McEwen has identified eight markers for damage resulting from poorly managed allostasis, including blood pressure, blood sugar levels, cholesterol and cortisol levels, and abdominal fat. Exercise and a balanced diet keep allostatic load in check, he notes, and so does social interaction.

Perhaps the best way of thinking about stress-related disease, explains Stanford University neuroendocrinologist Robert M. Sapolsky, "is not of sustained challenge to homeostasis making you sick, but sustained challenge to homeostasis making you more likely to get diseases that make you sick, and even more so, more likely to exacerbate preexisting disease."

Not all challenges to homeostasis are harmful. Some are needed to get you out of a dangerous situation, and some are even necessary for emotional and intellectual growth.

Princeton University neuroscientist Tracey J. Shors, for example, has shown in rats that "some stress actually enhances the ability to learn," but only in males. The unexpected gender difference is related to estrogen, and specifically to the high level of estrogen released during stress, she explains. (Some amount of estrogen is needed for learning and, in fact, estrogen helps to alleviate some of the memory deficits in patients with Alzheimer's disease.) She believes that the gender difference to an acute stress will reveal itself in humans as well, and she is now testing this.

But if a stressor is too severe, too protracted, and too uncontrollable, the stress hormones may be managed inefficiently, and the allostatic load may reach a point at which frank disease becomes manifest. For example, the fight or flight response, first described by U.S. physiologist Walter Cannon in the late 1920s, prompts a state of arousal and vigilance. If the reaction to the stressor is not turned off and homeostasis renewed, the person can become hyperaroused and hypervigilant-an anxious insomniac.

So homeostasis is a tricky balancing act, and the body's response to stress is an intricate cascade of neurophysiological and biochemical events that are triggered before an eye can blink but which have to be shut down quickly once the threat has passed.

At the first sign of a threat-physical or psychological-the brain activates a series of neural and hormonal pathways that make the animal or human more alert and aware; more focused and vigilant; more aggressive, if that is appropriate; but less perceptive to pain. In the rest of the body, the organs are commanded to redirect energy-in the form of oxygen and nutrients-to where it is most needed-the brain, heart, and skeletal muscles. The heart beats faster, blood pressure rises, breathing rate increases. Arteries to muscles relax to receive more sugar-packed blood while those supplying the skin constrict as a protection against loss of blood from a wound. Body functions unrelated to immediate survival-such as digestion, growth, and reproduction-are blunted. The stress response is in play: The animal or human is geared to do battle or to run.

Two key components mediate the stress response: corticotrophin-releasing hormone (CRH), released throughout the brain but mainly in the hypothalamus, a formation deep in the brain; and the locus ceruleus-norepinephrine/sympathetic nervous systems located in the brain stem. CRH is the central command post for stress.

CRH activates central neural pathways as well as the peripheral sympathetic nervous system to bring about the initial "call-to-arms" effects such as increased heart rate and alertness. The locus ceruleus-norepinephrine/sympathetic nervous systems act centrally to cause the release of the catecholamine norepinephrine from a network of neurons scattered throughout the brain.

The norepinephrine enhances the state of arousal and vigilance. Activated sympathetic nerves, acting peripherally, jolt the inner core of the adrenal gland-the medulla-into also secreting norepinephrine.

CRH also acts on the adrenal gland but via a circuitous route. Released from the hypothalamus, it travels a short distance to the pituitary gland located just under the brain. There it stimulates the release of still another hormone, the polypeptide ACTH, or adrenocorticotropin hormone. ACTH takes a longer journey, traveling the bloodstream to the adrenal gland. ACTH acts on the outer layer of the gland, the cortex, to prompt the release of cortisol, a steroid, to the bloodstream.

In humans, cortisol is the principal, but not the only, glucocorticoid released from the adrenal cortex. If everything is in working order, cortisol completes a feedback loop once the threat is perceived to be over. The steroid travels the bloodstream to the brain where it inhibits other stress hormones, including the nerve-signaling hormones (neurotransmitters) serotonin, norepinephrine, and dopamine. Via this inhibitory path, cortisol becalms the brain. Also in a feedback loop, circulating norepinephrine acts back on the brain to shut off further production of this catecholamine.

A properly functioning feedback mechanism ensures that the stress response is limited in duration. But according to the "glucocorticoid cascade hypothesis," if the stress system is not shut down, too much cortisol is produced, which eventually will damage the hippocampus-the site of emotion, consolidated memory, and learning. This damage impairs the hippocampus' ability to inhibit the hypothalamus-pituitary-adrenal gland (HPA) axis, throwing its response to stress out of whack. Without the hippocampal check, the HPA axis goes into overdrive, spewing out more and more glucocorticoids that further damage the hippocampus. At least that's what happens in animals. The sketchy evidence available indicates that this also happens in humans.

Neuroscientists and biochemists are searching for endogenous compounds that can act on any part of the HPA axis to modulate the behavioral and hormonal manifestations of stress. A group of researchers from Northwestern University Medical School, Chicago, and several other universities believe they have uncovered one, a peptide that is part of a larger molecule that also gives rise to thyrotropin-releasing hormone. They dubbed the peptide prepro-TRH178-199.

Last year, in the Journal of Neuroscience, the Northwestern scientists reported that in stressed lab animals, the peptide significantly reduced circulating levels of ACTH and prolactin, another pituitary hormone that is elevated in response to stress. The peptide also reduced fear and anxiety but enhanced arousal. Eva Redei, a biochemist from Northwestern who identified and isolated the peptide, believes it is an important endogenous component of the body's stress regulatory system. There are likely many others.

About 30 years ago, McEwen's lab started to look for stress hormone receptors in the brain and found a bounty in the hippocampus. The hippocampus isn't the only brain area involved in the stress response-the amygdala and the medial prefrontal cortex are involved as well. But the abundance of stress hormone receptors makes the hippocampus" the most vulnerable and sensitive area of the brain as far as the actions of stress hormones are concerned," McEwen says. Since that early finding decades ago, McEwen's lab has "been trying to understand what these receptors do."

As a graduate student in McEwen's lab in the late 1970s, Sapolsky discovered that chronic exposure to stress hormones damaged these hippocampal neurons, which resulted in memory deficits in rats. Sapolsky also showed that acute stress can impair memory, but for a very short time.

More recently in his own lab at Stanford, Sapolsky has refined earlier observations about the effects of acute stress. He has been able to show that acute exposure to excessive stress, or to stress hormones such as cortisol, probably won't kill neurons but may endanger them, making them less able to survive a coincident neurological insult such as a seizure or oxygen deprivation.

As Sapolsky explains, stress hormones significantly inhibit the uptake of glucose by neurons in the hippocampus. The neurons are left "energetically vulnerable" and much less able to contain the damage imposed on them by an excitatory amino acid neurotransmitter such as glutamate, or harmful ions such as calcium, or by oxygen radicals. There is little damage to neurons if the stress lasts only a few days. But if the glucocorticoids hang around for a couple of weeks and during that time the brain is deprived of a little oxygen or a little blood, neurons begin to atrophy. This atrophy is mediated by glucocorticoids acting synergistically with excitatory neurotransmitters, Sapolsky says.

While Sapolsky has continued to explore-both at the cellular level and in field studies of baboons in Kenya-the ways stress and stress hormones damage the brain, McEwen's lab has expanded its research scope to include a search for protective effects as well. As a result, McEwen's lab has found that repeated stress in rats causes the dendrites of pyramidal neurons in a specific region of the hippocampus to atrophy, resulting in cognitive deficits in spatial learning and memory. When the nerve cells atrophy, they pull back some of their dendrites: The extensive branching of these neural extensions is pruned, so to speak.

This finding does not intuitively appear to be beneficial, but McEwen and his coworkers, neuroscientists Ana María Magariños and José M. García Verdugo, suggest that the atrophy is actually protective. In essence, it lowers the profile of the nerve cells to the neurotoxic effects of cortisol and excitatory amino acids. The downside is temporary impairment of learning and memory.

At first, the Rockefeller researchers didn't know whether the stress-induced atrophy was reversible. But more recently, McEwen's lab has shown that it is in the tree shrew. Dendrites are merely shriveling up, so that the number of connections they make with other cells declines. The network of branching returns, however, once the stressor system is shut down.

Sapolsky's 1970s discovery that chronic stress-and thus elevated levels of glucocorticoid-lasting months or years is fatal to rat hippocampal neurons is important in defining "successful aging" in rodents. The aging hippocampus probably becomes more vulnerable to the destructive effects of the stress-induced glucocorticoid through a calcium-dependent mechanism and by the persistent release of the excitatory amino acid glutamate-both of which likely drive the nerve cells to commit suicide. And recent studies indicate that stress and cortisol might act similarly in elderly humans.

In a five-year longitudinal study of normal elderly humans, neuroscientist Sonia J. Lupien of McGill University and her coworkers there and at Rockefeller University, New York University Medical Center, and the University of California, San Diego, have found that long-term exposure to rising levels of cortisol may facilitate hippocampal aging in this population.

Lupien: cortisol can harm brain in elderly

Cortisol levels were tested for a 24-hour period once a year. One group was found to have rising cortisol levels that, at the fifth year, were very high. Another group had what turned out to be moderately rising cortisol levels that were normal in the fifth year. A third group had falling levels of the stress hormone over the five-year period.

When the brains of the extreme groups were scanned using magnetic resonance imaging (MRI), Lupien and colleagues were able to correlate hippocampal damage-a 14% decrease in hippocampal volume-and "significantly impaired memory" in the elderly with high cortisol levels, Lupien says. Some neuroscientists interpret these pathological findings as implying that what matters is long-term, rather than acute, exposure to cortisol.

The unexpected finding of declining cortisol levels in a subgroup has been replicated and "is real," Lupien says." These people, for some reason, are able to negotiate and manage stress," she explains. They are "successful agers" whose memory remains very good. "Now we are looking for the basic biological or sociological determinant."

Hippocampal atrophy has been noted in some depressives, and in patients-including veterans of the Vietnam and Gulf Wars-with PTSD and schizophrenia. It is seen in certain types of aging in which individuals aren't yet experiencing dementia but have memory loss-such as Lupien's subjects. It also is seen in patients with Cushing's syndrome (which causes obesity and muscle weakness due to excess cortisol), in which the atrophy is reversible. Except for Cushing's, it is not known "whether the atrophy is due to permanent cell loss or to the kind of reversible atrophy we have discovered," McEwen explains. That "has to be worked out in future research."

Lupien plans to find out-at least in depressives. She speculates that cortisol levels may be a biological marker of the frontier between normal and pathological hippocampal aging.

In a just-initiated study of elderly depressed humans, Lupien expects to find a significant number who have high circulating levels of cortisol and smaller than normal hippocampi. She plans on measuring hippocampal volume in this group before and at the end of a six-week treatment with an antidepressant. The antidepressant should lower the cortisol levels and, if the atrophy is reversible, hippocampal volume should increase.

It isn't absolutely certain that high circulating levels of cortisol cause hippocampal atrophy, but the prevailing view is that they are the precipitating factor. Not all types of depressives, for example, have high circulating cortisol levels, and PTSD patients actually have lower than normal circulating levels. What may be important is the level of cortisol at the time of the trauma, which may have been high.

In collaboration with McEwen, Lupien is measuring cortisol levels in rape victims in emergency rooms, very close to the time of the trauma. Over time, she estimates, about 30% of these victims will develop PTSD. "I think those women who will develop PTSD will be the ones who showed the largest cortisol response to the traumatic event."

It's still not known whether the hippocampal atrophy in PTSD patients is reversible, although Sapolsky guesses the passage of time makes it unlikely. If the neurons are dead, long-held dogma states that the hippocampus in humans will be unable to grow new nerve cells. Researchers at Princeton University and at the German Primate Center in Göttingen along with McEwen at Rockefeller, appear to be closer to overturning that doctrine, however.

Dark areas...

Princeton neuroscientist Elizabeth Gould and her colleagues have shown that a formation in the hippocampus of adult male marmosets, called the dentate gyrus, is able to grow significantly large numbers of new granule nerve cells, and that a single stressful event significantly decreases the number of new cells grown. "This is the first example in primates of neurons being produced in adulthood" and the first to demonstrate "that stress can prevent the production of these cells," Gould tells C&EN.

But marmosets, a New World primate, are not humans and are not even as closely related to humans as Old World primates such as rhesus monkeys. A decade ago, neuroscientist Pasko Rakic and colleagues at Yale University were unable to show new neuron development in the hippocampus of adult rhesus monkeys. Gould attributes the difference to the less sensitive detection methods used by Rakic.

Rakic, in turn, agrees that his methods were less sensitive but not enough to miss the formation of the hundreds of new cells a day that Gould found. Gould intends to replicate her study in New World monkeys.

New neuron growth is not hippocampuswide, and the dentate gyrus is not one of the areas known to be damaged by cortisol-mediated stress, Gould acknowledges. But, she argues, her findings "present another mechanism whereby stress could be resulting in a smaller hippocampus."

As it does with every other system in the body, the brain ultimately controls the immune system, even though the latter sometimes appears to act entirely alone. The immune system is the body's main chemical defense against the outside world. It is, in a sense, the body's eyes to the unseen world of pathogens. Even though the brain is the master gland, the immune system does influence the brain. The two share a common chemical language, and they complement each other in many ways. This bidirectional flow of influence is important to homeostasis.

That the brain and the immune system are inextricably linked and essentially act as an integrated system when the body comes under stress had long been suspected but never thoroughly probed. In the 1970s, psychoneuroimmunologists Karen Bulloch, then at the University of California, San Diego, and David L. Felten at the University of Rochester rediscovered sympathetic innervation of the immune system. (It was discovered earlier by Swedish and German scientists who showed innervation but didn't know what it did-so it was forgotten.)

Around the same time as Bulloch and Felten's work, Robert Ader and Nicholas Cohen, both at the University of Rochester School of Medicine & Dentistry, published a paper on the suppression of the rat's immune system by behavioral conditioning. Glimmers of an emerging field became evident.

In 1981, Ader, with colleagues Felten and Cohen, edited a book entitled "Psychoneuroimmunology," and a new field was tentatively launched. It has taken nearly two decades for the field to gain respectability and focus. As its name implies, the new discipline explores and exploits the interactions of behavior, the brain, and the endocrine and immune systems.

Indeed, the known and suspected causal links between the brain and immune system help to explain why the stress of a final exam period, or lack of control on the job, or loss of a loved one can bring on colds, other infectious diseases, and perhaps more serious disorders. For example, researchers are actively trying to unravel the role stress, via the immune system, may play in the development of cancer, AIDS, multiple sclerosis, and other devastating disorders. Psychologists are studying stress management techniques and the role exercise may play in preventing or ameliorating stress-induced disease.

The main organs of the immune system are the thymus and bone marrow, and white blood cells called lymphocytes are at the core of immune defenses. Lymphocytes are divided into two classes: T cells and B cells, which cruise the blood and lymphatic streams as sentinels on the lookout for trouble. The B cells give rise to antibodies, defensive proteins that bind to foreign proteins to defang them in a process called humoral immunity. Pathogen-destroying T cells stimulate B cells to secrete antibodies and also work in concert with scavenging macrophages to attack pathogens such as cancer cells in a process called cellular immunity.

In its role as the body's chief defender from hostile invaders, the immune system has had to develop memory. Once exposed to a pathogen, the immune system is able to remember it years later, and can attack it swiftly. Sometimes, however, the immune system declares war on the body's own tissues, and autoimmune diseases such as systemic lupus erythematosus develop. There is growing suspicion that autoimmune processes may be triggering some behavioral disorders.

The relationship between the nervous system and the immune system is intimate, intricate, and complicated. For example, organs of the immune system such as the thymus and lymph nodes are wired by sympathetic nerve fibers. The thymus, in turn, makes some hormones that are found in the hypothalamus and which are believed to regulate the pituitary-adrenal axis-the same axis involved in the stress response. Cytokines-signaling substances such as the interferons produced by activated immune cells-can inhibit endocrine and nerve cells. And cytokines produced in the brain can act as neurotransmitters to activate the HPA axis. Much of the brain's influence on the immune system is mediated via this axis.

Lymphocytes and macrophages produce and have receptors for the stress-associated polypeptides CRH-made by hypothalamic nerve cells-and ACTH- secreted by the pituitary. CRH and ACTH can modulate immune cells directly, but they exert more overt influence on the immune system indirectly, via the glucocorticoids.

Glucocorticoids generally act as immunosuppressors. But under acute stress, the steroids cause lymphocytes to "traffic" to blood vessel walls and to lymph nodes and bone marrow. Once at these outposts, if the white blood cells meet no challenge to the immune system, they return to the bloodstream.

Glucocorticoids also decrease cytokine production, possibly by regulating gene expression differentially, and therefore the type of cytokine receptor produced. This inhibitory activity-prompted by the call to arms of acute stress- may allow glucocorticoids to protect the animal or human from overactivity of the immune system that can lead to inflammatory and autoimmune reactions.

That glucocorticoids play a protective role has been clearly demonstrated by neuroimmunologist John Sheridan at Ohio State University, Columbus; viral immunologist Christine A. Biron at Brown University; and Andrew Miller, associate professor of psychiatry at Emory University School of Medicine, Atlanta, who have shown that the steroids enhance survival from acute viral infections. The hormones protect the body from itself by damping the vigorous defense cytokines launch to fight off the virus.

Epinephrine and norepinephrine released by the sympathetic nervous system also inhibit cellular immunity triggered by acute stress.

Influence, however, is bidirectional. Products of the immune system-cytokines, interleukins, and tumor necrosis factor-activate the HPA axis by triggering CRH release. This centrally produced CRH plus CRH produced in immune cells also can mediate local inflammation. The end result of the activation of the HPA axis is inhibition of the immune system by the released glucocorticoids.

So the loop triggered by cytokine-activation of the hypothalamus in response to an inflammatory stimulus ends with the secretion of the glucocorticoid, cortisol, "one of the most potent anti-inflammatory agents our bodies make," explains Esther M. Sternberg, chief of the section of neuroendocrine immunology at the National Institute of Mental Health. The end "effect of that signaling loop is that cortisol will shut off inflammation as soon as it is no longer needed." When the immune system is stuck in the "on" position, diseases resulting from chronic inflammation occur, she adds.

Sternberg: treating inflammatory disease

The prominent role the HPA axis plays in chronic inflammatory disease is best demonstrated in Lewis rats prone to arthritis. In 1989, Sternberg determined that the problem with these rodents "is that they had a blunted stress response. They couldn't make enough CRH."

As a result, ACTH was not released from the pituitary in amounts sufficient to elevate the secretion of glucocorticoids from the adrenal gland. The low level of circulating glucocorticoids failed to inhibit the immune system, inflammation continued, and arthritis developed.

Sternberg points out that Lewis rats have nearly immunological twins, Fisher rats, that "are resistant to inflammatory disease. The reason is they have a very potent stress response." By manipulating the HPA axis at different points, resistant rats can be turned into susceptible ones, "so very clearly this axis plays an important role in the susceptibility and resistance to inflammatory disease," she says.

Depression also may be a disregulation of the brain's stress response: Often, people with arthritis are depressed. "It may be that this association is not simply coincidental or secondary to the presence of chronic disease," Sternberg adds. Atopic dermatitis and asthma may also, like depression, be manifestations of a blunted cortisol response to psychological stressors, she says.

"It may be possible to design studies in humans to define which diseases fall into the category [of chronic inflammatory disease]," Sternberg says. "Then the principles uncovered can be employed to develop more precise treatments for these diseases."

More precise treatment would be salutatory because patients with chronic inflammatory diseases such as rheumatoid arthritis take long-term high doses of glucocorticoids. And researchers like psychiatrist John W. Newcomer at Washington University, St. Louis; neuropsychologist Pamela A. Keenan at Wayne State University, Detroit; and psychiatrist Owen Wolkowitz at the University of California, San Francisco, have been able to show that remarkably little glucocorticoid is needed to cause memory problems in humans.

It's currently unclear whether the cognitive impairments are due to dysfunction of nerve cells or to their atrophy or death. Postmortem studies are now being done. "I suspect," says Sapolsky, "they will generate some disturbing answers, meaning that chronic use of high-dose steroids such as these have some definite neurological costs."

Cytokine activation of the hypothalamus also explains the symptoms of "sickness behavior"-sleepiness, loss of appetite, and lethargy, for example. Rather than expressing an illness' debilitating effects, sickness behavior may be an energy-saving feature of the brain-immune system's counterattack on the foreign invader.

The direction, degree, and duration of stress-induced changes in immunity- and thus, susceptibility to disease-are influenced by a hodgepodge of interlacing factors. Some determinants include the individual's age, sex, and nutritional status; the strength of the stressor as well as the factor that triggers the immune response; and the ability of the individual to cope.

In a series of studies, Ronald Glaser, associate vice president for health sciences research at Ohio State University, and his wife, Janice Kiecolt-Glaser, a psychiatry professor in the university's medical school, have been teasing out the role of stress on the health of people caring for Alzheimer's disease patients. They found that, compared with controls, the caregivers "had a poor antibody response and an even poorer T-cell response" to an influenza virus vaccine, Glaser tells C&EN. The researchers also noted the same poor immune responses in another stressed group, medical students.

Kiecolt-Glaser and Glaser: studying stress and immune response

More recently, Glaser and Kiecolt-Glaser have shown that "psychological stress can affect wound healing" in Alzheimer's caregivers. "It took 24% longer to heal the same size wound" in caregivers than in controls, Glaser says. Levels of interleukin-2 (IL-2), an important cytokine for wound healing, "were significantly lower in cells from the caregivers as compared to the controls," he explains.

Lymphocytes can make growth hormone, and the level of this hormone in these cells in the caregivers was found to be significantly lower than in well-matched controls. "It is known that growth hormone can stimulate the production of IL-2 and we found reduced levels of IL-2 in these same people," Glaser explains. That doesn't prove that stress is the cause of the findings, "but all the pieces of the puzzle are, at least, going in the right direction."

Studies by psychiatrist David Spiegel at Stanford University have focused the spotlight on the protective role social support plays in prolonging survival in breast cancer patients. And Glaser's group is now beginning a study to probe the roles behavioral intervention and stress management may play in reducing anxiety and stress in breast cancer patients. Biological markers will be enhanced immune systems and improved survival rates.

The evidence for interactions between the brain and the immune system is steadily building. "Adaptive immunoregulatory processes involve these nervous and endocrine systems," Ader tells C&EN, "and an understanding of these networks is likely to have profound clinical implications."

As Ader explains, the immune system may be mediating the influence of psychosocial factors on some disease states." From this perspective, there is a need for research to establish that behavioral-neuroendocrine-immune system interactions do, in fact, influence the susceptibility to, or the course of, disease."

Which brings the tale back to stress and the illnesses some Gulf War veterans now have. With the caveat that he is speaking as a person who studies neurons in dishes, Sapolsky tells C&EN:" Stress may be the thing that takes a part of our body that was marginally damaged by exposure to God-knows-what sort of toxins and pushes it into overt disease."