KVD conceived the study, administered light therapy, performed the statistical analysis and wrote the paper. EAS provided gynaecological knowledge to the clinical interviews and final consultations, and did all investigational procedures including drawing venous blood and ultrasonography. Both authors designed the study, recruited and supervised the test participants, interpreted findings, and surveyed the literature on the topic.
The authors have declared that no competing interests exist.
Studies have shown a shortening of the menstrual cycle following light exposure in women with abnormally long menstrual cycles or with winter depression, suggesting that artificial light can influence reproductive hormones and ovulation. The study was designed to investigate this possibility.
Placebo-controlled, crossover, counterbalanced order.
Medical centres and participants' homes in Novosibirsk (55°N), Russia.
Twenty-two women, aged 19–37 years, with baseline menstrual cycle length 28.1–37.8 d and no clinically evident endocrine abnormalities completed the study. The study lasted for two menstrual cycles separated by at least one off-protocol cycle.
During one experimental cycle, bright light was administered at home for 1 wk with a light box emitting white light at 4,300 lux at 41 cm for 45 min shortly after awakening. During the other experimental cycle, dim light was <100 lux at 41 cm with a one-tube fluorescent source.
Blood samples and ultrasound scans were obtained in the afternoon before and after the week of light exposure, on day ∼7 and 14 after menstruation onset. Further ultrasound scans after day 14 documented ovulation. Serum was assayed for thyroid-stimulating hormone (TSH), prolactin (PRL), luteinizing hormone (LH), follicle-stimulating hormone (FSH), and estradiol (E2).
Concentrations of PRL, LH, and FSH were significantly increased with bright versus dim light exposure, as was follicle size (ANOVA, intervention × day,
Morning exposure to bright light in the follicular phase of the menstrual cycle stimulates the secretion of hypophyseal reproductive hormones, promotes ovary follicle growth, and increases ovulation rates in women with slightly lengthened menstrual cycles. This might be a promising method to overcome infertility.
ClinicalTrials.gov
Artificial (ocular) light administered at the appropriate time of day is used to overcome a variety of conditions associated with circadian rhythm abnormalities (shift work, jet lag, sleep disturbances), monthly cycles (premenstrual dysphoric disorder), seasonal variations (winter depression, bulimia nervosa), amongst others [
The above studies were not accompanied by the hormonal measurements that might elicit the mechanisms underlying the cycle-shortening effect of light. Surprisingly, there has been little investigation in humans into the effects of light on the two principal hormones regulating the menstrual cycle—follicle-stimulating hormone (FSH) and luteinizing hormone (LH)—compared with other hypophyseal hormones and melatonin. These studies showed an increase of LH and/or FSH, in men or women, following bright light 500 to 3,000 lux [
The aim of this study was to answer conclusively whether light can influence menstrual cycle and ovulation, and to determine the underlying hormonal changes taking place. The design combined several approaches used in previous studies and advanced some of them. For example, ultrasound scans were introduced as a method to objectively document ovulation. Bright light, during wakefulness was used since it has also been shown to shorten the menstrual cycle [
The study was performed in Novosibirsk (55°N), in three consecutive years 2003–2006, between October and April (the shortest day is 7 h). Test participants were recruited via advertisements in a local newspaper and a patient database of the gynaecologist (EAS). The advertisement was as follows: “Women with lengthened menstrual cycles (30–38 days) are invited to participate in a study. Ultrasound and hormonal investigations will be conducted free of charge”. Respondents provided dates of menstruation onset over the past year and suitable candidates were invited for clinical interview. During the interview with the investigators (both are clinicians), they received a detailed description of the study, completed a questionnaire concerning their general health and answered some medical questions from the clinicians. The inclusion criteria required that the majority of menstrual cycles, especially the last two baseline cycles, were free from any medication known to interfere with hormone release; the absence of clinically evident endocrine abnormalities; good general health; a normal sleep-wake regimen; and no transmeridian travel two months prior to the study.
After meeting the inclusion criteria, the candidates signed the consent form if they were willing to participate. The most common motive for participation was long-standing difficulty in conceiving and/or a physician's recommendation for a hormonal investigation, which is expensive. The consent included the statement: “The agreement provides the participant with free testing and is offered on a mutually beneficial basis: the participants receive information regarding their reproductive system, whilst the clinicians are able to investigate the effects of light”. The study was approved by the Ethics Committee of the Institute of Internal Medicine SB RAMS.
The study lasted for two menstrual cycles separated by an off-protocol episode of at least one menstrual cycle. Bright light was administered during one experimental cycle, dim light during the other (placebo-controlled, crossover, counterbalanced order design). Light treatment lasted for a week at home and was preceded by the day of baseline investigations and followed by the day of follow-up investigation. Both the baseline and follow-up investigations were planned to be performed in follicular phase before ovulation, because the dominant follicle ruptures and the hormonal profile changes very considerably on ovulation and this might confound consistency of our results, especially if not all cycles are ovulatory. On the other hand, light exposure is most effective when administered around the days of presumed ovulation [
As in study [
The figure shows a spectral power distribution of the light emitted from a Sunray2-Max box (Outside In). It uses phosphor type 835 lamps (data from Sylvania Lighting). The short wavelength up to ∼450 nm was significantly attenuated by the diffuser (filter characteristic from Interlux).
The objective of the study was to investigate the influence of morning bright light on ovarian function in women.
Blood was drawn from the antecubital vein, centrifuged within 1 h, and the serum was kept frozen until the hormonal assay. Serum was assayed for concentrations of thyroid-stimulating hormone (TSH), prolactin (PRL), LH, FSH by immunoluminometric assay (ILMA), and for estradiol (E2) by ELISA. Reagent ILMA kits were obtained from Immunotech (
Ultrasound examination was performed using a linear/sector ultrasonic compound scanner SSD-500 (Aloka [
No formal sample size calculation was done. In the previous light exposure studies, 8 to 27 participants were analysed regarding menstrual cycle length [
Entering the study was determined by participants' menstrual cycle onset and, therefore, was sequential and not dependent on investigators. The type of light intervention was allocated to participants alternately: the first received dim light; the next, bright light; the third, dim; etc. Alternating in this way helped to better balance between months (seasons) and also utilized fewer light devices.
The study participants were not told whether they were to received dim or bright light first and would have realised only during the second experimental cycle. Additionally, an interest in the study's diagnostic tests rather than the intervention helped to lower expectations regarding placebo and active light effects. One of the investigators (EAS) was blind to the intervention type.
Analysis of variance (ANOVA) with two repeated measures, intervention (bright versus dim light) and day (day 7 versus day 14), were the primary statistics in the study. Standard deviations of the means are reported in the text and tables while standard errors of the mean are in figures. Categorical or nominal variables (e.g., occurrence of ovulation) were analysed using nonparametric statistics.
Participant flow is shown in
The final study group of 22 were aged 19 to 37 years (
Baseline Characteristics of Women Completing the Study (
Actually, the participants entered the study on day 5–9 after menstruation onset (median = 7 d, 1-d maximal difference between the two experimental cycles, except in two cases of 2 d). Light exposure started at 07:43 ± 0:55 for the dim light session versus 07:59 ± 0:59 for the bright light session (
Deviations of hypophyseal hormones from the normal range were rare and mild in the studied group—not higher than twice the upper limit and not below the lower limit. One woman's levels of TSH were consistently above the norm, i.e., for all four measurements, indicating a subclinical hypothyroidism. Three women had increased values of PRL, one of whom also had an episodic small increase in TSH and LH. Two more women showed elevated LH or FSH during baseline measurements. The ovarian hormone E2 was typically around the lower limit of the normal range. Almost all women had multifollicle ovaries by ultrasound scan, indicating frequent anovulatory cycles [
The dynamics of the hormones' concentrations, as well as ovarian follicle growth following the repeated light stimuli, are demonstrated in
Twenty-two women with slightly lengthened menstrual cycles were exposed to bright or dim ocular light (45 min after waking, daily, for 1 wk between days ∼7 and 14 after menstruation onset, i.e., before ovulation) during two different menstrual cycles. Prolactin, luteinizing hormone (LH), follicle-stimulating hormone (FSH), and ovarian follicle size were significantly (*) increased in with bright versus dim light. The changes in thyroid-stimulating hormone (TSH) and estradiol levels were not significantly different between the two conditions (ANOVA, intervention × day interaction). Error bars indicate standard errors of the mean.
Main Outcomes and Statistical Testing for Comparisons between the Effects of Bright versus Dim Light
Separately analyzing the general ANOVA independent factors “order of treatment” (exactly half of the participants started the study with dim light first), “seasonality” (exactly half of the participants reported feeling worse in winter versus summer in the screening questionnaire), and “start month” (12 participants started the study before November) did not reveal any significant effect of these factors (either main or at interaction with repeated measures “intervention” and “day”) on the above outcomes.
The number of ovulatory cycles was higher with bright versus dim light (12 versus six cycles). Distribution was as follows: five participants ovulated during both dim and bright light cycles; nine had anovulatory cycles in both experimental conditions; one ovulated during the dim but not bright light cycle; and seven vice versa (tied
Menstrual cycle length was not significantly different between dim versus bright light (33.2 ± 4.8 d versus 32.7 ± 5.6 d,
In total, four women became pregnant during the study period (though all were asked to use condoms during intercourse)—three during the bright light experimental cycle and one during the dim light cycle. In the first three women, ovulation occurred at days 17–18 of the cycle, shortly after the light treatment ended. In the fourth case conception occurred late: at day 24, the dominant follicle was still not present, and the ultrasound examination was stopped. The prospective ultrasound scan 32 d later (because of absence of menstruation) and the temperature data suggested ovulation occurred on day ∼34. All pregnancies were desired.
Numerous significant intercorrelations were found between the studied hormones, follicle growth and ovary volume after light stimuli (
There were no adverse effects except two reports of transient mild eye strain known to occur during bright light therapy [
The study showed that PRL, LH, and FSH secretions, and ovarian follicle growth and the likelihood of ovulation are all increased following a week of morning bright light exposure compared with dim light administered to the same women, in the follicular phase of their cycles. The change of TSH and E2 secretions was not significantly different between two conditions.
This is the first study to focus on the effect of light on the ovulatory process with concomitant measurement of several hormones to understand the underlying mechanism. It is clear that the increased rate of ovulation following bright light exposure is a consequence of faster follicle maturation. Follicle maturation, in turn, is determined by the complex interrelated changes in the secretion of pituitary-ovarian hormones. The neuroanatomical basis for the effect of light may include the novel melanopsin-based photoreceptors in the eye conveying information via the retinohypothalamic tract (separate from the optic nerve) to the biological clock in the suprachiasmatic nucleus of the hypothalamus (reviewed in [
The effects of light discussed here should not be generalized to women with normal or very long/irregular menstrual cycles (as these groups are likely to have different endocrinology from that of the study participants), or to the summer season and Southern locations (both providing bright rather than dark conditions).
The literature search was done via our own database of publications on the topic, PubMed (
TSH secretion is generally rigidly coupled to artificial light exposure (e.g., [
The low rate of ovulation (27%) in the studied group is not surprising as menstrual cycles of over 35 days are associated with low rates of conception in a significant number of women [
The limited number of participants and somewhat narrow sample population are the major limitations of the study. Another consideration is that knowledge of the study results after each of the two consecutive interim analyses may have the potential to bias the results of the following year.
This study, using ultrasonographic examination and hormonal assays, shows conclusively, to our knowledge for the first time, that ovulation may be successfully potentiated by morning artificial bright light. This might be a promising method to overcome infertility in some women.
(44 KB DOC)
(63 KB DOC)
(58 KB DOC)
We thank Outside In for funding the study, study participants for their involvement, Drs. E. O. Styopkina and M. V. Ivanova for hormonal assays, Prof. N. M. Pasman who is a cosupervisor on Dr. Samoilova's scientific work, E. Price and S. Hayes of Outside In for comments and English editing of the manuscript, and Dr. Kripke for sharing his knowledge on the topic.
estradiol
follicle-stimulating hormone
luteinizing hormone
prolactin
thyroid-stimulating hormone