Liu, James H. MD; Reape, Kathleen Z. MD; Hait, Howard I.
Sleep is essential to daily living, and poor sleep can affect the quality of life. Perimenopausal and postmenopausal women often experience sleep problems with prevalence rates varying greatly from 15% to upward of 60%.1,2 The direct cause and effect associations for sleep disruption in this patient population have not been determined, but declining estrogen levels and the occurrence of vasomotor symptoms may be important contributing factors.3 Although subjective studies have found a relationship between sleep disturbances and hot flushes in women, there have been few studies using objective sleep assessment techniques to evaluate this relationship.4 Although not definitive, a number of studies suggest that estrogen use in symptomatic women experiencing sleep disturbances may decrease the frequency and severity of nocturnal vasomotor symptoms with a subsequent improvement in sleep parameters.5–7
In a previous pilot study, 19 postmenopausal women reporting sleep disturbance resulting from nocturnal diaphoresis and hot flushes received either 0.625 mg synthetic conjugated estrogens-A or matching placebo daily for up to 6 weeks to determine the effects on sleep using both subjective and objective (polysomnographic) measures. Significant mean reductions in galvanic skin responses and in the average number of hot flushes that interrupted usual sleep were observed for women receiving the active treatment.8
Synthetic conjugated estrogens-B differs from synthetic conjugated estrogens-A in that it contains 10 rather than nine estrogenic components, including the delta-8,9-dehydroestrone sulfate moiety. This compound has been shown to have unique estrogenic properties and high potency that may contribute to the overall biologic effects of conjugated estrogens, and when administered alone, it has been shown to significantly reduce vasomotor symptoms.9–11 Synthetic conjugated estrogens-B is approved for the treatment of moderate-to-severe vasomotor symptoms and vulvovaginal atrophy in postmenopausal women at doses ranging from 0.3 mg/day to 1.25 mg/day (Enjuvia prescribing information; Pomona, NY, Duramed Pharmaceuticals, Inc, a subsidiary of Barr Pharmaceuticals, Inc; revised April 2007).12,13 In this study, two lower doses of synthetic conjugated estrogens-B were compared with placebo to evaluate the change from baseline in the number of awakenings associated with nocturnal vasomotor symptoms. Secondary end points included assessments of changes in sleep quality and wakefulness measures as well as changes in objective sleep parameters (using actigraphy). The goal of this investigation was to estimate the magnitude and duration of treatment effects on vasomotor symptoms (including night sweats), specifically when nighttime sleep is disrupted.
MATERIALS AND METHODS
This randomized, double-blind, placebo-controlled multicenter trial was conducted at 19 sites in the United States from January 2008 to February 2009 with the primary objective of evaluating the efficacy of two different doses of oral synthetic conjugated estrogens-B for the treatment of awakenings resulting from nocturnal vasomotor symptoms in highly symptomatic postmenopausal women. This study was conducted in accordance with the Declaration of Helsinki (Republic of South Africa, 1996) and applicable guidelines for Good Clinical Practice; the Copernicus Group institutional review board in North Carolina and the Case Medical Center institutional review board for human investigation in Ohio reviewed the protocol and granted approval at each participating site. All participants signed informed consent forms.
Healthy, naturally or surgically menopausal women aged 30–65 years, inclusive, experiencing a minimum of three awakenings per night resulting from nocturnal vasomotor symptoms were eligible for screening. Patients with a uterus had normal endometrial biopsies at baseline. Those with any contraindications to estrogens were excluded, as were those using antidepressants, benzodiazepines, hypnotics, or other prescription or over-the-counter sleep aids and those with a history of fluctuating shift work in the 6 months before screening or current diagnosis of restless leg syndrome or sleep apnea. During screening, the frequency of hot flash episodes and number of awakenings resulting from nocturnal vasomotor symptoms were collected using a daily paper diary to determine eligibility. To qualify, women must have reported an average of at least seven daily or 50 weekly moderate-to-severe hot flash episodes and an average of at least three nocturnal awakenings per night as a result of hot flushes over any 7 consecutive days of the 2-week qualification period. Patients who met the entry criteria were randomized sequentially in a 1:1:1 ratio in blocks of six and assigned to one of three treatment groups: 0.3 mg daily (n=53), 0.625 mg daily (n=52), or placebo daily (n=52). Because the two active tablets were slightly different in size, a double-dummy design was used to maintain patient and investigator blinding. Each woman took two tablets daily (active plus placebo for each dose group or two placebo tablets for those randomized to placebo), one sized to match the 0.3-mg dose and one sized to match the 0.625-mg dose. Tablets were taken daily for up to 84 days (12 weeks). Clinic visits were scheduled at weeks 2, 4, 8, and 12. All women with a uterus underwent a posttreatment endometrial biopsy and received a 14-day course of progesterone at the end of treatment (three 100-mg capsules of micronized progesterone daily or one 10-mg tablet of medroxyprogesterone acetate daily for those women allergic to peanuts).
Patients used a daily paper diary to record the time pills were taken and the number of nocturnal awakenings resulting from hot flushes. Sleep quality was assessed on a three-point scale (poor, good, excellent) in the diary. At baseline and at three times during the treatment period (approximately monthly during the week before each clinic visit), women completed a rating three times daily on the seven-point Stanford Sleepiness Scale14,15 to assess any potential effect on daytime sleepiness. An actigraphy device was used during the same timeframe to objectively measure periods of wakefulness compared with periods of sleep and sleep efficiency. Actigraphy is a method used primarily in clinical research to study sleep–wake patterns by assessing limb movements over extended periods of time. An electronic device that looks similar to a watch is worn on the wrist while sleeping at home, and the device records period of activity and inactivity (based on movement), which are then used to estimate various sleep parameters such as total sleep time, duration and number of sleep bouts, and percentage of time spent sleeping.
The primary efficacy end point was a comparison of the mean reduction in frequency of awakenings resulting from nocturnal vasomotor symptoms at the end of the 12-week treatment period for each of the two active treatment groups compared with placebo using an intent-to-treat analysis. The weekly average of awakenings resulting from nocturnal hot flushes was calculated for each patient. For the baseline measurement, an average of 2 weeks of data before the randomization date was used. Hence, there were up to 13 weekly measurements, including one baseline measurement, for each patient.
The required sample size was calculated based on the primary efficacy evaluation of the mean reduction in the frequency of awakenings resulting from nocturnal vasomotor symptoms for either active treatment group compared with placebo. Although it was required that all enrolled women report at least three awakenings resulting from nocturnal hot flushes among the seven minimum total reported for a given day and at least 50 reported (day and night, total) for a given week, a conservative assumption was that all enrollees would report exactly three awakenings resulting from nocturnal hot flushes at baseline. Assuming a mean reduction of no greater than 0.5 awakenings daily (3.5 awakenings weekly) resulting from nocturnal hot flushes from baseline to the end of treatment for women taking the placebo compared with a mean reduction of 2.0 awakenings daily (14 awakenings weekly) for women on active treatment, with a standard deviation of the difference of 2.0, 90% power, and a two-sided 5% level of significance, a minimum of 39 analyzable cases per treatment group was required based on a two-sample t test criterion.
Analysis of covariance was used to evaluate the treatment effects of each active dose and placebo from baseline (week 0) to the end of treatment (week 12) with treatment and site serving as main effects and baseline as the covariate. Because the goal of the study was to evaluate treatment effect regardless of dose, no comparison was planned between the two active estrogen doses. Therefore, no adjustments for multiple comparisons were considered. An initial analysis included evaluation of a treatment-by-site interaction, which was found not to be statistically significant (P=.19), so it was removed from the analysis of covariance model and the analysis was conducted using the main effects only. Sites that did not have a sufficient number of women in the analysis (defined as at least three women in each treatment group) were pooled together as a single site for the purpose of analysis. This approach ensures that all data are included in the analysis, thereby maximizing the power of statistical tests. It does, however, limit the ability to estimate the magnitude of observed treatment effects on an individual site basis. The last observation carried forward procedure was used to impute the end-of-treatment efficacy data if data were missing or if the patient had discontinued the study before the end-of-treatment visit.
Secondary efficacy end points included a comparison between each active treatment group and placebo of the proportions of women in each group who, at the end of treatment, reported no nocturnal hot flushes, changes from baseline in sleep quality and individual sleep parameters (actigraphy), and changes from baseline in overall daytime sleepiness as assessed by the Stanford Sleepiness Scale.
For sleep quality and overall daytime sleepiness, analysis of covariance was used to compare treatment effects of each active dose and placebo from baseline (week 0) to the end of treatment (week 12) using baseline as the covariate. Although the principal analysis time point for all evaluations was week 12, changes from baseline (week 0) to weeks 2, 4, and 8 were also assessed to estimate the time to onset of treatment effects. A Cochran-Mantel-Haenszel model was used to test overall treatment effects for the proportion of women with no nocturnal hot flushes. Each treatment was also compared with placebo using exact probability tests.
Daily actigraphy data collected over all days of participation in the study were first summarized into weekly means for each patient and restricted to women who had actigraphy data through day 84 of treatment. For each patient, the simple arithmetic difference was obtained as the day 84 recorded mean value minus the baseline value. An analysis of covariance model evaluating treatment effect and including the baseline measure as the covariate was used to test for differences between each active dose and placebo. For each pairwise comparison of the respective active dose group compared with placebo, P values were reported.
A total of 157 women were randomized and received at least one dose of study medication (Fig. 1); of those, 145 women (92.4%) completed the study. The intent-to-treat cohort (n=152: 0.3 mg, n=52; 0.625 mg, n=49; and placebo, n=51) served as the primary analysis group from which all inferences regarding efficacy were drawn and included all women who were randomized and provided analyzable nocturnal vasomotor symptom data at baseline and at least one postbaseline visit. The demographics and baseline characteristics of the safety subject cohort by treatment group are summarized in Table 1. Overall, the treatment groups were comparable for all parameters. The majority of women (66.9%) were white. Ages ranged from 33.0 to 65.7 years, with a mean of 54.1 years. Patient weight varied from 98.0 to 325.0 lb, with a mean of 167.7 lb; body mass index (calculated as weight (kg)/[height (m)]2) ranged from 18.6 to 51.0, with a mean of 28.5.
The primary efficacy end point was the change from baseline in the frequency of awakenings (weekly average) resulting from nocturnal vasomotor symptoms based on a daily diary over the 12-week treatment period. The 12 women who discontinued prematurely (two on 0.3 mg, four on 0.625 mg, six on placebo) had their last observed value (before week 12) included in this analysis, but this was not considered a source of confounding with the week 12 inference. Significant mean reductions in weekly average awakenings resulting from nocturnal hot flushes from baseline to week 12 or the end of treatment were observed for both synthetic conjugated estrogens-B dosage groups compared with placebo. The 0.3-mg dose was associated with a mean reduction of 3.55 awakenings per week (P=.004) greater than placebo with a slightly larger differential mean reduction seen for the 0.625-mg dose compared with placebo (4.65; P<.001).
The mean reduction in the number of weekly awakenings resulting from nocturnal hot flushes during the time course of the treatment period was also assessed at each of the study visits conducted at weeks 2, 4, 8, and 12 (Fig. 2). Compared with placebo, 0.625 mg showed significant reductions in the frequency of awakenings as early as week 4, whereas 0.3 mg was associated with a significant mean reduction first observed at week 8. These reductions were maintained for the duration of the 12-week study.
The secondary efficacy analyses evaluated the changes from baseline in the proportions of women who, at the end of treatment, reported no nocturnal hot flushes, changes in individual sleep parameters, including sleep quality, and changes in overall daytime sleepiness as assessed using the Stanford Sleepiness Scale.
A responder analysis demonstrated that both the 0.3-mg and 0.625-mg doses were associated with significantly greater proportions of women (more than one-third) reporting the complete elimination of awakenings compared with placebo (19 of 52 [36.54%] for 0.3 mg, 16 of 47 [34.04%] for 0.625 mg, five of 51 [9.8%] for placebo; P≤.002 for each dose compared separately with placebo). As was seen in the primary analysis of mean reduction in nocturnal awakenings, any additional effect owing to the higher 0.625-mg dose was minimal compared with the 0.3-mg dose.
Sleep quality was measured on a subjective three-point scale (3=excellent, 2=good, 1=poor) as recorded in a daily diary. The weekly average of sleep quality was calculated and converted to a daily base value (divided by 7). There was no overall difference observed in weekly average sleep quality scores between either active dose and placebo during the treatment period. Overall daytime sleepiness was evaluated from the Stanford Sleepiness Scale Ratings based on a seven-point scale (1=most alert to 7=sleepiest) measured three times daily and collected for approximately 7 days before weeks 4, 8, and 12. The weekly average Stanford Sleepiness Scale score was calculated and converted to a daily base (divided by 7). Similar to sleep quality, changes in daytime sleepiness were also not significantly different between either active dose and placebo.
Analysis of the actigraphy data demonstrated no distinctive patterns. There was a trend for both estrogen doses exhibiting modest increases compared with placebo in sleep efficiency (percentage of time spent asleep during a particular sleep interval), the average duration of bouts of sleep, and the total percentage of time sleeping. Both doses were also associated with demonstrable reductions in the number of sleep bouts. However, except for the increases in percentage time sleeping (P=0.19 for the 0.3-mg dose, P<.002 for the 0.625-mg dose), none of these results were significant in any consistent manner, although a dose-related increase in the magnitude of treatment effects was apparent for the reduction in number of sleep bouts, increase in the average duration of sleep, and increase in the percentage of time sleeping. No differences between the active treatment groups and placebo were observed for total sleep time or sleep onset latency (time to onset of sleep).
Treatment-emergent adverse events (occurred on or after the first dose of study medication through the date of study completion) were reported by 18 women on 0.3 mg (34%), 27 women on 0.625 mg (52%), and 19 women on placebo (37%); and a total of 26 women (16.56%) reported treatment-related adverse events (events considered by the investigator to be possibly, likely, or definitely related to the study medication). No greater than two women in any treatment group reported any single adverse event. One placebo patient reported a serious adverse event (tibia and fibula fracture), and three women (two on 0.625 mg, one on placebo) discontinued the study as a result of adverse events. Events leading to patient discontinuation in the active treatment group were bloating and mood swings and lower extremity edema; arm and leg pain led to the discontinuation of the patient in the placebo group. The most commonly reported treatment-emergent adverse events were nasopharyngitis, dizziness, and headache. No clinically significant abnormalities were observed on endometrial biopsies.
In this multicenter, double-blind, parallel-group, placebo-controlled study, we have evaluated the effects of two doses of synthetic conjugated estrogens-B (0.3 mg/day and 0.625 mg/day) compared with placebo on the frequency of awakenings resulting from nocturnal vasomotor symptoms, individual sleep parameters, and overall daytime sleepiness in symptomatic postmenopausal women. Both estrogen doses were shown to significantly reduce the frequency of awakenings resulting from nocturnal vasomotor symptoms over a 12-week treatment period compared with placebo. The 0.625-mg dose significantly and consistently reduced the weekly average number of awakenings compared with placebo as early as week 4, with the lower 0.3-mg dose showing a significant reduction compared with placebo beginning at week 8. The 0.625-mg dose reduced weekly average awakenings to only a slightly greater extent than the 0.3-mg dose. This effect was also observed in the responder analysis in which both estrogen groups had somewhat similar, although significantly greater, proportions of women reporting complete elimination of nocturnal hot flushes at the end of treatment compared with placebo. Taken together, these findings suggest that the 0.3-mg dose is adequate for the treatment of nocturnal awakenings resulting from vasomotor symptoms, although higher doses may be required for those women who fail to respond at the lower dose. Use of hormone therapy should be consistent with treatment goals, benefits, and risks for the individual woman.
Patients in the active treatment groups did not demonstrate significant improvement in sleep quality or overall daytime sleepiness (based on Stanford Sleepiness Scale ratings) compared with placebo. However, objective actigraphy measurements suggested improved sleep efficiency for the 0.3-mg group compared with placebo. Both estrogen dosage groups demonstrated a significant improvement in the percentage of sleep time. The inconsistency of results for these secondary analyses may be related to the difficulty associated with formulating formal statistical analyses based on a priori determination of expected treatment effects. That is, a definitive understanding of the sensitivity of these analyses is unknown because they were secondary, supportive assessments undertaken to determine the value of actigraphy and its role in such studies. On the other hand, these outcomes may be viewed as supporting the use of estrogens as a specific treatment for nocturnal hot flushes; clearly, there is no evidence to support the use of estrogens as a broader sleep aid for postmenopausal women. Where the use of synthetic conjugated estrogens-B may have a beneficial effect in reducing the frequency of nighttime awakenings resulting from hot flushes, it is that specific benefit that may lead to better sleep as seen by an increase in the percentage of time women were asleep and the increase in the average number of minutes for each bout of sleep. Use as a sleep aid in women who do not have nighttime awakenings resulting from hot flushes is not supported by the results of this study.
Epidemiological studies have reported increased sleep disturbance associated with menopause, although the etiology is not clear.1 Simplistically, it has been postulated that nocturnal vasomotor symptoms result in frequent awakenings during the night, leading to impaired sleep, poor sleep quality, and subsequent daytime fatigue. However, controlled clinical trials that have been conducted more recently using objective polysomnographic end points and excluding women with other conditions that may impair sleep (eg, sleep apnea, restless leg syndrome, depression) as well as epidemiologic studies did not demonstrate an association between nocturnal vasomotor symptoms and sleep disturbance.16,17 Sleep disturbance in postmenopausal women is likely multifactorial in origin, and women may experience nocturnal vasomotor symptoms that are severe enough to result in awakenings and disruption of sleep while not necessarily demonstrating impaired sleep quality or daytime sleepiness on standard scales.
This study was designed and powered to primarily assess the reduction in the frequency of awakenings resulting from nocturnal vasomotor symptoms rather than changes in sleep parameters; thus, no minimal requirements for impaired sleep quality or daytime sleepiness were required for inclusion. The baseline levels of sleep quality and daytime sleepiness using the Stanford Sleepiness Scale showed that in general, the women in this study were not suffering from extremely poor sleep quality or excessive daytime fatigue; however, they were specifically bothered by nocturnal hot flushes that were causing them to awaken. Subjective self-reports of lower sleep quality may be suggestive of other problems unrelated to vasomotor symptoms and may differ from objective determinants of poor sleep quality.18 Although there was no strong observed effect of estrogen treatment on these various subjective sleep parameters, the objective actigraphy assessments suggest that treatment with synthetic conjugated estrogens-B may potentially improve sleep through its direct effect on reducing the number of nocturnal awakenings resulting from vasomotor symptoms.
In a highly symptomatic postmenopausal population of women experiencing sleep disruption resulting from nocturnal vasomotor symptoms, both the 0.3-mg and 0.625-mg doses of synthetic conjugated estrogens-B demonstrated a statistically significant reduction in the average weekly frequency of nocturnal awakenings resulting from hot flushes with significant differences being observed as early as week 4 for the 0.625-mg group and week 8 for the 0.3-mg group. A significant proportion (greater than one-third) of women in the active treatment groups reported complete elimination of nocturnal hot flushes, compared with placebo, and objective actigraphy measurements suggested favorable effects on sleep parameters in women receiving active treatment. Subjective assessments of sleep quality and daytime sleepiness did not reveal any differences between the active and placebo groups. Treatment with synthetic conjugated estrogens-B appears to be highly effective in treating moderate-to-severe vasomotor symptoms, including those that result in nocturnal awakenings, and may improve sleep.
1. Dennerstein L, Dudley EC, Hopper JL, Guthrie JR, Burger HG. A prospective population-based study of menopausal symptoms. Obstet Gynecol 2000;96:351–8.
2. Simon JA, Reape KZ. Understanding the menopausal experiences of professional women. Menopause 2009;16:73–6.
3. Ohayon MM. Severe hot flashes are associated with chronic insomnia. Arch Intern Med 2006;166:1262–8.
4. Moe KE. Hot flashes and sleep in women. Sleep Med Rev 2004;8:487–97.
5. Freedman MA. Quality of life and menopause: the role of estrogen. J Womens Health (Larchmt) 2002;11:703–18.
6. Sarti CD, Chiantera A, Graziottin A, Ognisanti F, Sidoli C, Mincigrucci M, et al.. Hormone therapy and sleep quality in women around menopause. Menopause 2005;12:545–51.
7. Montplaisir J, Lorrain J, Denesle R, Petit D. Sleep in menopause: differential effects of two forms of hormone replacement therapy. Menopause 2001;8:10–6.
8. Scharf MD, Berkowitz DV, Reape KZ. Effects of synthetic conjugated estrogens A on sleep quality in postmenopausal women with nocturnal diaphoresis and/or hot flushes: a pilot study. Fertil Steril 2007;88:654–6.
9. Bhavnani BR, Cecutti A, Dey MS. Biologic effects of delta-8-estrone sulfate in postmenopausal women. J Soc Gynecol Investig 1998;5:156–60.
10. Baracat E, Haidar M, Lopez FJ, Pickar J, Dey M, Negro-Vilar A. Estrogen activity and novel tissue selectivity of delta8,9-dehydroestrone sulfate in postmenopausal women. J Clin Endocrinol Metab 1999;84:2020–7.
11. Bhavnani BR, Cecutti A, Gerulath A. Pharmacokinetics and pharmacodynamics of a novel estrogen delta8-estrone in postmenopausal women and men. J Steroid Biochem Mol Biol 1998;67:119–31.
12. Utian WH, Lederman SA, Williams BM, Vega RY, Koltun WD, Leonard TW. Relief of hot flushes with new plant-derived 10-component synthetic conjugated estrogens. Obstet Gynecol 2004;103:245–53.
13. Simon J, Reape K, Wininger S, Hait H. Randomized, multicenter, double-blind, placebo-controlled trial to evaluate the efficacy and safety of synthetic conjugated estrogens B for the treatment of vulvovaginal atrophy in healthy postmenopausal women. Fertil Steril 2008;90:1132–8.
14. Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC. Quantification of sleepiness: a new approach. Psychophysiology 1973;10:431–6.
16. Freedman RR, Roehrs TA. Lack of sleep disturbance from menopausal hot flashes. Fertil Steril 2004;82:138–44.
17. Young T, Rabago D, Zgierska A, Austin D, Laurel F. Objective and subjective sleep quality in premenopausal, perimenopausal, and postmenopausal women in the Wisconsin Sleep Cohort Study. Sleep 2003;26:667–72.
18. Regestein QR, Friebely J, Shifren JL, Scharf MB, Wiita B, Carver J, et al.. Self-reported sleep in postmenopausal women. Menopause 2004;11:198–207.