Endogenous Circadian Regulation of Female Reproductive Hormones

Abstract

Context

Studies suggest that female reproductive hormones are under circadian regulation, although methodological differences have led to inconsistent findings.

Objective

To determine whether circulating levels of reproductive hormones exhibit circadian rhythms.

Design

Blood samples were collected across ∼90 consecutive hours, including 2 baseline days under a standard sleep-wake schedule and ∼50 hours of extended wake under constant routine (CR) conditions.

Setting

Intensive Physiological Monitoring Unit, Brigham and Women’s Hospital.

Participants

Seventeen healthy premenopausal women (22.8 ± 2.6 years; nine follicular; eight luteal).

Interventions

Fifty-hour CR.

Main Outcome Measures

Plasma estradiol (E2), progesterone (P4), LH, FSH, SHBG, melatonin, and core body temperature.

Results

All hormones exhibited significant 24-hour rhythms under both standard sleep-wake and CR conditions during the follicular phase (P < 0.05). In contrast, only FSH and SHBG were significantly rhythmic during the luteal phase. Rhythm acrophases and amplitudes were similar between standard sleep-wake and CR conditions. The acrophase occurred in the morning for P4; in the afternoon for FSH, LH, and SHBG; and during the night for E2.

Conclusions

Our results confirm previous reports of ∼24-hour rhythms in many female reproductive hormones in humans under ambulatory conditions but demonstrate that these hormones are under endogenous circadian regulation, defined as persisting in the absence of external time cues. These results may have important implications for the effects of circadian disruption on reproductive function.

Circadian regulation of female reproductive physiology is well characterized in rodent models [reviewed in Boden and Kennaway (1)] but is mainly supported by indirect evidence in humans. A recent meta-analysis of 16 female cohorts (123,403 women) showed that women who are engaged in shift work, which can disrupt circadian rhythms, have increased odds of menstrual cycle disruption, early spontaneous pregnancy loss, and infertility by ∼22%, ∼30%, and ∼80%, respectively, compared with women not engaged in shift work (2). Hormones related to these functional outcomes have been reported to exhibit 24-hour rhythms under normal sleep-wake, meal, and lighting conditions. Diurnal rhythms in estradiol (E2), progesterone (P4), LH, and FSH have been reported in premenopausal women (3–10). Findings of diurnal variability in these hormones have not been consistent, however (5, 11–13). Differences in the study population (e.g., the menstrual phase at which women were studied) and analytic techniques [e.g., the application of nonlinear cosinor regression in some (4–6, 9–12) but not others (3, 7, 8, 13)] likely contributed to the inconsistent findings between studies in the identification of rhythmic hormones.

Moreover, these studies cannot confirm whether the daily rhythm represents the output of an endogenous circadian pacemaker or is a result of evoked responses, such as sleep-wake and feeding-fasting cycles, or a combination of both (14). The presence of an endogenous circadian rhythm can only be confirmed when external time cues are absent or distributed uniformly, such as during a constant routine (CR) protocol, where activity, posture, light, food intake, and wakefulness remain constant (14). When healthy, premenopausal women were studied under such CR conditions, LH, FSH, and E2 levels were reported to be arrhythmic (15) and therefore, not under endogenous circadian regulation. This study included only women in the early follicular phase of their menstrual cycle (approximately cycle-day 4 in relation to menses), however, which precludes examining whether circadian regulation of these hormones differs between the follicular and luteal phase. Moreover, these studies did not directly compare rhythms under constant and ambulatory conditions, which is important to elucidate the masking effects of behavioral events, such as sleep and feeding-fasting on the underlying circadian regulation of physiologic rhythms (16). Therefore, in the current study, we assessed the endogenous circadian regulation of E2, P4, FSH, LH, and SHBG in the follicular and luteal phases of the menstrual cycle and compared the rhythms in these hormones between CR and ambulatory conditions.

Methods

Participants

We studied 17 healthy premenopausal women (mean age ± SD = 22.8 ± 2.6 years; range 19 to 29 years; nine women in the follicular phase). All participants underwent comprehensive physical, psychiatric, and medical screening. Participants self-reported a regular menstrual cycle lasting 27 to 35 days (mean cycle length ± SD = 28.9 ± 1.84 days) and were not using oral contraception at the time of the study or within the previous 3 months. Menstrual phase was determined based on each individual participant’s daily average P4 levels. Women with daily mean P4 concentrations >3 ng/mL were considered to be postovulatory (17, 18) and assigned to the luteal phase, and women with P4 <3 ng/mL were assigned to the follicular phase. Given the protocol duration, one participant (2251V) changed from the follicular phase during baseline to the luteal phase during CR based on levels of P4, with concomitant changes in FSH and LH, indicating ovulation. Demographic information, P4 concentrations, and allocation of menstrual phase for each of the participants on the different study days are presented in Table 1.

For ∼3 weeks before the laboratory study, all participants maintained a self-selected, constant 8-hour sleep/rest/dark schedule, confirmed with calls to a time- and date-stamped voicemail at bedtime and wake time and with actigraphy (Actiwatch-L; Philips Respironics, Bend, OR) and sleep diaries for at least 7 days before entering the laboratory. Participants were asked to refrain from using any prescription and nonprescription medications, supplements, recreational drugs, caffeine, alcohol, or nicotine for the duration of the study. Compliance was confirmed with a urine toxicology test during the 3-week screening phase and upon admission to the laboratory. The study was approved by the Partners Human Research Committee (2007P000566), and written, informed consent was given by participants before starting the study.

Study protocol

Participants were studied individually for 9 days in an environment free of time cues (no access to windows, clocks, live television, radio, or internet and continually supervised by staff trained not to reveal the time) per a standard protocol and as described in Lockley et al. (19) and Gooley et al. (20). Only data from the first 6 days were analyzed and reported herein (Fig. 1); data from a separate analysis of the relationship among reproductive hormones, temperature, and performance are reported elsewhere (21). The first 6 days of the protocol consisted of 3 baseline days with an 8-hour sleep opportunity timed according to the average for 7 days before admission. The baseline study days were followed immediately by a 49-hour, 40-minute CR, followed by an 8-hour recovery sleep (Fig. 1). During the CRs, participants remained awake in a semirecumbent posture in dim light (<3 lux) and were fed hourly isocaloric snacks [150 mEq Na⁺/100 mEq K⁺ (±20%); 1.3 × basal energy expenditure; 2000 mL fluids/24 h/d].

Lighting

Study lighting conditions have been described in detail previously (19, 20). During baseline days, maximum ambient light (ceiling mounted 4100K fluorescent lamps F96T12/41U/HO/EW, 95 W; F32T8/ADV841/A, 32 W; F25T8/TL841, 25 W; Philips Lighting, The Netherlands), during scheduled wake episodes, was ∼48 µW/cm² (∼190 lux) when measured vertically and ∼23 µW/cm² (∼88 lux) when measured horizontally at a height of 187 cm and 137 cm, respectively. Midway through day 3, maximum ambient light was reduced to <0.4 µW/cm² (<3 lux) when measured vertically and ∼0.6 lux when measured horizontally. This level of light was maintained for the remainder of the study except during scheduled sleep episodes that occurred in darkness.

Hormone measurement

Plasma was collected from an indwelling IV cannula inserted into a forearm vein and kept patent with a heparinized saline infusion (5 IU heparin/mL 0.45% NaCl infused at 40 to 42 mL/h). The IV was inserted on the second baseline day. Blood samples were transferred to EDTA tubes and kept on ice before centrifugation. The plasma fraction was transferred into plastic tubes and stored at −20°C. During the baseline days and the first CR, blood samples were collected every 30 to 60 minutes, and two hourly samples were assayed. Plasma melatonin was assayed using radioimmunoassay (Alpco Diagnostics, Salem NH). Plasma intra- and interassay coefficients of variation were <9% and <11%, respectively, at 1.94 and 16.59 pg/mL. Plasma total E2, FSH, LH, total P4, and SHBG were assayed using the Access Chemiluminescent Immunoassay (Beckman Coulter, Fullerton CA). Intra- and interassay coefficients of variation were 12% to 20% for E2, 3.1% to 5.6% for FSH, 4.3% to 6.4% for LH, 6.11% to 11.19% for P4, and 4.5% to 5.5% for SHBG.

Data analysis

Data are expressed as means ± SEM unless otherwise specified. The first 5 hours of hormone data were excluded from the analysis of CR data to remove any masking effects from the prior sleep episode and changes in posture (22). One participant (26H6V) had no blood collected during baseline days 2 and 3. In one participant (26F2V), LH was below the assay limit of detection, and for another participant (22A1V), FSH was excluded as a result of abnormally high levels (means ± SD, 44.6 ± 7.5 mIU/mL, 4 SDs above the mean), although other reproductive hormone data were retained, as they were within normal ranges.

Cosinor analysis was performed to assess the endogenous circadian regulation of sex hormones under standard sleep/wake baseline and CR conditions. For this analysis, all individual values for each participant were assigned a time relative to their scheduled wake time. For the group analysis, values from each individual were binned in 2-hour intervals, and values from matching bins were averaged across individuals. Hormone data were z-transformed for each individual using the means and SD across the entire data collection interval. Transformed data were fit for the baseline and CR intervals separately using cosinor regression analysis with a single harmonic model that included one 24-hour fundamental component and a linear term: $y=A cos {(2πτ)(x−ϕ)}+(mx+c)$⁠, where A = amplitude, τ = period (24 hours), ϕ = acrophase, m = slope of linear term, and c = vertical intercept of linear term (16, 23). The linear component was included in the model to estimate the influence of secular changes associated with the CR protocol (e.g., effects of accumulating time awake, repetitive meals, time in a semirecumbent posture, etc.). Melatonin data were z-transformed (24) before fitting with a single harmonic cosinor regression model with one 24-hour fundamental component (16): $y=A cos {(2πτ)+(x−ϕ)}$⁠. For both regression analyses, the regressions were considered significant if the amplitude of the fitted rhythm was significantly different from 0, with a one- and two-sided type I error threshold of 0.05 used for individual and group analyses, respectively. Regression analyses were conducted in SAS 9.4. (SAS Inc., Cary, NC). The single harmonic regression analysis of melatonin is included in figures for illustrative purposes. Individual phase estimates (Table 1 and Fig. 2) for melatonin are reported as the time at which melatonin secretion began under dim light (dim light melatonin onset), defined as the time at which the melatonin rhythm crossed the 25% threshold of the three-harmonic, peak-to-trough fitted amplitude (25).

Differences in rhythm acrophases and amplitudes between menstrual phases and between baseline and CR conditions were tested using the extra sum-of-squares F-tests (Prism 7; GraphPad Software, La Jolla CA). Differences in the proportion of participants displaying significant 24-hour rhythms between menstrual phases and baseline and CR conditions were tested using χ² tests.

Results

Circadian rhythms in female reproductive hormones during the follicular and luteal phases of the menstrual cycle

In group-fitted data, including only women in the follicular phase, there were significant 24-hour rhythms observed in all of the reproductive hormones examined under both baseline and CR conditions (Fig. 2), including E2 (baseline: P < 0.05; CR: P < 0.05), FSH (baseline: P < 0.001; CR: P < 0.01), LH (baseline: P < 0.001; CR: P < 0.05), P4 (baseline: P < 0.001; CR: P < 0.001), and SHBG (baseline: P < 0.01; CR: P < 0.001). In contrast to women in the follicular phase, significant 24-hour rhythms were only observed in group-fitted FSH (baseline: P < 0.05; CR: P < 0.01) and SHBG (baseline: P < 0.01; CR: P < 0.001) in women in the luteal phase under both baseline and CR conditions (Fig. 2). The acrophase and amplitude of the FSH and SHBG rhythms (the only hormones significant in both the follicular and luteal phases) were not different between menstrual phases under either condition (Fig. 2). The unadjusted hormone data are shown in the data repository (26).

At the individual level, the proportions of women in the follicular phase (n = 9) who exhibited a significant 24-hour rhythm under baseline conditions were 44% for E2, 100% for FSH, 75% for LH, 67% for P4, and 44% for SHBG (Table 2). Compared with women in the follicular phase, a similar proportion of women in the luteal phase exhibited significant 24-hour rhythms in E2 (57%), P4 (42%), and SHBG (57%; P > 0.34 for all). In contrast to these hormones, FSH exhibited significant 24-hour rhythms in a substantially smaller proportion of women (33%; χ² = 8.2, P < 0.01) in the luteal phase, and LH trended toward exhibiting significant 24-hour rhythms in a small proportion of women (29%; χ² = 3.2, P = 0.07) compared with the follicular phase under baseline conditions (Table 2).

Under CR conditions, the proportion of women in the follicular phase who exhibited significant 24-hour rhythms was 89% for E2, 56% for FSH, 63% for LH, 89% for P4, and 78% for SHBG (Table 3). Compared with women in the follicular phase, a similar proportion of women in the luteal phase exhibited significant 24-hour rhythms in E2 (75%), FSH (43%), LH (25%), and SHBG (50%; P > 0.13 for all), but a smaller proportion exhibited significant 24-hour rhythms in P4 (38%; χ² = 4.9, P < 0.05; Table 3). The phase distribution and time course of significant individual rhythms are shown in the data repository (26).

Comparison of circadian rhythms in female reproductive hormones under standard sleep-wake and CR conditions

For women in the follicular phase, the amplitude of the group-fitted FSH rhythm was significantly smaller (P < 0.001; Fig 2C) under CR compared with standard sleep-wake conditions (baseline), and a similar trend was observed for LH (P = 0.051; Fig. 2E). The acrophase of the SHBG rhythm was significantly earlier (∼4 hours) under CR conditions compared with baseline (P < 0.05; Fig. 2I). There were no differences in the phase or amplitude of the other hormones between baseline and CR in women in the follicular phase. At the individual level, compared with baseline, a larger proportion of women in the follicular phase exhibited 24-hour rhythms in E2 (baseline = 44%, CR = 89%; χ² = 4.0, P < 0.05) under CR conditions, and a smaller proportion exhibited 24-hour rhythms in FSH (baseline = 100%, CR = 56%; χ² = 5.1, P < 0.05). The proportion of participants in the follicular phase exhibiting significant rhythms at the individual level was similar for baseline and CR for LH, P4, and SHBG (P > 0.15 for all).

For women in the luteal phase, the acrophase of the SHBG rhythm was significantly earlier (∼5 hours) under CR compared with baseline (P < 0.01; Fig. 2J). At the individual level, the number of women in the luteal phase who exhibited 24-hour rhythms during CR was similar to under baseline conditions for all hormones (P > 0.46 for all).

Discussion

Our findings demonstrate that key endocrine regulators of female reproductive function, including sex-steroid hormones, gonadotropins, and SHBG, are under endogenous circadian regulation. Moreover, this regulation appears to depend on menstrual phase, such that most variables were robustly rhythmic during the follicular phase but not the luteal phase of the menstrual cycle. These results have important implications for understanding the regulation of reproductive function.

Our results show that the circadian clock differentially regulates female reproductive hormones in humans, depending on the menstrual phase. Although the data from our study preclude directly testing neuroendocrine mechanisms that can explain this differential circadian regulation of female reproductive hormones across the menstrual cycle, there are a number of possibilities. The loss of rhythmicity in P4 may be a result of a shift in the primary source of P4 from the adrenal cortex during the follicular phase to the corpus luteum during the luteal phase, which may not be rhythmic. Although the expression of steroidogenic acute regulatory protein, which catalyzes the rate-limiting step of steroidogenesis, is regulated by the circadian core clock gene brain and muscle aryl hydrocarbon receptor nuclear translocator-like l (BMAL1), P4 secretion by the corpus luteum is arrhythmic in both bmal1^−/− and heterozygote controls (27). Additionally, the arrhythmic and elevated P4 levels during the luteal phase may explain the loss in LH rhythmicity during the luteal phase, given that P4 suppresses the pulsatile secretion of LH and mean circulating levels of LH (28). The significant circadian regulation of sex steroids, gonadotropins, and SHBG during the follicular phase, preceding ovulation but not after ovulation, suggests that complex reproductive outcomes, such as ovulation, may be temporally coordinated by the circadian system.

Although our study did not directly assess the effect of circadian disruption of these rhythms, several epidemiologic studies have reported the adverse effects of shift work, which disrupts circadian rhythms, on female reproductive health. Data from the 1993 cohort of the Women’s Health Study that followed 71,077 women showed a 23% increased risk of irregular and abnormally long (>40 days) or short (<21 days) menstrual cycles with increasing months of rotating shift work (29). Other studies have reported that female shift workers self-report changes to cycle length, menstrual flow, menstrual pain, and duration of menses (30) and also report more problems conceiving [i.e., increased time to pregnancy (31)] and maintaining pregnancy [i.e., increased odds of miscarriage (32)]. Furthermore, two prospective studies have shown a higher rate of monophasic (i.e., no basal body temperature increase; anovulatory cycles) and irregular ovulatory cycles in rotating shift workers compared with day workers (33, 34). We hypothesize, therefore, that these circadian rhythms in reproductive hormones are disrupted during shift work and that their disruption contributes to the impaired reproductive function observed in female shift workers. Future studies should examine this hypothesis explicitly. Diurnal changes in gonadotropins and sex steroids have been reported previously [for example, see Veldhuis et al. (3), Rossmanith and Lauritzen (4), Mortola et al. (5), and Bao et al. (6)], although the rhythmic oscillations have been examined under constant environmental and behavioral conditions only once previously (15). We examined both ambulatory and constant conditions in the same study to compare directly evoked responses and the endogenous circadian regulation of these hormones. Our results corroborate previous reports of significant 24-hour oscillations under ambulatory conditions. In contrast, our results are not consistent with the previous study in which a CR was used, wherein no significant circadian oscillations in gonadotropins and E2 were detected in healthy young women in the early follicular phase of the menstrual cycle (15). Subtle differences in study protocols and possible differences in menstrual-phase estimation may have contributed to these different conclusions. For example, the ambient light levels were dimmer in our protocol (<3 lux) than in the previous protocol (50 to 150 lux). Whereas the effect of ambient light exposure on reproductive hormones in women is not well characterized, melatonin has been shown to modulate reproductive hormone levels (35), and melatonin is suppressed by light exposure. Therefore, it is possible that the higher intensity ambient light used in the prior study may have altered melatonin levels, which indirectly then modulated the levels of the reproductive hormones, masking the underlying circadian regulation. Future studies are necessary to characterize systematically the influence of environmental factors (e.g., light) on these reproductive rhythms.

Additionally, in our study, menstrual phase was determined using a threshold for circulating P4, which does not allow further categorization into the early or late follicular phase. A trend toward an increase in amplitude of the diurnal rhythm in free E2 between the early to late follicular phase has been previously reported, although the difference did not reach statistical significance in that study with a limited sample size (6). The study did, however, observe a significant difference in the acrophase of the rhythm between the early and late follicular phase. Additionally, the study by Bao et al. (6) did not characterize the rhythms under constant conditions, which may have masked the subtle differences in rhythm characteristics between the early and late follicular phase. It is possible that there was a wider distribution in the follicular phase in our study compared with the study by Klingman et al. (15), with more individuals in our study in the late follicular phase than in the early follicular phase. Therefore, there may be differences in circadian rhythms in reproductive hormones, not only between the follicular and luteal phases but also perhaps within a menstrual phase (e.g., early vs late follicular phase) that warrants further examination. Furthermore, we cannot rule out a possible order effect or the time-into-protocol effect that may have contributed to the differential hormone levels between CR and baseline conditions.

Another pertinent finding in the study was the remarkable variability in the proportion of women who were rhythmic for a given endocrine variable. The proportion ranged between 100% for FSH in the follicular phase under baseline conditions to 29% for LH in the luteal phase under baseline conditions. We and others have previously reported similarly large variability in the number of significant rhythms observed at the individual level for proinflammatory cytokines and chemokines (16), lipids (24), and metabolites (36). It is possible that the cosinor-based approach of the determination of rhythms is not adequately robust for individual-level data, which may not fit a sinusoid profile as well as group-averaged data. Additionally, this variability may also represent varying levels of susceptibility of each individual to rhythm perturbations caused by external or internal factors, which may, in turn, explain the differences in outcomes among individuals in response to external influences, such as shift work.

This study has several limitations. Whereas we collected blood samples every 2 hours longitudinally over ∼96 hours, most of these hormones are known to be pulsatile. Whereas a less frequent sampling over a long duration of time may be sufficient when trying to detect gross 24-hour rhythmicity, future studies with more a frequent sampling are needed to examine whether amplitude and frequency of gonadotropin pulsatility also exhibit circadian rhythms. Additionally, future studies with higher sensitivity assays (e.g., HPLC) are warranted to confirm the precise amplitude of the rhythms. The limited sample size and the age range in the current study also limit the generalizability of the results. Additionally, results from individual-based analyses should be interpreted with caution because of the limited sample size and limited statistical power. Additional studies with a greater sample size and statistical power are warranted, however, to test the interesting individual-level variability observed in the current study.

Overall, our study demonstrates that many female reproductive hormones are under endogenous circadian regulation. Our findings provide a potential mechanism by which shift work adversely affects reproductive function through disruption of the circadian regulation of reproductive hormones.

Acknowledgments

We thank the technical, dietary, and laboratory staff, nurses, and physicians; participant recruiters; and the study participants at the Center for Clinical Investigation and Division of Sleep and Circadian Disorders, Brigham and Women’s Hospital.

Financial Support: This work was supported by the National Institute of Mental Health (2R01 MH45130-11A1; to C.A.C. and S.W.L.), National Center for Complementary and Alternative Medicine (R01 AT002129; to C.A.C. and S.W.L.), and National Institute of Environmental Health Sciences (R21 ES017112-01A1; to S.W.L.). C.A.C. and S.W.L. were supported, in part, by the National Space Biomedical Research Institute through NASA NCC 9-58. The project was supported by Brigham and Women’s Hospital General Clinical Research Center Grant M01-RR02635.

Additional Information

Disclosure Summary: L.K.G. and J.J.G. have nothing to disclose. S.A.R. owns equity in Melcort Inc.; has provided paid consulting services to Sultan & Knight Limited and Bambu Vault LLC; and has received honoraria as an invited speaker and travel funds from Starry Skies Lake Superior, University of Minnesota Medical School, PennWell Corp., and Seoul Semiconductor Co. Ltd. S.M.W.R. is a program leader and serves as a consultant to the CRC for Alertness, Safety and Productivity, Australia; reports receiving research grants from the CRC for Alertness, Safety and Productivity, Philips Respironics, Rio Tinto, Shell, Linfox Australia, and Teva Pharma Australia; and has received equipment support and consultancy fees through his institution from Vanda Pharmaceuticals, Optalert, Tyco Healthcare, Compumedics, BHP, and Teva Pharmaceuticals, which are not related to this paper. C.A.C. reports grants from Cephalon Inc.; Jazz Pharmaceuticals Plc., Inc.; National Football League Charities; Optum; Philips Respironics, Inc.; Regeneron Pharmaceuticals; ResMed Foundation; San Francisco Bar Pilots; Sanofi S.A.; Sanofi-Aventis, Inc.; Schneider Inc.; Sepracor, Inc.; Mary Ann & Stanley Snider via Combined Jewish Philanthropies; Sysco; Takeda Pharmaceuticals; Teva Pharmaceuticals Industries, Ltd.; and Wake Up Narcolepsy and personal fees from Bose Corporation; Boston Celtics; Boston Red Sox; Cephalon, Inc.; Columbia River Bar Pilots; Ganésco Inc.; Institute of Digital Media and Child Development; Klarman Family Foundation; Samsung Electronics; Quest Diagnostics, Inc.; Teva Pharma Australia; Vanda Pharmaceuticals; Washington State Board of Pilotage Commissioners; and Zurich Insurance Company, Ltd. In addition, C.A.C. holds a number of process patents in the field of sleep/circadian rhythms (e.g., photic resetting of the human circadian pacemaker) and holds an equity interest in Vanda Pharmaceuticals, Inc. Since 1985, C.A.C. has also served as an expert on various legal and technical cases related to sleep and/or circadian rhythms, including those involving the following commercial entities: Casper Sleep Inc., Comair/Delta Airlines, Complete General Construction Company, Federal Express, Greyhound, HG Energy LLC, Purdue Pharma L.P., South Carolina Central Railroad Co., Steel Warehouse Inc., Stric-Lan Companies LLC, Texas Premier Resource LLC, and United Parcel Service. C.A.C. receives royalties from the New England Journal of Medicine; McGraw Hill; Houghton Mifflin Harcourt/Penguin; and Philips Respironics, Inc., for the Actiwatch 2 and Actiwatch Spectrum devices. C.A.C.’s interests were reviewed and managed by Brigham and Women’s Hospital and Partners HealthCare in accordance with their conflict of interest policies. S.W.L. has had a number of commercial interests in the last 24 months (2017 to 2019). No other interests are directly related to the research or topic reported in this paper but in the interests of full disclosure, are mentioned. S.W.L. has received consulting fees from BHP Billiton, EyeJust Inc., Noble Insights, and Team C Racing; honoraria and/or paid travel from BHP Billiton, DIN, IES, Ineos, SLTBR, and Teague; has current consulting contracts with Akili Interactive, Apex 2100 Ltd., Consumer Sleep Solutions, Headwaters Inc., Hintsa Performance AG, Light Cognitive, Lighting Science Group Corporation, Mental Workout, PlanLED, Six Senses, Stantec, and Wyle Integrated Science and Engineering; has received unrestricted equipment gifts from Bionetics Corporation and F. Lux Software LLC and royalties from Oxford University Press; and has served as a paid expert in legal proceedings related to light, sleep, and health.

Data Availability: Restrictions apply to the availability of data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

Abbreviations:

Abbreviations:
 
• CR

  constant routine

 
• E2

  estradiol

 
• P4

  progesterone

References and Notes