Before anything else: if you are in crisis
This page discusses depression and PTSD, and some people reach it because they feel unbearable and are looking for a physical explanation. If you are having thoughts of suicide or of harming yourself, stop reading and reach a person.
- Canada and the United States: call or text 9-8-8. The 9-8-8 Suicide Crisis Helpline answers 24 hours a day, every day of the year.
- If you are in immediate danger, call 9-1-1 or go to an emergency department.
- Anywhere else, find your local line through the international directory at findahelpline.com.
Depression is treatable, and it is treated by people, not by numbers. There is no cortisol reading, from any device or any laboratory, that tells you whether you are getting better. Nothing on this page should delay a call, an appointment, or a prescription.
The illnesses that hide in this symptom list
Exhaustion, low mood, broken sleep, weight change and "wired but tired" are the symptoms that send people looking for HPA axis dysfunction. They are also the opening symptoms of several treatable diseases, and of one that can kill you. The Endocrine Society's warning is blunt: do not spend time on an unproven label, because you may have adrenal insufficiency, depression, obstructive sleep apnea or another condition, and getting a real diagnosis matters.
Adrenal insufficiency is the dangerous one, because untreated it can progress to an adrenal crisis. StatPearls describes a crisis as weakness, fatigue, nausea, vomiting, diarrhoea, fever, abdominal or back pain, dizziness and low blood pressure that can progress rapidly to shock, confusion and coma, most often triggered by a gastrointestinal or flu-like illness, and reports that adrenal crisis contributed to 10% of deaths among people with adrenal insufficiency. Treatment must never wait for blood results.
Do not wait for a collapse. Vomiting is already the emergency. This applies to you if you have adrenal insufficiency, if you take or recently stopped steroids by any route, or if you have been taking an "adrenal" supplement. You do not need a diagnosis for this to be about you. If you start vomiting, have persistent diarrhoea, or cannot keep fluids or your tablets down, you cannot absorb an oral steroid, and you are in the danger zone. Use your emergency hydrocortisone injection if you have one, then call emergency services or go straight to an emergency department, and say the words: "possible adrenal crisis, may need hydrocortisone." The Endocrine Society guideline directs immediate treatment at suspected adrenal crisis, not confirmed crisis, and patient guidance is explicit: if you are not sure whether it is a crisis, treat it as one. A dose of hydrocortisone that turns out to have been unnecessary is a far smaller problem than a crisis that was waited out.
The prevention protocol, which almost nobody reading this page has been given. If you are on steroid replacement, or on long-term steroids of any kind, these come from the Endocrine Society guideline and from StatPearls, and you should be asking your doctor for every one of them:
- Sick-day rules, written down. StatPearls: double the oral dose for a fever above 38 °C (100.4 °F), triple it above 39 °C (102.2 °F), and increase it for acute illness. Your doctor sets your numbers. Get them on paper before you are ill, because you will not be reasoning well when you need them.
- An emergency injection kit. The Endocrine Society recommends that every patient be given one: 100 mg injectable hydrocortisone, saline and syringes, and that you and someone who lives with you are trained to use it. A kit in a drawer that nobody can use is not a kit.
- A steroid emergency card and a medical alert bracelet. Also an Endocrine Society recommendation. If you arrive unconscious, this is the only thing in the room that can speak for you.
Three signs people use as a filter. All three fail as one.
- Darkened skin. It is the classic sign of Addison's disease, and NIDDK describes it on scars, skin folds, knuckles, elbows and the lining of the mouth. Its absence rules nothing out. Darkening is driven by high ACTH, so it appears only in primary adrenal insufficiency. The forms caused by pituitary disease or by steroid medication run on low ACTH and produce no darkening at all. Unchanged skin is not reassurance.
- Fever. It is a common crisis sign, and StatPearls even sets steroid sick-day doses by it: double the dose above 38 °C, triple it above 39 °C. But high-dose steroids can suppress a fever, so a person taking them may be running a serious infection without one. A crisis with no fever is still a crisis, and so is an infection with no fever. If you take steroids by any route, including inhalers, creams and joint injections, read what a steroid taper does to your cortisol.
- A normal-looking morning cortisol. The Endocrine Society's guideline on primary adrenal insufficiency puts the diagnosis on the ACTH stimulation test, with a peak cortisol below 500 nmol/L (18 µg/dL) at 30 or 60 minutes indicating adrenal insufficiency. A mid-range basal morning cortisol on its own is not conclusive, which is the whole reason the stimulation test exists. Do not let one number close the question.
If you are certain your problem is physical and not psychiatric, this page is still about you. That is not an argument, it is the point. Most of the differential below is physical disease, and every item on it is found with a blood test or a sleep study rather than with a cortisol curve.
| What it could be | What points toward it | How it is actually found |
|---|---|---|
| Adrenal insufficiency | Fatigue, muscle weakness, loss of appetite and weight, abdominal pain, nausea, low blood pressure with dizziness, salt craving, low blood sugar | Morning cortisol with ACTH, confirmed by an ACTH stimulation test. Read low cortisol and how it is diagnosed |
| Cushing's syndrome | Weight centred on the trunk and face, thin skin, easy bruising, purple stretch marks, muscle weakness, high blood pressure | Overnight dexamethasone suppression, late-night salivary cortisol, or 24-hour urinary free cortisol, ordered and read by a clinician |
| Depression | Loss of interest and pleasure, hopelessness, change in appetite and sleep, guilt, thoughts of death | Clinical assessment. No blood or saliva test diagnoses it |
| Obstructive sleep apnea | Unrefreshing sleep, snoring, witnessed pauses in breathing, morning headache, daytime sleepiness | A sleep study |
| Thyroid disease, anemia, diabetes, coeliac disease | Any of them can present as nothing more than exhaustion. Wired but tired, a racing heart, tremor or unexplained weight loss can be an overactive thyroid. One exception is urgent: exhaustion with heavy thirst, frequent urination and unexplained weight loss can be new diabetes, and if vomiting, abdominal pain, deep rapid breathing or confusion join it, that is diabetic ketoacidosis, an emergency today, not an appointment | Routine blood work, which is cheap and fast. Ask for it. Unexplained weight loss on top of exhaustion always needs a doctor's eyes, whatever else you have concluded |
Do not let this page, or any page, be the reason you skipped that blood work. An article can tell you what a system does. It cannot tell you what is happening in yours.
The "adrenal support" supplement trap, and the way out of it
The phrase "HPA axis dysfunction" is attached to a large supplement market, and the products are not what their labels say. Researchers at the Mayo Clinic bought 12 over-the-counter adrenal support supplements and measured what was inside them. All 12 contained thyroid hormone (triiodothyronine, 63 to 394.9 ng per tablet). 25% contained budesonide, a synthetic glucocorticoid. Others contained pregnenolone, cortisone and cortisol. None declared any of it on the label.
Two separate dangers, and it matters that you do not confuse them. The amounts Mayo measured are small. At the label dose, that is up to about 1.3 µg of budesonide a day, a tiny fraction of a therapeutic steroid dose, so a tablet like that is unlikely, on its own, to have shut your axis down. The real problems are that you have been taking an undeclared drug at an unknown dose, that the dose and the contents vary between bottles and between brands, and that products in this market have been found adulterated with steroids at genuinely pharmacological strength. Separately, the Endocrine Society warns that taking adrenal hormone supplements you do not need can cause your adrenal glands to stop working normally, and that afterwards the glands can remain quiet for months, which can lead to a life-threatening adrenal crisis. That warning applies with full force to anything containing a real dose of glucocorticoid, which is precisely what you cannot tell from the label.
So here is the exit, because naming a trap and leaving it open is worse than saying nothing. If you have been taking an adrenal, cortisol or "adrenal support" supplement:
- Get seen within days, not months, and if you have taken it for more than about three weeks, do not just stop on your own. This is an unapproved product with undeclared hormones in it, and there is no version of "keep taking it while I wait for an appointment" that makes sense. But a product that contained a real glucocorticoid dose can leave your axis suppressed even while you still feel well, and stopping abruptly can trigger a crisis when the next illness or injury arrives. If you have taken it for weeks or longer, ask a doctor how to come off it rather than stopping cold. And if you already feel unwell, light-headed, nauseous, weak or vomiting, do not simply walk away from it. That is the pattern of a suppressed axis, the mechanism is the same as coming off prednisone, and you need medical advice now rather than an appointment next month. If you are vomiting or cannot keep fluids down, treat it as a possible adrenal crisis and go to an emergency department.
- Take the actual bottle to a doctor. Not a photo, not the name from memory. Say what you took, at what dose, and for how long.
- Say it again before any surgery, and any time you are seriously ill. Illness, infection and surgery raise the amount of cortisol your body needs. If your axis is suppressed and nobody in the room knows, that is the moment it matters.
- Mention the thyroid hormone too. The thyroid hormone doses Mayo measured were small, but they are undeclared and vary between bottles, and thyroid hormone is what raises heart-rhythm concern. Palpitations, a racing or irregular pulse, or chest pain are reasons to be seen the same day.
What is the HPA axis?
It is the loop that makes cortisol and then tells itself to stop. StatPearls sets it out in three steps. The hypothalamus releases corticotropin-releasing hormone (CRH). CRH tells the anterior pituitary to release adrenocorticotropic hormone (ACTH). ACTH tells the zona fasciculata of the adrenal cortex to make and release cortisol. Then comes the fourth step, and it is the one that matters: cortisol feeds back on both the hypothalamus and the pituitary and inhibits CRH and ACTH. The hormone switches off its own signal.
That is a thermostat. It is why cortisol does not run away with itself, and it is why the axis produces a daily rhythm rather than a flat line. Cortisol concentrations peak in the early morning and reach their lowest at night. The shape of that day is covered in Cortisol 101.
| Step | Where | Signal | What it does |
|---|---|---|---|
| 1 | Hypothalamus (brain) | CRH, corticotropin-releasing hormone | Starts the cascade |
| 2 | Anterior pituitary (brain) | ACTH, adrenocorticotropic hormone | Travels in the blood to the adrenal glands |
| 3 | Adrenal cortex (above the kidneys) | Cortisol | Raises blood glucose, modulates immune and inflammatory responses, supports blood pressure |
| 4 | Back to the hypothalamus and pituitary | Negative feedback | Cortisol inhibits CRH and ACTH, switching its own production off |
Nothing in that loop can get tired. The adrenal glands are not a battery that a hard year drains. When they genuinely fail, medicine has a name for it, a test for it and a treatment for it, and the causes are known ones: autoimmune destruction, pituitary disease, steroid medication, and less commonly infection such as tuberculosis, bleeding into the glands, cancer that has spread to them, and certain drugs including long-term opioids and some cancer immunotherapies.
Is "HPA axis dysfunction" a real diagnosis?
No, and the honest version of that answer has two halves. HPA axis disorders are real, and they are diagnosed every day: adrenal insufficiency, Cushing's syndrome, and the suppressed axis that follows long-term steroid treatment. If a doctor used the words "HPA axis" with you in one of those contexts, you have a real condition, it has a test, and this page is not arguing with you.
What does not exist is the state in between, the one supplement marketing describes: an axis quietly running down, readable on a saliva panel, correctable with a capsule. The Endocrine Society states that no scientific proof exists to support adrenal fatigue as a true medical condition, and that the tests sold for it are not based on scientific facts or supported by good scientific studies.
The systematic review behind that position is worth reading closely. Cadegiani and Kater screened 3,470 articles and included 58 studies. Among the studies measuring the cortisol awakening response, 51.9% found a normal response in fatigued people. Among those measuring the salivary cortisol rhythm across a day, 61.5% found no difference between fatigued patients and controls. The authors' conclusion, in their words: "adrenal fatigue is still a myth." The same paper warns that glucocorticoid therapy should be avoided in these patients, because even low doses raise cardiovascular and bone risk. The longer version of that literature is in our guide to whether adrenal fatigue is real.
What does the research show in depression, PTSD and chronic fatigue?
Real average differences, running in every direction, none of them a test. This is the part that gets flattened by both camps, by people selling a cure and by people waving the whole field away. The findings are genuine. They are also group findings with heavy overlap between groups, which is a different object from a diagnostic.
| Condition | Best available evidence | What it found | What it does not mean |
|---|---|---|---|
| Major depression | Stetler and Miller, meta-analysis of 361 studies, 18,454 people (2011) | Cortisol was higher in depressed people, d = 0.60 (95% CI 0.54 to 0.66). Restricted to higher-quality methods, the effect fell to d = 0.33 (95% CI 0.21 to 0.45) | A modest average shift. The better the method, the smaller the difference. It cannot sort one person from another |
| Major depression, feedback | Nelson and Davis, meta-analysis of 14 dexamethasone suppression test studies (1997) | 41% of non-psychotic and 64% of psychotic depressed patients failed to suppress cortisol after dexamethasone | It also means most non-psychotic depressed patients suppressed normally. A normal result is the commonest result |
| PTSD | Meewisse and colleagues, meta-analysis of 37 studies, 828 people with PTSD and 800 controls (2007) | No overall difference in basal cortisol. Pooled SMD -0.12 (95% CI -0.32 to 0.08). Lower cortisol appeared only in subgroups: plasma or serum sampling, female participants, physical or sexual abuse, afternoon samples | The popular claim that PTSD means low cortisol. The authors concluded that low cortisol in PTSD is found only under certain conditions |
| Anxiety disorders | Zorn and colleagues, meta-analysis of stress-test studies: 14 in depression (n = 1,129), 9 in anxiety disorders (n = 732) (2017) | Direction depended on sex. Women with current depression or an anxiety disorder showed a blunted cortisol response to a social stress test. Men with current depression or social anxiety showed an increased one | That a single "anxiety cortisol pattern" exists. Opposite results in men and women cancel out in any test that ignores sex |
| Chronic fatigue syndrome | Papadopoulos and Cleare, review in Nature Reviews Endocrinology (2012) | Mild hypocortisolism, attenuated daily variation, enhanced negative feedback, and blunted responsiveness of the axis | That the axis is the cause. The same review reports that low activity levels, depression and early-life stress reduce cortisol while psychotropic medication raises it, and states that steroid replacement is not recommended |
Look at the shape of that table. Cortisol runs high in depression, low in chronic fatigue, nowhere on average in PTSD, and in opposite directions by sex in anxiety. A signal that points every way is a real biological finding about populations. It is not a marker of an individual. And in chronic fatigue, the review that documents the abnormality also reports that it sits downstream of inactivity, low mood and medication. Direction of causation is not a technicality. It decides whether treating the hormone would help at all, and the field's own answer, in the same paper, is that steroid replacement is not recommended.
Why doesn't psychiatry measure your cortisol?
Because it was tried, at length, and it did not work as a test. The dexamethasone suppression test was the great hope of biological psychiatry in the 1980s. The American Psychiatric Association task force on laboratory tests reviewed the whole literature and reported sensitivity of about 40% to 50% in major depression, rising to 60% to 70% in severe, psychotic or melancholic cases, with specificity above 90% against healthy controls but ranging from below 70% to above 90% against the other psychiatric conditions it would actually need to be told apart from. Its conclusion: "the clinical utility of the DST as currently understood is limited."
The second finding is the one this page has a duty to carry, because it is the sentence people most want to be false. The same task force reported that a positive test does not add significantly to the likelihood of responding to an antidepressant, and a negative test is not an indication for withholding antidepressant treatment.
Cortisol is not a scoreboard for whether treatment is working. Not the dexamethasone test, not a saliva panel, not a wearable, not anything. Recovery is not a curve moving one way and relapse is not a curve moving the other. Anyone offering you a hormone number as an objective measure of psychiatric progress is selling something that the field examined carefully and set down 40 years ago. If the anxiety itself is what brought you here, the physiology of it is covered in cortisol and anxiety.
Can a wearable or an at-home test measure HPA axis function?
No. We make a cortisol sensor, so it should be us saying so.
Assessing the HPA axis is a clinical procedure. It means paired ACTH and cortisol, and usually a dynamic test: an injection of synthetic ACTH, or a dose of dexamethasone, followed by timed blood draws, interpreted against your history by a physician. It is the challenge that produces the information, not the sitting-still measurement. No consumer device performs a dynamic test, ours included, and no saliva panel sold direct to the public substitutes for one.
A cortisol reading from a general wellness device is not on the same scale as a clinical laboratory result and cannot be compared with one. It is not a screen, not a finding, and not a second opinion. If you want to understand what the hormone does across a day, that is Cortisol 101. If you want to know what is wrong with you, that is a doctor.
This page is general medical information, not medical advice, and not a diagnosis. It summarizes published clinical guidance and peer-reviewed literature from the sources below. Depression, PTSD, anxiety disorders, chronic fatigue syndrome and every disorder of the HPA axis are diagnosed and treated by qualified healthcare professionals. Never start or stop a hormone, a steroid or a supplement on your own. If you are in crisis, call or text 9-8-8.
References
- StatPearls (NCBI Bookshelf). Physiology, Cortisol. (Source for the CRH, ACTH and cortisol cascade, the negative feedback on the hypothalamus and pituitary, and the circadian pattern.)
- Endocrine Society. Adrenal Fatigue. (Source for "no scientific proof exists", the warning about the tests, the risk that unnecessary adrenal hormone supplements leave the glands quiet for months, and the list of conditions that get missed.)
- Cadegiani FA, Kater CE. Adrenal fatigue does not exist: a systematic review. BMC Endocrine Disorders. 2016. (Source for the 58 included studies, the 51.9% and 61.5% figures, and the glucocorticoid warning.)
- Stetler C, Miller GE. Depression and hypothalamic-pituitary-adrenal activation: a quantitative summary of four decades of research. Psychosomatic Medicine. 2011. (Source for 361 studies, 18,454 individuals, d = 0.60 and d = 0.33.)
- The APA Task Force on Laboratory Tests in Psychiatry. The dexamethasone suppression test: an overview of its current status in psychiatry. American Journal of Psychiatry. 1987. (Source for the sensitivity and specificity figures, and for the finding that the test does not predict antidepressant response.)
- Nelson JC, Davis JM. DST studies in psychotic depression: a meta-analysis. American Journal of Psychiatry. 1997. (Source for the 64% and 41% non-suppression rates.)
- Meewisse ML, Reitsma JB, de Vries GJ, Gersons BPR, Olff M. Cortisol and post-traumatic stress disorder in adults: systematic review and meta-analysis. British Journal of Psychiatry. 2007. (Source for the 37 studies, the 828 and 800 participants, and the pooled SMD of -0.12.)
- Zorn JV, et al. Cortisol stress reactivity across psychiatric disorders: a systematic review and meta-analysis. Psychoneuroendocrinology. 2017. (Source for the sex-dependent blunting and the study counts.)
- Papadopoulos AS, Cleare AJ. Hypothalamic-pituitary-adrenal axis dysfunction in chronic fatigue syndrome. Nature Reviews Endocrinology. 2012. (Source for mild hypocortisolism, the moderating factors, and "steroid replacement is not recommended".)
- Akturk HK, et al. Over-the-Counter "Adrenal Support" Supplements Contain Thyroid and Steroid-Based Adrenal Hormones. Mayo Clinic Proceedings. 2018. (Source for the 12 supplements, the triiodothyronine in all of them, and the budesonide in 25%.)
- Bornstein SR, et al. Diagnosis and Treatment of Primary Adrenal Insufficiency: An Endocrine Society Clinical Practice Guideline. JCEM. 2016. (Source for the ACTH stimulation test, the 500 nmol/L, 18 µg/dL peak cortisol threshold, the instruction to treat suspected adrenal crisis immediately with parenteral hydrocortisone, and the recommendations for an emergency injection kit, a steroid emergency card and medical alert identification.)
- StatPearls (NCBI Bookshelf). Adrenal Crisis. (Source for the crisis presentation, the triggers, "the treatment of adrenal crises should never be delayed to obtain blood work", the 10% of deaths figure, and the sick-day rule of doubling the oral dose above 38 °C and tripling it above 39 °C.)
- Addison's Disease Self Help Group. Sick day rules. (Source for the patient guidance that vomiting is the danger zone, and that if you are unsure whether you are in a crisis you should treat it as one.)
- NIDDK, National Institutes of Health. Adrenal Insufficiency & Addison's Disease: Symptoms & Causes. (Source for the symptom list and for skin darkening in primary adrenal insufficiency.)
- 9-8-8 Suicide Crisis Helpline, Canada. Call or text 9-8-8, 24 hours a day, every day of the year.