Low-Dose Melatonin: What the Clinical Data Actually Says

Low-Dose Melatonin: What the Clinical Data Actually Says

By Jack Zheng, MS Pharmacy — Founder of MIHIYO Labs

Summary

The clinical evidence for melatonin supports smaller doses than US retail products usually carry. In a 2001 trial, 0.3 mg restored plasma melatonin to normal nighttime levels in adults over 50 and improved sleep efficiency (Zhdanova et al.). EFSA's authorized claim for reduced sleep onset latency is set at 1 mg taken close to bedtime (EFSA, 2011). A 2005 meta-analysis found melatonin reduced sleep onset latency by about 4 minutes on average (Brzezinski et al.). A 2017 Canadian survey found melatonin product content ranging from -83% to +478% of the label. The MIHIYO Labs Sleep-Support oral dissolving strip (ODS) is engineered for a precise low dose.

How much melatonin do you actually need to support sleep?

Much less than most US supplement bottles imply. The dose-response curve for melatonin in sleep onset is not a "more is better" curve, and the clinical literature has been consistent on that point for two decades. The figures the evidence actually points to are 0.3 mg to 1 mg taken close to bedtime, not the 3 mg, 5 mg, and 10 mg products that dominate American shelves.

I want to walk through what the data actually shows, because the gap between "what the studies tested" and "what the gummy on the nightstand contains" is one of the largest mismatches I have seen between the science and the supplement market.

The Zhdanova low-dose study and what it found

The single most important reference point for low-dose melatonin is the Zhdanova trial in the Journal of Clinical Endocrinology and Metabolism. The team enrolled adults over 50, including 15 with confirmed age-related sleep efficiency decreases, and tested 0.1 mg, 0.3 mg, and 3.0 mg of oral melatonin against placebo, taken 30 minutes before bedtime.1

The 0.3 mg dose elevated plasma melatonin levels to within the normal nighttime physiological range and significantly restored sleep efficiency. The 3.0 mg dose pushed plasma melatonin well above the normal nighttime range, which the authors interpreted as a supraphysiologic level — and importantly, the higher dose did not deliver a larger sleep-efficiency benefit on the primary measure. The authors' framing was that a physiologically scaled dose, not a pharmacologic one, is what restored sleep efficiency in the older adults studied.

For a formulator, that paper is a doctrine, not a data point. It says the dose should be calibrated to physiologic melatonin levels, not to whatever fits in a 3 mg or 5 mg manufacturing slot.

What the EFSA-authorized claim sets as the regulatory dose

EFSA, the European Food Safety Authority, evaluated the human evidence for melatonin and sleep onset latency and issued an authorized health claim. The claim language is specific: melatonin contributes to the reduction of sleep onset latency, and the claim may be used for foods that provide at least 1 mg of melatonin per quantified portion, taken close to bedtime.2

The 1 mg figure is not a target. It is a regulatory minimum that EFSA considered substantiated by the human trials available at the time. Doses lower than 1 mg can also work physiologically — the Zhdanova paper shows this directly1 — but the EFSA claim required a defensible dose for general label use. The relevant point for a US reader is that this is the floor a European regulator considered evidence-supported, not the ceiling.

The popular US dose range of 3 mg to 10 mg sits well above this. It is not anchored to a clinical efficacy reason. It is anchored to manufacturing convention and to the assumption that more equals better.

The Brzezinski meta-analysis and the size of the effect

If you summarize the human trials that have asked the simple question "does melatonin reduce sleep onset latency," the answer is yes, on average, but the effect size is modest.

Brzezinski and colleagues pooled 17 studies with 284 subjects in a 2005 meta-analysis in Sleep Medicine Reviews. Melatonin reduced sleep onset latency by 4.0 minutes (95% CI 2.5 to 5.4). It also increased sleep efficiency by 2.2 percent and total sleep duration by 12.8 minutes.3 Those are real, statistically significant effects. They are not large effects. They are the kind of effect you would expect from a small physiologic prompt, not a sedative.

I think the honest framing of this is important. Melatonin is not a sleeping pill. It is a circadian-system signal molecule that, when given in roughly physiologic amounts at the right time, can shorten the gap to falling asleep by a few minutes and modestly improve sleep efficiency. Calling it more than that is a marketing claim, not a clinical one.

Melatonin dose vs plasma level and sleep-efficiency response Low doses of melatonin (around 0.3 to 1 mg) raise plasma melatonin into the normal nighttime physiologic range and capture the sleep-efficiency benefit; higher doses push plasma well above physiology without a proportional additional benefit. Dose vs plasma melatonin — physiologic vs supraphysiologic zone Sleep-efficiency benefit plateaus once plasma is back to nighttime range. Plasma melatonin level Normal nighttime range (physiologic) Supraphysiologic zone 0.1 mg 0.3 mg 1 mg 3 mg 10 mg Oral melatonin dose Zhdanova 2001 restored to normal EFSA 2011 claim supports sleep onset No greater benefit Zhdanova 2001 Common US OTC dose Sources: Zhdanova et al., 2001 (PMID 11600532); EFSA, 2011; Brzezinski et al., 2005 (PMID 15649737); Harpsoe et al., 2015 (PMID 26008214). Stylized — sleep-efficiency benefit plateaus once plasma is in the normal nighttime range.

Why oral melatonin pharmacokinetics make dose precision hard

The dose-response complication is compounded by how oral melatonin actually behaves in the body. Harpsoe and colleagues' systematic review of melatonin pharmacokinetics reported wide between-study variability in Tmax, Cmax, and bioavailability for immediate-release oral melatonin in healthy adults.4 The same labeled dose does not produce the same plasma curve in every person or every product.

That variability matters for two reasons.

First, a dose at the high end of the OTC range will reliably overshoot physiologic plasma levels, sometimes by an order of magnitude, but the overshoot does not translate into a proportional sleep-onset benefit.13 The dose-response is closer to a plateau than a slope above the physiologic range.

Second, if the product itself is mislabeled, the customer cannot calibrate at all. Erland and Saxena (2017) analyzed 31 melatonin natural health products sold in Canada and found measured content ranging from -83 percent to +478 percent of the label.5 In other words, a "5 mg" tablet could contain anything from less than 1 mg to roughly 29 mg in this dataset, and lot-to-lot variation was visible within a single product. Cohen and colleagues' 2023 JAMA analysis of US melatonin gummies came to a similar conclusion: 22 of 25 (88 percent) were inaccurately labeled, with measured content from 74 to 347 percent of claim.6

If the dose-response is flat above physiologic levels, and the labels are unreliable, the practical question becomes: how do you give a customer a dose they can trust?

The honest dose-effect picture in one table

Dose What the literature shows Source
0.1 mg Below clear sleep-efficiency effect in older adults Zhdanova et al., 2001
0.3 mg Restored plasma melatonin to normal nighttime range; improved sleep efficiency in older adults Zhdanova et al., 2001
1 mg EFSA-substantiated dose for "reduction of sleep onset latency" health claim EFSA, 2011
3 mg Plasma melatonin well above physiologic range; no greater sleep-efficiency benefit than 0.3 mg in Zhdanova trial Zhdanova et al., 2001
Pooled across trials Mean sleep onset latency reduction of 4.0 min; sleep efficiency +2.2%; total sleep +12.8 min Brzezinski et al., 2005
Wild range Single OTC products measured at -83% to +478% of label Erland & Saxena, 2017
Melatonin product label accuracy and pooled sleep-onset effect size Independent surveys of melatonin natural-health products show measured content ranging from minus 83 to plus 478 percent of the label, while the pooled clinical effect on sleep onset latency from the Brzezinski meta-analysis is about a 4 minute reduction. A. Measured melatonin content vs label claim — independent surveys 100% (label) -83% +478% Erland & Saxena, 2017 (J Clin Sleep Med) — 31 products tested 74% 347% Cohen et al., 2023 (JAMA) — 25 US melatonin gummies, 88% inaccurate 0% 100% 200% 300% 400% 500% Measured content as % of labeled dose B. Pooled meta-analysis effect — sleep onset latency 0 3 6 9 12 15 Mean reduction in sleep onset latency (minutes) 4.0 min 95% CI 2.5 - 5.4 Source: Brzezinski et al., 2005 (Sleep Med Rev, PMID 15649737) — pooled across 17 studies, 284 subjects.

The pattern is clear. Smaller doses, taken close to bedtime, line up with the physiology. Larger doses do not unlock larger sleep effects in healthy adults, and unreliable labels mean the actual dose the customer takes may be far from what the bottle says.

What this means for MIHIYO Sleep-Support

For the Sleep-Support ODS, the calculation runs in this order:

  1. Set the dose where the human evidence is strongest. That is the 0.3 mg to 1 mg band — Zhdanova for the low end and EFSA for the regulated minimum claim.12
  2. Pick a delivery format that protects that low dose from the variability problems documented in the gummy and tablet surveys.56
  3. Avoid the unit-dose accuracy failure mode the gummy literature describes by using a film with a known polymer mass and a measured active load per strip.

That last point is why I keep coming back to the strip format for melatonin specifically. A precision low-dose film does not need to be "stronger" than a 10 mg gummy. It needs to be reliably what it says it is, dosed at a level the literature supports.

The honest companion to that argument is the absorption-route conversation. Some sublingual and buccal melatonin formulations have shown higher Cmax and AUC than swallowed tablets at equivalent doses, consistent with partial bypass of hepatic first-pass.4 That is a real route advantage, and it argues for keeping the dose low rather than high, because the same delivered exposure can be achieved with less starting material. For readers who want the dose-and-format detail in one place, our earlier piece on melatonin strip dosage and the strip versus gummy comparison are the natural follow-ups.

Where the low-dose argument has limits

I do not want to oversell low-dose melatonin either.

It does not treat insomnia, in the clinical sense. The Brzezinski meta-analysis effect size of a few minutes on sleep onset latency is statistically reliable, but it is not what insomnia therapy looks like.3 For chronic insomnia, the first-line evidence-based intervention is cognitive behavioral therapy for insomnia, not a melatonin gummy.

It also does not solve circadian misalignment by itself. For jet lag or shift-work phase shifting, the dose, timing relative to the user's internal night, and direction of the shift matter at least as much as the milligrams. Off-protocol melatonin can move the circadian phase in the wrong direction.

Finally, the data on long-term use of nightly melatonin in healthy adults is thinner than the marketing implies. Standard cautions about pregnancy, hormone-sensitive conditions, and interactions with sedatives and immunosuppressants apply. None of the trials cited here justify daily melatonin as a default.

The bottom line

For most adults using melatonin to nudge sleep onset, the evidence base supports doses near 0.3 mg to 1 mg, taken close to bedtime, with realistic expectations of a few minutes' reduction in time to fall asleep on average. Larger doses do not clear a higher bar in the controlled trials, and they push plasma melatonin well above the body's normal nighttime range. With labels showing measured content from -83 percent to +478 percent of claim in independent surveys, dose precision matters at least as much as the milligram printed on the bottle. The MIHIYO Labs Sleep-Support strip is built around exactly that constraint.

References

  1. Zhdanova IV, Wurtman RJ, Regan MM, Taylor JA, Shi JP, Leclair OU. Melatonin treatment for age-related insomnia. J Clin Endocrinol Metab. 2001;86(10):4727-4730. PMID: 11600532. DOI: 10.1210/jcem.86.10.7901. <https://pubmed.ncbi.nlm.nih.gov/11600532/>
  2. EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA). Scientific Opinion on the substantiation of a health claim related to melatonin and reduction of sleep onset latency. EFSA Journal. 2011;9(6):2241. DOI: 10.2903/j.efsa.2011.2241. <https://www.efsa.europa.eu/en/efsajournal/pub/2241>
  3. Brzezinski A, Vangel MG, Wurtman RJ, Norrie G, Zhdanova I, Ben-Shushan A, Ford I. Effects of exogenous melatonin on sleep: a meta-analysis. Sleep Med Rev. 2005;9(1):41-50. PMID: 15649737. DOI: 10.1016/j.smrv.2004.06.004. <https://pubmed.ncbi.nlm.nih.gov/15649737/>
  4. Harpsøe NG, Andersen LP, Gögenur I, Rosenberg J. Clinical pharmacokinetics of melatonin: a systematic review. Eur J Clin Pharmacol. 2015;71(8):901-909. PMID: 26008214. DOI: 10.1007/s00228-015-1873-4. <https://pubmed.ncbi.nlm.nih.gov/26008214/>
  5. Erland LA, Saxena PK. Melatonin Natural Health Products and Supplements: Presence of Serotonin and Significant Variability of Melatonin Content. J Clin Sleep Med. 2017;13(2):275-281. PMID: 27855744. DOI: 10.5664/jcsm.6462. <https://pubmed.ncbi.nlm.nih.gov/27855744/>
  6. Cohen PA, Avula B, Wang YH, Katragunta K, Khan I. Quantity of Melatonin and CBD in Melatonin Gummies Sold in the US. JAMA. 2023;329(16):1401-1402. PMID: 37097362. DOI: 10.1001/jama.2023.2296. <https://pubmed.ncbi.nlm.nih.gov/37097362/>
  7. Andersen LP, Werner MU, Rosenkilde MM, Harpsøe NG, Fuglsang H, Rosenberg J, Gögenur I. Pharmacokinetics of oral and intravenous melatonin in healthy volunteers. BMC Pharmacol Toxicol. 2016;17:8. PMCID: PMC4759723. DOI: 10.1186/s40360-016-0052-2. <https://pubmed.ncbi.nlm.nih.gov/26893170/>

By — Founder of MIHIYO Labs. Focused on the R&D of high-bioavailability, fast-absorption oral dissolving strips.

0 comments

Leave a comment

Please note, comments need to be approved before they are published.