The HPA Axis — Three Glands, One System
The hypothalamic-pituitary-adrenal axis is a three-tier hormonal cascade that coordinates the body’s response to stress — physical, psychological, inflammatory, or metabolic. The three components communicate through a sequential chemical signalling chain: the hypothalamus instructs the pituitary, the pituitary instructs the adrenal glands, and the resulting cortisol output feeds back to both the hypothalamus and pituitary to regulate its own production. This feedback loop is the central regulatory mechanism — and its failure is the mechanism of HPA axis dysfunction.
The hypothalamus is the integrating centre — it receives input from the cerebral cortex (conscious thought, worry, anticipation), the limbic system (emotional processing, memory, the amygdala’s threat assessment), the brainstem (physiological signals from the body), and the circadian system (time of day via the suprachiasmatic nucleus). All of these inputs converge on the paraventricular nucleus (PVN) of the hypothalamus, which synthesises and releases corticotrophin-releasing hormone (CRH) in response to perceived demand.
CRH travels through the hypothalamic-portal blood system to the anterior pituitary, where it binds to CRH receptors on corticotroph cells, stimulating the synthesis and release of adrenocorticotrophic hormone (ACTH). ACTH is released into the systemic circulation and travels to the adrenal glands.
The adrenal glands — small, pyramid-shaped glands sitting atop each kidney — respond to ACTH at the zona fasciculata of the adrenal cortex, stimulating cholesterol mobilisation from lipid droplets and the enzymatic cascade that converts cholesterol to cortisol via pregnenolone, progesterone, and 17-hydroxyprogesterone. The entire sequence from hypothalamic CRH to circulating cortisol takes approximately 15–30 minutes in an acute stress response.
Pregnenolone is the first step in making every steroid hormone, which is why the popular “steal” idea says chronic stress diverts a shared pool of it into cortisol. There is no shared pool. Each steroid-making cell makes its own pregnenolone from cholesterol and uses it locally; cortisol (zona fasciculata) and DHEA (zona reticularis) are made in different adrenal layers with different regulators, and progesterone and most testosterone come from the ovaries and testes (Endotext, Adrenal Androgens).
Low progesterone, low DHEA-S and falling testosterone do occur together in chronically stressed people. The supported mechanism is signalling: stress turns down GnRH, LH and FSH (see Section 6), and DHEA falls with age and in some people under long-term stress. Low DHEA metabolites alongside high cortisol metabolites on a DUTCH is a pattern worth noting; it isn’t evidence of a steal.
Cortisol Synthesis and Metabolism — What the DUTCH Measures
Cortisol circulates in two forms: approximately 90–95% is bound to corticosteroid-binding globulin (CBG) and albumin — physiologically inactive, a circulating reservoir. The remaining 5–10% is free cortisol — the bioavailable fraction that enters cells, binds glucocorticoid receptors, and produces biological effects. Standard blood cortisol tests measure total cortisol (bound + free), which is primarily a measure of CBG levels rather than bioavailable cortisol. This is why blood cortisol is an unreliable indicator of functional cortisol status — oestrogen significantly raises CBG, increasing total cortisol without increasing free cortisol. The DUTCH measures free cortisol in urine, providing a direct measure of bioavailable cortisol output.
Cortisol is metabolised primarily in the liver by two enzymes: 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) regenerates active cortisol from inactive cortisone in peripheral tissues — particularly important in adipose tissue, liver, and brain where local cortisol concentrations are regulated independently of circulating levels. 11β-HSD2 converts active cortisol to inactive cortisone — a protective mechanism in tissues like the kidney that should not be exposed to glucocorticoid signalling continuously.
The metabolised forms — the “spent” cortisol and cortisone after they have been reduced and conjugated by the liver — appear in urine as tetrahydrocortisol (THF), allo-THF, and tetrahydrocortisone (THE). The DUTCH measures these metabolites as the total metabolised cortisol output — the most complete picture of the body’s total 24-hour cortisol production available from a non-invasive test. High total metabolised cortisol with normal free cortisol indicates high production but efficient clearance. High free cortisol with lower metabolites indicates reduced clearance. Low everything indicates reduced production — the late-stage burnout pattern.
Elevated urinary cortisone relative to cortisol (as in the pattern described in the constipation post — saliva cortisone 64, range 36–55) indicates that 11β-HSD2 is actively converting excess cortisol to cortisone across tissues. The body is producing more cortisol than its receptor sensitivity can comfortably manage and is attempting to inactivate the excess. This is early-stage HPA axis overactivation — the system is working hard but is beginning to protect peripheral tissues from cortisol overexposure through enzymatic inactivation.
The Cortisol Awakening Response — The Most Informative 60 Minutes of the Day
The cortisol awakening response (CAR) is a discrete, specific component of the diurnal cortisol rhythm — distinct from the overall morning rise. It is a rapid 50–100% surge in cortisol that occurs in the first 20–30 minutes after waking, triggered by the environmental cue of light exposure and the circadian transition from sleep to wakefulness through the hypothalamic suprachiasmatic nucleus.
The CAR is not simply “high morning cortisol.” It requires four measurements at specific time points to capture properly: immediately on waking (0 minutes), at +15 or +30 minutes (the peak), and declining measurements at +45 and +60 minutes. The magnitude of the rise from waking to peak is the CAR. A healthy CAR represents a 50–100% increase from the waking value. The DUTCH Plus captures this through four timed saliva samples taken by the client on the morning of collection.
The CAR has distinct physiological functions. It primes the immune system for the day — mucosal sIgA production is partially dependent on the morning cortisol surge, which is why depleted sIgA on the GI-MAP and a blunted CAR on the DUTCH frequently co-occur. It mobilises glucose for cognitive function — the early morning blood glucose rise is partially cortisol-driven. It prepares the cardiovascular system — heart rate and blood pressure peak in the early morning partly through the CAR. It regulates inflammatory tone for the day ahead.
The CAR also reflects the brain’s anticipatory assessment of the day’s demands. People with high work demands, high perceived stress, or significant anticipatory anxiety show consistently exaggerated CARs — the HPA axis is preparing for a day it expects to be difficult. Shift workers, people with circadian misalignment, burnout, PTSD, and chronic fatigue syndrome all show significantly disrupted CAR patterns. The CAR is not measuring how stressed someone is in that moment — it is measuring how responsive the HPA system is, which is a different and more informative clinical question.
Strict timing is mandatory. The client must take the first sample the moment they wake — before getting up, before drinking anything, before checking their phone. Any delay of even 10 minutes significantly distorts the waking value and therefore the calculated CAR magnitude.
Workday collection preferred for most clients. A day off produces a different CAR than a workday — lower anticipatory arousal, later waking time, reduced perceived demand. The workday CAR is more clinically representative of habitual HPA activation patterns.
Alarm vs natural waking. Alarm-waking produces a more exaggerated CAR than natural waking — the abrupt transition creates a larger acute cortisol response. If the client always uses an alarm, alarm-waking collection is more representative of their typical pattern.
The Diurnal Pattern — Reading the Full Day
After the CAR, cortisol follows a predictable diurnal decline — high in the morning (typically peaking 30–45 minutes after waking), declining progressively through the morning, reaching a mid-point in early afternoon, and falling to its lowest point in the hours before midnight. This diurnal rhythm is driven by the circadian clock in the suprachiasmatic nucleus and is the physiological basis for the advice to align the most cognitively demanding work with the morning cortisol window.
The DUTCH Plus provides four saliva cortisol measurements across the day (waking, +30 minutes, afternoon, evening) plus the metabolised cortisol from the urine collection. This gives both the pattern (the four-point diurnal shape) and the total output (the 24-hour metabolised cortisol sum). Both are clinically necessary — a person with a flat, low diurnal pattern but high total metabolised cortisol has a different clinical picture from a person with a dramatically high morning spike and rapid afternoon crash.
Elevated evening cortisol — the clinical consequence
Elevated cortisol in the evening window (the third or fourth DUTCH saliva sample) is one of the most clinically consequential single findings on the DUTCH. Evening cortisol elevation directly suppresses melatonin secretion from the pineal gland — melatonin synthesis requires darkness and low cortisol, as the two are physiologically antagonistic. Elevated evening cortisol therefore produces delayed sleep onset, fragmented sleep architecture, and reduced delta wave deep sleep. The person lies awake with a busy mind — a direct neurological consequence of cortisol’s arousal effects on the locus coeruleus and prefrontal cortex — not a psychological habit of worrying. The sleep disruption then drives morning fatigue, impairs the cortisol clearance that occurs during deep sleep, and perpetuates elevated evening cortisol the following night. The cycle is self-reinforcing.
Elevated evening cortisol is not “stress.” It is a specific hormonal state with documented neurological consequences — melatonin suppression, locus coeruleus activation, impaired prefrontal inhibition of the default mode network. The person is not choosing to lie awake thinking. Their cortisol is keeping them awake.
Cortisol Patterns — Three Shapes, Not a Staircase
A DUTCH collection day usually falls into one of three shapes. Naming the shape matters, because high output and low output call for different emphasis. What the evidence doesn’t support is the popular four-stage “burnout trajectory”, in which everyone climbs from activation to resistance to exhaustion to burnout as the adrenal glands wear out. The adrenals don’t wear out (Cadegiani & Kater 2016).
There is some support for a drift over time: in a meta-analysis, cortisol tended to be higher soon after a stressor began and lower the longer it had gone on (Miller 2007), and low cortisol after long-running stress has been proposed as an adaptation of the axis rather than a failure (Fries 2005). But people move between these patterns and back, and one collection day can be thrown by a bad night. Evidence: mixed for a high-to-low drift; none for fixed stages or for adrenal depletion.
High output points to reducing load: sleep timing, evening wind-down, workload, stimulants, and blood sugar swings.
Low output points to ruling out medical causes first, then eating enough, pacing activity rather than pushing hard training, morning light, and treating whatever is keeping the axis turned down (poor sleep, inflammation, depression, under-eating).
Disrupted rhythm points to light, sleep and meal timing before anything else.
What the pattern doesn’t do is pick supplements. No trial has chosen an adaptogen or nutrient by DUTCH pattern, so rules like “ashwagandha only for high output” are untested. Evidence: none for pattern-matched supplement rules.
Downstream Consequences — What Cortisol Dysregulation Drives
Immune suppression and gut vulnerability
Glucocorticoid receptors are present on virtually every immune cell type. Cortisol at chronically elevated levels suppresses Th1 immune function (cellular immunity, viral defence), promotes Th2 dominance (allergic and antibody responses), and reduces secretory IgA production at mucosal surfaces. The practical consequence: the chronically stressed person is more susceptible to viral infections, develops more food sensitivities (as reduced sIgA allows antigens to reach the gut immune system that would normally be neutralised in the lumen), and shows progressively worsening gut dysbiosis as mucosal immunity becomes inadequate to maintain commensal ecology.
Insulin resistance and metabolic consequences
Covered in detail in Study Guide 01 (Insulin Resistance). Briefly: cortisol drives hepatic gluconeogenesis, promotes adipocyte lipolysis, and directly impairs peripheral insulin receptor signalling. The DUTCH CAR magnitude is one of the best available predictors of fasting insulin elevation and HOMA-IR trajectory — a person with a 67% CAR is driving morning hepatic glucose output that precedes the first meal of the day and sets the insulin tone for the morning.
Thyroid conversion impairment
Covered in Study Guide 02 (Thyroid Conversion). Cortisol at elevated levels upregulates type 3 deiodinase (T4 → reverse T3) and downregulates type 1 deiodinase (T4 → T3). The combination of high cortisol output and impaired T3 production produces the tissue hypothyroidism pattern — fatigue, cold intolerance, weight gain, brain fog — with normal TSH and borderline Free T4. The DUTCH CAR sits upstream of the thyroid conversion problem in most cases of stress-related functional hypothyroidism.
Sex hormone disruption
This, not a pregnenolone steal, is the main route by which stress lowers sex hormones. Stress hormones suppress gonadotrophin-releasing hormone (GnRH) from the hypothalamus, reducing LH and FSH from the pituitary and consequently reducing ovarian and testicular hormone production. The evolutionary logic: reproduction is not a priority during sustained threat. The clinical consequence in the modern context: low testosterone in men under chronic occupational stress, irregular cycles in women, anovulatory cycles, and worsening PCOS as the cortisol-insulin loop amplifies the androgen excess. At the extreme this is functional hypothalamic amenorrhoea, a recognised diagnosis (Endocrine Society guideline 2017; mechanism reviewed in Whirledge & Cidlowski 2013).
Gut motility — the sympathetic override
As described in detail in the constipation and gut motility clinical post: CRH from the hypothalamus reaches enteric mast cells directly via the HPA-enteric nervous system connection, triggering mast cell degranulation and modifying gut permeability and motility independently of circulating cortisol. The gut knows about stress through the nervous system before the adrenal glands have had time to respond. This is the mechanism behind stress-induced IBS, urgency with acute anxiety, and the chronic motility suppression of sustained sympathetic dominance.
Feedback Failure — Why the HPA Axis Loses Regulation
Under healthy physiology, rising cortisol feeds back to suppress its own production. Cortisol binds to glucocorticoid receptors in the hippocampus, the hypothalamus (reducing CRH release), and the anterior pituitary (reducing ACTH release). This negative feedback loop is the regulatory mechanism that prevents cortisol from rising indefinitely.
Under chronic stress, this feedback mechanism fails progressively through two concurrent mechanisms. First, glucocorticoid receptor downregulation: chronic cortisol exposure reduces the number and sensitivity of glucocorticoid receptors on hippocampal neurons and hypothalamic cells — the feedback signal becomes weaker as the receptor population shrinks. Second, hippocampal damage: cortisol at chronically elevated levels is directly neurotoxic to hippocampal CA3 neurons, reducing the hippocampal volume that normally constrains HPA axis activity. Chronic stress literally damages the brain region responsible for switching off the stress response. MRI studies of people with chronic burnout, PTSD, and major depression consistently show reduced hippocampal volume.
The clinical implication: HPA overactivation, if sustained for long enough, progressively impairs the feedback mechanism, making it harder for the axis to self-regulate regardless of the stressor load reducing. This is why people who have been in chronic stress for years do not simply recover when the external stressor resolves — the regulatory architecture itself has been compromised and requires specific intervention to restore feedback sensitivity.
Phosphatidylserine (PS) is the best-studied intervention for glucocorticoid receptor sensitisation. PS (300–800mg daily) has documented effects on cortisol response to psychological and physical stress in RCTs — blunting the ACTH response to CRH and reducing the cortisol overshoot to standardised stress protocols. The mechanism: PS is a structural component of neuronal cell membranes and improves glucocorticoid receptor function in hippocampal tissue. Particularly relevant at Stages 1 and 2.
Sleep is the primary mechanism of hippocampal restoration — deep sleep (delta wave, slow-wave sleep) is when hippocampal neurogenesis occurs and cortisol clearance is maximised. Addressing sleep architecture is therefore not a symptom-level intervention but a mechanism-level one for HPA feedback restoration. Melatonin (0.5–1mg, not the 5–10mg doses commonly used), magnesium glycinate (400mg evening), and phosphatidylserine together address the evening cortisol excess that is preventing deep sleep.
Omega-3 fatty acids (EPA 2–3g daily) reduce hypothalamic NF-κB activation, reducing CRH secretion and dampening the inflammatory drive to HPA axis activation that perpetuates the cycle in people with both chronic stress and gut-derived inflammation.
What Moves the Pattern — Graded
Sleep, light, meal timing, eating enough and workload come first for every pattern. The supplements below are the ones most often used for the HPA axis, graded on the evidence; none has been tested by DUTCH pattern.
Phosphatidylserine
In small trials, 400 mg a day of a soy phosphatidylserine/phosphatidic acid complex blunted the ACTH and cortisol rise to a lab stress test; 600 and 800 mg did not (Hellhammer 2004). In chronically stressed men, 400 mg normalised an exaggerated response (Hellhammer 2014). The trials are small and linked to the manufacturer, and none selected people by DUTCH pattern.
Ashwagandha
Root extract, 300 mg twice a day for 60 days, lowered stress scores and serum cortisol against placebo in stressed adults (Chandrasekhar 2012). Trials are small and mostly industry-funded. Liver injury has been reported, usually cholestatic and self-limiting, but fatal in people with existing liver disease (Björnsson 2020; Philips 2023). Avoid with liver disease. The idea that it harms people with low output is untested.
Rhodiola
Not checked against the evidence for this guide. Treat claims about pattern-specific use as untested.
Adrenal glandulars
No trials. Products sold for “adrenal support” have been found to contain undeclared thyroid hormone and steroids, including a prescription corticosteroid in a quarter of those tested (Akturk 2018). Not recommended.
Melatonin
Low doses (0.5–1 mg, 30–60 minutes before bed) are closer to the body’s own levels than the 5–10 mg commonly sold. Its effect on cortisol is untested. Prescription-only in the UK.
Exercise snacks (VILPA)
One to two-minute bursts of vigorous everyday activity, a few times a day, were associated with lower mortality in UK Biobank non-exercisers (Stamatakis 2022). Exercise raises cortisol briefly rather than “metabolising” it. With a low-output pattern and post-exertional symptoms, pace activity and build slowly.
Vitamin C
3 g a day of sustained-release vitamin C for 14 days reduced blood pressure and subjective stress responses to a lab stressor and sped cortisol recovery, without changing adrenal responsiveness (Brody 2001). There is no evidence it “supports” adrenal output.
Pantothenic acid (B5)
B5 is part of coenzyme A, which steroid synthesis uses, but there are no human trials of B5 for cortisol patterns, and deficiency is rare.
What to Test — The Complete HPA Axis Assessment
Step 1: DUTCH Plus — complete HPA picture in one test. CAR (pattern identification), total output (production capacity), free cortisol (bioavailability), cortisone (inactivation pattern), DHEA-S (adrenal androgens), melatonin (sleep impact). This is the non-negotiable assessment — no clinical HPA intervention should be made without it.
Step 2: GI-MAP sIgA — immune consequence of the adrenal pattern. Low sIgA with blunted CAR confirms the gut immune consequence and guides restoration sequencing.
Step 3: HTMA Na/Mg and Na/K — the 8–12 week adrenal mineral picture that contextualises the DUTCH snapshot. DUTCH tells you what is happening in 24 hours. HTMA tells you what has been happening over months.
Step 4: Blood chemistry — hsCRP (inflammatory driver), HOMA-IR (metabolic consequence), Free T3 (thyroid conversion consequence), ferritin (immune and adrenal substrate). These four blood markers complete the downstream consequence picture.
Name the pattern, and rule out medical causes of very low output, before choosing interventions. Read it with the symptoms and the collection day: it describes that day, not a stage.