Pinealectomy in humans removes virtually all plasma melatonin (65). Other consequences of the operation consist of diffuse neurological problems that do not add up to a consistent functional effect as yet and may be more related to non-specific effects of operation. Some preliminary evidence suggests that pinealectomised humans are less ‘seasonal’ than healthy subjects, underlining the predominantly photoperiodic role of melatonin if confirmed (66). There is good evidence that the neural and biochemical pathways known to control pineal function in rats are similar in humans. Pathological or traumatic denervation of the pineal abolishes the plasma melatonin rhythm. Beta-adrenergic antagonists suppress melatonin production, and increased availability of norepinephrine and serotonin are stimulatory (for references see reviews (12, 21)). The melatonin content of pineals obtained post-mortem is related to the time of death with, as expected, higher values at night (67).
In a "normal" environment, melatonin is secreted during the night in healthy humans as in all other species. The average maximum levels attained in plasma in adults are of the order of 60 to 70 pg/ml when measured with high-specificity assays. The concentrations in saliva are approximately one third of those in plasma. Minimum concentrations in both fluids are usually below 5 pg/ml. The peak concentrations of melatonin in plasma normally occur between 0200 and 0400 hours. The onset of secretion is usually around 2100 to 2200 hours and the offset at 0700 to 0900 hours in adults in temperate zones. The appearance and peak levels of 6-sulfatoxymelatonin (aMT6s) in plasma are delayed by 1 to 2 hours, and the morning decline by 3 to 4 hours. The mean concentrations of plasma and saliva melatonin together with urinary 6-sulfatoxymelatonin (aMT6s) are shown in Fig. 3. There are strong correlations between the timing and amplitude of the plasma melatonin and urinary aMT6s rhythms, such that aMT6s is a useful measure of circadian phase in field situations (12, 21). In urine 50 to 80 per cent of aMT6s appears in the overnight sample (2400 to 0800 hours), and it is low but rarely undetectable in the afternoon and early evening. Possibly the most striking characteristic of the normal human melatonin rhythm is its reproducibility from day to day and from week to week in normal individuals, rather like an hormonal fingerprint. There is however a large variability in amplitude of the rhythm between subjects. A small number of apparently normal individuals have no detectable melatonin in plasma at all times of day (12, 21) .
Figure 3. Average concentrations of melatonin in plasma (black, average N=133), saliva (blue, average N=28) and 6-sulphatoxymelatonin (aMT6s) in urine (red, average N=88), all measurements by radioimmunoassay. Diagrammatic representation of mean normal values (healthy men and women over 18 years old) from the author's laboratory.
As stated previously, even domestic intensity light can suppress human melatonin production at night. Exposure to light during 'biological night' has been perceived as deleterious to health (for example in night shift work) (68). This hypothesis is based on the beneficial effects of melatonin in some situations (see later). For example, the progression or spontaneous appearance of cancer in animals is enhanced by continuous light (69) and in human breast cancer xenografts exogenous melatonin is reported to reverse this effect (70). Whether or not the suppression of endogenous melatonin has undesirable consequences in the long term remains to be evaluated.
The melatonin peak is closely associated with the nadir in core body temperature, together with maximum tiredness/fatigue, lowest alertness and performance (Fig. 4) (71). Causal links are suggested by a number of observations. For example, bright light at night suppresses melatonin, simultaneously increasing body temperature, alertness and performance, and decreasing sleepiness (72). Exogenous melatonin during the daytime acutely increases sleepiness and decreases core body temperature (73). This latter observation is dependent on posture: subjects must be seated or recumbent (74). The ovulatory rise in temperature during the menstrual cycle is associated with a reported decline in amplitude of melatonin, but the decline in melatonin is not a consistent observation.
Figure 4. Relationship of plasma melatonin to other major circadian rhythms. Note the close correspondence between the core temperature nadir and the melatonin peak. Sleep propensity closely follows the melatonin rhythm. Reproduced from Rajaratnam SMW and Arendt J. Lancet 358:999-1005, 2001 by permission.
Sleep deprivation does not abolish the melatonin rhythm and in dim light does not affect secretion (71). In controlled experimental conditions it is clear that the evening rise of melatonin corresponds closely to the opening of the 'sleep gate', following a period of wake maintenance which has been called the 'forbidden zone for sleep' (75). Few associations have emerged between melatonin production and sleep stages, with the exception of a relationship between the timing of sleep spindles and certain other EEG characteristics, and the circadian phase of melatonin (76). Possibly the best correlative evidence for a role of melatonin in human sleep is the appearance of daytime naps, in free-running blind subjects when the peak of melatonin (and of course the temperature nadir) occurs during the daytime (77).
The relationships of stress, exercise, and some other non-pharmacological interventions in modification of melatonin production are somewhat unclear and do not appear to play a major physiological role in humans.
Shortly after birth very little melatonin or aMT6s is detectable in body fluids. A robust melatonin rhythm appears around 6 to 8 weeks of life (78). The plasma concentration of melatonin increases rapidly thereafter and reaches a lifetime peak on average at 3 to 5 years old (79). The increment is much greater at night. Subsequently a steady decrease is seen, reaching mean adult concentrations in mid to late teens with the major decline occurring before puberty. Values remain relatively unchanged until 35 to 40 years, and a final decline in amplitude then takes place until (on average) low levels are seen in old age (see (12) for references). Exceptionally healthy elderly may not show this age-related decline. Reports of differences in secretion in adults with gender, height, or body weight are not consistent. However, the measured plasma concentrations of melatonin in children are probably related to body weight.
Although lower melatonin has been reported in precocious puberty and higher concentrations in delayed puberty and hypothalamic amenorrhea compared with age-matched controls (80, 81), these remain correlative not causal associations, and there is no good evidence for a causal role of melatonin in primate pubertal development. Ovarian suppression with a GnRH analogue in precocious girls is not accompanied by changes in melatonin secretion (82). However, induction of sexual development with estrogen was associated with a very rapid decline of melatonin metabolite excretion in one case report (83).
Circulating melatonin may or may not vary during the menstrual cycle, the existing data are inconsistent. There is (84) evidence for abnormal melatonin secretion in patients with pre menstrual tension .
Low melatonin is reported to associate (inter alia) with cardiovascular disease and diabetic autonomic neuropathology (85-87). Studies of intensive care unit patients have shown very abnormal rhythms- but the data are confounded by the concomitant medication.
Clearly, the importance of the pineal in humans depends on the importance of light in human physiology. It is reasonable to assume that the pineal conveys information concerning light-dark cycles for the organization of seasonal and circadian rhythms in humans as in animals.