evipedia.ai
← Back

Melatonin for Health & Longevity

Evidence Review created on 10/21/2026 using AI4L / Opus 5.5

Also known as: N-acetyl-5-methoxytryptamine, Circadin, Slenyto

Motivation

Melatonin is a hormone made mainly by a small gland deep in the brain as daylight fades. It works as the body’s darkness signal, telling the internal clock that night has begun and helping set the timing of sleep. Synthetic melatonin is sold as an inexpensive supplement in some countries and as a prescription medicine in others, and it has become one of the most widely used sleep aids.

Interest now reaches well beyond sleep. The nightly peak of melatonin tends to fall with age, which led researchers to ask whether topping it up could support healthy aging, and laboratory work describing it as an antioxidant drew attention from the longevity community. Use of melatonin supplements has also risen sharply over the past two decades, and with it new questions about long-term nightly use and about whether products contain what their labels state.

This review examines the evidence on what melatonin does and does not do for sleep, the body clock and markers of aging, together with its risks, interactions, dosing practices and product quality, so that health-focused adults can weigh its place in a long-term health strategy.

Benefits - Risks - Protocol - Conclusion

Expert articles, podcasts and overviews that discuss melatonin’s role, dosing and safety in depth.

  • Concerning findings on melatonin content in over-the-counter supplements - Angela Misic, Kathryn Birkenbach & Peter Attia

    Reviews a laboratory analysis showing most melatonin gummies were mislabeled and sets out Attia’s case for occasional, very low doses (about 0.3 mg) rather than nightly high-dose use.

  • RHR: The Melatonin Problem - Chris Kresser

    A 2026 podcast framing melatonin as a clock signal rather than a sedative, covering the heart failure association, label inaccuracy, use in children and light-based alternatives to nightly supplementation.

  • Melatonin - FoundMyFitness

    An evidence summary on age-related decline in melatonin, sleep effect sizes by population, antioxidant actions and blood pressure and neurodegeneration research, noting that sleep benefits are modest.

  • Melatonin: Benefits, Side Effects, and Research - Larisa Sheloukhova

    A longevity-oriented overview of melatonin’s discovery, dietary sources, antioxidant and immune roles and aging research, useful for mapping the preclinical rationale behind longevity interest.

  • A Case for Higher-Dose Melatonin - William Faloon

    Presents the high-dose position, with the author reporting nightly doses of 50 mg, later about 20 mg, for immune and cancer-related aims; Life Extension sells melatonin supplements, a direct financial interest in this position.

Andrew Huberman (hubermanlab.com) covers melatonin only within broader sleep episodes, so no Huberman item is listed; the five places went to sources dedicated to melatonin.

Grokipedia

Melatonin

An AI-written encyclopedia entry covering melatonin’s production, receptors, circadian role and supplement uses; useful as a broad orientation, though its claims need checking against primary sources.

Examine

Melatonin benefits, dosage, and side effects

Examine’s independent supplement page summarizes human trial outcomes for melatonin, including sleep onset, jet lag and other conditions, with dosing guidance and safety notes from a team without industry funding.

ConsumerLab

Melatonin Supplements Review & Top Picks

Independent laboratory testing of melatonin products, including gummies and time-release forms, for label accuracy and purity; full results and Top Picks require a paid membership.

Systematic Reviews

Five systematic reviews and meta-analyses (analyses pooling results across trials) covering melatonin’s sleep, body clock and cardiometabolic effects and its safety.

Mechanism of Action

Melatonin is made from serotonin in the pineal gland at night. Light reaching the eye suppresses its release through the suprachiasmatic nucleus (SCN, the brain’s master clock).

  • MT1 and MT2 receptors (melatonin’s two main cell-surface receptors): MT1 dampens wake-promoting signals from the SCN, easing sleep onset; MT2 shifts the clock’s timing. Evening doses move the clock earlier and morning doses later.
  • Sleep gating, not sedation: melatonin lowers core body temperature and opens the nightly sleep window rather than forcing sleep.
  • Antioxidant and mitochondrial actions: melatonin and its breakdown products neutralize reactive oxygen species (unstable molecules that damage cells) and raise antioxidant enzyme activity, the main longevity rationale, resting largely on laboratory and animal data.
  • Pancreatic signaling: MT2 receptors on insulin-producing beta cells reduce insulin release; a common variant of MTNR1B (the gene encoding MT2) raises type 2 diabetes risk and amplifies this effect (Qian et al., 2026).

Competing view: the “replacement” rationale assumes melatonin declines with age, but a systematic review found total 24-hour production may not change in healthy aging, only the nightly peak (Scholtens et al., 2016).

Pharmacology: about 15% of an oral dose reaches the bloodstream, peak levels arrive about 50 minutes after immediate-release doses, and the half-life is about 45 minutes (Harpsøe et al., 2015). It binds MT1 and MT2 selectively and crosses the blood-brain barrier. The liver enzyme CYP1A2 (which also breaks down caffeine), helped by CYP2C19 (a second drug-metabolizing liver enzyme), converts it to 6-hydroxymelatonin, excreted in urine as 6-sulfatoxymelatonin.

Historical Context & Evolution

Melatonin was isolated in 1958 by Aaron Lerner and colleagues, who were searching for a skin-lightening factor in the pineal gland (Lifespan.io). It was later recognized as the body’s darkness hormone governing circadian timing.

Interest in healthy aging grew when mouse studies reported longer survival with melatonin in drinking water (Pierpaoli & Regelson, 1994). Popular books in the mid-1990s promoted it as a way to slow aging. In the United States it is sold as a dietary supplement (National Center for Complementary and Integrative Health (NCCIH)), and adult use rose markedly between 1999 and 2018 (Li et al., 2022).

Later research complicated these claims: one mouse study found longer lifespan but more tumors (Anisimov et al., 2001), and human trials focused on sleep rather than aging. In 2007 the European Commission authorized a prolonged-release 2 mg tablet (Circadin) as a prescription medicine for insomnia in people aged 55 and over (European Medicines Agency (EMA)).

Meta-analyses confirmed small but consistent sleep effects (Ferracioli-Oda et al., 2013), while a 2017 American Academy of Sleep Medicine guideline suggested not using melatonin for chronic insomnia (Sateia et al., 2017); the society’s members do not profit from melatonin sales but bill for behavioral insomnia treatment and sleep testing. Newer evidence pulls both ways: a receptor gene variant was tied to worse glucose handling (Qian et al., 2026) and a heart failure signal appeared (American College of Cardiology news story), while bone (Amstrup et al., 2015) and cardiometabolic (Mohammadi et al., 2025) trials reported modest gains.

Expected Benefits

High 🟩 🟩 🟩

Faster sleep onset and modestly longer sleep ⚠️ Conflicted

Melatonin shortens time to fall asleep and slightly lengthens sleep across independent groups (Ferracioli-Oda and Bloch; Cruz-Sanabria and Faraguna; Maruani and Geoffroy), including longer daytime sleep, but not faster onset, after night shifts (Liira, Verbeek). Effects are larger with prolonged-release forms from age 55, doses near 4 mg, and dosing three hours before bed. A network meta-analysis (comparing many treatments together) found no material benefit on self-rated sleep quality in insomnia (De Crescenzo et al., 2022). Net reading: the benefit is consistent but small, well below hypnotics (sleeping medicines).

Magnitude: Sleep onset 7.06 minutes faster than placebo, with a 95% confidence interval (CI, the range likely to contain the true effect) of 4.37 to 9.75, and total sleep 8.25 minutes longer (95% CI 1.74 to 14.75) across 19 trials (Ferracioli-Oda et al., 2013); daytime sleep after night shifts 24 minutes longer than placebo (95% CI 9.8 to 38.9) across seven trials (Liira et al., 2014).

Faster resetting of the body clock after jet lag or with delayed sleep timing ⚠️ Conflicted

A Cochrane review (Herxheimer, Petrie) found nine of ten trials reduced jet lag after flights crossing five or more time zones, and an Australian trial (Sletten, Rajaratnam) advanced sleep onset in delayed sleep-wake phase disorder (a body clock running persistently late). A trial in 257 physicians (Spitzer et al., 1999) and a meta-analysis measuring only time to fall asleep (Buscemi et al., 2006) found no effect. Net reading: the review measuring jet lag symptoms directly is the stronger evidence, with benefit clearest when doses are taken at destination bedtime.

Magnitude: Number needed to treat (NNT, how many people must take it for one to benefit) of 2 for jet lag (Herxheimer & Petrie, 2002); sleep onset 34 minutes earlier than placebo (95% CI 8 to 60) in delayed sleep-wake phase disorder (Sletten et al., 2018).

Relief of irritable bowel syndrome symptoms ⚠️ Conflicted

In irritable bowel syndrome (IBS, recurrent abdominal pain with altered bowel habits), melatonin 3–6 mg daily reduced abdominal pain or overall symptom scores versus placebo in trials from independent groups in Singapore (Song et al., 2005), India (Saha et al., 2007) and Iran (Faghih Dinevari et al., 2023). A 2026 meta-analysis of seven placebo-controlled trials found fewer moderate-to-severe symptoms but no significant change in pain intensity, quality of life or sleep quality. The trials were small and short. Net reading: overall symptom relief is consistent; pain intensity gains are unconfirmed.

Magnitude: Moderate-to-severe IBS symptoms: risk ratio (RR, the risk relative to placebo) 0.52 (95% CI 0.39 to 0.71) versus placebo across pooled trials (Mahdi et al., 2026).

Less anxiety before surgery ⚠️ Conflicted

Taken shortly before surgery, melatonin lowered pre-surgery anxiety versus placebo in trials from independent groups in Saudi Arabia (Naguib & Samarkandi, 1999) and Turkey (Acil et al., 2004), and in a Cochrane review of 12 randomized trials (Hansen, Gögenur) rating this evidence high quality; two trials found it similar to midazolam (a benzodiazepine sedative). Two placebo-controlled trials found no difference, and effects after surgery were weaker. Net reading: the pooled trials support a short-term pre-surgery benefit despite the two null trials.

Magnitude: Pre-surgery anxiety 13.36 mm lower than placebo on a 0–100 mm visual analogue scale (a line on which patients mark their anxiety) (95% CI 10.58 to 16.13) across seven pooled trials (Hansen et al., 2015).

Medium 🟩 🟩

Higher bone density in postmenopausal osteopenia ⚠️ Conflicted

In a one-year Danish trial in 81 postmenopausal women with osteopenia (bone density below normal but above the osteoporosis range), nightly melatonin raised femoral neck bone mineral density (BMD, a validated marker of fracture risk) in a dose-dependent way. A six-month trial in 18 perimenopausal women (in the years around menopause) (Kotlarczyk et al., 2012) and a 2025 pilot (Jahan et al., 2025) found no difference between groups. No trial measured fractures. Net reading: one adequately sized trial supports a modest effect that smaller trials have not confirmed.

Magnitude: Femoral neck BMD rose 1.4% more than with placebo over one year, and 2.3% with 3 mg nightly; the abstract reports no confidence interval (Amstrup et al., 2015).

Fewer migraine days ⚠️ Conflicted

In a Brazilian trial of 196 adults (Gonçalves, Peres), 3 mg immediate-release melatonin reduced migraine frequency versus placebo. A Norwegian crossover trial (each participant receives both treatments in turn) of 2 mg prolonged-release melatonin in 48 adults found no benefit (Alstadhaug et al., 2010). A 2026 meta-analysis of nine small trials, several in children, found fewer headache days overall (Abouelmagd et al., 2026). Net reading: pooled trials favor immediate-release melatonin, but they are small and mixed in population, and the prolonged-release trial was null.

Magnitude: 1.54 fewer headache days per month than placebo (95% CI 0.58 to 2.50) across pooled trials (Abouelmagd et al., 2026); in the largest adult trial, migraine days fell by 2.7 per month with melatonin versus 1.1 with placebo (p = 0.009; p-value, the probability of a difference this large arising by chance alone) (Gonçalves et al., 2016).

Small improvements in cholesterol and blood pressure ⚠️ Conflicted

A 2025 meta-analysis of 63 trials (Mohammadi, Reiter) found small reductions in LDL (low-density lipoprotein, “bad” cholesterol) and systolic (upper-number) blood pressure. For nighttime hypertension, an earlier meta-analysis co-authored by scientists of Neurim Pharmaceuticals, maker of prolonged-release Circadin, found benefit only with controlled-release forms (Grossman et al., 2011); an independent review found no significant reduction (Lee et al., 2022). Net reading: effects are small and inconsistent across formulations.

Magnitude: LDL cholesterol 6.28 mg/dL lower (95% CI 2.03 to 10.53) and systolic blood pressure 2.34 mmHg lower (95% CI 0.55 to 4.13) than control (Mohammadi et al., 2025).

Better heart pumping function in heart patients ⚠️ Conflicted

In a 2026 meta-analysis of 14 randomized trials in people with heart disease (Ang, Iglesias), melatonin improved the change in left ventricular ejection fraction (LVEF, the share of blood the heart pumps out with each beat), most consistently around coronary bypass surgery. Final LVEF, troponin (a heart-damage marker) and heart-attack size did not improve consistently. A meta-analysis of four heart failure trials found no significant LVEF gain, though quality of life improved (Daliri et al., 2025). Net reading: a possible modest gain around heart surgery, unconfirmed in chronic heart failure.

Magnitude: LVEF change 3.95 percentage points greater than control (95% CI 1.70 to 6.20) across 14 trials (Ang et al., 2026); LVEF 2.39 points higher in heart failure (95% CI −1.82 to 6.59, not significant) (Daliri et al., 2025).

Lower risk of delirium after surgery ⚠️ Conflicted

Delirium (sudden confusion) after surgery fell in adults over 60 in a meta-analysis of 12 melatonin-only randomized trials (Borges, Nascimento), with larger effects at 5 mg or more; a network meta-analysis of melatonin-class drugs agreed (Luney et al., 2026). Certainty was rated low because trial results varied widely. A meta-analysis of intensive care trials found no reduction. Net reading: a likely benefit around surgery in older adults, not seen in critical illness.

Magnitude: Delirium in 22.4% (163 of 729) with melatonin versus 31.4% (230 of 732) with control, RR 0.67 (95% CI 0.52 to 0.88) (Borges et al., 2026); RR 0.89 (95% CI 0.73 to 1.09) versus placebo in intensive care (Lakbar et al., 2025).

Small cognitive gains in mild cognitive impairment ⚠️ Conflicted

In older adults with mild cognitive impairment (memory or thinking problems short of dementia) or dementia, two independent meta-analyses of randomized trials (Mdluli, Chang; Leung et al., 2025) found modest gains on the Mini-Mental State Examination (MMSE, a 30-point thinking test), clearest in mild cognitive impairment after 13–24 weeks. An earlier meta-analysis in Alzheimer’s disease found no cognitive improvement (Wang et al., 2017). Net reading: a small benefit in existing impairment, unconfirmed for protection in healthy aging.

Magnitude: MMSE 1.8 points higher than placebo (95% CI 0.9 to 2.7) across 10 randomized trials in adults 65 and older with cognitive impairment (Mdluli et al., 2025).

In a US health-record study of adults aged 50 and older (Jeong, Singh), repeated recorded melatonin use was linked to fewer new diagnoses of age-related macular degeneration (a leading cause of central vision loss in older adults) and less progression to its advanced “wet” form than in matched non-users. Laboratory studies suggest melatonin opposes several processes that drive the disease. The design is observational and open to confounding (bias from other differences) by lifestyle and health-care use, and no randomized trial has tested this outcome.

Magnitude: New macular degeneration diagnoses RR 0.42 (95% CI 0.28 to 0.62) and progression to the wet form RR 0.44 (95% CI 0.34 to 0.56) versus matched non-users; the abstract reports no per-group event rates (Jeong et al., 2024).

Low 🟩

Supportive care during cancer treatment ⚠️ Conflicted

Unblinded trials from a single hospital network reported lower one-year mortality with melatonin added to cancer therapy. A 2025 Cochrane review found less fatigue versus standard care alone but no difference versus placebo. Net reading: placebo-controlled data do not confirm the benefit.

Magnitude: Fatigue RR 0.46 (95% CI 0.39 to 0.55) versus standard treatment alone, but RR 1.02 (95% CI 0.90 to 1.17) versus placebo (Yu et al., 2025); one-year mortality RR 0.66 (95% CI 0.59 to 0.73) in unblinded trials (Mills et al., 2005).

Better recovery from COVID-19 ⚠️ Conflicted

Added to standard hospital care for COVID-19, melatonin raised clinical recovery in a meta-analysis of three randomized trials (Lan, Lai). A larger meta-analysis of nine randomized trials found no significant effect on mortality, symptom recovery or inflammatory markers. Net reading: no reliable benefit on hard outcomes.

Magnitude: Clinical recovery in 94.2% (81 of 86) with melatonin versus 82.4% (70 of 85) with control, odds ratio (the odds of recovery relative to control) 3.67 (95% CI 1.21 to 11.12) (Lan et al., 2022); mortality RR 0.72 (95% CI 0.47 to 1.11, not significant) versus control across five trials (Huang et al., 2023).

Relief of reflux symptoms

A systematic review of dietary approaches to gastroesophageal reflux disease (GERD, stomach acid flowing back into the esophagus) found symptom relief in one trial of melatonin combined with amino acids and B vitamins. The combination makes this evidence indirect for melatonin alone.

Magnitude: Not quantified in available studies. The review synthesized the single melatonin-containing trial qualitatively, without an effect estimate (Martin et al., 2022).

In a Brazilian trial in 40 women with endometriosis (uterine-lining tissue growing outside the uterus; Schwertner, Caumo), 10 mg nightly reduced pelvic and menstrual pain versus placebo. A Swedish 20 mg trial (Söderman, Marions) found no difference. Net reading: two small trials disagree, leaving the benefit unconfirmed.

Magnitude: Daily pain scores 39.80% lower than placebo (95% CI 12.88 to 43.01%) over eight weeks (Schwertner et al., 2013); mean pain 2.9 versus 3.3 with placebo on a 0–10 scale (p = 0.45) (Söderman et al., 2023).

Relief of chronic musculoskeletal pain ⚠️ Conflicted

For chronic musculoskeletal pain (pain in muscles, joints and bones), a 2026 meta-analysis of 23 randomized trials (Wu, Ferreira) found melatonin no better than placebo overall, but better in low-bias trials and after surgery. Most effects were below what patients notice. Net reading: a possible modest benefit, not confirmed overall.

Magnitude: Chronic pain 6.76 points lower than placebo on a 0–100 scale (95% CI −2.36 to 15.89, not significant), and 10.04 points lower (95% CI 2.39 to 17.68) in low-bias trials (Wu et al., 2026).

Fewer dream-enactment episodes in rapid eye movement sleep behavior disorder ⚠️ Conflicted

In rapid eye movement (REM) sleep behavior disorder (acting out dreams during sleep), a 2023 American Academy of Sleep Medicine guideline conditionally suggests immediate-release melatonin (Howell et al., 2023); its members bill for sleep testing, not melatonin sales. Two placebo-controlled prolonged-release trials found no benefit. Net reading: benefit unconfirmed.

Magnitude: Dream-enactment episodes 3.4 versus 3.6 per week with placebo, a difference of 0.2 (95% CI −3.2 to 3.6, not significant), with 4 mg prolonged-release melatonin in Parkinson’s disease (Gilat et al., 2020); no difference from placebo in clinical improvement with 2 or 6 mg (Jun et al., 2019).

Improved mood and depressive symptoms ⚠️ Conflicted

Melatonin has been proposed to lift mood through better sleep and clock timing. A systematic review of ten trials found no clear effect on depression scores (Hansen et al., 2014), and a later one found improvement only on one questionnaire. Net reading: randomized trials do not support a mood benefit.

Magnitude: Depression scores showed a standardized mean difference (SMD, the effect expressed in standard-deviation units) of −0.17 versus placebo (95% CI −0.38 to 0.05, not significant) across 19 trials (Li et al., 2022).

Lower body weight ⚠️ Conflicted

A 2021 systematic review of 23 placebo-controlled trials found lower pooled body weight with melatonin, though results varied widely between trials (Delpino & Figueiredo, 2021). A larger 2025 meta-analysis (Mohammadi et al., 2025) and an earlier one found no significant change. Net reading: no reliable weight benefit.

Magnitude: Body weight SMD −0.48 versus placebo (95% CI −0.94 to −0.02) (Delpino & Figueiredo, 2021); pooled SMD 0.09 (95% CI −0.17 to 0.34, not significant) across seven trials (Mostafavi et al., 2017).

Speculative 🟨

Antioxidant and anti-inflammatory effects ⚠️ Conflicted

Meta-analyses report lower interleukin-6 (IL-6, an inflammatory signaling protein) (Cho et al., 2021) and oxidative-stress markers (Mohammadi et al., 2025); only the latter found lower C-reactive protein. Net reading: inconsistent, clinically unvalidated markers.

Lower liver enzymes in fatty liver disease ⚠️ Conflicted

Meta-analyses in fatty liver disease not caused by alcohol found lower liver enzymes but disagreed on alanine aminotransferase (a liver-damage marker) (Mansoori et al., 2020; Akhavan Rezayat et al., 2021). Net reading: inconsistent, unvalidated markers.

Longer lifespan

Nightly melatonin extended survival in aging mice (Pierpaoli & Regelson, 1994), and another mouse study found longer lifespan alongside more tumors (Anisimov et al., 2001). The basis is animal data; no human lifespan data exist.

Benefit-Modifying Factors

  • Age 55 and older: In a six-month trial co-authored by Neurim scientists, prolonged-release melatonin shortened time to fall asleep in adults aged 55–80 but not across the full 18–80 range (Wade et al., 2011); older adults’ lower nightly peak may leave more room for benefit.
  • Baseline melatonin level: Low urinary 6-sulfatoxymelatonin (the main breakdown product, a measure of nightly output) did not predict response to prolonged-release melatonin in a Neurim-co-authored trial, so measuring one’s own levels does not identify likely responders (Wade et al., 2010).
  • Genetic polymorphisms: No variant has been shown to predict sleep benefit. Lower CYP1A2 activity raises melatonin exposure and may strengthen effects; MTNR1B risk-variant carriers face more glucose impairment, which shifts their benefit-to-risk balance (Qian et al., 2026).
  • Sex: Trials have not reported consistent sex differences in sleep benefit. Oral contraceptives raise melatonin levels, which may enlarge both effect and drowsiness in women using them (Harpsøe et al., 2015).
  • Beta-blocker use: Beta-blockers (blood pressure drugs such as atenolol and metoprolol) suppress nightly melatonin; 2.5 mg melatonin added 36 minutes of sleep versus placebo in hypertensive users (Scheer et al., 2012).
  • Pre-existing conditions: Benefit is clearest for jet lag and delayed sleep timing, small for chronic insomnia, and little or none, on low-certainty evidence, for sleep problems in Alzheimer’s disease (McCleery & Sharpley, 2020).
  • Older end of the age range: Even at 75 and over, the remaining nightly peak is lower (Scholtens et al., 2016); a physiological 0.3 mg dose restored sleep efficiency in insomniacs over 50 (Zhdanova et al., 2001).

Potential Risks & Side Effects

High 🟥 🟥 🟥

Next-day drowsiness, headache, dizziness and nausea ⚠️ Conflicted

Reported adverse events, usually mild, include daytime sleepiness, headache, dizziness and nausea (NCCIH). They, with lowered body temperature, were documented across 37 sleep-disorder trials (Besag, Wong) and, at 10 mg or more, in a pooled analysis of low-bias trials (Menczel Schrire, Hoyos). A review of 17 short-term trials (Buscemi et al., 2006) and a six-month 2 mg prolonged-release trial (Wade et al., 2011) found no excess over placebo. Net reading: excess events are clearest at higher doses, which leave melatonin elevated into daylight (Zhdanova et al., 2001).

Magnitude: Adverse events such as drowsiness, headache and dizziness: rate ratio (how many times more often events occurred) 1.40 (95% CI 1.15 to 1.69) versus placebo at 10 mg or more, with per-group rates not reported in the abstract (Menczel Schrire et al., 2022); daytime sleepiness in 1.66% of trial participants, reported without a comparison-group figure (Besag et al., 2019).

Medium 🟥 🟥

Impaired glucose tolerance when doses overlap meals ⚠️ Conflicted

Melatonin reduces insulin release via pancreatic MT2 receptors. In controlled studies from Murcia and Boston (Garaulet, Scheer), 5 mg impaired glucose tolerance, in intravenous testing only in MTNR1B risk-variant carriers (Qian et al., 2026); in Aarhus (Lauritzen, Støy), 10 mg for three months reduced insulin sensitivity in type 2 diabetes. Meta-analyses found lower fasting glucose (Mohammadi et al., 2025) and improved HbA1c (average blood sugar over three months) and insulin resistance (Dias et al., 2026). Net reading: direct tests show harm when melatonin overlaps meals, which fasting markers miss.

Magnitude: Evening glucose area under the curve (total glucose exposure after the test drink) 54% higher than with placebo after 5 mg (Rubio-Sastre et al., 2014); insulin sensitivity 3.6 versus 4.1 mg/kg/min with placebo, a 12% reduction, after three months of 10 mg (Lauritzen et al., 2022).

Menstrual cycle changes

In a six-month placebo-controlled trial in 18 perimenopausal women, 3 mg nightly reduced the number of menstrual cycles and lengthened the interval between them, consistent with melatonin’s known influence on reproductive hormones. This is a single small trial.

Magnitude: 4.3 versus 6.5 menstrual cycles over six months with placebo, and 51.2 versus 24.1 days between cycles (p < 0.05) (Kotlarczyk et al., 2012).

Low 🟥

Heart failure signal with long-term use ⚠️ Conflicted

A 2025 conference abstract linked over a year of use to heart failure in insomnia patients, an observational finding open to confounding. A survey found no association (Khan & Lima, 2026); pooled heart-patient trials showed improved heart function (Ang et al., 2026). Net reading: unconfirmed signal.

Magnitude: Incident heart failure in 5% (3,021) of users versus 3% (1,797) of matched non-users over five years, hazard ratio (HR, the relative rate over time) 1.89, absolute difference 1.9% (American College of Cardiology news story).

Worsening of seizures ⚠️ Conflicted

A 1998 case series in neurologically disabled children reported more seizures on melatonin (Sheldon, 1998). A meta-analysis of 10 add-on epilepsy trials found improved seizure severity. Net reading: controlled trials do not support the concern.

Magnitude: Seizure severity RR 0.33 (95% CI 0.04 to 0.62) versus placebo in add-on trials; the case series had no comparison group (Liu et al., 2024).

Fracture signal in middle-aged and older adults ⚠️ Conflicted

In UK primary care, adults aged 45 and older prescribed melatonin had more fractures than matched non-users, an observational finding open to confounding (Frisher et al., 2016). A later cohort found no excess hip fractures versus benzodiazepine users. Net reading: unconfirmed signal, possibly via drowsiness-related falls.

Magnitude: Fracture HR 1.44 (95% CI 1.01 to 2.04) versus matched non-users after adjustment, per-group rates not reported in the abstract (Frisher et al., 2016); hip fracture in 41 versus 51 of 4,519 people per group, HR 0.78 (95% CI 0.51 to 1.17) versus benzodiazepine users (Yang et al., 2023).

Nightmares, agitation and mood changes

Vivid dreams, nightmares, agitation and mood swings appear sporadically in trial reports, usually resolving on stopping. Reviews list them without pooled rates.

Magnitude: Not quantified in available studies. The reviews describe these events narratively, without per-group rates (Besag et al., 2019).

Heart rhythm disturbances

Palpitations are listed among the rare clinically significant adverse events in sleep-disorder trials (Besag et al., 2019). A 2026 systematic review collected case reports of arrhythmia (irregular heartbeat) and heart block (interrupted electrical signals between heart chambers), mostly resolving after stopping. The evidence consists of isolated reports.

Magnitude: Not quantified in available studies. Only case reports exist, with no comparison group (Sola et al., 2026).

Speculative 🟨

Tumor promotion

In female mice, long-term melatonin increased spontaneous tumor incidence despite extending lifespan (Anisimov et al., 2001). The basis is animal data only.

Immune activation in autoimmune disease

Melatonin stimulates some immune functions, raising a theoretical concern for autoimmune conditions. The basis is mechanistic, with no controlled human data.

Risk-Modifying Factors

  • MTNR1B risk variant: Carriers of the G allele (rs10830963) showed worse glucose tolerance and blunted insulin responses after 5 mg, while non-carriers did not (Qian et al., 2026).
  • CYP1A2 activity: Drug-induced low CYP1A2 activity (fluvoxamine) multiplies exposure and smoking lowers levels (Harpsøe et al., 2015); inherited low-activity variants would act similarly (theoretical).
  • Baseline glucose and HbA1c: In men with type 2 diabetes, 10 mg nightly for three months reduced insulin sensitivity by 12% (Lauritzen et al., 2022); people with prediabetes may be similarly affected (theoretical).
  • Sex: Oral contraceptives raise melatonin levels; perimenopausal women showed menstrual cycle changes (Kotlarczyk et al., 2012).
  • Age: Melatonin may stay active longer in older adults, causing daytime drowsiness (NCCIH); next-morning sleepiness may add to fall risk at the older end of the range (theoretical).
  • Epilepsy and anticoagulant use: Medical supervision is advised for people with epilepsy or taking anticoagulants (drugs that slow blood clotting) (NCCIH).
  • Critical illness and liver impairment: Critically ill patients show compromised melatonin elimination (Harpsøe et al., 2015); in liver impairment, reduced clearance markedly raises daytime melatonin levels, and the Circadin label does not recommend use.

Key Interactions & Contraindications

Prescription drugs:

  • Fluvoxamine and quinolone antibiotics (ciprofloxacin), CYP1A2 inhibitors: Avoid fluvoxamine: it raised melatonin exposure 17-fold, risking prolonged sedation (Härtter et al., 2000); citalopram had no effect (von Bahr et al., 2000). Caution with quinolones, which the Circadin label states may increase exposure. Mitigation: a non-inhibiting antidepressant.
  • Sedative-hypnotics (sleep-inducing drugs: zolpidem, eszopiclone, benzodiazepines such as temazepam): Caution. Prolonged-release melatonin worsened zolpidem’s impairment of coordination and driving performance, without changing blood levels of either drug, in a Neurim-co-authored trial (Otmani et al., 2008). Mitigation: separate nights or the lowest doses of each.
  • Calcium channel blockers (blood pressure drugs such as nifedipine): Caution. Evening melatonin 5 mg raised 24-hour blood pressure by 6.5/4.9 mmHg in patients controlled on nifedipine (Lusardi et al., 2000). Mitigation: home blood pressure checks for two to four weeks.
  • Anticoagulants (warfarin): Monitor. Case reports suggest people taking warfarin may come to harm, with altered clotting control the presumed consequence (Herxheimer & Petrie, 2002). Mitigation: INR (international normalized ratio, a clotting-time test) check within two weeks of starting.
  • Glucose-lowering drugs (metformin, sulfonylureas, an older diabetes drug class, such as glimepiride, insulin): Monitor (theoretical). Melatonin’s reduction of insulin sensitivity could blunt glucose control (Lauritzen et al., 2022). Mitigation: dosing well after meals and tracking fasting glucose.
  • Hormonal contraceptives and estrogens (ethinylestradiol): Monitor. Oral contraceptives raised melatonin exposure four- to five-fold by slowing its breakdown, which could prolong its effect, though alertness did not change measurably (Hilli et al., 2008). Mitigation: the lowest available dose.
  • CYP1A2 inducers (drugs that speed up this enzyme, such as carbamazepine and rifampicin): Monitor. The Circadin label states they may lower melatonin levels, weakening its effect. Mitigation: a dose review if the sleep benefit fades after starting them.
  • Cimetidine (an acid-reducing drug): Caution. The Circadin label states it raises melatonin levels by slowing its breakdown, risking prolonged drowsiness. Mitigation: the lowest available dose.
  • Immunosuppressants (anti-rejection and immune-dampening drugs such as tacrolimus, cyclosporine): Caution (theoretical). Melatonin’s immune-stimulating actions could oppose them, risking reduced immunosuppression. Mitigation: use limited to cases with the prescriber’s agreement.

Over-the-counter medications:

  • Sedating antihistamines (diphenhydramine, doxylamine): Caution (theoretical). Additive sedation and next-day impairment. Mitigation: no same-night combination.
  • Alcohol: Avoid. The EMA advises avoiding alcohol before, during and after taking prolonged-release melatonin (EMA), and the Circadin label states that alcohol reduces its effectiveness on sleep. Mitigation: no alcohol within several hours of dosing.
  • Caffeine: Monitor. Caffeine (600 mg) more than doubled melatonin exposure by blocking CYP1A2 (Härtter et al., 2003), while its stimulant effect counteracts sleepiness. Mitigation: no caffeine after midday when melatonin is used for sleep.

Supplements:

  • Sedating supplements (valerian, kava, magnesium, L-theanine, glycine, cannabidiol): Caution (theoretical). Additive sedation, with kava also adding liver strain. Mitigation: introducing one product at a time.
  • Blood pressure-lowering supplements (garlic extract, potassium, beetroot nitrate): Monitor (theoretical). Additive lowering of blood pressure, risking light-headedness. Mitigation: home blood pressure checks.
  • Glucose-lowering supplements (berberine, chromium): Monitor (theoretical). Opposing or unpredictable effects on glucose control. Mitigation: tracking fasting glucose.

Other interventions:

  • Tobacco smoking: Monitor. Smoking induces CYP1A2; oral melatonin exposure in smokers was about one-third of that after a week without smoking (Ursing et al., 2005), weakening effect. Mitigation: dose reduction after quitting, when levels rise.
  • Bright-light therapy and evening screens: Monitor (theoretical). Morning light and evening melatonin advance the clock together; evening bright light opposes melatonin, reducing its effect. Mitigation: dim evening light around dosing.

Populations who should avoid Melatonin:

  • Pregnant or breastfeeding women, because safety research is lacking (NCCIH)
  • Children, except under medical supervision, given rising accidental ingestions and related hospitalizations (Lelak et al., 2022)
  • People with epilepsy or taking anticoagulants, unless medically supervised (NCCIH)
  • People with dementia using it for sleep, per a 2015 American Academy of Sleep Medicine guideline recommending against it (Auger et al., 2015; NCCIH); the society’s members do not profit from melatonin sales but do bill for behavioral insomnia treatment and sleep testing
  • People hypersensitive to melatonin or product ingredients (EMA)
  • People with autoimmune disease, for whom the Circadin label does not recommend use because no clinical data exist
  • People with liver impairment, for whom the Circadin label does not recommend use because reduced clearance markedly raises melatonin levels

Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

  • Lowest effective dose: Common starting doses are 0.3–1 mg; 0.3 mg restored sleep efficiency, while 3 mg caused hypothermia (lowered body temperature) and daytime elevation (Zhdanova et al., 2001). Prevents next-day drowsiness, headache and dizziness.
  • Separation from meals: Dosing at least two to three hours after the last meal avoids overlap, since melatonin near meals impairs glucose tolerance (Rubio-Sastre et al., 2014). Mitigates impaired glucose tolerance, especially in MTNR1B risk carriers.
  • Glucose checks for those at metabolic risk: Fasting glucose and HbA1c before starting and after three months, with use reviewed if values rise from baseline. Mitigates impaired glucose tolerance in people with prediabetes or diabetes.
  • Verified products: Third-party-verified products reduce dose uncertainty, since most tested products deviated from label claims (Erland & Saxena, 2017). Prevents unintended overdosing and serotonin contamination.
  • No fluvoxamine combinations: Combining with fluvoxamine is avoided (Härtter et al., 2000), and quinolone antibiotics such as ciprofloxacin are combined only with caution. Prevents excessive exposure and prolonged sedation.
  • Caution with driving and alcohol: Driving or operating machinery is approached with caution after dosing, and alcohol is avoided before, during and after use (EMA). Prevents drowsiness-related accidents.
  • Correct timing for jet lag: Doses at destination bedtime; early-day doses cause sleepiness and delay adaptation (Herxheimer & Petrie, 2002). Prevents worsened jet lag and daytime sleepiness.
  • Periodic reassessment of long-term use: Nightly use reviewed every 3–6 months, with a pause to test whether it is still needed. Limits cumulative exposure while the heart failure signal remains unresolved.
  • Child-proof storage: Products kept out of children’s reach; pediatric ingestions rose 530% over a decade (Lelak et al., 2022). Prevents accidental poisoning of children in the household.
  • Blood pressure checks with nifedipine: Home blood pressure readings for two to four weeks after starting (Lusardi et al., 2000). Mitigates loss of blood pressure control.

Therapeutic Protocol

Timing, titration steps and cycling below reflect common practice unless cited.

  • Physiological low dose: 0.3 mg about 30 minutes before bedtime restored sleep efficiency in adults over 50 (Zhdanova et al., 2001); Peter Attia popularized a ceiling of about 0.3 mg nightly (Attia).
  • Dose-optimized sleep approach: Pooled trial data suggest effects peak near 4 mg daily taken about three hours before desired bedtime (Cruz-Sanabria et al., 2024).
  • Prolonged-release regimen: 2 mg taken one to two hours before bedtime after food, for up to 13 weeks, per the Circadin label for adults 55 and over (EMA).
  • Jet lag: 0.5–5 mg near target bedtime at the destination; 5 mg brought faster, better sleep than 0.5 mg, with no gain above 5 mg (Herxheimer & Petrie, 2002).
  • Delayed sleep timing: 0.5 mg fast-release one hour before desired bedtime, with a fixed bedtime schedule, for four weeks (Sletten et al., 2018).
  • High-dose approach: William Faloon of Life Extension, a melatonin seller, reports moving from 50 mg to about 20 mg nightly (Life Extension). Safety data at 10 mg or more are limited (Menczel Schrire et al., 2022).
  • Best time of day: Evening only. For sleep, 30 minutes to three hours before bedtime; morning doses delay the clock and cause daytime sleepiness (Herxheimer & Petrie, 2002).
  • Half-life: About 45 minutes for immediate-release forms (Harpsøe et al., 2015); prolonged-release tablets release melatonin slowly over a few hours (EMA).
  • Single versus split dosing: A single nightly dose is standard; for middle-of-night waking, a prolonged-release form replaces splitting rather than a second immediate-release dose.
  • Genetic polymorphisms: MTNR1B risk carriers may prefer the lowest dose taken well after the last meal (Qian et al., 2026); people with low CYP1A2 activity may need lower doses.
  • Sex-based differences: No sex-specific dosing exists; women on oral contraceptives reach higher levels (Harpsøe et al., 2015), making the 0.3 mg physiological dose (Zhdanova et al., 2001) a common starting point.
  • Age: Adults 55 and older show the clearest prolonged-release benefit (Wade et al., 2011); the oldest adults may favor the lowest doses to limit morning drowsiness.
  • Baseline biomarkers: Endogenous melatonin output does not guide dosing (Wade et al., 2010); baseline fasting glucose and HbA1c identify people who need closer metabolic monitoring.
  • Pre-existing conditions: For beta-blocker users, 2.5 mg nightly improved sleep (Scheer et al., 2012); people with diabetes favor low doses away from meals; dementia-related sleep problems showed little or no response in low-certainty evidence (McCleery & Sharpley, 2020).

Discontinuation & Cycling

  • Short-term versus long-term use: Jet lag use is short-term; the Circadin label covers up to 13 weeks (EMA). Lifelong nightly use for longevity lacks long-term safety data.
  • Withdrawal effects: No withdrawal symptoms or rebound insomnia occurred after six months of prolonged-release melatonin (Wade et al., 2011); in beta-blocker users, faster sleep onset persisted the night after stopping (Scheer et al., 2012).
  • Tapering: Not required at usual doses, given the absence of rebound; people on 10 mg or more often step down over one to two weeks.
  • Cycling for efficacy: Not needed to maintain effect, since no tolerance appeared over six months (Wade et al., 2010). Some practitioners, such as Peter Attia and Chris Kresser, favor occasional rather than nightly use.

Sourcing and Quality

  • Label accuracy: In 31 supplements, melatonin content ranged from 83% below to 478% above the label, and 26% contained serotonin (Erland & Saxena, 2017).
  • Gummies: Only 3 of 25 US gummy products were within 10% of the labeled melatonin amount; one contained none and several contained cannabidiol (Cohen et al., 2023).
  • Third-party testing: USP Verified (United States Pharmacopeia) or NSF (NSF International) certification marks, or a pass in ConsumerLab testing (ConsumerLab), indicate that contents were checked against the label.
  • Synthetic versus animal-derived: Most supplements use synthetic melatonin (NCCIH); synthetic forms avoid the contamination concerns of animal pineal extracts.
  • Formulation: Immediate-release forms suit sleep onset and jet lag; prolonged-release forms suit night waking. Pharmaceutical products such as Circadin (Neurim Pharmaceuticals) offer regulated content where prescriptions are available (EMA).
  • Dose availability: Low doses (0.3–1 mg) are harder to find than 3–10 mg products; splitting scored tablets or liquid forms allows low dosing.

Practical Considerations

  • Time to effect: Sleep onset effects appear from the first nights; jet lag relief within the trip; prolonged-release benefits by three weeks (EMA); bone density changes only after a year (Amstrup et al., 2015).
  • Common pitfalls: Taking 5–10 mg when 0.3–1 mg may suffice, dosing right after a late dinner, taking it at the wrong time for jet lag, relying on gummies, expecting sedative-like effects, and continuing nightly indefinitely without reassessment.
  • Regulatory status: In the US, melatonin is regulated as a dietary supplement (NCCIH); in the EU, prolonged-release melatonin (Circadin) is a prescription medicine for insomnia in adults 55 and over (EMA).
  • Guideline positions: A 2017 American Academy of Sleep Medicine guideline suggests not using melatonin for chronic insomnia (Sateia et al., 2017); the society’s members do not profit from melatonin sales but do bill for behavioral insomnia treatment and sleep testing.
  • Cost and accessibility: Supplements are inexpensive and widely available, but several countries restrict melatonin to prescription use (NCCIH). Insurers rarely reimburse over-the-counter melatonin but cover behavioral therapy and hypnotics, a payer asymmetry that may structurally bias guidelines and research funding.

Interaction with Foundational Habits

  • Sleep: Direct and potentiating. Melatonin signals darkness, while evening light suppresses it. Practical considerations: dim evenings, limited screens and a regular sleep schedule support it, and it serves as a timing aid rather than a substitute for sleep habits, a view argued by Chris Kresser.
  • Nutrition: Direct and blunting for glucose. High melatonin during eating impairs glucose tolerance; late dinners worsened glucose responses, especially in MTNR1B risk carriers (Garaulet et al., 2022). Practical considerations: dinner finished three or more hours before dosing; pistachios and mushrooms are among the richest dietary sources (Lifespan.io).
  • Exercise: None for performance. A meta-analysis found no effect on aerobic or strength performance, though recovery markers improved (Drummond et al., 2024). Practical considerations: dosing before morning training risks drowsiness; timing relative to evening training has not been studied.
  • Stress management: Indirect. Melatonin reduced anxiety before surgery in a Cochrane review (Hansen et al., 2015), and a one-year trial found no change in nighttime cortisol (Jahan et al., 2025). Practical considerations: it pairs with wind-down routines but does not replace stress-reduction practices.

Monitoring Protocol & Defining Success

Baseline testing before starting covers fasting glucose and HbA1c, blood pressure, and, for warfarin users, INR; people pursuing bone benefits add a dual-energy X-ray absorptiometry (DXA, a low-dose X-ray scan that measures bone density) scan. A one- to two-week sleep diary records typical sleep onset time, night waking and morning alertness as a reference point.

Ongoing monitoring follows a cadence of 2 weeks (INR and blood pressure where relevant), 3 months (fasting glucose, HbA1c, blood pressure and lipids), then every 6–12 months, with DXA repeated after one year. Success is defined as faster sleep onset or quicker jet lag recovery without morning grogginess, with glucose and blood pressure unchanged or improved. A rise in fasting glucose or HbA1c from baseline, or persistent morning impairment, is a reason to lower the dose, move it further from meals, or stop.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Fasting glucose 70–99 mg/dL (standard reference range) Safety check: melatonin can impair glucose handling Morning draw after an 8-hour fast; no dosing within three hours of the prior evening meal; pair with HbA1c (Rubio-Sastre et al., 2014)
HbA1c 4.0–5.6% (standard reference range) Safety check for longer-term glucose effects No fasting needed; reflects about three months; most relevant for MTNR1B risk carriers and people with prediabetes (Lauritzen et al., 2022)
Blood pressure Below 120/80 mmHg, the “normal” category of the 2017 ACC/AHA guideline Marker expected to change (small decrease) and safety check with nifedipine ACC/AHA: American College of Cardiology/American Heart Association (Whelton et al., 2018); their members do not profit from melatonin sales but do bill for treating high blood pressure; home or 24-hour readings; no conventional laboratory reference range exists for blood pressure
LDL cholesterol Below 100 mg/dL, the “optimal” category of NCEP ATP III Marker expected to change (small decrease) NCEP ATP III: National Cholesterol Education Program Adult Treatment Panel III (NCEP, 2002); conventional laboratory reference range: below 100 mg/dL on most laboratory reports; fasting not required
INR (warfarin users only) No established target; track change from the individual’s own stable INR Safety check: case reports of altered clotting control Check within two weeks of starting or changing dose follows common practice; warfarin case reports noted by Herxheimer & Petrie, 2002
Femoral neck bone mineral density No established target; track change from the individual’s own baseline Marker expected to change (increase) DXA scan at baseline and after one year, when the trial benefit appeared (Amstrup et al., 2015)

Qualitative markers:

  • Time to fall asleep and night waking, from a sleep diary or wearable tracker
  • Morning grogginess, alertness and mood
  • Days needed to adjust after long-haul travel
  • Dream intensity and nightmares
  • Headache or migraine frequency

Emerging Research

  • Cognition in aging (ongoing): A 230-participant trial testing 5 mg nightly for nine months on episodic memory and blood Alzheimer’s biomarkers, recruiting, with estimated completion in February 2028 (NCT03954899). A positive result would extend the cognitive evidence from existing impairment to prevention; a null result would confine it to people already impaired.
  • Cardiovascular events and cancer in older adults (registered): A planned 10,000-person Shanghai trial of melatonin on total cancer and cardiovascular incidence, listed as not yet recruiting despite a 2021 start date (NCT04631341). Positive results would provide the first hard-outcome longevity evidence; null results would remove a core rationale.
  • Head-to-head insomnia drugs (ongoing): A 1,200-adult Phase 4 trial (testing already-marketed drugs) comparing 3–5 mg immediate-release melatonin with zolpidem, trazodone, doxepin, diphenhydramine and placebo over six months, completing 2032 (NCT07542756). A positive result would strengthen melatonin’s sleep grade; a null result would support the guideline position against it.
  • Behavioral therapy plus melatonin in older adults (ongoing): A 183-participant trial in adults 60 and older adding 3 mg melatonin to cognitive behavioral therapy for insomnia, measuring the Insomnia Severity Index, completing 2029 (NCT07695246). A positive result would support combined use; a null result would favor therapy alone.
  • Bone and high-impact exercise (completion date passed): The MelaOstrong trial of 5 mg plus osteogenic loading (brief high-force exercise that stimulates bone), listed as active, not recruiting, passed its September 2025 estimated completion date with no results posted (NCT04233112); its publication found no between-group bone density difference (Jahan et al., 2025).
  • Heart failure safety: The 2025 heart failure signal (American College of Cardiology news story) awaits peer-reviewed publication; an NHANES (US national health survey) analysis found no association (Khan & Lima, 2026), and a pathway through prolactin (a milk-production hormone) has been proposed (Savvidis et al., 2026). Randomized safety data would settle the question.
  • Genotype-guided use: Evidence that MTNR1B risk carriers respond with impaired insulin secretion (Qian et al., 2026) points toward genotype-informed dosing and meal-timing rules, which could either confine or broaden safe use.

Conclusion

Melatonin is the body’s own darkness signal, sold cheaply as a supplement in some countries and as a prescription medicine in others. For health-focused adults, its clearest value is as a timing tool: it reliably helps reset the body clock after long flights and with late-running sleep schedules, and it helps people fall asleep slightly faster. That sleep benefit is consistent but small, far short of a sleeping medicine, and seems to do little for dementia-related sleep problems.

Evidence beyond sleep is thinner, apart from consistent trials easing anxiety before surgery and irritable bowel symptoms. Modest gains in bone density, cholesterol, blood pressure, migraine frequency and thinking in people with memory problems rest on few or inconsistent trials, and claims that melatonin slows aging rest only on animal and laboratory work.

The main risks are next-morning drowsiness, headache and dizziness, plus worse blood sugar handling when melatonin overlaps with meals, especially in people carrying a common diabetes-linked gene variant. A heart failure link with long-term nightly use remains unconfirmed. Product labels are often inaccurate, so the true dose may differ widely from what is printed.

Financial interests sit on several sides. Much of the long-acting tablet research was co-authored by its manufacturer, supplement sellers promote high doses, the sleep-medicine society advising against melatonin for insomnia has members who bill for behavioral insomnia treatment, and heart societies setting blood pressure targets have members treating high blood pressure. Long-term safety of nightly use, particularly at higher doses, remains uncertain.

Top - Benefits - Risks - Protocol