High Cortisol Levels: Causes and When to Worry

Your cortisol came back above the upper limit printed on your report. Before you go looking for a disease, one question outranks every other: what time was the blood drawn? Cortisol swings enormously across the day, and a number without its clock time cannot be read at all. This page goes through the causes in the order they actually occur, why estrogen inflates the result in many women, why steroid pills can push it up rather than down, and when the result genuinely warrants a visit.

One more thing, since it may be what brought you here: "high cortisol" is also a wellness-industry topic — cortisol face, cortisol detoxes, adrenal-fatigue supplements. This page treats that framing on the same evidence as the lab number.

First: is this really abnormal?

⚠️ Without the time of the draw, the number means nothing

This is the single most important line on the page. Cortisol is released in pulses, and the size of those pulses follows a strong circadian rhythm: highest in the last hours of sleep and the early morning, falling through the day, lowest around midnight.¹ The steepest climb comes in the 30 to 45 minutes after you wake — so consistent it has its own name, the cortisol awakening response.²

The practical consequence: the reference interval on your report almost always corresponds to a morning draw, roughly 7 to 9 a.m., at the top of the curve — typically somewhere around 6–23 µg/dL, depending on the assay.³⁴ A "high" value at 8 a.m. may be nothing more than the normal peak. And if your blood was drawn later in the day, comparing it to that morning range is meaningless — an afternoon value should be closer to 3–13 µg/dL, so a result flagged "normal" at 4 p.m. can be more informative than one flagged "high" at 8 a.m. How the hormone works, and what a low result means, is covered on the parent page: cortisol blood test.

That rhythm is anchored to your wake time, not the wall clock. Night shifts, or getting up at 4 a.m. to make a 7 a.m. draw, shift the whole curve with you.

The cutoff changes from one lab to the next

Cortisol is measured on different analyzers, and each laboratory publishes its own interval. The same serum can read "slightly high" at one lab and "normal" at another. Compare yourself to the range on your report, never to a number found online.³⁵

What happened in the 30 minutes before the draw — not the needle itself

You will read that the needle itself raises cortisol. That does not hold up mechanically: stress acts through a burst of pituitary ACTH, and only then does the adrenal gland respond.¹ The stick at time zero cannot change blood already entering the tube. What can have moved the number is the preceding half hour — an anxious wait, running to make the appointment, pain, a sleepless night, an acute illness. Real, but modest: it explains a mild overshoot, not a tripled result.

⚠️ Birth control and estrogen: the most common false alarm in women

Most circulating cortisol travels bound to a carrier protein, corticosteroid-binding globulin (CBG), and the routine test measures total cortisol — bound plus free. Estrogen raises CBG, so on a combined oral contraceptive, in pregnancy, or on oral menopausal hormone therapy, more carrier means more measured total cortisol with no excess of active hormone.

This is not theoretical. In a study comparing 41 healthy women on combined oral contraceptives with 46 women not taking exogenous estrogen, plasma cortisol was significantly higher in the contraceptive group, while salivary cortisol — the free fraction — was slightly lower.⁶ It is why guidelines warn against interpreting suppression testing on oral estrogen without stopping it first.⁷⁸

If you take an estrogen-containing pill, say so. It is often the entire explanation.

The causes, from most common to rarest

1. The timing and conditions of the draw

By a wide margin the leading cause: a late draw compared against a morning range, a very early wake-up, shift work, stress or pain in the preceding half hour, an acute illness, a hard workout the day before. A properly timed repeat settles most of these.

2. Estrogen — pill, pregnancy, hormone therapy

The second most frequent cause in women, through the CBG mechanism above.⁶ Very common, and entirely benign.

3. ⚠️ Steroid medication — and it does not do what most people assume

The most counter-intuitive point on this page — and in the US, not a rare scenario: NHANES data covering 1999–2008 put oral glucocorticoid use at 1.2% of American adults, about 2.5 million people at any given time.⁹ Prednisone is the workhorse of that group.

Synthetic steroids suppress your own cortisol production — true, and it is why stopping them abruptly is dangerous. In a meta-analysis of 74 studies and 3,753 participants, adrenal insufficiency appeared in 6.8% of people on inhaled steroids for asthma, 4.2% on nasal sprays, and up to 52.2% after a joint injection.¹⁰

But what the laboratory reports depends entirely on which molecule you take — and several of them are read by the assay as if they were cortisol:

Read that middle row again if you take prednisone. Prednisone itself cross-reacts only about 0.3%, but it is metabolized to prednisolone in the body, and prednisolone is what circulates.¹¹ So a high cortisol on prednisone, prednisolone, or methylprednisolone is most often a measurement artifact, not a disease. Tell your clinician about every steroid you use — pills, inhalers, sprays, creams, drops, injections — and never stop one on your own.

4. Genuinely raised cortisol without adrenal disease

The non-neoplastic hypercortisolism states, historically called pseudo-Cushing: obesity, poorly controlled diabetes, depression, alcohol use, sleep apnea, severe chronic stress, an ongoing illness. Cortisol is truly elevated here, sometimes enough to disturb screening tests, but there is no tumor and no adrenal disease, and treating the underlying cause is the treatment.⁸ This is why blood glucose and A1c often appear on the same order.

5. An adrenal nodule found by accident

A CT done for something else sometimes turns up an adrenal nodule producing cortisol with mild autonomy. It has its own follow-up protocol, but a single blood test does not uncover it.⁸

6. Endogenous Cushing's syndrome — rare

This is the diagnosis everyone types into a search bar, so here is the number you need to place yourself. The estimated incidence of Cushing's syndrome from endogenous cortisol overproduction is 2 to 8 cases per million people per year.¹² Across the US population that works out to roughly 700 to 2,700 new cases a year — against tens of millions of cortisol measurements. In about 60–70% of those cases the cause is a benign pituitary tumor making too much ACTH (Cushing's disease); most of the rest are adrenal.¹²¹³

It never announces itself as one mildly high number. It shows up as an evolving clinical picture: a rounded, flushed face, a fat pad between the shoulders, arms and legs thinning while the abdomen thickens, wide purple stretch marks, thin skin, spontaneous bruising, new high blood pressure, new diabetes, weak thigh muscles.¹²¹³

What should send you to a doctor promptly

It is not the number that is urgent — it is the company it keeps. Make an appointment soon if you notice:

• wide purple stretch marks that appeared recently on the abdomen, thighs, or arms;
• muscle loss in the thighs and upper arms alongside a thickening abdomen and face;
• new high blood pressure or new diabetes, especially before 40 and with no family history;
• bruises appearing without any knock, on skin that has become thin;
• in a woman, facial hair and periods stopping over the course of a few months;
• in a child, weight gain together with slowed growth in height — never a normal combination.

A cortisol slightly over the line with none of these is not an emergency.

What your doctor will do next

The path is well marked, and it rarely starts with another blood cortisol.

1. Go through your medication list. Guidelines are explicit: exogenous steroid use is ruled out before anything else is considered.⁷ Same for the pill and hormone therapy.
2. Check the draw conditions, and repeat at the right hour if needed.
3. If a cluster of clinical signs is present, move to the tests that actually decide. Three first-line options measure different things: the 1 mg overnight dexamethasone suppression test (one tablet at 11 p.m., cortisol at 8 a.m. — it should fall below about 1.8 µg/dL); 24-hour urine free cortisol, which captures total daily output independent of the rhythm and is reported against an assay-specific ceiling, commonly in the range of 45–50 µg/24 h; and late-night salivary cortisol, which probes the evening trough that disappears in Cushing's.⁷⁸
4. Confirm with a second test. A meta-analysis of 139 studies and 14,140 participants found all three perform well and comparably: sensitivity 98.6% for the overnight suppression test, 95.8% for late-night salivary cortisol, 94.0% for urine free cortisol, with specificities of 90–93%.¹⁴ But 90% specificity means roughly one false positive in every ten healthy people tested — one abnormal result is never a diagnosis, and two concordant results are required.⁷

Only then does the search for a source begin — ACTH, then imaging — with an endocrinologist.⁸ The wider hormone panel may add DHEA-S, aldosterone, renin, or an electrolyte panel with potassium, which tends to run low in frank hypercortisolism.

What this result does NOT mean

It is not a measurement of your stress level. A single draw samples one instant of a pulsatile curve. It does not quantify how much you have on your plate.

It is not "cortisol face." The idea that a puffy face signals excess cortisol circulates widely on social media. The rounded face of Cushing's syndrome is real — but it never travels alone: purple striae, muscle wasting, hypertension, and diabetes come with it, and the condition affects a handful of people per million per year.¹² A face that looks puffy on waking is far more often about sleep, salt, alcohol, or an allergy.

It is not "adrenal fatigue." The notion of adrenals "exhausted" by chronic stress is recognized by no endocrine society. A systematic review screened 3,470 articles and kept the 58 studies that actually measured cortisol in fatigued people: the findings were systematically contradictory, and no evidence supported the entity.¹⁵ If you are exhausted, the useful workup is the one in blood tests for fatigue, starting with TSH and ferritin.

No cleanse or supplement "lowers cortisol." None has an established effect. Regular sleep, physical activity, and less alcohol act on your overall physiology, not as an anti-cortisol drug — and real hypercortisolism is never treated with diet.

Should you repeat the test, and when?

Usually yes — but under better conditions.

• Drawn outside 7–9 a.m., or while ill, sleep-deprived, or acutely stressed: redo it at the right hour after an ordinary night. A few days to two weeks is enough.
• On a combined oral contraceptive: total cortisol will stay elevated as long as you take it. Repeating the identical test achieves nothing; your clinician decides between a test unaffected by CBG (salivary cortisol) and a supervised interruption.⁶⁷
• On steroids, or just after: the result is not usable, and its direction depends on the molecule. Change nothing on your own — abruptly stopping long-term steroids risks acute adrenal insufficiency.¹⁰
• With a suggestive clinical picture: the baseline test is not repeated at all; you move to the specific tests, two of which are done at home (24-hour urine, evening saliva).⁷

Cortisol has no place in a routine health checkup: it is ordered on clinical suspicion, precisely because it distorts so easily.

Frequently asked questions

Sources

Official US health sources and peer-reviewed publications (PubMed) used for this page: