Dexamethasone suppression test: how it works and what the result means

A dexamethasone suppression test checks whether your body's cortisol production switches off when it should. You take a tablet of dexamethasone, a synthetic steroid, late in the evening, and have a blood test early the next morning. In most people the tablet makes the morning cortisol fall to a very low level. If it does not, the body may be making cortisol it cannot turn off, which is the hallmark of Cushing's syndrome. The overnight 1 mg version is one of the first-line tests for Cushing's syndrome,¹² and the standard test for a milder form of cortisol excess found in some people with a lump on an adrenal gland.³ An abnormal result is a reason for further tests, not a diagnosis.

How the test works

Cortisol is controlled by a feedback loop. The pituitary gland at the base of the brain makes ACTH, a hormone that tells the adrenal glands how much cortisol to make.⁴ When steroid levels rise, they act back on the brain to switch the system down; this negative feedback is how the body turns off a stress response.⁵ Dexamethasone is a synthetic steroid that triggers the same feedback, so in a healthy person ACTH falls and, by the next morning, so does cortisol.

In Cushing's syndrome, cortisol is being driven by something that ignores this signal: a tumour of an adrenal gland making cortisol on its own, a pituitary tumour making ACTH (Cushing's disease), or a tumour elsewhere in the body making ACTH.⁶ The morning cortisol stays high despite the dexamethasone. The most common cause of Cushing's syndrome overall, though, is taking high-dose steroid medicine for a long time, which is why your doctor will want to know about any steroids you use.⁷

The overnight 1 mg test, step by step

1. You take 1 mg of dexamethasone by mouth between 11pm and midnight.¹
2. Blood is taken the next morning between 8am and 9am to measure cortisol.¹
3. Some laboratories also measure dexamethasone in the same sample, to check that enough was absorbed, although that test is not available everywhere.¹

Higher doses, of 1.5 or 2 mg, do not make the test significantly more accurate.¹ Take the tablet at the time you are given, and tell the clinic if you forget it, take it late or are sick soon afterwards, because the result depends on the tablet having been taken and absorbed. Before the test, your doctor may ask you to stop medicines that can affect it, such as some antibiotics, anti-seizure medicines, steroids, oestrogen, the contraceptive pill and water tablets.⁶ Do not stop anything without their advice.

Other versions

A longer low-dose test gives 0.5 mg every 6 hours for 48 hours, 2 mg a day in total, with blood cortisol measured afterwards against the same 50 nmol/L cut-off.¹ The Endocrine Society says it may be more specific than the overnight test for people with depression or other psychiatric conditions, severe obesity, alcohol dependence or diabetes.¹ High-dose versions, using 2 mg every 6 hours, are used by specialists to help find the cause once cortisol excess is established.⁶

What the result means

A normal result is a morning cortisol below 50 nmol/L (1.8 µg/dL).⁶¹ One NHS laboratory states that a 9am cortisol below 50 nmol/L after overnight dexamethasone excludes endogenous Cushing's syndrome, meaning cortisol excess made by the body itself.⁸

The cut-off is set low on purpose. At 50 nmol/L the test picks up more than 95 per cent of people with Cushing's syndrome, but its specificity is about 80 per cent, so some people without the condition fail to suppress.¹ Raising the cut-off to 140 nmol/L (5 µg/dL) pushes specificity above 95 per cent, but in studies it misclassified up to 15 per cent of people with Cushing's disease as normal.¹ The 2021 Pituitary Society consensus puts the sensitivity of all the standard first-line tests above 90 per cent, with lower specificity.² In plain terms, a normal result is reassuring, while an abnormal one needs checking.

A result above 50 nmol/L means cortisol did not suppress fully. Your doctor will usually check it with other first-line tests, such as 24-hour urine or late-night saliva collected at least twice, and look for other explanations.¹² If Cushing's syndrome is confirmed, specialists use further tests and scans to find the cause.⁷ See the urine cortisol test and saliva cortisol test guides.

Adrenal incidentalomas and mild autonomous cortisol secretion

An adrenal incidentaloma is a lump on an adrenal gland found by chance on a scan done for another reason. Some make a little cortisol of their own. European guidelines use the overnight 1 mg test to check for this, with the same cut-off of 50 nmol/L (1.8 µg/dL).³ Someone with a result above 50 nmol/L who does not have the outward features of Cushing's syndrome is classed as having mild autonomous cortisol secretion (MACS).³ Your specialist will explain what the result means for you and whether further tests are needed.

Hair cortisol has been studied in this setting and does not help. In 67 people with adrenal incidentalomas, hair cortisol and cortisone did not differ from healthy controls, or between people with and without cortisol autonomy, and did not correlate with blood, saliva or urine cortisol.⁹ The authors concluded hair is not suitable for detecting MACS.⁹

What can give a false result

The test depends on dexamethasone reaching the right level in your blood and on cortisol being measured accurately. Several things can upset it.

• Oral oestrogen. Oestrogen in the contraceptive pill and in oral HRT raises corticosteroid-binding globulin (CBG), the protein that carries cortisol, and blood tests measure total cortisol including the bound part. In one study, women on oral oestrogen had 67 per cent higher total cortisol than women who were not.¹⁰ In women taking the pill, the false-positive rate of the overnight test is 50 per cent, and the Endocrine Society advises stopping oestrogen-containing medicines for 6 weeks before testing where possible.¹ The 2021 consensus warns that the test may be unreliable in women taking oral oestrogen.² Pregnancy raises CBG too.¹¹ Only stop contraception or HRT on medical advice.
• Medicines that speed up dexamethasone breakdown. Phenobarbital, phenytoin, carbamazepine, primidone, rifampicin, rifapentine, ethosuximide and pioglitazone boost the enzyme (CYP3A4) that clears dexamethasone. That lowers its level and makes a false positive more likely.¹
• Medicines that slow it down. Itraconazole, ritonavir, fluoxetine, diltiazem, cimetidine and aprepitant block the same enzyme, which can raise dexamethasone levels and push the result the other way.¹
• Absorption. People absorb and break down dexamethasone at different rates, even healthy people on no medicines, and this can affect the result.¹
• Other conditions. Depression, anxiety disorders, obsessive-compulsive disorder, poorly controlled diabetes and alcohol dependence can cause mild cortisol excess and abnormal suppression without Cushing's syndrome.¹ Obesity and stress can also cause false results, which are more common in women.⁶
• Cyclic Cushing's. If cortisol excess comes and goes, the Endocrine Society suggests urine or late-night saliva tests rather than dexamethasone tests,¹ and the 2021 consensus recommends repeated late-night saliva tests over time.²

What happens next

If Cushing's syndrome is suspected, a GP can arrange initial cortisol tests, and specialists then carry out further tests and scans.⁷ Signs that should prompt a check include more fat on the neck, upper back, chest and tummy with thinner arms and legs, a red, round face, easy bruising and stretch marks, and muscle weakness at the top of the arms and legs.⁷ Our guides to high cortisol symptoms and testing for Cushing's syndrome explain the full route, and the blood cortisol test guide covers how blood cortisol is measured and reported.

There is no home version of this test. A home blood, saliva or hair kit cannot show whether your cortisol suppresses, and none of them replaces it. Hair cortisol, which reflects average cortisol over months, is not part of the diagnostic guidelines for Cushing's syndrome.¹²