Skip to main content

The Human Operating Manual

By Jay Downes14 min readPublished 18 November 2026

Physiology and Neurobiology of Stress and Adaptation

Central Role of the Brain

Quick answers5 questions this page answers

What is this paper about?

Bruce McEwen’s 2007 review of how the brain decides what is stressful and controls the body’s response, and how that response both protects and wears down body and brain.

What is allostatic load?

The wear and tear on body and brain from running the stress response, especially when the stress mediators aren’t switched off when the stress is over, or aren’t switched on enough when they’re needed.

Stress, allostasis and allostatic load →

What are the four types of allostatic load?

Repeated hits from several stressors, failing to get used to the same stressor, failing to switch the response off once the stressor is over, and an inadequate response that leaves other systems to overcompensate.

Stress, allostasis and allostatic load →

Are the brain changes from chronic stress permanent?

In rats, the shrinkage of hippocampal neurons and the memory problems disappeared within one to two weeks of the stress ending. McEwen sees this as structural flexibility rather than damage, though there is very little evidence on ordinary stress and the human brain.

The brain as a target of stress →

What does McEwen suggest for reducing chronic stress?

Better sleep, more social support, a positive outlook, a healthy diet, not smoking and regular moderate exercise, with medication useful but limited, and policy that makes healthy choices easier.

Managing chronic stress →

Author: Bruce S. McEwen

Source: Physiological Reviews, 87(3), 873–904 (2007). doi:10.1152/physrev.00041.2006

Topics: Stress, allostatic load, the hippocampus, amygdala and prefrontal cortex, sleep, early life, status, exercise, social support

All information is attributed to the author. Except in the case where we may have misunderstood a concept and summarized incorrectly. These notes are only for reference and we always suggest reading from the original source.

Contents

1: Stress, Allostasis and Allostatic Load

2: Stress and Ageing Across the Life Span

3: Being “Stressed Out”: The Case of Sleep

4: The Brain as a Target of Stress

5: Stress in People

6: Managing Chronic Stress

7: Conclusions

8: Reading This in 2026 (HOM note)


1: Stress, Allostasis and Allostatic Load

McEwen starts with the everyday split. “Good stress” is short, something you can master, and leaves you feeling you’ve achieved something. “Bad stress” is where you lack control, it drags on or keeps coming back, and it leaves you drained.

The core of the stress response is the autonomic nervous system and the HPA (hypothalamic–pituitary–adrenal) axis. We need this response to survive a predator, an accident or a mugger, and both too little and too much of it are bad for health. The problem, borrowing from Robert Sapolsky’s Why Zebras Don’t Get Ulcers, is that humans keep these systems switched on for long stretches: through anxiety, noise, pollution, conflict with other people, and the changes in lifestyle that chronic stress brings.

Because the word “stress” is vague, McEwen uses two terms instead:

  • Allostasis: “achieving stability through change” (a term from Sterling and Eyer). The body actively keeps itself stable by releasing stress hormones and other mediators.
  • Allostatic load (or overload): the wear and tear on body and brain from running allostasis, especially when the mediators are dysregulated, meaning they aren’t switched off when the stress is over, or aren’t switched on enough when they’re needed.

The same chemicals protect and damage. Adrenaline and noradrenaline raise heart rate and blood pressure so you can get out of bed, climb the stairs or deal with a threat. Kept high for years, the same rise contributes to atherosclerosis, which can lead to strokes and heart attacks.

The four types of allostatic load

  • Repeated hits: one stressor after another, from several sources.
  • Lack of adaptation: failing to get used to the same stressor when it comes back, so the response never dampens.
  • Prolonged response: failing to switch the response off once the stressor is over.
  • Inadequate response: not mounting enough of a response in the first place, so other systems overcompensate. For example, too little cortisol lets inflammatory cytokines, which cortisol normally keeps in check, run high.

These mediators form a network rather than a chain. Cortisol, adrenaline and noradrenaline, inflammatory and anti-inflammatory cytokines, and the parasympathetic nervous system all regulate one another. The parasympathetic system generally opposes the sympathetic: it slows the heart and has anti-inflammatory effects. Change one mediator and the others shift to compensate, which is why a study that measures only one or two of them has to be read with care.

Stress in the natural world

  • A sudden spring snowstorm stresses birds, and their stress hormones redirect them to stop breeding, find food and move somewhere better. That is allostasis working as designed.
  • Bears fatten up before hibernating, which they need to survive winter and carry young. Zoo bears that overeat out of boredom and don’t exercise put on the same fat with none of the purpose. McEwen calls this allostatic overload.
  • Overload can even serve the species: salmon, and male marsupial mice, die after mating, freeing up food for the next generation.

2: Stress and Ageing Across the Life Span

The idea that hardship “ages” a person has a name, the weathering hypothesis, and animal and population research has started to show how it might work. McEwen’s view is that the brain is the key organ of stress, because it decides what is threatening and controls the response.

The hippocampus

  • McEwen’s lab found receptors for adrenal stress hormones in the hippocampus in the late 1960s. The hippocampus is important for spatial, episodic and contextual memory.
  • The hippocampus also helps switch off the HPA response. When it is damaged or shrinks, the response to psychological stress lasts longer, which led to Sapolsky’s “glucocorticoid cascade hypothesis” of stress and ageing.
  • In older people studied by Lupien and colleagues, cortisol that rose year on year over five years predicted a smaller hippocampus and poorer memory.
  • Animal work suggests ageing in the hippocampus is mostly a loss of connections between neurons rather than a loss of neurons, though the ageing human hippocampus may lose some neurons too.

How much cortisol reaches the brain

Three things control it: a binding protein in the blood (only unbound hormone gets in), a pump at the blood–brain barrier that pushes some steroids back out, and an enzyme (11β-HSD1) that turns inactive cortisone back into cortisol inside brain tissue. Mice without the enzyme show less age-related decline in memory, and short-term treatment with a drug that blocks it improved cognition in people with metabolic syndrome and high cortisol.

Metabolic hormones and the hippocampus

  • The hippocampus responds to insulin, IGF-1, ghrelin and leptin. IGF-1 carries part of the effect of exercise on new neurons. Ghrelin, the hunger hormone, increased new synapses and improved memory in animals. Leptin infused into the hippocampus had antidepressant effects in rats.
  • A high-fat diet impaired hippocampal memory in rats. A high-fat diet plus three weeks of exposure to a predator shrank neurons in the hippocampus, when neither did on its own.
  • In people in their 60s and 70s with mild cognitive impairment, poor glucose control goes with a smaller hippocampus and worse memory.

Experience and early life

  • Past experience biases how we meet new situations. Someone treated badly by a domineering boss walks into the next job differently.
  • Early abuse carries a lifelong burden. Cold, uncaring families leave lasting emotional problems, and chaos in the home goes with poor self-regulation, helplessness, distress, higher body mass and higher blood pressure in children.
  • Prenatal stress is thought to contribute to preterm birth and low birth weight, which in turn raises the risk of heart disease and higher body mass.
  • In rats, mothers who lick and groom their pups a lot raise “neophilic” offspring: curious about new places, less emotionally reactive, with a smaller and better-contained stress response. Poor care produces “neophobic” offspring. Daughters pass the pattern on through how they mother, with changes in DNA methylation playing a role.
  • Neophobic male rats lived about 200 days less than neophilic ones. Neophobic females died about six months sooner, with tumours the likely cause.
  • Pups taken briefly into a new environment away from home grew into adults with better spatial memory and social competition, regardless of how much care their mother gave them.
  • Rhesus monkey mothers with an unpredictable food supply raised more emotionally reactive offspring.

Genes

McEwen notes that common gene variants (in the serotonin transporter, MAOA, BDNF and the glucocorticoid receptor) were thought to change how people respond to stressful experiences. See section 8: several of these findings have not held up.

3: Being “Stressed Out”: The Case of Sleep

Being “stressed out” comes down to raised cortisol, raised sympathetic activity and raised inflammatory cytokines, with less parasympathetic activity. Poor sleep is both a common result of this state and a driver of it.

  • Shorter sleep was linked to higher body mass and obesity in the US NHANES study.
  • Restricting sleep to 4 hours a night raised blood pressure, evening cortisol and insulin, lowered parasympathetic tone, and increased appetite, possibly through more ghrelin and less leptin.
  • Even a modest restriction to 6 hours a night raised inflammatory cytokines and worsened performance on vigilance tests.
  • In rats, sleep deprivation cut the production of new cells in the hippocampus by as much as half, impaired several kinds of memory and learning, raised oxidative stress, and lowered brain glycogen by up to 40% in 24 hours, which recovery sleep reversed.
  • Lost REM sleep increased fighting in rats. Yet some forms of sleep deprivation reduced anxiety-like and depression-like behaviour, which may relate to the short-lived antidepressant effect sleep deprivation can have in people.

4: The Brain as a Target of Stress

Stress and memory: timing matters

  • Stress hormones act on the hippocampus through two receptor types (MR and GR), and their effects follow an inverted U: moderate levels help memory and high levels hinder it.
  • Joëls and colleagues proposed a rule that McEwen calls very plausible. Stress helps learning when it happens in the context of, and around the time of, the thing to be remembered, and when the stress hormones act on the same circuits the situation is using. Stress that comes some time before learning sets off a slower, gene-driven process (taking more than an hour) that raises the threshold for new input, and then it impairs learning.
  • In one study, the same dose of stress hormone that improved memory tested 24 hours later impaired memory tested 1 hour later.

New neurons

In adult rats, the dentate gyrus of the hippocampus makes around 9,000 new neurons a day, with a half-life of about 28 days. Exercise, learning, oestrogen, IGF-1 and antidepressants increase this. Many chronic stressors suppress it. (See section 8 on whether this happens in adult humans.)

Remodelling, not damage

  • Three weeks of daily restraint stress makes the branches (dendrites) of neurons in the CA3 region of the hippocampus shorter and simpler. The same happens in both dominant and subordinate rats living in a social hierarchy, and in tree shrews.
  • Hibernating hamsters and ground squirrels go through the same retraction within hours of entering hibernation and reverse it within hours of waking. McEwen takes this as evidence that the change is structural flexibility, not damage.
  • In stressed rats, both the shrinkage and the memory problems disappeared within one to two weeks of the stress ending.
  • He suggests the retraction may protect neurons by reducing excitatory input, though chronically stressed rats were also more vulnerable to one kind of chemical damage. The question was open.
  • Glutamate, the main excitatory neurotransmitter, sits at the centre of these changes, along with stress hormones, BDNF, CRF and other molecules.

Prefrontal cortex and amygdala

  • Chronic stress shrinks dendrites in the medial prefrontal cortex. Stressed rats were worse at shifting attention to new rules, and one report found they were worse at extinguishing a learned fear.
  • In the amygdala, the opposite happens: dendrites grow. Chronically stressed rats show more fear, more fear learning, and more aggression towards cage-mates. Chronic stress hormone in drinking water made mice more anxious.
  • These three regions talk to each other. The prefrontal cortex damps down the amygdala and is needed for extinguishing fear, while the amygdala and hippocampus together process emotional memories in context.

Puberty and sex differences

  • Rats around puberty have a longer stress response than adults. Chronic stress before puberty stunted the growth of the hippocampus, with effects that showed up weeks later.
  • Female rats did not show the CA3 shrinkage unless their ovaries were removed. Chronic stress reduced new cell production in males housed alone but increased it in females, and group housing prevented both effects.

5: Stress in People

McEwen is careful here: there is very little evidence about whether ordinary life stress changes the structure of the human brain. Most of what we know comes from illness.

  • Depression: recurrent depression goes with a smaller hippocampus and prefrontal cortex, while the amygdala can be larger in a first episode. Autopsy studies point to loss of glial cells rather than neurons.
  • Cushing’s disease: the body makes too much cortisol, and depressive symptoms improve when this is corrected surgically. In both Cushing’s and major depression, how long the illness has lasted, not age, predicts how much the hippocampus shrinks.
  • PTSD and borderline personality disorder have also been linked to a smaller hippocampus.
  • Too little can be a problem too: an inadequate cortisol response at the time of trauma appears to make PTSD more likely. In a rat strain that makes little stress hormone, giving it before a predator-odour stress reduced anxiety afterwards.
  • Chronic fatigue, fibromyalgia and irritable bowel syndrome look like an imbalance across the network (the “inadequate response” pattern): in chronic fatigue, low cortisol and aldosterone alongside raised inflammatory cytokines. There is no single, uniform HPA fault.
  • Burnout overlaps with these: emotional exhaustion, depersonalisation, low sense of accomplishment and low self-esteem. Lower cortisol has been reported, though not consistently, and chronic burnout has been linked to a higher risk of type 2 diabetes.
  • Pain: distress and strong emotion drive the symptoms of IBS, fibromyalgia and jaw-joint disorder, with changes in brain regions involved in arousal, pain and emotion. The placebo effect, where expectation changes how much pain is felt, shows how much the brain regulates the body from the top down.
  • Food: stress often triggers comfort eating. The hippocampus helps limit food intake, and people who are or have been obese show less activity there after a satisfying meal than lean people do.

Outlook, self-esteem and company

  • Positive mood across a working or leisure day went with lower cortisol, higher heart rate variability (a sign of parasympathetic activity) and a smaller clotting-factor response to a stress test.
  • Asked to give a public speech more than once, people with good self-esteem showed a smaller cortisol response after the first time. People with poor self-esteem didn’t get used to it.
  • Poor self-esteem and a low sense of control over one’s life have been linked to a 12–13% smaller hippocampus and higher cortisol during a mental arithmetic task. McEwen notes the cortisol may be both cause and result of the smaller hippocampus.
  • Loneliness went with a bigger cortisol rise on waking, larger immune and clotting responses to a stress test, and poorer sleep. People with three or more regular social contacts had lower allostatic load scores than those with two or fewer.

Socioeconomic status

Health follows income and education in a straight line: low status fares worse than middle, and middle worse than high, for disease, obesity, metabolic syndrome, substance abuse, anxiety and mood disorders. Where people rank themselves predicts their health too. Likely links include worry about money, sense of control, and self-esteem.

6: Managing Chronic Stress

Because the brain is the central organ of the stress response, McEwen says it is the main target for reducing chronic stress. The interventions are familiar: better sleep, more social support, a positive outlook, a healthy diet, not smoking, and regular moderate exercise. He admits these are easier said than done, and argues that government and employers can make them easier, as they did with smoking and seat belts.

Medication

Sleep medicines, anti-anxiety drugs, beta blockers, antidepressants, and drugs for inflammation, cholesterol, insulin resistance and pain all help to a degree, and all have side effects and limits. Because the stress network is interlinked, a drug can block the useful side of the system it targets or disturb another system. Fever is his example: it is part of fighting an infection, so suppressing it completely is unwise, yet the runaway response of septic shock has to be contained.

Physical activity

  • A sedentary life is a major risk factor for obesity, diabetes, heart disease, depression and dementia, and moderate activity benefits the brain, heart and metabolism.
  • In animals, voluntary exercise raises growth factors in the brain and increases new neurons, partly through IGF-1. Eating less (dietary restriction) also increased new neurons and BDNF in the hippocampus.
  • Context matters: rats housed alone had raised stress hormone, and running didn’t increase their new neurons. When their hormone levels were brought back to normal, the benefit of exercise returned.

Social support

Regular contact with supportive friends, family or health professionals, giving both emotional support and useful information, lowers allostatic load scores. Social support also eased a kind of caregiver stress that has been linked to shorter telomeres in white blood cells. How support benefits the stressed brain circuits themselves was unknown.

Policy

The UK’s Acheson Report (1998) held that no public policy should pass without considering its effect on health: education, housing, tax, the minimum wage, workplace safety and pollution all reach health. Affordable good food, safer and friendlier neighbourhoods, and employers who support healthy habits all lower the load.

McEwen’s favourite example combines activity, company and purpose. Experience Corps trains older volunteers as teachers’ aides in local schools. The children’s education improves, and so does the volunteers’ physical and mental health, with a slower age-related decline.

7: Conclusions

  • The adult brain is malleable. It changes its structure and function in response to stress, and those changes are not necessarily damage, though they can last.
  • Reversing those changes, whether they recover on their own or with behavioural or drug treatment, may be the key to treating anxiety, mood and other stress-related disorders.
  • Early life shapes how the brain responds to stress for the rest of life, and affects ageing and the risk of heart disease, diabetes and depression.
  • Social factors and physical activity have a powerful effect on the brain across life. Policy plus individual change can counter chronic stress, as an addition to medication and in some cases an alternative.

8: Reading This in 2026 (HOM note)

This section is ours, not the paper’s. The paper is from 2007, and a few things have moved since.

  • Most of the mechanisms come from rats and mice. McEwen says so himself: there is very little evidence about whether everyday stress changes the structure of the human brain. Treat the remodelling findings as a strong working model for humans, not a demonstrated fact.
  • The “stress genes” didn’t hold up. A 2019 study by Border and colleagues tested 18 of the most-studied depression candidate genes, including the serotonin transporter variant and its interaction with stress, in samples of tens to hundreds of thousands of people. It found no support for them. Read the paper’s genetics section as history.
  • New neurons in adult humans are disputed. In 2018, Sorrells and colleagues reported that new neurons in the human hippocampus drop to undetectable levels by adulthood, while Boldrini and colleagues reported that they persist into old age. The rodent findings on exercise and new neurons are solid; whether they carry over to adult humans is unsettled.

References

  • Boldrini, M., et al. (2018). Human hippocampal neurogenesis persists throughout aging. Cell Stem Cell, 22(4), 589–599. doi:10.1016/j.stem.2018.03.015
  • Border, R., et al. (2019). No support for historical candidate gene or candidate gene-by-interaction hypotheses for major depression across multiple large samples. American Journal of Psychiatry, 176(5), 376–387. doi:10.1176/appi.ajp.2018.18070881
  • McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews, 87(3), 873–904. doi:10.1152/physrev.00041.2006
  • Sorrells, S. F., et al. (2018). Human hippocampal neurogenesis drops sharply in children to undetectable levels in adults. Nature, 555(7696), 377–381. doi:10.1038/nature25975

More to explore