| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Departments of Obstetrics and Gynecology, Psychiatry, Center for Clinical Epidemiology and Biostatistics, and the Department of Medicine, University of Pennsylvania Medical Center, Philadelphia, Pennsylvania.
Address reprint requests to: Lori Hollander, BA, Department of Obstetrics and Gynecology, 2 Dulles/Mudd Suite, University of Pennsylvania Medical Center, 3400 Spruce Street, Philadelphia, PA 19104-4283; E-mail: loriholl{at}mail.med.upenn.edu.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: A cohort of 218 black and 218 white women aged 35–47 years at enrollment (aged 37–49 at final follow-up) with regular menstrual cycles was identified through random digit dialing for a longitudinal study of ovarian aging correlates. Data obtained at four assessment periods, including enrollment, over a 2-year interval were collected between days 1 and 6 (mean = 3.9) of the menstrual cycle. The primary outcome measure was subjects’ rating of sleep quality at each assessment period. Associations of sleep quality with hormone levels (estradiol, follicle-stimulating hormone, luteinizing hormone, testosterone, and dehydroepiandrosterone sulfate) and other clinical, behavioral, and demographic variables were examined in bivariable and multivariable analyses.
RESULTS: Approximately 17% of subjects reported poor sleep at each assessment period. Significant independent associations with poor sleep included greater incidence of hot flashes (odds ratio [OR] 1.52; 95% confidence interval [CI] 1.08, 2.12, P = .02), higher anxiety levels (OR 1.03; 95% CI 1.00, 1.06, P = .04), higher depression levels (OR 1.05; 95% CI 1.02, 1.07, P < .001), greater caffeine consumption (OR 1.25; 95% CI 1.04, 1.49, P = .02), and lower estradiol levels in women aged 45–49 (OR 0.53; 95% CI 0.34, 0.84, P = .006), after adjustment for current use of sleep medications.
CONCLUSION: Both hormonal and behavioral factors were associated with sleep quality. Estradiol levels are an important factor in poor sleep reported by women in the 45–49 age group. Further evaluation of estrogen treatment for poor sleep of women 45 years and older is warranted.
Poor sleep is one of the most common symptoms in the transition through menopause.1 Women in their 40s and 50s report poor sleep at a greater rate than women at other ages,2,3 and the report of poor sleep during this time increases at a significantly greater rate among women when compared with men.4,5 Although both age and gender differences suggest that the sex steroids might be involved in women’s poor sleep, the relationships of these variables with sleep problems are poorly understood.6,7
Problems sleeping upset many important aspects of healthy functioning4 but are rarely queried by clinicians and seldom diagnosed.5,8 Poor sleep may result in poor job performance, discord in relationships, mood swings, reduced memory and learning, impaired concentration, fatigue, irritability, and decreased general well-being.4,9 Poor sleep is estimated to be responsible for motor vehicle accidents resulting in more than 1000 fatalities, 45,000 serious injuries, and $1.75 billion in costs annually.10 Cardiovascular diseases are associated with insomnia, and insomnia is a predictor of myocardial infarction and coronary death.2
The present study examined sleep quality using longitudinal data from four assessment periods in a community-based cohort of women who had regular menstrual cycles at study enrollment, were aged 35–47 years at enrollment, and 37–49 at the final follow-up 2 years later. The aims of the present study were to estimate the prevalence of poor sleep in these generally healthy women, identify associations of sex steroid hormone levels and other previously identified predictors with poor sleep, and determine the independent contributions to sleep quality from among the hormonal, behavioral, and demographic background variables associated with the quality of sleep in this cohort.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Eligibility for enrollment in the cohort included black or white women, aged between 35 and 47 years (37–49 at last follow-up 2 years later), menstrual cycles in the normal range (22–35 days) for the previous 3 months, with the uterus and at least one ovary intact. Exclusion criteria included any serious illness that might compromise ovarian or hormonal function, such as diabetes, liver disease, breast or endometrial cancer; use of hormonal medication including hormonal contraception; use of psychotropic drugs; alcohol or drug abuse within the past year; and pregnancy, lactation, or intention of becoming pregnant. The Institutional Review Board of the University approved the study, and participants gave written informed consent.
A total of 1420 women in Philadelphia County (563 black, 633 white, and 224 race unknown) were identified as potentially eligible for the study. Of these, 402 refused further screening, and 1018 were fully screened to yield 580 eligible and 438 ineligible subjects. Seventy-five percent (436 of 580) of the eligible women participated, and 144 declined to participate. The stratified sampling for race yielded 218 subjects in each racial group.
This report included the full cohort at baseline (N = 436) and all available subjects at each of the next three assessment periods. During the 2-year interval, 56 subjects dropped from the study for the following reasons: three moved from the area, 16 were lost to follow-up, one died, six had family conflicts, four cited time constraints, two had medical reasons, three withdrew consent, and 21 did not provide an explanation. Another 48 subjects had sporadic missing data points resulting from missing hormone data or missing questionnaire information. In addition, during the course of the study two women reported a bilateral oophorectomy, and three subjects became pregnant; blood was not drawn during pregnancy or breastfeeding. Two subjects used hormonal contraception, and nine additional subjects reported other hormone use during the follow-up period; these subjects remained in the study, but we conducted further analyses excluding them at the times of their hormone use. There was no substantive change in the estradiol sleep association when these subjects were excluded from the analysis. The final report excluded these subjects.
After the first assessment period at enrollment, subjects participated in three follow-up assessment periods at approximately 8-month intervals over 2 years. Within each assessment period, there were two visits 1 month apart to obtain blood samples for hormone measurements, providing a maximum of eight blood samples per subject. All visits were scheduled within the first 6 days (mean = 3.9) of the menstrual cycle. At each visit a trained research interviewer obtained the blood sample, collected anthropometric measures, conducted a standardized interview, and participants completed self-administered standard questionnaires.
The interview was described as a general health assessment and consisted of questions concerning sociodemographic factors, reproductive history, medical and gynecologic history, and menstrual cycle characteristics. At each interview we asked the women to report all medications, including over-the-counter and herbal preparations. The following variables from the extensive interview were included in this report. Caffeine and alcohol use was queried at alternate assessment periods by asking women about their typical consumption over the past year. We asked women whether they currently smoked cigarettes or had smoked in the past, the average number of cigarettes smoked per day, and the duration of smoking. In addition, we asked women whether they experienced menopausal hot flashes and the frequency and severity of hot flashes. Using standard procedures, we measured body mass index (BMI) and waist/hip ratio. We queried cycle length in the interview and confirmed these data by having subjects maintain daily symptom ratings for one menstrual cycle at each assessment period.
Sleep quality was assessed using the St. Mary’s Hospital Sleep Questionnaire,11 a validated self-report questionnaire that has been demonstrated to describe insomnia both quantitatively and qualitatively.12 The 12 items addressed the previous night’s sleep rather than retrospective sleep habits generally, which was important for evaluating the associations of sleep with concurrent levels of hormones and symptoms.13 Six of the 12 items in the questionnaire were quantitative questions about sleep, and the remaining six items were qualitative items that assessed how well the subject slept, how clear-headed she felt in the morning, how satisfied she was with her sleep, if she was troubled by early morning awakenings, and how last night’s sleep compared with usual sleep. Because previous studies established that sleep quality was better than sleep quantity for assessing general health and feelings of depression, fatigue, tension, confusion, stress, and sleep medications,14 we selected the item, "How well did you sleep last night?" which was rated on a 6-point scale ranging from 1 (very badly) to 6 (very well) as the primary outcome variable. After inspection of the response distribution, a dichotomous variable was established for "slept badly" (responses 1–3) or "slept well" (responses 4–6). We also examined the comparison of last night’s sleep with usual sleep and found that most subjects reported last night’s sleep as representative (69% representative, 14% better, 17% worse). We believe that the subject’s assessment of the quality of sleep incorporates the other qualitative aspects that were queried.
Information on sleep medications during the previous month was evaluated as part of the St. Mary’s Hospital questionnaire. The reported use of sleep medications was used as a control variable in the data analysis. Nineteen percent of subjects reported use of sleep medications, and these subjects reported poor sleep. We examined the analytic models with and without the inclusion of this information and found no difference in risk factor estimates.
Other standard self-report questionnaires included the Center for Epidemiological Studies’ Depression Scale to assess depressive symptoms,15 the Zung Anxiety Scale,16 and the Cohen Perceived Stress Scale.17 The Center for Epidemiological Studies’ Depression Scale is a well-established self-report scale that was developed for epidemiologic research and used in numerous studies including reproductive aging. Subjects rated 20 items that relate to depressed mood from 0 (rarely or none of the time) to 4 (most of the time). Positive item scores were reversed and the ratings were summed for a total score, with scores of at least 16 generally classified as indicating depression. The Zung Anxiety Scale is a validated self-report measure that is sensitive to the frequency of affective and somatic anxiety symptoms17. Each of the 20 items is a well-recognized symptom of anxiety. Subjects rated each item from 1 (none or a little of the time) to 4 (most or all of the time), and the ratings were summed for a total anxiety score. The Perceived Stress Scale is a 14-item validated self-report measure of the degree to which situations are perceived as stressful. Subjects rated each item from 0 (never) to 4 (very often). Total scores were obtained by reverse-scoring the seven positive items and summing all ratings. The Perceived Stress Scale has been correlated with depression, physical symptomatology, and anxiety and has been shown to measure a different and independently predictive construct of appraised stress.
We collected all blood samples during the early follicular phase between days 1 and 6 of the menstrual cycle or 1 month apart in nonmenstruating women. Blood samples were centrifuged and plasma frozen in aliquots at -70C. Assays were conducted in the General Clinical Research Center at the Hospital of the University of Pennsylvania in batches that included four visits per subject to reduce the within-subject variability resulting from assay conditions. Estradiol (E2), follicle-stimulating hormone (FSH), luteinizing hormone (LH), dehydroepiandrosterone sulfate (DHEAS), and testosterone were measured by radioimmunoassay using Coat-A-Count commercial kits (Diagnostic Products, Los Angeles, CA). Assays were performed in duplicate for all hormones and repeated if values differed by more than 15%. The inter- and intra-assay coefficients of variation were calculated from the assays conducted for the present study as follows: E2: 2.7%, 0.8%; FSH: 1.2%, 1.0%; LH: 2.5%, 1.0%; testosterone: 5.3%, 4.6%; DHEAS: 2.7%, 0.8%.
To minimize the variability and propensity to skewness inherent in hormones values, the natural logarithms of plasma hormone values and the average value of two natural log transformed values for a hormone taken 1 month apart were used in the analysis. Decisions to use categorical variables for age, caffeine use, and BMI were based on the evidence of nonlinear relationships of these variables with the main outcome variable and on the statistical evaluations of the data in the model as described below. The age categories are those used in population studies and census data divisions, that is, 35–39, 40–44, and 45–49 years.
At each assessment period, women who reported poor sleep were compared with the remaining women at that time period using means and Student t test for continuous variables and frequencies with the 
2 test of association for categorical variables.
To evaluate the variables over the 2-year period of this study, we used a multivariable logistic regression for repeated measures to estimate effects of the covariates on subjects’ rating of their sleep quality the previous night (good or poor). Variance estimates for the Wald statistics of the true logistic coefficients were adjusted for the repeated observations from each participant using Generalized Estimating Equations.18 We included all available data for each subject in the model, which retained all observations and did not drop subjects with missing data. The initial model included all potential predictors with values less than .20 in preliminary bivariate analyses.
To evaluate the relationship for quantitative risk factors such as age, caffeine use, and BMI, the variables were categorized and entered as a set of indicator (dummy) variables. Each variable was also modeled as continuous, assuming a linear relationship with poor sleep. To test this linearity assumption, we tested the significance of a quadratic term for each factor (eg, age squared). When the quadratic term was significant, the decision was made to model the risk factor using the set of dummy variables for ease of interpretation. The association of age with sleep was nonlinear, and age categories were used in the model. Based on hypotheses of associations of decreasing E2 levels with age, hot flashes, and depression,19 we tested the interactions of E2 with these variables.
Because caffeine consumption was assessed at only the second and fourth time periods, we used an imputation in which we used the subject value from caffeine consumption for the previous period. We constructed models with and without the caffeine variable and with and without assessment periods 1 and 3 to test the sensitivity of imputing caffeine. The results of these sensitivity analyses showed that the imputation of caffeine had no substantive effects on the parameters in the model. We included a variable for time in the model and tested the time by covariate interactions to determine if the associations between the given variable and sleep changed over time. Neither the time variable nor the interactions were significant, and we excluded the time variable from the final model.
We developed the final model using backward selection from the set of potential risk factors with the final selection of covariates based on whether the variable remained statistically significant at P 
.05, and whether the inclusion of the variable modified other significant associations in the model by 15% or more. In addition, we evaluated the subgroup of subjects who reported changes in their sleep quality over time (n = 170) using conditional logistic regression.20 We performed all analyses using SAS 6.12 (SAS Institute Inc., Cary, NC), with two-tailed interpretation of P values.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Subjects reported sleep status at 8-month intervals. Seventeen percent of the sample (72 of 436) reported poor sleep at baseline. Although individuals who reported poor sleep changed over the 2-year study, the overall proportion of subjects who reported poor sleep remained similar at each assessment period: 17% (68 of 392) at 8 months; 16% (57 of 359) at 16 months; and 15% (49 of 332) at 24 months. Among the 364 subjects (83%) who reported good sleep at baseline, 207 (48% of the total sample) reported good sleep throughout the observation period, and another 76 subjects (17%) reported good sleep at all observed visits with one or more visits missing information; 60 (14%) subjects reported poor sleep at one visit, 19 (3%) reported poor sleep at two visits, and only two subjects reported poor sleep at three visits. Among the 72 (17%) subjects who reported poor sleep at baseline, only five (1%) subjects reported poor sleep at all four visits; nine (2%) reported poor sleep at three visits, 28 (6%) reported poor sleep at two visits, and 30 (7%) reported poor sleep at only one visit.
The bivariable associations of hormone values, behavioral variables, and background variables with poor sleep over the 2-year study period are presented in Table 1
. In these analyses, poor sleep was associated significantly with race (blacks were more likely to report poor sleep as compared with whites), increased hot flashes, high BMI, high caffeine consumption, increased perceived stress, increased depression, increased anxiety, less employment, less education, and the use of sleep medication.
|
). Adjusting for sleep medications had no effect on the other variables in the model. It is also noteworthy that in multiple sensitivity tests for the caffeine variable (described above), the addition of caffeine to the model did not change the associations of the other variables; that is, the final model retained the same variables with no change in odds ratios with and without inclusion of the caffeine variable.
|
. Other interactions were tested in the model but were not statistically significant, notably the interaction of E2 and depressive symptoms (P = .84), and the interaction of E2 and hot flashes (P = .34). Because hormone levels may influence the occurrence of hot flashes, we considered models with and without hot flashes. The association between E2 and poor sleep was slightly stronger (P = .049) in a model that did not adjust for the prevalence of hot flashes.
We examined the within-subject changes in sleep quality during the study. Thirty-nine percent of subjects (n = 170) reported changes in their sleep patterns over the four assessment periods. The associations among the risk factors shown in Table 2 remained consistent in this subgroup when we considered only the within-subject comparisons using conditional logistic regression.21 Decreases in a subject’s E2 level (P = .05) and/or increases in the Zung anxiety score (P = .007) were significantly associated with poor sleep over time for these subjects. The associations were present in age group–specific analyses as well, with the oldest age group exhibiting the strongest decreases in E2 levels over time.
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As reported in numerous studies,11,22,23 there was a significant association between women’s reports of hot flashes and poor sleep, regardless of whether the hot flashes occurred during the actual sleep cycle. The association was maintained independent of all other factors in the multivariable model including E2 levels, depressive symptoms, anxiety, and caffeine consumption.24 There was no significant interaction of hot flashes and E2. It is again noteworthy that hot flashes were experienced in the late 30s and 40s—before observable irregularity in menstrual cycles—and were associated with poor sleep earlier than is commonly assumed.25 Both hot flashes and poor sleep were also reported in a cross-sectional survey conducted in the Study of Women’s Health Across the Nation.26 Gold et al showed that among the women defined as "early perimenopausal," 37% reported "hot flashes/night sweats," and 41% reported "sleep difficulties." Although the studies cannot be compared directly because of differences in design, methodology, and definitions, both studies provide evidence that women experience these common menopausal symptoms in their late 30s and 40s, before the accepted definitions of the menopausal transition.
As hypothesized, depressive symptoms had a strong association with poor sleep,4,27,28 independent of the other variables in the multivariable model, including hot flashes. The depressive symptoms were reported across the age range of the sample and were not associated with E2 levels. The association of depressive symptoms with poor sleep remained after adjusting for age, hot flashes, E2, caffeine consumption, and anxiety in the final model.
This study is important because of its longitudinal assessment of hormones associated with ovarian aging together with mood and behavioral variables in women in their late 30s and 40s. The report is based on the subject’s perceptions of poor sleep, which were assessed longitudinally using a standard and validated sleep questionnaire. Subjective bias in the assessment of sleep was minimized inasmuch as the study was described as a general health study, and sleep was only one of many variables assessed. However, physiologic measures obtained in sleep laboratory studies would be important to confirm and extend the present findings. Other limitations of the study include those of unassessed variables that were beyond the focus of the project, such as sleep-related disordered breathing.5 We believe that this cohort is a representative sample of urban women, given that the cohort was identified randomly by population-based recruitment, but the results may not be generalized to women in nonurban areas or racial groups other than those of this study. Both the response rates and the follow-up rates were sufficient for the conclusions of this study.
The results indicate that reports of poor sleep from women in their late 30s and 40s were stable over a 2-year period, and that both hormonal and behavioral factors played a role in the oldest age group. Estradiol levels may be more important than previously thought for women aged 45–49 who report poor sleep. In this circumstance, it is possible that hormone therapy may be helpful earlier in the menopausal transition than is currently the practice. Future research might include a randomized, controlled trial to evaluate the effect of estrogen treatment on poor sleep in women aged 45 and older.
|
Footnotes |
|---|
The authors thank Shiv C. Kapoor, PhD, for the hormone assays; and Beatriz Garcia-Espagna, MA, and Paula Martin for their assistance with computer analysis.
Received November 13, 2000. Received in revised form June 11, 2001. Accepted June 15, 2001.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Owens JF, Matthews K. Sleep disturbance in healthy middle-aged women. Maturitas 1998;30:41–50.[Medline]
3. Baker A, Simpson S, Dawson D. Sleep disruption and mood changes associated with menopause. J Psychosom Res 1997;43:359–69.[Medline]
4. Clark AJ, Flowers J, Boots L, Shettar S. Sleep disturbance in mid-life women. J Adv Nurs 1995;22:562–8.[Medline]
5. Shaver JLF, Zenk SN. Sleep disturbance in menopause. J Womens Health Gend Based Med 2000;9:109–18.[Medline]
6. Mamber R, Armitage R. Sex, steriods, and sleep: A review. Sleep 1999;22:540–55.[Medline]
7. Empson JA, Purdie DW. Effects of sex steroids on sleep. Ann Intern Med 1999;31:141–5.
8. Everitt DE, Avron J. Clinical decision making in the evaluation and treatment of insomnia. Am J Med 1990;89: 357–62.[Medline]
9. Peterson DD, Schmidt RM. Longitudinal and cross-sectional analysis of HEALTH WATCH data with a subset of perimenopausal women and matched controls. J Gerontol A Biol Sci Med Sci 1999;54:B160–70.[Abstract]
10. Webb W. The cost of sleep-related accidents: A reanalysis. Sleep 1995;18:276–80.[Medline]
11. Ellis BW, Johns MW, Lancaster R, Raptopoulos P, Angelopoulos N, Priest RG. The St. Mary’s Hospital sleep questionnaire: A study of reliability. Sleep 1981;4:93–7.[Medline]
12. Patel AG, Gurer R, Kurian T, Lambert MT, Steinert J, Priest RG. A comparison of the hypnotic efects of temazepam elixir. Int Clin Psychoparm 1991;6:1–9.
13. Carskadon MA, Dement WC, Mitler MM, Guilleminault C, Zarcone VP, Spiegel R. Self-reports vs. sleep laboratory findings in 122 drug-free subjects with complaints of insomnia. Am J Psychiatry 1976;133:1382–8.
14. Pilcher JJ, Ginter DR, Sadowsky B. Sleep quality versus sleep quantity: Relationships between sleep and measures of health, well-being and sleepiness in college students. J Psychosom Res 1997;42:583–96.[Medline]
15. Radloff LS. The CES-D Scale: A new self-report depression scale for research in the general population. Applied Psychology Measurement 1977;1:385–401.
16. Zung WWK. A rating instrument for anxiety disorders. Psychosomatics 1971;12:371–9.
17. Cohen S, Kamarck TW, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96.[Medline]
18. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42: 121–30.[Medline]
19. Polo-Kantola P, Erkkola R, Helenius H, Irjala K, Polo O. When does estrogen replacement therapy improve sleep quality? Am J Obstet Gynecol 1998;178:1002–9.[Medline]
20. Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol 1989;129:125–37.
21. Breslow NE, Day NE. Statistical methods for cancer research, Volume 1. Lyon, France: International Agency for Research on Cancer, 1980.
22. Shaver J, Giblin E, Lentz M, Lee K. Sleep patterns and stability in perimenopausal women. Sleep 1988;11: 556–61.[Medline]
23. Erlik Y, Tataryn IV, Meldrum DR, Lomax P, Bajorek JG, Judd HL. Association of waking episodes with menopausal hot flushes. JAMA 1981;245:1741–4.[Abstract]
24. Freedman RR, Woodward S. Core body temperature during menopausal hot flushes. Fertil Steril 1996;65: 1141–4.[Medline]
25. Freeman E, Sammel M, Grisso JA, Kapoor S, Garcia-Espagna B, Hollander L. Risk factors for hot flashes: African American and Caucasian women in the late reproductive years. J Womens Health Gend Based Med 2000: 10;69–78.
26. Gold E, Sternfeld B, Kelsey J, Brown C, Moulton C, Reame N, et al. Relation of demographic and lifestyle factors to symptoms in a multiracial/ethnic population of women 40-55 years of age. Am J Epidemiol 2000;152: 463–73.
27. Bhatia SC, Bhatia SK. Depression in women: Diagnostic and treatment considerations. Am Fam Physician 1999;6: 225–34, 239–40.
28. Leger D, Scheuermaier K, Phillip P, Paillard M, Guilleminault C. Evaluation of quality of life in severe and mild insomniacs compared with good sleepers. Psychosom Med 2001;63:49–55.
This article has been cited by other articles:
|
|
 |
|
  M. R. Taylor, C. Booth-LaForce, E. Elven, B. B. McGrath, and R. C. Thurston Participant Perspectives on a Yoga Intervention for Menopausal Symptoms Complementary Health Practice Review, October 1, 2008; 13(3): 171 - 181. [Abstract] [PDF]
|
|
|
|
 |
|
 C. R. Gracia, E. W. Freeman, M. D. Sammel, H. Lin, and M. Mogul Hormones and Sexuality During Transition to Menopause Obstet. Gynecol., April 1, 2007; 109(4): 831 - 840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Roenneberg and M. Merrow Entrainment of the Human Circadian Clock Cold Spring Harb Symp Quant Biol, January 1, 2007; 72(0): 293 - 299. [Abstract] [PDF]
|
|
|
|
|
|
K. Visvanathan, L. Gallicchio, C. Schilling, J. K. Babus, L. M. Lewis, S. R. Miller, H. Zacur, and J. A. Flaws Cytochrome Gene Polymorphisms, Serum Estrogens, and Hot Flushes in Midlife Women Obstet. Gynecol., December 1, 2005; 106(6): 1372 - 1381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. Kravitz, I. Janssen, N. Santoro, J. T. Bromberger, M. Schocken, S. A. Everson-Rose, K. Karavolos, and L. H. Powell Relationship of Day-to-day Reproductive Hormone Levels to Sleep in Midlife Women Arch Intern Med, November 14, 2005; 165(20): 2370 - 2376. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. SCHILLING, L. GALLICCHIO, S. R. MILLER, J. K. BABUS, L. M. LEWIS, H. ZACUR, and J. A. FLAWS CURRENT ALCOHOL USE IS ASSOCIATED WITH A REDUCED RISK OF HOT FLASHES IN MIDLIFE WOMEN Alcohol Alcohol., November 1, 2005; 40(6): 563 - 568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. W. Freeman, M. D. Sammel, L. Liu, C. R. Gracia, D. B. Nelson, and L. Hollander Hormones and Menopausal Status as Predictors of Depression in Women in Transition to Menopause Arch Gen Psychiatry, January 1, 2004; 61(1): 62 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Poor Sleep Occurs During Perimenopause, Even When Cycles are Normal Journal Watch Psychiatry, January 24, 2002; 2002(124): 14 - 14. [Full Text]
|
|
|
|
 |
|
 Poor Sleep Occurs During Perimenopause, Even When Cycles Are Normal Journal Watch Women's Health, November 20, 2001; 2001(1120): 5 - 5. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
