Cortisol and sleep: how each affects the other

Cortisol and sleep run on the same daily clock, and each affects the other. In healthy people cortisol falls to its lowest point around midnight, starts rising again in the early hours and peaks within about three hours of waking, and falling asleep itself damps the stress system down.¹ Losing sleep pushes cortisol up: after a short night or a missed night, cortisol the next evening was 37 and 45 per cent higher.² Over months, poorer sleep has been linked with higher long-term cortisol measured in hair,³ and in one trial treating insomnia was followed by a fall in hair cortisol.⁴ Shift work and jet lag, which pull sleep away from the body clock, disturb the whole pattern.¹

The nightly low and the morning rise

Cortisol is released in pulses every one to two hours, on top of a daily rhythm.¹ In a group of healthy adults, blood cortisol peaked at about 8.30 in the morning and reached its low point, below 2 µg/dL (about 55 nmol/L), at around a quarter past midnight. It stayed below its daily average from about 7.45 in the evening until 5.30 the next morning.⁵ Falling asleep, and deep (slow-wave) sleep in particular, suppresses the hypothalamic-pituitary-adrenal (HPA) axis, the chain of signals from the brain to the adrenal glands that controls how much cortisol is made.¹

Waking sets off a second, sharper event, the cortisol awakening response: in healthy adults, salivary cortisol rises by 50 to 160 per cent in the first 30 minutes after waking.⁶ The rise lasts 30 to 45 minutes and depends on the exact moment you wake, which is why expert guidelines stress exact timing of samples from waking.⁷ With age, the gap between the daily high and low narrows, the early-morning rise comes earlier and the night-time low rises.¹ There is more on the rhythm in the daily cortisol rhythm.

What a bad night does to cortisol

In a classic laboratory study, healthy adults either had a short night or went without sleep altogether. The following evening, between 6 and 11 pm, their cortisol was 37 and 45 per cent higher respectively than after a normal night, and the evening quiet period, when cortisol is normally at its lowest, began at least an hour later.² The cost of lost sleep shows up the next day, at the very time cortisol should be winding down.

That matters in two ways. If you sleep badly night after night, those repeated evening rises may add to your overall cortisol exposure. And if you are having a blood or saliva cortisol test, a bad night beforehand is one more thing that can move a single result, so it is worth mentioning to whoever arranges the test.

Shift work and jet lag

Sudden changes to the sleep-wake cycle, such as night shifts or crossing time zones, activate the stress system and change the usual daily pattern of cortisol.¹ Hair studies suggest the effect can last. In a study of 33 shift workers and 89 day workers, the shift workers had higher hair cortisol, 47.32 against 29.72 pg/mg. The difference was found only in younger workers (48.53 against 26.42 pg/mg), and the younger shift workers also had a higher body mass index.⁸ It was one study, so it shows a link rather than proof that shift work is the cause.

Shift work also makes the usual cortisol tests harder to use. Late-night saliva tests assume a regular night-time sleep and are not suitable for night-shift workers or people with variable bedtimes,⁹¹⁰ and NICE advises getting specialist advice before testing shift workers, because it is not clear when to test and the results can be hard to interpret.¹¹ Hair cortisol, by contrast, does not vary with the time of day the sample is taken.¹² See cortisol and shift work.

Long-term cortisol and sleep quality

Most sleep research measures cortisol in blood or saliva over a night or a day. Hair adds a longer view. In a pilot study of 145 women aged 50 to 64 who were going through or past the menopause, higher hair cortisol, and higher hair glucocorticoids overall, went with worse scores on the Pittsburgh Sleep Quality Index, a standard questionnaire about the past month's sleep. Taking longer to fall asleep was also linked with higher hair cortisol.³

That study was small and took a single snapshot, so it cannot say which came first. Poor sleep may raise cortisol, higher cortisol may disturb sleep, or something else, such as ongoing stress or night sweats, may drive both. See cortisol and the menopause for more on this stage of life.

Can better sleep lower cortisol?

The most direct evidence comes from a randomised trial in 74 shift-working nurses with shift work sleep disorder, a form of insomnia linked to working shifts. A digital course of cognitive behavioural therapy for insomnia (CBT-I, a structured programme that changes sleep habits and unhelpful thoughts about sleep) improved their insomnia markedly, a large effect (d = 1.11) by the end of the programme. Hair cortisol in the treated group fell by 1.8 pg/mg, or 44 per cent from where it started.⁴ That fall was measured within the treated group rather than against the control group, so it needs confirming in further trials.

More broadly, a review of 14 studies with 1,916 participants that used hair cortisol to measure the effect of psychological treatments found mixed results.¹³ So better sleep is a reasonable route to lower cortisol, with real but limited evidence behind it. The Endocrine Society notes that a regular sleep routine, exercise and a healthy diet help whatever the cause of tiredness.¹⁴ For other approaches, see how to lower cortisol.

Measuring cortisol when sleep is the question

A single blood or saliva test describes one moment, and that moment is shaped by when you woke, how you slept and what happened that morning. Even the waking rise is largely situational on any one day: researchers need two to six days of samples to measure it reliably.¹⁵

If your question is whether months of poor sleep or shift work have raised your cortisol, hair answers it more directly. It does not vary with the time of day,¹² and it is far more consistent from test to test: the same people's hair cortisol correlated at r = 0.84 from one month to the next, against 0.38 to 0.61 from week to week for saliva.¹⁶ Because hair grows about 1 cm a month, the 3 cm nearest the scalp cover roughly the last three months.¹⁷ That makes hair the most accurate way to measure long-term cortisol from a single sample. It cannot diagnose a sleep disorder or an adrenal condition.

When to see a doctor

See a GP if poor sleep goes on for weeks, or if you are very sleepy during the day despite enough time in bed. Tiredness and unrefreshing sleep can come from conditions such as obstructive sleep apnoea or depression, which need their own assessment.¹⁴ Extreme tiredness with dizziness on standing, weight loss, salt craving or darker patches of skin can point to Addison's disease,¹⁸ while a red, round face, easy bruising, stretch marks and weak upper arms and thighs can point to Cushing's syndrome.¹⁹ Both are diagnosed with blood, urine or saliva tests arranged by a doctor, not with a hair test. See high cortisol symptoms and low cortisol symptoms.