Evidence File 031 · Melatonin · Method v0.14 of 4 claims researched

MelatoninImmediate release / gummies · 4 claims

The real effect is about seven minutes. The bigger issue is that the dose in the bottle frequently is not the dose on the label.

4Claims scored

5Sources

0.5–4 mgStudied dose

Jul 26Reviewed

Claim 01

Reduces the time it takes to fall asleep

7.0/10

Moderate·19 studies · 1,683 subjects · about 7 minutes

Across 19 randomised trials and 1,683 subjects with primary sleep disorders, melatonin reduced sleep onset latency by 7.06 minutes (95% CI 4.37 to 9.75, p<0.001) and improved overall sleep quality (SMD 0.22, 95% CI 0.12–0.32).

The reviewers were direct about magnitude: the effects are modest, the absolute benefit is smaller than other pharmacological insomnia treatments, and melatonin’s case rests on its comparatively benign side-effect profile rather than its potency. Notably, the effects did not dissipate with continued use and melatonin showed little dependence potential or habituation.

Seven minutes is the honest number. If your expectation is set by the marketing, it will not match.

The working

7Quantity

6Quality

7Consistency

9Directness

(7 × 0.25) + (6 × 0.35) + (7 × 0.25) + (9 × 0.15) = 7.0
Consistent and well replicated. The score reflects a real, reliable, small effect — which is a different thing from a large one.

MetaFerracioli-Oda E, Qawasmi A, Bloch MH. Meta-analysis: melatonin for the treatment of primary sleep disorders PLoS One 2013;8(5) · 19 studies, 1,683 subjects

Claim 02

Meaningfully increases total sleep time

5.8/10

Emerging·Statistically significant · about 8 minutes

Total sleep time increased by 8.25 minutes (95% CI 1.74 to 14.75, p=0.013). The lower bound of that interval is under two minutes.

Trials with longer duration and higher doses showed greater effects on both sleep latency and total sleep time — though the dose-response analysis in the next claim complicates that picture.

The working

6Quantity

6Quality

5Consistency

6Directness

(6 × 0.25) + (6 × 0.35) + (5 × 0.25) + (6 × 0.15) = 5.8
Directness scores 6. The result is statistically real and the magnitude is unlikely to be what anyone means by “meaningfully.”

MetaFerracioli-Oda E, et al. Total sleep time outcome WMD 8.25 minutes, 95% CI 1.74–14.75

Claim 03

Retail products contain the dose on the label

2.2/10

Evidence against·88% of gummies inaccurately labelled

A 2023 JAMA analysis tested 25 melatonin gummy products sold in the US. 22 of 25 (88%) were inaccurately labelled. Only three contained melatonin within ±10% of the declared quantity. Actual content ranged from 74% to 347% of the label. One product contained no detectable melatonin at all, but did contain 31.3 mg of CBD. Around a quarter of products contained unlabelled serotonin.

This is not new. A Canadian analysis found content ranging from 83% to 478% of label, with lot-to-lot variation within a single brand reaching 465%.

The structural cause is regulatory. Melatonin is sold as a dietary supplement in the US, so the FDA does not review label accuracy before products go to market. The JAMA analysis was prompted by paediatric melatonin calls to US Poison Control Centres rising 530% between 2012 and 2021. Adult use rose more than fivefold from 1999 to 2018.

The working

1Quantity

1Quality

1Consistency

9Directness

(1 × 0.25) + (1 × 0.35) + (1 × 0.25) + (9 × 0.15) = 2.2
Scored as support for the claim. A direct laboratory analysis answered this and the answer was no.

LabCohen PA, et al. Quantity of melatonin and CBD in melatonin gummies sold in the US JAMA 2023;329(16) · 25 products, 88% inaccurately labelled
RevCanadian supplement content analysis and format comparison, summarising StatPearls melatonin entry Content 83–478% of label; lot-to-lot variation to 465%

Claim 04

Higher doses work better

4.1/10

Emerging·Efficacy peaked near 4 mg · timing beat dose

A 2024 dose-response analysis across 26 RCTs and 1,689 observations found efficacy for reducing sleep onset latency peaked at approximately 4 mg/day regardless of format, and that timing was a stronger predictor than dose — specifically, taking it 2–3 hours before intended sleep rather than 30 minutes before.

That cuts against both the retail market, which sells 10 mg, and consumer habit, which is to take it at bedtime. It also cuts against the earlier meta-analysis, which found higher doses gave greater effects on sleep latency.

The working

3Quantity

4Quality

3Consistency

8Directness

(3 × 0.25) + (4 × 0.35) + (3 × 0.25) + (8 × 0.15) = 4.1
One dose-response analysis against a general trend in the older trial literature. Enough to raise the question, not enough to settle it.

MetaCruz-Sanabria F, et al. dose-response findings, summarised J Pineal Res 2024 · doi 10.1111/jpi.12985 · 26 RCTs, 1,689 observations

Dosage as studied

Most-studied regimen

0.5–4 mg per day

Efficacy for sleep onset peaked around 4 mg/day in dose-response analysis. Timing mattered more than dose.

Studied doses: Trials span a wide range. A dose-response analysis across 26 RCTs and 1,689 observations found efficacy for reducing sleep onset latency peaked around 4 mg/day regardless of format.
Timing: Taking melatonin 2–3 hours before intended sleep onset was a stronger predictor of effect than dose. Most consumers take it 30 minutes before bed.
Typically sold: 10 mg is a common retail strength — well above the point where added dose stopped helping.
Format: Gummies show the worst label accuracy of any format. They are manufactured as food products, and heat and sugar degrade melatonin in ways tablets avoid.