Article Text

Download PDFPDF
▼Melatonin for primary insomnia?
  • Relevant BNF section: 4.1.1


Melatonin, a hormone produced by the pineal gland, has a key role in regulating circadian rhythms, most importantly, the sleep-wake cycle.1 Melatonin's action has led to its being tried as a treatment for a wide range of sleep disorders, such as jet lag, primary insomnia, sleep-wake cycle disruption and sleep problems in children with neuro-developmental disorders.26 Until recently, it had not been licensed in the UK for any indication. Prolonged-release melatonin (▼Circadin – Lundbeck) has now been licensed as a treatment for primary insomnia. Here we consider whether this product has a place in the management of people with this condition.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

  • Relevant BNF section: 4.1.1

About sleep problems in general

Sleep disorders are common. For example, in one UK longitudinal study, around one-third (870 of 2,363) of adults reported having one or more sleep problems at night and, of these, two-thirds continued to have such problems 12 months later.7 People with these problems should be advised about lifestyle adjustments to promote good sleep hygiene (e.g. moderate caffeine and alcohol consumption; regular sleeping hours; a suitable environment for sleep).8,9 Non-drug treatments for insomnia include cognitive and behavioural approaches, and relaxation therapy.8,9 Commonly used drug treatments include hypnotics such as benzodiazepines and so-called Z-drugs (i.e. zaleplon, zolpidem, zopiclone), as well as sedative antihistamines and antidepressants.8 Although hypnotics are effective in promoting sleep in the short term, long-term use has not been proven to be effective for patients with chronic insomnia, and there are concerns about dependence.8 Consequently, hypnotics should, in general, be given for no more than a few days, preferably intermittently, and only to alleviate acute distressing insomnia caused by short-lasting events, illnesses or upsets.8,10 Older people (i.e. those aged 60 years or over) are particularly likely to experience unwanted effects with hypnotic drugs, such as falls and cognitive impairment.11

What is primary insomnia?

There are multiple classifications and definitions of insomnia, which include those in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV);12 the International Classification of Diseases, 10th edition (ICD-10);13 and the International Classification of Sleep Disorders (ICSD).14 DSM-IV defines primary insomnia as “a complaint of difficulty initiating or maintaining sleep or of nonrestorative sleep that lasts for at least 1 month… and causes clinically significant distress or impairment in social, occupational, or other important areas of functioning”.12 In this definition, the sleep disturbance does not occur exclusively during the course of another sleep or mental disorder, and is not due to the direct physiological effects of a substance or general medical condition.12 ICD-10 refers to “nonorganic insomnia” rather than primary insomnia, and defines this as “a condition of unsatisfactory quantity and/or quality of sleep, which persists for a considerable period of time, including difficulty falling asleep, difficulty staying asleep or early final wakening”, and in which the insomnia dominates the clinical picture.13 Similarly, ICSD does not recognise a category of primary insomnia but lists three criteria needed to make the diagnosis of insomnia: complaint of difficulty in falling and maintaining sleep or early awakening, or chronically non-restorative or unqualified sleep; occurrence of insomnia in the presence of adequate opportunity and circumstance for sleep; and impairment due to overnight insomnia.15

About melatonin

Endogenous melatonin (N-acetyl-5-methoxytryptamine) is synthesised and secreted by the pineal gland from soon after the onset of darkness, with secretion peaking at 2a.m. to 4a.m. and falling during the second half of the night.16 The hormone is associated with the control of circadian rhythms and sleep regulation.16 It is thought to have an effect on sleep disorders primarily by acting at MT1 and MT2 receptor subtypes in the suprachiasmatic nucleus of the anterior hypothalamus.1,17

What is Circadin?

Circadin comprises a prolonged-release tablet containing 2 mg of melatonin.18 It is licensed as a monotherapy for the short-term treatment of patients aged 55 years or more who have primary insomnia characterised by poor quality of sleep.18 The recommended dose is 2 mg once daily, taken around 1–2 hours before bedtime and after a small quantity of food.10 The summary of product characteristics (SPC) recommends that the dose should be continued only for 3 weeks, a duration based on the main trial data.1,18

Clinical efficacy


A systematic review pooled data from 14 randomised controlled trials that had compared melatonin (various formulations, ranging from slow- to standard-release products, but not Circadin) with placebo in a total of 279 patients with primary insomnia.3 Outcome measures assessed in the trials were sleep onset latency (time from lying down to sleep to onset of sleep); sleep efficiency (time asleep as a proportion of total time spent in bed); sleep quality (perceived quality of sleep); wakefulness after sleep onset (time spent awake in bed after first falling asleep); total sleep time; and percentage of time spent in rapid eye movement (REM) sleep (i.e. dreaming). There was considerable heterogeneity among the studies due to differences in study design and in the patients recruited. Melatonin doses ranged from below 1 mg daily to 5 mg daily, and treatment duration ranged from a few days to 3–4 weeks. Most studies involved adults with primary insomnia, but two studies included only children (aged up to 18 years), and two studies involved patients with delayed sleep phase syndrome (as can occur in patients working shifts or following long-distance travel). Overall, patients treated with melatonin had a shorter sleep onset latency (weighted mean difference [WMD] -11.72 minutes, 95% CI -18.24 to -5.20), but this difference was not considered clinically important by the authors of the review. There was no difference between melatonin and placebo in the other outcome measures. Also, there was no significant difference between melatonin and placebo in terms of adverse events.

In a second systematic review, which included four double-blind randomised placebo-controlled trials of melatonin (doses ranging from 0.3 mg to 75 mg; not Circadin) involving a total of 48 adults aged 18–65 years with primary insomnia, only one of the trials provided some suggestion of an improvement in sleep quality following melatonin, but the reviewers concluded that a meaningful meta-analysis was impossible due to the lack of reported data.4

Trials of Circadin

Two primary-care based, double-blind randomised placebo-controlled trials, published after the systematic reviews discussed above, provide evidence on prolonged-release melatonin (Circadin) in the treatment of patients aged 55 years and over with primary insomnia.19,20 We know of no published trials that have compared Circadin with other non-drug or drug treatments in people with primary insomnia.

The first study involved 170 patients (mean age 68.5 years) with primary insomnia (as defined by DSM-IV) for at least 1 month and who also had consistent complaints of poor sleep quality at the end of a 2-week single-blind placebo run-in period (i.e. responders to placebo were excluded from randomisation).19 Patients were screened for concurrent hypnotic drug use; a positive screen led to immediate exclusion. Patients received prolonged-release melatonin 2 mg or placebo once daily at night for 3 weeks, followed by a 2-week single-blind run-out phase. The outcome measures used were changes in the Leeds Sleep Evaluation Questionnaire (LSEQ; see Box) and in sleep diaries (patient-assessed quality of sleep on 5-point rating scale from 1 [very bad] to 5 [very good]). The primary outcome measure was quality of sleep (derived from the mean of questions 4 and 5 of the LSEQ, which relate to quality of sleep).1 Outcomes were measured on the last 3 days of each stage of the study (run-in, treatment, run-out). The results of each period were averaged for each of the four parameters and the changes in each parameter from run-in placebo to treatment and to run-out placebo were calculated for each patient. Compared with patients on placebo, those treated with melatonin had more of an improvement from baseline on the primary outcome of quality of sleep (−22.5 mm vs. −16.5 mm), but this was of marginal significance (p = 0.047). They had also had greater improvement in morning alertness (−15.67 mm vs. −6.79 mm, p = 0.002).

The Leeds Sleep Evaluation Questionnaire (lSEQ)

The LSEQ is a validated outcome measure of sleep and daytime effects in studies assessing treatments for insomnia.19,20 It includes ten 100 mm visual analogue scales divided into four domains of sleep and morning behaviour: ease of getting to sleep; quality of sleep; hangover on awakening from sleep; and alertness and behavioural integrity the following morning.

In the second study, 354 patients (mean age 65.7 years) with primary insomnia (based on DSM-IV and ICD-10 classifications) and persistent symptoms at the end of a 2-week single-blind placebo run-in period received prolonged-release melatonin 2 mg or placebo once daily, 2 hours before bedtime for 3 weeks.20 Patients were excluded if they had taken hypnotics within the previous 2 weeks or any psychoactive treatment within the previous 3 months. The primary outcome measure was the proportion of patients who completed the study (the “full analysis set”) with an improvement of 10 mm from baseline in both the quality of sleep (QOS) and the “behaviour following wakening” (BFW) domains of the LSEQ. Results were reported for a total of 334 patients who completed the study (94% of total randomised; the other 6% had no outcome data at visit 3). In this group, more of those treated with melatonin reported an improvement of 10 mm or more in both the QOS and BFW domains of the LSEQ (26.0% vs. 15.2%, p = 0.014); and there was a similar finding for the intention-to-treat population (25.0% with melatonin vs. 14.0% with placebo, odds ratio 2.01, 95% CI 1.17 to 3.46; p = 0.011). The European Medicines Agency (EMEA) has classified this as a relatively small effect size.1

The use of an initial placebo run-in period in both these studies may have induced selection bias:1 excluding patients with the potential to respond to a placebo could obviously worsen the overall treatment outcome in the randomised placebo group, so exaggerating the relative benefit of Circadin.

UK marketing authorisation*

An application to the EMEA for a marketing authorisation for Circadin was made in 2000, but withdrawn in January 2002 following an adverse report from the Agency.1 The EMEA's principal concern at that time was that the efficacy of the product had not been demonstrated. Advice was given to the applicant on the deficiencies in the application and on what actions had to be taken. These included the recommendation that a further pivotal study should be conducted, ideally involving three arms, including an active control arm. However, no such study was performed. Instead, a double-blind randomised controlled trial comparing zolpidem (10 mg) with placebo in the improvement of sleep quality in insomnia patients aged 55 years or over was conducted,1 and the results compared with one of trials of prolonged-release melatonin described above.20 Indirect comparisons were considered to provide evidence that melatonin was comparable to zolpidem 10 mg.1 When finally granting a marketing authorisation, the EMEA noted that the overall results from the different studies submitted with the application “suggest that the product is efficacious with a small effect size and in a relatively small fraction of patients”. Despite its qualms about the data, the EMEA concluded that “the risk-benefit balance of Circadin, as monotherapy, in the short-term treatment of primary insomnia characterised by poor quality of sleep in patients who are aged 55 or over, was favourable”, and recommended the granting of the marketing authorisation.1

Unwanted effects

In clinical trials, the rate of discontinuation of treatment due to unwanted effects was similar between patients taking Circadin and those taking placebo (37% and 31%, respectively).1,1820 The most frequent unwanted effects in the trials (i.e. affecting over 1 in 100 patients but fewer than 1 in 10 patients) were headache, pharyngitis, back pain, and weakness; these were common with both Circadin and placebo.1,1820 According to the SPC, uncommon effects (i.e. in over 1 in 1,000 patients but fewer than 1 in 100 patients) defined as possibly, probably or definitely related to treatment with Circadin include irritability, nervousness, restlessness, insomnia, abnormal dreams, migraine, psychomotor hyperactivity, dizziness, somnolence, abdominal pain, constipation, dry mouth, hyperbilirubinaemia, hyperhidrosis, asthenia and weight increase. There was no evidence of dependence, withdrawal effects or rebound insomnia associated with melatonin use in the clinical trials.1

Cautions and contraindications

Circadin is not licensed for use in children and adolescents under 18 years due to a lack of data on safety and efficacy.18 The SPC states that the drug should be used only with “caution” in patients with impaired renal function, and avoided in those with hepatic impairment (due to decreased clearance of the drug). The SPC also states that melatonin is not recommended for use in women who are pregnant or breastfeeding. Melatonin is mainly metabolised by CYP1A enzymes, so there are potential drug interactions with other drugs, such as fluvoxamine, that are metabolised by the same pathway. The SPC states that the drug should be used only “with caution” in patients taking 5- or 8-methoxypsoralen, cimetidine, oestrogens and quinolone antibacterials, which increase concentrations of melatonin by inhibiting its metabolism. Enzyme induction by cigarette smoking and drugs such as carbamazepine and rifampicin may reduce plasma concentrations of the drug. According to the SPC, alcohol should be avoided because it reduces the effectiveness of Circadin on sleep. Circadin may cause drowsiness and has moderate influence on the ability to drive and use machines.

Unlicensed use of Circadin

In 2008, about 94,000 prescriptions for unlicensed melatonin-containing products were dispensed in primary care in England, at a cost of around £6.5 million.21 This compares with 6,500 prescriptions for Circadin at a cost of around £160,000.21 However, the Medicines and Healthcare products Regulatory Agency (MHRA) now advises that, since a licensed formulation of melatonin is available in the UK, this should be used wherever possible, even for unlicensed indications if such use is deemed suitable by the clinician.22 The MHRA currently requests all importers to provide evidence of special clinical need (e.g. the need for alternative dosage forms or strengths, or for an immediate release product) for imported melatonin products and states that prescribers should be aware that, whereas melatonin products are classed as medicines in the UK, this is not so in the USA (where they are classed as supplements). Melatonin products from the USA should, therefore, only be used if no other product will suffice, as they are not subject to pharmaceutical Good Manufacturing Practice requirements. Although routine notification to the MHRA about special clinical need is not required for unlicensed melatonin products manufactured in the UK to such standards, UK manufacturers of such products should have these details available in case they are requested by the MHRA, for example, during inspection.23


The basic cost to the NHS of a 3-week course of prolonged-release melatonin (2 mg once daily) is £10.77. By comparison, a 3-week course of temazepam 10 mg daily costs £0.60, zolpidem 10 mg daily £1.24, and zopiclone 7.5 mg daily £1.16.


Prolonged-release melatonin (▼Circadin) is licensed as short-term (3 weeks) treatment for primary insomnia in patients aged 55 years and over. In contrast to other hypnotic drugs, there appears to be no evidence of dependence, withdrawal effects or rebound insomnia associated with its use. However, while published trials suggest some benefit with Circadin compared with placebo for people with primary insomnia, the limited evidence on efficacy does not provide a convincing basis for recommending its use. Furthermore, Circadin has not been directly compared with other non-drug or drug treatments in published trials.


    [R=randomised controlled trial; M=meta-analysis]

  1. 1.
  2. M 2.
  3. M 3.
  4. 4.
  5. M 5.
  6. M 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. M 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. R 19.
  20. R 20.
  21. 21.
  22. 22.
  23. 23.


Linked Articles

  • Articles
    BMJ Publishing Group Ltd