| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Centre for Chronobiology, School of Biomedical and Life Sciences, University of Surrey, Guildford, Surrey GU27XH, United Kingdom
Address all correspondence and requests for reprints to: Dr. Guy Warman, Department of Anaesthesiology, Faculty of Medical Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand. E-mail: g.warman{at}auckland.ac.nz.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Subsequent epidemiological studies (reviewed in Ref.2) of both 50- and 60-Hz exposure, primarily investigating the proposed link between MF exposure and cancer, have yielded inconclusive and contradictory results (3, 4, 5, 6, 7). MF flux densities produced by 60-Hz fields are moderately higher than those produced by the same current at 50 Hz. Despite this fact, there does not appear to be any correlation between frequency of exposure and observation of an effect, with contradictory data at both 50 and 60 Hz.
A hypothetical mechanism of MF-induced carcinogenesis that has received favor in recent years is the so-called melatonin hypothesis of Stevens. The cornerstone of this hypothesis relies on the reported oncostatic properties of melatonin (8, 9).
Stevens (10) proposed that MF acts in a similar way to light at night, suppressing the body’s nighttime production of melatonin from the pineal gland and that this reduction in nighttime melatonin could result in increased incidence of breast cancer.
Controlled laboratory studies clearly indicate that light at night acutely suppresses melatonin in humans with conscious light perception (11, 12, 13) and causes shifts in the phase of circadian rhythms on subsequent days (14, 15). The neuroanatomical pathways by which light results in melatonin suppression are clear. Data supporting the effect of MFs on melatonin levels are, however, very limited, and where available, often poorly controlled. Furthermore, the potential neuroanatomical pathways by which MF could exert an action on melatonin remain obscure.
Field-based MF exposure studies (16, 17, 18) are inadequate to assess accurately the acute effects of MF exposure on melatonin levels because precise dosimetry and control of confounding factors is impossible. Correctly determining the effects of MFs on melatonin thus necessitates controlled laboratory-based studies. Results from the laboratory-based studies that have been conducted (19, 20, 21, 22, 23, 24, 25) are as inconclusive and contradictory in nature as the epidemiological data. Although some studies suggest an effect of MF exposure on the production or timing of the nightly melatonin rise (22, 23), others have found no effect (20, 24, 25). One pertinent point raised by Wood et al. (22) is that if MFs were acting on circadian rhythms in a similar way to light, then the timing of MF exposure relative to the phase of the circadian clock may be important to observe an effect. Limited support for this hypothesis can be found in their data (22), which suggests an effect of MF exposure on the timing of the melatonin rise (a phase shift) in a responsive subgroup before or during the nightly melatonin onset but not after it.
Some of these studies have suffered from a paucity of correct sham controls and lack of control for non-MF-related factors such as posture or light that alter melatonin production.
As a direct result of the melatonin hypothesis, the majority of studies that have been conducted to investigate the potential effects of MFs on hormone secretion have used melatonin suppression as a marker. There are, however, a limited number of studies looking at circulating hormone levels of other endocrine systems including the pituitary, thyroid, adrenal cortex, and reproductive organs. TSH, T4, GH, cortisol, FSH, LH, and testosterone have been studied in men (26, 27), and GH, cortisol, and prolactin in men and women (28). As is the case with the melatonin studies, no conclusive effect of MF exposure has been shown.
Here, we describe results from a highly controlled laboratory-based study designed to detect minor changes in the human melatonin profile after exposure to MFs. The rationale for this study was to determine whether high-level exposure [200–300 microTesla (µT)] to MFs of a similar field configuration to those produced by the United Kingdom power distribution system (circularly polarized, 50 Hz) is capable of acutely altering the production or timing of human melatonin secretion and whether this response can be used as an indicator of the interaction of alternating current (AC) MF on human biology.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous work has suggested that women in different stages of the menstrual cycle and those on oral contraceptives can display different melatonin and temperature profiles; thus, to reduce potential confounding effects associated with variable profiles, only male subjects were chosen. Double-blind, sham-controlled data were obtained in a balanced crossover design from 23 healthy male subjects aged between 18 and 35 yr. Ethical approval for the study was obtained from the University of Surrey Advisory Committee on Ethics and was in accordance with the Helsinki Declaration, 1975. Before participation subjects were screened to ensure they were healthy and not taking medication likely to alter melatonin levels (e.g. ß-blockers, monoamine oxidase inhibitors). The subjects’ general practitioners certified they were suitable for participation. Urinary 6-sulfatoxymelatonin levels were measured over 48 h to ensure normative melatonin production and timing.
Participating subjects completed two 7-d study legs (one sham, one MF exposure) separated by 7 d. During the first 5 d of each leg (i.e. Monday through Friday), subjects maintained a regular sleep-wake regimen. Each subject wore an Actiwatch (Cambridge Neurotechnologies, Cambridge, UK) to assess compliance to the sleep schedule and an Emdex II magnetic field meter (Enertech Consultants, Campbell, CA) to assess environmental AC MF exposure. Between 1600 h Friday and 1000 h Sunday of each leg, subjects remained in the clinical investigation unit, University of Surrey. Light levels were maintained at less than 10 lux (horizontal angle of gaze < 3 lux). Each night 5-ml blood samples were drawn at 30- to 60-min intervals between 1700 h and 1000 h the following morning through an indwelling cannula (Wallace 19G Y-cann, Simms-Portex, Hythe, UK) inserted into the cephalic or median cubital vein under local anesthetic. Core body temperature was monitored in 60-sec epochs by constant-wear soft insertion temperature probes connected to Squirrel MQ32–2U/2C data loggers (Grant Instruments Ltd., Cambridge, UK).
Exposure paradigm
Subjects were assigned to one of four 2-h exposure bins starting between 1700 and 2300 h. Baseline melatonin profiles were determined on the first night of each leg by placing subjects in the exposure equipment with no current flow. The second night of each leg served as the exposure or sham night.
Subjects consumed a standard meal (
600 kcal) 90 min before exposure. They remained semirecumbent for 60 min before treatment. Fifteen minutes before treatment they sat up on the edge of their beds and were exposed in this position to avoid confounding effects of major postural changes. Of the total 23 subjects, the initial eight were exposed to a 200-µT MF. Initial results from n = 8 after a 200 µT MF indicated no significant effect; however, to avoid type II error, 15 more subjects were exposed to a 300-µT MF. The rationale for this flux density was to determine whether 200 µT was below the effective threshold for an effect because it has been suggested that 300 µT is the level required to induce currents in the pineal body (29). Exposure to the MFs (exposure condition), less than 0.2 µT (sham condition), or no field (baseline condition) occurred for 120 min, during which subjects were asked to report whether they considered they had been exposed to a MFs. Fifteen minutes after treatment subjects returned to a semirecumbent position and were allowed to sleep in a supine position 90 min after treatment.
Exposure system
Three identical coil systems were used to expose subjects to MFs and sham fields. Each coil (Fig. 1
) was designed to ensure maximal field uniformity (total field uniformity in the central 20%). Windings, supported by acrylic formers, fastened together with nylon bolts (Stanley Plastics, Midhurst, UK) produced an exposure system of dimensions 900 mm x 900 mm x 900 mm, which was supported by an open-bay trolley to allow in-bed subject exposure. Two sets of orthogonally oriented coils (one north-south, one east-west), each consisting of two sets of 50 windings in series, enabled production of circularly polarized magnetic fields of up to 300 µT. To enable the production of correct sham controls and eliminate confounding effects of vibration, electric fields, noise, and heat, windings were bifilar. MFs and sham fields were thus created by unidirectional and bidirectional flow of current, respectively. Windings were embedded in silicon sealant (RTVC encapsulant EN1032C, Dow Corning, Midland, MI) to minimize vibration.
|
4.5 A at 300 µT). Field characteristics were measured before experimentation with a search coil attached to an oscilloscope. During experiments MF exposure levels were measured at head height in 60-sec epochs using two Emdex II magnetic field meters (one high field, one low field) (Enertech Consultants). Unidirectional and bidirectional flow of current for sham and exposure conditions was controlled by a switching box locked into a secure container until the results of the experiment had been analyzed and unblinded. The 120-min exposure occurred with subjects’ heads centered in the apparatus. No shielding from ambient magnetic fields was present; however, background AC fields were constantly monitored and were all less than 0.2 µT. Measurement of melatonin
Immediately after collection, blood samples were centrifuged (3000 x g, 10 min) and plasma aspirated and frozen at -20 C. Melatonin RIA was conducted by Stockgrand Ltd. (University of Surrey, UK). Average inter- and intraassay variation was less than 12%.
Data analysis
Complete data were obtained and analyzed from 19 subjects. Incomplete data from four subjects were discarded.
In those subjects who were treated on the rising phase of their melatonin profile (n = 10), melatonin suppression during treatment (2 h) and for a period of 2 h after treatment was calculated for each subject for each time point. This was achieved by aligning subjects with respect to treatment start and calculating the ratio of melatonin during the MF vs. sham treatment. The mean of these data was then plotted against time after treatment onset.
Phase-shifting effects of MF exposure were assessed by plotting plasma melatonin levels for each night (two baseline nights, one exposure night, one sham night) as a function of clock time (Greenwich mean time, decimal hours) of sample. The phase reference point employed here to define melatonin onset was midrange crossing [defined as the midpoint (50%) between baseline and melatonin peak (acrophase)] (Mid X) (30).
Statistical analysis was conducted using the SPSS for Windows software package (SPSS Inc., Chicago, IL). Overall variability of melatonin onset between baseline and treatment nights of each leg was determined by F test. Statistical significance of changes in melatonin onset between baseline and treatment nights was assessed by paired t test. Determination of the statistical power of the protocol was conducted on the cumulative number of subjects (i.e. all of the subjects irrespective of whether they received 200 or 300 µT) and on those subjects exposed only to 300 µT using on-line software (http://hedwig.mgh.harvard.edu/size.html).
Because the phase shifting effects of putative time cues (zeitgebers) depend on the circadian time (CT) (time relative to the phase of the circadian clock) of administration, data were also normalized to the same circadian time of MF administration. This was achieved by denoting melatonin onset (Mid X) on the baseline night of each leg as CT 14 and the time melatonin onset (Mid X) occurred on the treatment night (i.e. sham or exposure night) as a change from this value. These data were then plotted as a phase response curve with circadian time of treatment [determined as the midpoint of the MF pulse relative to the baseline melatonin onset (Mid X)] on the abscissa and the change in phase between baseline and treatment nights on the ordinate.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. From these data it is clear that relative to the baseline, there is no mean change in timing or overall amplitude of the melatonin rhythm after exposure to MFs, compared with the sham condition. The lack of a MF-induced alteration in the mean melatonin profile is corroborated by uniform core body temperature profiles on all legs of the study (data not shown). Furthermore, in the 10 subjects who received MFs during the rising leg of their melatonin profile, there is no evidence of a suppressive effect of MF exposure relative to sham, either during or after exposure (data not shown).
|
The lack of a consistent change in the mean timing of the melatonin rise (phase shift) was confirmed by the analysis of the mean melatonin onset as determined by midrange crossing. Timing of the mean melatonin rise and mean peak melatonin levels on the treatment night of each leg is statistically indistinguishable from mean timing on the baseline night of each leg [sham leg: mean baseline melatonin onset time 2330 ± 0.24 h, acrophase 81 pg/ml (349 pmol/liter) ± 11 pg/ml (47 pmol/liter); mean sham treatment melatonin onset time 2347 ± 0.26 h, acrophase 78 pg/ml (336 pmol/liter) ± 10 pg/ml (43 pmol/liter); MF exposure leg: mean baseline melatonin onset time 2262 ± 0.21 h, acrophase 81 pg/ml (349 pmol/liter) ± 11 pg/ml (47 pmol/liter); mean MF treatment melatonin onset time 2383 ± 0.20 h, acrophase 77 pg/ml (332 pmol/liter) ± 11 pg/ml (47 pmol/liter); paired t tests, P > 0.05].
When changes in the melatonin onsets from individual subjects were analyzed, a trend toward increased variability after MF exposure (compared with sham exposure) is evident in the data (Fig. 3); however, this trend is not significant (F test, sham F = 0.69, MF F = 0.97). The trend is evident only in the individual data because variation is bidirectional (advances and delays in melatonin onset).
|
), there may appear to be an interaction between circadian time of exposure and direction of phase shift, with exposures before CT 11 resulting in a phase advance and those after CT 11 eliciting a delay. However, when compared with the sham data, there is no significant difference between the treatments.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
If MFs were to elicit responses comparable with light, suppression of melatonin should provide a more sensitive index than a phase shift (12). Our findings provide no support for the direct suppression of melatonin levels by MF exposure as proposed by Stevens (10).
In addition to direct suppression, a second way in which MFs could result in a change in the overall production of melatonin is by altering the timing of the melatonin rise without eliciting a concomitant change in the melatonin decline, as suggested by Wood et al. (22). Furthermore, overall phase shifts in the circadian clock as illustrated by a phase shift of the melatonin onset are associated with changes in timing of core body temperature and cortisol secretion in humans (31, 32), with sleep disorders such as advanced sleep phase syndrome, delayed sleep phase syndrome, and other health problems associated with shift work (reviewed in Ref.33). If MFs were acting to phase shift the melatonin rhythm in a similar way to light (i.e. as a zeitgeber) the direction and magnitude of any shift would depend on the circadian timing of administration. In this case we would expect to observe both a significant increase in the variability of individual melatonin onset after exposure to the MF, compared with the sham, and an influence of circadian time of exposure on the magnitude and direction of the shift.
Statistical ANOVA in melatonin onset data indicate that the current protocol possessed 87% power to detect a change in the timing of melatonin onset as small as 20 min (25 min if the power calculation is restricted to those subjects receiving 300 µT). Despite this, no significant alteration in the timing of individual melatonin onset times was detected (Fig. 3
). Furthermore, there is no significant effect of circadian time of MF administration on the observation of a response relative to the sham data (Fig. 4).
When compared with light, this lack of effect becomes yet more evident. Even low levels of light, commonly encountered in laboratory and home settings, administered during the rising phase of the melatonin profile (200–0200) suppress melatonin by more than 50% (13, 14) and result in phase shifts of up to 3 h (34).
The potential consequences of not including appropriate baseline and sham controls in previous protocols is illustrated by Fig. 4. Without the sham data, a circadian biologist might assume that the MF results suffice to claim that MF acts as a weak zeitgeber; acting to phase advance before CT 11, and delay after CT 11. However, when considered in conjunction with the sham data, this effect can be seen to be due either to random variability or the regression to the mean effect (35).
The claims of some authors that there is an effect of MF exposure on the human melatonin profile may be attributable to the presence of confounding factors. There are two main areas of inadequacy in the current literature. The first is the use of inappropriate MF-generating equipment that cannot deliver well-characterized MFs or appropriate sham controls. Because heating, vibration, and electric and MFs are all associated with current flow, incomplete correction for any of these factors could influence results. It is by no means sufficient to turn off MF-generating equipment in the sham condition, as is frequently the case (19, 20, 21, 23, 24, 25).
The second major confounding factor inherent in most of these studies is the lack of control for non-MF-related factors that are known to alter melatonin levels. Light is the most obvious of these, but other more subtle nonphotic factors including posture, medication, and circadian phase of subjects entering the study can also profoundly influence apparent plasma melatonin levels. Although the work of Wood et al. (22) provides some of the most controlled data concerning the potential effect of 50-Hz MFs on the human melatonin profile published to date, the disparity between the data presented here, suggesting the absence of an effect, and those presented by Wood et al., suggesting an effect in a 25% responsive subgroup, may come down to methodological issues. In their study Wood et al. did not control for postural and lighting effects, and although they conducted baseline nights, they were not carried out immediately before each treatment as would be appropriate. Furthermore, different subjects appear to have received exposures of vastly different lengths (1.5–4 h), and there are no data on whether exposure duration was correlated with observation of an effect.
With tens of thousands of kilometers of 50-Hz power distribution networks in the United Kingdom alone and electrical equipment omnipotent in our everyday lives, people living in industrialized societies are regularly exposed to the time varying (AC) MFs produced by the flow of current through these wires. We have shown clearly that acute exposure to circularly polarized 50-Hz MFs of 200–300 µT does not result in suppression of melatonin, alteration in peak melatonin levels, or changes in the timing of the nighttime melatonin rise. When put in perspective, these exposure levels are some 4–6 times higher than are likely to be encountered by all but the most highly exposed electricity utility workers. From these data we can conclude that acute exposure to 50-Hz MFs does not significantly alter the melatonin profile in young male subjects. The lack of a significant effect of MFs on the melatonin profile does not, however, negate the possibility that MF is affecting the circadian clock at a more subtle level. Clarification of this issue will rely on the analysis of more sensitive indicators of minor changes in the rhythm generating system, such as changes in gene expression in the suprachiasmatic nucleus. Clarification of possible gender and age differences and/or effects of MFs on other neuroendocrine systems will rely on highly controlled laboratory based studies such as the one described here.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Current address for G.W.: Department of Anaesthesiology, University of Auckland, Private Bag 92019, Auckland, New Zealand.
Abbreviations: AC, Alternating current; CT, circadian time; MF, magnetic field; Mid X, midpoint between baseline and melatonin peak.
Received February 10, 2003.
Accepted September 2, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  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 |
