Stress and the HPA Axis
A student stares at an exam question she cannot answer. Within seconds her heart pounds, her palms sweat, and her stomach knots. Meanwhile, unseen, a slower hormonal cascade is only just beginning to unfold. Why does the body need two stress systems running on two different clocks? The fast one is built on sympathetic nerves and adrenal catecholamines. It buys milliseconds-to-minutes. The slow one is built on the hypothalamic-pituitary-adrenal (HPA) axis. It reshapes metabolism, immunity, and brain function over minutes to days. Understanding both — and how they talk to each other — is central to physiology. "Stress" is not a vague feeling. It is a precisely wired, quantifiable neuroendocrine program. The HPA axis matters far beyond exam anxiety. It governs how the body mobilizes fuel during illness, trauma, and starvation. It also dampens inflammation. Chronic activation of this axis contributes to hypertension, immunosuppression, depression, and metabolic disease. Nearly every organ system carries Cortisol Receptors. This makes it one of the most physiologically pervasive signaling systems in the body. Dysregulation of this single axis — too much, too little, or mistimed — underlies many conditions. These range from Cushing syndrome to Addison disease to post-traumatic stress disorder. The big picture is a layered defense system. A stressor is detected, and the brain interprets it. Two parallel outputs then fire. One is a fast neural output (sympathetic-adrenomedullary, SAM). The other is a slower endocrine output (HPA). Both converge on the same goal. They redirect the body's resources toward immediate survival. But they do it on different timescales, through different chemistries. They also carry different consequences if left switched on too long. Negative feedback loops normally shut the HPA axis off once the threat passes. Chronic stress is, mechanistically, a failure of that shutoff. This article builds the HPA axis from the hypothalamus down to the target tissue. It traces the classic experiments that established each link in the chain. It also works through quantitative examples — from a diurnal cortisol curve to a clinical case of adrenal failure. The goal is for the mechanism, not just the vocabulary, to become usable.
- Stress activates two systems: fast sympathetic-adrenomedullary (epinephrine, seconds) and slow HPA (cortisol, hours).
- The HPA cascade is CRH (hypothalamus) → ACTH (anterior pituitary) → cortisol (adrenal cortex).
- Cortisol has permissive, metabolic, anti-inflammatory, and immunosuppressive effects, and feeds back to shut off its own axis.
- Classic experiments (Cannon, Selye, Harris, Vale) established the neural, hormonal, and molecular links in this chain.
- Cortisol follows a diurnal rhythm, peaking near waking and troughing near midnight, on top of stress-induced pulses.
- Chronic HPA activation — or feedback failure — links stress to hypertension, immune suppression, depression, and metabolic syndrome.
The HPA axis converts a perceived threat into a hormonal cascade — CRH → ACTH → cortisol — whose job is to mobilize energy and dampen non-essential processes, then shut itself off via negative feedback once the threat resolves.
The core idea
Stress, in physiological terms, is any real or perceived challenge to homeostasis — a stimulus (the stressor) that triggers a coordinated neural and hormonal response (the stress response). Hans Selye, who popularized the term, distinguished a general adaptation syndrome with three stages: alarm (rapid mobilization), resistance (sustained adaptation), and exhaustion (if the stressor persists too long, systems fail). Selye's insight was that the physiological response is remarkably similar regardless of whether the stressor is a predator, a pathogen, a burn, or a cold room — a nonspecific, stereotyped reaction layered on top of specific responses to the particular insult. The HPA axis is the endocrine backbone of that response. It is a three-tier cascade: the paraventricular nucleus (PVN) of the hypothalamus secretes corticotropin-releasing hormone (CRH), which stimulates corticotrophs in the anterior pituitary to secrete adrenocorticotropic hormone (ACTH), which stimulates the zona fasciculata of the adrenal cortex to secrete cortisol (in humans; corticosterone predominates in rodents). This is a textbook example of a hypothalamic-pituitary axis, sharing its portal-vessel wiring and tropic-hormone logic with the thyroid and gonadal axes, but its output — cortisol — is uniquely tuned for mobilizing glucose, fat, and amino acid reserves and for restraining immune and inflammatory activity. Running in parallel, and activated by the same hypothalamic stress circuitry, is the sympathetic-adrenomedullary (SAM) system: descending sympathetic fibers activate the adrenal medulla directly (as a modified sympathetic ganglion), releasing epinephrine (about 80% of adrenal medullary output) and norepinephrine into the blood within seconds. The SAM system is fast and short-lived; the HPA axis is slower (cortisol rises over 15–30 minutes and can stay elevated for hours) but longer-lasting and more far-reaching in its metabolic and immune effects. Both are stress responses; neither alone is "the" stress response.
Mechanism
Step 1 — Perception and integration. Stressors are of two broad kinds: physical/systemic (hemorrhage, hypoglycemia, infection, pain, cold) and psychological/anticipatory (fear, novelty, social threat). Physical stressors reach the hypothalamus via direct sensory and visceral afferent pathways (e.g., baroreceptor and chemoreceptor input via the nucleus tractus solitarius, cytokine signaling from the immune system). Psychological stressors are processed higher up — the amygdala amplifies threat perception and excites the PVN, while the hippocampus and prefrontal cortex exert an inhibitory, context-dependent brake, damping the response once a stressor is appraised as non-threatening. This limbic gating is why the same event (a loud noise) can be stressful or not depending on context and prior experience.
Step 2 — CRH release. Parvocellular neurons of the PVN synthesize CRH (a 41-amino-acid peptide) and co-release arginine vasopressin (AVP), which strongly synergizes with CRH at the pituitary. These neurons project axons to the median eminence, releasing CRH into the hypophyseal portal circulation — a specialized vascular link that carries hypothalamic hormones directly to the anterior pituitary at high local concentration, bypassing systemic dilution.
Step 3 — ACTH release. CRH binds CRH receptor type 1 (CRHR1), a Gs-protein-coupled receptor on pituitary corticotrophs, raising cAMP and triggering release of ACTH — a 39-amino-acid peptide cleaved from the large precursor pro-opiomelanocortin (POMC). (POMC is also cleaved to melanocyte-stimulating hormone and β-endorphin, which is why severe stress and heavy exercise are accompanied by endorphin release.) ACTH enters the systemic circulation.
Step 4 — Cortisol synthesis and release. ACTH binds the melanocortin-2 receptor (MC2R) on cells of the adrenal cortex's zona fasciculata, again via Gs/cAMP signaling. This up-regulates StAR protein (steroidogenic acute regulatory protein), which shuttles cholesterol into the mitochondrion — the rate-limiting step — and induces the enzymes of steroidogenesis (cholesterol desmolase, 17α-hydroxylase, 21-hydroxylase, 11β-hydroxylase) that convert cholesterol to cortisol. Unlike peptide hormones, cortisol is not stored in vesicles; its release rate is set directly by its synthesis rate, which is why the cortisol response lags ACTH by roughly 15–30 minutes. Cortisol is lipophilic and circulates mostly (≈90%) bound to corticosteroid-binding globulin (CBG, transcortin); only the free fraction is biologically active.
Step 5 — Peripheral action. Free cortisol diffuses into cells and binds intracellular glucocorticoid receptors (GR); the hormone-receptor complex translocates to the nucleus and acts as a transcription factor, altering gene expression over tens of minutes to hours (genomic effects), with some faster non-genomic membrane effects also documented. Key actions:
| System | Cortisol effect |
|---|---|
| Glucose metabolism | ↑ hepatic gluconeogenesis, ↓ peripheral glucose uptake ("glucose-sparing" for brain) |
| Protein/fat metabolism | ↑ proteolysis (muscle), ↑ lipolysis (adipose), supplying gluconeogenic substrate |
| Cardiovascular | Permissive: potentiates catecholamine action on vasculature, maintaining blood pressure |
| Immune/inflammatory | ↓ cytokine production, ↓ prostaglandin synthesis, stabilizes lysosomal membranes — net anti-inflammatory and immunosuppressive at high levels |
| CNS | Alters mood, memory consolidation (via hippocampal GR/mineralocorticoid receptors), and — via feedback — its own further secretion |
Step 6 — Negative feedback. Cortisol feeds back at three levels: it inhibits CRH release from the hypothalamus, inhibits ACTH release from the pituitary, and inhibits its own further stimulation, forming a classic closed-loop negative feedback system. This feedback operates on multiple timescales: fast feedback (within minutes, via non-genomic membrane GR effects) and delayed/genomic feedback (over hours, suppressing POMC transcription). Chronic stress, or diseases that impair GR sensitivity, blunt this feedback and allow cortisol to remain pathologically elevated.
Step 7 — Diurnal rhythm superimposed on stress reactivity. Independent of acute stress, the HPA axis has an intrinsic circadian rhythm driven by the suprachiasmatic nucleus (SCN): CRH/ACTH/cortisol pulse in a pattern that peaks shortly after waking (the cortisol awakening response) and reaches its nadir around midnight. A stressor superimposes an additional pulse on top of this baseline rhythm — so the same stressor produces a different absolute cortisol level depending on time of day.
The evidence
Walter Cannon and "fight-or-flight" (1910s–1930s). Cannon, studying cats exposed to barking dogs, showed that sympathetic activation released adrenal medullary secretions that mobilized glucose, redirected blood flow to muscle, and increased heart rate — coining the term "fight-or-flight" and the broader concept of homeostasis. Crucially, Cannon showed that denervating the adrenal medulla, or removing it, abolished the glucose and cardiovascular surge normally seen with acute emotional stress, establishing that the sympathetic-adrenal link, not some diffuse "vital force," drove the acute response.
Hans Selye and the General Adaptation Syndrome (1936 onward). Selye's classic (initially serendipitous) experiments involved injecting rats with impure ovarian extracts, expecting to find a new hormone. Instead, he found that virtually any noxious stimulus — cold, surgical trauma, forced exercise, toxic substances — produced the same triad of changes: adrenal cortex enlargement, thymic and lymphoid involution (shrinkage), and gastric/duodenal ulcers. Because the effect was nonspecific to the stimulus, Selye concluded the body mounts a generic "stress response" independent of the particular insult, which he formalized as the General Adaptation Syndrome. This work first drew attention to the adrenal cortex, rather than only the medulla, as central to the stress response — implicating what would later be identified as the CRH-ACTH-cortisol axis.
Geoffrey Harris and the discovery of the hypophyseal portal system (1940s–1950s). Harris demonstrated, via careful vascular anatomy and transection/regeneration experiments, that the anterior pituitary has no direct nerve supply controlling hormone secretion, but instead depends on a portal blood supply from the median eminence. Cutting the pituitary stalk and interrupting the portal vessels abolished stress-induced ACTH release even though the pituitary gland itself remained intact and vascularized by other routes — proving that hypothalamic control of the pituitary is humoral (chemical, via portal blood) and not merely neural. This established the anatomical basis for hypothalamic releasing hormones controlling pituitary tropic hormones.
Roger Guillemin and Andrew Schally — isolating the releasing hormones (1950s–1970s). Working independently and competitively, Guillemin's and Schally's groups processed literally tons of sheep and pig hypothalamic tissue to isolate minute quantities of hypothalamic releasing factors, first proving such factors existed by showing hypothalamic extracts could trigger pituitary hormone release in vitro and in hypophysectomized/portal-intact animals. This work (Nobel Prize, 1977) established TRH and GnRH first; CRH itself was finally isolated and sequenced by Wylie Vale in 1981, using ovine hypothalamic extract, confirming it as a 41-residue peptide sufficient to stimulate ACTH release from cultured pituitary cells — the direct molecular proof of the top tier of the HPA cascade.
Demonstrating negative feedback — adrenalectomy and glucocorticoid replacement studies. Classic endocrine "remove-and-replace" experiments showed that adrenalectomy in animals causes ACTH and CRH to rise sharply (loss of feedback inhibition), while replacement with exogenous glucocorticoids suppresses ACTH back toward baseline in a dose-dependent manner. This logic — surgical removal, observation of compensatory hypertrophy/hypersecretion upstream, and reversal by hormone replacement — is the general experimental template used throughout endocrinology (also used to establish, e.g., thyroid and gonadal axis feedback) and directly proved that cortisol suppresses its own upstream secretagogues, closing the loop diagrammed above.
Worked examples
Example 1 — Reading a dexamethasone suppression test. A patient has an elevated morning cortisol (32 µg/dL; normal ≈ 5–23 µg/dL). She is given a low dose of dexamethasone, a synthetic glucocorticoid that occupies GR and CRH/ACTH-suppressing feedback sites but is not measured by the cortisol assay. Two outcomes are possible:
- If cortisol suppresses to <1.8 µg/dL the next morning, the HPA axis's negative feedback is intact — implying any elevated cortisol was due to a normally-regulated cause upstream (e.g., physiological stress, obesity, depression — so-called pseudo-Cushing states) rather than autonomous adrenal or pituitary disease.
- If cortisol fails to suppress, the feedback loop itself is broken — pointing to Cushing syndrome: either a pituitary corticotroph adenoma secreting ACTH independent of feedback (Cushing disease), an ectopic ACTH-secreting tumor, or a cortisol-secreting adrenal adenoma acting independently of ACTH altogether.
A high-dose dexamethasone test can further distinguish these: pituitary adenomas typically retain some residual feedback sensitivity and suppress at high doses, whereas ectopic ACTH sources and adrenal tumors generally do not. This reasoning directly exploits the feedback architecture built into the HPA cascade — the test is a controlled perturbation of the very loop diagrammed in the mechanism section, and its logical structure mirrors the classic adrenalectomy/replacement experiments above.
Example 2 — Quantifying the two-timescale response to a hemorrhage. A patient loses 1200 mL of blood (roughly 25% of a 5 L blood volume) in a trauma. Trace the two stress systems on their respective clocks:
- 0–30 seconds (SAM system): falling blood pressure is sensed by aortic and carotid baroreceptors; decreased baroreceptor firing disinhibits the vasomotor center, triggering sympathetic outflow. Adrenal medullary epinephrine and norepinephrine surge within seconds, producing tachycardia, vasoconstriction in skin/gut/kidney, and glycogenolysis — buying time by preserving perfusion pressure to heart and brain.
- Minutes concurrently: falling blood pressure and pain/tissue-damage signals also activate the PVN, releasing CRH; ACTH rises within roughly 5–10 minutes.
- 15–30 minutes (HPA axis): cortisol rises measurably, peaking over the following 1–2 hours. Its contribution is not to raise blood pressure acutely (that's catecholamines and vasopressin) but to permit those catecholamines to work — cortisol's permissive effect up-regulates α₁-adrenergic receptor expression on vascular smooth muscle. Without adequate cortisol (e.g., in untreated adrenal insufficiency), the same catecholamine surge produces a blunted pressor response, and patients can develop refractory hypotension despite a maximal sympathetic effort — a life-threatening scenario called adrenal crisis. Cortisol also begins mobilizing hepatic glycogen and gluconeogenic substrate over the following hours to sustain glucose delivery to the brain as the acute glycogen reserves from epinephrine's fast action run down.
This example shows why both systems are necessary and non-redundant: SAM acts in seconds but cannot sustain itself metabolically or maintain vascular reactivity alone; HPA acts more slowly but is required to permit and prolong the cardiovascular and metabolic support that survival demands.
Common misconceptions & traps
| Misconception | Correction |
|---|---|
| "Stress" means only psychological distress | Physiologically, a stressor is any homeostatic threat — hemorrhage, cold, infection, surgery, or fear all activate the same HPA machinery |
| Cortisol and adrenaline are the same thing / interchangeable | Cortisol (adrenal cortex, steroid, slow, genomic) and epinephrine (adrenal medulla, catecholamine, fast, receptor-mediated) are chemically and kinetically distinct hormones |
| The HPA axis is "bad" and stress hormones are inherently harmful | Acute cortisol/catecholamine responses are adaptive and essential for survival; chronic, unresolved activation is what causes pathology |
| ACTH acts directly on target tissues to cause stress symptoms | ACTH's main job is stimulating adrenal cortisol synthesis; most "stress symptoms" reflect cortisol and catecholamine action on peripheral tissues, not ACTH itself |
| Cortisol is only "released" like a stored neurotransmitter | Cortisol is synthesized on demand from cholesterol (StAR-dependent) rather than stored in secretory vesicles, so its release kinetics reflect synthesis, not exocytosis of a preformed pool |
| The HPA axis has one fixed "resting" cortisol level | Cortisol has a strong circadian rhythm (peak near waking, trough near midnight); a single random cortisol value is uninterpretable without knowing time of day |
The scientist behind it
Origins · 1 minHans Selye and how this was discovered
Selye trained as a physician in Prague and Vienna before emigrating to Canada, where — while trying to isolate a hypothetical new ovarian hormone at McGill University in the mid-1930s — he noticed that his control-group rats, injected with all manner of irrelevant and even toxic substances, developed the identical triad of enlarged adrenals, shrunken thymus/lymph nodes, and gastric ulcers as his experimental animals. Rather than discard this as noise, Selye recognized it as a general, stereotyped physiological program and spent the rest of his career characterizing it as the General Adaptation Syndrome, popularizing the very word "stress" in its modern biomedical sense. He was a prolific, sometimes polarizing figure — nominated for the Nobel Prize multiple times but never awarded it — and his framework, though later refined by more molecular work on CRH, ACTH, and glucocorticoid receptors, remains the conceptual scaffold on which the modern HPA-axis narrative is built.
Where this matters today
Endocrinology and diagnosis. The dexamethasone suppression test, ACTH stimulation (cosyntropin) test, and diurnal cortisol/ACTH profiling are routine clinical tools for diagnosing Cushing syndrome, Addison disease, and congenital adrenal hyperplasia — all direct applications of the feedback logic described above.
Critical care. Recognizing relative adrenal insufficiency in septic shock — where cortisol fails to rise adequately despite severe physiologic stress — guides decisions about stress-dose glucocorticoid therapy in ICU patients, and is grounded in the permissive-effect mechanism worked through in Example 2.
Psychiatry and behavioral medicine. Chronic HPA dysregulation (either hyperactivity with blunted feedback, or in some cases hypocortisolism) is implicated in major depression, PTSD, and burnout; the cortisol awakening response is used as a research biomarker of chronic stress exposure.
Pharmacology. Synthetic glucocorticoids (prednisone, dexamethasone, hydrocortisone) exploit cortisol's anti-inflammatory and immunosuppressive actions therapeutically in autoimmune disease, transplant rejection, asthma, and allergic reactions — but chronic use requires tapering because exogenous glucocorticoids suppress endogenous CRH/ACTH via the same feedback loop, risking adrenal atrophy and crisis on abrupt withdrawal.
Agriculture and animal welfare. Cortisol (and corticosterone) assays from hair, feathers, or feces are used as non-invasive welfare indicators in livestock and wildlife management, quantifying chronic stress exposure in transport, crowding, or habitat disturbance.
Conservation biology. Field endocrinologists measure fecal glucocorticoid metabolites in endangered species to assess population-level stress from habitat loss, human disturbance, or climate change, linking this classic laboratory physiology directly to ecosystem-scale management decisions — an application that also foreshadows how hormonal axes integrate with broader systems, a theme picked up more fully when considering the Integration of Hormonal Control across multiple endocrine axes simultaneously.
Ask the tutor: How does chronic HPA axis activation specifically alter hippocampal structure and memory, and is this reversible?
What is the difference between the HPA axis and the sympathetic-adrenomedullary (SAM) system?
Both respond to stress but differ in chemistry, speed, and duration. The SAM system is a direct sympathetic neural pathway to the adrenal medulla, releasing catecholamines (epinephrine, norepinephrine) within seconds for immediate cardiovascular and metabolic mobilization. The HPA axis is a three-tier hormonal cascade (CRH → ACTH → cortisol) acting over 15 minutes to hours, producing sustained metabolic, anti-inflammatory, and permissive cardiovascular effects. They are triggered together by the same hypothalamic stress circuitry but operate on different timescales and via different receptors — one fast/neural/reversible, one slow/hormonal/genomic.
Why does cortisol have anti-inflammatory effects, and why is this clinically important?
Cortisol suppresses transcription of pro-inflammatory cytokines, inhibits phospholipase A2 (reducing prostaglandin and leukotriene synthesis), stabilizes lysosomal membranes, and reduces immune cell trafficking and antibody production. Physiologically this limits collateral tissue damage during a strong stress/immune response. Clinically, synthetic glucocorticoids exploit this pathway to treat asthma, autoimmune disease, allergic reactions, and transplant rejection — but chronic high-dose use also causes immunosuppression (increased infection risk), impaired wound healing, and, through prolonged negative feedback, suppression of the patient's own CRH/ACTH/cortisol axis.
What happens if the adrenal glands are surgically removed (adrenalectomy)?
Removing the adrenal cortex eliminates cortisol production entirely. Because cortisol normally suppresses CRH and ACTH via negative feedback, its absence causes ACTH (and CRH) to rise dramatically — a phenomenon used experimentally and clinically to prove the feedback loop exists. Without cortisol replacement, the animal or patient cannot mobilize glucose adequately during stress, loses the permissive support for catecholamine-driven vasoconstriction, and can die from hypotensive shock during even mild stressors — this is essentially an induced adrenal crisis, mirroring Addison disease.
Why does cortisol follow a daily (circadian) rhythm rather than staying constant?
The suprachiasmatic nucleus, the brain's master circadian clock, drives rhythmic CRH/ACTH/cortisol secretion independent of acute stress, producing a peak shortly after waking (the cortisol awakening response) and a trough around midnight. This anticipates the metabolic demands of the coming waking, active period — mobilizing glucose and priming cardiovascular tone before the organism even becomes active — an efficient predictive strategy rather than a purely reactive one. Acute stressors superimpose additional pulses on top of this baseline rhythm, which is why a single stress-related cortisol reading must always be interpreted relative to time of day.
How does chronic stress lead to disease, mechanistically?
Chronic or repeated stressor exposure keeps CRH/ACTH/cortisol chronically elevated, or can blunt the negative feedback loop so cortisol fails to return to baseline. Sustained high cortisol causes prolonged gluconeogenesis and insulin resistance (contributing to metabolic syndrome and hyperglycemia), sustained proteolysis (muscle wasting), immune suppression (increased infection susceptibility, impaired wound healing), and altered hippocampal/prefrontal function (implicated in memory impairment, anxiety, and depression). Selye's "exhaustion" stage in the General Adaptation Syndrome describes this failure state, when an initially adaptive response becomes pathological because the stressor is never resolved.
Why is a single random cortisol measurement often clinically uninformative?
Cortisol is not one fixed number; it fluctuates with circadian rhythm, ultradian pulsatility (roughly hourly pulses), acute stress, illness, and even the stress of blood draw itself. A "normal" reference range (commonly cited around 5–23 µg/dL in the morning) only applies at a specific time of day under baseline conditions. Clinicians therefore rely on dynamic tests — ACTH stimulation tests, dexamethasone suppression tests, or 24-hour urinary/late-night salivary cortisol — that probe the axis's responsiveness and feedback, rather than a single static value, to distinguish normal variation from true HPA pathology.
Is the amygdala-hippocampus interaction part of the HPA axis, or separate from it?
It is the upstream regulatory layer that determines whether and how strongly the HPA axis fires for psychological stressors. The amygdala, detecting threat, excites the PVN and promotes CRH release, while the hippocampus and prefrontal cortex normally exert an inhibitory brake, damping the response once a stimulus is appraised as non-threatening or once cortisol feedback signals sufficiency. This limbic gating is not part of the core CRH-ACTH-cortisol hormonal cascade itself but critically shapes its magnitude and duration — and its dysregulation (e.g., an overactive amygdala or an impaired hippocampal brake, sometimes worsened by cortisol-induced hippocampal changes) is a leading hypothesis for how chronic psychological stress becomes self-perpetuating.