Cortisol

What it actually measures

Cortisol is called the stress hormone, which is true and misleading in roughly equal measure.

Most of what it does has nothing to do with stress. It regulates blood sugar, blood pressure, immune activity and inflammation, and it is central to how you wake up in the morning. Without it you could not function.

The defining feature for anyone interpreting a result is its rhythm. Cortisol peaks sharply within about half an hour of waking, falls steeply through the morning, continues declining through the day, and bottoms out around midnight.

That swing is enormous. A perfectly healthy person can have a morning cortisol several times their evening value, and both are entirely normal.

So a single measurement is one point plotted on a steep curve, and the time it was taken is most of the interpretation rather than a footnote.

A blood test measures total cortisol, both the protein-bound majority and the small free fraction that is biologically active.

What it is genuinely for, and what it cannot do

Let me be direct about scope, because cortisol is one of the most over-ordered and over-interpreted tests in wellness.

A morning serum cortisol is a real diagnostic test with real uses. It is part of the workup for Cushing's syndrome, where cortisol is genuinely too high, and for adrenal insufficiency, where it is genuinely too low. Both are serious conditions properly diagnosed with dynamic testing that a single measurement feeds into.

The 2008 Endocrine Society clinical practice guideline on diagnosing Cushing syndrome sets out which measurements actually have diagnostic accuracy, recommending urine cortisol, late-night salivary cortisol, or a dexamethasone suppression test as valid initial tests, with a midnight serum cortisol reserved as second-line. It also recommends two concordant abnormal tests before anyone goes looking for a cause.

Notice what is in that list and what is not. A single daytime serum cortisol is not the recommended first-line test even for the condition this marker genuinely diagnoses.

What it is not, in any version, is a measure of how stressed you have been. A raised morning cortisol in someone under pressure at work is not a finding.

There is a whole category of interpretation built around cortisol curves, usually salivary, offered as evidence of adrenal fatigue.

A 2016 systematic review in BMC Endocrine Disorders searched 3,470 articles and analysed 58 studies, examining specifically the tests that concept rests on: direct awakening cortisol, the cortisol awakening response and salivary cortisol rhythm. It found conflicting results almost systematically across those studies and concluded there was no substantiation for adrenal fatigue as a condition.

I want to be precise about what that does and does not say, because it is easy to hear as dismissal.

It does not say stress has no physical effects. A 2021 randomized crossover study found that clamping cortisol and testosterone during four nights of short sleep cut the resulting insulin resistance roughly in half, which is direct evidence that cortisol changes have real metabolic consequences.

What the review says is narrower: the specific cortisol testing used to diagnose adrenal fatigue does not produce consistent findings, so building a treatment plan on a cortisol curve is building on unreliable ground.

So cortisol matters, and measuring it in a single sample tells you far less about your life than it appears to.

If long-term exposure is genuinely the question, there is a better tool than any single draw. A 2015 review in the European Journal of Endocrinology noted that blood, saliva and urine cortisol all vary with acute stress, diurnal rhythm and pulsatile secretion, so none reflect long-term exposure, whereas scalp hair cortisol captures cumulative exposure over months and can even build a retrospective timeline. That is a research and specialist tool rather than a routine one, and it is the honest answer to what a single number cannot do.

What moves it

Down: adrenal insufficiency, which is a medical diagnosis requiring dynamic testing. Long-term corticosteroid use suppressing your own production, which is why steroids are tapered rather than stopped. Pituitary problems, which lower the signal. And simply the time of day, since evening values are normally a fraction of morning ones.

Up: the normal morning peak, which is the commonest reason a result looks high. Acute physical or psychological stress. Cushing's syndrome. Corticosteroid medication. Pregnancy and oestrogen, which raise the binding protein and therefore total cortisol. Significant illness. And short sleep, which raises it measurably.

The honest hedge

I am not a doctor and I am not diagnosing anyone, and there are two presentations here that need one promptly.

Features suggesting Cushing's syndrome include rapid central weight gain with purple stretch marks, easy bruising, a rounded face and muscle weakness. Features suggesting adrenal insufficiency include profound fatigue, low blood pressure, salt craving, weight loss and skin darkening. Both need proper assessment rather than a wellness plan.

Short of those, this is a poor first test for ordinary fatigue, and I would rather say so than let someone spend on it.

If exhaustion is your question, a full thyroid panel, ferritin with hs-CRP, B12 and folate, vitamin D3 and fasting insulin answer considerably more from the same draw. And if you never wake refreshed, a sleep study answers more than any of them.

The stress load itself is worth addressing whether or not a test confirms it, and it does not require a lab result to be worth working on.

That is not me telling you your exhaustion is imaginary. It is me pointing at the tests that will actually tell you something.

So. What to actually do.

Tonight, at no cost. If you have had a cortisol drawn, check whether the time was recorded. Without it, the result is close to uninterpretable given how steeply the curve falls.

Do not build a supplement protocol on a cortisol curve. The systematic review evidence does not support it, and the money buys more elsewhere.

Test the things that answer more. A thyroid panel, ferritin with hs-CRP, B12 and folate and fasting insulin, all from one draw.

If you never wake refreshed, ask about a sleep study. That is the investigation exhaustion most often actually needs.

Address the load regardless of any number. Stress affects sleep, blood sugar and recovery whether or not a test confirms it.

See a doctor promptly for the specific patterns. Central weight gain with purple stretch marks and easy bruising, or profound fatigue with low blood pressure, salt craving and skin darkening.

If a cortisol is ordered, draw it early morning and record the time. Between 7am and 9am is the convention.

You are not failing a stress test. You are a rhythm, sampled once, at an hour that decided most of what the number said.

Your body is not broken. It is blocked. And often the block is thyroid, iron or sleep, sitting underneath a question everyone agreed to call adrenal.

Go check whether the time was even recorded.