Cortisol and Adrenal Function Tests

Cortisol is a glucocorticoid made in the adrenal zona fasciculata under the control of ACTH from the pituitary, which is itself controlled by hypothalamic CRH. Secretion follows a circadian rhythm, with a peak at about 8 am and a trough near midnight, and rises with stress. About 90% is bound to cortisol binding globulin and albumin, and the free fraction is active. It is measured by immunoassay or liquid chromatography with mass spectrometry in serum, saliva or urine.

Cortisol tests diagnose adrenal insufficiency (too little cortisol) and Cushing syndrome (too much). A single random cortisol is rarely diagnostic, and dynamic tests or timed samples are used. Interpretation depends on whether the axis is under suppression or stimulation.

Adrenal insufficiency affects about 100 to 140 per million (primary) and 150 to 280 per million (secondary), and the commonest cause overall is glucocorticoid withdrawal. Cushing syndrome has an incidence of about 2 to 3 per million each year, and the commonest cause by far is exogenous glucocorticoid use.

Indications.

• Insufficiency: fatigue, weight loss, hypotension, hyponatraemia, hyperkalaemia, hypoglycaemia, pigmentation, unexplained shock, and long term glucocorticoid use.
• Excess: central obesity, proximal myopathy, striae, easy bruising, resistant hypertension, diabetes, osteoporosis, and an incidental adrenal mass.

Specimen. Serum or plasma drawn at 8 to 9 am for basal cortisol. Late night salivary cortisol is collected by passive drool or cotton swab at about 11 pm to midnight. Twenty four hour urine for free cortisol uses a complete collection checked by creatinine. Oestrogens (oral contraceptives, pregnancy) raise total cortisol by raising binding globulin.

Reference ranges are method specific. Conversion: nmol/L divided by 27.6 gives microg/dL.

Basal morning cortisol (8 to 9 am).

• Over 400 nmol/L (over 15 microg/dL): adrenal insufficiency effectively excluded.
• 150 to 400 nmol/L: indeterminate, so a stimulation test is needed.
• Under 100 nmol/L (under 3 microg/dL) in an unwell patient: adrenal insufficiency very likely.
• Under 50 nmol/L: diagnostic when the clinical picture is compatible.

Short ACTH (tetracosactide) stimulation test. Cortisol is measured at 0 and 30 (and 60) minutes after 250 microg intravenous or intramuscular tetracosactide. A peak over 450 to 550 nmol/L (assay specific, about 18 to 20 microg/dL) excludes primary and established secondary adrenal insufficiency. A low peak confirms insufficiency, and ACTH distinguishes primary (high) from secondary (low or inappropriately normal). In recent onset secondary insufficiency the test can be normal.

Plasma ACTH. Collected in an EDTA tube on ice, centrifuged cold, and analysed promptly. Raised in primary adrenal insufficiency (over 2 times the upper limit) and ectopic ACTH syndrome, and low in secondary insufficiency and adrenal Cushing syndrome.

Overnight dexamethasone suppression test. Dexamethasone 1 mg orally at 11 pm to midnight, and serum cortisol at 8 to 9 am. Cortisol under 50 nmol/L (under 1.8 microg/dL) excludes Cushing syndrome. A higher value needs further testing, and false positives occur with oestrogens, enzyme inducers (carbamazepine, phenytoin, rifampicin), depression, alcohol, obesity and acute illness. A paired dexamethasone level confirms adequate absorption.

Late night salivary cortisol. Raised in Cushing syndrome because circadian nadir is lost. Two samples on separate nights, with a raised result in both (over the assay upper limit, for example 4 to 5 nmol/L), have a sensitivity and specificity of 90 to 95%. Unreliable in shift workers.

24-hour urinary free cortisol. Raised to over 3 times the upper limit in Cushing syndrome, with 2 to 3 collections. Unreliable with severe renal impairment (eGFR under 30 mL/min), excessive fluid intake and incomplete collections.

Further tests for cause of Cushing syndrome. ACTH (suppressed in adrenal causes, normal or raised in pituitary and ectopic), high dose dexamethasone suppression (8 mg), CRH stimulation, inferior petrosal sinus sampling with CRH, and adrenal and pituitary imaging.

Primary adrenal insufficiency. Low cortisol, high ACTH, high renin, low aldosterone, hyponatraemia, hyperkalaemia, hypoglycaemia, and pigmentation. Causes: autoimmune (21-hydroxylase antibodies), tuberculosis, adrenal haemorrhage, metastases, infections (HIV, cytomegalovirus), adrenoleukodystrophy, and congenital adrenal hyperplasia.

Secondary and tertiary insufficiency. Low cortisol, low or normal ACTH, normal aldosterone, and no pigmentation. Causes: glucocorticoid therapy (including inhaled, topical and intra-articular), pituitary tumour, surgery, radiotherapy, apoplexy, and hypophysitis (including checkpoint inhibitor therapy).

Cushing syndrome. Failed suppression on dexamethasone testing, raised late night salivary cortisol or urinary free cortisol, and loss of circadian rhythm. Causes: pituitary adenoma (70%), ectopic ACTH (10 to 15%, often lung neuroendocrine tumour and a very high cortisol with hypokalaemia), adrenal adenoma or carcinoma (15 to 20%), and exogenous glucocorticoids. Pseudo-Cushing states (alcohol use, depression, obesity, diabetes) give false positives.

Critical illness. A random cortisol under 276 nmol/L in septic shock or a rise under 250 nmol/L after tetracosactide indicates critical illness related corticosteroid insufficiency, though diagnosis and treatment remain debated. Total cortisol underestimates free cortisol in hypoalbuminaemia.

Glucocorticoid treated patients. Prednisolone cross-reacts with cortisol assays (hydrocortisone is measured, and dexamethasone is not). Testing for adrenal suppression is done at least 24 hours after the last hydrocortisone dose (and later for longer acting steroids).

Critical values and findings needing immediate action.

• Suspected adrenal crisis (hypotension, vomiting, abdominal pain, hyponatraemia, hyperkalaemia, hypoglycaemia): treatment with intravenous hydrocortisone 100 mg and saline is given at once, and a cortisol and ACTH sample is taken beforehand if this does not delay treatment.
• Cortisol under 100 nmol/L in a patient who is acutely unwell.
• Very high cortisol with hypokalaemia, sepsis, psychosis, severe hypertension or proximal weakness (severe Cushing or ectopic ACTH).
• Patient on long term glucocorticoids who is vomiting or acutely ill, needing stress dose cover.

Interferences and pitfalls.

• Oestrogen therapy, pregnancy and oral contraceptives raise cortisol binding globulin and total cortisol and cause false positive suppression tests. Oestrogens are stopped for 6 weeks.
• Low binding proteins (nephrotic syndrome, cirrhosis, critical illness) lower total cortisol.
• Biotin interferes with some immunoassays.
• Cross-reactivity with prednisolone, 6-methylprednisolone and fludrocortisone alters immunoassay results, and hydrocortisone doses are held before sampling.
• Spironolactone and carbenoxolone can interfere with some assays, and liquorice and exogenous steroids alter test results.
• Collection time matters: afternoon cortisol is lower than morning.
• Contamination of saliva with hydrocortisone creams, blood or food gives false results.
• Dynamic tests are done at least 24 to 48 hours after any change in steroid treatment.

Special groups.

• Neonates: cortisol levels are higher at birth, and transient adrenal insufficiency occurs in sick preterm infants.
• Children: congenital adrenal hyperplasia needs 17-hydroxyprogesterone, and age specific ranges apply.
• Pregnancy: total cortisol rises 2 to 3 times, and Cushing syndrome is rare.
• Older adults: cortisol is unchanged, but atypical presentations are frequent.
• Regions: tuberculosis and HIV remain important causes of adrenal insufficiency.

Follow up tests. ACTH, renin and aldosterone, 21-hydroxylase antibodies, sodium, potassium, glucose, 17-hydroxyprogesterone, DHEA sulphate, androgen profile, urinary steroid profile, adrenal CT or MRI, pituitary MRI, other pituitary hormones, bone density, and screening for diabetes and hypertension in Cushing syndrome.