Objective sleep duration and timing predicts completion of in vitro fertilization cycle

Purpose To examine associations between objectively measured sleep duration and sleep timing with odds of completion of an in vitro fertilization (IVF) cycle.

This prospective cohort study enrolled 48 women undergoing IVF at a large tertiary medical center between 2015 and 2017. Sleep was assessed by wrist-worn actigraphy, 1–2 weeks prior to initiation of the IVF cycle. Reproductive and IVF cycle data and demographic and health information were obtained from medical charts. Sleep duration, midpoint, and bedtime were examined in relation to IVF cycle completion using logistic regression models, adjusted for age and anti-Müllerian hormone levels. A sub-analysis excluded women who worked non-day shifts to control for circadian misalignment.

The median age of all participants was 33 years, with 29% of women >35 years. Ten women had an IVF cycle cancelation prior to embryo transfer. These women had shorter sleep duration, more nocturnal awakenings, lower sleep efficiency, and later sleep timing relative to those who completed their cycle. Longer sleep duration was associated with lower odds of uncompleted IVF cycle (OR = 0.88; 95%CI 0.78, 1.00, per 20-min increment of increased sleep duration). Women with later sleep midpoint and later bedtime had higher odds of uncompleted cycle relative to those with earlier midpoint and earlier bedtime; OR = 1.24; 95%CI 1.09, 1.40 and OR = 1.33; 95%CI 1.17, 1.53 respectively, for 20-min increments. These results were independent of age, anti-Müllerian hormone levels, or sleep duration, and remained significant after exclusion of shift-working women.

Shorter sleep duration and later sleep timing increase the odds of uncompleted cycles prior to embryo transfer.

In vitro fertilization, Assisted reproductive technology, Infertility, Subfertility, Actigraphy, Sleep, Embryo transfer, Bedtime

The prevalence of infertility among reproductive-age couples in the USA is between 7 and 16% [1, 2]. According to global estimates, more than half of infertile couples seek medical care, but less than a quarter receive treatment [3]. In vitro fertilization (IVF), while commonly pursued by women who have been unsuccessful at natural conception, is costly and provides age-dependent pregnancy rates that average at 50% [4, 5]. Infertility and IVF treatments may provoke stress and anxiety, which could negatively impact clinical pregnancy rates [6]. Stress and anxiety are inextricably linked to sleep disturbances [7], and these relationships have been suggested to influence infertility [8]. A recent investigation of nearly 7000 women demonstrated that self-reported “trouble sleeping” was associated with increased time to pregnancy, while a weak association was found between short sleep duration and decreased fecundability [9]. Women undergoing IVF may be particularly vulnerable to sleep disturbances. Indeed, short sleep duration and poor sleep quality have been reported in 30–50% of women undergoing IVF, and disturbed sleep has been linked to stress and anxiety during the IVF cycle [10–12]. Data regarding the impact of sleep on non-psychometric IVF outcomes are limited. A small pilot study has demonstrated a link between sleep duration and oocytes retrieved, with a 50% increase in the number of oocytes retrieved for every 1-h increase in total sleep time [12]. Taken together, these findings suggest that sleep could be relevant in IVF cycles. Optimal sleep quality and duration requires appropriate entrainment of the circadian timekeeping system. A growing body of literature implicates sleep and circadian rhythms in fertility. However, a previous work has been limited by the use of cross-sectional studies, self-reported sleep measures, or use of shift work as a marker for circadian disruption that excludes widespread circadian misalignment (“social jetlag”), experienced by larger populations [13, 14]. Moreover, examination of sleep duration and the circadian timing of sleep specifically in relation to IVF outcomes is lacking. Sleep is a critical and modifiable determinant of wellness, and thus, identification of its contribution to successful IVF cycles is necessary. Therefore, the objective of this study was to examine associations between objectively measured sleep duration and timing with completion of IVF cycles.

Study participants Women who underwent IVF at a large Midwestern center for Reproductive Medicine from 2015 to 2017 were invited to participate. Women who used donor oocytes or were undergoing IVF for the purpose of fertility preservation were excluded. Written informed consent was obtained from all women, and this study was approved by the local Institutional Review Board (HUM 00071566). Demographic and reproductive data Demographic and reproductive information were extracted from medical records, including baseline age; race/ethnicity; body mass index (BMI) measured clinically during the initial visit; reproductive history; indication for IVF; and baseline levels of anti-Müllerian hormone (AMH, ng/ml), antral follicle count (AFC), day 3 follicle-stimulating hormone (FSH, mIU/ml), and estradiol (pg/ml). Cycle characteristics and outcomes were also extracted, including stimulation protocol, stimulation duration, cumulative gonadotropin dose (IU), peak estradiol level (pg/ml), number of oocytes retrieved, number of mature oocytes, method of fertilization (conventional IVF or intracytoplasmic sperm injection (ICSI)), number of fertilized oocytes, preimplantation genetic testing for aneuploidies (PGT-A) (yes or no), type of embryo transfer (fresh or frozen), and number of embryos transferred. Work status and work hour information was collected during the initial visit through questionnaires. Mental health data During the initial visit, all women completed the Concerns of women undergoing Assisted Reproductive Technology (CART) scale. The CART is a 10-item, 3-point Likert scale that assesses concern levels related to different components of the IVF process. Possible responses range from “little or no concern,” “moderate concern,” and “extremely concerned” corresponding to 1, 2, and 3 points, respectively. IVF cycle completion An IVF cycle that did not reach the embryo transfer stage was defined as an uncompleted cycle. Reasons for uncompleted cycles were [1] poor ovarian response to stimulation resulting in cancelation prior to oocyte retrieval, [2] embryos that did not reach the blastocyst stage for biopsy/cryopreservation in women planning PGT-A and/or a frozen transfer; or [3] no euploid blastocysts available for transfer following PGT-A. Clinical protocols of IVF cycles Women underwent controlled ovarian hyperstimulation using either the gonadotropin-releasing hormone (GnRH) agonist protocol or the GnRH antagonist protocol. Protocol selection was at the discretion of the treating physician. Initial gonadotropin dosage was determined by age and ovarian reserve and adjusted according to ovarian response to stimulation, which was assessed by regular ultrasound examinations and serum estradiol measurements. When ultrasound examination revealed at least two follicles measuring 18 mm or greater, final follicular maturation was triggered with a subcutaneous injection of human chorionic gonadotropin (250 μg). Thirty-six hours after trigger, women underwent transvaginal oocyte retrieval under ultrasound guidance. Fertilization was performed by conventional in vitro fertilization or ICSI using standard techniques. In women who opted for PGT-A, trophectoderm biopsy was performed on embryos that reached the blastocyst stage. Following biopsy, blastocysts were then vitrified in liquid nitrogen. The number of embryos transferred was based on the American Society for Reproductive Medicine guidelines [15]. Women planning a fresh transfer underwent an embryo transfer either 3 or 5 days following the retrieval. All women took both estradiol and natural micronized progesterone for luteal support. A serum beta hCG (human chorionic gonadotropin) assay was performed 9–11 days after embryo transfer. Biochemical pregnancy was defined as the presence of a positive serum hCG. Women with a positive beta hCG subsequently underwent a transvaginal ultrasound between 6 and 7 weeks gestation to determine fetal viability. Clinical pregnancy was defined as the presence of a gestational sac and fetal cardiac activity on ultrasound. Live birth was defined as a birth of one or more live born infant(s) at a gestational age ≥20 weeks. Sleep assessment Sleep measures were objectively collected using an actigraph , a wrist-worn device that quantifies patterns of movement. This device continuously records motion and discerns sleep from wake state, under the premise that movement is more infrequent during sleep than in wakefulness [16]. Comparable to polysomnography, the gold-standard device for sleep assessment, actigraphy is a validated instrument for measurement of sleep in individuals both with and without sleep disorders [17]. To evaluate baseline sleep patterns, women were asked to wear an actigraph (Actiwatch 2, Philips Respironics) on their non-dominant wrist continuously for seven days in addition to completing a sleep diary, prior to the start of the stimulation protocol. Activity and light exposure were recorded and summarized over 1-min epochs. Rest intervals, i.e., periods when women were attempting sleep, or were scored based on validated criteria [18]. Briefly, rest intervals were defined by four hierarchical inputs: event markers, records from sleep diary, light intensity, and activity levels respectively. For all four inputs, concordance within 15 min of at least two inputs were assessed and the higher rank input was selected for rest interval start or end time. If no pair input was concordant within 15 min, the data were then reassessed for a 30-min concordance interval. When none of the inputs fit into a 30-min agreement, activity level was used to define rest interval start or rest interval end. Based on the sleep diary, all rest intervals across the day were marked either as the main sleep bout or as a nap. Rest intervals were manually defined within the Actiware software platform (version 6.0.9, Philips Respironics, Murrysville, PA, USA). Within each rest interval, sleep and wake were differentiated by the validated, proprietary algorithm within the Actiware software platform set at medium sensitivity (wake threshold of 40) and the following sleep parameters were computed: total sleep time, sleep onset latency, i.e., time from start of rest interval to sleep onset, awakenings, wake after sleep onset (WASO), and sleep efficiency, i.e., the proportion of time in bed that was spent sleeping, bedtime, and rise time. The midpoint of sleep was determined from the half-way point between bedtime and rise time. Statistical analysis Demographic measures, BMI, working hours, and reproductive and IVF cycle characteristics were evaluated overall and by cycle completion status. Among women with completed cycles, we further evaluated embryo characteristics and pregnancy outcomes. Violin and box plots were generated to describe the probability density and quartiles of mean sleep duration and mean sleep midpoint of all women. We examined associations between several sleep measures and IVF cycle completion status. Specifically, using the cycle-completed group as the referent, weighted logistic regression models were implemented to assess the odds of uncompleted IVF cycle according to sleep duration, midpoint, and bedtime. Weights represented the number of actigraphy-recorded nights per woman. Sleep midpoint and bedtime were defined in relation to midnight on a minute scale. For example, a midpoint or bedtime of 11 PM was transformed to −60, representing time point that is 60 min prior to midnight. For this analysis, we divided sleep duration, midpoint, and bedtime by 20 to create 20-min intervals, suggested as clinically meaningful sleep quantity. Unadjusted and adjusted odds ratios and 95% confidence intervals were estimated (R statistical software 4.0.2); all adjusted models included age and AMH level, a sensitive biomarker of ovarian reserve [19]. Weights representing the number of actigraphy-recorded nights per woman were also included in adjusted models. To examine the influence of sleep timing independent of sleep duration, we accounted for sleep duration in regression models with sleep midpoint or sleep bedtime as the exposures. Given that shift work has been shown to influence sleep characteristics and IVF outcomes [20], we performed a sensitivity analysis restricting the sample to women who did not report shiftwork.

This study enrolled 48 women who met the inclusion criteria. Demographic information of study participants is shown in Table 1. Most women were younger than 35 years, predominantly white and non-obese. On average, women who completed their IVF cycle were 4 years younger than those whose cycle was canceled prior to embryo transfer. Only 15% of women reported non-day shift work. Of the 48 participants, 38 women (79%) underwent an embryo transfer (completed cycles group). Women <35 years compared to those older than 35 years were less likely to have an uncompleted cycle, while the proportion of obesity was similar between groups. The median CART score was similar in women with uncompleted vs. with completed IVF cycles. Table 1. | All women n=48 (%) | 95% CI | Uncompleted cycle an=10 (%) | 95% CI | Completed cycle n=38 (%) | 95% CI | | |---|---|---|---|---|---|---| | Age (y, range) | 33 (25-42) | (32, 34) | 36 (32-42) | (35, 39) | 32 (25-37) | (31, 33) | | Age group (years) | |||||| | <35 c | 34 (71) | (57, 82) | 2 (20) | (5, 52) | 32 (84) | (69, 93) | | 35-37 c | 11 (23) | (13, 37) | 5 (50) | (24, 76) | 6 (16) | (7, 31) | | 38-42 c | 3 (6) | (2, 18) | 3 (30) | (11, 61) | 0 (0) | (0, 11) | | Race | |||||| | White c | 45 (94) | (82, 98) | 8 (80) | (48, 95) | 37 (97) | (85, 100) | | Non-white c | 3 (6) | (2, 18) | 2 (20) | (5, 52) | 1 (3) | (0, 15) | | BMI (kg/m2) d | 25.7 (20-40) | (24.5, 26.9) | 26.3 (21-36) | (14.2, 38.5) | 25.7 (20-40) | (24.6, 26.7) | | BMI30 c | 8 (17) | (9, 30) | 2 (20) | (5, 52) | 6 (16) | (7, 31) | | Work hours (h/week) b, d | 40 (8-65) | (40, 40) | 40 (36-65) | (40, 40) | 40 (8-65) | (40, 40) | | Non-day shift workers b, c | 7 (15) | (7, 29) | 1 (13) | (1, 50) | 6 (16) | (7, 31) | | CART Score (Median) d, e | 1.8(1.1-2.6) | (1.8, 1.8) | 1.8(1.1-2.6) | (1.7, 1.8) | 1.7(1.3-2.4) | (1.5, 2.0) | aCycle canceled prior to embryo transfer bWork status data is missing for two women cConfidence intervals for proportions are computed using the Wilson method dConfidence intervals for medians are calculated with the Jackknife method eCART score data is missing for three women Median AMH and AFC were significantly correlated (Pearson’s correlation coefficient = 0.76), and higher in the completed IVF cycle group in comparison to women with an uncompleted cycle (Table 2). Most of the women (77%) were nulliparous. The most common infertility diagnoses were ovulatory dysfunction (44%), male factor (40%), and diminished ovarian reserve (19%). The proportion of completed cycles was higher among women reporting ovulatory dysfunction or male factor diagnoses compared with other infertility diagnoses. Table 2. | Median (range) | All womenn=48 (%) | 95%CI | Uncompleted cycle an=10 (%) | 95%CI | Completed cyclen=38 (%) | 95%CI | |---|---|---|---|---|---|---| | Anti-Müllerian hormone (ng/ml) | 3.15 (0.5-38) | (2.48, 3.82) | 1.0 (0.5-6.8) | (1, 1) | 3.4 (0.6-38) | (2.2, 4.6) | | Antral follicle count | 21.0 (4-84) | (14.8, 28.2) | 14.5 (4-45) | (11.6, 17.4) | 24.0 (11-84) | (18.5, 30.5) | | Estradiol, day 3 (pg/ml)b | 38.5 (6-104) | (31.9, 45.1) | 40.5 (26-69) | (31.7, 49.3) | 38.0 (6-104) | (38, 38) | | Follicle Stimulating hormone, day 3 (mIU/ml)b | 7.1 (2-13) | (6.4, 7.7) | 8.55 (6-13) | (8.3, 8.8) | 6.80 (2-12) | (6.8, 6.8) | | Gravidity | 0 (0-5) | (0, 0) | 0.5 (0-3) | (0, 3) | 0 (0-5) | (0, 0) | | No prior pregnancies | 26 (54) | (40, 67) | 5 (50) | (24, 76) | 21 (55) | (40, 70) | | Parity | 0 (0-3) | (0, 0) | 0 (0-2) | (0, 0) | 0 (0-3) | (0, 0) | | No prior deliveries | 37 (77) | (63, 87) | 8 (80) | (48, 95) | 29 (76) | (61, 87) | | Indication for IVF c | |||||| | Ovulatory dysfunction | 21 (44) | (31, 58) | 3 (30) | (11, 61) | 18 (47) | (33, 63) | | Male factor | 19 (40) | (29, 56) | 3 (30) | (11, 61) | 16 (42) | (30, 60) | | Diminished ovarian reserve | 9 (19) | (9, 30) | 6 (60) | (31, 83) | 3 (8) | (0, 18) | | Unexplained | 7 (15) | (7, 28) | 1 (10) | (0, 43) | 6 (16) | (7, 31) | | Recurrent pregnancy loss | 4 (8) | (3, 20) | 1 (10) | (0, 43) | 3 (8) | (2, 22) | | Uterine factor | 3 (6) | (2, 18) | 1 (10) | (0, 43) | 2 (5) | (1, 18) | | Endometriosis | 3 (6) | (2, 18) | 0 (0) | (0, 33) | 3 (8) | (2, 22) | | Tubal dysfunction | 2 (4) | (0, 15) | 1 (10) | (0, 43) | 1 (3) | (0, 15) | aCycle canceled prior to embryo transfer bFollicle-stimulating hormone and estradiol is missing in two women cSixteen women had more than one indication for IVF Most women who completed their IVF cycle had, on average, sufficient sleep in comparison to those with uncompleted cycles (Fig. 1). Similarly, an earlier sleep midpoint, that is, within the recommended 2 AM–4 AM range, was more common among women who completed their IVF cycle (Fig. 2). The GnRH antagonist protocol was the most commonly used stimulation protocol (Table 3). Despite having a similar number of stimulation days, women with uncompleted cycles received a higher median cumulative dose of gonadotropins relative to women with completed IVF cycles (>3600 IU vs. 2775 IU). Furthermore, median peak estradiol levels at the end of stimulation were lower in women with uncompleted vs. completed cycles, 724 pg/ml and 2669 pg/ml, respectively. The median number of retrieved oocytes, mature oocytes, and fertilized oocytes, and use of ICSI were higher in women who completed their IVF cycle in comparison to those of women with uncompleted cycles. Of the 38 women who completed their IVF cycle, i.e., underwent an embryo transfer, the majority (60%) had fresh transfers, and all but two had a single embryo transfer (Table 4). Table 3. | Median (range) | All women n=48 (%) | 95% CI | Uncompleted cycle an=10 (%) | 95% CI | Completed cycle n=38 (%) | 95% CI | |---|---|---|---|---|---|---| | Stimulation protocol | |||||| | GnRH antagonist | 36 (75) | (61, 85) | 9 (90) | (57, 100) | 27 (71) | (55, 83) | | GnRH agonist | 12 (25) | (15, 39) | 1 (10) | (0, 43) | 11 (29) | (17, 45) | | Number of stimulation days | 11 (8-17) | (11, 11) | 10.5 (8-14) | (7.6, 13.4) | 11 (8-17) | (11, 11) | | Cumulative gonadotropin dose (IU)b | 2963 (1650-5250) | (2459, 3466) | 3638(2023-5250) | (2535, 4740) | 2775(1650-5175) | (2775, 2775) | | Peak estradiol level (pg/mL) | 2539(23-5825) | (2351, 2727) | 723.5(23-2331) | (0, 1679) | 2669(717-5825) | (2538, 2800) | | Number of oocytes retrieved c | 18 (3-54) | (18, 18) | 8.5 (3-26) | (6.3, 10.7) | 19.5 (4-54) | (13.5, 25.5) | | Number of mature oocytes c | 14 (3-44) | (14, 14) | 6.5 (3-22) | (4.3, 8.7) | 14.5 (3-44) | (8.5, 20.5) | | Fertilization method c | |||||| | In-vitro fertilization | 14 (32) | (20, 47) | 3 (50) | (19, 81) | 11 (29) | (17, 45) | | Intracytoplasmic sperm injection | 30 (68) | (53, 80) | 3 (50) | (19, 81) | 27 (71) | (55, 83) | | Number of fertilized Oocytes c | 11 (2-34) | (4, 17) | 4.0 (2-15) | (4, 4) | 11 (2-34) | (11, 11) | | PGT-Ad | 12 (27) | (19, 47) | 1 (17) | (2, 54) | 11 (30) | (17, 45) | aCycle canceled prior to embryo transfer bInternational units cNumber of oocytes retrieved, mature oocytes, fertilization method, fertilized oocytes, and PGT-A were calculated in 44 women who had successful oocyte retrieval dPreimplantation genetic testing for aneuploidies Table 4. | Completed cyclen=38 (%) | 95% CI | | |---|---|---| | Type of embryo transfer | || | Fresh | 22 (58) | (42, 72) | | Frozen | 16 (42) | (28, 58) | | Number of embryos transferred a | 1 (1-2) | (1, 1) | | Pregnancy outcomes | || | Biochemical pregnancy | 28 (74) | (58, 85) | | Clinical pregnancy | 28 (74) | (58, 85) | | Live birth b | 23 (61) | (45, 74) | a Median (range) b Representing pregnancy rate of 48% among all women (n=48) Women with uncompleted cycles were more likely to sleep <7 h, have more nocturnal awakenings, greater wake after sleep onset, lower sleep efficiency, and later mean midpoint of sleep, later bedtime and rise time in comparison to those in the completed cycle group (Table 5). Table 5. | N (%) b | Uncompleted cycle b n=10 (%) | Completed cyclen n=38 (%) | |---|---|---| | Mean Nocturnal Sleep Time, min (SD) | 428 (91) | 433 (41) | | Mean Sleep Duration group | || | <420 min | 6 (60) | 14 (37) | | ≥420 min | 4 (40) | 24 (63) | | Mean Nocturnal Sleep Time variability, min (SD) | 60 (28) | 65 (35) | | Mean Sleep Onset Latency, min (SD) | 9 (8) | 12 (12) | | Mean Wake After Sleep Onset, min (SD) | 60 (21) | 57 (27) | | Mean Number of Awakening (SD) | 33 (9) | 28 (7) | | Mean Sleep efficiency, min (SD) | 82 (4) | 84 (6) | | Nap | || | Yes | 2 (20) | 7 (18) | | No | 8 (80) | 31 (82) | | Mean Nap Duration, min (SD) | 49 (40) | 84 (88) | | Mean Sleep Midpoint (range) | 3:43 AM (2:01 AM – 5:17 AM) | 3:29 AM (1:45 AM – 8:57 AM) | | Mean Sleep Midpoint variability, min (SD) | 60 (25) | 59 (56) | | Mean Bedtime (range) | 11:23 PM (9:55 PM – 01:22 AM) | 11:11 PM (8:43 PM – 04:16 AM) | | Mean Bedtime variability, min (SD) | 61 (23) a | 64 (68) | | Mean Rise Time (range) | 8:03 AM (6:01 AM – 11:03 AM) | 07:47 AM (5:58 AM – 01:39 PM) | | Mean Rise Time variability, min (SD) | 73 (38) | 74 (51) | aExcluding one woman who had one day of sleep data bCycle cancelation prior to embryo transfer; a midpoint or bedtime of 11 PM was transformed to −60, representing time point that is 60 min prior to midnight Analyses adjusted for AMH and age suggested significant associations between sleep characteristics and cycle completion status, among all women and in a restricted sample of non-shift workers (n = 39) (Table 6). Increased sleep duration, in 20-min intervals, was associated with lower odds of canceled IVF cycles prior to embryo transfer; (OR = 0.88; 95% CI, 0.78, 1.00). Similarly, later sleep midpoint and bedtime were associated with increased odds for uncompleted cycles, OR = 1.24 (95% CI, 1.09, 1.40) and 1.33 (95% CI 1.17, 1.53), respectively, per 20-min intervals. Exclusion of non-day shift workers (n = 9) produced more pronounced associations; ORs for sleep duration, midpoint, and bedtime were 0.67 (95% CI, 0.54, 0.81), 1.32 (95% CI 1.10, 1.60), and 1.47 (95% CI, 1.24, 1.77), respectively. Further adjustment for sleep duration, in regression models with sleep midpoint or sleep bedtime as the exposures, had little impact on the odds of uncompleted cycles (Table 6). Table 6. | Odds Ratio of uncompleted Cycle (95% CI)a | |||| |---|---|---|---|---| | All Women (N=48) | Non Shift Workers (N=39) | ||| | Unadjusted | Adjusted b | Unadjusted | Adjusted b | | | Sleep Duration | 0.90(0.80, 1.01) | 0.88(0.78, 1.00) | 0.73(0.60, 0.86) | 0.67(0.54, 0.81) | | Sleep Midpoint | 1.05(0.99, 1.12) | 1.24(1.09, 1.40) | 1.16(1.00, 1.34) | 1.32(1.10, 1.60) | | Sleep Midpoint adjusted for Sleep Duration | 1.06(0.99, 1.13) | 1.26(1.11, 1.43) | 1.11(0.96, 1.30) | 1.29(1.06, 1.57) | | Bedtime | 1.07(1.00, 1.14) | 1.33(1.17, 1.53) | 1.19(1.05, 1.35) | 1.47(1.24, 1.77) | | Bedtime adjusted for Sleep Duration | 1.06(0.99, 1.13) | 1.32(1.15, 1.52) | 1.08(0.93, 1.25) | 1.31(1.07, 1.62) | aFrom a logistic model with cycle cancelation as binary outcome and sleep measurement as predictors; sleep measures formulated into 20-min intervals; a midpoint or bedtime of 11 PM was transformed to −60, representing time point that is 60 min prior to midnight bAdjusted for AMH and age

This study examined associations between objectively measured sleep and IVF cycle completion in women initiating IVF treatment. Our findings suggest that women who slept <7 h had a higher likelihood of canceled cycles prior to embryo transfer in comparison to those who slept at least 7 h per night, and an increase of 20-min in sleep duration was associated with lower odds of cycle cancelations prior to embryos transfer. Furthermore, women who had later sleep midpoints or later bedtimes had 1.24 and 1.33 times the odds of uncompleted cycles, respectively. These findings provide evidence for an association of key sleep behaviors on a successful completion of the IVF cycle. Optimal sleep duration for adults is defined as 7–9 h per day [21]. In the present study, objectively assessed insufficient sleep was prevalent in almost half of women, consistent with data from the National Health and Nutrition Examination Survey in reproductive-aged women [22]. Reports on the influence of sleep duration on reproductive function are lacking, though tend to suggest that longer sleep is associated with changes in several reproductive hormones and potentially fecundability as well. A study of 106 reproductive-aged women observed a 20% increase in FSH levels among long sleepers (≥8 h/night, self-reported) in comparison to women who slept less than 8 h daily [23]. Furthermore, a recent report in 259 reproductive-aged women has linked increased sleep duration with higher levels of estradiol and progesterone [24]. Associations between sleep duration, trouble sleeping, and fecundability have been examined in a large sample of 6873 North American women who were attempting a pregnancy for less than 6 months [25]. Lower fecundability was suggested among short sleepers (<6 h) in comparison to that among sufficient sleepers (8 h/night), though statistically insignificant. Sleep duration of 6, 7, or 9 h per night had negligible or no influence on fecundability. Among women undergoing IVF, a pilot study demonstrated a linear relationship between objective assessment of sleep duration and oocyte retrieval, such that an additional 1 h of sleep duration was associated with a 50% increase in the number of oocytes retrieved [12]. The present study showed that a larger proportion of women with uncompleted IVF cycles slept less than the recommended 7 h compared to those with completed cycles. Furthermore, increases in sleep duration were significantly associated with lower odds of an uncompleted IVF cycle. These data highlight the promising influence of sufficient sleep on IVF outcomes and reinforce the need to elucidate its impact on the reproductive axis. The present study also demonstrates that sleep timing — independent of sleep duration — is associated with uncompleted IVF cycles. Shift workers manifest the most profound degree of circadian misalignment, given the overlap of the work period with the endogenous time when sleep is promoted. Supportive of the circadian contribution to reproduction, ample evidence demonstrates decreased fertility and increases in miscarriage among female shift workers [26–32]. Most recently, in the largest investigation to date, more than 20,000 pregnant women were found to have an increased hazard of miscarriage when they participated in more than 2 nights per week of shift work (hazard ratio 1.32 [95% CI 1.07 to 1.62]) [28] The impact of circadian disturbance on infertility treatments is less clear, but emerging evidence suggests circadian disturbance may also be detrimental to outcomes in women undergoing IVF. The start of daylight savings time (i.e., the spring time change) represents a ubiquitous and abrupt misalignment between the endogenous circadian rhythm and external schedule, and was associated with pregnancy loss after embryo transfer in women undergoing IVF [33]. However, a large cohort study reported no association between seasonal variation and IVF success [34]. A meta-analysis demonstrated higher odds of infertility and menstrual disruptions in shift-working women compared with non-shift workers, likely attributed to short sleep duration and circadian misalignment [9, 20]. In women undergoing controlled ovarian hyperstimulation, evening, night, or rotating shift work was associated with fewer mature oocytes relative to women with day work [35]. Our finding is in line with prior evidence of poor reproductive function among shift-working women while extending the negative impact of circadian misalignment to women undergoing IVF treatment. Additional work is needed to fully elucidate the influence of “social jetlag” — defined as the difference between preferred sleep timing according to a molecular clock and actual sleep time dictated by work schedules — [36] on reproductive outcomes. Biological mechanisms Insufficient sleep has been suggested to affect fertility through multiple pathways including alterations in hypothalamic-pituitary-adrenal axis (HPA), changes in hormone levels, sympathetic hyperactivation, and increase in oxidative stress [8, 37]. These phenomena are associated with poor IVF success rate [38–41]. The circadian rhythm aligns physiological functions with the day-night cycle. The suprachiasmatic nucleus (SCN) of the hypothalamus generates circadian oscillations that tightly regulate sleep-wake behavior, feeding, and endocrine rhythms and time, and various other physiological functions, including the reproduction. Extensive work has established circadian control of the luteinizing hormone (LH) surge in rodents which occurs just before the active locomotor period [42]. Like rodents, humans also demonstrate a circadian rhythm of ovulatory hormone secretion with the LH surge occurring prior to the active period [43–45]. Animal studies have also demonstrated clock gene expression in the ovaries, oviducts, uterus, and oocytes [46]. This peripheral clock machinery in reproductive organs (and early embryos) and central coordination by the SCN may underlie the observation that, in animals, clock gene mutations, alterations in mating time, increasing day length, and externally imposed phase shifts are detrimental to conception and maintenance of pregnancy [47–50]. Indeed, disruption of this tightly synchronized rhythm is a contributor to impaired fertility. Late sleep midpoint is a marker for circadian misalignment, representing a discrepancy between sleep-wake cycles in relation the biological night. This study has several strengths. The cohort of 48 women undergoing IVF was well-characterized with objective sleep assessment, reproductive profile, and controlled stimulation protocols (Table 6). Moreover, their sleep duration and midpoint were similar to that reported by the American Time Use Survey, a nationally representative sample of US residents [51]. The single site location provided consistency of protocols, procedures, and clinical care. While our sample demonstrated a homogeneous sociodemographic profile that differs from that of all US women who experience infertility, it represents US women who pursue IVF. However, this study has several limitations. Women in this cohort are mostly white, which reduces the generalizability of the findings. Fifteen women had their first IVF cycle, and information on prior IVF cycles for the other women was not available. While the present study has controlled for common underlying causes in the analysis — i.e., age, AMH, and shift work — its observational design cannot indicate causal relationship. We acknowledge that PGT-A procedures could impact the completion of IVF cycles. However, we were unable to adjust for PGT-A procedures in the analysis, as only one woman who used PGT-A had an uncompleted IVF cycle. Furthermore, while PGT-A utilized cryopreserved embryos, the blastocyst thaw rate was 100% which did not influence the likelihood of cycle completion.

This novel study of women undergoing IVF demonstrates a significant association between shorter sleep duration, later sleep midpoint, later bedtime, and lower odds of IVF cycle completion. As sleep is a modifiable risk factor, strategies to improve sleep may contribute to optimal implementation of IVF protocols. Future investigations should involve larger samples and evaluate the role of anxiety and depression in the relationship between sleep-wake timing and successful completion of IVF cycles. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

- 1.Thoma ME, McLain AC, Louis JF, King RB, Trumble AC, Sundaram R, et al. Prevalence of infertility in the United States as estimated by the current duration approach and a traditional constructed approach. Fertil Steril. 2013;99:1324–31.e1. doi: 10.1016/j.fertnstert.2012.11.037. [DOI] [PMC free article] [PubMed] [Google Scholar] - 2.Louis JF, Thoma ME, Sørensen DN, McLain AC, King RB, Sundaram R, et al. The prevalence of couple infertility in the United States from a male perspective: evidence from a nationally representative sample. Andrology. 2013;1:741–748. doi: 10.1111/j.2047-2927.2013.00110.x. [DOI] [PMC free article] [PubMed] [Google Scholar] - 3.Boivin J, Bunting L, Collins JA, Nygren KG. International estimates of infertility prevalence and treatment-seeking: potential need and demand for infertility medical care. Hum Reprod. 2007;22:1506–1512. doi: 10.1093/humrep/dem046. [DOI] [PubMed] [Google Scholar] - 4.De Geyter C, Calhaz-Jorge C, Kupka MS, Wyns C, Mocanu E, Motrenko T, et al. ART in Europe, 2014: results generated from European registries by ESHRE: the European IVF-monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE) Hum Reprod. 2018;33:1586–1601. doi: 10.1093/humrep/dey242. [DOI] [PubMed] [Google Scholar] - 5.SART. Society of Assisted Reproductive Technology national summary report 2017. In, 2019. - 6.Matthiesen S, Frederiksen Y, Ingerslev HJ, Zachariae RJHr. Stress, distress and outcome of assisted reproductive technology (ART): a meta-analysis 2011;26:2763-76. [DOI] [PubMed] - 7.Kalmbach DA, Anderson JR, Drake CLJJosr. The impact of stress on sleep: pathogenic sleep reactivity as a vulnerability to insomnia and circadian disorders 2018;27:e12710. [DOI] [PMC free article] [PubMed] - 8.Kloss JD, Perlis ML, Zamzow JA, Culnan EJ, Gracia CR. Sleep, sleep disturbance, and fertility in women. Sleep Med Rev. 2015;22:78–87. doi: 10.1016/j.smrv.2014.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar] - 9.Willis SK, Hatch EE, Wesselink AK, Rothman KJ, Mikkelsen EM, Wise LAJF et al. Female sleep patterns, shift work, and fecundability in a North American preconception cohort study 2019;111:1201-10. e1. [DOI] [PMC free article] [PubMed] - 10.Lin YH, Chueh KH, Lin JLJJocn. Somatic symptoms, sleep disturbance and psychological distress among women undergoing oocyte pick-up and in vitro fertilisation–embryo transfer 2016;25:1748-56. [DOI] [PubMed] - 11.Huang L-H, Kuo C-P, Lu Y-C, Lee M-S, Lee S-HJTJO. Gynecology. Association of emotional distress and quality of sleep among women receiving in-vitro fertilization treatment. 2019;58:168–172. doi: 10.1016/j.tjog.2018.11.031. [DOI] [PubMed] [Google Scholar] - 12.Goldstein CA, Lanham MS, Smith YR, O’Brien LM. Sleep in women undergoing in vitro fertilization: a pilot study. Sleep Med. 2017;32:105–113. doi: 10.1016/j.sleep.2016.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar] - 13.Mills J, Kuohung W. Impact of circadian rhythms on female reproduction and infertility treatment success. Curr Opin Endocrinol Diabetes Obes. 2019;26:317–321. doi: 10.1097/MED.0000000000000511. [DOI] [PubMed] [Google Scholar] - 14.Willis SK, Hatch EE, Wise LA. Sleep and female reproduction. Curr Opin Obstet Gynecol. 2019;31:222–227. doi: 10.1097/GCO.0000000000000554. [DOI] [PubMed] [Google Scholar] - 15.Penzias A, Bendikson K, Butts S, Coutifaris C, Falcone T, Fossum G, et al. ASRM standard embryo transfer protocol template: a committee opinion. Fertil Steril. 2017;107:897–900. doi: 10.1016/j.fertnstert.2017.02.108. [DOI] [PubMed] [Google Scholar] - 16.Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP. The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 2003;26:342–392. doi: 10.1093/sleep/26.3.342. [DOI] [PubMed] [Google Scholar] - 17.Smith MT, McCrae CS, Cheung J, Martin JL, Harrod CG, Heald JL, et al. Use of actigraphy for the evaluation of sleep disorders and circadian rhythm sleep-wake disorders: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2018;14:1231–1237. doi: 10.5664/jcsm.7230. [DOI] [PMC free article] [PubMed] [Google Scholar] - 18.Patel SR, Weng J, Rueschman M, Dudley KA, Loredo JS, Mossavar-Rahmani Y, Ramirez M, Ramos AR, Reid K, Seiger AN, Sotres-Alvarez D, Zee PC, Wang R. Reproducibility of a standardized actigraphy scoring algorithm for sleep in a US Hispanic/Latino population. Sleep. 2015;38:1497–1503. doi: 10.5665/sleep.4998. [DOI] [PMC free article] [PubMed] [Google Scholar] - 19.Jirge PR. Poor ovarian reserve. Journal of human reproductive sciences. 2016;9:63–69. doi: 10.4103/0974-1208.183514. [DOI] [PMC free article] [PubMed] [Google Scholar] - 20.Stocker LJ, Macklon NS, Cheong YC, Bewley SJJO. Influence of shift work on early reproductive outcomes: a systematic review and meta-analysis. Gynecology. 2014;124:99–110. doi: 10.1097/AOG.0000000000000321. [DOI] [PubMed] [Google Scholar] - 21.NSF. National Sleep Foundation recommends new sleep times. In, 2015. - 22.Amyx M, Xiong X, Xie Y, Buekens P. Racial/ethnic differences in sleep disorders and reporting of trouble sleeping among women of childbearing age in the United States. Matern Child Health J. 2017;21:306–314. doi: 10.1007/s10995-016-2115-9. [DOI] [PMC free article] [PubMed] [Google Scholar] - 23.Touzet S, Rabilloud M, Boehringer H, Barranco E, RJF E. Relationship between sleep and secretion of gonadotropin and ovarian hormones in women with normal cycles. Sterility. 2002;77:738–744. doi: 10.1016/s0015-0282(01)03254-x. [DOI] [PubMed] [Google Scholar] - 24.Michels KA, Mendola P, Schliep KC, Yeung EH, Ye A, Dunietz GL et al. The influences of sleep duration, chronotype, and nightwork on the ovarian cycle 2020;37:260-71. [DOI] [PMC free article] [PubMed] - 25.Willis SK, Hatch EE, Wesselink AK, Rothman KJ, Mikkelsen EM, Wise LA. Female sleep patterns, shift work, and fecundability in a North American preconception cohort study. Fertil Steril. 2019;111:1201–10. e1. doi: 10.1016/j.fertnstert.2019.01.037. [DOI] [PMC free article] [PubMed] [Google Scholar] - 26.Grajewski B, Whelan EA, Lawson CC, Hein MJ, Waters MA, Anderson JL, MacDonald LA, Mertens CJ, Tseng CY, Cassinelli RT, II, Luo L. Miscarriage among flight attendants. Epidemiology. 2015;26:192–203. doi: 10.1097/EDE.0000000000000225. [DOI] [PMC free article] [PubMed] [Google Scholar] - 27.Fernandez RC, Marino JL, Varcoe TJ, Davis S, Moran LJ, Rumbold AR, Brown HM, Whitrow MJ, Davies MJ, Moore VM. Fixed or rotating night shift work undertaken by women: implications for fertility and miscarriage. Semin Reprod Med. 2016;34:74–82. doi: 10.1055/s-0036-1571354. [DOI] [PubMed] [Google Scholar] - 28.Begtrup LM, Specht IO, Hammer PEC, Flachs EM, Garde AH, Hansen J, Hansen ÅM, Kolstad HA, Larsen AD, Bonde JP. Night work and miscarriage: a Danish nationwide register-based cohort study. Occup Environ Med. 2019;76:302–308. doi: 10.1136/oemed-2018-105592. [DOI] [PubMed] [Google Scholar] - 29.Labyak S, Lava S, Turek F, Zee P. Effects of shiftwork on sleep and menstrual function in nurses. Health care for women international. 2002;23:703–714. doi: 10.1080/07399330290107449. [DOI] [PubMed] [Google Scholar] - 30.Bisanti L, Olsen J, Basso O, Thonneau P, Karmaus W. Shift work and subfecundity: a European multicenter study. European Study Group on Infertility and Subfecundity. J Occup Environ Med. 1996;38:352–358. doi: 10.1097/00043764-199604000-00012. [DOI] [PubMed] [Google Scholar] - 31.Aspholm R, Lindbohm ML, Paakkulainen H, Taskinen H, Nurminen T, Tiitinen A. Spontaneous abortions among Finnish flight attendants. J Occup Environ Med. 1999;41:486–491. doi: 10.1097/00043764-199906000-00015. [DOI] [PubMed] [Google Scholar] - 32.Knutsson A. Health disorders of shift workers. Occupational medicine (Oxford, England) 2003;53:103-8. [DOI] [PubMed] - 33.Liu C, Politch JA, Cullerton E, Go K, Pang S, Kuohung W. Impact of daylight savings time on spontaneous pregnancy loss in in vitro fertilization patients. Chronobiol Int. 2017;34:571–577. doi: 10.1080/07420528.2017.1279173. [DOI] [PubMed] [Google Scholar] - 34.Kirshenbaum M, Ben-David A, Zilberberg E, Elkan-Miller T, Haas J, Orvieto R. Influence of seasonal variation on in vitro fertilization success. PLoS One. 2018;13:e0199210. doi: 10.1371/journal.pone.0199210. [DOI] [PMC free article] [PubMed] [Google Scholar] - 35.Mínguez-Alarcón L, Souter I, Williams PL, Ford JB, Hauser R, Chavarro JE et al. Occupational factors and markers of ovarian reserve and response among women at a fertility centre 2017;74:426-31. [DOI] [PMC free article] [PubMed] - 36.Sen A, Hoffmann HM. Role of core circadian clock genes in hormone release and target tissue sensitivity in the reproductive axis. Mol Cell Endocrinol. 2020;501:110655. doi: 10.1016/j.mce.2019.110655. [DOI] [PMC free article] [PubMed] [Google Scholar] - 37.Lateef OM, Akintubosun MO. Sleep and reproductive health. J Circadian Rhythms. 2020;18:1. doi: 10.5334/jcr.190. [DOI] [PMC free article] [PubMed] [Google Scholar] - 38.Smeenk JMJ, Verhaak CM, Vingerhoets AJJM, Sweep CGJ, Merkus JMWM, Willemsen SJ, van Minnen A, Straatman H, Braat DDM. Stress and outcome success in IVF: the role of self-reports and endocrine variables. Hum Reprod. 2005;20:991–996. doi: 10.1093/humrep/deh739. [DOI] [PubMed] [Google Scholar] - 39.du Plessis SS, Makker K, Desai NR, Agarwal A. Impact of oxidative stress on IVF. Expert Review of Obstetrics & Gynecology. 2008;3:539–554. [Google Scholar] - 40.Nikolettos N, Asimakopoulos B, Nicolettos N, Efthimiadou A, Mourvati E, Demirel C. Evaluation of leptin, interleukin-1β, tumor necrosis factor-α and vascular endothelial growth factor in serum and follicular fluids of women undergoing controlled ovarian hyperstimulation as prognostic markers of ICSI outcome. In vivo (Athens, Greece) 2004;18:667-74. [PubMed] - 41.Singh AK, Chattopadhyay R, Chakravarty B, Chaudhury K. Markers of oxidative stress in follicular fluid of women with endometriosis and tubal infertility undergoing IVF. Reprod Toxicol. 2013;42:116–124. doi: 10.1016/j.reprotox.2013.08.005. [DOI] [PubMed] [Google Scholar] - 42.Williams WP, 3rd, Kriegsfeld LJ. Circadian control of neuroendocrine circuits regulating female reproductive function. Front Endocrinol (Lausanne) 2012;3:60. doi: 10.3389/fendo.2012.00060. [DOI] [PMC free article] [PubMed] [Google Scholar] - 43.Cahill GM, Hurd MW, Batchelor MM. Circadian rhythmicity in the locomotor activity of larval zebrafish. Neuroreport. 1998;9:3445–3449. doi: 10.1097/00001756-199810260-00020. [DOI] [PubMed] [Google Scholar] - 44.Khattab AF, Mustafa FA, Taylor PJ. The use of urine LH detection kits to time intrauterine insemination with donor sperm. Hum Reprod. 2005;20:2542–2545. doi: 10.1093/humrep/dei098. [DOI] [PubMed] [Google Scholar] - 45.Rossmanith WG. Ultradian and circadian patterns in luteinizing hormone secretion during reproductive life in women. Hum Reprod. 1993;8(Suppl 2):77–83. doi: 10.1093/humrep/8.suppl_2.77. [DOI] [PubMed] [Google Scholar] - 46.Kennaway DJ, Boden MJ, Varcoe TJ. Circadian rhythms and fertility. Mol Cell Endocrinol. 2012;349:56–61. doi: 10.1016/j.mce.2011.08.013. [DOI] [PubMed] [Google Scholar] - 47.Sakai N, Endo A. Effects of delayed mating on preimplantation embryos in spontaneously ovulated mice. Gamete Res. 1988;19:381–385. doi: 10.1002/mrd.1120190409. [DOI] [PubMed] [Google Scholar] - 48.Endo A, Watanabe T. Effects of non-24-hour days on reproductive efficacy and embryonic development in mice. Gamete Res. 1989;22:435–441. doi: 10.1002/mrd.1120220409. [DOI] [PubMed] [Google Scholar] - 49.Summa KC, Vitaterna MH, Turek FW. Environmental perturbation of the circadian clock disrupts pregnancy in the mouse. PLoS One. 2012;7:e37668. doi: 10.1371/journal.pone.0037668. [DOI] [PMC free article] [PubMed] [Google Scholar] - 50.Boden MJ, Varcoe TJ, Voultsios A, Kennaway DJ. Reproductive biology of female Bmal1 null mice. Reproduction. 2010;139:1077–1090. doi: 10.1530/REP-09-0523. [DOI] [PubMed] [Google Scholar] - 51.Fischer D, Lombardi DA, Marucci-Wellman H, Roenneberg T. Chronotypes in the US – influence of age and sex. PLoS One. 2017;12:e0178782. doi: 10.1371/journal.pone.0178782. [DOI] [PMC free article] [PubMed] [Google Scholar]