Cortisol high at night: What the evidence shows

Cortisol should be at its lowest between midnight and roughly 3am. When it isn't, sleep suffers — and the research is clear on why. Late-night salivary cortisol above 4.3 nmol/L is used clinically as a diagnostic threshold for excess cortisol production, according to Goyal et al. (2023). This piece covers what drives elevated night-time cortisol, what the biology actually looks like, and what — if anything — you can do about it.

What the evidence actually shows

The first thing worth saying: cortisol high at night is not always a sign of something serious. It exists on a spectrum. At one end, you have physiological Cushing's syndrome — a pathological state of chronic cortisol excess. At the other, you have the much more common experience of a stressed, under-slept person whose cortisol rhythm has simply shifted in the wrong direction. The clinical literature focuses heavily on the former, which means the evidence base for the latter is thinner than I'd like.

That said, the diagnostic tools are useful even for non-clinical contexts. Late-night salivary cortisol is considered one of the most sensitive screening tests for hypercortisolism. Ceccato et al. (2017) describe sensitivity figures of around 92–100% for late-night salivary cortisol in detecting Cushing's syndrome — which tells you something important: the test works because healthy cortisol at night is reliably low. When it isn't, it stands out.

More recently, a cross-sectional study in patients with type 2 diabetes found that a disrupted cortisol circadian rhythm — specifically a flattened morning-to-evening slope — was independently associated with lower muscle mass, suggesting the timing of cortisol matters beyond just sleep quality. Liu et al. (2025) reported this association even after adjusting for age, BMI, and disease duration. The human data on whether this applies to otherwise healthy people with mildly elevated night cortisol is thin, and I'd be overstating it to claim otherwise.

For people without a diagnosed condition, Besedovsky et al. (2012) provide useful context: the normal nocturnal cortisol nadir appears to be functionally tied to sleep architecture, particularly slow-wave sleep. Disrupted sleep and elevated night-time cortisol appear to reinforce each other — it's genuinely bidirectional, which makes it hard to identify a clean starting point. I've written more on this dynamic in my piece on high cortisol levels and sleep.

The biology: what's actually happening at night

Cortisol is secreted by the adrenal cortex in response to adrenocorticotropic hormone (ACTH), which is itself regulated by corticotropin-releasing hormone (CRH) from the hypothalamus. This HPA axis — hypothalamic-pituitary-adrenal — operates on a circadian clock. In a healthy rhythm, cortisol peaks roughly 30–45 minutes after waking (the cortisol awakening response) and then declines across the day, reaching its nadir in the hours around midnight.

The suprachiasmatic nucleus (SCN) in the hypothalamus drives this rhythm. Light is the primary synchroniser. When light exposure patterns are disrupted — late-night screens, shift work, travel across time zones — the SCN signal weakens, and the HPA axis can lose its normal suppression cue at night. Lissak (2019) documented associations between evening screen exposure and altered cortisol patterns in younger populations, though the effect sizes in that literature are modest and the confounders are significant.

There's also a feedback loop worth understanding. Elevated night-time cortisol may suppress melatonin — the two hormones appear to have an inverse relationship, though the mechanistic detail in humans is still being worked out. What's clearer is that cortisol at night keeps the brain in a state of relative alertness. It raises blood glucose, increases heart rate, and maintains a degree of metabolic activation that is directly antagonistic to sleep onset and maintenance. Cryer et al. (1989) established early on that cortisol is a key counter-regulatory hormone in glucose metabolism — which is why night-time cortisol spikes can sometimes feel like waking up wired and slightly anxious at 3am.

In pathological hypercortisolism, the normal feedback inhibition of cortisol on the HPA axis breaks down. Hinojosa-Amaya et al. (2024) highlight in their systematic review that differentiating true Cushing's syndrome from functional hypercortisolism — driven by obesity, depression, or chronic stress — is genuinely difficult, even for endocrinologists. The biology overlaps substantially.

When elevated night cortisol is a clinical concern

I want to be direct here: if you're consistently waking between 2am and 4am, feeling wired, experiencing unexplained weight gain around the abdomen, or noticing skin changes, it's worth talking to a doctor. These can be features of pathological cortisol excess, not just lifestyle stress.

The clinical screening tools for Cushing's syndrome include late-night salivary cortisol, 24-hour urinary free cortisol, and the 1mg overnight dexamethasone suppression test. Nieman et al. (2008) recommend using at least two of these tests in combination, given the risk of false positives with any single measure. Balomenaki et al. (2022) add that the late-night salivary cortisol test is particularly useful because it captures the nadir — the point where cortisol should be lowest — making an elevated result harder to dismiss.

Shiwa et al. (2014) found significantly elevated late-night urinary free cortisol-to-creatinine ratios in patients with subclinical hypercortisolism compared to healthy controls — a finding that suggests even sub-clinical disruption of the nocturnal nadir is detectable and meaningful. For those diagnosed with Cushing's disease specifically, Fleseriu et al. (2023) outline the current management options, which range from surgery to medical therapy depending on the underlying cause.

None of this is meant to alarm you. Most people reading this don't have Cushing's syndrome — it's rare, affecting roughly 10–15 per million people annually. But the clinical literature is the best evidence we have on what disrupted nocturnal cortisol actually looks like, and ignoring it entirely would be dishonest.

Lifestyle factors that may influence night-time cortisol

For the majority of people, elevated night-time cortisol is a functional problem — driven by stress, poor sleep habits, or circadian disruption — rather than a pathological one. The evidence on modifiable factors is real, if not always large in effect size.

Light exposure

The association between evening light exposure and HPA axis dysregulation is plausible and biologically coherent. Lissak (2019) documented that screen use in the two hours before bed may disrupt cortisol rhythm, though the study population was younger and the causal direction isn't fully established. Reducing bright light exposure in the evening is low-risk and consistent with what the biology suggests.

Psychological stress

Chronic psychological stress activates the HPA axis via CRH release from the hypothalamus. The research on acupuncture and anxiety — while not directly measuring cortisol — offers some context on non-pharmacological approaches to stress physiology. Amorim et al. (2022) conducted a double-blinded RCT and found that electroacupuncture may reduce anxiety scores compared to sham, though the sample was small (n=58) and the clinical significance for cortisol specifically remains unclear. I mention it not to oversell it, but because it's the kind of non-supplement intervention that deserves honest acknowledgement.

If you want a deeper look at how chronic stress and disrupted sleep interact at the hormonal level, my piece on high cortisol insomnia goes into more detail.

Sleep architecture itself

As noted above, slow-wave sleep appears to be the phase most associated with cortisol suppression. Anything that fragments sleep — alcohol, caffeine taken too late, irregular sleep timing — may indirectly contribute to a less pronounced nocturnal cortisol nadir. The evidence here is largely observational, and I'd be cautious about overstating the effect sizes. What I can say is that the relationship between insomnia cortisol and sleep quality is well-documented enough to take seriously.

What supplements are being studied — and what the evidence actually supports

I'll be honest: the supplement evidence for night-time cortisol specifically is not strong. There are ingredients with plausible mechanisms and some human data, but nothing I'd describe as definitive. Here's where things actually stand.

Glycine

Glycine is an inhibitory neurotransmitter and a precursor to several important compounds. Some research suggests it may support sleep quality, possibly by lowering core body temperature via peripheral vasodilation. The human data on cortisol specifically is thin, and large-scale RCTs are limited. The KōJō Daily Formula includes 2000mg of crystalline glycine — a dose consistent with the preliminary human studies — but I include it for its broader studied properties, not as a cortisol-specific intervention.

Taurine

Taurine has been studied for its potential role in modulating the stress response, with some animal data suggesting it may interact with GABA receptors. The human evidence is early-stage, large-scale trials are limited, and I wouldn't claim it directly addresses elevated night cortisol on the basis of current data.

Aged Garlic Extract

Aged Garlic Extract has been studied in the context of cardiovascular and oxidative stress markers. Research into its effects on HPA axis activity in humans is ongoing, and the evidence base for any cortisol-specific claim is not yet sufficient to draw firm conclusions.

Olive Leaf Extract, Grape Seed Extract, Pine Bark Extract

These polyphenol-rich extracts are studied primarily for their antioxidant properties. Vitamin C contributes to the protection of cells from oxidative stress — that's a registered claim I can make with confidence. For the polyphenol extracts, the research into stress physiology is ongoing, large-scale human trials are limited, and I won't overstate what the current data supports.

Vitamin C

Vitamin C contributes to the reduction of tiredness and fatigue, and Vitamin C contributes to normal energy-yielding metabolism — both are registered claims. There's also some older research suggesting adrenal tissue has one of the highest concentrations of vitamin C in the body, and that the adrenal glands may release vitamin C alongside cortisol during stress. The mechanistic interest is real. Whether supplemental vitamin C meaningfully influences cortisol output in humans is a different question, and the evidence for that specific claim is not strong enough for me to make it.

Frequently asked questions

My honest take

I started thinking seriously about night-time cortisol after a period of genuinely bad sleep — the kind where you wake at 3am feeling alert in the worst possible way, heart going, mind already running. I assumed it was just stress. It probably was. But researching this properly made me realise how much I'd been glossing over the biology.

The thing that struck me most is how little the supplement industry engages with the actual clinical literature on cortisol. Most of what's sold in this space is built on animal studies, small pilot trials, or proxy outcomes that may have nothing to do with cortisol specifically. I've tried to be straight about that here. The ingredients I include in KōJō have reasons behind them — but I'm not going to tell you that glycine or taurine will fix elevated night-time cortisol, because the human evidence for that specific claim isn't there yet.

What the evidence does support, fairly consistently, is that the timing and pattern of cortisol matters — not just the absolute level. A disrupted circadian rhythm, whether from stress, poor sleep, or light exposure, appears to have downstream effects that go beyond just feeling tired. Liu et al. (2025) on cortisol rhythm and muscle mass, Besedovsky et al. (2012) on sleep and immune function — these aren't fringe findings. The rhythm matters.

If I had one practical suggestion, it would be this: before spending money on supplements, get serious about light exposure in the evening, consistent sleep timing, and — if symptoms persist — a conversation with your GP about whether testing is appropriate. The clinical tools exist. Balomenaki et al. (2022) lay out the diagnostic workup clearly. Most people don't need it. But some do, and a supplement won't help them.

I built KōJō because I wanted a product I could actually justify on the evidence. That means being honest when the evidence is thin. Night-time cortisol is a real phenomenon with real biology behind it. The intervention story is still being written.

This article is for informational purposes only and does not constitute medical advice. Consult your healthcare provider before starting any supplement regimen.