Melatonin

What it is

Melatonin is a hormone made mainly by the pineal gland, mostly at night [1]. The daily rhythm of its release is generated by the suprachiasmatic nuclei, the body's central clock, and light can suppress or synchronise its production [1]. Supplements are taken in several forms. Immediate-release tablets reach peak blood levels in about 50 minutes [2]. Prolonged-release tablets, such as Circadin 2 mg, are taken once a day about 2 hours before bedtime [3]. Some trials have tested a lipid-based oral form and a transdermal jelly applied to the skin in critically ill adults [4].

What the research says

In a meta-analysis of 19 randomised controlled trials with 1,683 people, melatonin reduced the time to fall asleep by about 7 minutes compared with placebo and added about 8 minutes of total sleep time [5]. Overall sleep quality also improved, by a small amount [5]. In a trial of 354 adults aged 55 to 80 with primary insomnia, 2 mg of prolonged-release melatonin cut the time to fall asleep by 24.3 minutes, compared with 12.9 minutes on placebo [3].

For children with neurodevelopmental disorders, a meta-analysis of 21 trials with 984 children found that melatonin improved sleep onset and total sleep time compared with placebo, with tolerability no different from placebo [6]. A separate meta-analysis of four trials in children with autism (238 children) found shorter time to fall asleep and fewer night-time awakenings [7]. A 2026 systematic review of young children found trials showing improved sleep onset in children with neurologic conditions such as autism, but no efficacy data for children with typical development [8].

In people with cancer receiving standard treatment, a Cochrane review rated the evidence as moderate certainty that melatonin reduces fatigue: the risk of fatigue was about half that with standard treatment alone (risk ratio 0.46, 10 trials, 1,359 people) [9].

In adults aged 55 and older with primary insomnia, a trial of 170 people found that 2 mg of prolonged-release melatonin improved sleep quality and morning alertness compared with placebo [10].

In people with delayed sleep phase disorder, a meta-analysis of nine randomised trials (317 people) found that melatonin advanced the timing of the body's own melatonin rise by about 1.2 hours and sleep onset by about 0.7 hours [11]. In a meta-analysis of 12 trials, melatonin regulated sleep-wake patterns in blind people better than placebo [12]. In a 1986 crossover study of 12 healthy volunteers, 2 mg a day advanced the body's own melatonin release by 1 to 3 hours in five of them [13].

A meta-analysis of 11 randomised trials, drawn from four systematic reviews, concluded that oral melatonin probably reduces jet lag symptoms [14].

Melatonin did not help people taper off benzodiazepines. In six trials with 322 people (average age about 64), it made no difference to successful discontinuation, and its effect on sleep quality was inconsistent [15].

Doses used in studies

Doses vary widely. In a study of adults over 50, 0.3 mg restored sleep efficiency and nightly melatonin levels, while 3 mg kept levels high into the daylight hours [16]. A pharmacokinetic review covered doses from 0.3 mg to 100 mg across 22 studies [2]. A meta-analysis found that longer trials and higher doses gave larger effects on time to fall asleep [5].

Trials in adults with insomnia aged 55 and older used 2 mg of prolonged-release melatonin, one tablet about 2 hours before bedtime, for 3 weeks [3] [10]. One trial continued this for 6 months [17]. In children with neurologic conditions, the trials lasted a mean of 12.7 weeks (range 2 weeks to 2 years) [8].

Safety and interactions

Common side effects are mostly mild. A critical review of 50 controlled studies found that 24 reported at least one statistically significant adverse event; most were minor and short-lived, and fatigue, mood changes and effects on psychomotor and thinking performance were the most common [18]. In a trial of prolonged-release melatonin in older adults, most adverse events were mild, with no significant difference from placebo [17]. In a meta-analysis of 34 trials in children and adults with sleep or mental health disorders, tolerability did not differ from placebo [6]. The review found that most side effects could likely be avoided by timing doses to natural circadian rhythms [18].

No safe upper limit was found in the studies read for this page. At 3 mg, a pharmacological dose, one trial found hypothermia and melatonin levels that stayed raised into daylight [16].

Blood melatonin levels are changed by fluvoxamine, oral contraceptives, smoking and caffeine [2]. A systematic review reported that effects on blood pressure and heart rate may depend on dose, dose timing and possible interactions with antihypertensive drugs [18]. The studies read for this page did not report interaction data for sedatives, blood thinners or other sleep medicines.

Pregnancy and breastfeeding: a 2022 scoping review of 15 human studies found no major safety concerns in clinical use, and concluded that melatonin is probably safe, but it found no randomised trials of melatonin for sleep in these groups [19]. People who are pregnant or breastfeeding should check with a doctor first.

Children: in young children, melatonin is a leading cause of unsupervised medicine exposure and overdose in emergency departments, and prescriptions have risen; efficacy data for children with typical development are absent [8]. Parents should keep it out of children's reach.

Older adults: a narrative review describes a favourable safety profile in people over 65, but notes little data on prolonged use and several factors that raise the risk of side effects [20].

The studies read for this page did not report outcomes for people with kidney or liver disease, bleeding disorders or epilepsy. People with these conditions, and people taking sedatives or blood pressure medicines, should ask a doctor first.

Caveats and open questions

The benefit in time to fall asleep is modest. Compared with placebo, its absolute benefit is smaller than that of other drugs for insomnia [5]. The body-clock evidence comes mostly from small studies, including one with 12 volunteers [13]. Its use in children with autism rests on few trials [7]. For visually impaired children, there is no high-quality trial data [21]. In adults in intensive care, a Cochrane review found insufficient evidence on whether melatonin improves sleep [22]. The longest trial read for this page ran for 6 months [17], and a review notes little data on prolonged use [20]. In a meta-analysis of trials for benzodiazepine discontinuation, the results varied widely between trials [15].