Cortisol and sleep: what the evidence actually tells us

What cortisol is for

Cortisol is a hormone made by the adrenal glands, and one of its jobs is helping the body respond to stress.[24] That job description has turned it into the most quoted hormone in wellness writing, usually cast as the villain.

The more useful fact is that cortisol is not supposed to sit at one level. It follows a daily cycle, and how much of it a person releases is influenced by sex, health, everyday habits and how stressed they feel.[9] A hormone that moves this much, for this many reasons, does not fit into a one-line explanation. That is partly why so many people try to squeeze it into one.

The daily curve

In healthy adults, cortisol is typically at its lowest around midnight. It starts to rise around 2 to 3 AM and peaks at roughly 8:30 in the morning.[8] Those times come from a 24-hour study of 33 healthy people, so they are group averages, and they move with when people sleep.[8]

around midnight

the lowest point of the day[8]

2 to 3 AM

the slow climb begins[8]

about 8:30 AM

the daily peak, on average[8]

30 to 45 min

the extra rise right after waking[10]

On top of that slow climb sits a sharper rise in the first 30 to 45 minutes after waking, which researchers call the cortisol awakening response.[10] What that morning surge is for is still an open question.[9] Over the rest of the day levels fall, and cortisol normally drops at night.[24]

Two practical points follow. A high reading in the morning is what the rhythm is meant to produce, not a warning sign in itself. And for most people the climb toward morning is already under way in the small hours, asleep or awake. That timing is a normal feature of the curve. It does not explain why any particular person wakes at any particular time, and it should not be read as if it does.

What a short night does

The study most often quoted here is old and small, and still worth knowing. Researchers restricted or removed one night of sleep in healthy young men, then measured cortisol the following day. After partial sleep loss, evening cortisol was 37% higher; after a night with no sleep, it was 45% higher. The usual evening slide toward the quiet night-time level also began later.[12]

It was a lab study of short-term sleep loss in young men only, so it cannot tell us what happens to women, to older adults, or to people who sleep a little short for months.[12] What it does show is that the connection can run from sleep to cortisol. Popular writing tends to put it the other way round, with cortisol keeping people awake. Of the two directions, sleep affecting cortisol is the one our sources test directly.

Caffeine and cortisol

Caffeine raises cortisol. In a randomized study of 96 healthy young adults, caffeine increased cortisol during waking hours, and in people who used it regularly the response was smaller but still present.[14] So the idea that a daily coffee habit leaves you fully tolerant is not what this study found.

The limits matter just as much. Samples were taken between 7:30 AM and 7 PM, and sleep was not measured at all, so the study says nothing about what happens overnight.[14] Caffeine and sleep is a separate question with separate evidence: in a small trial, a large single dose disrupted sleep even when it was taken six hours before bed.[13] If you want to look at your own caffeine, the timing of your last cup is something you can see. Your cortisol response to it is not.

The “flattened rhythm” research, and its limits

You may have read that a “flat” cortisol curve, one that falls only a little from morning to evening, sits behind fatigue and poor sleep. There is real research underneath that claim. A meta-analysis pooling 80 studies found that flatter daytime slopes were associated with poorer mental and physical health.[11]

Three details keep it in proportion.

• It is an association. It cannot say whether a flatter curve came first, followed from illness, or whether both trace back to something else.[11]
• The average effect was small, a correlation of about 0.15.[11]
• The studies measured cortisol in different ways, which makes them harder to compare.[11]

For researchers studying large groups, it is a meaningful pattern. For one tired person, it is not an explanation, and it is not something anyone can see by asking how they slept.

Why this can’t be checked at home

If cortisol moves this much across a day, a single number means little unless you know exactly when it was taken. Expert guidance on measuring the morning rise says valid results depend on precisely timed saliva samples, starting at the moment of waking.[10] That guidance was written for research studies, not for diagnosing individuals.[10]

Clinical testing is a different process. Diagnosing a cortisol disorder involves a medical history, a physical exam and lab tests, and because no single test is perfect, doctors usually use more than one.[24] Guidance for suspected cortisol excess says anyone with an abnormal first result should see an endocrinologist and have a second test before any diagnosis is made.[25]

A home kit, a symptom quiz or a description of your nights does not get you to the same answer by a shorter route. If you have reason to think your cortisol is out of range, that is a conversation for a clinician, who can decide whether testing makes sense at all.

About “adrenal fatigue”

The phrase usually comes with a story: years of stress wear the adrenal glands down, cortisol runs low, and you end up tired and on edge at once. A systematic review of 58 studies found no substantiation that adrenal fatigue is an actual medical condition.[22] The Endocrine Society states that no scientific proof supports it as a true medical condition.[23]

None of that means the tiredness people describe is imaginary. It means this particular explanation has not held up. Our concern with the label is practical: a name without evidence can pull attention away from things that can actually be checked. Genuine disorders of cortisol do exist, and they are diagnosed through the clinical testing described above.[24]

Myth and evidence, side by side

What you can actually observe

You can’t see your cortisol. You can see a great deal about your days and nights: when you went to bed and got up, how often you woke and roughly when, when your last caffeine and any alcohol landed, when you first saw daylight, and what filled the evening. Each of those has evidence of its own. Caffeine timing affects sleep,[13] alcohol tends to unsettle the second half of the night,[15] and light is the main signal that keeps the body clock in line with the day.[17]

Writing these down for one to two weeks is what the NHLBI suggests doing before a doctor’s appointment about sleep.[3] A diary like that will tell you nothing about hormones. It can show patterns in things you have some say over, and it gives a clinician something concrete to work from.