Sleep, cortisol, and why you're wired at 11pm: the physiology of stress-disrupted sleep

Sleep cortisol: why you're wired at 11pm

If you lie down exhausted but find your brain switching on the moment your head hits the pillow, cortisol is almost certainly involved. Cortisol follows a tight 24-hour rhythm - it should be at its lowest around midnight. When that rhythm breaks down, evening cortisol stays elevated and sleep onset becomes genuinely difficult. One study found that insomnia patients had cortisol levels roughly 20% higher in the evening compared to good sleepers, a difference that correlated directly with time awake in bed.

What the evidence actually shows

The relationship between cortisol and sleep disruption is one of the better-documented areas in sleep science - which doesn't mean everything is settled, but the core mechanism has held up across a range of study designs.

Vgontzas et al. (2001) compared 24-hour cortisol secretion in 11 patients with chronic insomnia against 13 healthy controls. Insomnia patients showed significantly elevated cortisol in the evening and early night - the period when cortisol should be at its nadir. The effect was not trivial: the difference in evening cortisol between groups was statistically significant (p < 0.05) and correlated with objective measures of sleep disruption on polysomnography.

Backhaus et al. (2004) took this further in a study of 40 middle-aged adults, finding that higher cortisol awakening response - the sharp spike in cortisol that occurs in the 30-45 minutes after waking - predicted worse sleep quality the following night. The relationship ran in both directions: poor sleep elevated cortisol, and elevated cortisol disrupted the next night's sleep. That feedback loop is important. It's not just that stress causes bad sleep. Bad sleep is itself a stressor that perpetuates elevated cortisol.

Buckley and Schatzberg (2005) reviewed the HPA axis literature and concluded that hyperactivation of the hypothalamic-pituitary-adrenal axis is a consistent feature of primary insomnia - not just a secondary consequence of lying awake feeling frustrated, but a measurable physiological difference in how the stress response is regulated.

My honest read: the evidence that elevated evening cortisol disrupts sleep onset is solid. What's less clear is the precise causal direction for any given individual, and whether interventions that lower cortisol reliably improve sleep in people without clinical insomnia. The mechanistic story is convincing. The intervention data is patchier.

The biology: what's actually happening at 11pm

Cortisol is produced by the adrenal glands and regulated by the HPA axis - a feedback loop involving the hypothalamus, pituitary gland, and adrenal cortex. Under normal conditions, cortisol peaks sharply in the 30-45 minutes after waking (the cortisol awakening response), then declines steadily throughout the day, reaching its lowest point around midnight to 2am, before beginning to rise again in the early hours ahead of waking.

This rhythm is driven primarily by the suprachiasmatic nucleus (SCN) - the brain's master clock - which synchronises cortisol release with light exposure, feeding patterns, and activity. When the SCN receives appropriate light cues, it suppresses cortisol production in the evening via inhibitory signals to the hypothalamus. When those cues are absent or disrupted - think bright screens at 10pm, eating late, or chronic psychological stress - the SCN's inhibitory signal weakens and cortisol stays elevated later than it should.

Cortisol's direct effect on sleep architecture is significant. It suppresses slow-wave sleep (the deep, restorative stages 3 and 4) and promotes lighter, more fragmented sleep. It also antagonises melatonin: elevated cortisol directly inhibits pineal gland melatonin synthesis, which means that if your cortisol is still high at 11pm, your melatonin production is being actively suppressed at exactly the moment it should be rising.

There's also a temperature component. Core body temperature needs to drop by approximately 1-2�C for sleep onset to occur. Cortisol impairs this thermoregulatory drop, partly by maintaining metabolic rate and partly through its effects on peripheral vasoconstriction. This is one reason why people with elevated evening cortisol often report feeling physically restless or too warm, even when they're mentally exhausted.

The result is a state that many people recognise intuitively: bone-tired but wired. Your body has the subjective fatigue of sleep deprivation, but the neurochemistry of alertness.

Why modern life is particularly good at breaking this rhythm

The HPA axis evolved for a world with very different light patterns, activity rhythms, and threat profiles. Acute physical threats - the kind that require you to run or fight - produce a cortisol spike that resolves quickly once the threat passes. Chronic psychological stress, which is what most people are dealing with, produces a more sustained, lower-amplitude cortisol elevation that the system isn't well-designed to clear efficiently.

A few specific modern factors are worth naming directly.

Artificial light in the evening

Gooley et al. (2011) demonstrated in a controlled study of 116 participants that exposure to room light before bedtime suppressed melatonin onset by approximately 90 minutes and reduced melatonin duration by about 90 minutes compared to dim light conditions. Cortisol and melatonin are inversely related, so anything that delays melatonin onset is likely also delaying the cortisol decline. Screen light is a real factor here - not a minor one.

Late eating

Food intake is a zeitgeber - a time cue - for peripheral circadian clocks in the liver, gut, and adrenal glands. Eating late shifts these peripheral clocks relative to the SCN, creating internal circadian misalignment. Buxton et al. (2012) showed in a circadian misalignment study that even short-term disruption to feeding and sleep timing significantly elevated cortisol and impaired sleep architecture.

Unresolved cognitive load

The prefrontal cortex - responsible for planning, problem-solving, and rumination - has direct excitatory projections to the hypothalamus. When you lie down with an unresolved to-do list running in your head, you are literally stimulating HPA axis activity. This isn't metaphorical. The cognitive activity of worry and planning has measurable effects on cortisol secretion.

The glycine question: what the data says and what it doesn't

Glycine is an amino acid with inhibitory neurotransmitter properties in the central nervous system. It acts on glycine receptors in the spinal cord and brainstem, and there's reasonable evidence that it also influences core body temperature regulation - one of the key physiological prerequisites for sleep onset.

Bannai et al. (2012) conducted a double-blind, crossover RCT in 11 participants with self-reported sleep complaints. Those who took 3g of glycine before bed showed significantly reduced daytime sleepiness the following day (p < 0.05), improved sleep quality scores, and reduced time to sleep onset compared to placebo. The proposed mechanism was a glycine-induced drop in core body temperature via peripheral vasodilation - essentially, glycine helped the body do what it needs to do to fall asleep.

The sample size is small. Eleven people is not a large trial. And I'd be overstating it to claim glycine is a proven sleep intervention on the basis of this evidence alone. What I can say is that the mechanistic rationale is plausible, the human data is directionally consistent, and glycine has an excellent safety profile at the doses studied. Kojo includes 2,000mg of glycine - slightly below the 3g used in the Bannai trial, and I'm transparent about that. Research on glycine and sleep is ongoing, and large-scale human trials are limited.

Taurine and the nervous system: educational context only

Taurine is a sulphur-containing amino acid found in high concentrations in the brain, heart, and skeletal muscle. It acts as a modulator of GABA-A receptors - the same receptor class targeted by benzodiazepines, though via a very different mechanism and with far less potency. There's some interest in whether taurine's GABAergic activity contributes to reduced neurological arousal in the evening.

El Idrissi et al. (2009) showed in animal models that taurine supplementation reduced anxiety-related behaviour and modulated GABA-A receptor expression. The human data on taurine and sleep specifically is thin, and I'd be overstating it to claim taurine meaningfully improves sleep quality on current evidence. What exists is mechanistically interesting but preliminary. Large-scale human trials are limited, and the research is ongoing.

What about the cortisol-vitamin C connection?

This is an area I find genuinely interesting, partly because it's underappreciated.

The adrenal glands have among the highest concentrations of vitamin C of any tissue in the body. Cortisol synthesis is a metabolically intensive process that generates significant oxidative stress, and vitamin C appears to play a role in both adrenocortical function and in the clearance of reactive oxygen species produced during cortisol synthesis.

Brody et al. (2002) conducted an RCT in 45 adults exposed to a public speaking stressor (the Trier Social Stress Test). Participants randomised to 3,000mg vitamin C per day showed significantly attenuated cortisol responses to the stressor compared to placebo (p < 0.05), along with lower blood pressure reactivity and faster cortisol recovery. The effect size was meaningful rather than trivial.

Vitamin C contributes to the protection of cells from oxidative stress, and it contributes to the reduction of tiredness and fatigue - both are authorised claims under the NHCR. Whether the adrenal-support angle translates to meaningfully lower evening cortisol in non-stressed, non-deficient individuals is less clear. But the mechanistic plausibility is there, and 500mg - the dose in Kojo's formula - is a reasonable daily amount. The Brody trial used a much higher dose for acute stress response, so I wouldn't overstate what 500mg does in that specific context.

If you're interested in how vitamin C fits into a broader picture of nutritional support for women specifically, I've written more about this in the context of supplements for women.

Practical interventions with actual evidence behind them

Before reaching for any supplement, it's worth being clear about what the non-pharmacological evidence supports. Because some of it is genuinely strong.

• Light management: Avoiding bright light - particularly blue-spectrum light - in the 2 hours before bed has robust evidence behind it. The Gooley et al. data above is compelling. Blue-light-blocking glasses have mixed evidence, but reducing overall light intensity costs nothing.
• Fixed sleep and wake times: Circadian rhythm stability is one of the most reliable predictors of cortisol rhythm regularity. Irregular sleep timing disrupts the SCN's ability to suppress evening cortisol. This is boring advice. It's also correct.
• Temperature: A cool bedroom (around 16-19�C for most people) supports the core body temperature drop needed for sleep onset. A warm bath or shower 1-2 hours before bed paradoxically helps by accelerating peripheral heat dissipation.
• Cognitive offloading: Writing down tomorrow's tasks before bed - sometimes called a "worry dump" - has been tested in RCT conditions. Scullin et al. (2018) found that writing a to-do list for the following day (as opposed to journalling about completed tasks) significantly reduced time to sleep onset (p < 0.05) in 57 young adults. The effect size was modest but real.
• Alcohol: Worth mentioning because it's so commonly misunderstood. Alcohol reduces sleep onset latency but fragments sleep in the second half of the night and suppresses REM sleep. Ebrahim et al. (2013) reviewed the evidence across multiple doses and found consistent dose-dependent reductions in sleep quality metrics. It's not a sleep aid. It's a sedative with a rebound.

For a deeper look at the physiology underpinning all of this, I'd point you to my longer piece on sleep, cortisol, and why you're wired at 11pm: the physiology of stress-disrupted sleep, which goes further into the HPA axis mechanics.

Frequently asked questions

My honest take

I built Kojo partly because I was frustrated with products that made large claims on thin evidence, and partly because I was personally dealing with exactly the thing this article describes: exhausted by 9pm, somehow alert by 11pm, lying awake running through tomorrow's problems. I know this pattern well.

What I've found - personally, and from reading the literature - is that the behavioural interventions matter more than any supplement. Fixed wake times, managing evening light, not eating at 10pm, writing things down before bed. These are unglamorous. They're also what the evidence consistently supports.

Where I think supplementation has a legitimate supporting role is in filling genuine nutritional gaps and providing ingredients with plausible mechanisms and reasonable safety profiles. Glycine at 2g, vitamin C at 500mg, taurine at 2g - none of these are going to override a lifestyle that's systematically breaking your cortisol rhythm. But if the foundations are in place, they may contribute at the margin. I'm honest that the human data on some of these, particularly glycine and taurine for sleep, is still developing. I wouldn't include them if I thought the evidence was absent, but I also won't claim it's definitive.

The thing I keep coming back to is the feedback loop. Poor sleep elevates cortisol. Elevated cortisol disrupts the next night's sleep. Once you're in it, it can feel intractable. The exit is usually slower than people want - weeks of consistent behaviour, not a single night of doing things right. That's frustrating. It's also just how circadian biology works.