Full mechanism, in text

Why small doses are enough

1 · Endogenous rhythm — dark triggers

As light fades the releases its inhibition of the pineal gland, which converts serotonin into melatonin and tells every cell that it is night.The pineal gland, in the back of the brain, is the body's night clock. Its activity is controlled by the suprachiasmatic nucleus (, the central clock in the hypothalamus):

· Daytime: retinal ipRGCs detect blue light (~ 480 nm) → signal via optic nerve → SCN → inhibits the pineal → melatonin ~ 0
· Dusk, light falls: SCN releases the inhibition → the pineal turns on N-acetyltransferase (AANAT) → converts serotonin () into melatonin
· Night (about 2 h after DLMO, dim-light melatonin onset): plasma melatonin rises from < 2 pg/mL to its nocturnal peak — how high that peak runs varies widely between people (roughly tens to a couple of hundred pg/mL), which is why the next step does not use it as the denominator for "how many times over" → tells every cell: "it's night"
· 2-4 am: peak
· Morning light → SCN re-inhibits → melatonin falls back to < 2 pg/mL.

Key: melatonin is not a sleeping pill, it is the "biological-night signal":
· It does not directly "flip the off switch" to put you to sleep
· It tells the body "it is night" + synchronises circadian rhythm + slightly advances sleep onset
· Actual sleep is driven by adenosine pressure + pathways + falling core temperature.

Receptors:
· MT1 — mainly on the SCN itself → feedback regulation + suppresses neuronal firing
· MT2 — mediates circadian phase shifting
· Both are GPCRs, in the same family as the A1 / A2A receptors that caffeine blocks.

2 · 0.1-0.3 mg already saturates

This step rests on two papers, and they measured different things.Dollins 1994 (PNAS) — the four-dose design:

· N = 20 healthy men
· Compared 0.1 / 0.3 / 1.0 / 10 mg taken at noon vs placebo
· Result: from 0.1 mg upward, daytime serum melatonin was raised into the normal nocturnal range and sleep latency shortened; larger doses brought no dose-response advantage.

Zhdanova 1995 (Clinical Pharmacology & Therapeutics) — tested only two doses:

· 0.3 mg and 1.0 mg, given at 6, 8 or 9 pm vs placebo
· Both shortened sleep onset
· This paper reports that neither dose induced hangover effects — no residue was detectable on next-morning mood and performance testing.

Feeling as if you never woke up shows up at larger doses: Zhdanova 2001 (JCEM) compared 0.1 / 0.3 / 3.0 mg in people over 50, and the 3.0 mg arm left plasma melatonin elevated into the daylight hours and caused hypothermia, while 0.3 mg already restored sleep efficiency in the poor sleepers.

Core finding: melatonin receptors are essentially saturated at very low doses — around 0.1 mg is already enough to put plasma back in the nocturnal range. More dose is wasted and only prolongs the side-effect window.

To quantify the inversion, use dose as the denominator, not plasma concentration. The physiological nocturnal peak varies too much between individuals to divide by; the 0.1 mg ≈ nocturnal range anchor is firm:

· 1 mg ≈ 10 × that dose
· 3 mg ≈ 30 ×
· 5-10 mg ≈ 50-100 × — and in the Dollins trial the larger doses brought no dose-response advantage.

Dose-response curve is "steep early, flat later" (a classic saturation curve) — common to all receptor drugs, but extreme in melatonin because of its signalling nature.

Brzezinski 2005 (Sleep Medicine Reviews) across 17 / 284 subjects reproduces: sleep onset latency −4.0 min (95% 2.5-5.4); total sleep duration +12.8 min (2.9-22.8); sleep efficiency +2.2% (0.2-4.2) — a small but consistent real effect.

3 · Market dose 5-10 mg — backwards

Common melatonin doses on North American shelves: 3 / 5 / 10 mg, with some extra-strength products higher still.This is one of the most dramatic dose inversions in the supplement industry:

· Putting plasma back into the physiological night-time range takes about 0.1 mg (the anchor from the previous step, Dollins 1994)
· 0.3 mg: 3 × that dose
· 3 mg: 30 ×
· 5 mg: 50 ×
· 10 mg: 100 × — and in the Dollins trial the larger doses brought no dose-response advantage.

Why did marketing become bigger is better?
· Mega-doses have no immediate, visible side effect (unlike sleeping pills)
· The intuition that dose = effect (it fails for a signalling molecule that works at tiny doses)
· Industry margins = big dose × big bottle × marketing
· The US classes melatonin as a dietary supplement with no dose regulation; many EU countries, the UK and Australia mainly treat it as a prescription medicine (some also allow low-dose products as food supplements).

Real consequences of high doses:

· Next-day grogginess: in Zhdanova 2001 the 3.0 mg arm left plasma melatonin elevated into the daylight hours and caused hypothermia
· Vivid dreams, brief mood swings: reported
· Reproductive hormone axis: melatonin is linked to it (mostly animal research), so some worry that high doses affect puberty or periods; human data are scarce, so this is a concern, not a proven harm
· Children: the CDC 2022 MMWR counted poison-control reports of pediatric melatonin ingestions; the annual number rose 530% from 2012 to 2021, and in the same report five children needed mechanical ventilation and two died.

Safer use:

· Starting dose: 0.3-0.5 mg, 1 mg if needed, rarely more than 3 mg — a principle drawn from the dose trials above; the AASM 2015 guideline itself gives no dose
· Timing: in most cases 30-60 min before sleep
· Insomnia: AASM 2015 covers only intrinsic circadian rhythm disorders; for ordinary chronic insomnia the first choice is , and melatonin is at most a short trial
· Jet lag: the Cochrane review (Herxheimer 2002) pooled 10 trials, and 9 found that melatonin taken close to bedtime at the destination (about 10 pm to midnight) reduced jet lag after crossing 5 or more time zones, 2; 0.5-5 mg worked about equally, 5 mg gave faster sleep, and more than 5 mg was no better. It warns that taking it too early in the day causes sleepiness and slows adaptation — so do not move it up to dusk
· Delayed sleep phase / shift changes (moving the clock earlier): a small dose about 4-6 h before the target sleep time, based on the Lewy 1998 phase-response curve; the exact time is best set by a doctor
· Long term: review whether you still need it after 2-4 weeks and again at 3 months; there are few trial data on nightly use beyond 6 months (a cautious practice, not a trial-derived number).

Most common mistake: seeing a 5 mg label, swallowing it, feeling groggy the next morning and calling it a natural reaction — more likely it was too much.

4 · Label chaos + decision tree

Store-bought melatonin often does not contain what the label says and some products hold unlabelled serotonin, so choose third-party certified, simple formulations.Erland 2017 (J Clin Sleep Med) sampled 31 North American over-the-counter melatonin products:

· Measured content vs label: deviation range -83% to +478%; more than 71% of products fell outside ±10% of the label
· Eight contained unlabelled serotonin, at 1 to 75 μg
· Batch-to-batch variability within one product reached 465%.

Read that μg figure. These are microgram amounts, and oral serotonin is poorly absorbed and does not cross the blood-brain barrier. So the paper's point is not that these products cause serotonin syndrome; it is that the bottle contains an active molecule that appears nowhere on the label. For someone on an or , the real problem is not knowing what you swallowed.

Gummies were measured by a different study: Cohen 2023 in JAMA analysed 25 US melatonin gummy products — one contained no melatonin at all, and the rest ranged from 74% to 347% of label.

Why: in the US melatonin is regulated as a dietary supplement, which does not have to prove content or purity before sale, and a listing is not a batch-by-batch assay; many EU countries and the UK mainly treat it as a prescription medicine made to pharmaceutical standards.

Quality choices (if you decide to use melatonin):

· Look for USP / NSF / ConsumerLab certification — independently assayed
· Avoid sleep-formula blends: they usually add , , valerian and more → you cannot tell which ingredient is acting, and the risks stack
· Choose simple formulations: pure melatonin, a clearly labelled dose, no serotonin / 5-HTP
· Tablets / capsules are more accurate than gummies.

Decision tree:

Q1: what is my sleep problem?

· Trouble falling asleep: melatonin 0.3-0.5 mg, 30-60 min before bed — but first try (cognitive behavioural therapy for insomnia), the first choice for chronic insomnia
· Waking in the night or too early: ordinary immediate-release melatonin helps little; it does not stay in the blood long enough
· Jet lag: the best short-term use — 0.5-5 mg taken close to local bedtime (about 10 pm to midnight) after arrival, not moved up to dusk
· Shift work / delayed sleep phase (common in teenagers): a low dose taken early (about 4-6 h before target sleep) to move the clock earlier
· Total blindness (non-24): prescription dose under a doctor's supervision.

Q2: I have been taking 5-10 mg every night for months — what now?

· Taper gradually rather than stopping and then climbing back up (a cautious practice; no trial sets a fixed schedule)
· If sleep improves → stop
· Let CBT-I take over (a 6-8 week course, backed by several randomized trials).

Q3: children / adolescents?

· First choice: routine, less bright screen time in the evening, a fixed schedule
· If needed: short-term use under a doctor, who sets the dose
· Avoid: high doses and sugary gummies (pediatric ingestion reports have surged, mostly accidental ingestions in children aged 5 or under).

Q4: pregnancy / breastfeeding?

· Not recommended — safety data are limited, and the body's own melatonin has signalling roles in pregnancy.

Melatonin is a time signal, not a sleeping pill. About 0.1 mg puts plasma back in the night-time range; bigger is better is marketing, and a high dose plus next-day grogginess most likely means too much.

Full mechanism, in text

1 · Endogenous rhythm — dark triggers

2 · 0.1-0.3 mg already saturates

3 · Market dose 5-10 mg — backwards

4 · Label chaos + decision tree

Inverted dose-response — 0.3 mg saturates

Full mechanism, in text

1 · Endogenous rhythm — dark triggers

2 · 0.1-0.3 mg already saturates

3 · Market dose 5-10 mg — backwards

4 · Label chaos + decision tree

Inverted dose-response — 0.3 mg saturates