This paper begins with a complaint we could not explain. A person sleeps a full seven or eight hours. Their tracker confirms it. And they still wake up tired — not sleepy, exactly, but unrestored, as if the night had not counted.

If sleep were simply time spent unconscious, that would not happen. So sleep is not simply time. Something else decides whether you wake refreshed — and most sleep products optimise for the wrong thing entirely. They try to keep you asleep longer, when the outcome people actually want is to wake up clear. This paper follows that distinction, from its physiology to its formulation, one step at a time.

We were not looking for another sleep ingredient. We were looking for what everyone selling sleep had missed.

Step one — define the outcomeWaking refreshed is an event, not an absence

We tend to think a good morning is what happens when nothing drags us down. The physiology says the opposite. Waking refreshed is an active, orchestrated event that the body has to be allowed to perform — and three things have to go right.

The dawn cortisol rise. In the thirty to forty-five minutes around waking, cortisol should surge sharply. This is the body's built-in alarm, the signal to get up and go. Anything that suppresses cortisol too far into the morning blunts this rise and produces exactly the groggy, hard-to-start morning we are trying to prevent. This single fact rules out a whole class of popular sleep ingredients.

Waking from the right stage. Wake out of deep slow-wave sleep and you get sleep inertia — that thick, disoriented grogginess that can last an hour. Wake from lighter sleep and you feel refreshed. Good architecture front-loads deep sleep into the first half of the night, so that by morning you are naturally in the lighter stages a clean wake-up needs.

No chemical residue. Anything still circulating at wake time is working against you. A sleep aid with a long half-life keeps sedating you past the point of usefulness. The design rule is strict: everything in the formula should be cleared, or clearing, by the target wake time.

Notice what has already happened. Before choosing a single ingredient, the definition of the outcome has disqualified whole categories of what the industry sells. The goal is not more sleep. It is a clean landing.

Step two — the 3 AM wake-upWhat actually pulls you awake in the back half of the night

The most common complaint we hear is not "I can't fall asleep." It is "I fall asleep fine, then I'm wide awake at 3 or 4 AM." There are several real causes — circadian misalignment, alcohol, late caffeine, sleep apnoea among them — and any persistent, distressing pattern deserves proper medical evaluation rather than a supplement. But underneath the common cases sits one physiological driver that the sleep-supplement category almost never addresses: cortisol in the wrong place at the wrong time.

Cortisol is meant to be at its lowest in the evening and early night, so that deep sleep can proceed, then rise gently before dawn to prepare you to wake. Under chronic stress that rhythm distorts. Insomnia is increasingly understood as a disorder of hyperarousal rather than of sleep loss — associated with an overall increase in cortisol secretion that tracks the degree of objective sleep disturbance, with elevated nighttime cortisol proportional to the amount of wakefulness across the night. In practical terms, the pre-dawn cortisol rise arrives too early and too strong, crosses the internal arousal threshold ahead of schedule, and forces a full awakening at 3 or 4 AM instead of 7.

Worse, it feeds on itself: a bad night raises the next morning's arousal, which worsens the next night — a measurable, self-reinforcing loop. That single detail matters enormously for what follows, because a loop that builds over days is not fixed by a molecule that acts for one night. It is fixed by lowering the drive itself, over weeks.

There is also a self-inflicted version worth naming, because it is so common. Standard over-the-counter melatonin (3–10 mg) is immediate-release: it spikes far above any natural level within forty minutes, then clears within hours. The brain can read that sudden disappearance of the melatonin signal as a "morning" cue and trigger alertness — and taken at bedtime, that crash lands around 3–4 AM. The product sold to keep you asleep can be the very thing waking you.

A fair caveat: the cortisol-hyperarousal picture is well supported but not unanimous — some studies of insomniacs without objective sleep disturbance find no elevation. We therefore build for the stressed, sleep-disturbed person this describes, not as if it were a universal law.

Step three — the blind spotThe category treats onset, and ignores the cortisol drive

Once nocturnal cortisol is the lens, the gap in every premium sleep formula becomes plain. Consider the actives a good one relies on: glycine, L-theanine, apigenin, magnesium. They are, almost without exception, onset-and-calm ingredients — they help you fall asleep and quiet the pre-sleep mind. Glycine's own mechanism sheds core body heat to initiate sleep; theanine and apigenin damp arousal at the front of the night. All valuable. But not one of them lowers the elevated nighttime cortisol that actually drives the 3 AM wake-up. The category sells onset-and-calm and labels it maintenance. Those are different jobs, and almost nobody is doing the second one.

The right question is not "which ingredient calms you best at bedtime." It is "what lowers the cortisol drive that wakes you at 3 AM."

Here is the part that is almost ironic. There is a well-studied natural ingredient whose primary, best-evidenced action is lowering cortisol — ashwagandha. And the sleep-supplement world has mostly mishandled it, in one of two ways. Some ignore it, filing it under "stress, not sleep." Others reach for the wrong extract at the wrong dose and time, in a way that risks flattening the healthy morning cortisol rise along with the harmful nighttime one. The opportunity was never another calming herb. It was to use the cortisol-lowering ingredient correctly — the right extract, the right dose, timed to the night.

Step four — what we build insteadOne hero on the cortisol drive, one cofactor on the early night

If the problem is a cortisol drive late and architecture early, the answer has a clear hierarchy — not two co-equal stars. Shoden is the hero: it carries the maintenance objective the category misses, and on its own evidence it improves the back-half endpoints and morning alertness. Glycine is the cofactor that completes the early-night architecture the hero doesn't act on. Two clocks, yes — but one hero and one supporting role, each checked against human evidence rather than mechanism alone.

Step five — now the formulaWhat earns a place — and at exactly what dose

Only now does the formula make sense, because every choice is downstream of a step already walked. Two principles govern it. First, every dose is the dose the trials actually used — not a higher number for label appeal, not a lower one for margin. The body responds to a studied dose; overshoot and effects can reverse, undershoot and they vanish. Second, ingredients are tiered by the strength of their human evidence, and labelled honestly where that evidence is thin. Glycine at a real 3 g is the line most products won't cross; Shoden at its studied 120 mg is the pillar most miss entirely.

Read top to bottom, the table is not a shopping list — it is the argument made concrete. One hero and one cofactor on two clocks: Shoden carries the cortisol-drive and maintenance objective across weeks, glycine completes the architecture of a single night. Three distinct calming pathways that do not overlap — theanine (alpha-wave), saffron (serotonergic), apigenin (GABA-A). And an honest line drawn at magnesium and apigenin, where the evidence is supporting rather than headline.

Step five, continued — the disciplineWhat we refused, and why the refusals matter

A formula is defined as much by what it refuses to include. Each exclusion below is a decision made for a physiological reason, not a cost one — and each traces back to a step above.

Step six — how it works across a nightTwo clocks, one clean morning

A person taking this does not experience a table of ingredients. They experience a night — one shaped by two mechanisms running on two different clocks. Follow it, and the design becomes concrete.

The slow clock, running underneath. By the time any single night arrives, the person has been taking Shoden each evening for weeks, and their elevated nighttime cortisol has been steadily coming down. This is the part no single-night ingredient can do. The drive that would have pulled them awake at 3 AM has already been turned down at the source. This is why the promise is about the thirtieth morning, not the first.

At bedtime, the fast clock takes over. Glycine sheds core body heat and eases the person into sleep, while theanine and saffron quiet arousal along two separate pathways — alpha-wave and serotonergic. They fall asleep cleanly.

Through the first half, glycine does its quieter, more important work — front-loading deep slow-wave sleep into the early hours, so that by morning the person is naturally in the lighter stages a clean wake needs.

Through the back half — the 2 to 4 AM window. There is no cortisol surge crossing the arousal threshold early, because weeks of Shoden have lowered it. There is no melatonin crash to jolt them, because there is no melatonin in the formula. The most vulnerable hours pass without the thing that usually breaks them.

At dawn, the design does the part that matters most for waking fresh: it stays out of the way. Glycine has cleared, nothing sedating lingers — and, critically, Shoden has lowered the nighttime cortisol drive without flattening the morning rise. The person surfaces from light sleep into a clean, intact cortisol surge, onto the seven or eight hours a working life allows — and wakes refreshed.

One ingredient works on the night. One works on the weeks behind it. Neither sedates you harder — together they remove the reason you were waking at all.

Step six, resolved — the delivery questionWhy a single bedtime dose is the right answer, not a clever one

It is tempting to reach for timed-release engineering here — a layer that drip-feeds an active across the 2–4 AM window. We considered it seriously, and concluded it would be solving a problem this formula does not have. Look at how the two pillars actually work.

Glycine does not need to still be circulating at 3 AM to have done its job. Its role is to set sleep architecture early — front-load the deep sleep. That work is finished in the first half of the night by design. Sustaining it late would fight its own mechanism, and a 3 g load is the hardest thing in the formula to controlled-release anyway.

Shoden does not act on a nightly release curve at all. It is an HPA-axis modulator whose benefit accrues over weeks of daily dosing — its trial effect shows at six weeks, not on night one. Taking 120 mg in the evening and letting it accumulate systemically is exactly how it was studied. A timed-release version would change nothing about how it works.

Step six, proven — the study we runHow we settle it with human evidence

A claim this specific has to be tested on the endpoints it names — including the one honest risk. So the validation study is designed, not deferred:

• Design. Randomised, double-blind, placebo-controlled, in adults on a normal schedule who fall asleep adequately but wake in the back half of the night — the exact person from Step two.the right population, not general insomniacs
• Co-primary endpoints. Objective wake-after-sleep-onset, and next-morning psychomotor vigilance. One number for the night held together, one for the morning earned — the two the category avoids.WASO + PVT, measured objectively
• The safety endpoint that is really the whole thesis. Salivary cortisol awakening response, measured to confirm the healthy morning rise is preserved while nighttime cortisol falls. This is how the one genuine worry about a cortisol-lowering ingredient becomes a tested claim rather than an assumption.preserved morning CAR — proven, not asserted
• Duration. Thirty nights — because Shoden's mechanism builds over weeks, and the promise is about the thirtieth morning, not the first.30 nights, matched to the mechanism
• The bar. Beat placebo on objective measures, not just questionnaires. This category is littered with trials where placebo won on actigraphy; we design against that failure from the start.objective superiority or nothing

The conclusionWhy 30 nights of this wakes you better than anything else on the shelf

Read the argument in sequence and the answer is straightforward. The question was why a full night can still leave you tired. The physiology of waking refreshed set the rules. The real complaint pointed to the back half of the night — and the driver turned out to be nighttime cortisol, an arousal problem, not a sedation one. The category treats onset and misses the cortisol drive entirely, and fumbles the one ingredient that addresses it. One hero and one cofactor answer it on two clocks: Shoden supports the nighttime cortisol rhythm across weeks and carries the maintenance objective, while glycine sets the early-night architecture — with three distinct calming pathways beneath them and every dose matched to the trials.

The moat is the mechanism and the proof, not a secret ingredient. A competitor can match a label overnight. They cannot easily match a two-clock design validated on the endpoints they never measure — objective WASO, next-morning performance, and a preserved cortisol awakening response — over the thirty nights the mechanism actually needs.

And where a supplement is not the answer — anxiety, apnoea, a displaced body clock — we will say so. That honesty is the product.