Melatonin: Effects, Dosing, and the Most Common Myths
Melatonin is not a sleep medication in the classic sense, but primarily a time-giving signal. Anyone who understands that doses smaller, times more wisely, and expects the right thing. What the studies actually show, where melatonin has a place, and what the BfR 2024 warns about.
Melatonin is sold like a herbal sleep aid and expected to work like knockout drops. Both miss the biology. Melatonin is the body's darkness signal, a clock, not a hammer. The meta-analysis by Ferracioli-Oda 2013 in PLoS One shows an effect on sleep onset latency of about 7 minutes, small but real. The phase response curve by Burgess 2007 in the Journal of Physiology shows that the time of intake decides between effect and counter-effect. Forsling 2002 shows that high doses act differently and in part oppositely to low doses. The BfR warns in 2024 against uncritical continuous use. The most important question with melatonin is not how much, but when and for what. In this spoke I separate marketing from mechanism.
This spoke is the pharmacological workshop of the sleep cluster. We go through the mechanism of action (what melatonin actually does in the body), the crucial distinction between the circadian and the hypnotic effect, the question of a sensible versus an excessive dose, timing according to the phase response curve, the good areas of use (jet lag, shift work, late type, people from 55 onward), the typical myths, the safety and BfR situation, the KPNI lenses, and three concrete levers for practice.
What melatonin really does in the body
Melatonin is a hormone that the pineal gland releases in darkness. It is the biochemical signal It is night. The rise begins in the evening a few hours before falling asleep; this moment is called Dim Light Melatonin Onset (DLMO) and is the most important marker of the internal clock. Over the night the level stays high and falls toward morning. Bright light, above all with a high blue component, suppresses the release. That is why light management is the foundation of every sleep strategy and the preparation is only the supplement.
Exogenous melatonin (as a tablet, spray, or capsule) mimics this signal. It binds to the receptors MT1 and MT2, above all in the suprachiasmatic nucleus, the central clock in the hypothalamus. From this arise two different effects that are constantly confused in everyday life and that I separate cleanly in the next section.
Many people treat melatonin like a stronger valerian. More apt is the image of a clock you set. A set clock does not make you tired; it ensures that the tiredness comes at the right time. Precisely for this reason the same tablet can help or disrupt depending on the time of intake.
Circadian versus hypnotic effect
This is the central distinction of this whole topic. Anyone who has understood it understands melatonin.
Hypnotic effect
The direct, weak drowsiness effect shortly after intake. It slightly lowers core body temperature and dampens the wake signal of the internal clock. This effect is small: in the meta-analysis by Ferracioli-Oda 2013 about 7 minutes shorter sleep onset latency. No knockout, more a gentle nudge.
Circadian effect
The shifting of the internal clock itself. It follows the phase response curve (Burgess 2007): melatonin in the early evening shifts the clock earlier, melatonin in the morning shifts it later. This effect is the real lever in a shifted sleep rhythm and can take hold independently of the direct drowsiness effect.
Why this matters
A large dose at bedtime uses almost only the weak hypnotic effect and lies in the dead zone of the phase response curve, where hardly any shift takes place. Anyone with a rhythm problem thereby wastes the actual effect.
The practical consequence
In pure difficulty falling asleep without a rhythm shift, a small dose shortly before bedtime can have a modest effect. With a shifted clock (late type), the small early-evening dose is the key, timed by the internal clock, not by the desired clock time.
Three-pulse phase response curve to 3 mg melatonin in humans
Human study Helen Burgess, Victoria Revell, and Charmane Eastman published a phase response curve to exogenous melatonin in the Journal of Physiology in 2007. 27 healthy young participants took 3 mg melatonin or placebo over three days at the same clock time, in a controlled dim-light setting, with saliva measurement of the body's own melatonin onset. Result: the advance portion of the curve peaked about 5 hours before the Dim Light Melatonin Onset, that is in the afternoon to early evening. The delay portion peaked about 11 hours after the DLMO, shortly after the usual wake time. In the middle of the first half of the night there was a dead zone with minimal shift. The maximum shifts were 1.8 hours (advance) and 1.3 hours (delay). The authors concluded explicitly that melatonin as a pure nightly sleep aid has only minimal time-shifting effect.
Burgess HJ, Revell VL, Eastman CI. J Physiol. 2007;586(2):639-47. DOI: 10.1113/jphysiol.2007.143180 · PMID: 18006583
How strongly melatonin can affect falling asleep
Here a sober number helps more than any advertising promise. The largest solid summary comes from Ferracioli-Oda 2013.
Melatonin for primary sleep disorders, 19 studies
Meta-analysis Eduardo Ferracioli-Oda and colleagues analyzed 19 randomized placebo-controlled studies with a total of 1683 participants in PLoS One in 2013. Melatonin significantly shortened sleep onset latency by an average of 7.06 minutes (95 percent confidence interval 4.37 to 9.75), extended total sleep time by 8.25 minutes, and improved sleep quality moderately (standardized mean difference 0.22). The meta-regression showed: longer use and higher doses were associated with somewhat larger effects on sleep onset latency and total sleep time, but not on sleep quality. The authors' key statement: the absolute effect is smaller than with classic sleep medications, but it does not disappear with continued use, and the side effect profile is comparatively favorable.
Ferracioli-Oda E, Qawasmi A, Bloch MH. PLoS One. 2013;8(5):e63773. DOI: 10.1371/journal.pone.0063773 · PMID: 23691095
For context by comparison: in the large network meta-analysis of insomnia medications (Yue 2023 in Sleep Medicine Reviews, 69 studies) and in the overview of 153 studies by Pan 2023 in Drugs, orexin antagonists and Z-drugs performed better than melatonin receptor agonists for pure symptom strength. Melatonin receptor agonists showed above all an effect on sleep-onset insomnia with a good safety profile. This fits the overall picture: melatonin is the gentle, weak, but safe option, not the strongest.
Sensible versus excessive dose
The most common mistake is not too little but too much. Over-the-counter products often contain 3, 5, or even 10 mg. Physiologically the body reaches nighttime peaks that are already mimicked with fractions of a milligram. Higher doses produce unphysiologically high levels that can reach into the late morning and then cause daytime drowsiness.
Low and high doses act differently
Human study Mary Forsling and colleagues showed in 2002 in Clinical Endocrinology, in a double-blind, randomized cross-over study in 24 young men, that melatonin has dose-dependent and partly opposite effects. On stimulated hormonal responses (vasopressin, oxytocin) a low dose of 0.5 mg had an enhancing effect, whereas a high dose of 5 mg had an attenuating effect. The authors conclude that low, near-physiological doses show a different pharmacological behavior than high doses. Translated into practice this means: more melatonin is not simply more of the same, but can act qualitatively differently.
Forsling ML, Williams AJ. Clin Endocrinol (Oxf). 2002;57(5):615-20. DOI: 10.1046/j.1365-2265.2002.01644.x · PMID: 12390335
A sensible orientation from the study landscape: for circadian control 0.5 to 1 mg, precisely timed, are usually enough. For jet lag, according to the Cochrane review, daily doses between 0.5 and 5 mg were similarly effective, and doses above 5 mg brought no additional benefit. In Germany melatonin medicines are dosed by a physician. For food supplements the BfR warning applies. The rule of thumb is: as little as possible, as precisely timed as necessary.
Myth 1: More helps more. Wrong. Higher doses do not reliably improve sleep quality and can trigger daytime drowsiness. Myth 2: Melatonin is a strong sleep medication. Wrong. The mean effect on sleep onset latency is about 7 minutes (Ferracioli-Oda 2013), that is gentle, not strong. Myth 3: It does not matter when you take it. Wrong. Timing decides between effect and counter-effect (Burgess 2007). Myth 4: It is harmless because it is endogenous. Incomplete. The BfR warns in 2024 explicitly against uncritical continuous use, especially in risk groups.
Timing in practice
Because timing is the real lever, here are the most important constellations, each without rigid numerical prescriptions, since the individual internal clock varies.
Timing by application goal
- Pure sleep aid: small dose about 30 to 60 minutes before the desired bedtime. Expectation: modest effect, above all on sleep onset latency.
- Late type (delayed sleep phase): small dose in the early evening, a few hours before one's own usual tiredness, to shift the clock forward. Combined with bright light in the morning.
- Jet lag eastward: intake close to the target bedtime (10 pm to midnight local time) at the destination, starting on the day of arrival (Herxheimer and Petrie 2002).
- Shift work: a complex special case, intake before the planned daytime sleep after the night shift, always combined with consistent darkening. Here individual guidance is sensible.
In all cases the rule is: bright light at the wrong time (screen in the evening, no daylight in the morning) can override any melatonin effect. Light behavior is the stronger time cue. More on this in the circadian spoke (Spoke 18).
Where melatonin has a good place
Three areas of use are best documented, a fourth is rather disappointing.
Melatonin for the prevention and treatment of jet lag
Systematic review Andrew Herxheimer and Keith Petrie evaluated 10 randomized studies in the Cochrane review in 2002. Nine out of ten showed that melatonin, taken close to the target bedtime, reduces jet lag on flights across five or more time zones. The number needed to treat was 2. Daily doses of 0.5 to 5 mg were similarly effective, with people falling asleep somewhat faster after 5 mg than after 0.5 mg. Doses above 5 mg brought no additional benefit. The effect is greater with more time zones crossed and with eastward flights. The authors warn: taken too early in the day, melatonin can cause sleepiness and delay adaptation. They also mention caution with epilepsy and a possible link to warfarin.
Herxheimer A, Petrie KJ. Cochrane Database Syst Rev. 2002;(2):CD001520. DOI: 10.1002/14651858.CD001520 · PMID: 12076414
Prolonged release melatonin over 6 months, age as predictor
RCT Alan Wade and colleagues investigated prolonged release melatonin versus placebo in 791 adults with primary insomnia in BMC Medicine in 2010, double-blind, over up to 6 months. In older patients prolonged release melatonin markedly shortened sleep onset latency (minus 19.1 versus minus 1.7 minutes under placebo, P equals 0.002), independent of the measured body's own melatonin level. Interesting: a low own melatonin level alone did not predict success, age did so better. The effects persisted over 6 months without tolerance development, without clinically relevant safety problems. Already Wade 2007 in Current Medical Research and Opinion had shown significant improvements in sleep quality, morning alertness, and sleep onset latency in 55- to 80-year-olds.
Wade AG, Ford I, Crawford G, et al. BMC Med. 2010;8:51. DOI: 10.1186/1741-7015-8-51 · PMID: 20712869
The fourth, rather disappointing area: sleep disturbances in dementia. The Cochrane review by McCleery and Sharpley 2020 found no convincing benefit of melatonin up to 10 mg on relevant sleep parameters, whereas orexin antagonists and, in small studies, trazodone tended to show effects. Melatonin is therefore not an all-purpose remedy but has clearly defined indications.
Well documented: Circadian rhythm disorder
Late type and delayed sleep phase, small dose timed in the early evening according to the phase response curve (Burgess 2007).
Well documented: Jet lag
Across five or more time zones, especially eastward, close to the target bedtime (Herxheimer and Petrie 2002, NNT 2).
Documented: Insomnia from about 55 years
Prolonged release melatonin, since one's own production declines with age (Wade 2007, Wade 2010).
Rather weak: Dementia, young classic insomnia
In dementia no convincing benefit (McCleery 2020). In young people without a rhythm problem only a small effect.
The KPNI lenses on melatonin
In Clinical Psychoneuroimmunology melatonin does not stand in isolation but embedded in light biology, metabolism, and inflammation regulation. Four lenses.
Light and the internal clock
Melatonin is downstream. The primary lever is light: bright in the morning, dimmed in the evening. Anyone who does not change their light behavior can only work against their own time cue with melatonin. The preparation supplements but does not replace light.
Tryptophan and serotonin
Melatonin is formed via tryptophan and serotonin. Adequate protein supply and a functioning day-night contrast in the serotonin-melatonin axis are the foundation. Without daytime activity and light the substrate timing is missing.
Inflammation and sleep
Low-grade inflammation and stress can disturb sleep and are not remedied by a single pill. Melatonin can be a building block; addressing the causes (stress, metabolism, exercise) remains the core.
Individual chronobiology
Early and late types respond differently. There is no blanket recommendation. The history of one's own internal clock (when do I become tired on my own) decides dose timing more than any standard figure.
Safety, risk groups, and the BfR position
At usual amounts melatonin is not acutely toxic, that is the good news. The differentiated news is provided by the German Federal Institute for Risk Assessment.
BfR warns against uncritical use of melatonin-containing food supplements
Authority document The Federal Institute for Risk Assessment made clear in its statement 42/2024 (17 September 2024): the available data argue against uncritical, uncontrolled use of melatonin-containing food supplements, above all over longer periods. Possible adverse effects include, among others, daytime sleepiness, headache, prolonged reaction time, drop in blood pressure, and vivid dreams or nightmares. Some studies in adults point to a possible influence on the body's own hormones with developmental or reproductive biological significance. The BfR advises in particular pregnant women, breastfeeding women, children, adolescents, and people with certain pre-existing conditions against independent, uncontrolled use. Unlike medicines, food supplements are not subject to an approval review for efficacy, safety, and tolerability.
Bundesinstitut für Risikobewertung. Melatoninhaltige Nahrungsergänzungsmittel: BfR weist auf mögliche Gesundheitsrisiken hin. Stellungnahme Nr. 42/2024. bfr.bund.de
In practice this means: no driving or operating machinery shortly after intake (reaction ability may be impaired), caution with interactions with blood-thinning medications and with epilepsy, no self-application in children or during pregnancy and breastfeeding without medical guidance. Anyone who wants to use melatonin for longer than the short term should clarify this with a physician, not least so as not to overlook the actual cause of the sleep disturbance.
Set a clock, do not swing a hammer
Melatonin can help when the internal clock is shifted or weakens with age. It can work when timing and dose are right. It remains a building block alongside light, rhythm, and addressing the causes, not a replacement for them.
Three levers for practice
First clarify the timing, then the dose
Ask yourself: do I have a rhythm problem (becoming tired too late) or pure difficulty falling asleep? With a rhythm problem, a small dose in the early evening; with a pure sleep aid, shortly before sleep. The dose is secondary, the timing decides.
Start small, not large
0.5 to 1 mg is often enough. Higher doses do not reliably improve quality and can cause daytime drowsiness (Forsling 2002, Ferracioli-Oda 2013). More is no advantage here.
Light first, pill afterward
Out into daylight in the morning, reduce screens and bright light in the evening. That is the stronger time cue. Melatonin can best take hold as a supplement to good light behavior, not as a replacement for it (see Spoke 18).
Frequently asked questions about melatonin
How strongly can melatonin really affect falling asleep?
Weakly, but measurably. The meta-analysis by Ferracioli-Oda 2013 in PLoS One (19 studies, 1683 participants) showed an average shortening of sleep onset latency by about 7 minutes and an extension of total sleep time by about 8 minutes. Sleep quality improved moderately. That is markedly less than with classic sleep medications, but persists with continued use and comes with a favorable side effect profile. Melatonin is primarily a time-giving signal, not a strong sleep medication.
What is the difference between the circadian and the hypnotic effect?
The hypnotic effect is the weak, direct drowsiness effect shortly after intake. The circadian effect shifts the internal clock and follows the phase response curve. Burgess 2007 in the Journal of Physiology showed: melatonin in the early evening (about 5 hours before the body's own rise) shifts the clock forward, in the morning backward, in the middle of the night hardly at all. Anyone who cannot fall asleep because of a shifted clock needs a small dose in the early evening, not a large one at bedtime.
How much melatonin is sensible and when is it too much?
For circadian control 0.5 to 1 mg at the right time are usually enough. Higher doses of 3 to 10 mg produce unphysiologically high levels that tend to overshoot. Forsling 2002 showed that low and high doses can even have opposite effects. The most common mistake is too high a dose at the wrong time. In Germany melatonin medicines are dosed by a physician; for food supplements the BfR (statement 42/2024) warns against uncritical use. Never dose by the principle that a lot helps a lot.
When is the best time to take melatonin?
That depends on the goal. As a pure sleep aid usually 30 to 60 minutes before bedtime. To shift a displaced clock, timing follows the phase response curve (Burgess 2007): late types benefit from a small dose in the early evening. For jet lag the Cochrane review by Herxheimer and Petrie 2002 recommends intake close to the target bedtime at the destination. Wrong timing can delay adaptation and cause daytime drowsiness. Light management is at least as important as the preparation.
Does melatonin help with jet lag?
Here the evidence is comparatively good. The Cochrane review by Herxheimer and Petrie 2002 included 10 studies, nine found a reduction of jet lag on flights across five or more time zones when taken close to the target bedtime. The number needed to treat was 2. Daily doses of 0.5 to 5 mg were similarly effective. The effect is greater with more time zones and with eastward flights. If mistimed, melatonin can delay adaptation. It can help with jet lag but does not replace well-considered light and sleep management.
Is melatonin addictive or does it have withdrawal effects?
According to current data, melatonin is not considered classically addictive, unlike benzodiazepines or Z-drugs. In the meta-analysis by Ferracioli-Oda 2013 the effect did not diminish with continued use, and the 6-month study by Wade 2010 showed no tolerance and no withdrawal symptoms. This does not mean that unlimited continuous use is harmless. The BfR emphasizes in 2024 that long-term data on food supplements are lacking. A time-limited, targeted use is the serious option.
Is melatonin dangerous? What side effects are there?
In controlled studies melatonin was comparatively well tolerated. The BfR nonetheless lists in 2024 possible adverse effects: daytime sleepiness, headache, prolonged reaction time, drop in blood pressure, vivid dreams. In particular pregnant women, breastfeeding women, children, adolescents, and people with pre-existing conditions should not use it independently. Relevant are interactions with blood-thinning medications and caution with epilepsy. No driving shortly after intake. Acutely toxic melatonin is not at usual amounts; rather the problem is uncritical continuous use.
Who is melatonin most likely suited for?
Best documented: in circadian rhythm disorders such as delayed sleep phase syndrome (small dose in the early evening, Burgess 2007), in jet lag across several time zones especially eastward (Herxheimer and Petrie 2002), and in difficulty falling and staying asleep from about 55 years of age, where one's own production declines (Wade 2007 and 2010 with prolonged release melatonin). In Wade 2010 age predicted success better than the melatonin level. In young people with classic insomnia without a rhythm problem the effect is small; for dementia-related sleep disturbances McCleery 2020 found no convincing benefit.
All spokes in the sleep cluster
- Pillar: Treating sleep disorders holistically
- Spoke 5: Melatonin effects, dosing, myths (you are here)
- Spoke 6: Sleep hygiene that really works
- Spoke 18: Circadian rhythm and light
Connections to other topics
The pillar article of the cluster. Where melatonin fits into the overall strategy of light biology, rhythm, metabolism, and addressing the causes.
Why light is the stronger time cue and how the phase response curve is used in practice. The foundation for any sensible melatonin timing.
The behavioral foundations that move more than any preparation. Melatonin can best take hold on a solid sleep foundation.
Sleep disorders have causes that a single tablet does not solve. When a medical examination instead of self-medication is the right path.
Sources and further reading
- Ferracioli-Oda E, Qawasmi A, Bloch MH. Meta-analysis: melatonin for the treatment of primary sleep disorders. PLoS One. 2013;8(5):e63773. DOI: 10.1371/journal.pone.0063773 · PMID: 23691095 [Meta-analysis]
- Burgess HJ, Revell VL, Eastman CI. A three pulse phase response curve to three milligrams of melatonin in humans. J Physiol. 2007;586(2):639-47. DOI: 10.1113/jphysiol.2007.143180 · PMID: 18006583 [RCT]
- Forsling ML, Williams AJ. The effect of exogenous melatonin on stimulated neurohypophysial hormone release in man. Clin Endocrinol (Oxf). 2002;57(5):615-20. DOI: 10.1046/j.1365-2265.2002.01644.x · PMID: 12390335 [RCT]
- Herxheimer A, Petrie KJ. Melatonin for the prevention and treatment of jet lag. Cochrane Database Syst Rev. 2002;(2):CD001520. DOI: 10.1002/14651858.CD001520 · PMID: 12076414 [Systematic review]
- Wade AG, Ford I, Crawford G, et al. Nightly treatment of primary insomnia with prolonged release melatonin for 6 months. BMC Med. 2010;8:51. DOI: 10.1186/1741-7015-8-51 · PMID: 20712869 [RCT]
- Wade AG, Ford I, Crawford G, et al. Efficacy of prolonged release melatonin in insomnia patients aged 55-80 years. Curr Med Res Opin. 2007;23(10):2597-605. DOI: 10.1185/030079907X233098 · PMID: 17875243 [RCT]
- Yue JL, Chang XW, Zheng JW, et al. Efficacy and tolerability of pharmacological treatments for insomnia in adults: a systematic review and network meta-analysis. Sleep Med Rev. 2023;68:101746. DOI: 10.1016/j.smrv.2023.101746 · PMID: 36701954 [Meta-analysis]
- Pan B, Ge L, Lai H, et al. The comparative effectiveness and safety of insomnia drugs: a systematic review and network meta-analysis of 153 randomized trials. Drugs. 2023;83(7):587-619. DOI: 10.1007/s40265-023-01859-8 · PMID: 36947394 [Meta-analysis]
- McCleery J, Sharpley AL. Pharmacotherapies for sleep disturbances in dementia. Cochrane Database Syst Rev. 2020;11(11):CD009178. DOI: 10.1002/14651858.CD009178.pub4 · PMID: 33189083 [Meta-analysis]
- Bundesinstitut für Risikobewertung. Melatoninhaltige Nahrungsergänzungsmittel: BfR weist auf mögliche Gesundheitsrisiken hin. Stellungnahme Nr. 42/2024 vom 17. September 2024. bfr.bund.de [Authority document]