| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Circadian, Neuroendocrine, and Sleep Disorders Section, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02115
Address all correspondence and requests for reprints to: Dr. E. B. Klerman, Circadian, Neuroendocrine, and Sleep Disorders Section, Brigham and Women’s Hospital, 221 Longwood Avenue, Boston, Massachusetts 02115. E-mail: ebklerman{at}hms.harvard.edu
Abstract
The daily rhythm of melatonin influences multiple physiological measures, including sleep tendency, circadian rhythms, and reproductive function in seasonally breeding mammals. The biological signal for photoperiodic changes in seasonally breeding mammals is a change in the duration of melatonin secretion, which in a natural environment reflects the different durations of daylight across the year, with longer nights leading to a longer duration of melatonin secretion. These seasonal changes in the duration of melatonin secretion do not simply reflect the known acute suppression of melatonin secretion by ocular light exposure, but also represent long-term changes in the endogenous nocturnal melatonin episode that persist in constant conditions. As the eyes of totally blind individuals do not transmit ocular light information, we hypothesized that the duration of the melatonin secretory episode in blind subjects would be longer than those in sighted individuals, who are exposed to light for all their waking hours in an urban environment. We assessed the melatonin secretory profile during constant posture, dim light conditions in 17 blind and 157 sighted adults, all of whom were healthy and using no prescription or nonprescription medications. The duration of melatonin secretion was not significantly different between blind and sighted individuals. Healthy blind individuals after years without ocular light exposure do not have a longer duration of melatonin secretion than healthy sighted individuals.
IN MAMMALS, melatonin has a number of documented influences, including reproductive changes (1), hypothermic responses (2, 3), hypnotic effects (4, 5, 6), and induction of phase shifts of the circadian system (7, 8, 9, 10, 11). Claims have also been made for immunoenhancing, anticancer, antiaging, and antioxidant effects.
In humans and other mammals, the observed daily pattern in melatonin levels reflects both an endogenous circadian rhythm, with levels high during biological nighttime and low during biological daytime, and an evoked response to nocturnal ocular light exposure, which causes a rapid reduction in melatonin levels for the duration of the light exposure (12, 13). Consistent with this light-evoked suppression, the duration of the nightly melatonin secretion decreases with increased duration of daytime light exposure (14, 15, 16, 17, 18). In photoperiodic animal species, this information about the number of hours of daylight is conveyed by the duration of melatonin secretion that, in turn, affects reproductive function (1). In hamsters, less than 12 h of daytime light exposure per day causes a reduction of testes size (19). However, if these rodents are maintained in total darkness or on long nights for more than a few months, the testes spontaneously recrudesce even at a time when the animals’ melatonin profile remains expanded (15, 20). Data from blinded animals are similar to those from animals maintained in constant darkness in measures of N-acetyltransferase activity (the rate-limiting enzyme for synthesizing melatonin), the amount of melatonin in whole pineal glands, and gonadal regression (21, 22, 23).
Over weeks of exposure to a specific photoperiod, the duration of melatonin secretion becomes an endogenous property of pineal secretion rather than only an evoked response to lighting conditions. In in vitro studies of the suprachiasmatic nucleus, the site of the mammalian circadian pacemaker that influences the timing of melatonin secretion, c-fos expression and multiunit neural activity reflect the prior in vivo photoperiod (24, 25, 26). In humans, similar photoperiodic responses have been reported. In one study the length of nightly endogenous melatonin secretion increased by more than 1.5 h when the light/dark conditions changed: melatonin secretion lasted approximately 10.3 h when individuals were studied under dim light conditions after several weeks of daily exposure to 8 h of dark and 16 h of natural and artificial light ("summer") and lasted approximately 11.9 h after 4 weeks of daily exposure to 14 h of dark and 10 h of natural and artificial lighting ("winter") (18). A shorter duration of the nocturnal melatonin profile during summer compared with winter in field studies of humans is consistent with this finding (27, 28). However, in these latter two field studies, the ambient light levels present during collection of samples for melatonin determinations were both sufficient to suppress melatonin secretion and different in the summer and winter data collections; these experimental conditions may have affected the results.
The reproductive system of humans may be sensitive to the duration of daily light exposure. Numerous studies have documented seasonality in births and, by implication, conception rates (29, 30, 31, 32, 33, 34, 35, 36, 37). In sighted women, the increased melatonin duration and daytime excretion during winter at northern latitudes were correlated with increased pituitary hormones (FSH and LH) during the luteal phase and decreased ovarian activity, as manifested by decreased estradiol at ovulation and during luteal phase, decreased testosterone during the luteal phase, and increased sex hormone-binding globulin throughout the menstrual cycle (28). Nonhuman mammals that display photoperiodism have specific melatonin receptors in the pars tuberalis (38). However, in eight of nine human brains examined postmortem, there were no melatonin receptors localized in the pars tuberalis (39). The mechanisms of any of these possible reproductive effects are unknown, including whether ocular light exposure information is conveyed via the duration of melatonin secretion in humans.
To continue evaluation of whether humans exhibit changes in melatonin
secretion with photoperiod, we considered an extreme example of
prolonged long night conditions: the absence of ocular light exposure
that is experienced by totally blind individuals. We hypothesized that
blind persons would have a greater duration of melatonin secretion than
sighted individuals exposed to a long history of normal light/dark
cycles. Previous studies in blind subjects have not verified total
blindness or have not performed frequent enough sampling to determine
the duration of the melatonin secretory episode. More stringent
criteria for total ocular blindness have been used recently (40, 41). For an individual to be classified as totally blind, he or
she must have no conscious light perception, no visual light reflexes,
and no evidence of light information transmission to the circadian
system (circadian photoreception), as evidenced by a lack of melatonin
suppression in response to an appropriately timed ocular bright
(
10,000 lux) light exposure (40). As sighted
individuals, including those in this study, had been exposed to both
natural and artificial lighting for several weeks, and even the level
of artificial indoor light exposure is sufficient to suppress melatonin
secretion (42, 43, 44), the sighted subjects were exposed to
effectively 8 h of darkness and 16 h of light before starting
the study. We analyzed the duration of the daily episode of plasma
melatonin secretion in healthy totally blind and sighted individuals
using General Clinical Research Center in-patient protocols.
Subjects and Methods
Seventeen totally blind subjects (aged 23–66 yr; 5 women and 12 men) and 157 sighted subjects (aged 18–81 yr; 43 women and 114 men) were studied. Of the 17 blind subjects, 8 were entrained (i.e. their circadian rhythms appeared to be phase-locked to the 24-h environmental cycle, probably through nonphotic stimuli), 7 were free-running, and the entrainment status of 2 was not known. Data from some of the subjects have been reported previously (40, 45, 46, 47, 48). All subjects were healthy, as determined by history, physical exam, and biochemistry and hematology screening. All denied using prescription or nonprescription medications; this was verified by urine toxicological examination on the day of admission. Blindness was verified by history, neuroophthalmological exam (including the absence of pupillary light reflexes), and the absence of melatonin suppression in response to exposure to approximately 10,000 lux of light at the time of the daily melatonin peak (40). All had been blind for at least 5 yr.
The protocol for each subject began with the subject living at home on a self-selected 8-h sleep and 16-h wake schedule. They were instructed not to vary sleep or wake times by more than 30 min for any sleep episode during the 3 weeks before admission, compliance of which was determined through analysis of wrist actigraphy and phone logs. Subjects were then admitted to a special suite in the Intensive Physiological Monitoring Unit or the Environmental Scheduling Facility of Brigham and Women’s Hospital General Clinical Research Center for 1–3 baseline days, during which their habitual sleep/wake schedule was maintained, with 16 h awake in room light (<15 or <150 lux during day of admission; <150 lux during remaining 2 days) and 8 h in bed in the dark (<0.03 lux). These suites have no external windows, and all lighting was controlled by the investigators. After the baseline days, all individuals were studied under constant routine (CR) conditions that included semirecumbent enforced wakefulness in dim light (<15 lux) with frequent small meals. This protocol was designed to minimize the effects of sleep or wake state, changes in posture or activity, or large meals on core body temperature, hormone concentrations, and other physiological variables (40, 49). During the CR, blood was drawn every 20–60 min through an indwelling iv catheter. Plasma melatonin was assayed by RIA (Diagnostech, Inc., Osceola, WI) with an assay sensitivity of 0.7–5 pg/mL (depending on the year in which the assay was performed), an intraassay coefficient of variation of 8%, and an interassay coefficient of variation of 13%. As ambient light levels and posture changes (12, 42, 43, 50, 51, 52) can affect melatonin levels, only those hormone samples collected during a CR were included in these analyses.
Within each CR, there was at least one complete episode of melatonin secretion in all subjects. The mean melatonin value for each subject was defined as the average value of h 5–29 of the CR. The maximum value was defined as the maximum melatonin level during the entire CR. The duration of the melatonin peak was defined in three ways: the total time plasma melatonin concentrations were greater than the 24-h mean level (45), the total time plasma melatonin concentrations were greater than half the maximum level during the CR, or the total time plasma melatonin concentrations were greater than the dim light melatonin onset level (10 pg/L = 43.05 pmol/L) (53). The integrated value of melatonin was computed by the trapezoid method and then normalized for the number of hours of melatonin collection. Groups were compared using unpaired t tests or general linear models ANOVA using SAS software (version 6.12, SAS Institute, Inc., Cary, NC). Standard power calculations were performed with the PS (Power and Sample Size Calculations) program (version 1.0.15).
For statistical purposes, we used a method for showing bioequivalence
described by Blackwelder (54). This method was selected
because we had an a priori assumption that the duration of
melatonin secretion would be longer in blind individuals. For this
method, the null hypothesis was: true mean of blind subjects >
true mean of sighted subjects + 
, where was an arbitrary small
number. For a first estimation, we chose as a meaningful a value of
1.0 h, which is approximately 1 SD of the
data (reported below). The test statistic was:
t = (observed mean of blind subjects - observed mean of
sighted subjects - )/SEM.
For this statistical test of bioequivalence, we used an 
of 0.1 and
a ß of 0.05, because the rules of type I and type II errors are
reversed in these statistics (54).
All protocols were reviewed and approved by the Brigham and Women’s Hospital institutional review board, and all subjects gave written informed consent.
Results
The patterns of melatonin secretion were similar in sighted and
blind subjects (Fig. 1
) when the data
were plotted relative to the upward crossing of the mean value. The
average, maximum, and integrated melatonin concentrations were not
significantly different in blind and sighted subjects (Table 1).
|
|
). There
was no significant difference in average duration or amplitude of the
melatonin secretory episode by all three measures of width among blind
subjects when they were subdivided into those with circadian rhythms
that were free running, entrained, or of unknown status. We did not
find a significant effect of age (F3,170 = 0.17;
P = 0.68) or a significant interaction between age and
visual status (blind/sighted; F3,170 = 0.000;
P = 0.94) on the duration of the melatonin profile
using the crossing of mean criteria, consistent with a previous report
that included data from some of these sighted subjects
(45).
According to standard power calculations, we had the power to detect a
difference of 0.2 h in melatonin duration using the crossing of
the mean criteria and 0.3 h using the crossing of half-maximum
criteria (Table 1). With the Blackwelder bioequivalence statistics, the
90% confidence interval for the difference in means for the duration
of the melatonin secretory episode using the crossing of mean criteria
was (-1.0,0.14). This indicates that the blind and sighted individuals
did not have significantly different durations of melatonin secretion.
The statistical procedure for these tests of the null hypothesis
indicates that given the variability in this population, we had the
power to detect a difference of = 0.35 h in the duration
of melatonin secretion (t 0.90,172;
P = 0.047). Similar results were obtained for the other
two measures of width. Therefore, we can reject the null hypothesis of
nonequivalence (duration in blind longer than that in sighted) and
conclude that the duration of the melatonin peak was not significantly
longer in blind than in sighted subjects.
Discussion
Contrary to our hypothesis, we found that the duration of the melatonin peak was not significantly longer in totally blind individuals than in sighted individuals. The average durations of melatonin secretion observed in this study suggest that prolonged absence of light exposure, including loss of both conscious vision and circadian photoreception, has little influence on the average duration of the melatonin rhythm in humans. However, it should be noted that the duration of melatonin secretion in these blind individuals before and immediately after they became blind is unknown, thus precluding evaluation of the acute response to loss of ocular regulatory input to the pineal gland. Other, currently unknown mechanisms may be involved in regulating the duration of the daily episode of endogenous melatonin secretion in humans, which has been shown to be increased after the individual’s exposure during 4 weeks of 14-h nights and the nonphotic stimuli associated with this schedule (18, 55). It is possible that the duration of melatonin secretion was shortened by nonphotic entrainment of circadian processes that regulate the duration of melatonin secretion, such as the postulated morning and evening oscillators (56, 57).
It is possible that the duration of melatonin secretion became longer in the sighted subjects during the 1–3 baseline days before the CR. However, this is unlikely because 1) the sighted subjects were exposed to a 150:0 lux 16:8 h photoperiod for only 1–3 days; and 2) 150 lux is capable of suppressing melatonin levels (42) and presumably thereby altering the duration of melatonin secretion. We also did not find any evidence of the seasonality of the duration of melatonin secretion in sighted subjects (data not shown). We believe that the absence of seasonality of the duration of melatonin duration observed in sighted individuals may be due to the ubiquitous use of indoor lighting at levels known to suppress melatonin and therefore exposure of all individuals to at least 16 h of light, 8 h of darkness during all seasons. Although it has been shown that healthy individuals are exposed to surprisingly low levels of light in their usual environments (58, 59), the light intensity required for a saturated circadian response curve to both phase shifting and melatonin suppression is in the range of typical indoor lighting (42). Therefore, it is likely that even these low levels of ambient light are sufficient to maintain circadian synchrony.
There was a near-significant trend (P = 0.06) toward a higher amplitude of melatonin secretion in sighted compared with blind individuals. In measures of duration of melatonin secretion that have an absolute threshold, such as dim light melatonin onset level, it might be expected that a larger amplitude of melatonin secretion would cause a longer calculated duration of nocturnal melatonin secretion due to a delay in the clearance of elevated melatonin concentrations. This would be expected to be less of a factor in calculations of melatonin duration that are based on a relative threshold, such as average value or half the maximal value. If blind individuals had a larger amplitude of melatonin secretion, then we would once again expect a longer duration of melatonin secretion in blind compared with sighted subjects. However, this was not observed.
The observed melatonin rhythm includes both an endogenous rhythm that may reflect a prior photoperiod and acute changes dependent on the conditions under which the melatonin samples are collected. Whether these chronic or acute changes affect the general health of blind subjects is unknown. Given the demonstrated relationships between the duration of melatonin secretion and reproductive function in some species, one possible change in blind subjects would be related to reproductive function. Reports of changes in fertility of blind individuals based on questionnaires and surveys vary from no effect (60), to impaired fertility (61), early menarche (62, 63), or no changes in the timing of menarche (64). Difficulties in interpreting these results arise from the lack of documentation of whether individuals were totally blind, with no light perception and no circadian photoreception. In seasonally reproductive animals, it is not necessarily the absolute duration of melatonin secretion that regulates reproductive function, but the relative change in this duration. Thus, the same duration can stimulate or inhibit reproduction depending on the prior photoperiodic history of the animal (65). As discussed above, we do not know the duration of the melatonin secretory episode before the individuals became blind or whether the duration was altered after the individual became blind. Therefore, we cannot at present hypothesize a relationship between a change in reproductive competence of subclinical changes in reproductive function or other physiological functions due to a change in the duration in melatonin secretion.
Animal studies comparable to these studies in humans have not been performed. Experiments in seasonally breeding animals are needed to determine the duration of melatonin secretion (requiring frequent sampling of blood for hormone analysis) and changes in reproductive function under conditions of extended constant darkness and in blinded animals compared with summer conditions. These results may elucidate the mechanisms involved in melatonin secretion and the photoperiodic reproductive response in both seasonally reproductive animals and humans.
In conclusion, although sighted humans may have photoperiodic responses to changes in light and dark duration over 1 month, we did not find a difference in the duration of melatonin secretion in healthy blind individuals who had been blind for at least 5 yr. The baseline melatonin duration of humans may be similar to that seen during summer conditions. We found that the duration of melatonin secretion in both blind and sighted subjects, using the crossing of 43.05 pmol/L criteria, was about 9.4 h, shorter than the data from summer conditions reported by Wehr et al. (18), who used similar crossing criteria. Exposure to winter conditions may temporarily increase the duration of melatonin secretion and alter reproductive competence. However, under prolonged conditions of darkness, melatonin secretion duration may revert to that during summer-like photoperiods. The mechanisms and implications of this finding merit further exploration.
Acknowledgments
We thank Dr. K. P. Wright for data from some of the sighted subjects and Dr. R. J. Hughes and J. Olivo for comments on the manuscript.
Footnotes
1 This work was supported in part by NIMH Grants 1-R01-MH-45130 and
MH-18825-12 (to J.M.Z.), NIA Grants 1-R01-AG-06072 and K01-AG-00661 (to
E.B.K.), NIH General Clinical Research Center Grant
NCRR-GCRC-M01-RR-02635, a fellowship from Grant T32-HL-07901 (to
J.F.D.), NASA Cooperative Agreement NCC9-58 with the National Space
Biomedical Research Institute, NHLBI Grant F33-HL-09588 (to
S.B.S.K.). 
Received November 17, 2000.
Revised January 12, 2001.
Revised February 27, 2001.
Accepted March 9, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  J. A. Knight, S. Thompson, J. M. Raboud, and B. R. Hoffman Light and Exercise and Melatonin Production in Women Am. J. Epidemiol., December 1, 2005; 162(11): 1114 - 1122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Arendt Melatonin: Characteristics, Concerns, and Prospects J Biol Rhythms, August 1, 2005; 20(4): 291 - 303. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
