Walk into any pharmacy and the melatonin section will have bottles ranging from 1mg to 10mg, with 5mg being roughly the standard. Walk into a health food store and you’ll find 10mg tablets marketed as premium sleep support. Most people who try melatonin and find it works start there, or go higher when it seems to stop working, or take it every night as part of their wind-down ritual.
All of this is backwards. Your body naturally produces somewhere between 0.1 and 0.3 milligrams of melatonin at the peak of its nightly release. Standard over-the-counter doses are 10 to 50 times what your body makes on its own. Research shows no meaningful improvement in sleep quality above about 1mg, and higher doses consistently increase next-day grogginess and vivid, disruptive dreams. The mass market for melatonin has essentially trained people to massively overdose a hormone with a ceiling effect they hit somewhere around 0.5mg.
This isn’t a fringe position. It’s what the research says. And understanding why melatonin is the wrong tool for most sleep problems opens the door to compounds that work better for what most people are actually dealing with.
The melatonin misunderstanding
Melatonin is a timing hormone, not a sedative. It doesn’t knock you out. It signals to your brain that it’s dark out, which nudges your circadian rhythm toward sleep. That’s a useful thing to do when you’re jet-lagged, shifting time zones, or trying to advance your sleep phase. It’s less useful for people whose core problem is an anxious mind at 10pm, difficulty staying asleep, or poor sleep quality — all of which are about the nervous system’s state at bedtime, not about the timing signal itself.
Clinical trials on melatonin for primary insomnia are not impressive. A systematic review found that melatonin shortened sleep onset by an average of four minutes and increased total sleep time by about thirteen minutes. That’s real but modest, and it’s the population-level average including people whose problem was actually a timing mismatch. For people with a general sleep quality or anxiety-related problem, the effect is even smaller.
The dose problem makes this worse. In studies where dose response was examined, effects on blood serum levels, circadian phase-shifting, and core body temperature were essentially identical for doses above 1mg. You get no additional benefit from 5mg over 1mg, but you do get more next-day impairment. The half-life of melatonin is less than an hour for fast-release formulas, which means even a large dose clears quickly — but it can spike blood levels high enough to produce residual effects that bleed into morning alertness.
The right use case for melatonin: 0.5–1mg of fast-release, taken 30–60 minutes before the desired sleep time when you’re shifting your schedule or recovering from travel. Not every night. Not 5mg. Not as a substitute for fixing the conditions that are actually disrupting your sleep.
Glycine: the quiet compound with real polysomnographic data
Glycine is the smallest amino acid, found in high concentrations in collagen and connective tissue. It’s also an inhibitory neurotransmitter in the central nervous system, and it has a specific effect on sleep that’s been verified with polysomnographic monitoring — the gold standard of sleep measurement that records actual brain waves, eye movements, and muscle activity throughout the night.
The mechanism: glycine activates NMDA receptors in the suprachiasmatic nucleus, the brain’s master clock. This activation produces peripheral vasodilation — blood vessels in the extremities widen, releasing body heat, which causes core temperature to drop. Core body temperature dropping is one of the primary physiological signals that initiates deep sleep. Lower core temp means faster sleep onset, more time in slow-wave sleep, and higher sleep efficiency.
A 2007 study published in Sleep and Biological Rhythms gave participants 3g of glycine before bed and measured polysomnographic sleep on the third night. Compared to placebo: significant shortening of sleep onset latency, improved sleep efficiency, reduced time in light non-REM sleep, and less wakefulness. Crucially, the participants also reported less daytime fatigue and better cognitive performance the following morning. A follow-up study in sleep-restricted adults found the same pattern — glycine didn’t just improve the number; it improved how people felt and performed the next day.
A 2024 systematic review in Nutrients covering the available human data concluded that glycine consistently improved subjective sleep quality and reduced fatigue in people with insufficient sleep. There’s also evidence that collagen peptide supplementation (which is rich in glycine) before bed improves sleep fragmentation in athletes — the mechanism is the same glycine content doing the same temperature regulation work.
The dose in essentially every study: 3 grams, taken 30–60 minutes before bed. Glycine is extremely safe, essentially non-toxic even at much higher doses, cheap, available without a prescription, and tasteless enough to dissolve in water. It’s one of the more evidence-backed sleep interventions that almost nobody outside the biohacking community knows about.
Apigenin: chamomile’s active compound, but not at chamomile tea doses
Apigenin is a flavonoid found in chamomile, parsley, celery, and various other plants. Chamomile tea has a folk tradition as a sleep aid going back centuries. The active reason is apigenin, which is a partial agonist at GABA-A receptors — the same receptor type targeted by benzodiazepines and many sleep medications, though with dramatically lower binding affinity and without the dependence risk.
The GABA-A mechanism reduces neuronal excitability across the brain, producing a calming, anxiolytic effect that lowers the arousal threshold for sleep. If the obstacle to sleep is a nervous system that’s still running hot at bedtime — racing thoughts, background anxiety, difficulty disengaging — apigenin addresses that underlying state rather than forcing sedation.
Andrew Huberman includes 50mg of apigenin in his nightly sleep stack, and the research basis for that dose is legitimate. The key distinction between apigenin supplementation and chamomile tea is concentration: a cup of chamomile tea contains roughly 0.5–1mg of apigenin. The 50mg supplement dose is 50–100x higher than what you’d get from tea. The folk tradition was pointing at the right compound; the delivery mechanism was just too dilute.
One flag worth mentioning: apigenin is a mild aromatase inhibitor, meaning it mildly reduces the enzyme that converts testosterone to estrogen. At 50mg, Huberman has noted this is unlikely to be clinically meaningful, and the research doesn’t show significant hormonal effects at supplemental doses. It’s still something to be aware of for women who need their estrogen levels stable, and it’s worth not dramatically exceeding the studied dose on the theory that more is better.
Magnesium: the anchor of the stack
Magnesium doesn’t quite belong in the “underknown” category anymore — it’s become a standard sleep supplement recommendation across the biohacking community. But the form matters more than most coverage acknowledges, and the mechanism is worth understanding.
Magnesium activates GABA receptors and regulates the HPA axis stress response, which is why it produces a calming effect and improves sleep quality in people who are deficient or insufficient. Crucially, up to 79% of American adults don’t meet the recommended daily allowance for magnesium through diet alone, which means the majority of people taking magnesium for sleep are correcting a genuine gap rather than just adding something extra on top of adequate levels.
The forms matter significantly. Magnesium oxide, the most common and cheapest form, has about 4% bioavailability — most of it passes through unabsorbed and causes GI distress. Magnesium glycinate (bisglycinate) has substantially higher bioavailability, crosses into the central nervous system more effectively, and is gentler on digestion. Magnesium L-threonate, used by Huberman and Peter Attia, was specifically developed to cross the blood-brain barrier and is most studied for cognitive and sleep effects. Both glycinate and L-threonate are legitimate; L-threonate is more expensive.
The dose that appears consistently in sleep studies is 200–400mg of elemental magnesium, taken 30–60 minutes before bed.
What the Huberman stack actually is
For completeness: the sleep supplement protocol Andrew Huberman has described publicly consists of magnesium glycinate (200mg) or magnesium L-threonate (145mg), apigenin (50mg), and L-theanine (100–400mg), all taken 30–60 minutes before bed. On nights where he needs additional support, he adds 3–4g of glycine and 100mg of GABA. He explicitly recommends against making melatonin a nightly staple, suggesting it be reserved for travel and schedule adjustments at doses of 0.1–0.3mg.
This stack works through complementary rather than redundant mechanisms. Magnesium addresses the mineral deficiency and GABA receptor modulation. Apigenin further modulates GABA-mediated neuronal calm. L-theanine (from green tea, also available standalone) promotes alpha wave activity in the brain associated with relaxed alertness without sedation. Glycine handles core body temperature regulation. None of these compete or interfere with each other, which is why the combination is more reliable than any single compound.
What this stack doesn’t fix
Sleep supplements of any kind are downstream of the behaviors that actually govern sleep quality. The stack above will improve sleep at the margins. It won’t compensate for: inconsistent sleep timing that keeps your circadian rhythm unstable; screens within the hour before bed that suppress melatonin and raise cortisol; caffeine consumed after noon in people who metabolize it slowly; a room that’s too warm (68°F or below is consistently associated with better sleep architecture); significant chronic stress that keeps cortisol elevated into the evening.
These aren’t minor factors. The effect size of going to bed at the same time every night on sleep quality is larger than any supplement. The effect of cutting off screens 60 minutes before bed is larger than the whole stack above combined.
Supplements work best as adjuncts to good sleep hygiene, not substitutes for it. Which is also exactly why they’re worth understanding properly — because when the conditions are already reasonable and sleep is still difficult, glycine, apigenin, and low-dose magnesium address real neurobiological obstacles in ways that high-dose melatonin simply doesn’t.
Sources:
- Yamadera, W. et al. (2007). Glycine ingestion improves subjective sleep quality in individuals with insufficient sleep. Sleep and Biological Rhythms, 5, 126–131.
- Kawai, N. et al. (2015). The sleep-promoting and hypothermic effects of glycine are mediated by NMDA receptors in the suprachiasmatic nucleus. Neuropsychopharmacology, 40(6), 1405–1416. PMC4397399.
- Inose, H. et al. (2024). Effects of glycine ingestion on human sleep and health: a systematic review. Nutrients, 16(2), 319.
- Bannai, M. et al. (2012). The effects of glycine on subjective daytime performance in partially sleep-restricted healthy volunteers. Frontiers in Neurology, 3, 61.
- Shaw, D.M. et al. (2024). Collagen peptide supplementation before bedtime reduces sleep fragmentation. PMC10799148.
- Zick, S.M. et al. (2011). Preliminary examination of the efficacy and safety of a standardized chamomile extract for chronic primary insomnia. BMC Complementary Medicine and Therapies.
Leave a Reply