage of implantation of the fertilized egg and changes in blood flow to the fetus generate numerous ways in which stress can make it less likely that a pregnancy will be carried to term. With all these different mechanisms implicated, it seems as if even the mildest of stressors would shut down the reproductive system completely. Surprisingly, however, this is not the case; collectively, these mechanisms are not all that effective. One way of appreciating this is to examine the effects of chronic low-grade stress on reproduction. Consider traditional nonwesternized agriculturists with a fair amount of background disease (say seasonal malaria), a high incidence of parasites, and some seasonal malnutrition thrown in—farmers in Kenya, for example. Before family planning came into vogue, the average number of children born to a Kenyan woman was about eight. Compare this with the Hutterites, nonmechanized farmers who live a life similar to that of the Amish. Hutterites experience none of the chronic stressors of the Kenyan farmers, use no contraceptives, and have an almost identical reproductive rate—an average of nine children per woman. (It is difficult to make a close quantitative comparison of these two populations. The Hutterites, for example, delay marriage, decreasing their reproductive rate, whereas Kenyan agriculturists traditionally do not. Conversely, Kenyan agriculturists typically breast-feed for at least a year, decreasing their reproductive rate, in contrast to the Hutterites, who typically nurse far less. The main point, however, is that even with such different lifestyles, the two reproductive rates are nearly equal.) How about reproduction during extreme stress? This has been studied in a literature that always poses problems for those discussing it: how to cite a scientific finding without crediting the monsters who did the research? These are the studies of women in the Third Reich’s concentration camps, conducted by Nazi doctors. (The convention has evolved never to cite the names of the doctors, and always to note their criminality.) In a study of the women in the Theresienstadt concentration camp, 54 percent of the reproductive-age women were found to have stopped menstruating. This is hardly surprising; starvation, slave labor, and unspeakable psychological terror are going to disrupt reproduction. The point typically made is that, of the women who stopped menstruating, the majority stopped within their first month in the camps— before starvation and labor had pushed fat levels down to the decisive point. Many researchers cite this as a demonstration of how disruptive even psychological stress can be to reproduction. To me, the surprising fact is just the opposite. Despite starvation, exhausting labor, and the daily terror that each day would be their last, only 54 percent of those women ceased menstruating. Reproductive mechanisms were still working in nearly half the women (although a certain number may have been having anovulatory cycles). And I would wager that despite the horrors of their situation, there were still many men who were reproductively intact. That reproductive physiology still operated in any individual to any extent, under those circumstances, strikes me as extraordinary.

Reproduction represents a vast hierarchy of behavioral and physiological events that differ considerably in subtlety. Some steps are basic and massive—the eruption of an egg, the diverting of rivers of blood to a penis. Others are as delicate as the line of a poem that awakens your heart or the whiff of a person’s scent that awakens your loins. Not all the steps are equally sensitive to stress. The basic machinery of reproduction can be astoundingly resistant to stress in a subset of individuals, as evidence from the Holocaust shows. Reproduction is one of the strongest of biological reflexes—just ask a salmon leaping upstream to spawn, or males of various species risking life and limb for access to females, or any adolescent with that steroid-crazed look. But when it comes to the pirouettes and filigrees of sexuality, stress can wreak havoc with subtleties. That may not be of enormous consequence to a starving refugee or a wildebeest in the middle of a drought. But it matters to us, with our culture of multiple orgasms and minuscule refractory periods and oceans of libido. And while it is easy to make fun of those obsessions of ours, those nuances of sexuality, the Cosmos and GQs and other indices of our indulged lives, matter to us. They provide us with some of our greatest, if also our most fragile and evanescent, joys. Chapter 8

Immunity, Stress, and Disease

The halls of academe are filling with a newly evolved species of scientist—the psychoneuroimmunologist—who makes a living studying the extraordinary fact that what goes on in your head can affect how well your immune system functions. Those two realms were once thought to be fairly separate—your immune system kills bacteria, makes antibodies, hunts for tumors; your brain makes you do the bunny hop, invents the wheel, has favorite TV shows. Yet the dogma of the separation of the immune and nervous systems has fallen by the wayside. The autonomic nervous system sends nerves into tissues that form or store the cells of the immune system and eventually enter the circulation. Furthermore, tissue of the immune system turns out to be sensitive to (that is, it has receptors for) all the interesting hormones released by the pituitary under the control of the brain. The result is that the brain has a vast potential for sticking its nose into the immune system’s business. The evidence for the brain’s influence on the immune system goes back at least a century, dating to the first demonstration that if you waved an artificial rose in front of someone who is highly allergic to roses (and who didn’t know it was a fake), they’d get an allergic response. Here’s a charming and more recent demonstration of the brain influencing the immune system: take some professional actors and have them spend a day doing either a depressing negative scene, or an uplifting euphoric one. Those in the former state show decreased immune responsiveness, while those in the latter manifest an increase. (And where was such a study carried out? In Los Angeles, of course, at UCLA.) But the study that probably most solidified the link between the brain and the immune system used a paradigm called conditioned immunosuppression. Give an animal a drug that suppresses the immune system. Along with it, provide, à la Pavlov’s experiments, a “conditioned stimulus”—for example, an artificially flavored drink, something that the animal will associate with the suppressive drug. A few days later, present the conditioned stimulus by itself—and down goes immune function. In 1982 the report of an experiment using a variant of this paradigm, carried out by two pioneers in this field, Robert Ader and Nicholas Cohen of the University of Rochester, stunned scientists. The two researchers experimented with

a strain of mice that spontaneously develop disease because of overactivity of their immune systems. Normally, the disease is controlled by treating the mice with an immunosuppressive drug. Ader and Cohen showed that by using their conditioning techniques, they could substitute the conditioned stimulus for the actual drug—and sufficiently alter immunity in these animals to extend their life spans. Studies such as these convinced scientists that there is a strong link between the nervous system and the immune system. It should come as no surprise that if the sight of an artificial rose or the taste of an artificially flavored drink can alter immune function, then stress can, too. In the first half of this chapter, I discuss what stress does to immunity and how this might be useful during a stressful emergency. In the second half, I’ll examine whether sustained stress, by way of chronic suppression of immunity, can impair the ability of a body to fight off infectious disease. This is a fascinating question, which can be answered only with a great deal of caution and many caveats. Although evidence is emerging that stress-induced immunosuppression can indeed increase the risk and severity of some diseases, the connection is probably relatively weak and its importance often exaggerated. In order to evaluate the results of this confusing but important field, we need to start with a primer about how the immune system works. Immune System Basics The primary job of the immune system is to defend the body against infectious agents such as viruses, bacteria, fungi, and parasites. The process is dauntingly complex. For one thing, the immune system must tell the difference between cells that are normal parts of the body and cells that are invaders—in immunologic jargon, distinguishing between “self” and “non-self.” Somehow, the immune system can remember what every cell in your body looks like, and any cells that lack your distinctive cellular signature (for example, bacteria) are attacked. Moreover, when your immune system does encounter a novel invader, it can even form an immunologic memory of what the infectious agent looks like, to better prepare for the next invasion—a process that is exploited when you are vaccinated with a mild version of an infectious agent in order to prime your immune system for a real attack. Such immune defenses are brought about by a complex array of circulating cells called lymphocytes and monocytes (which are collectively known as white blood cells; cyte is a term for cells). There are two classes of lymphocytes: T cells and B cells. Both originate in the bone marrow, but T cells migrate to mature in the thymus (hence the T), while B cells mature in the bone marrow. B cells principally produce antibodies, but there are several kinds of T cells (T helper and T suppressor cells, cytotoxic killer cells, and so on). The T and B cells attack infectious agents in very different ways. T cells bring about cell- mediated immunity (illustration). When an infectious agent invades the body, it is recognized by a type of monocyte called a macrophage, which presents the foreign particle to a T helper cell. A metaphorical alarm is now sounded, and T cells begin to proliferate in response to the invasion. This alarm system ultimately results in the activation and proliferation of cytotoxic killer cells,

which, as their name implies, attack and destroy the infectious agent. It is this, the T-cell component of the immune system, that is knocked out by the AIDS virus. By contrast, B cells cause antibody-mediated immunity (illustration). Once the macrophage–T helper cell collaboration has occurred, the T helper cells then stimulate B-cell proliferation. The main task of the B cells is to differentiate and generate antibodies, large proteins that will recognize and bind to some specific feature of the invading infectious agent (typically, a distinctive surface protein). This specificity is critical—the antibody formed has a fairly unique shape, which will conform perfectly to the shape of the distinctive feature of the invader, like the fit between a lock and key. In binding to the specific feature, antibodies immobilize the infectious agent and target it for destruction.

The cascade of cell-mediated immunity. (1) An infectious agent is encountered by a type of

monocyte called a macrophage. (2) This stimulates the macrophage to present the infectious agent to a T helper cell (a type of white blood cell) and to release interleukin-1 (IL-1), which stimulates T helper cell activity. (3) The T helper cell, as a result, releases

interleukin-2 (IL-2), which triggers T-cell proliferation. (4) This eventually causes another type of white blood cell, cytotoxic killer cells, to proliferate and destroy the infectious agent.

The cascade of antibody-mediated immunity. (1) An infectious agent is encountered by a macrophage. (2) This encounter stimulates it to present the infectious agent to a T helper cell and to release interleukin-1 (IL-1), which stimulates T helper cell activity. (3) The T

helper cell then secretes B-cell growth factor, triggering differentiation and proliferation of another white blood cell, B cells. (4) The B cells make and release specific antibodies that

bind to surface proteins on the infectious agent, targeting it for destruction by a large group of circulating proteins known as complement.

There is an additional twist to the immune system. If different parts of the liver, for example, need to coordinate some activity, they have the advantage of sitting adjacent to each other. But the immune system is distributed throughout the circulation. In order to sound immune alarms throughout this far-flung system, blood-borne chemical messengers that communicate between different cell types, called cytokines, have evolved. For example, when macrophages first recognize an infectious agent, they release a messenger called interleukin-1. This triggers the T helper cell to release interleukin-2, which stimulates T-cell growth (to make life complicated, there are at least half a dozen additional interleukins with more specialized roles). On the antibody front, T cells also secrete B-cell growth factor. Other classes of messengers, such as interferons, activate broad classes of lymphocytes. The process of the immune system sorting self and non-self usually works well (although truly insidious tropical parasites like those that cause schistosomiasis have evolved to evade your immune system by pirating the signature of your own cells). Your immune system happily spends its time sorting out self from non-self: red blood cells, part of us. Eyebrows, our side. Virus, no good, attack. Muscle cell, good guy…. What if something goes wrong with the immune system’s sorting? One obvious kind of error could be that the immune system misses an infectious invader; clearly, bad news. Equally bad is the sort of error in which the immune system decides something is a dangerous invader that really isn’t. In one version of this, some perfectly innocuous compound in the world around you triggers an alarm reaction. Maybe it is something that you normally ingest, like peanuts or shellfish, or something airborne and innocuous, like pollen. But your immune system has mistakenly decided that this is not only foreign but dangerous, and kicks into gear. And this is an allergy. In the second version of the immune system overreacting, a normal part of your own body is mistaken for an infectious agent and is attacked. When the immune system erroneously attacks a normal part of the body, a variety of horrendous “autoimmune” diseases may result. In multiple sclerosis, for example, part of your nervous system is attacked; in juvenile diabetes, it’s the cells in the pancreas that normally secrete insulin. As we’ll see shortly, stress has some rather confusing effects on autoimmune diseases. So far in this overview of the immune system, we’ve been concentrating on something called acquired immunity. Suppose you’re exposed to some novel, dangerous pathogen, pathogen X, for the first time. Acquired immunity has three features. First, you acquire the ability to target pathogen X specifically, with antibodies and cell-mediated immunity that specifically recognize that pathogen. This really works to your advantage—a bullet with pathogen X’s name written on

it. Second, it takes some time to build up that immunity when you are first exposed to pathogen X—this involves finding which antibody has the best fit and generating a zillion copies of it. Finally, while you will now be geared up to specifically go after pathogen X for a long time to come once that specific defense is on line, repeated exposure to pathogen X will boost those targeted defenses even more. Such acquired immunity is a pretty fancy invention, and it is found only in vertebrates. But we also contain a simpler, more ancient branch of the immune system, one shared with species as distant as insects, called innate immunity. In this realm, you don’t bother with acquiring the means to target pathogen X specifically with antibodies that will be different from those that would target, say, pathogen Y. Instead, the second any sort of pathogen hits your system, this nonspecific immune response swings into action. This generalized immune response tends to occur at the beachhead where a pathogen gets its first foothold, like your skin, or moist mucosal tissue, like in your mouth or nose. As a first step, your saliva contains a class of antibodies that generically attack any sort of microbe that it encounters, instead of acquiring a means of targeting specific invaders. These antibodies are secreted and coat your mucosal surfaces like an antiseptic paint. In addition, at the site of infection, capillaries loosen up, allowing cells of the innate immune response to slip out of the circulation to infiltrate the immediate area of infection. These cells include macrophages, neutrophils, and natural killer cells, which then attack the microbe. The loosening of the capillaries also allows fluid containing proteins that can fight the invasive microbes to flow in from the circulation. And what happens as a result of that? The proteins fight the microbe, but the fluid also makes the area swell up, causing edema. This is your innate immune system leaping into action, causing inflammation.* This gives us a broad overview of immune function. Time to see what stress does to immunity. Naturally, as it turns out, a lot more complicated things than used to be suspected.

Photomicrograph of a natural killer T cell attacking a tumor cell.

How Does Stress Inhibit Immune Function? It’s been almost sixty years since Selye discovered the first evidence of stress-induced immunosuppression, noting that immune tissues like the thymus gland atrophied among rats

subjected to nonspecific unpleasantness. Scientists have learned more about the subtleties of the immune system since then, and it turns out that a period of stress will disrupt a wide variety of immune functions. Stress will suppress the formation of new lymphocytes and their release into the circulation, and shorten the time preexisting lymphocytes stay in the circulation. It will inhibit the manufacturing of new antibodies in response to an infectious agent, and disrupt communication among lymphocytes through the release of relevant messengers. And it will inhibit the innate immune response, suppressing inflammation. All sorts of stressors do this—physical, psychological, in primates, rats, birds, even in fish. And, of course, in humans, too. The best-documented way in which such immune suppression occurs is via glucocorticoids. Glucocorticoids, for example, can cause shrinking of the thymus gland; this is such a reliable effect that in olden days (circa 1960), before it was possible to measure directly the amount of glucocorticoids in the bloodstream, one indirect way of doing so was to see how much the thymus gland in an animal had shrunk. The smaller the thymus, the more glucocorticoids in the circulation. Glucocorticoids halt the formation of new lymphocytes in the thymus, and most of the thymic tissue is made up of these new cells, ready to be secreted into the bloodstream. Because glucocorticoids inhibit the release of messengers like interleukins and interferons, they also make circulating lymphocytes less responsive to an infectious alarm. Glucocorticoids, moreover, cause lymphocytes to be yanked out of the circulation and stuck back in storage in immune tissues. Most of these glucocorticoid effects are against T cells, rather than B cells, meaning that cell-mediated immunity is more disrupted than antibody-mediated immunity. And most impressively, glucocorticoids can actually kill lymphocytes. This taps into one of the hottest topics in medicine, which is the field of “programmed cell death.”* Cells are programmed to commit suicide sometimes. For example, if a cell begins to become cancerous, there is a suicide pathway that gets activated to kill the cell before it starts dividing out of control; a few types of cancers involve the failure of the programmed cell death to occur. It turns out that glucocorticoids can trigger those suicide pathways into action in lymphocytes, through a variety of mechanisms. Sympathetic nervous system hormones, beta-endorphin, and CRH within the brain also play a role in suppressing immunity during stress. The precise mechanisms by which this happens are nowhere near as well understood as with glucocorticoid-induced immune suppression, and these other hormones have traditionally been viewed as less important than the glucocorticoid part of the story. However, a number of experiments have shown that stressors can suppress immunity independently of glucocorticoid secretion, strongly implicating these other routes. Why is Immunity Suppressed During Stress? Figuring out exactly how glucocorticoids and the other stress hormones suppress immunity is a very hot topic these days in cell and molecular biology, especially the part about killing lymphocytes. But amid all this excitement about cutting-edge science, it would be reasonable to begin to wonder why you should want your immune system suppressed during stress. In chapter 1, I offered an explanation for this; now that the process of stress-induced immunosuppression

has been explained in a little more detail, it should be obvious that my early explanation makes no sense. I suggested that during stress it is logical for the body to shut down long-term building projects in order to divert energy for more immediate needs—this inhibition includes the immune system, which, while fabulous at spotting a tumor that will kill you in six months or making antibodies that will help you in a week, is not vital in the next few moments’ emergency. That explanation would make sense only if stress froze the immune system right where it was—no more immune expenditures until the emergency is finished. However, that is not what happens. Instead, stress causes the active expenditure of energy in order to disassemble the preexisting immune system—tissues are shrunk, cells are destroyed. This cannot be explained by a mere halt to expenditures—you’re paying, energetically, to take apart the immune system. So out goes this extension of the long-term versus short-term theory. Why should evolution set us up to do something as apparently stupid as disassembling our immune system during stress? Maybe there isn’t a good reason. This actually isn’t as crazy of a response as you might think. Not everything in the body has to have an explanation in terms of evolutionary adaptiveness. Maybe stress-induced immunosuppression is simply a by-product of something else that is adaptive; it just came along for the ride. This is probably not the case. During infections, the immune system releases the chemical messenger interleukin-1, which among other activities stimulates the hypothalamus to release CRH. As noted in chapter 2, CRH stimulates the pituitary to release ACTH, which then causes adrenal release of glucocorticoids. These in turn suppress the immune system. In other words, under some circumstances, the immune system will ask the body to secrete hormones that will ultimately suppress the immune system. For whatever reason the immunosuppression occurs, the immune system sometimes encourages it. It is probably not just an accident.* Various ideas have floated around over the years to explain why you actively disassemble immunity during stress with the willing cooperation of the immune system. Some seemed fairly plausible until people learned a bit more about immunity and could rule them out. Others were quite nutty, and I happily advocated a few of these in the first edition of this book. But in the last decade, an answer has emerged, and it really turns this field on its head. Surprise It turns out that during the first few minutes (say, up to about thirty) after the onset of a stressor, you don’t uniformly suppress immunity—you enhance many aspects of it (phase A on the accompanying graph). This is shown with all realms of immunity, but in particular for innate immunity. This makes sense—it may be helpful to activate parts of your immune system that are going to make some swell antibodies for you over the next few weeks, but it makes even more sense to immediately activate parts of the immune system that are going to help you out right now. More immune cells are rushed into the circulation and, in the injured nervous system, more inflammatory cells infiltrate the site of injury. Moreover, circulating lymphocytes are better at releasing and responding to those immune messengers. And more of those generic antibodies of the innate immune system are released into your saliva. This boosting of immunity doesn’t occur only after some infectious challenge. Physical stressors, psychological stressors, all appear to cause an early stage of immune activation. Even more surprisingly, those immunosuppressive

villains, glucocorticoids, appear to play a major role in this (along with the sympathetic nervous system). So, with the onset of all sorts of stressors, your immune defenses are enhanced. And now we are ready for our usual other side of the two-edged sword, when the stress goes on longer. By the one-hour mark, more sustained glucocorticoid and sympathetic activation begins to have the opposite effect, namely, suppressing immunity. If the stressor ends around then, what have you accomplished with that immunosuppression? Bringing immune function back to where it started, back to baseline (phase B). It is only with major stressors of longer duration, or with really major exposure to glucocorticoids, that the immune system does not just return to baseline, but plummets into a range that really does qualify as immunosuppressing (phase C). For most things that you can measure in the immune system, sustained major stressors drive the numbers down to 40 to 70 percent below baseline.

Stress turns out to transiently stimulate the immune system.

The idea of temporarily perking up your immune system with the onset of a stressor makes a fair amount of sense (certainly at least as much as some of the convoluted theories as to why suppressing it makes sense). As does the notion that what goes up must come down. And as does the frequent theme of this book, namely, that if you have a stressor that goes on for too long, an adaptive decline back to baseline can overshoot and you get into trouble. Why did it take people so long to figure this out? Probably for two reasons. First, because many of the techniques for measuring what’s happening in the immune system have only recently become sensitive enough to pick up small, rapid differences, the thing needed to catch phase A, that fast immunostimulatory blip at the beginning of a stressor. Thus, for decades, people thought they were studying the immune response to stress, whereas they were actually studying the recovery of the immune response to stress. As a second reason, most scientists in this field study major, prolonged stressors, or administer major amounts of glucocorticoids for prolonged periods. This represents a reasonable bias in how experiments are done—start with a sledgehammer of an experimental manipulation. If nothing happens, pick a new field to study. If something does happen and it’s been replicated enough times that you’re confident about it, only then begin to think about more subtle elaborations. So in the early years, people were only

studying the sorts of stressors or patterns of glucocorticoid exposure that pushed into phase C, and only later got around to the subtler circumstances that would reveal phase B. This reorientation of the field represents a triumph for Allan Munck of Dartmouth University, one of the godfathers of the field, who predicted most of these new findings in the mid-1980s. He also predicted what turns out to be the answer to a question that pops up after a while. Why would you want to bring immune function back down to the prestress level (phase B in the diagram)? Why not just let it remain at the enhanced, improved level achieved in the first thirty minutes and get the benefits of an activated immune system all the time? Metaphorically, why not have your military that defends you always on maximal alert? For one thing, it costs too much. And, even more important, a system that’s always on maximal, hair-trigger alert is more likely to get carried away at some point and shoot one of your own guys in a friendly fire accident. And that’s what can happen with immune systems that are chronically activated—they begin to mistake part of you for being something invasive, and you’ve got yourself an autoimmune disease. Such reasoning led Munck to predict that if you fail to have phase B, if you don’t coast that activated immune system back down to baseline, you’re more at risk for an autoimmune disease. This idea has been verified in at least three realms. First, artificially lock glucocorticoid levels in the low basal range in rats and then stress them. This produces animals that have phase A (mostly mediated by epinephrine), but there isn’t the rise in glucocorticoids to fully pull off phase B. The rats are now more at risk for autoimmune disease. Second, doctors have to occasionally remove one of the two adrenal glands (the source of glucocorticoids) from a patient, typically because of a tumor. Immediately afterward, circulating glucocorticoid levels are halved for a period, until the remaining adrenal bulks up enough to take on the job of two. During that period of low glucocorticoid levels, people are more likely than normal to flare up with some autoimmune or inflammatory disease—there’s not enough glucocorticoids around to pull off phase B when something stressful occurs. Finally, if you look at strains of rats or, weirdly, chickens, that spontaneously develop autoimmune diseases, they all turn out to have something wrong with the glucocorticoid system so that they have lower than normal levels of the hormone, or have immune and inflammatory cells that are less responsive than normal to glucocorticoids. Same for humans with autoimmune diseases like rheumatoid arthritis.

A schematic representation of how a failure to inhibit immune function during stress can bias you toward autoimmune disease.

Thus, early on in the stress-response, the immune system is being activated, rather than inhibited, and a big thing that the stress-response does is make sure that immune activation doesn’t spiral into autoimmunity. So that has forced some revisionism in this field. But just to add to this, once stress has gone on long enough to begin to suppress immunity, some of what have classically been taken to be aspects of immune suppression are actually more subtle versions of immune enhancement. This is seen in two ways. Give someone massive amounts of glucocorticoids, or a huge stressor that has gone on for many hours, and the hormones will be killing lymphocytes indiscriminately, just mowing them down. Have a subtle rise in glucocorticoid levels for a short time (like what is going on at the start of phase B), and the hormones kill only a particular subset of lymphocytes— older ones, ones that don’t work as well. Glucocorticoids, at that stage, are helping to sculpt the immune response, getting rid of lymphocytes that aren’t ideal for the immediate emergency. So that indirectly counts as a version of immune enhancement. A second subtlety reflects reinterpretation of something people have known since the dawn of humans (or at least during Selye’s prime). As noted, glucocorticoids not only kill lymphocytes, but also yank some remaining lymphocytes out of the circulation. Firdhaus Dhabhar of Ohio State University asked, Where do those immune cells go when they are pulled out of the circulation? The assumption in the field had always been that they all go into immune storage tissues (like the thymus gland)—they’re taken out of action, so that they aren’t much use to you. But Dhabhar’s work shows that they don’t all get mothballed. Instead, glucocorticoids and epinephrine are diverting many of those lymphocytes to the specific site of infection, such as the skin. The immune cells aren’t being deactivated—they’re being transferred to the front lines. And a consequence of this is that wounds heal faster. Thus, early on during exposure to a stressor, glucocorticoids and other stress-responsive hormones transiently activate the immune system, enhancing immune defenses, sharpening them, redistributing immune cells to the scenes of infectious battle. Because of the dangers of the systems overshooting into autoimmunity, more prolonged glucocorticoid exposure begins to reverse these effects, bringing the system back to baseline. And during the pathological scenario of truly major, sustained stressors, immunity is suppressed below baseline. These new findings help to explain one of the persistent paradoxes in this field. It concerns autoimmune diseases. Two facts about autoimmunity:

1. Insofar as autoimmune diseases involve over activation of the immune system (to the point of considering a healthy constituent of your body to actually be something invasive), the most time-honored treatment for such diseases is to put people “on steroids”—to give them massive amounts of glucocorticoids. The logic here is obvious: by dramatically suppressing the immune system it can no longer attack your pancreas or nervous system, or whatever is the inappropriate target of its misplaced zeal (and, as an

obvious side effect to this approach, your immune system will also not be very effective at defending you against real pathogens). Thus, administration of large amounts of these stress hormones makes autoimmune diseases less damaging. Moreover, prolonged major stressors decrease the symptoms of autoimmune diseases in lab rats.

2. At the same time, it appears that stress can worsen autoimmune diseases. Stress is among

the most reliable, if not the most reliable, factor to worsen such diseases. This has often been reported anecdotally by patients, and is typically roundly ignored by clinicians who know that stress hormones help reduce autoimmunity, not worsen it. But some objective studies also support this view for autoimmune diseases such as multiple sclerosis, rheumatoid arthritis, Grave’s disease, ulcerative colitis, inflammatory bowel disease, and asthma. There have been only a handful of such reports, and they suffer from the weakness of relying on patient-reported retrospective data, rather than on prospective data. Nevertheless, their findings are relatively consistent—there is a subset of patients whose initial onset of an autoimmune disease and, to an even greater extent, their intermittent flare-ups of bad symptoms are yoked to stress. Moreover, there is, by now, a pretty hefty literature showing that stress can worsen autoimmunity in animal models of these diseases.

So, do glucocorticoids and stress worsen or lessen the symptoms of autoimmunity? The graph below gives an answer that wasn’t clear in earlier years. We’ve now seen two scenarios that increase the risk of autoimmune disease. First, it seems as if numerous transient stressors (that is, lots of phases A and B) increase the risk of autoimmunity—for some reason, repeated ups and downs ratchet the system upward, biasing it toward autoimmunity. Second, while it seems not to be great to have lots of instances of phase A followed by phase B, having phase A not followed by phase B increases the risk of autoimmunity as well. If you don’t have an adequate phase B, that pushes the immune system spiral upward into autoimmunity (diagram).

A schematic representation of how repeated stress increases the risk of autoimmune disease.

As we would now expect, if you instead have massive prolonged stressors, or are administered big hefty doses of glucocorticoids, you put the system in phase C—dramatic immune suppression, which decreases the symptoms of autoimmunity. Supporting this summary is the finding that while acute stress puts rats more at risk for a model of multiple sclerosis, chronic stress suppresses the symptoms of that autoimmune disease. The system apparently did not

evolve for dealing with numerous repetitions of coordinating the various on-and-off switches, and ultimately something uncoordinated occurs, increasing the risk that the system becomes autoimmune. Chronic Stress and Disease Risk A repeated theme in this book is how some physiological response to your average, run-of-the- mill mammalian stressor, if too long or too frequent, gets you into trouble. The ability of major stressors to suppress immunity below baseline certainly seems like a candidate for this category. How damaging is stress-induced immunosuppression when it actually occurs? As the AIDS virus has taught us, if you suppress the immune system sufficiently, a thirty-year-old will fester with cancers and pneumonias that doctors used to see once in an elderly patient during a fifty-year career. But can chronic stress suppress the immune system to the point of making you more susceptible to diseases you wouldn’t otherwise get? Once you have a disease, are you now less capable of fighting it off? Evidence pouring in from many quarters suggests that stress may indeed impair our immune systems and increase the risk of illness. But despite these striking findings, it remains far from clear just how much chronic stress makes you more vulnerable to diseases that would normally be fought off by the immune system. In order to appreciate the current disarray of the research, let us try to break down the findings into their component parts. Essentially, all these studies show a link between something that increases or decreases stress and some disease or mortality outcome. The approach of many psychoneuroimmunologists is based on the assumption that this link is established through the following steps: 1. The individuals in question have been stressed, 2. causing them to turn on the stress-response (the secretion of glucocorticoids, epinephrine, and so on). 3. The duration and magnitude of the stress-response in these individuals is big enough to suppress immune function, 4. which increases the odds of these individuals getting some infectious disease, and impairs their ability to defend themselves against that disease once they have it. Thus, suppose you see that a certain immune-related disease is more common in circumstances of stress. You now have to ask two critical questions. First, can you show that steps 1 to 4 occurred in those stressed individuals with that disease? Second, is there some alternative route that explains starting with stress and getting to the disease? Let’s begin by analyzing those four separate steps, in order to see how tough it is to demonstrate that all four have occurred.

Step 1, “The individuals in question have been stressed.” In studies of nonhuman animals, the general consensus is that with enough stress, you are going to get to steps 2 through 4. But a problem in extrapolating to humans is that the experimental stressors used in animal studies are usually more awful than what we typically experience. Not only that, but we differ tremendously among ourselves as to what we experience as truly stressful—the whole realm of individual differences that will be the focus of the last chapter of this book. Therefore, if you try to study the effects of stressors on people’s immune systems, you must wrestle with the problem of whether these things actually seem stressful to a given individual or not. What that winds up meaning is that step 1 is probably satisfied in stress/immune-related disease studies that involve events that most everyone would consider pretty awful—the death of a loved one, divorce, financially threatening unemployment. But if the external reality is one that a lot of people would not consider to be stressful, you can’t automatically accept that you’re at step 1. There is another problem with step 1: it’s often not clear whether humans are really exposed to the stressors to which they claim they’re exposed. We tend to be notoriously bad reporters of what goes on in our lives. An imaginary experiment: take one hundred lucky people and slip them a drug that will give them bad stomachaches for a few days. Then send them to a doctor secretly participating in this experiment, who tells them that they have developed stomach ulcers. The doctor asks innocently, “Have things been particularly stressful for you recently?” Perhaps ninety of those subjects will come up with something or other putatively stressful to which they will now attribute the ulcer. In retrospective studies, people confronted with an illness are very likely to decide there were stressful events going on. When you rely heavily on retrospective studies with humans, you are likely to get a falsely strong link between stress and disease; and the trouble is, most studies in this field are retrospective (a problem that popped up in the chapter on digestive disorders as well). The expensive and lengthy prospective studies are only recently becoming more common—pick a bunch of healthy people and follow them for decades to come, recording as an objective outsider when they are being exposed to stressors and whether they become sick. We move to the next step: from the stressor to the stress-response (step 1 to step 2). Again, if you give an organism a massive stressor, it will reliably have a strong stress-response. With more subtle stressors, we have more subtle stress-responses. The same thing holds for the move from step 2 to step 3. In experimental animal studies, large amounts of glucocorticoids will cause the immune system to hit the floor. The same occurs if a human has a tumor that causes massive amounts of glucocorticoids to be secreted (Cushing’s syndrome), or if a person is taking huge doses of synthetic glucocorticoids to control some other disease. But as we now know, the moderate rises in glucocorticoid levels seen in response to many more typical stressors stimulate the immune system, rather than suppress it. Moreover, in a few types of cancers elevated levels of glucocorticoids should be protective. As we saw in the last chapter, very high levels of glucocorticoids will suppress levels of estrogens in females and testosterone in males, and certain types of cancers are stimulated by these hormones (most notably “estrogen-sensitive” forms of breast cancer and “androgen-sensitive” prostate cancers). In these cases, lots of stress equals lots of glucocorticoids equals less estrogen or testosterone equals slower tumor growth.

Moving from step 3 to step 4, how much does a change in immune profile alter patterns of disease? The odd thing is that immunologists are not sure about this. If your immune system is massively suppressed, you are more likely to get sick, no doubt about that. People taking high doses of glucocorticoids as medication, who are thus highly immunocompromised, are vulnerable to all sorts of infectious diseases, as are people with Cushing’s syndrome. Or AIDS. The more subtle fluctuations in immunity are less clear in their implications, however. Few immunologists would be likely to assert that “for every tiny decrease in some measure of immune function, there is a tiny increase in disease risk.” Their hesitancy is because the relationship between immune competence and disease may be nonlinear. In other words, once you pass a certain threshold of immunosuppression, you are up the creek without a paddle; but before that, immune fluctuations may not really matter much. The immune system is so complex that being able to measure a change in one little piece of it in response to stress may mean nothing about the system as a whole. Thus, the link between relatively minor immune fluctuation and patterns of disease in humans winds up being relatively weak.

There is another reason why it may be difficult to generalize from findings in the laboratory to the real world. In the laboratory, you might be studying the effects of steps 1, 2, and 3 on disease outcome 4. It is inconvenient for most scientists to manipulate a rat’s levels of stress, glucocorticoids, or immunity and then wait for the rest of the rat’s lifetime to see if it is more likely to become ill than is a control rat. That’s slow and expensive. Typically, instead, scientists study induced diseases. Manipulate step 1, 2, or 3 in a rat that has been exposed to a certain virus; then see what happens. When you do that, you get information about steps 1 through 3 that have to do with step 4 when dealing with severe, artificially induced disease challenges. But it should be obvious that an approach like this misses the point that we don’t get sick because some scientist deliberately exposes us to disease. Instead, we spend our lives passing through a world filled with scattered carcinogenic substances, occasional epidemics, someone sneezing from

across the room. Relatively few experimental animal studies have looked at spontaneous diseases, rather than induced ones. These are a lot of caveats. Let’s consider some areas where there are links between stress and diseases associated with immune dysfunction. This will let us evaluate to what extent these links are a function of progressing from steps 1 through 4, what we will call the “Psychoneuroimmune Route,” which links stress and disease. In each case, we’ll consider if there is an alternative sequence, what we’ll loosely call the “Lifestyle Route,” which can link stress and immune- related disease while bypassing the sequence of steps 1 to 4.

Testing the Stress-Disease Link Social Support and Social Isolation What the data show: the fewer social relationships a person has, the shorter his or her life expectancy, and the worse the impact of various infectious diseases. Relationships that are medically protective can take the form of marriage, contact with friends and extended family, church membership, or other group affiliations. This is a fairly consistent pattern that cuts across a lot of different settings. Moreover, these general findings are based on some careful prospective studies and are seen in both sexes and in different races, in American and European populations living in both urban and rural areas. Most important, this effect is big. The impact of social relationships on life expectancy appears to be at least as large as that of variables such as cigarette smoking, hypertension, obesity, and level of physical activity. For the same illness, people with the fewest social connections have approximately two-and-a-half times as much chance of dying as those with the most connections, after controlling for such variables as age, gender, and health status. Very exciting. And what might explain this relationship? Maybe it’s through the Psychoneuroimmune Route of steps 1 to 4, which would run something like: socially isolated people are more stressed for lack of social outlets and support (step 1); this leads to chronic activation of stress-responses (step 2); leading to immune suppression (step 3); and more infectious diseases (step 4). Let’s see what support there is for each of these steps. First, just because someone is socially isolated doesn’t mean they are stressed by it—there are lots of hermits who would be happy to pass on yet another crowded Twister party. Social isolation as a stressor is a subjective assessment. In many of these studies, however, the subjects who fit the bill as socially isolated rate themselves as lonely, certainly a negative emotion. So we can check off step 1. On to step 2—do these people have chronically overactive stress-responses? We have little evidence for or against that. How about step 3—is social isolation associated with damping down some aspect of immune function? There’s a lot of evidence for that: lonelier, more socially isolated individuals having less of an antibody response to a vaccine in one study; in another study of people with AIDS, having a faster decline in a key category of lymphocytes; in another, of women with breast cancer, having less natural killer cell activity.

Then on to step 4—can you actually show that that degree of immune suppression played a role in the disease occurring? The facts are relatively weak. Some studies show social isolation and step 3; others show isolation and step 4, but few show both and also explicitly show that the magnitude of step 3 has something to do with the transition to step 4. Still, there is relatively good evidence for this pathway being relevant. What about the Lifestyle Route? What if the problem is that socially isolated people lack that special someone to remind them to take their daily medication? It is known that isolated people are less likely to comply with a medical regime. What if they’re more likely to subsist on reheated fast food instead of something nutritious? Or more likely to indulge in some foolish risk-taking behavior, like smoking, because there’s no one to try to convince them to stop? Many lifestyle patterns could link social isolation with more infectious disease, bypassing this sequence of steps. Or what if the causality is reversed—what if the linkage occurs because sickly people are less likely to be able to maintain stable social relationships? Numerous studies have controlled for these lifestyle risk factors like smoking, diet, or medication compliance and have shown that the isolation/poor health outcome relationship is still there. Moreover, critically, you can show the same in nonhuman primates, who don’t confound their health with Big Macs, alcohol, and smoking. Infect monkeys with SIV (the simian equivalent of HIV) and more socially isolated animals had higher glucocorticoid levels, fewer antibodies against the virus, more virus in their system, and a greater mortality rate—in other words, steps 1 to 4. Overall, I’d say a pretty good case can be made that social isolation can impact health through the effects of stress on immunity. But the case isn’t airtight. Bereavement Bereavement, an extreme version of social isolation, is, of course, the loss of a loved one. An extensive literature shows that while bereavement often coincides with depression, it is distinct from it. A common belief is that the one left behind—the grieving spouse, the bereft parent, even the masterless pet—now pines away to an early death. A number of studies uggest that bereavement does indeed increase the risk of dying, although the effect is not all that strong. This is probably because the risk occurs only in a subset of grievers, amid those people who have an additional physiological or psychological risk factor coupled with the bereavement. In one careful prospective study, the parents of all the Israeli soldiers who died in the Lebanese war were followed for ten years afterward. Loss of a child did not affect mortality rates in the population of grieving parents in general. However, significantly higher mortality rates occurred among parents who were already widowed or divorced. In other words, this stressor is associated with increased mortality in the subset of parents with the added risk factor of minimal social support. Thus, we are turfed back to the subject of social isolation. Again, the evidence for the Psychoneuroimmune Route occurring is decent but, again, there are many potential Lifestyle Routes—grieving people are unlikely to be eating, sleeping, exercising in a healthy manner. Sometimes the confound is more subtle. People tend to marry people who are ethnically and

genetically quite similar to themselves. Intrinsic in this trend toward “homogamy” is a tendency of married couples to have higher-than-random chances of sharing environmental risk factors (as well as to disproportionately share disease-related genes, making this component of the Lifestyle Route not really related to lifestyle). This makes it more likely that they will get sick around the same time. Nonetheless, amid those confounds, the Psychoneuroimmune Route’s steps 1 to 4 are probably relevant to the increased mortality rates among bereaved individuals lacking social support. The Common Cold Everybody knows that being stressed increases your chances of getting a cold. Just think back to being run down, frazzled, and sleep-deprived during final exams, and, sure enough, there’s that cough and runny nose. Examine the records at university health services and you’ll see the same thing—students succumbing to colds left and right around exam period. Many of us continue to see the same pattern decades later—burn the candle at both ends for a few days and, suddenly, there’s that scratchy throat. Psychoneuroimmune Route steps 1 to 4 seem quite plausible. Some of the studies involve some pretty hefty external events that most people would consider stressful, like financially disastrous unemployment (step 1). But few have looked at the magnitude of the stress-response (step 2). Changes in relevant immune measures have been documented, however—for example, in studies in which stress increases the risk of a cold, those same individuals are shown to have less of the cold-fighting class of antibodies that are secreted in your saliva and nasal passageways (steps 3 and 4). But we have to consider some possible Lifestyle Route confounds. Maybe the disruptive effects of stress on memory (stay tuned for chapter 10) cause us to forget to button up our overcoats. Or maybe when we are under stress due to social isolation, we are more willing to consort with people who sneeze recklessly without covering their faces. Okay, maybe those aren’t confounds you have to worry about too much. But stress changes lifestyle and different lifestyles mean differing degrees of exposure to the viruses that cause colds. That possibility has been controlled for in a celebrated series of studies. In one version, some cheerfully compliant volunteers were housed under conditions where some major lifestyle confounds were controlled for. They then filled out questionnaires regarding how stressed they were. Subjects were then spritzed up their noses with equal amounts of rhinovirus, the bug that causes the common cold. Note that everyone was exposed to the same amount of pathogen. And the results? (Fanfare.) More stress equaled about three times the likelihood of succumbing to a cold after being exposed to the virus. Prolonged stressors more than a month long that were social in nature provided the greatest risk.* Moreover, the same thing works in laboratory mice and nonhuman primates—spritz them with rhinovirus, and it is the stressed, socially subordinate animals who get their species’ equivalent of the sniffles.

Collectively, it seems pretty convincing that stress makes the common cold more common at least partially along the Psychoneuroimmune Route. Aids Given that AIDS is a disease of profound immunosuppression, and that major stressors suppress the immune system, can stress increase the likelihood that someone who is HIV positive develops AIDS? And once AIDS is established, can stress worsen its course? These questions have been aired since the AIDS epidemic began. Since the last edition of this book, the triple combination antiretroviral therapy has turned AIDS from a fatal disease to an often manageable chronic one, making these questions even more relevant.* There is some good indirect evidence to think that stress can alter the course of AIDS. Suppose you grow human lymphocytes in a petridish and expose them to HIV. If you expose the cells to glucocorticoids as well, they become more likely to be infected with the virus. Moreover, norepinephrine can also make it easier for the virus to invade a lymphocyte and, once inside, enhances replication of the virus. Support also comes from a study with nonhuman primates, discussed earlier, which suggests that steps 1 to 4 might apply to HIV. To reiterate, the monkeys were infected with SIV, the simian version of HIV. The authors then showed that the monkeys who were more socially isolated (step 1) had higher glucocorticoid levels (step 2), fewer antibodies against the virus (step 3), and a higher mortality rate (step 4). How about humans? To begin, starting with the same amount of HIV in your system, a faster decline and a higher mortality rate occur, on average, among people who have any of the following: (a) a coping style built around denial; (b) minimal social support; (c) a socially inhibited temperament; (d) more stressors, particularly loss of loved ones. These are not huge effects but, nevertheless, there seems a fair consistency in the findings on this. So that seems to qualify for step 1. Do these individuals also have overactive stress-responses (step 2)? Glucocorticoid levels are not particularly predictive of the course of HIV. However, the more at-risk people with the socially inhibited temperaments have elevated activity of their sympathetic nervous system, and the extent of that overactivity is an even better predictor of decline than is the personality itself. So that seems to get us to step 2. Does lots of stress, an inhibited temperament, denial, or lack of social support not only predict higher mortality rates (step 4) but a faster decline of immune function (step 3)? That seems to be the case as well. So AIDS seems to follow the Psychoneuroimmune Route. How about the Lifestyle Route? The medication regimes for dealing with HIV can be enormously complex, and it is quite plausible that people who are more stressed are less likely to take their antiviral medication, or to take it correctly. My sense is that lifestyle risk factors have not been all that well controlled for in these studies. How about if the connection runs in the opposite direction—what if having a faster decline with the disease makes you more socially inhibited, makes for fewer social connections?

That seems quite plausible but, as an important control, the personality style has been shown to predict immune profiles many months later. In summary, psychoneuroimmune aspects could well contribute to a link between stress and worsening of aspects of AIDS. But more research needs to be done to examine how much stress influences whether people comply with their treatment regimes, versus how well their treatment regimes work. Latent Viruses After rhinoviruses and the AIDS virus, there is one last category of viruses—those that, after initially infecting you, can go latent. “Latency” means that the virus, once burrowing into some cells of yours, goes into hibernation for a while, just lurking near your own cellular DNA, but not yet replicating itself. At some later point, something triggers the dormant virus out of latency and it reactivates. After going through a couple of rounds of replication the by now larger number of viral particles burrow in and go latent again. The classic example are herpes viruses which, after infecting some of your neurons, can go latent for years, even decades, before flaring up out of latency. This is a clever tactic that viruses have evolved. Infect some cells, replicate, burst the cells open in the process, make the sort of mess of things that sets off all sorts of alarms in the immune system and, just as those activated immune cells are about to pounce, burrow into another round of cells. While the immune cells are cleaning up, the virus goes latent again. The next clever thing that viruses have done? They don’t reactivate at any old time. They wait until the immune system of the host organism is lousy, and then gun for some quick rounds of replication. And when are immune systems often at their lousiest? You got it. It’s been endlessly documented that latent viruses like herpes flare up during times of physical or psychological stress in all sorts of species. It’s the same thing with some other viruses that go latent, like Epstein-Barr virus and varicella-zoster (which causes chicken pox and shingles). So hats off to these highly evolved viruses. Now a key question. How does a latent herpes virus that, after all, is just some unschooled little stretch of DNA sitting mothballed inside a bunch of your neurons, know that you are immunosuppressed? One possibility is that herpes is always attempting to come out of latency and, if your immune system is working fine, it snuffs out the attempt. A second possibility is that herpes can somehow measure how the immune system is doing. Amazingly, the answer has emerged in the last few years. Herpes doesn’t measure how your immune system is doing. It measures something else that, for its purposes, gives it the information it needs—it measures your glucocorticoid levels. Herpes DNA contains a stretch that is sensitive to elevated glucocorticoid signals, and when levels are up, that DNA sensor activates the genes involved in coming out of latency. Epstein-Barr and varicella-zoster contain this glucocorticoid-sensitive stretch as well.

And now for something even more fiendishly clever. You know what else herpes can do once it infects your nervous system? It causes your hypothalamus to release CRH which releases ACTH which raises glucocorticoid levels. Unbelievable, huh? So you don’t even need a stressor. Herpes infects you, artificially pushes you to step 2 with your elevated glucocorticoid levels, which gets you to step 3, and allows the virus to come out of latency. Moreover, elevated glucocorticoid levels impair your immune defenses against activated herpes. This leads to step 4—a cold sore flare-up. And we think we’re so clever with our big brains and opposable thumbs. We’ve now looked at several favorite topics in psychoneuroimmunology, and can see that stress can increase the likelihood, the severity, or both of some immune-related diseases. All of this is a prelude for considering the most contentious subject in this whole field. The punch line is one of the most important in this book, and runs counter to what is distressingly common folk wisdom.

Stress and the Big C What does stress have to do with getting cancer? The first piece of evidence suggesting stress may increase the risk of a cancer diagnosis comes from animal studies. There is, by now, a reasonably convincing animal-experimentation literature showing that stress affects the course of some types of cancer. For example, the rate at which some tumors grow in mice can be affected merely by what sort of cages the animals are housed in—the more noisy and stressful, the faster the tumors grow. Other studies show that if you expose rats to electric shocks from which they can eventually escape, they reject transplanted tumors at a normal rate. Take away the capacity to escape, yet give the same total number of shocks, and the rats lose their capacity to reject tumors. Stress mice by putting their cages on a rotating platform (basically, a record player), and there is a tight relationship between the number of rotations and the rate of tumor growth. Substitute glucocorticoids for the rotation stressor, and tumor growth is accelerated as well. These are the results of very careful studies performed by some of the best scientists in the field. Does stress work through the Psychoneuroimmune Route in these animals? Seemingly at least partially. These stressors raise glucocorticoid levels in these studies. And these glucocorticoids directly influence tumor biology through both immune and non-immune realms. As a first mechanism, the immune system contains a specialized class of cells (most notably, natural killer cells) that prevent the spread of tumors. Stress suppresses the numbers of circulating natural killer cells in these studies. A second route is probably non-immunologic. Once a tumor starts growing, it needs enormous amounts of energy, and one of the first things that tumors do is send a signal to the nearest blood vessel to grow a bush of capillaries into the tumor. Such angiogenesis allows for the delivery of blood and nutrients to the hungry tumor. Glucocorticoids, at the concentration generated during stress, aid angiogenesis. A final route may involve glucose delivery. Tumor cells are very good at absorbing glucose from the bloodstream. Recall the zebra sprinting away from the lion: energy storage has stopped in order to increase concentrations of circulating glucose to be used by the muscles. But, as my own lab reported some years back, when circulating glucose concentrations are elevated in rats during stress, at least one kind of experimental tumor can grab the glucose before the muscle does. Your storehouses of energy,

intended for your muscles, are being emptied and inadvertently transferred to the ravenous tumor instead. So we have some stress-cancer links in animals, and some psychoneuroimmune mechanisms to explain those effects. Does this apply to humans? Two big features of these animal studies dramatically limit their relevance to us. First, these were studies of induced tumor, where tumorous cells are injected or transplanted into the animal. So we’re not looking at stress causing cancer in these animals, we’re looking at stress altering the course of cancers introduced by artificial routes. No animal studies to my knowledge have shown that stress increases the incidence of spontaneous tumors. Furthermore, most of these studies have relied on tumors that are caused by viruses. In such cases, viruses take over the replication machinery of a cell and cause it to start dividing and growing out of control. In humans most cancers arise from genetic factors or exposure to environmental carcinogens, rather than from viruses, and those have not been the subject of study with laboratory animals. So a cautionary note from the animal studies: stress can accelerate the growth of a number of tumors, but these are types of cancers of limited relevance to humans, and introduced through completely artificial means. Thus, we turn our attention to humans. Our first, simplest question: Is a history of major stressors associated with an increased risk of having cancer somewhere down the line? A number of studies seemed to show this, but they all suffered from the same problem, namely, that they were retrospective. Again, someone with a cancer diagnosis is more likely to remember stressful events than someone with a bunion. How about if you do a retrospective study where you rely upon a history of verifiable stressors, like the death of a family member, loss of a job, or a divorce? A couple of studies have reported a link between such major stressors and the onset of colon cancer five to ten years later. A number of studies, especially of breast cancer patients, have had a “quasi-prospective” design, assessing stress histories of women at the time that they are having a biopsy for a breast lump, comparing those who get a cancer diagnosis with those who don’t. Some of these studies have shown a stress-cancer link, and this should be solid—after all, there can’t be a retrospective bias, if the women don’t know yet if they have cancer. What’s the problem here? Apparently, people can guess whether it will turn out to be cancer at a better than chance rate, possibly reflecting knowledge of a family history of the disease, or personal exposure to risk factors. Thus, such quasi-prospective studies are already quasi-retrospective, and of the least reliable kind. When you rely on the rare prospective studies, there turns out not to be good evidence for a stress-cancer link. For example, as we will see in chapter 14 on depression, having a major depression is closely linked to both stress and excessive glucocorticoid secretion, and one famous study of two thousand men at a Western Electric plant showed that depression was associated with doubling the risk of cancer, even up to decades later. But a careful reexamination of those data showed that the depression-cancer link was attributable to a subset of men who were depressed as hell because they were stuck working with some major carcinogens. Subsequent prospective studies of other populations have shown either no depression/cancer link, or a tiny one that is biologically meaningless. Moreover, these studies have not ruled out the alternative Lifestyle Route, in that depressed people smoke and drink more, two routes to

increase the risk of cancer. Similar findings emerge from the careful prospective studies of bereavement as a stressor—no link with subsequent cancer. Thus, we shift to a different literature. We’ll be seeing in chapter 11 how sleep deprivation and altered sleep patterns (such as with night shifts) are major stressors. In searching for a link between stress and increased risk of cancer, it may not be surprising to find that women who have spent long periods (decades in these studies) working night shifts have an increased risk of breast cancer. However, the most plausible explanation here has nothing to do with stress. Instead, a shifted day/night schedule dramatically decreases the level of a light-responsive hormone called melatonin, and depletion of this hormone greatly increases the risk of a number of types of cancer, including breast cancer. More suggestive links go by the wayside as well. As discussed earlier, individuals who get organ transplants are at risk for rejecting them, and one of the prevention strategies is to give them glucocorticoids in order to suppress the immune system past the point of being able to reject the organ. In a small subset of such individuals, there is an increased incidence of a few types of skin cancer (of the less serious, non-melanoma kind). Moreover, as noted, if someone’s immune system is massively suppressed because of AIDS, there is an increased incidence of a handful of types of cancers. So do these findings tighten the links between cancer and stress? No. This is because: (a) stress never suppresses the immune system to that extent; (b) even when the immune system is suppressed that much, only a small subset of organ transplant or AIDS patients get cancer; and (c) it is only a tiny subset of cancers that now become more common. So besides those two reports about colon cancer, there is no particular support for the idea that stress increases the risk of cancer (and, it should be noted, this conclusion includes numerous studies of breast cancer, the type of cancer most frequently assumed by people to be stress related). But is there a subset of individuals who have a particular (and poor) style of coping with stress that puts them more at risk for cancer? We already saw, in chapter 5, the notion of there being personality types that are more prone toward functional gastrointestinal disorders. Is there a cancer-prone personality, and can it be interpreted in the context of coping poorly with stress? Some scientists think so. Much of the work in this area has been done with breast cancer, in part because of the prevalence and seriousness of the disease. However, the same pattern has been reported for other cancers as well. The cancer-prone personality, we’re told, is one of repression—emotions held inside, particularly those of anger. This is a picture of an introverted, respectful individual with a strong desire to please—conforming and compliant. Hold those emotions inside and it increases the likelihood that out will come cancer, according to this view. Most of these studies have been retrospective or quasi-prospective, and we have seen the problems endemic to such studies. Nonetheless, the prospective studies have shown there to be some link, though a small one. Are we in the realm of Psychoneuroimmune Route steps 1 through 4? No one has shown that yet, in my opinion. As we will see in chapter 15, a repressed personality is associated with elevated glucocorticoid levels, so we’re in the range of step 2. But, to my knowledge, no one has shown evidence for step 3—some sort of immune suppression—occurring, let alone it being of a

magnitude relevant to cancer. In addition, none of the good prospective studies have ruled out the Lifestyle Route (such as smoking, drinking, or, in the case of breast cancer, more fat consumption). So the jury remains out on this one. So collectively, we have, with the exception of two studies concerning one type of cancer, no overall suggestion that stress increases the risk of cancer in humans. Stress and Cancer Relapse What if your cancer has been cured? Does stress increase the risk of it coming back? The handful of studies on this subject don’t suggest that there’s a connection—a few say yes, an equal number, no. Stress and the Course of Cancer Now on to the most complex and controversial issue of all. Sure, stress may not have anything to do with whether you come down with cancer, but once you have cancer, will stress make a tumor grow faster, increasing your risks of dying from the disease? And can stress reduction slow down tumor growth, extending survival times? As we saw above, stress will accelerate tumor growth in animals, but those types of instigated tumors and their biology are of limited relevance to the way humans get cancer. So we have to look at studies of humans. And here the subject is a mess. We begin by looking at whether different coping styles predict different cancer outcomes. When you compare patients who respond to their cancer with a “fighting spirit” (that is, they are optimistic and assertive) with those who collapse into depression, denial, and repression, the former live longer, after controlling for cancer severity. Findings like these prompted studies in which clinicians attempted to intervene, to reduce stress and inculcate more of that fighting spirit in people, in order to influence the patient’s cancer outcome. The landmark study of this type was carried out in the late 1970s by the psychiatrist David Spiegel of Stanford University. Women who had just gotten a metastatic breast cancer diagnosis were randomly assigned to either a group that received standard medical care or a group that, in addition, had intensive supportive group psychotherapy with other breast cancer patients. As Spiegel has emphasized in his accounts of this famous study, he went into it anticipating that the group therapy intervention might decrease psychological distress in patients, but he certainly didn’t expect that it would affect the biology of the cancer. Amid his skepticism, what he found was that the group therapy intervention extended life span an average of eighteen months, a whopping great effect. This made front-page news. But there’s been a big problem since then—it’s just not clear if a psychosocial intervention actually works. Since the Spiegel study, there have been roughly a dozen others, and they are evenly split as to whether there is any protective effect from group therapy. In what was probably the most thorough attempt at a replication of Spiegel’s findings, a

study published in 2001 in the prestigious New England Journal of Medicine, there was no effect on survival time. Why has this finding been so difficult to replicate? Spiegel and others give a plausible explanation, having much to do with the massive changes that have occurred over the years in the “culture of cancer.” Not that many decades ago, getting cancer had a weirdly shameful quality to it—doctors wouldn’t want to tell their patients about the embarrassing and hopeless diagnosis; patients would hide having the disease. As one example, in a 1961 survey, a boggling 90 percent of American physicians said they did not typically reveal a cancer diagnosis to their patients; within two decades, the number was down to 3 percent. Moreover, over the years, doctors have come to consider the psychological well-being of their patients as essential to fighting the cancer, and see the course of medical treatment as a collaboration between themselves and the patient. As Spiegel says, when he began his work in the 1970s, the biggest challenge was to get patients in the “experimental” group to be willing to waste their time with something as irrelevant as group therapy. In contrast, by the 1990s versions of these studies, the biggest challenge was to convince the “control” subjects to forgo group therapy. In this view, it has become difficult to show that introducing a stress-reducing psychosocial intervention extends cancer survival over control subjects because everyone, including control subjects, now recognizes the need to reduce stress during cancer treatment, and seeks psychosocial support all over the place, even if it doesn’t come with an official stamp of “twice weekly group psychotherapy.” Let’s assume this explanation is correct, and I do find it to be convincing. Thus we accept the premise that psychosocial interventions that reduce stress do extend cancer survival. Let’s grind through the steps of the Psychoneuroimmune Route to see if we can understand why the group therapy has such an effect. Are the psychosocial interventions perceived as being stress reducing by the patients (step 1)? There are striking individual exceptions, but the studies, overall, show this resoundingly to be the case. Are those psychosocial interventions associated with a damping of the stress-response (step 2)? A few studies have shown that psychosocial interventions can lower glucocorticoid levels. Flip the question the other way—does having an overactive stress-response predict shorter cancer survival? No. In the most detailed study of this, following a subsequent population of Spiegel’s metastatic breast cancer patients, having high glucocorticoid levels around the time of diagnosis didn’t predict a shorter survival time.* So while psychosocial interventions can reduce glucocorticoid levels, there’s little evidence that elevated glucocorticoid levels predict shorter cancer survival. But do cancer patients with more psychosocial support have better immune function (step 3)? Seemingly. Breast cancer patients who reported more stress had lower activities of those natural killer cells, while there’s higher NK cell activity in women who report more social support or who received some sort of group therapy intervention. Were those immune changes relevant to the change in survival time (step 4)? Probably not, since someone’s levels of NK cell activity didn’t predict survival times in these studies.

So there’s not much evidence for a Psychoneuroimmune Route. How about the Lifestyle Route? There are lots of reasons to think lifestyle plays a key role in the link between stress and the course of cancer, but it’s very hard to show, for a subtle reason. One of the great confounds in cancer therapy is that about a quarter of cancer patients don’t take their medications as often as prescribed, or miss chemotherapy appointments. Go figure, when these treatments make you feel so so awful. And what happens in a group therapy setting, when you’re surrounded by people going through the same hell as you? “You can go the extra round of chemo, I know you can— yeah, I felt awful the whole time during mine, but you can do it, too,” or “Have you eaten today? I know, I have no appetite either, but we’re going to get something to eat right after this,” or “Have you taken your meds today?” Compliance goes up. Any sort of intervention that increases compliance will increase the success rates of treatments. And because a cancer patient, reasonably, would often be very uncomfortable about admitting that she’s not completely complying with a treatment regime, it’s hard to detect accurately whether any of the protective effects of psychosocial therapy are kicking in through this route.* What we have here are some extremely interesting but murky waters. There appears to be virtually no link between a history of a lot of stress and a greater incidence of cancer, or a greater risk of relapse out of remission. There seems to be a link between a certain personality type and a somewhat greater cancer risk, but no studies have shown where stress physiology fits into that story, nor have lifestyle confounds been ruled out. Next, the findings are about evenly divided as to whether psychosocial interventions that reduce stress improve cancer outcomes. Finally, when considering the cases where psychosocial intervention is effective, there’s little support for a Psychoneuroimmune Route to explain the efficacy, and good reasons to think that an alternative route involving issues of lifestyle and compliance is important. What does one do with these findings? Right on cue—more research, of course. Lots more. However, it is time to discuss what one should not do with these findings in the meantime. Cancer and Miracles This leads to a tirade. Once we recognize that psychological factors, stress-reducing interventions, and so on can influence something like cancer, it is often a hopeful, desperate leap to the conclusion that such factors can control cancer. When that proves to be false, there is a corrosive, poisonous flip side: if you falsely believe you had the power to prevent or cure cancer through positive thinking, you may then come to believe that it is your own fault if you are dying of the disease. The advocates of a rather damaging overstatement of these psychology-health relationships are not always addled voices from the lunatic fringe. They include influential health practitioners whose medical degrees appear to lend credence to their extravagant claims. I will focus my attention here on the claims of Bernie S. Siegel, a Yale University surgeon who has been wildly effective at disseminating his ideas to the public as the author of a bestseller. The premise of Siegel’s still-popular magnum opus, Love, Medicine and Miracles (New York: Harper & Row, 1986), is that the most effective way of stimulating the immune system is

through love, and that miraculous healing happens to patients who are brave enough to love. Siegel purports to demonstrate this. As the book unfolds, you note that it is a strange world that Siegel inhabits. When operating on anesthetized patients, “I also do not hesitate to ask the [anesthetized] patient not to bleed if circumstances call for it,” he asserts. In his world, deceased patients come back as birds, there are unnamed countries in which individuals consistently live for a century, and best of all, people who have the right spirituality not only successfully fight cancer but can drive cars that consistently break down for other people. This is relatively benign gibberish, and history buffs may even feel comforted by those among us who live the belief system of medieval peasants. Where the problems become appallingly serious is when Siegel concentrates on the main point of his book. No matter how often he puts in disclaimers saying that he’s not trying to make people feel guilty, the book’s premise is that (a) cancer can be caused by psychosocial factors in the person; (b) cancer (or any other disease, as far as I can tell) is curable if the patient has sufficient courage, love, and spirit; (c) if the patient is not cured, it is because of insufficient amounts of those admirable traits. As we have just seen, this is not how cancer works, and a physician simply should not go about telling seriously ill people otherwise. His book is full of descriptions of people who get cancer because of their uptightness and lack of spirituality. He speaks of one woman who was repressed in her feelings about her breasts: “Naturally [my emphasis], Jan got breast cancer”—this seems an indication that Siegel is aware of the literature on cancer-prone personality, but this constitutes a caricature of those mostly careful studies. Of another patient: “She held all her feelings inside and developed leukemia”. Or, in an extraordinary statement: “Cancer generally seems to appear in response to loss…I believe that, if a person avoids emotional growth at this time, the impulse behind it becomes misdirected into malignant physical growth”. Naturally, those who do have enough courage, love, and spirit can defeat cancer. Sometimes it takes a little prodding from Siegel. He advises in chapter 6 that people with serious diseases consider the ways in which they may have wanted their illness because we are trained to associate sickness with reward; Siegel cites our receiving cards and flowers in Chapter 6. Sometimes Siegel has to be a bit more forceful with a recalcitrant cancer patient. One woman was apparently inhibited about drawing something Siegel requested her to, being embarrassed about her poor drawing skills. “I asked [her] how she expected to get over cancer if she didn’t even have the courage to do a picture”. You know whose fault it was if she eventually died. But once the good patients overcome their attitude problems and get with the program, miracles just start popping up everywhere you look. One patient with the proper visualizing techniques cured his cancer, his arthritis, and, as long as he was at it, his twenty-year problem with impotency as well. Of another, Siegel writes: “She chose the path of life, and as she grew, her cancer shrank”. Consider the following exchange:

I came in, and he said, “Her cancer’s gone.”

“Phyllis,” I said, “Tell them what happened.” She said, “Oh, you know what happened.” “I know that I know,” I said, “But I’d like the others to know.” Phyllis replied, “I decided to live to be a hundred and leave my troubles to God.” I really could end the book here, because this peace of mind can heal anything.

Thus, presumably, people who die from cancer never got around to deciding to live to be a hundred. According to Siegel, cancer is curable with the right combination of attributes, and those people without them may get cancer and die of it. An incurable disease is the fault of the victim. He tries to soften his message now and then: “Cancer’s complex causes aren’t all in the mind,” he says, and in chapter 5 he tells us he’s interested in a person gaining understanding of his or her role in a disease rather than in creating guilt. But when he gets past his anecdotes about individual patients and states his premise in its broadest terms, its toxicity is unmistakable: “The fundamental problem most patients face is an inability to love themselves” “I feel that all disease is ultimately related to a lack of love”. Siegel has a special place in his book for children with cancer and for the parents of those children trying to understand why it has occurred. After noting that developmental psychologists have learned that infants have considerably greater perceptual capacities than previously believed, Siegel says he “wouldn’t be surprised if cancer in early childhood was linked to messages of parental conflict or disapproval perceived even in the womb”. In other words, if your child gets cancer, consider the possibility that you caused it.* And perhaps most directly: “There are no incurable diseases, only incurable people”. (Compare the statement by the late stress researcher Herbert Weiner: “Diseases are mere abstractions; they cannot be understood without appreciating the person who is ill.” Superficially, Siegel’s and Weiner’s notions bear some resemblance to each other. The latter, however, is a scientifically sound statement of the interactions between diseases and individual makeups of sick people; the former seems to me an unscientific distortion of those interactions.) Since at least the Middle Ages, there has been a philosophical view of disease that is “lapsarian” in nature, characterizing illness as the punishment meted out by God for sin (all deriving from humankind’s lapse in the Garden of Eden). Its adherents obviously predated any knowledge about germs, infection, or the workings of the body. This view has mostly passed (although see the endnote for this page for an extraordinary example of this thinking that festered in the Reagan administration), but as you read through Siegel’s book, you unconsciously wait for its reemergence, knowing that disease has to be more than just not having enough groovy New Age spirituality, that God is going to be yanked into Siegel’s world of blame as well. Finally, it bubbles to the surface in chapter 8: “I suggest that patients think of illness not as God’s will but as our deviation from God’s will. To me it is the absence of spirituality that leads to difficulties.” Cancer, thus, is what you get when you deviate from God’s will.

Oh, and one other thing about Siegel’s views. He founded a cancer program called Exceptional Cancer Patients, which incorporates his many ideas about the nature of life, spirit, and disease. To my knowledge there have been only two published studies of his program and its effects on survival time. Both reported that the program has no significant effect on survival. And one last word from the good doctor, washing his hands of the first study (the second had not yet been published when he wrote his book): “I prefer to deal with individuals and effective techniques, and let others take care of the statistics.” Why is it worth going on at length about this subject, to pay so much attention to a more-than- fifteen-year-old book? Because of how influential Siegel’s style of thinking has been. Here is but one chilling example: in one study, breast cancer patients were asked what they thought had caused their cancer. Among the hundreds of participants, answers came back such as genetics, environment, hormones, diet, and breast trauma. And what was the most common attribution, by a wide margin? Stress. This, in a paper published in 2001, at the dawn of our new millennium. This topic is one that I will return to in the final chapter of the book when I discuss stress management theories. Obviously, a theme of this book is just how many things can go wrong in the body because of stress and how important it is for everyone to recognize this. However, it would be utterly negligent to exaggerate the implications of this idea. Every child cannot grow up to be president; it turned out that merely by holding hands and singing folk songs we couldn’t end all war, and hunger does not disappear just by visualizing a world without it. Everything bad in human health now is not caused by stress, nor is it in our power to cure ourselves of all our worst medical nightmares merely by reducing stress and thinking healthy thoughts full of courage and spirit and love. Would that it were so. And shame on those who would profit from selling this view. Postscript: A Grotesque Piece of Medical History The notion that the mind can influence the immune system, that emotional distress can change resistance to certain diseases, is fascinating; psychoneuroimmunology exerts a powerful pull. Nevertheless, it sometimes amazes me just how many psychoneuroimmunologists are popping up. They are even beginning to speciate into subspecialties. Some study the issue only in humans, others in animals; some analyze epidemiological patterns in large populations, others study single cells. During breaks at scientific conferences, you can even get teams of psychoneuroimmunological pediatricians playing volleyball against the psychoneuroimmunological gerontologists. I am old enough, I confess frankly, to remember a time when there was no such thing as a psychoneuroimmunologist. Now, like an aging Cretaceous dinosaur, I watch these new mammals proliferating. There was even a time when it was not common knowledge that stress caused immune tissues to shrink—and as a result, medical researchers carried out some influential studies and misinterpreted their findings, which indirectly led to the deaths of thousands of people. By the nineteenth century, scientists and doctors were becoming concerned with a new pediatric disorder. On certain occasions parents would place their apparently perfectly healthy infant in bed, tuck the blankets in securely, leave for a peaceful night’s sleep—and return in the morning

to find the child dead. “Crib death,” or sudden infant death syndrome (SIDS), came to be recognized during that time. When it happened, one initially had to explore the unsettling possibility that there was foul play or parental abuse, but that was usually eliminated, and one was left with the mystery of healthy infants dying in their sleep for no discernible reason. Today, scientists have made some progress in understanding SIDS. It seems to arise in infants who, during the third trimester of fetal life, have some sort of crisis where their brains do not get enough oxygen, causing certain neurons in the brain stem that control respiration to become especially vulnerable. But in the nineteenth century, no one had a clue as to what was going on. Some pathologists began a logical course of research in the 1800s. They would carefully autopsy SIDS infants and compare them with the normal infant autopsy material. Here is where the subtle, fatal mistake occurred: “normal infant autopsy material.” Who gets autopsied? Who gets practiced on by interns in teaching hospitals? Whose bodies wind up being dissected in gross anatomy by first-year medical students? Usually, it has been poor people. The nineteenth century was the time when men with strong backs and a nocturnal bent could opt for a career as “resurrectionists”—grave robbers, body snatchers, who would sell corpses to anatomists at the medical schools for use in study and teaching. Overwhelmingly, the bodies of the poor, buried without coffins in shallow mass graves in potter’s fields, were taken; the wealthy, by contrast, would be buried in triple coffins. As body-snatching anxiety spread, adaptations evolved for the wealthy. The “patent coffin” of 1818 was explicitly and expensively marketed to be resurrectionist-proof, and cemeteries of the gentry would offer a turn in the dead- house, where the well-guarded body could genteelly putrefy past the point of interest to the dissectors, at which time it could be safely buried. This period, moreover, gave rise to the verb burking, named after one William Burke, the aging resurrectionist who pioneered the practice of luring beggars in for a charitable meal and then strangling them for a quick sale to the anatomists. (Ironic-ending department: Burke and his sidekick, after their execution, were handed over to the anatomists. Their dissection included particular attention to their skulls, with an attempt to find phrenological causes of their heinous crimes.) All very helpful for the biomedical community, but with some drawbacks. The poor tended to express a riotous displeasure with the medico-body snatcher complex (to coin a phrase). Frenzied crowds lynched resurrectionists who were caught, attacked the homes of anatomists, burned hospitals. Concerned about the mayhem caused by unregulated preying on the bodies of the poor, governments moved decisively to supervise the preying. In the early nineteenth century, various European governments acted to supply adequate bodies to the anatomists, put the burkers and resurrectionists out of business, and keep the poor in line—all with one handy little law: anyone who died destitute in a poorhouse or a pauper’s hospital would now be turned over to the dissectors. Doctors were thus trained in what the normal human body looked like by studying the bodies and tissues of the poor. Yet the bodies of poor people are changed by the stressful circumstances of their poverty. In the “normal” autopsy population of six-month-olds, the infants had typically died of chronic diarrheal disorders, malnutrition, tuberculosis. Prolonged, stressful diseases. Their thymus glands had shrunk.

We now return to our pathologists, comparing the bodies of SIDS infants with those of “normal” dead infants. By definition, if children had been labeled as having died of SIDS, there was nothing else wrong with them. No prior stressors. No shrinking of the thymus gland. The researchers begin their studies and discover something striking: SIDS kids had thymuses much larger than those of “normal” dead infants. This is where they got things backward. Not knowing that stress shrinks the thymus gland, they assumed that the thymuses in the “normal” autopsy population were normal. They concluded that some children have an abnormally large thymus gland, and that SIDS is caused by that large thymus pressing down on the trachea and one night suffocating the child. Soon this imaginary disorder had a fancy name, “status thymicolymphaticus.” This supposed biological explanation for SIDS provided a humane substitute for the usual explanation at the time, which was to assume that the parents were either criminal or incompetent, and some of the most progressive physicians of the time endorsed the “big thymus” story (including Rudolph Virchow, a hero of chapter 17). The trouble was, the physicians decided to make some recommendations for how to prevent SIDS, based on this nonsense. It seemed perfectly logical at the time. Get rid of that big thymus. Maybe do it surgically, which turned out to be a bit tricky. Soon, the treatment of choice emerged: shrink the thymus through irradiation. Estimates are that in the ensuing decades it caused tens of thousands of cases of cancers in the thyroid gland, which sits near the thymus. When I lecture on this subject, I regularly encounter people whose parents, as late as the 1950s, had their throats irradiated for this reason. What recommendations does one offer from the history of status thymicolymphaticus? I could try for some big ones. That so long as all people are not born equal and certainly don’t get to live equally, we should at least be dissected equally. How about something even more grandiose, such as that something should be done about infants getting small thymuses from economic inequality. Okay, I’ll aim for something on a more manageable scientific scale. For example, while we expend a great deal of effort doing extraordinary things in medical research—say, sequencing the human genome—we still need smart people to study some of the moronically simple problems, like “how big is a normal thymus?” Because they are often not so simple. Maybe another lesson is that confounds can come from unexpected quarters—bands of very smart public health researchers wrestle with that idea for a living. Perhaps the best moral is that when doing science (or perhaps when doing anything at all in a society as judgmental as our own), be very careful and very certain before pronouncing something to be the norm—because at that instant, you have made it supremely difficult to ever again look objectively at an exception to that supposed norm. Chapter 9

Stress and Pain

In Joseph Heller’s classic novel about World War II, Catch-22, the antihero, Yossarian, has an unlikely argument with someone about the nature of God. Unlikely because they are both atheists, which would presumably lead to agreement about the subject. However, it turns out that

while Yossarian merely does not believe in the existence of a God and is rather angry about the whole concept, the God that she does not believe in is one who is good and warm and loving, and thus she is offended by the vehemence of his attacks.

“How much reverence can you have for a Supreme Being who finds it necessary to include such phenomena as phlegm and tooth decay in His divine system of creation? What in the world was running through that warped, evil, scatological mind of His when He robbed old people of the power to control their bowel movements? Why in the world did He ever create pain?”

“Pain?” Lieutenant Scheisskopf’s wife pounced upon the word victoriously. “Pain is a useful symptom. Pain is a warning to us of bodily dangers.” “And who created the dangers?” Yossarian demanded. He laughed caustically. “Oh, He was really being charitable to us when He gave us pain! Why couldn’t He have used a doorbell instead to notify us, or one of his celestial choirs? Or a system of blue-and-red neon tubes right in the middle of each person’s forehead. Any jukebox manufacturer worth his salt could have done that. Why couldn’t He?” “People would certainly look silly walking around with red neon tubes in the middle of their foreheads.” “They certainly look beautiful now writhing in agony or stupefied with morphine, don’t they?”

Unfortunately, we lack neon lights in the middle of our foreheads, and in the absence of such innocuous signs, we probably do need pain perception. Pain can hurt like hell, but it can inform us that we are sitting too close to the fire, or that we should never again eat the novel item that just gave us food poisoning. It effectively discourages us from trying to walk on an injured limb that is better left immobilized until it heals. And in our westernized lives, it is often a good signal that we had better see a doctor before it is too late. People who congenitally lack the ability to feel pain (a condition known as pain asymbolia) are a mess; because they can’t feel pain when they step down with too much force, their feet may ulcerate, their knee joints may disintegrate, and their long bones may crack; they burn themselves unawares; in some cases, they even lose a toe without knowing it. Pain is useful to the extent that it motivates us to modify our behaviors in order to reduce whatever insult is causing the pain, because invariably that insult is damaging our tissues. Pain is useless and debilitating, however, when it is telling us that there is something dreadfully wrong that we can do nothing about. We must praise the fact that we have evolved a physiological system that lets us know when our stomachs are empty. Yet at the same time we must deeply rue our evolving physiological system that can wrack a terminal cancer patient with unrelenting pain. Pain, until we get the lights on our foreheads, will remain a necessary but highly problematic part of our natural physiology. What is surprising is how malleable pain signals are—how readily the intensity of a pain signal is changed by the sensations, feelings, and thoughts that coincide with

the pain. One example of this modulation, the blunting of pain perception during some circumstances of stress, is the subject of this chapter. The Basics of Pain Perception The sensation of pain originates in receptors located throughout our body. Some are deep within the body, telling us about muscle aches, fluid-filled, swollen joints, or damage to organs. Or even something as mundane as a distended bladder. Others, in our skin, can tell us that we have been cut, burned, abraded, poked, or compressed.* Often, these skin receptors respond to the signal of local tissue damage. Cut yourself with a paring knife, and you will slice open various cells of microscopic size that then spill out their proverbial guts; and, typically, within this cellular soup now flooding out of the area of injury is a variety of chemical messengers that trigger pain receptors into action. The tissue injury also triggers an influx of cells of the immune system, which are there to scarf up and dispose of those sliced-up cells. The swelling around the injury site because of this infiltration is what we call inflammation, and those inflammatory cells release chemicals that make pain receptors more sensitive.

George Cruikshank, The Headache, hand-colored etching, 1819.

Some pain receptors carry information only about pain (for example, the ones responding to cuts); others carry information about both pain and everyday sensations. How are the two differentiated? By intensity. For example, by way of various tactile receptors on my back, I am greatly pleased to have my back scratched and rubbed by my wife. However, as evidence that there are limits to all good things, I would not at all enjoy it if she vigorously scratched my back with coarse sandpaper. Similarly, we may be pleased to have our thermal receptors stimulated by warm sunlight but not by boiling water. Sometimes pain consists of everyday sensations writ large.

Regardless of the particular type of pain and the particular receptor activated, all these receptors send nerve projections to the spinal cord. This can activate a spinal reflex, where spinal neurons rapidly send commands to your muscles (and thus, for example, you jerk your finger away from the flame). Information about the painful stimulus is also sent up to the brain (a lot more on this later). Sensory Modulation of Pain Perception A striking aspect of the pain system is how readily it can be modulated by other factors. The strength of a pain signal, for example, can depend on what other sensory information is funneled to the spine at the same time. This, it turns out, is why it feels great to have a massage when you have sore muscles. Chronic, throbbing pain can be inhibited by certain types of sharp, brief sensory stimulation. The physiology behind this is one of the most elegant bits of wiring I know of in the nervous system, a circuit sorted out some decades ago by the physiologists Patrick Wall and Ronald Melzack. It turns out that the nervous projections—the fibers carrying pain information from your periphery to the spinal cord—are not all of one kind. Instead, they come in different classes. Probably the most relevant dichotomy is between fibers that carry information about acute, sharp, sudden pain and those that carry information about slow, diffuse, constant, throbbing pain. Both project to spinal cord neurons and activate them, but in different ways (see part A of the figure 190). Two types of neurons found in the spinal cord are being affected by painful information (see part B of the illustration). The first (X) is the same neuron diagrammed before, which relays pain information to the brain. The second neuron (Y) is a local one called an interneuron. When Y is stimulated, it inhibits the activity of X.

The Wall-Melzack model of how pain information is passed to the brain, and how it can be modulated by the brain. (A) A neuron (X) in the spinal cord sends a signal to the brain that something painful has happened, once it is stimulated by a pain fiber. Such pain fibers can

carry information about sudden pain or slow, diffuse pain. (B) A more realistic version of how the system actually works, showing why sudden and slow pain information is differentiated. In

the case of sudden pain, the sudden pain fiber stimulates neuron X, causing a pain signal to be relayed to the brain. The sudden pain fiber also stimulates an interneuron (Y) that inhibits neuron X, after a brief delay. Thus, neuron X sends a pain signal to the brain for only a short time. In contrast, the slow pain fiber stimulates neuron X and inhibits interneuron Y. Thus, Y

does not inhibit X, and X continues to send a pain signal to the brain, producing a slow, diffuse pain. (C) Both stimulatory and inhibitory fibers come from the brain and send

information to neuron X, modulating its sensitivity to incoming pain information. Thus, the brain can sensitize neuron X to a painful signal, or blunt its sensitivity.

As things are wired up, when a sharp, painful stimulus is felt, the information is sent on the fast fiber. This stimulates both neurons X and Y. As a result, X sends a painful signal up the spinal cord, and an instant later, Y kicks in and shuts X off. Thus the brain senses a brief, sharp burst of pain, such as after stepping on a tack. By contrast, when a dull, throbbing pain is felt, the information is sent on the slow fiber. It communicates with both neurons X and Y, but differently from the way it does on the fast fiber. Once again the X neuron is stimulated and lets the brain know that something painful has occurred. This time, however, the slow fiber inhibits the Y neuron from firing. Y remains silent, X keeps firing, and your brain senses a prolonged, throbbing pain, the type you’d feel for hours or days after you’ve burned yourself. The pain physiologist David Yeomans has framed the functions of the fast and slow fibers in a way that fits perfectly with this book: what the fast fibers are about is getting you to move as quickly as possible (from the source of the piercing pain). What the slow fibers are about is getting you to hunker down, immobile, so you can heal. The two classes of fibers can interact, and we often intentionally force them to. Suppose that you have some sort of continuous, throbbing pain—sore muscles, an insect bite, a painful blister. How can you stop the throbbing? Briefly stimulate the fast fiber. This adds to the pain for an instant, but by stimulating the Y interneuron, you shut the system down for a while. And that is precisely what we often do in all of those circumstances. Experiencing a good vigorous mauling massage inhibits the dull throbbing pain of sore muscles for a while. An insect bite throbs and itches unbearably, and we often scratch hard right around it to dull the pain. Or we’ll pinch ourselves. In all these cases, the slow chronic pain pathway is shut down for up to a few minutes. This model has had important clinical implications. For one thing, it has allowed scientists to design treatments for people with severe chronic pain syndromes (for example, a patient who has had a nerve root crushed in his back). By implanting a little electrode into the fast pain pathway and attaching it to a stimulator on the person’s hip, they enable the patient to buzz that pathway now and then to turn off the chronic pain; works wonders in many cases. Pain that Goes on Longer than Normal

If someone pokes you over and over, you will continue to feel pain each time. Similarly, if you get an injury that causes days of inflammation, there are likely to be days of pain as well. But sometimes, something goes wrong with pain pathways somewhere between those pain receptors and your spine, and you feel pain long after the noxious stimulus has stopped or the injury has healed, or you feel pain in response to stimuli that shouldn’t be painful at all. Now you’ve got problems—allodynia, which is feeling pain in response to a normal stimulus. Some versions of allodynia can arise down at the level of the pain receptors themselves. Recall how when there is tissue injury, inflammatory cells infiltrate into the area and release chemicals that make those local pain receptors more excitable, more easily stimulated. Now those inflammatory cells are pretty indiscriminate as to where they dump these chemicals, and some of them can leach over in the direction of receptors outside the area of injury, thereby making them more excitable. And suddenly the perfectly healthy tissue surrounding the injured area starts to hurt as well. Allodynia can also occur when neurons in the pain pathway are injured. If nerve endings are severed near the pain receptors, those inflammatory cells release growth promoting factors that prompt the nerves to regenerate. Sometimes the regeneration is bollixed up so that the nerve endings rewire into a tangle called a neuroma, which tends to be hyperexcitable, sending pain signals from perfectly healthy tissue. And if the nerve projections carrying pain information are severed near the spine, this can lead to a cascade of inflammatory events that results in a hyperexcitable spinal cord. A mere touch now feels excruciating. The Wall-Melzack pathway model explains another instance of allodynia, as seen in severe cases of both types of diabetes. As we saw in chapter 4, elevated levels of glucose in the bloodstream can increase the risk of atherosclerotic plaques, clogging up blood vessels. As a result, insufficient energy gets through those vessels, potentially damaging nerves that depend on that energy. In general it is the fast fibers, which take more energy to operate than the lower- maintenance slow fibers, that are damaged. Thus, the person loses the ability to shut down the Y interneuron in that pathway, and what would be a transient pain for anyone else becomes a constant throbbing one for a diabetic. No Brain, No Pain We started with pain receptors scattered all over the body, and have gotten as far as the spinal cord receiving projections from them. From there, a lot of those spinal neurons that are activated by pain send projections up into the brain. This is where things become really interesting. Consider three scenarios involving pain. First, a soldier is in the middle of some appalling battle, people being slaughtered all around. He is injured—not life-threatening, but serious enough to warrant evacuation. Second, consider someone with advanced liver cancer, administered an experimental drug. Within a few days, her gut hurts like hell, a sign of the drug killing the tumor cells. Or third, someone is abrading their rear end raw while enthusiastically having sex on a rough carpet. What do they all have in common? Their pain’s not going to seem all that painful—the war’s over for me; the drug’s working; what carpet? The brain’s interpretation of pain can be extremely subjective.

A study conducted in the 1980s provides a striking example of this subjectivity. A scientist examined a decade’s worth of records at a suburban hospital, noting how many painkillers were requested by patients who had just had gallbladder surgery. It turned out that patients who had views of trees from their windows requested significantly less pain medication than those who looked out on blank walls. Other studies of chronic pain patients show that manipulating psychological variables such as the sense of control over events also dramatically changes the quantity of painkillers that they request (this important finding will be elaborated upon in the final chapter of the book). This is because the brain is not a mindless pain-ometer, simply measuring units of ouchness. Certainly some parts of the brain allow you to make some objective assessments (“Whoa, this water is WAY too hot for the baby’s bath”). And there are factors that can modulate how much those pain-ometer areas register pain—for example, oxytocin, the hormone released in connection with birth and maternal behavior in mammals, will blunt pain responsiveness in these pathways. But most of what the brain’s responses to pain are about is generating emotional responses and giving contextual interpretations about the pain. This is how being shot in the thigh, gasping in pain, can also leave you gasping in euphoric triumph—I’ve survived this war, I’m going home. Three important things about the emotional ways the brain interprets and responds to pain: First, the emotional/interpretative level can be dissociated from the objective amount of pain signal that is coursing up to the brain from the spine. In other words, how much pain you feel, and how unpleasant that pain feels, can be two separate things. That’s implicit in the war, cancer, and tush-abrading scenarios. An elegant study shows it more explicitly. In it, volunteers dipped their hands into hot water before and after being given a hypnotic suggestion that they feel no pain. During both hand dips, brain imaging was carried out to show which parts of the brain were becoming active. The sensation-processing part of the cortex (kind of a pain-ometer in this case) was activated to identical extents in both cases, reflecting the similar number of heat-sensitive pain receptors being triggered to roughly the equivalent extent in both cases. But the more emotional parts of the brain activated only in the pre-hypnosis case. The pain was the same in both cases; the response to it was not. As a second point, those more emotive parts of the brain not only can alter how you respond to pain information coming up the spinal cord; those areas of the brain can alter how the spinal cord responds to pain information. And the third point: this is where stress comes in big time. Stress-Induced Analgesia Chapter 1 recounted anecdotal cases of people who, highly aroused during battle, did not notice a severe injury. This is certainly a useful thing for a soldier, or a zebra, who still needs to function despite those circumstances. One of the first to document this phenomenon of stress-induced analgesia was an anesthesiologist, Henry Beecher, who examined injured soldiers as a battlefront medic in World War II and compared them with civilian populations. He found that for injuries

of similar severity, approximately 80 percent of civilians requested morphine, while only a third of the soldiers did. Few of us experience stress-induced analgesia in the midst of battle. For us, it is more likely to happen during some sporting event where, if we are sufficiently excited and involved in what we are doing, we can easily ignore an injury. On a more everyday level, stress-induced analgesia is experienced by the droves who exercise. Invariably the first stretch is agony, as you search for every possible excuse to stop before you suffer the coronary that you now fear. Then suddenly, about half an hour into this self-flagellation, the pain melts away. You even start feeling oddly euphoric. The whole venture seems like the most pleasant self-improvement conceivable, and you plan to work out like this daily until your hundredth birthday (with all vows, of course, forgotten the next day when you start the painful process all over again).* Traditionally many hard-nosed laboratory scientists, when encountering something like stress- induced analgesia, would relegate it to the “psychosomatic” realm, dismissing it as some fuzzy aspect of “mind over matter.” The analgesia, however, is a real biological phenomenon. One bit of evidence for that assertion is that stress-induced analgesia occurs in other animals as well, not just in humans emotionally invested in the success of their nation’s army or their office’s softball team. This can be shown in animals with the “hot-plate test,” Put a rat on a hot plate; then turn it on. Carefully time how long it takes for the rat to feel the first smidgen of discomfort, when it picks up its foot for the first time (at which point the rat is removed from the hot plate). Now do the same thing to a rat that has been stressed—forced to swim in a tank of water, exposed to the smell of a cat, whatever. It will take longer for this rat to notice the heat of the plate: stress-induced analgesia. The best evidence that such analgesia is a real phenomenon is the neurochemistry that has been discovered to underlie it. The tale begins in the 1970s, with the subject that interested every ambitious, cutting-edge neurochemist of the time. It concerned the various opiate drugs that were being used recreationally in vast numbers: heroin, morphine, opium, all of which have similar chemical structures. In the early 1970s, three groups of neurochemists almost simultaneously demonstrated that these opiate drugs bound to specific opiate receptors in the brain. And these receptors tended to be located in the parts of the brain that process pain perception. This turned out to solve the problem of how opiate drugs block pain—they activate those descending pathways that blunt the sensitivity of the X neuron shown in the illustration. Terrific—but two beats later, something puzzling hits you. Why should the brain make receptors for a class of compounds synthesized in poppy plants? The realization rushes in; there must be some chemical—a neurotransmitter? a hormone?—made in the body that is structurally similar to opiates. Some kind of endogenous morphine must occur naturally in the brain. Neurochemists went wild at this point looking for endogenous morphine. Soon they found exactly what they were looking for: endogenous compounds with chemical structures reminiscent of the opiate drugs. They turned out to come in three different classes—enkephalins, dynorphins, and the most famous of them all, endorphins (a contraction for “endogenous morphines”). The opiate receptors were discovered to bind these endogenous opioid compounds,

just as predicted. Furthermore, the opioids were synthesized and released in parts of the brain that regulated pain perception, and they would make some of the neurons that relay pain signals in the spine less excitable. (Opiate refers to analgesics not normally made by the body, such as heroin or morphine. Opioid refers to those made by the body itself. Because the field began with the study of the opiates—since no one had discovered the opioids as yet—the receptors found then were called opiate receptors. But clearly, their real job is to bind the opioids.) Chapter 7 introduced the finding that the endorphins and enkephalins also regulate sex hormone release. An additional intriguing finding concerning opioid action emerged: release of these compounds explained how acupuncture worked. Until the 1970s, many Western scientists had heard about the phenomenon, but most had written it off, dumping it into a bucket of anthropological oddities—inscrutable Chinese herbalists sticking needles into people, Haitian shamans killing with voodoo curses, Jewish mothers curing any and all diseases with their secret-recipe chicken soup. Then, right around the time of the explosion in opiate research, Nixon ventured to China, and documentation started coming out from there about the reality of acupuncture. Furthermore, scientists noted that Chinese veterinarians used acupuncture to do surgery on animals, thereby refuting the argument that the painkilling characteristic of acupuncture was one big placebo effect ascribable to cultural conditioning (no cow on earth will go along with unanesthetized surgery just because it has a heavy investment in the cultural mores of the society in which it dwells). Then, as the corker, a prominent Western journalist (James Reston of the New York Times) got appendicitis in China, underwent surgery, and was administered acupuncture for pain relief. He survived just fine. Hey, this stuff must be legit—it even works on white guys. Acupuncture stimulates the release of large quantities of endogenous opioids, for reasons no one really understands. The best demonstration of this is what is called a subtraction experiment: block the activity of endogenous opioids by using a drug that blocks the opiate receptor (most commonly a drug called naloxone). When such a receptor is blocked, acupuncture no longer effectively dulls the perception of pain. Endogenous opioids turn out to be relevant to explaining placebos as well. A placebo effect occurs when a person’s health improves, or the person’s assessment of their health improves, merely because they believe that a medical procedure has been carried out on them, regardless of whether it actually has. This is where patients in a study either get the new medicine being tested or, without knowing it, merely a sugar pill, and sugar pill folks get somewhat better. Placebo effects remain controversial. A highly publicized paper in the New England Journal of Medicine a few years back surveyed the efficacy of placebo treatments across the board in all realms of medicine. The authors examined the results of 114 different studies, and concluded that, overall, receiving a placebo treatment had no significant effects. The study irritated me no end, because the authors included all sorts of realms where it seemed crazy to expect a placebo effect to occur. For example, the study informed us that believing you’ve received an effective medical treatment when you actually have not has no beneficial effects for epilepsy, elevated cholesterol levels, infertility, a bacterial infection, Alzheimer’s disease, anemia, or schizophrenia.

Thus, the placebo effect got trashed and, amid the triumphant chest-thumping by all sorts of dead-white-male elements of the medical establishment, what was lost in that paper was a clear indication that placebo effects are highly effective against pain. This makes a great deal of sense, given what we have now seen about pain processing in the brain. As an example of such a placebo effect, IV infusion of painkillers is more effective if the patient sees the infusion occurring than if it is done on the sly—knowing that a pain-reducing procedure is being carried out adds to its effectiveness. I saw a great example of this a few years back when my then two-year-old daughter came down with an ear infection. She was miserable beyond consolation, clearly in tons of pain. Off to the pediatrician and, amid much wailing and protestations of pain, she had her ears examined. Yup, she’s got a huge infection, both ears, said the doc, disappearing to get an injection of antibiotics. We turn to find our daughter looking serene. “My ears feel much better now that the doctor fixed them,” she announced. Placeboed by having some instrument stuck in her ears. Not surprisingly, it turns out that they work by releasing endogenous opioids. As but one example of the evidence for that, block opiate receptors with naloxone, and placebos no longer work. All of this is a prelude to the discovery that stress releases opioids as well. This finding was first reported in 1977 by Roger Guillemin. Fresh from winning the Nobel Prize for the discoveries described in chapter 2, he demonstrated that stress triggers the release of one type of endorphin, beta-endorphin, from the pituitary gland. The rest is history. We all know about the famed runner’s high that kicks in after about half an hour and creates that glowing, irrational euphoria, just because the pain has gone away. During exercise, beta-endorphin pours out of the pituitary gland, finally building up to levels in the bloodstream around the 30-minute mark that will cause analgesia. The other opiates, especially the enkephalins, are mobilized as well, mostly within the brain and spine. They activate the descending pathway originating in the brain to shut off the X neurons in the spinal cord, and they work directly at the spinal cord to accomplish the same thing. Moreover, they also work at the pain receptors in the skin and organs, blunting their sensitivity. All sorts of other stressors produce similar effects. Surgery, low blood sugar, exposure to cold, examinations, spinal taps, childbirth—all do it.* Certain stressors also cause analgesia through “nonopioid-mediated” pathways. No one is quite sure how those work, nor whether there is a systematic pattern as to which stressors are opioid-mediated. So stress blocks pain perception, enabling you to sprint away from the lion despite your mauling, or at least to put up with the muscle ache of smiling obsequiously non-stop during the stressful meeting with the boss. This explains everything. Unless it happens to be the sort of stressful situation that makes pain worse instead of better. Why is Muzak in the Dentist’s Office Painful?

All that stress-induced analgesia stuff may be swell for that disemboweled zebra, but what if you’re the sort of person where just seeing the nurse taking the cap off the hypodermic needle for the blood draw makes your arm throb? What we’ve got now is stress-induced hyperalgesia. The phenomenon is well documented, if studied less than stress-induced analgesia. What is known about it makes perfect sense, in that stress-induced hyperalgesia does not actually involve more pain perception, and has nothing to do with pain receptors or the spinal cord. Instead, it involves more emotional reactivity to pain, interpreting the same sensation as more unpleasant. So stress-induced hyperalgesia is just in your head. On the other hand, so is stress-induced analgesia, just a different part of your head. The pain-ometer parts of your brain respond to pain normally in people with stress-induced hyperalgesia. It’s the more emotional parts of the brain that are hyperreactive, the parts of the brain that are the core of our anxieties and fears.

Vic Boff, New York Polar Bear Club member known as “Mr. Iceberg,” sitting in the snow after

a swim during the blizzard of 1978. This can be shown with brain-imaging studies, showing what parts of pain circuitry in the brain become overly active during such hyperalgesia. Moreover, anti-anxiety drugs like Valium and Librium block stress-induced hyperalgesia. People who score high on tests for neuroticism and anxiety are most prone toward hyperalgesia during stress. Amazingly, so are rat strains that have been bred for high anxiety. So we’re at one of those crossroads that makes science look kind of lame. Just like, “Stress can increase appetite. And it can decrease it, too,” we’ve got, “Stress can blunt pain perception. But sometimes it does the opposite.” How to combine these opposing effects of stress? My sense from the literature is that the analgesia arises more in circumstances of massive, physical injury. Half your body is burned and your ankle’s sprained, and you’re trying to carry a loved one out of some inferno—that’s when stress-induced analgesia is going to dominate. Discover some weirdo growth on your shoulder that hurts a bit, decide in a panic that you’ve got fatal melanoma, be informed by an unsympathetic answering machine that your doctor has just left for a three-day weekend. That’s when the stress-induced hyperalgesia will dominate, as you lie awake for three nights, thanks to how painful you’ve now decided the spot feels.

This brings up a subject that needs to be treaded on carefully. So carefully in fact that in the last edition of the book, I bravely made a point of not mentioning a word about it. Fibromyalgia. This is the mysterious syndrome of people having markedly reduced pain tolerance and multiple tender spots throughout the body, often paralyzing extents of pain, and no one can find anything wrong—no pinched nerve, no arthritis, no inflammation. Mainstream medicine has spent decades consigning fibromyalgia to the realm of psychosomatic medicine (that is, “Get out of my office and go see a shrink”). It doesn’t help that fibromyalgia is more likely to strike people with anxious or neurotic personalities. There’s nothing wrong, is the typical medical conclusion. But this may not quite be the case. For starters, sufferers have abnormally high levels of activity in parts of the brain that mediate the emotional/contextual assessments of pain, the same areas activated in stress-induced hyperalgesia. Moreover, their cerebral spinal fluid contains elevated levels of a neurotransmitter that mediates pain (called Substance P). And, as noted in chapter 2, unexpectedly, glucocorticoid levels are below normal in people with fibromyalgia. Maybe these are highly stressed people with some sort of defect in glucocorticoid secretion, and because of that deficiency, instead of getting stress-induced analgesia, they get hyperalgesia.* I don’t know. No one knows, as far as I can tell. But there is increasing evidence that there is something biologically real going on in these cases. There, I’ve broken the ice on this subject; stay tuned for the next edition. Pain and Chronic Stress Time now for our usual question. What happens with pain perception when there is chronic stress? With stress-induced hyperalgesia, the answer seems to be, the pain just keeps going, maybe even worsens. But what about stress-induced analgesia? In the acute, lion-mauling scenario, it is adaptive. To follow the structure laid out in previous chapters, this represents the good news. So what’s the bad news? How does an excess of opioid release make us sick in the face of the chronic psychological stressors that we specialize in? Does chronic stress make you an endogenous opioid addict? Does it cause so much of the stuff to be released that you can’t detect useful pain anymore? What’s the downside in the face of chronic stress? Here the answer is puzzling because it differs from all the other physiological systems examined in this book. When Hans Selye first began to note that chronic stress causes illness, he thought that illness occurs because an organism runs out of the stress-response, that the various hormones and neurotransmitters are depleted, and the organism is left undefended to the pummelings of the stressor. As we’ve seen in previous chapters, the modern answer is that the stress-response doesn’t become depleted; instead, one gets sick because the stress-response itself eventually becomes damaging. Opioids turn out to be the exception to the rule. Stress-induced analgesia does not go on forever, and the best evidence ascribes this to depletion of opioids. You are not permanently out of business, but it takes a while for supply to catch up with demand. Thus, to my knowledge, there is no stress-related disease that results from too much opioid release during sustained stressors. From the standpoint of this book and our propensity toward chronic psychological stressors, that is good news—one less stress-related disease to worry

about. From the standpoint of pain perception and the world of real physical stressors, the eventual depletion of the opioids means that the soothing effects of stress-induced analgesia are just a short-term fix. And for the elderly woman agonizing through terminal cancer, the soldier badly injured in combat, the zebra ripped to shreds but still alive, the consequence is obvious. The pain will soon return.