Sleep medicine's own guideline suggests clinicians not use melatonin for insomnia

The American Academy of Sleep Medicine's clinical practice guideline on drugs for chronic insomnia in adults contains one line about melatonin, and it is a negative one: the guideline suggests that clinicians not use melatonin as a treatment for sleep onset or sleep maintenance insomnia, versus no treatment, in adults [s1]. Where a pooled effect on falling asleep has been measured, it is about seven minutes [s2].

Both of those statements need immediate qualification, and the qualifications run in opposite directions. This is one of those questions where the confident answer — in either direction — is the wrong one.

What the guideline says, and how much weight it carries

The 2017 AASM guideline evaluated individual drugs rather than broad classes, using the GRADE process, and every single recommendation it issued came out WEAK [s1]. That includes the eight agents it suggests clinicians use — suvorexant and doxepin for sleep maintenance insomnia; zaleplon, triazolam and ramelteon for sleep onset insomnia; eszopiclone, zolpidem and temazepam for both — and the six it suggests clinicians not use: trazodone, tiagabine, diphenhydramine, melatonin, tryptophan and valerian [s1].

The guideline is explicit that a weak recommendation "reflects a lower degree of certainty in the outcome and appropriateness of the patient-care strategy for all patients, but should not be construed as an indication of ineffectiveness" [s1]. It goes further and explains why weakness was expected in advance: downgrading is predictable under GRADE given the funding source for most pharmacological trials and the attendant risk of publication bias, the relatively small number of eligible trials for each individual agent, and the heterogeneity observed in the data [s1].

So the honest summary of the guideline position is not "melatonin has been shown not to work." It is that a task force of four sleep-medicine experts looked at the randomised trials for each agent one at a time, found the melatonin evidence insufficient to justify recommending it, and had low confidence in that judgment too [s1].

The size of the effect

The most-cited pooled estimate comes from a meta-analysis of 19 randomised, placebo-controlled trials covering 1,683 adults and children with primary sleep disorders [s2]. Melatonin reduced the time taken to fall asleep by a weighted mean difference of 7.06 minutes (95% CI 4.37 to 9.75), and increased total sleep time by 8.25 minutes (95% CI 1.74 to 14.75) [s2]. Overall sleep quality improved, with a standardized mean difference of 0.22 (95% CI 0.12 to 0.32) [s2].

Those are statistically significant and small. Seven minutes is inside the margin of error of most people's own estimate of how long they lay awake. The meta-analysis authors describe the effects as modest, note that they do not appear to dissipate with continued use, and observe that the absolute benefit is smaller than that of other pharmacological treatments for insomnia [s2].

What the evidence says about dose

Less than the packaging implies. The meta-analysis performed a meta-regression on dose and duration and found that trials using longer durations and higher doses of melatonin showed greater effects on decreasing sleep latency and increasing total sleep time — but no significant effect of either dose or trial duration on sleep quality [s2]. That is a signal, not a dosing schedule, and it comes from comparing across trials rather than from randomising people to different doses within one.

No dose appears in the AASM recommendation at all, because the recommendation is against use [s1]. There is no guideline-endorsed melatonin dose for chronic insomnia in adults to report, and this article is not the place to infer one.

The problem underneath the dose question

Even a well-specified dose assumes the bottle contains what the label says. An analysis published in the Journal of Clinical Sleep Medicine quantified melatonin in commercial supplements using ultraperformance liquid chromatography and found that melatonin content ranged from −83% to +478% of the labelled content [s3]. Lot-to-lot variability within a single product varied by as much as 465% [s3]. More than 71% of supplements did not meet their label claim within a 10% margin [s3].

Separately, serotonin — a controlled substance used in the treatment of several neurological disorders — was identified in eight of the supplements, at levels of 1 to 75 μg [s3].

One bookkeeping note, in the interest of reporting sources as they are: the paper's abstract states in one sentence that melatonin was quantified in 30 commercial supplements and in the next that a total of 31 supplements were analyzed [s3]. The serotonin figures are internally consistent with a denominator of 31 — eight products is the "additional 26%" the abstract reports [s3] — but the two sample sizes are not reconciled in the abstract itself, and readers should treat the exact denominator as unresolved.

Where this leaves the question

Melatonin is not the same kind of object as a prescription hypnotic, and it is not the same kind of object as a placebo. The measured effect on falling asleep is real, replicated across 19 trials, and roughly the size of a coffee break [s2]. The body that writes the American guidelines looked at that evidence and declined to recommend it, while grading its own recommendation weak and warning readers not to read weakness as proof of ineffectiveness [s1]. And the supplement in a given bottle may contain a small fraction, or nearly five times, the dose printed on it [s3].

Those three findings do not cancel out into a verdict. They describe a substance with a small effect, a guideline body that is unconvinced, and a supply chain that makes the dose question partly unanswerable in practice. Anyone weighing melatonin for persistent insomnia has a clinician-shaped question on their hands, not a shopping one — particularly since the same guideline that declines to recommend melatonin does suggest several prescription options for the same indication, at the same weak level of certainty [s1].

This article is informational and is not medical advice, and nothing in it should be used to start, stop or change a medication or supplement.

Sources

• Clinical Practice Guideline for the Pharmacologic Treatment of Chronic Insomnia in Adults: An American Academy of Sleep Medicine Clinical Practice Guideline — Journal of Clinical Sleep Medicine, 2017-02-15
• Meta-analysis: melatonin for the treatment of primary sleep disorders — PLOS One, 2013-05-17
• Melatonin Natural Health Products and Supplements: Presence of Serotonin and Significant Variability of Melatonin Content — Journal of Clinical Sleep Medicine, 2017-02-15