Melatonin

Circadin (Sigma)
2 mg prolonged-release tablets
Approved indication: primary insomnia
Australian Medicines Handbook section 18.4

Melatonin is a hormone which is secreted by the pineal gland at night-time. Its secretion is part of the normal circadian rhythm and promotes sleep. This resulted in a theory that low concentrations of melatonin may be associated with difficulty sleeping.

A study of 59 volunteers and 517 patients with insomnia found that the patients had lower urinary concentrations of a melatonin metabolite. When 396 of the patients were given an evening dose of melatonin, those with lower concentrations of the metabolite had a greater clinical response than those with higher concentrations. They had a better quality of sleep and they found it easier to get to sleep. The following morning they were more alert than the patients with higher urinary concentrations.¹

Several randomised controlled trials then looked at using melatonin to treat primary sleep disorders. A meta-analysis of 16 of these studies found that melatonin was as well tolerated as placebo, but was not very efficacious. Patients given melatonin fell asleep 12 minutes earlier than those given a placebo. The effect was greater (39 minutes) in the small sub-group with delayed sleep phase syndrome. Melatonin did not increase sleep efficiency (the proportion of time in bed spent asleep) significantly more than placebo.²

Another meta-analysis looked at sleep disorders secondary to other conditions or sleep restriction, for example jet lag. It found no evidence that melatonin was of any benefit.³

The meta-analysis of primary insomnia concluded that larger controlled trials were needed.²One subsequent trial in general practice randomised 170 patients, over the age of 55 years, with primary insomnia to take 2 mg modified-release melatonin or placebo. After three weeks there was no significant difference in getting to sleep, but sleep quality and alertness the next day were significantly improved with melatonin.⁴

A similar trial in general practice randomised 177 patients to take 2 mg modified-release melatonin and 177 to take a placebo. After three weeks, patients given melatonin fell asleep approximately nine minutes faster than the placebo group. They also had greater improvements in their quality of sleep and morning alertness, however total sleep time was not significantly improved.⁵

Adverse events occurred in 37% of the patients given melatonin and 32% of the patients given placebo. The most frequently reported symptoms were headache, back pain, asthenia and pharyngitis.

Melatonin undergoes significant first pass metabolism and most of the dose is excreted in the urine as metabolites. This metabolism involves cytochrome P450 1A1, 1A2 and possibly 2C19. It may be inhibited by drugs such as cimetidine, fluvoxamine, oestrogen and the quinolones, and induced by smoking and drugs such as carbamazepine and rifampicin. Melatonin is not recommended for patients with liver impairment and the effect of renal impairment is unknown. As the half-life of melatonin is less than an hour a modified-release formulation is needed. After a meal it takes three hours to reach the maximum plasma concentration, so it is recommended that the modified-release tablet is taken one or two hours before bedtime and after food. Patients should not drink alcohol with melatonin, as alcohol may cause the immediate release of the drug from the modified-release formulation.

There appear to have been no direct comparisons with benzodiazepines, but the results of a separate placebo-controlled trial with zolpidem have been used to assess the relative efficacy of melatonin. Overall patients given zolpidem fall asleep sooner than those given melatonin, but both drugs improve sleep quality. Melatonin should not be used in combination with other hypnotics. Stopping melatonin does not appear to cause more withdrawal symptoms than placebo,⁴but its use is restricted to a maximum of three weeks. It can only be prescribed to patients with primary insomnia over the age of 55 years. Many of these patients will be disappointed with the effect, as only about 30% respond to treatment. When the placebo effect is discounted, nine people would need to be treated for three weeks for one person to have improved sleep quality and to function better the next morning.⁵

manufacturer provided additional useful information

The Transparency Score ( ) is explained in New drugs: transparency', Vol 37 No 1, Aust Prescr 2014;37:27.

Notes on references

At the time the comment was prepared, a scientific discussion about this drug was available on the website of the European Medicines Agency (www.emea.europa.eu).

At the time the comment was prepared, information about this drug was available on the website of the Therapeutic Goods Administration (www.tga.gov.au/industry/pm-auspar.htm).