Read this before anything else
1. If you take a steroid, or recently stopped one, do not change it because of this page. You are about to read that corticosteroid medication is the most common cause of Cushing's syndrome, and the honest risk of writing that sentence is that someone stops their prescription. Do not. MedlinePlus states it plainly: suddenly stopping corticosteroids after taking them for a long time, usually more than 2 weeks, can result in a life-threatening condition called adrenal crisis. The dose belongs to the doctor who prescribed it. If you recently finished or tapered a course, this still applies to you: your own adrenal glands may take months to restart.
Know what an adrenal crisis looks like, because it is mistaken for a stomach bug. Severe weakness, drowsiness or confusion; severe vomiting and diarrhea; severe abdominal, back or leg pain; dizziness or collapse; low blood sugar; and often a fever. It is set off by an ordinary vomiting or flu-like illness, any infection, injury or surgery. And if you are vomiting, you cannot absorb an oral steroid tablet, so being unable to keep medication down is itself an emergency.
If you suspect an adrenal crisis, call emergency services now. Say these words: "I think I may have adrenal insufficiency. I may need hydrocortisone." Adrenal crisis is missed in emergency departments, and saying it directly is the fastest route to treatment. Treatment must not wait for test results. Separately, ask the doctor who prescribes your steroid for sick-day rules (doses need to go up during illness, not stay the same) and whether you should carry an emergency hydrocortisone injection kit. The full picture is in what coming off prednisone does to your own cortisol production and adrenal insufficiency.
2. These need care today, not at your next appointment. This list is not a checklist to pass. Matching none of it does not mean you are fine, it means none of these specific emergencies is running:
- Sudden breathlessness, chest pain, or a swollen, painful calf. Cortisol excess makes blood clot more readily. Cushing's syndrome carries roughly 10 times the general-population risk of venous thromboembolism in the first year after diagnosis. Call emergency services.
- Rapidly worsening muscle weakness, especially alongside blood pressure that will not come down. Ask for a potassium level. It is on a routine blood panel, it is not a cortisol test, and severe low potassium is dangerous in its own right. See the box below on why this pattern matters.
- A sudden severe headache, any change in your vision, or double vision. Most endogenous Cushing's syndrome starts in a pituitary tumour. A growing one presses on the nerves to the eyes and can take your peripheral vision permanently; a sudden bleed into one (pituitary apoplexy) is a surgical emergency that also shuts down cortisol production. Sudden headache with vision change means emergency care, not an eye appointment.
- Feeling suddenly and seriously unwell, with or without a fever. The Endocrine Society notes that severe cortisol excess "predisposes to severe, systemic infection and/or sepsis". Cortisol also blunts the inflammatory response, so the absence of a fever does not rule out a serious infection. Never use a normal temperature to talk yourself out of getting seen.
- Thoughts of harming yourself. Depression is reported in 55 to 81% of people with Cushing's syndrome, and psychosis in about 8%. This is not a data problem and it is not something to sit on while tests are arranged. In Canada and the United States, call or text 988. Elsewhere, call your local emergency number.
The dangerous thing that hides in this symptom set
There is a form of cortisol excess that does not look like Cushing's syndrome, and it is the one that moves fastest. In ectopic ACTH syndrome, a tumour somewhere other than the pituitary makes the hormone that drives the adrenal glands. It accounted for 17% of the 212 cyclic cases in a 2023 systematic review. When the tumour is a small cell lung cancer, the classic picture is largely absent.
In a published series of 7 such patients, all 7 had low potassium (mean 2.12 mmol/L), all 7 had high blood pressure, 4 of 7 needed 3 to 5 drugs to control it, all 7 had profound muscle weakness, all 7 lost weight (median 5 kg), and only 1 of the 7 had stretch marks. 5 of 7 developed bacterial infections and 2 developed sepsis within a month. Mean survival was 2.3 months.
Weight loss, severe weakness, stubborn hypertension and low potassium are the pattern to act on. The absence of the round face and the purple stretch marks does not make this less likely. In that series it was the norm. If that is your picture, this is not a testing puzzle to work through over months. Get assessed now, and ask for electrolytes and a chest image, not another cortisol result.
Speed and virilization are the second pattern to act on. Among adrenal causes sits adrenocortical carcinoma, a rare and aggressive cancer. It tends to announce itself by moving fast and by producing male hormones as well as cortisol: Cushingoid change over weeks to a few months rather than years, with new coarse facial or body hair, a deepening voice, male-pattern balding, or flank or abdominal pain. In a woman that combination is frequently written off as polycystic ovary syndrome. Rapid onset plus virilization is not PCOS until someone has looked. Ask for adrenal imaging and androgen levels, and ask soon.
What a normal cortisol result does not rule out
Every red flag above is useful as a positive. None of them is safe as a negative, and neither is a normal test.
- It does not rule out Cushing's syndrome. That is the entire subject of this page.
- It does not rule out anything else, because it did not look. A cortisol measurement says nothing about your thyroid, your potassium, your blood count, your heart, your lungs or your sleep. A normal number is one hormone at one moment, not a clean bill of health.
- A normal temperature does not rule out infection when cortisol is high, because high cortisol suppresses the response that produces the fever.
- Normal potassium does not rule out a lung tumour. It removes one clue, not the possibility.
- Two normal tests do not close the question if you are getting worse. The Endocrine Society's own guideline says to keep testing when cyclic disease is suspected, and to re-evaluate at 6 months if signs or symptoms progress. Deterioration outranks a filed result.
If you are calling for help, say which steroids you take, by which route, and for how long, and say so even if the steroid is an inhaler or a cream. If you have recently stopped or reduced one, say that too, and say you may have adrenal insufficiency.
Three groups of people quietly decide this page is not about them
The first: people whose steroid does not come in a tablet. Cushing's syndrome is cortisol excess from any source, and iatrogenic Cushing's syndrome has been reported from oral, injected, intra-articular, inhaled, intranasal, ocular and topical steroids. Risk climbs steeply when a drug that inhibits CYP3A4 is taken alongside an inhaled or nasal steroid, ritonavir and itraconazole being the usual reported culprits, because the steroid then reaches the bloodstream in far higher amounts. In a meta-analysis of 74 studies, measured adrenal insufficiency ran to 52.2% in people receiving intra-articular joint injections (that figure is adrenal suppression, the flip side of the same coin: the steroid still reached your bloodstream). Its conclusion is worth quoting: "there is no administration form, dosing, treatment duration, or underlying disease for which adrenal insufficiency can be excluded with certainty."
The second: people who do not look like the photographs. The images attached to this condition are of a round, flushed face and wide purple stretch marks. Ectopic ACTH from a fast-growing tumour frequently produces none of that. Mild disease often produces none of it. And because Cushing's syndrome is reported roughly 3 times more often in women than in men, men read the material and put it down. Absence of the classic appearance is not evidence of absence.
The third: parents. In a child the single most discriminating sign is not a round face, it is putting on weight while growth slows, crossing down through height percentiles while crossing up through weight. The Endocrine Society names it specifically and says to test for it. Children are also the group most exposed to potent topical steroids and inhalers, and the group in whom adrenal suppression is most easily missed. If your child is on any steroid and their growth has flattened, that is a conversation with their doctor, not a wait-and-see.
Can you have Cushing's syndrome with a normal cortisol test?
Yes, and the guideline that endocrinologists work from says so directly. The 2008 Endocrine Society clinical practice guideline on the diagnosis of Cushing's syndrome recommends "additional testing in patients with discordant results, normal responses suspected of cyclic hypercortisolism, or initially normal responses who accumulate additional features over time." A normal result is a data point in a process, not the end of one.
Three separate mechanisms produce a normal number in a person who has the syndrome.
- The test was not one of the tests. The guideline recommends against using a random serum cortisol or a plasma ACTH level to screen for Cushing's syndrome, along with urinary 17-ketosteroids, the insulin tolerance test and the loperamide test. A one-off morning blood cortisol, which is what many people are actually given, is not a screening test for this condition at all.
- The sample landed in a trough. On 24-hour urinary free cortisol, the guideline states the result "can be normal if a patient has cyclic disease" and "may be normal in some patients with mild Cushing's syndrome." Cortisol is not a fixed quantity. It is a curve, and where you sample it decides what you see. That is the whole of cyclic Cushing's syndrome, below.
- The cortisol excess is not your cortisol. In steroid-induced Cushing's syndrome the medication does the work and your own adrenal glands go quiet, so a cortisol measurement can read low or normal while the syndrome is florid. This is why the guideline recommends "obtaining a thorough drug history to exclude excessive exogenous glucocorticoid exposure leading to iatrogenic Cushing's syndrome before conducting biochemical testing." The drug history comes first, before any test, and it is the step most often skipped.
That third mechanism is the most common one, and it is the least discussed. Both the NIH's NIDDK and StatPearls state that the most common cause of Cushing's syndrome is long-term glucocorticoid medication, ahead of every tumour combined. Of the endogenous cases that remain, StatPearls reports that a pituitary source accounts for up to 80%.
What is cyclic Cushing's syndrome?
Cortisol production that comes in episodes. Peaks of real cortisol excess, separated by troughs in which secretion is normal or even low. Meinardi and colleagues, reviewing 65 reported cases in the European Journal of Endocrinology, set the formal bar: three peaks and two troughs of cortisol production must be demonstrated. In their series the phases between cycles ranged from days to years, and the origin was a pituitary adenoma in 54% of cases, an ectopic ACTH-producing tumour in 26%, and an adrenal tumour in about 11%.
Their assessment of the diagnostic problem is blunt: "Cortisol stimulation or suppression tests may give spurious results owing to spontaneous falls or rises in serum cortisol at the time of testing."
The largest systematic review of the condition, published in The Lancet Diabetes & Endocrinology in 2023, pooled 212 cases from 118 articles. Pituitary tumours accounted for 67%, ectopic tumours 17%, adrenal tumours 11%. Two of its findings matter to anyone holding a normal result:
Patients with cyclic Cushing's syndrome "might be turned away from physicians when presenting during a trough phase (and hence with physiological cortisol concentrations)." Overall, 6% of the patients (12 of 212) had unnecessary surgery due to misclassification.
Nowak E, et al. Lancet Diabetes & Endocrinology, 2023.
Both of those are errors of timing, and they cut in both directions. The trough sends a sick person home. The peak, in the wrong patient, sends a well person to theatre. This is why the guideline's answer to cyclic disease is not a single better test. It is more tests, in the right form, spread across time, ordered by someone who is following you. To be explicit, since we sell a cortisol sensor: "spread across time" means repeated laboratory assays with clinical thresholds, not continuous consumer readings. A wellness trace cannot substitute for one of these tests, however many points are on it. The Endocrine Society specifically suggests using urinary free cortisol or late-night salivary cortisol rather than a dexamethasone suppression test when cyclic Cushing's syndrome is suspected, precisely because those two can be repeated and can be timed.
Which tests are used, and what can each of them miss?
Three first-line screening tests, per the Endocrine Society. Note how many of them are specified as repeated measurements. That is not administrative caution. It is the guideline building the timing problem into its own method.
| Test | What it does | How the guideline specifies it | What it can miss |
|---|---|---|---|
| Late-night salivary cortisol | Samples cortisol at its natural low point, late at night | Two measurements. Reported sensitivity 92 to 100% and specificity 93 to 100% | A night that falls in a trough of cyclic disease |
| 24-hour urinary free cortisol | Totals the cortisol excreted across a full day | At least two collections. Compared against the upper limit of normal for the specific assay | The guideline states it "may be normal in some patients with mild Cushing's syndrome" and "can be normal if a patient has cyclic disease" |
| Dexamethasone suppression test | Gives a synthetic steroid and checks whether your cortisol switches off in response | 1 mg overnight, or 2 mg/day for 48 hours. A cortisol below 1.8 µg/dL (50 nmol/L) is a normal suppression | Specificity at that threshold is around 80%. The guideline states it is not recommended where cyclic disease is suspected |
| Random serum cortisol or plasma ACTH | A single blood draw at whatever hour you happened to attend | Recommended against. Not a screening test for Cushing's syndrome | Everything. If this is the "normal cortisol test" you were given, you have not yet been screened |
The path after screening is fixed. The Endocrine Society recommends that anyone with an abnormal result see an endocrinologist and undergo a second test, and that people with two abnormal results go on to testing for the cause. It also acknowledges the limit of its own toolkit in one line: "no test has optimally high specificity, so that false-positive results may occur." A test can be wrong in either direction, which is exactly why no single one of them is allowed to decide.
Why is Cushing's syndrome so hard to diagnose?
Because almost every symptom of it is also a symptom of something ordinary. A meta-analysis of 5,367 patients across 44 studies found a mean time to diagnosis of 34 months. It varied by cause: 14 months for ectopic disease, 30 months for adrenal, 38 months for pituitary. Nearly 3 years, on average, and the figure had not improved between the cohorts diagnosed before and after 2000.
That delay is the reason a normal result is worth questioning rather than filing. The NIH's NIDDK puts the stakes in one sentence: "Although Cushing's syndrome can usually be cured, it can be fatal if not treated." It lists the untreated complications as heart attack, stroke, blood clots, infections, bone loss and fractures, high blood pressure, depression, memory problems and type 2 diabetes.
The guideline splits the features in two, and the split is the most useful table in the whole document.
| Features that best discriminate Cushing's syndrome | Features common in Cushing's syndrome and in the general population |
|---|---|
| Easy bruising Facial plethora (a persistent red flush) Proximal myopathy (weakness in the thighs, shoulders and hips) Striae, especially reddish-purple and wider than 1 cm In children: weight gain with slowing growth |
Depression Fatigue Weight gain and obesity Back pain Hypertension Vertebral osteoporosis Thin skin Type 2 diabetes Menstrual changes Hirsutism or balding Acne Insomnia and irritability |
Read the left-hand column as a reason to be taken seriously, never as a checklist you must pass. These features are more predictive. They are not required, they were mostly absent in the lung cancer series described above, and the guideline itself directs testing at people with "unusual features for age" and "multiple and progressive features", and at "children with decreasing height percentile and increasing weight", which is a description of a trajectory, not a look.
The right-hand column is where the misdiagnoses live, and it is also where the honest counterweight to this page sits. Most people who search this question do not have Cushing's syndrome. The same cluster of fatigue, weight gain, low mood, high blood pressure and disturbed sleep is produced by conditions that are far more common and that a cortisol test does not investigate at all: obstructive sleep apnea, thyroid disease, polycystic ovary syndrome, poorly controlled diabetes, alcohol use, and depression itself. The Endocrine Society lists several of these as causes of genuinely raised cortisol without Cushing's syndrome, alongside pregnancy and severe obesity. If your cortisol tests are normal and you still feel unwell, that is not the end of the investigation. It is a reason to widen it.
Your test was normal and you still have symptoms. What now?
Go back, and go back with specifics. The point of everything above is not that you should doubt your doctor. It is that the guideline your doctor works from already anticipates this situation and tells them what to do about it. Bring it up by name.
- Ask which test you actually had. If the answer is a single morning blood cortisol, the guideline recommends against using that to screen for Cushing's syndrome. You have not been screened yet.
- Ask about repeating it. Two late-night salivary cortisol measurements, or at least two 24-hour urine collections, is what the guideline specifies. One of anything is not the protocol.
- Say the word "cyclic" out loud. The guideline recommends additional testing for normal responses in someone suspected of cyclic hypercortisolism, and suggests urinary free cortisol or late-night salivary cortisol rather than a dexamethasone suppression test in that situation.
- Give a complete drug history, including the steroids you do not think of as steroids. Inhalers, nasal sprays, creams, eye drops, joint injections. Add any antifungal or HIV medication, which can multiply the amount of an inhaled steroid reaching your blood.
- Ask what else is being considered. Sleep apnea, thyroid function, potassium, blood glucose. A normal cortisol result is a reason to look elsewhere, not a reason to stop looking.
- Ask for a follow-up interval, and keep it. The guideline suggests re-evaluation at 6 months if signs or symptoms progress. Write down what is progressing, and when.
An endocrinologist is the specialist for this. If your features are unusual for your age, or multiple and getting worse, that is precisely the group the guideline says to test. Ask for the referral.
And on our own product, plainly: no wearable and no at-home kit can screen for, detect, diagnose, monitor or rule out Cushing's syndrome, ours included. Nothing you see on a wellness device is a reason to delay or skip getting seen, and a reassuring-looking trace is not a negative test. We make a cortisol sensor, so we should be the ones to say it. The tests above are laboratory assays with defined clinical thresholds, ordered by a physician, interpreted against your history, and followed by further testing to find a cause. A general wellness reading is not on that scale and cannot be compared to a clinical result. There is no shortcut around the doctor here, and anyone selling you one is selling you something. If you want to understand what cortisol is and how it moves through an ordinary day, that is Cortisol 101.
This page is general medical information, not medical advice, and not a diagnosis. It summarizes published clinical guidance from the sources below. Cushing's syndrome is diagnosed and treated by qualified healthcare professionals, never by a wearable or an at-home test. Never start, stop or change a steroid medication without speaking to the doctor who prescribed it. If you suspect an adrenal crisis, a blood clot, or a serious infection, seek emergency care immediately.
References
- Nieman LK, et al. The Diagnosis of Cushing's Syndrome: An Endocrine Society Clinical Practice Guideline. JCEM, 2008. (Source for the first-line tests, the repeat-measurement specifications, the 1.8 µg/dL threshold, the recommendation against random serum cortisol and plasma ACTH, the drug-history-first recommendation, the cyclic-hypercortisolism recommendation, the discriminatory-features table, and the 6-month re-evaluation.)
- Nieman LK, et al. Treatment of Cushing's Syndrome: An Endocrine Society Clinical Practice Guideline. JCEM, 2015. (Source for severe hypercortisolism predisposing to systemic infection and sepsis, and for the mortality and thrombosis context.)
- Meinardi JR, Wolffenbuttel BHR, Dullaart RPF. Cyclic Cushing's syndrome: a clinical challenge. European Journal of Endocrinology, 2007. (Source for the three-peaks-and-two-troughs definition, the 65 cases, the 54/26/11% origins, and the days-to-years intercyclic phases.)
- Nowak E, et al. Diagnostic challenges in cyclic Cushing's syndrome: a systematic review. Lancet Diabetes & Endocrinology, 2023. (Source for the 212 cases from 118 articles, the 67/17/11% origins, and the 6% who had unnecessary surgery due to misclassification.)
- Rubinstein G, et al. Time to Diagnosis in Cushing's Syndrome: A Meta-Analysis Based on 5367 Patients. JCEM, 2020. (Source for the 34-month mean, and the 14 / 30 / 38-month split by cause.)
- NIH / NIDDK. Cushing's Syndrome. (Source for glucocorticoid medicine being the most common cause, for "no one test is perfect, so doctors usually do two", and for the roughly 3-to-1 female-to-male ratio.)
- Chaudhry HS, Singh G. Cushing Syndrome. StatPearls (NCBI Bookshelf). (Source for iatrogenic corticosteroid use being the most common cause, and for pituitary ACTH accounting for up to 80% of endogenous cases.)
- MedlinePlus (NLM). Exogenous Cushing syndrome. (Source for the warning that suddenly stopping corticosteroids after long-term use can cause a life-threatening adrenal crisis.)
- Gamrat-Żmuda A, et al. Ectopic Cushing's Syndrome in Advanced Small-Cell Lung Cancer: Clinical Challenges and Therapeutic Insights. Cancers, 2025. (Source for the 7-patient series: potassium, hypertension, weakness, weight loss, striae in 1 of 7, sepsis, and the 2.3-month mean survival.)
- Koraćević G, et al. Cushing's Syndrome, a Risk Factor for Venous Thromboembolism, Is a Candidate for Guidelines. Acta Endocrinologica (Bucharest), 2020. (Source for roughly 10 times the risk of venous thromboembolism in the first year after diagnosis.)
- Lin TY, Hanna J, IsHak WW. Psychiatric Symptoms in Cushing's Syndrome: A Systematic Review. Innovations in Clinical Neuroscience, 2020. (Source for depression at 55 to 81% and psychosis at about 8%.)
- Broersen LHA, et al. Adrenal insufficiency in corticosteroids use: systematic review and meta-analysis. JCEM, 2015. (Source for the 52.2% intra-articular figure and for the conclusion that adrenal insufficiency cannot be excluded with certainty for any route, dose or duration.)
- Raveendran AV. Inhalational Steroids and Iatrogenic Cushing's Syndrome. The Open Respiratory Medicine Journal, 2014. (Source for inhaled and intranasal steroids causing iatrogenic Cushing's syndrome, and for the CYP450-inhibitor interaction.)