Cortisol 101

“My cortisol must be through the roof” has turned into something people say about a rough week, a tired face or a stubborn belly. It sounds like a diagnosis. Usually it’s a guess, and nobody can check it by looking in a mirror. A psychotherapy practice isn’t the obvious place for an article about a hormone, but the word is now part of how people talk about stress, sleep and their bodies, and it helps to know what’s true about it before anything gets bought or feared.

What cortisol is

Cortisol is a hormone, which means a chemical messenger that travels in your blood. Your adrenal glands make it: two small glands that sit on top of your kidneys. The signal to make it starts in the brain, in a chain that runs from a small control center at the base of the brain, the hypothalamus, to the pituitary gland, which releases ACTHACTHA hormone from the pituitary gland, at the base of the brain. It tells the adrenal glands to make cortisol., and from there to the adrenal glands. Our article Your Brain on Stress shows how this slow chain sits beside the fast, adrenaline-driven one.

Cortisol doesn’t arrive in a steady drip. It’s released in pulses, roughly once an hour, and the daily pattern comes from how big those pulses are.

What it does when you aren’t stressed

Think about what your body needs on an ordinary Tuesday. Cortisol is at work in at least four places.

• Energy and blood sugar. Cortisol tells your liver to make sugar and makes muscle and fat a little less responsive to insulin, which keeps sugar available for your brain between meals. A little of this goes on every day. Too much for too long raises blood sugar, as part six explains.
• Inflammation and defense. Cortisol is part of how your body turns inflammation back down once it has done its job, which is why cortisol-like medicines have been used against inflammation since 1948.
• Blood pressure and fluid balance. Cortisol helps keep your blood pressure steady. You notice that job mostly when it goes wrong: too much raises blood pressure, and too little lets it fall, which makes people dizzy when they stand up.
• Sleep and waking. Cortisol is lowest in the first half of the night, starts climbing before you wake and gets an extra push just after. It’s part of how your body gets you going, as part two shows.

It also touches bone, muscle, mood and growth, and you tend to meet those jobs only when something goes wrong. One more detail: under long stress, the cells that respond to cortisol can start responding less. In two studies in which volunteers were exposed to a cold virus, that went with more colds in the first and more inflammatory signals in the second. So the familiar story, that stress means too much cortisol and therefore more inflammation, is simpler than the research is.

Why “stress hormone” is only part of the story

The nickname has some truth in it, and four things missing.

• It has a rhythm without any stress. Part two draws it.
• Plenty of things that raise it aren’t stressful. Waking up, bright morning light, exercise and a cup of coffee all do.
• It doesn’t rise for every stressor. In more than 200 lab studies, it went up mainly when a task felt out of the person’s control or when other people could judge how they did, and the biggest, longest rises came when both were true.
• Too little is dangerous. Without enough cortisol, the body can’t cope with illness or injury, and it can be life-threatening (part six). Hydrocortisone, the medicine, is cortisol, and people with Addison’s disease take it for life. That’s one reason a “cortisol detox” would be the wrong goal for anybody.

Adrenaline often gets mixed up with cortisol, so here is the difference. Adrenaline comes from the inner part of the same glands and is out within seconds: it’s the pounding heart and the sudden alertness. Cortisol comes from the outer layer, rises over minutes and acts for hours. Adrenaline is the sprint, and cortisol is the supply line that keeps you stocked afterward.

There’s a stretch just after you wake when you aren’t quite there yet. The room is a little too bright, your phone is a little too loud, and your body is already a few steps ahead of you. Your cortisol began climbing before your eyes opened, and it’s about to get an extra push.

The shape of a day

In a careful study of 33 healthy volunteers, blood cortisol peaked at about half past eight in the morning, then fell through the day and bottomed out around midnight, roughly eight times lower. It sat below its daily average from about a quarter to eight in the evening until half past five the next morning. That’s one small group, so take the numbers as the shape of a healthy day and not as a target to hit. The figure lets you see that shape, and what happens to it in four common situations.

The morning surge

On top of that daily curve, there’s an extra burst in the first half hour or so after you wake up, called the cortisol awakening response. Researchers still aren’t sure what it’s for; one idea is that it prepares you for the day ahead. Morning light helps push it up (more under “What bends the curve”). A big rise after waking is normal, so a high morning reading isn’t a warning sign by itself.

It’s also hard to measure well. The expert consensus says samples have to be timed strictly from the moment of waking, and even research studies often fall short of that, so a casual morning reading at home tells you very little.

What bends the curve

Several things move the curve. Here is each one, with how sure the research is.

• Sleep. After a night of too little sleep, the next evening’s cortisol was about 40% higher in young men, and the quiet phase of the night began at least an hour later. A larger review of 24 studies, though, found no significant overall change after one short night, so the nudge is modest and doesn’t show up every time. Years of short sleep went with a flatter daily pattern in a large group of British civil servants. Our article The Surprisingly Busy Business of Sleeping covers the sleep side, including what caffeine does to it.
• Light. In a small, old study, an hour of bright light after waking raised the morning rise by about a third, while the same light in the late evening did nothing.
• Shift work. Night work moves and disrupts the curve, and studies disagree about how far people adapt. Sleeping through the day doesn’t switch cortisol down the way night sleep does, and timed bright light and darkness moved the peak toward a night schedule in a small group of night-shift workers.
• Jet lag. After a long flight east across more than five time zones, travelers’ cortisol stayed on home time for a while: higher than usual late in the evening and lower in the morning.
• Meal timing. A weak lever. In a small study, pushing meals back by five hours for six days shifted blood sugar rhythms but not cortisol, so “eat at this hour to fix your cortisol” has no good support yet.
• Age. Between 20 and 80, average cortisol rose by 20 to 50%, the night-time low rose, and the swing between day and night got smaller and earlier.

Caffeine and exercise nudge it too, and part four covers both.

What a flat curve does and doesn’t mean

Researchers have also looked at the slope of the day’s curve, meaning how steeply it falls from morning to night. Across 80 studies, a flatter slope went with poorer health on average, but the link was small. In just over 4,000 British civil servants, a flatter slope predicted dying from any cause over the years of follow-up.

What nobody knows is whether a flat slope causes harm or only marks it. As far as we could find, no trial has shown that changing the slope changes health, no guideline uses a person’s slope to diagnose anything, and working it out from a few saliva samples at home is shaky. So a flat curve on a home kit isn’t a diagnosis of anything.

Think of the hour before a hard conversation or a job interview. You rehearse, you check the time, and your body is clearly doing something. Then it’s over and your system settles. Now think of the same body a year into low-grade strain: a job that never lets up, a money worry that comes back every month. Those are two very different situations for a stress system, and cortisol doesn’t behave the same way in both.

A short spike is the system working

In the lab, a stressful task makes cortisol rise, peak and then come back down. That’s the system doing its job: short rises help the body mobilize, and it’s long exposure that turns costly. Recovery is slowest when the situation is both out of your control and judged by other people.

Long stress isn’t one thing

You might expect long stress to mean a body soaked in cortisol. The research is less tidy. In a meta-analysismeta-analysisA study of studies. It pools the results of many separate studies of one question, which gives a steadier answer than any single study can. of the human studies, cortisol was raised when a stressor started and then declined as time went by.

Stressors that threatened physical safety, involved trauma or couldn’t be controlled produced a high, flat daily pattern, and people with PTSD tended to have lower output. Hair, which stores cortisol over months, tells a similar story: people who were still living with a stressor had hair cortisol roughly 40% higher, while people whose stress had passed did not.

Depression: a real link, smaller than the headlines

Depression is the diagnosis most often tied to cortisol. People with depression have higher cortisol on average, but the difference shrinks by nearly half when only the studies that met basic standards are counted. In teenagers and young adults, higher morning cortisol came before depression and didn’t differ once it had started, which points to a risk factor and not a marker of current illness. So the link is real, it’s smaller than the headlines make it, and you can’t see depression in a cortisol test. Nothing here says cortisol causes it.

Some conditions show low cortisol

At the other end, some people under long stress have lower cortisol than usual. In PTSD, morning and 24-hour cortisol were lower on average than in people without it, though not in every study. Burnout is a word to handle carefully. The World Health Organization’s ICD-11 calls it an occupational phenomenon and not a medical condition, and reviews haven’t found a reliable cortisol marker for it. A smaller morning rise has also been found in chronic fatigue syndrome. One idea, and so far only an idea, is that low cortisol can follow a long stretch of high output. What stress does to your thinking and memory is in our article Your Brain on Stress.

You’re scrolling late, and somebody with a calm voice explains that your puffy face is cortisol, and that a drink, a supplement or a morning routine can fix it. By the end there’s a link. These claims spread because they are tidy and because the tiredness behind them is real. This part gives each one a fair hearing: what is true in it first, then what the research says, then what to do instead. Pick a phrase in the figure to see all three, or read on.

“Cortisol face”

A rounder, puffier face is a real sign of Cushing’s syndrome, a condition in which the body has had too much cortisol for a long time, and of long-term steroid tablets. Doctors call it moon face. Studies of faces in Cushing’s find a real pattern, with fat settling toward the cheeks and lower face.

What nothing supports is the idea that everyday stress does it. Cushing’s is rare (part six), steroid medicine is its most common cause, and doctors don’t diagnose it from a rounder face alone. They look for several signs together, such as easy bruising, a pad of fat over the upper back and high blood pressure. Many ordinary things change how a face looks from one day to the next, including sleep, salt, alcohol, allergies and aging.

If you find yourself checking your face in the mirror for signs, that’s worth telling someone. It’s a good reason to talk, and it isn’t a reason to buy anything.

“Cortisol belly”

There’s a small grain of truth here. In Cushing’s, fat gathering around the middle is a hallmark. In everyone else, long-term cortisol (measured in hair) goes with body size only slightly, and nobody knows which way it runs: whether cortisol changes the body or the body changes cortisol. A study of several hundred adults with overweight or obesity found that cortisol didn’t track weight, and the authors concluded the data don’t support a strong link between cortisol or stress and obesity. There’s no cortisol plan that targets belly fat, and weight changes for many reasons.

“Adrenal fatigue”

The idea is that years of stress wear out the adrenal glands, causing tiredness, cravings and brain fog, which a saliva panel can show and a supplement can fix. The Endocrine Society, a US professional body for hormone specialists, says there’s no scientific proof that adrenal fatigue is a real medical condition and no test that detects it, and that the tests some providers use aren’t backed by solid research. A review of 58 studies reached the same place: it found no substantiation. Supplements sold for it can do harm, too. Adrenal hormones taken when they aren’t needed can suppress the adrenal glands and risk an adrenal crisis when they are stopped (part six).

None of that makes the tiredness less real. Tiredness that has gone on for weeks deserves a proper look from a doctor, because the explanations the Endocrine Society and the UK’s national health service list include adrenal insufficiency, depression, sleep apnea and thyroid disease.

Cortisol cocktails and detoxes

A “cortisol cocktail” is a drink, often orange juice and coconut water with a pinch of salt and sometimes magnesium, that’s said to balance your adrenals. We found no controlled trial of it. Dietitians and doctors quoted in news pieces say there are no clinical data behind it and that at best it supplies fluid and salts. If you like the taste, it’s a drink. And the idea it rests on, that your adrenals are worn out, hasn’t held up.

A “cortisol detox” fails for a different reason. Cortisol isn’t a toxin: your body needs it for blood pressure, blood sugar and energy, and it’s meant to be highest in the morning and lowest at night. A critical review of detox diets in general found little clinical evidence for them and no randomized trials of the commercial ones. Products promising to control cortisol have been sold on similar promises for more than twenty years. By 2007 the US Federal Trade Commission had settled with the makers of two of them over false and unproven claims about weight loss and disease prevention.

Supplements

Ashwagandha is the one most readers meet, so it gets the most space. Many people say it helps them feel calmer, and that deserves to be taken seriously, because calm is what they were after.

What the trials show is more mixed. In short trials it lowered a lab measure of cortisol by a small amount. But in a 2025 meta-analysis of seven trials, people’s own ratings of how stressed they felt didn’t change significantly, even though cortisol did, while an earlier review of nine trials had found lower stress and anxiety scores as well. Many of the trials were small, and we didn’t check who paid for them. Put together, that’s a modest effect on a lab number and an unclear effect on how people feel.

The bigger question is safety, and that is disputed. There are reports of serious liver injury in people who took it: in one series from Iceland and the US, five people developed jaundice within weeks (none went into liver failure, tests returned to normal in four, and one was lost to follow-up), and in a series from India, three people who already had long-term liver disease died. Case reports can’t say how often this happens.

Germany’s federal risk assessment institute advises children, pregnant and breastfeeding women and people with liver disease not to take ashwagandha supplements, says others should be cautious, and says it couldn’t set a safe intake. Denmark banned it in food supplements in 2023, and the Netherlands, France, Australia and the UK have reviewed it or issued warnings. Some researchers in the Ayurvedic field say the Danish report was flawed and that no causal link to liver injury has been shown. As of October 2026, we found no decision from the European Union as a whole.

The other names that come up are phosphatidylserine, magnesium, L-theanine and rhodiola. They have little or no good evidence for lowering cortisol itself. Some have small trials for anxiety or tiredness, usually without a cortisol measure. Magnesium, for example, looked suggestive for anxiety in people prone to it, but the studies were of poor quality, and that review looked at how anxious people felt, not at cortisol. In the US, the Food and Drug Administration doesn’t oversee supplements the way it does medicines, so what is on a label may not match what is in the bottle.

“Low-cortisol workouts”

“Cardio spikes your cortisol, so train gently” has a true core. A single session of moderate to hard exercise does raise cortisol, more the harder you go, and a gentle one barely moves it. The rise is brief and normal, and trained bodies adapt. Across ten studies of physical activity programs, cortisol went down on average and sleep got better. An expert consensus statement says no hormonal marker, cortisol included, can diagnose overtraining. So “cardio raises cortisol” is true for a few minutes and is no reason to skip it. If you feel exhausted from training, rest matters, and a cortisol test isn’t how to find out.

Coffee

Coffee raises cortisol for a while after you drink it. People who drink several cups a day are partly tolerant: in 96 adults, daily drinkers lost the cortisol response to a morning dose, although later doses still raised it, and by evening cortisol was back to normal levels. Nothing here says when to have your first cup, because we didn’t look into that.

A lab result is on the screen: a number, a printed range, and a small arrow beside your value pointing up or down. What does the arrow mean? Often much less than it looks like, and with cortisol especially.

Why one number says little

Cortisol swings roughly eightfold across a day, comes in pulses, and moves with sleep, posture, food, pain, illness, light, exercise, the pill and pregnancy. So whether a reading counts as high or low depends on when and how it was taken. Try the slider in Figure 1: pick a time of day and watch the same body give very different numbers. A doctor does order a morning blood cortisol, for a specific reason, such as checking whether your adrenal glands have recovered after steroid medicine. It’s a tool for a question and not a score for stress.

What doctors actually use

Doctors use different tests for different questions, and knowing the names helps if you ever need to ask for the right one.

In a pooled review of 139 studies, each of the main tests for too much cortisol picked up the great majority of people who had Cushing’s and flagged fewer than 1 in 10 people who didn’t, among people who had already been sent for testing. That’s why they are trusted, and why a second test usually confirms the first. No cut-off numbers are quoted here, because they differ between labs. Testing is for suspected Cushing’s or adrenal insufficiency and is ordered by a doctor. Stress on its own isn’t diagnosed by a cortisol test.

Hair cortisol

Cortisol is also stored in hair, and a few centimeters of it show roughly the past few months. In research it picks up long stress exposure, with hair cortisol about a fifth higher in stress-exposed groups. But it needs adjusting for age, sex, hair washing, hair treatments and the pill, and it remains mainly a research tool, though clinics are starting to use it for rare conditions in which cortisol rises and falls in cycles.

Home kits

Mail-order saliva, urine and hair kits promise to show your cortisol pattern, your adrenal health or your stress level. We found no peer-reviewed check of any mail-order panel. The one peer-reviewed validation we found, of a portable saliva device, agreed only moderately with lab results and read low at rest. The morning surge needs samples timed strictly from waking, which most studies fail to manage. The Endocrine Society (US) says the tests some providers use for “adrenal fatigue” aren’t backed by solid research. A kit can’t tell you what to do. If you’re worried about a hormone problem, the route is the doctor’s tests above.

Why your number isn’t mine

Two healthy people can have different cortisol for ordinary reasons. Sex and age shift averages a little: in one large Swedish study women had higher morning cortisol than men, and cortisol tends to creep up with age. Hard early experiences leave a small and inconsistent mark, nowhere near enough to read a history from a number. Genes play a part too, though no single common gene variant explains much.

The biggest reason is a protein. Most of the cortisol in your blood rides on a carrier protein, and only the unattached, or free, part is active. Estrogen from the pill, from oral estrogen medicine and from pregnancy raises the carrier protein, so a standard blood test reads higher while the active part usually moves much less. Pick a situation in the figure to see total and free side by side.

Swallowed as a tablet, estrogen raised total cortisol by about two-thirds in one small study, and a low-dose pill nearly tripled it in another, while free cortisol barely moved. In pregnancy, total cortisol climbs to about three times the usual level by the third trimester, the active part rises too, and that’s normal. How your cycle and your stress system talk to each other is in our article So, Is It Your Hormones?.

Imagine someone who has been on steroid tablets for a few months for a long-term condition. They feel much better, and they wonder, quite reasonably, whether it’s time to stop. That question is why this part is here. These conditions are rare. If you take steroid medicine, here is what to ask, and if a group of real signs is getting worse, here is when it’s worth seeing a doctor. It isn’t a checklist for finding something wrong with yourself.

Too much: Cushing’s syndrome

Cushing’s syndrome means the body has been exposed to too much cortisol, or to a cortisol-like medicine, for a long time. The most common cause is steroid medicine, taken as tablets and sometimes in high doses by other routes. The much rarer kind starts in the body itself, most often with a small tumor on the pituitary gland and less often on an adrenal gland. In European registries it appears in about two or three people per million per year, and most of the data come from Europe, with little from the rest of the world.

It comes on gradually. Doctors look for a group of signs that build over months or years, such as easy bruising, wide purple stretch marks, muscle weakness and high blood pressure, along with the changes in the face and middle that people talk about online. Each sign on its own has many other causes. It matters to find, because it’s linked with high blood pressure, diabetes and weaker bones, and it’s treatable: usually by treating its cause, which is often surgery for a tumor, or by gradually reducing a steroid medicine under a doctor. The tests are in part five.

The common one: steroid medicines

This is the part of the story that reaches the most people. About 1 in 100 adults takes steroid medicine for a long time. Steroid medicines are cortisol-like drugs, also called corticosteroids or glucocorticoids. While you take them, your body senses plenty of cortisol-like hormone and turns down its own production. After weeks of tablets the adrenal glands may be switched off, and for a while after you stop they may not restart fully. It can happen with steroid injections too, and less often with inhalers, nasal sprays and creams. In a pooled review of 74 articles, no route, dose or length of treatment could rule it out completely.

Coming off steroids can feel rough, and the feelings of withdrawal overlap with those of low cortisol, so working out which is which is a job for a doctor and not for guessing. How fast the adrenal glands recover varies a great deal from person to person. The guideline suggests seeing an endocrinologist, a hormone specialist, if there’s no recovery after a year on a low replacement dose, or after an adrenal crisis.

Too little: Addison’s disease and adrenal insufficiency

Adrenal insufficiency means the body doesn’t make enough cortisol. When the adrenal glands themselves are the problem, it’s called Addison’s disease, and in richer countries it’s usually autoimmune, which means the immune system attacks the glands. When the pituitary gland doesn’t send enough ACTH, it’s called secondary, and steroid medicine is one of its leading causes. In Iceland, about 22 in every 100,000 adults had Addison’s disease, more than older reports suggested. Most of the data on this come from Europe.

Because it begins with ordinary-sounding things like lasting tiredness, adrenal insufficiency is often diagnosed late. The help box below says which combination is worth a visit. It’s treatable. People take daily hydrocortisone and, with Addison’s disease, a second medicine called fludrocortisone, take extra doses when they are ill or having surgery, carry a steroid card, and many have an emergency injection and training in how to use it. With treatment, people live normal lives.

Adrenal crisis

Adrenal crisis is a life-threatening shortage of cortisol, usually set off by an infection, an injury, surgery, or vomiting and diarrhea. In a survey of 444 people with adrenal insufficiency, 42% had had at least one crisis, and the most common trigger was a stomach infection or fever. Vomiting matters for a particular reason: tablets can’t work if they don’t stay down, which is why injectable hydrocortisone is prescribed.

It’s nine at night, the day took everything, and some part of you is looking for the one thing that would fix it. A hormone makes a very tempting single thing, because it has a number and, apparently, an off switch. Here is what the research can and can’t say about the ordinary things that do help.

A lower cortisol number isn’t the goal here. The goal is how you feel and function, and cortisol is a side measure that trials record inconsistently. Across 78 trials that took saliva samples over the day, the methods varied so much that the results were hard to compare. So the figure shows how much each approach moves cortisol on average, drawn so you can see how small “small” is.

Sleep

Most of the research is about what happens when sleep goes wrong (part two), and not about whether fixing sleep lowers cortisol. We found no trial that shows that. Sleep still earns its place on this list, because it’s tied to nearly everything else, including mood and memory.

Movement

The World Health Organization recommends 150 to 300 minutes of moderate activity a week for adults, or 75 to 150 minutes of vigorous activity, plus muscle-strengthening. In ten studies of activity programs, cortisol came down by a small to moderate amount and sleep improved. How movement helps mood, anxiety and sleep is in our article Motion and Emotion.

People and touch

In one lab study, men who had their best friend with them before a stressful task had a smaller cortisol rise, and a smaller one still when they also had oxytocin, a hormone linked to bonding. In another, a hug or a hand held over the heart lowered cortisol after a stressful task in 159 adults, even though their heart rate and their own ratings of stress didn’t differ. These are small lab effects, so we don’t know yet how much of this carries into daily life.

Time in nature

Cortisol was lower after time in forests than in cities in a pooled analysis of eight studies, though expectation may play a part. In a study of 36 city dwellers, a nature break of 10 minutes or more lowered cortisol, most efficiently at 20 to 30 minutes. The effect is small and short-lived, and the second study had no comparison group.

Breathing, mindfulness and yoga

Slow breathing reliably calms the body in the moment, but its effect on cortisol isn’t established: one review found only three studies of belly breathing, all of them showing less stress, and one with lower saliva cortisol. Mindfulness, meditation and yoga did better, with small to moderate drops in cortisol across dozens of trials, though the results varied a lot. How to do slow breathing is in our article Slow Breathing.

Therapy

Talking therapies had a small effect on cortisol that wasn’t statistically significant in the large review of 58 trials, and psychotherapy for PTSD improved people’s symptoms without changing their baseline cortisol. Therapy works for people whether or not their cortisol changes. Cortisol is a side measure, and it isn’t what therapy is for.

Laughter and music

Laughter looks strong on paper: eight small studies found cortisol about a third lower than in comparison groups. Small studies tend to overstate, though, so that number is probably too high. Music had a small to moderate effect on stress hormones across a large review. In one lab study, relaxing music before a stressful task didn’t lower cortisol, but the sound of water did.

Free help

The World Health Organization has a free illustrated guide, Doing What Matters in Times of Stress, in about 50 languages and with audio.

These are the questions people type into a search bar late at night, or ask a friend instead of a doctor. None of them is a silly question, and each has a short answer.