Cortisol: what the stress hormone does to skin, sleep and energy

Cortisol: what the stress hormone does to skin, sleep and energy

Take a saliva sample from a healthy adult at seven in the morning, then another from the same person at ten at night. The two numbers can differ by a factor of ten. Nothing has gone wrong. That is what a working cortisol rhythm looks like.

Which is the first useful thing to say about cortisol: a single measurement tells you almost nothing. The shape of the curve across the day carries the information, and the shape is what the popular conversation leaves out.

The short version

• Cortisol is a daily curve, not a single number. It peaks around your habitual waking time and should reach its lowest point late in the evening, and that trough is the first part of the curve to go when something is genuinely wrong.

• The nuance that matters most: the link between a flattened curve and poorer health is small, r = 0.147, and drawn mostly from cross sectional work. Home saliva and hair tests sell you a value without the interpretation that makes it mean anything.

• Health Royals sells no cortisol product. Peak Performance is a daily nutritional base with 25 µg vitamin D3 and B12 as MecobalActive per daily dose. Vitamin D contributes to the normal function of the immune system, and vitamin B12 contributes to the reduction of tiredness and fatigue. It is not a substitute for sleep.

Health Royals Peak Performance, supplement with vitamin D3, K2 as MenaQ7, folate as Quatrefolic and omega-3 from algal oil

No cortisol product, but a nutritional base

Peak Performance is formulated as a daily nutritional base, not for the HPA axis. It provides 25 µg vitamin D3, B12 as MecobalActive and folate as Quatrefolic per daily dose. Vitamin D contributes to the normal function of the immune system, and vitamin B12 contributes to the reduction of tiredness and fatigue. It does not lower cortisol, and it is not a substitute for sleep.

See Peak Performance

What is cortisol and why is it not the enemy?

Cortisol is a glucocorticoid hormone made in the adrenal cortex, and its job is to hold blood glucose, blood pressure and immune activity inside a workable range from one hour to the next.

It is released by the hypothalamic pituitary adrenal axis, and belongs to the signalling molecules usually called stress hormones. Adrenaline handles the first seconds of a threat, cortisol the hours that follow. We have written about that longer arc in a separate piece on the biology of stress.

Both extremes are genuine disease. Sustained glucocorticoid excess is Cushing's syndrome, which the Endocrine Society clinical practice guideline puts at roughly two to three cases per million people per year in Europe. Sustained deficiency is adrenal insufficiency. Both are diagnosed with laboratory tests, not with lifestyle. The interesting question is not how much cortisol you have. It is when you have it.

How does a normal cortisol curve look over 24 hours?

It peaks around the time you habitually wake, declines through the day, and reaches its lowest point in the late evening and the first hours of the night.

On top of that slope sits a separate event. Within the first hour after waking there is an additional rise, the cortisol awakening response. A short review in Current Opinion in Endocrine and Metabolic Research describes it as separate from the increase that occurs during the second half of the night, and, unusually, as largely independent of circadian control.

The floor matters as much as the ceiling. The same review places the circadian nadir across roughly four to six hours in the late evening and early night. That trough is when most overnight repair happens, and it is the first part of the curve to go when something is genuinely wrong. Clinicians use exactly that: one of the four recommended first line tests for cortisol excess is a late night salivary sample.

What actually raises cortisol during the day?

Waking, acute psychological stress, hard training, large doses of caffeine and short sleep, and only the last two are easy to change.

Acute social stress is the best characterised. A meta-analysis of 76 sub-studies covering 5,171 people used the Trier Social Stress Test, in which people give an unprepared talk in front of an unresponsive panel. The average salivary cortisol response was d = 0.93, with peaks averaging 17.15 nmol/l in men and 10.87 nmol/l in women.

That is a large and reliable rise. It is also normal, and in healthy people it resolves within the hour. A cortisol response to a stressful event is the system working, not failing.

Caffeine is more specific than the internet allows. A randomised crossover trial in Sleep gave 23 healthy men 100 mg or 400 mg of caffeine at 12, 8 and 4 hours before bed. The 100 mg dose produced no significant sleep disruption at any timing. The 400 mg dose, roughly four strong coffees, delayed sleep onset by 14 to 25 minutes and cut deep sleep by 15 to 30 minutes. That is a sleep effect rather than a direct cortisol effect, but short sleep is itself a cortisol input.

Why do you wake up tired after a full night in bed?

Usually because the evening trough never arrived, not because the morning peak was too low.

In a controlled study in Sleep, young men deprived of sleep had cortisol levels 37 and 45 percent higher the following evening, measured between six and eleven at night, with the normal nightly decline delayed. The morning looked ordinary. The evening did not.

Chronic restriction produces the same shape. Sleeping 5.5 hours or less raises late afternoon and early evening cortisol without necessarily changing the 24 hour total. The area under the curve looks unremarkable while the curve has flattened.

Sleep loss is not a niche problem here. In the 2024 national public health survey run by Folkhälsomyndigheten and sent to more than 110,000 adults, around 45 percent reported sleep problems. The response rate was close to 40 percent, down from 61 percent in 2004, which is worth knowing before treating the figure as precise. The agency publishes the underlying sleep statistics.

As for what a flattened curve is associated with, a meta-analysis of 80 studies and 179 separate associations found flatter slopes linked to poorer outcomes in 10 of 12 categories of emotional and physical health, with an overall effect of r = 0.147. The strongest link was with immune and inflammatory markers, r = 0.288. Read that carefully: r = 0.147 is small, most of the work is cross sectional, and association is not causation. The flattened curve may be a readout of poor health rather than a driver of it. More on recovery sits in our wellbeing section.

Can cortisol and stress be seen in the skin?

There is a measurable effect, but on how quickly the skin barrier repairs itself, not on how tired your face looks the next morning.

A study of 60 healthy women in Clinical and Experimental Dermatology compared poor sleepers with good sleepers. Poor sleepers had higher baseline transepidermal water loss. Seventy two hours after the barrier was deliberately disrupted by tape stripping, good sleepers had recovered about 30 percent more. They also scored better on intrinsic ageing measures and recovered faster from ultraviolet induced redness.

The caveats matter. Sixty participants is few, sleep quality was self reported, the sample was ethnically homogeneous, and nothing separates cortisol from everything else that differs between sleeping five hours and sleeping eight.

What it does support is a mechanism people underrate: barrier repair is largely an overnight job. That is a reason to protect sleep, not a reason to buy anything. The biology of the barrier is covered in our article on the biology of radiant skin.

What does the evidence not show about lowering cortisol?

It does not show that measuring your own cortisol is informative, and it does not show that any supplement lowers it in a way you would notice.

Begin with measurement. A 2025 review of the cortisol axis in Pharmacological Reports concludes that cortisol alone may not serve as a reliable biomarker, because what a value means depends heavily on the condition and the person. A home kit sells you a number without the interpretation that makes it mean anything.

Hair cortisol, marketed as a reading of chronic stress, has its own problems. A review in the Journal of Sports Science and Medicine lists them: only weak correlations with perceived stress, substantial elevation from regular vigorous exercise independent of any psychological stress, and low to moderate agreement with saliva, plasma and urine measures.

Then supplements. Ashwagandha is the most studied candidate. A 2025 systematic review in BJPsych Open pooled 15 randomised trials and 873 participants and reported reductions in both perceived stress and measured cortisol at eight weeks. The trials are small, short and heterogeneous, and a change in a laboratory value is not the same as a change you experience. No adaptogen holds an authorised health claim in the EU register of nutrition and health claims, so under Regulation 1924/2006 no product sold in Sweden may tell you it lowers cortisol, whatever an individual trial suggests.

Two further ideas do not survive the evidence. There is no recognised condition called cortisol face, and facial fullness from real glucocorticoid excess is a clinical sign of Cushing's syndrome. And no food may state or imply that it prevents, treats or cures a disease, which is why Swedish labelling rules require every supplement to carry the statement that it is not a substitute for a varied diet.

Health Royals sells no cortisol product. Nothing in our range is formulated for the HPA axis, and we would rather write that here than let the article imply otherwise.

What actually helps an overloaded nervous system?

The measures with the best support are unglamorous: one consistent wake time, daylight early, caffeine kept to the first half of the day, and enough sleep for the evening trough to happen.

  • Keep one wake time, weekends included. The awakening response is anchored to habitual waking, so a wake time that moves by three hours leaves nothing to anchor to.
  • Get outside early. Daylight in the first hour is the strongest signal that the day has started, and in a Swedish winter the scarcest.
  • Move the caffeine rather than removing it. In the trial above, 100 mg carried no measurable cost at any timing. 400 mg needed a full 12 hours.
  • Train, and stop worrying that it raises cortisol. It does, acutely, and that rise is part of the training stimulus rather than a side effect of it.
  • Defend the last two hours of the evening. The trough is the point of the whole curve.

Nutrition is the smallest lever on this list, and it is worth saying so. The Nordic Nutrition Recommendations 2023 set reference values for nutrients, a floor for normal function rather than a strategy for a demanding month.

If you want a daily nutritional base while you work on the parts that move the curve, Health Royals makes Peak Performance, with vitamin D3 at 25 ug, vitamin B12 as MecobalActive, folate as Quatrefolic and 375 mg of omega-3 from OMEGATEX algal oil. Under the EU register, vitamin D contributes to the normal function of the immune system, and vitamin B12 contributes to the reduction of tiredness and fatigue. Those are the claims that may lawfully be made. It is not a cortisol product, and it is not a substitute for sleep. We have written separately about melon SOD, oxidative stress and recovery.

Further reading

Frequently asked questions

How do I know if my cortisol is high?
A single value tells you almost nothing, because the shape of the curve across the day carries the information. Genuine cortisol excess is Cushing's syndrome, an uncommon condition investigated in healthcare with tests that include a late night salivary sample. If you suspect something, ask for an investigation rather than buying a home kit.

Are home saliva or hair cortisol tests worth the money?
In most cases no. Hair cortisol correlates only weakly with perceived stress and is raised by regular vigorous exercise independent of any psychological load, and a 2025 review concludes that cortisol alone may not serve as a reliable biomarker. You get a value without the interpretation that makes it mean anything.

Does ashwagandha lower cortisol?
A systematic review of 15 randomised trials and 873 participants reported lower measured cortisol at eight weeks. The trials are small, short and heterogeneous, and a laboratory value that changes is not the same as a change you experience. No adaptogen holds an authorised health claim in the EU, so no product sold in Sweden may say it lowers cortisol.

Is cortisol face a real thing?
No, there is no recognised condition by that name. Facial fullness from genuine glucocorticoid excess is a clinical sign of Cushing's syndrome, which is diagnosed with laboratory tests. Puffiness after a bad night is something else.

Should I give up coffee because of cortisol?
Dose and timing matter far more than coffee itself. In a randomised crossover trial, 100 mg of caffeine produced no significant sleep disruption at any timing, while 400 mg delayed sleep onset by 14 to 25 minutes and needed a full twelve hours of margin. Move the caffeine to the first half of the day rather than removing it.

Can a supplement help with stress?
Not in the way the marketing implies. Health Royals sells no cortisol product, and the measures with the best support are one consistent wake time, daylight early and enough sleep for the evening trough to happen. A supplement can contribute nutrition, not recovery.

Sources