Melatonin (N-acetyl-5-methoxytryptamine) has been credited with many properties since its discovery in 1958. Foremost among these are an ability to provide a biological signal for dawn and dusk, to change the timing of the ‘internal clock’, which for mammals is located in the suprachiasmatic nuclei of the hypothalamus, and to promote sleepiness or sleep. ^{Reference Arendt1,Reference Wehr, Duncan, Sher, Aeschbach, Schwartz, Turner, Postolache and Rosenthal2}

Endogenous melatonin

Melatonin is primarily synthesised by the pineal gland with a marked circadian rhythm normally peaking during darkness. Its major function is to act as a ‘photoneuroendocrine transducer’ – providing information about day length by its profile of secretion. The time of melatonin secretion, rising in the evening and falling during the morning hours, has been referred to as ‘biological night’. ^{Reference Arendt1,Reference Wehr, Duncan, Sher, Aeschbach, Schwartz, Turner, Postolache and Rosenthal2}

In humans peak values of melatonin concentration are associated with the lowest point in rhythms of core body temperature, alertness, mental performance and many metabolic functions, and with maximum sleep propensity. ^{Reference Rajaratnam and Arendt3} If sleep is timed to occur outside the biological night, its quality and duration are compromised. For these reasons endogenous melatonin is thought to facilitate and reinforce sleep and other aspects of night-time physiology, when it is appropriately timed. ^{Reference Dijk and Lockley4}

Abnormal rhythms of melatonin have frequently (but inconsistently) been reported in psychiatric disorder, with reports of advanced or delayed phase and a possible decline or increase in amplitude in depression. There appears to be no consensus as to precisely the abnormality present in various types of depression, or indeed if there is one. In most patients with seasonal affective disorder there is probably a delay in melatonin secretion; ^{Reference Lewy, Lefler, Emens and Bauer5} for other affective disorders, the pattern remains unclear.

Exogenous melatonin

Exogenous melatonin has both acute and phase-shifting effects on sleep ^{Reference Rajaratnam, Middleton, Stone, Arendt and Dijk6} and on more strongly endogenous circadian rhythms such as core body temperature. ^{Reference Arendt1} These effects depend critically on the time of administration. ^{Reference Arendt1} For example, taken in the afternoon or early evening exogenous melatonin will advance, and taken in the early morning it is reported to delay, the circadian clock. Melatonin treatment during night-time, when endogenous melatonin is secreted, has little effect on sleep but during daytime it induces sleepiness and promotes sleep. There is some evidence that these effects are dose dependent.

Melatonin has obvious therapeutic potential and there has been some success in treating circadian rhythm sleep disorders such as delayed sleep phase syndrome, non-24-h sleep–wake cycle (frequent in people with blindness with no light perception) and, less consistently, the sleep problems of jet lag and shift work. ^{Reference Arendt1} Limited data suggest that treatment with melatonin itself may be useful in various types of depression ^{Reference Serfaty, Osborne, Buszewicz and Raven7} particularly if rhythm abnormalities are present. ^{Reference Lewy, Lefler, Emens and Bauer5} It is of interest to note that a number of antidepressant drugs increase peripheral melatonin levels.

Why develop melatonin analogues?

Melatonin is available over the counter in the USA. However, in most other countries its availability is restricted, and only recently has one particular formulation been approved for prescription use in Europe (www.emea.europa.eu/humandocs/Humans/EPAR/circadin/circadin.htm). Other formulations are also being developed. Pharmacokinetic properties differ depending on formulation. Melatonin is probably safe in low doses.

The discovery, cloning and characterisation of two principal melatonin membrane receptors (Mel_1a and Mel_1b, now renamed MT₁ and MT₂), found in numerous locations both in the central nervous system (notably within the suprachiasmatic nuclei) and the periphery, has led to much activity designed to synthesise and evaluate indolic and non-indolic analogues specific for either or both of these receptors. ^{Reference Dubocovich8} The MT₃ receptor is an enzyme, quinone reductase 2, ^{Reference Dubocovich8} concerned with detoxification mechanisms (not considered here).

Dosage and timing considerations

It is curious and interesting that doses of agomelatine and ramelteon, the two compounds for which the most information is available, appear to be high compared with the effective dose range for melatonin used as a chronobiotic or soporific. The daily evening dose of agomelatine for depression is at present 25 mg, the usual dose of ramelteon appears to be 8 mg daily in the evening, with reports of 4–64 mg being effective for sleep latency, sleep efficiency and total sleep time. In contrast, melatonin is able to entrain the free-running rhythms of most people with blindness at a timed dose of 0.5 mg daily. When given in the early evening, before the endogenous secretion rises, it has soporific activity in healthy adults in the dose range 0.1–10 mg, comparable with temazepam (see Arendt ^{Reference Arendt1} ).

Melatonin must be timed appropriately for maximum efficacy. If melatonin agonists are, like melatonin, most efficacious when given during the ‘biological day’, the question arises as to whether the compounds are most suitable for the treatment of primary insomnia rather than circadian rhythm sleep disorders. The interest of agomelatine is that it addresses the issue of depression as a primary target. In limited comparisons with available antidepressant medication it appears to have advantages with respect to safety profile, tolerability and normalisation of sleep. However, the National Prescribing Centre website states that, should agomelatine be licensed, ‘On the basis of currently available evidence, initial use of agomelatine by prudent clinicians will be limited’ (www.npci.org.uk/blog/?p=38).

Comparisons with melatonin

There are few comparisons regarding the effects of melatonin for any of these analogues. The chronobiotic and thermoregulatory properties of agomelatine were investigated in humans and proved to be very similar to melatonin. The effects of these compounds on sleep also appear to be similar. However, there are insufficient data to decide whether the agonists are preferable to the ‘natural’ compound.

Conclusions

Melatonin has proved useful in the treatment of certain sleep disorders. Melatonin agonists provide the clinician with new tools to address insomnia and, in the case of agomelatine, depression; however, their use in circadian rhythm sleep disorders has not been fully explored. Their main advantage is their good safety profile. Whether efficacy is improved compared with the natural compound remains to be determined. Large randomised controlled trials in this area are warranted.