How Do I Know If My Puffy Face Is Cortisol or Something Else?

Your face looked different this morning. Rounder in the mirror, a bit tight under the eyes, and the phone in your hand has an explanation ready: cortisol. Stress hormones. Here is a routine that fixes it.

So let us do the thing the videos never do, and check.

“Cortisol face” is not a diagnosis. There is no such entry in any medical reference, no test for it, no treatment, because the phrase was built for a caption rather than for a clinic. But it is not pure invention either, and that is what makes it stick. Genuine, sustained cortisol excess really does change a face. Doctors have had words for it for a long time: moon facies for the rounding, facial plethora for the flushed fullness. Those are clinical signs of Cushing’s syndrome, a rare and serious disease.

The meme took the vocabulary of that disease and stapled it onto something ordinary and harmless. Half of it is real, which is exactly why the whole of it sounds credible.

What actually separates the two

Puffiness and pathological fullness behave differently, and you can watch the difference without any equipment.

Ordinary puffiness moves. It is worst when you first stand up, better an hour later, worse after a salty meal or a late drink or a short night, and it comes and goes across a week. It is fluid obeying gravity and sodium and sleep. The Better Health Channel links excess sodium intake to fluid retention, and points out that the average Australian eats close to twice the sodium the body needs — so most of us are running a mild version of this most days.

Cushing’s facial change does not move. It builds. It is there in the evening as much as the morning, it is still there next month, and crucially it does not arrive alone. The Endocrine Society’s diagnostic guideline singles out the features that actually discriminate the disease from everything it can be confused with: reddish-purple stretch marks, plethora, weakness in the large muscles close to the trunk, bruising with no obvious cause, and unexplained thinning of the bones. Around those sit weight settling on the torso while the limbs thin, a fat pad between the shoulders, blood pressure that will not settle, new diabetes.

The guideline also suggests something you can do yourself tonight: look at old photographs. Its wording is that reviewing them may help a clinician appreciate whether physical change has occurred over time. If your face two years ago looks like your face now under different lighting, that tells you something. If there is a slow, steady march across the photos, that tells you something else.

What you are watching	Ordinary fluid	Worth investigating
Time of day	Worst on waking, eases	Constant, all day
Direction over months	Fluctuates	Progresses
Company it keeps	Nothing else	Bruising, purple striae, weakness

How rare are we talking

Cushing’s syndrome arising from the body’s own overproduction runs at roughly 1.8 new cases per million people each year. That is not a hedge word — it is genuinely uncommon, and the arithmetic on a puffy-faced adult with no other signs is overwhelmingly in favour of salt, sleep, alcohol or a head cold.

Rare does not mean harmless. Left untreated, around half of people with endogenous Cushing’s die within five years, mostly from cardiovascular causes. Both things are true at once, and holding both is the whole skill here.

And there is a cause the wellness feed never mentions, because there is nothing to sell against it. The commonest reason anyone develops Cushingoid features is not a tumour. It is medication — corticosteroids in every form, oral tablets, high-dose inhalers, potent topical creams, joint injections, nasal sprays, depot shots. The guideline’s very first recommendation is to take a thorough drug history and exclude that before any biochemical testing begins. If your face has changed and you are on a steroid in any form, that is the conversation to have with your GP, and not a reason to stop anything on your own.

The timeline nobody gives you

This is the gap in almost every article ranking for this question. They say “prolonged exposure” and leave you to guess whether that means a stressful fortnight or a decade.

Here is a number. A meta-analysis pooling 45 studies and 5,560 patients found the average time from symptoms beginning to Cushing’s syndrome being diagnosed was 34 months, and that this had not improved. Roughly three years, in a disease being actively looked for by multiple doctors.

Read that in the direction that helps you. It means the real thing is slow, insidious and easy to miss — which is a reason to take a progressive change seriously. It also means a face that puffed up on Tuesday and settled by Thursday is not that disease. Whatever your face did this week, cortisol excess did not do it that fast.

The tests, named

The vague “get a cortisol panel” advice is worse than useless, because there is a booming market in panels that answer nothing.

The Endocrine Society guideline lists four initial tests: urine free cortisol measured on at least two collections, late-night salivary cortisol on two separate nights, the 1 mg overnight dexamethasone suppression test, and a longer low-dose dexamethasone protocol run over 48 hours. Then it says something the direct-to-consumer market would rather you did not know. It recommends against random serum cortisol for testing Cushing’s syndrome, on the grounds that the accuracy of such measurements is too low.

That is the single most useful sentence in this article. A one-off morning blood cortisol — the exact thing people order for themselves and then panic over — cannot answer this question. It was never going to. Cortisol is supposed to be near its daily peak in the morning, so a high morning reading is your body working, and the number arrives with no clinical question attached for it to answer.

The same applies to the mail-order saliva and hair kits, and to the four-point salivary curve marketed for “adrenal fatigue” — that particular test is not validated for that particular purpose, and “adrenal fatigue” is not a diagnosis endocrinology recognises. Late-night salivary cortisol, done properly for a specific suspicion, is real medicine. The same molecule measured with no question behind it is a graph you paid for.

When to stop reading and book an appointment

Go and see your own GP, promptly, if you have facial change alongside any of: wide purplish stretch marks, bruising you cannot account for, muscle weakness that makes standing from a chair hard, blood pressure or blood sugar that has suddenly stopped behaving, unexplained weight loss, dizziness every time you stand, or skin darkening without sun.

Go too if the change is simply progressive and it worries you. Rare does not mean never, and the only way to rule a serious thing in or out is to have someone look. This article is general health education, it cannot examine you, and nothing in it is advice about your particular body.

What we still do not know, and one thing we got wrong

Nobody can tell you how much puffiness is too much. There is no threshold, no measurement, no cut-off — the discrimination in real practice comes from the cluster of signs and the direction of travel over time, which is a fuzzier tool than anyone would like.

There is also no good evidence base on ordinary facial fluid retention itself. How fast it resolves, how much sodium moves it, how much a bad night contributes: the honest answer is that this has barely been studied, because it is not a disease and there is no funding in it. What is written above about salt and sleep is mechanism and general guidance, not trial data.

And a correction, since this site publishes those. In an earlier draft of the book this article draws on, the machine wrote that morning daylight helps by lowering cortisol. That is backwards. The literature examines bright morning light acutely nudging cortisol up, as part of the normal waking rise — its value is anchoring your body clock, not damping a hormone. The verification pass caught it and the line was rewritten. The books are machine-written and doctor-checked, and that check exists precisely to catch the machine’s own overreach before you read it.