Freeman, Ellen W. PhD; Sammel, Mary D. ScD; Lin, Hui MS; Liu, Ziyue PhD; Gracia, Clarisa R. MD
Relief for hot flushes, a predominant symptom of menopause,1 is among the most common reasons for clinical visits of midlife women and a major cost in healthcare expenditures.2–4 Hot flushes are associated with poor sleep,5,6 depressed mood,7,8 decreased quality of life,9,10 may worsen depressive symptoms, and signal the onset or relapse of a major depressive episode.11,12 Hot flushes may possibly mark underlying vascular changes that are associated with subclinical cardiovascular disease,13,14 increased aortic calcification among users of hormone therapy,15 greater incident coronary heart disease,16 and may be a risk factor for poor bone health.17
The peak prevalence of hot flushes occurs approximately 1 year after menopause,18 but the overall duration of hot flushes is unclear. Clinical guidelines indicate that the duration of hot flushes for most women is approximately 6 months to 2 years.19 However, the average duration of hot flushes was more than 5 years in the Melbourne Women's Midlife Health Project,20 and the median duration was 4 years in a meta-analysis of 10 previous hot flush studies18 with 10% of the women reporting hot flushes 12 years after the final menstrual period.
Increased information about the duration of hot flushes is important for the clinical management of menopausal symptoms. Clinicians need to make evidence-based decisions about the duration of symptoms to counsel patients more effectively about the risk-to-benefit ratios of treatments. This is particularly important for treatment with hormone therapy inasmuch as informed decisions about the duration of therapy are not possible without accurate estimates of the duration of the symptoms and knowledge of cofactors that influence symptom duration.
The purpose of this study was to estimate the duration of moderate-to-severe hot flushes in a population-based cohort of late reproductive-aged women and identify risk factors associated with hot flush duration. We hypothesized that menopausal stage at the onset of hot flushes predicted the duration of hot flushes. A secondary analysis was conducted to include all women who reported hot flushes, including mild hot flushes.
MATERIALS AND METHODS
The cohort of 436 women was identified by random-digit dialing, stratified to obtain equal numbers of African American and white women (218 in each group), and monitored from 1995 to 2009. At enrollment, the participants were ages 35–47 years, premenopausal with regular menstrual cycles in the normal range (22–35 days) for the previous three cycles, had an intact uterus, and had at least one ovary. Exclusion criteria included current use of hormonal contraception, hormonal therapies, or psychotropic medications; pregnancy or breast feeding; serious health problems known to compromise ovarian function such as diabetes mellitus or liver disease; breast or endometrial cancer; and alcohol or drug abuse in the past year.
The present study evaluated 259 women in the cohort who did not report hot flushes in interview at baseline but did experience moderate-to-severe hot flushes between assessment 2 and assessment 14. An additional 90 participants reported only mild hot flushes in this interval and were included in the secondary analysis that evaluated a total of 349 women. Fifty-five women reported no hot flushes in the entire study interval and were not included in the analyses of hot flush duration (Table 1). The remaining 32 women in the cohort had data only in assessment 1 (at study enrollment) and were not included in this study.
The cohort was monitored for 13 years. The start point at assessment 2 was the first follow-up assessment, conducted 9 months after cohort enrollment, and was the first assessment with ratings of the severity of hot flushes. Subsequent assessments were conducted at approximately 9-month intervals for the first 5 years of the study and then annually for the next 9 years with a 2-year gap between assessments 10 and 11. At each assessment, there were two in-home visits to collect study data and blood samples for hormone assays. All visits were timed to the early follicular phase (days 1–6 of the menstrual cycle) and were conducted in two consecutive menstrual cycles or 1 month apart in noncycling women (yielding a possible maximum of 26 hormone samples per participant in this study). The study was approved by the institutional review board of the University of Pennsylvania and all participants provided written informed consent.
At each assessment, trained research interviewers obtained structured interview and questionnaire data, blood samples for the hormone assays, and anthropometric measures (height, weight, waist and hip circumference). The study was described to participants as a general women's health study. The structured interview questionnaire focused on overall health, and participants also completed a set of validated self-report measures to assess health and other behavioral variables of the study.
The primary outcome variable was the duration of moderate-to-severe hot flushes. The duration of hot flushes was calculated for each woman from her first report of moderate or severe hot flushes to her last observed assessment of moderate or severe hot flushes. The participants were queried about hot flushes at each assessment using a validated menopausal symptom list that was embedded in the structured interview questionnaire.21 The questions asked whether hot flushes or night sweats occurred in the past month, whether hot flushes occurred in the past year (asked at assessments 12–14), and the severity of reported hot flushes (severity ratings were 0, none; 1, mild; 2, moderate; 3, severe). Moderate or severe hot flushes were defined as those rated 2 or 3 by the participants. The cessation of hot flushes was defined as no moderate or severe hot flushes for at least 1 year. Because the interview questions queried only the past 30 days at assessments 2–11, two consecutive assessments with no report of moderate or severe hot flushes were required for the cessation of hot flushes in this interval.
Menopausal stage was determined at each assessment using the menstrual date at the assessment (visits were conducted within the first 6 days of bleeding) and the two previous menstrual dates reported in the interview. Other confirmatory data were obtained from the daily symptom diaries that participants recorded for one menstrual cycle at each assessment and other interview data that included the number of menstrual periods between assessments, menstrual cycle length, and the number of bleeding days.
Stages of the menopause transition were adapted from the consensus statement of the Stages of Reproductive Aging Workshop22 to capture early changes in bleeding patterns. Significant associations between these stages and reproductive hormones changes were reported previously.23,24 The stages were as follows: 1) premenopausal: regular menstrual cycles in the 22- to 35-day range; 2) late premenopausal: a change 7 days or more in either direction in the participant's own cycle length observed at one study assessment; 3) early transition: changes in cycle length 7 days or more in either direction from the participant's baseline at enrollment and observed for at least two consecutive menstrual cycles in the study or 60 days amenorrhea; 4) late transition: 90 days to 11 months amenorrhea; and 5) postmenopause: 12 months or more amenorrhea excluding hysterectomy.
Hormones were measured in the Clinical and Translational Research Center of the University of Pennsylvania. Assays were conducted in batches that included four visits per participant to reduce the within-participant variability resulting from assay conditions. All assays were performed in duplicate and repeated if values differed by greater than 15%. Estradiol and follicle-stimulating hormone (FSH) were measured by radioimmunoassay using Coat-A-Count commercial kits (Siemens). Inter-assay and intra-assay coefficients of variation were less than 5%. Inhibin b was measured by enzyme-linked immunosorbent assay using the commercial kit from Diagnostic Systems Laboratories. The sensitivity of the assay was 7 pg/mL (range, 5–531 pg/mL). The intercoefficients and intracoefficients of variation were less than 5% and 7.3%, respectively. For the first 10 assessments, Patrick Sluss, PhD, Massachusetts General Hospital, Boston, performed the inhibin b assays with a solid-phase sandwich enzyme-linked immunosorbent assay based on the use of plates coated with a monoclonal antibody specific for the α subunit for detection.25,26 The limit of measurement for the assay was 15 pg/mL (coefficient of variation, 20%). The assay was controlled in triplicate using samples with mean concentrations of 155.3, 316.3, and 919.3 pg/mL with interassay coefficients of variation of 11.6%, 7.6%, and 9.7%, respectively.
Other covariates were selected for their significance in previous studies and the goals of this study and were obtained at the study visits: age, race (African American or white), currently employed (yes, no), education (high school or less, more than high school), body mass index (calculated as weight in kilograms divided by the square of height in meters), current smoking (yes, no), and alcohol use daily (yes, no). The measures of depressed mood (Center for Epidemiologic Studies–Depression Scale),27 anxiety,28 and perceived stress29 provided continuous scores for analysis with higher scores indicating more symptoms.
Statistical analyses were performed using Kaplan-Meier curves,30 log-rank tests,31 and Cox proportional hazards models32 to compare duration of hot flushes among menopausal stages and age groups. Proportionality of hazards was evaluated by evaluating plots of transformed hazard estimates and smoothed residuals33,34 and identified no violations of model assumptions.
The primary analysis included all women who reported moderate or severe hot flushes in the study interval. A secondary analysis included all women who reported any hot flushes in the same study interval. We used the value of each covariate at the time of the first report of moderate-to-severe hot flushes to emulate the clinical setting. The mean of four hormone measurements immediately preceding the onset of hot flushes for each participant was used and transformed to the natural log value to reduce the influence of large values. Mean hormone values are expressed in the report as the geometric mean (back-transformed from the natural log values) with 95% confidence intervals (CIs). Hot flush duration was censored at times of reported hormone use, pregnancy, and breast feeding; observations of hot flushes at the last available assessment (assessment 14 or the point of dropout if sooner) were also considered censored. Of the 259 women who comprised the primary study sample, 50 were considered censored from the point of discontinuation in the 13-year study interval (17 in years 2–5; 16 in years 6–9; 17 in years 10–13). Reasons for discontinuation were death (n=11), medical problems (n=7), withdrew consent (n=6), moved from the area (n=2), no reason given (n=8), and lost to follow-up (n=16).
Each covariate was evaluated individually; those meeting the significance level of P≤.15 were then included in multivariable models to identify their independent contributions after adjusting for the presence of all other variables. The final selection of covariates used a backward elimination strategy and was guided by whether the variable was associated with the response variable at P≤.05 or its inclusion in the model modified other significant associations by 15% or more.
A sensitivity analysis of the primary outcome (duration of moderate-to-severe hot flushes) was conducted omitting 24 women who indicated hot flushes in daily diaries that were collected at assessment 1. The results for the remaining 235 women who had no moderate-to-severe hot flushes at assessment 1 were nearly identical to the results for the 259 women that are presented in this report.
The study variables were compared among groups at baseline using χ2 or F tests as appropriate for the data. All analyses were conducted using the SAS 9.2 statistical package. Statistical tests were two-sided with P<.05 considered significant.
At the study baseline, the mean (standard error) age was 42.2 (0.18) years; 91% were premenopausal or late premenopausal and 9% were in the early transition stage. Age, menopausal status, race, hormone measures, body mass index, and alcohol use did not differ among the hot flush severity groups (Table 1). The moderate-to-severe hot flush group had significantly higher baseline scores on measures of depressed mood, anxiety, and stress; fewer were employed, education levels were lower, and more were smokers. Thirty-four percent of white participants and 49% of African Americans were smokers (χ2=5.58; df=1, P=.02). Smoking did not differ by menopausal stage at the onset of hot flushes. At hot flush onset, 44% in the premenopausal group were smokers compared with 46% in the early transition group and 35% in the late transition or postmenopause group (χ2=2.23; df=2, P=.33).
The median duration of moderate-to-severe hot flushes was 10.2 years. Thirty-seven percent of the women who reported moderate-to-severe hot flushes (96 of 259) reported hot flush cessation in the study interval (ie, no moderate-to-severe hot flushes for at least 1 year), whereas the remaining women continued to report moderate-to-severe hot flushes.
The duration of hot flushes was strongly associated with menopausal stage at hot flush onset (P<.001; Fig. 1). The median duration of moderate-to-severe hot flushes for women who reported onset in the premenopausal or late premenopausal stage was more than 11.25 years, and only 21% in this group reported cessation in the study interval (Table 2). Median duration when onset was in the early transition stage was 7.35 years with 51% reporting cessation in the study interval. When onset was in the late transition or postmenopause stage, the median duration was 3.84 years with 52% reporting cessation during the follow-up interval.
The most common age at onset of moderate-to-severe hot flushes was 45–49 years (35%); 30% were ages 40–44 years, 21% were older than age 50 years, and 14% were younger than 40 years. Age at onset was inversely associated with duration of hot flushes (P<.001) (Table 2). The median duration of hot flushes was longest when onset occurred at ages younger than 40 years (11.57 years) and decreased with onset at older ages: 11.25-year duration with onset at ages 40–44 years; 8.1-year duration with onset at ages 45–49 years; and 3.8-year duration with onset at 50 years of age or older.
Mean hormone levels were significantly associated with the duration of hot flushes in unadjusted analyses. Higher levels of FSH and lower levels of inhibin b and estradiol indicated a greater likelihood of hot flush cessation (P<.001, P<.001, and P=.08, respectively; Table 3). Higher FSH levels remained a significant predictor of cessation of hot flushes after adjusting for all other variables in the multivariable model (P=.003).
Menopausal stage at onset of moderate-to-severe hot flushes was the strongest predictor of duration after adjusting for the presence of all other variables in the model (P<.001) (Table 4). When the onset of hot flushes was in the premenopausal or late premenopause stages, the duration was longer compared with onset in the late transition or postmenopause stages (hazard ratio, 5.14; 95% CI 2.70–9.77; P<.001). We then investigated whether the proportion of women with severe hot flushes was smaller in the earlier stages of the transition by comparing the proportion of women with severe hot flushes at onset in each menopausal stage. The proportion of severe hot flushes at onset was similar at each menopausal stage with no significant association between hot flush severity and menopausal stage. At onset in the premenopausal to late premenopausal stage, 24% reported severe hot flushes; at onset in the early transition stage, 35% reported severe hot flushes; at onset in the late transition or postmenopausal stage: 20% reported severe hot flushes (chi square=4.29; df=2, P=.12).
Other independent predictors of the duration of moderate-to-severe hot flushes in adjusted analysis were age (the younger the age at onset, the longer the duration, P=.02); race (African American women had longer duration of hot flushes than white women, P=.02); and body mass index (nonobese women had longer duration of hot flushes than obese women, P=.003).
Figure 2 shows the influence of age, race, and body mass index on the estimated duration of moderate-to-severe hot flushes for the early-onset and late-onset groups. For example, African American, nonobese women who were early starters (hot flush onset in the premenopausal or late premenopausal stage) had an estimated duration of over 10 years regardless of age at onset (Fig. 2A). In contrast, white, nonobese women who were late starters (onset in late transition or postmenopausal stages) and age 45 years or older at onset had an estimated duration of approximately 4 years (Fig. 2B).
Of the remaining study variables, significant unadjusted associations with hot flush duration were observed for depressed mood (P=.005), anxiety (P=.006), stress (P=.04), and current employment (P=.008), but none was significantly associated with the duration of hot flushes in the multivariable model. Smoking, alcohol use, and number of children had no association with duration of hot flushes in either adjusted or unadjusted models.
Current smoking was not associated with the duration of hot flushes in unadjusted analysis (hazard ratio, 0.83; 95% CI 0.55–1.25; P=.36) but was examined in the final multivariable model to determine whether it was a confounder of the other risk factor associations with the duration of hot flushes. In the multivariable model, smoking remained nonsignificant (P=.41), and model results were nearly identical to those shown (Table 4).
The analyses were repeated to include all women who reported any hot flushes (mild, moderate, severe) in the study interval (n=349) from the first report of any hot flushes to cessation of any hot flushes. By including mild hot flushes in the analysis, the median duration increased to 11.6 years. The duration of any hot flushes was associated with menopausal stage (P<.001), age (P<.001), and anxiety (P=.006) in unadjusted analysis (data not shown). Mean hormone levels were associated with hot flush duration: FSH (P<.001), inhibin b (P<.001), and estradiol (P=.09), with hazard ratios similar to those shown for the unadjusted analysis in Table 3. Menopausal stage was the only significant predictor of duration of any hot flushes after adjusting for age, anxiety, and each hormone separately in the multivariable model (P<.001).
In this 13-year follow-up of hot flushes in a randomly identified, population-based cohort, the median duration of moderate-to-severe hot flushes was 10.2 years, well beyond the duration considered in clinical guidelines.19 When women who reported mild hot flushes were included, the median duration of hot flushes increased to 11.6 years.
Previous reports indicated that the duration of hot flushes increased with the length of the follow-up.18,20 However, the present results indicate that it is not simply the length of follow-up but the menopausal stage at onset that more accurately identifies hot flush duration. Women who reported onset of moderate-to-severe hot flushes as they entered the menopause transition had a duration greater than 11 years; in contrast, women whose onset of moderate-to-severe hot flushes was in the late transition or postmenopausal stages had a median duration of approximately 4 years. It is noteworthy that a 4-year median duration is consistent with the results of a meta-analysis of 10 hot flush studies that enrolled women who were generally older than the present cohort and were likely in stages closest to menopause when they were observed in studies.18
As indicated in this study, the initiation of hot flushes is highly variable in relation to menopausal stage, but we know of no biological explanation for this. Although hot flushes appear to originate in the central nervous system and are linked with a lowered thermoregulatory set point, their pathophysiology is not well understood.35,36 Possibly women who report hot flushes in the late premenopause or earliest transition stages have greater sensitivity to hormone fluctuations or perception of symptoms, but this has not been demonstrated. In a previous report from this cohort, anxiety predicted hot flushes and the most anxious women had the most severe hot flushes.37 However, another study that compared anxiety scores between women with hot flushes and women with no hot flushes found significantly higher anxiety in peri- and postmenopausal women but not in the premenopausal women with hot flushes.38 Clear relationships among anxiety, initiation of hot flushes, and biological mechanisms in the same women remain unknown. Age at hot flush onset was a strong predictor of hot flush duration. It is noteworthy that the great majority of women (79%) were younger than age 50 years when they reported the onset of moderate-to-severe hot flushes with 45–49 years the most common ages. The younger the age at onset, the longer the median duration of hot flushes. Hot flush duration was approximately 8 years when onset occurred in the 45- to 49-year age group compared with less than 4 years when onset occurred at ages 50 years or older.
Race was a significant predictor of hot flush duration in the multivariable model that adjusted for body mass index with African American women having a longer duration of hot flushes than white women. These findings extend previous information that indicated African American women were more likely to report hot flushes than other racial groups39–44, but the reasons for racial differences remain unclear.
Obesity was significantly associated with the duration of moderate-to-severe hot flushes, but there was no significant interaction between obesity and race. Overall, obese women had a shorter duration of hot flushes compared with nonobese women. Nonobese African American women had the longest duration of moderate-to-severe hot flushes, followed by nonobese white women. Obese white women had the shortest duration, followed by obese African American women. These data suggest that the shorter duration of hot flushes among obese women may be associated with the well-recognized changes in estrogen metabolism that occur with ovarian senescence when the contribution of estradiol from fat becomes dominant and obese women have higher estradiol levels than nonobese women. We and others have previously reported a postmenopausal shift in the body mass index effect, in which obese women have lower estradiol levels premenopause (and earlier onset of hot flushes) but higher estradiol levels postmenopause compared with nonobese women.45,46
Although a number of studies have reported that women who smoke are more likely to report hot flushes, possibly as a result of the antiestrogenic effects of smoking,43,47,48 current smoking had no significant association with the duration of hot flushes in this cohort. Whether hot flushes started very early or in the late stages of the menopause transition also was not significantly associated with smoking.
Several strengths of this study include the prospective assessment of hot flushes and menopausal stage in a randomly identified cohort over the entire menopause transition. Another strength is the use of survival analysis, although this method may overestimate hot flush duration as a result of interval evaluation of hot flushes in the long follow-up interval. The annual assessments of hot flushes could potentially overestimate the duration by 1 year. On the other hand, although we believe that we evaluated the first onset of moderate-to-severe hot flushes, hot flush duration would be underestimated if hot flushes occurred before the study.
It is possible that hot flushes were the result of conditions other than menopause, although the participants were generally in good physical health, and the strong associations of hot flushes with menopausal stages provide evidence of their association with reproductive aging. The study did not include women using hormone therapy, and studies to determine the effect of hormone therapy on hot flush duration are needed. Finally, the participants represent African American and white urban women in the United States, and the results cannot be generalized to women in nonsimilar areas or other racial groups.
The findings indicate that the median duration of moderate-to-severe hot flushes considerably exceeds the timeframe that is generally accepted in clinical practice. Indeed, a sizable number of women experienced hot flushes in the early stages of the menopause transition and approximately 80% were younger than age 50 years at hot flush onset. This information is important for the clinical management of hot flushes. Perhaps treatments for vasomotor symptoms should be targeted more commonly to younger, irregularly menstruating women, although it must be recognized that traditional hormone therapy may not be the ideal choice for this population given, for example, the problems of breakthrough bleeding and the need for contraception. Other treatments for hot flushes need to be evaluated, particularly for women who have not reached menopause. Race and body mass index also significantly influenced the duration of hot flushes and are important considerations in individualizing treatment, particularly when the predicted duration of hot flushes is substantially longer than the length of hormone therapy that is currently recommended.
1. Williams RE, Kalilani L, DiBenedetti DB, Zhou X, Granger AL, Fehnel SE, et al. Frequency and severity of vasomotor symptoms among peri- and postmenopausal women in the United States. Climacteric 2008;11:32–43.
2. Guthrie JR, Dennerstein L, Taffe JR, Donnelly V. Health care-seeking for menopausal problems. Climacteric 2003;6:112–7.
3. Hoerger TJ, Downs KE, Lakshmanan MC, Lindrooth RC, Plouffe L Jr, Wendling B, et al. Healthcare use among U.S. women aged 45 and older: total costs and costs for selected postmenopausal health risks. J Womens Health Gend Based Med 1999;8:1077–89.
4. Owens GM. Gender differences in health care expenditures, resource utilization, and quality of care. J Manag Care Pharm 2008;14(suppl):2–6.
5. Joffe H, Massler A, Sharkey KM. Evaluation and management of sleep disturbance during the menopause transition. Semin Reprod Med 2010;28:404–21.
6. Kravitz HM, Ganz PA, Bromberger J, Powell LH, Sutton-Tyrrell K, Meyer PM. Sleep difficulty in women at midlife: a community survey of sleep and the menopausal transition. Menopause 2003;10:19–28.
7. Bromberger JT, Kravitz HM, Matthews K, Youk A, Brown C, Feng W. Predictors of first lifetime episodes of major depression in midlife women. Psychol Med 2009;39:55–64.
8. Woods NF, Smith-DiJulio K, Percival DB, Tao EY, Mariella A, Mitchell S. Depressed mood during the menopausal transition and early postmenopause: observations from the Seattle Midlife Women's Health Study. Menopause 2008;15:223–32.
9. Avis NE, Ory M, Matthews KA, Schocken M, Bromberger J, Colvin A. Health-related quality of life in a multiethnic sample of middle-aged women: Study of Women's Health Across the Nation (SWAN). Med Care 2003;41:1262–76.
10. Nappi RE, Lachowsky M. Menopause and sexuality: prevalence of symptoms and impact on quality of life. Maturitas 2009;63:138–41.
11. Reed SD, Ludman EJ, Newton KM, Grothaus LC, LaCroix AZ, Nekhlyudov L, et al. Depressive symptoms and menopausal burden in the midlife. Maturitas 2009;62:306–10.
12. Stahl SM. Vasomotor symptoms and depression in women, part I. Role of vasomotor symptoms in signaling the onset or relapse of a major depressive episode. J Clin Psychiatry 2009;70:11–2.
13. Thurston RC, Sutton-Tyrrell K, Everson-Rose SA, Hess R, Matthews KA. Hot flashes and subclinical cardiovascular disease: findings from the Study of Women's Health Across the Nation Heart Study. Circulation 2008;118:1234–40.
14. Thurston RC, Christie IC, Matthews KA. Hot flashes and cardiac vagal control: a link to cardiovascular risk? Menopause 2010;17:456–61.
15. Thurston RC, Kuller LH, Edmundowicz D, Matthews KA. History of hot flashes and aortic calcification among postmenopausal women. Menopause 2010;17:256–61.
16. Rossouw JE, Prentice RL, Manson JE, Wu L, Barad D, Barnabei VM, et al. Postmenopausal hormone therapy and risk of cardiovascular disease by age and years since menopause. JAMA 2007;297:1465–77.
17. Crandall CJ, Tseng CH, Crawford SL, Thurston RC, Gold EB, Johnston JM, et al. Association of menopausal vasomotor symptoms with increased bone turnover during the menopausal transition. J Bone Miner Res 2010 Sep 27 [Epub ahead of print].
18. Politi MC, Schleinitz MD, Col NF. Revisiting the duration of vasomotor symptoms of menopause: a meta-analysis. J Gen Intern Med 2008;23:1507–13.
19. North American Menopause Society. Treatment of menopause-associated vasomotor symptoms: position statement of The North American Menopause Society. Menopause 2004;11:11–33.
20. Col NF, Guthrie JR, Politi M, Dennerstein L. Duration of vasomotor symptoms in middle-aged women: a longitudinal study. Menopause 2009;16:453–7.
21. Freeman EW, Sammel MD, Liu L, Martin P. Psychometric properties of a menopausal symptom list. Menopause 2003;10:258–65.
22. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Executive summary: Stages of Reproductive Aging Workshop (STRAW). Climacteric 2001;4:267–72.
23. Freeman EW, Sammel MD, Gracia CR, Kapoor S, Lin H, Liu L, et al. Follicular phase hormone levels and menstrual bleeding status in the approach to menopause. Fertil Steril 2005;83:383–92.
24. Gracia CR, Sammel MD, Freeman EW, Lin H, Langan E, Kapoor S, et al. Defining menopause status: creation of a new definition to identify the early changes of the menopausal transition. Menopause 2005;12:128–35.
25. Pitteloud N, Dwyer AA, DeCruz S, Lee H, Boepple PA, Crowley WF Jr, et al. The relative role of gonadal sex steroids and gonadotropin-releasing hormone pulse frequency in the regulation of follicle-stimulating hormone secretion in men. J Clin Endocrinol Metab 2008;93:2686–92.
26. Welt CK, Taylor AE, Fox J, Messerlian GM, Adams JM, Schneyer AL. Follicular arrest in polycystic ovary syndrome is associated with deficient inhibin A and B biosynthesis. J Clin Endocrinol Metab 2005;90:5582–7.
27. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1977;1:385–401.
28. Zung WWK. A rating instrument for anxiety disorders. Psychosomatics 1971;12:371–9.
29. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96.
30. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81.
31. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–70.
32. Cox DR. Regression models and life-tables (with discussion). J Roy Statist Soc Ser B 1972;34:187–220.
33. Kleinman DG, Klein M. Survival analysis: a self-learning text. 2nd ed. New York (NY): Springer-Verlag; 2005, p. 140.
34. Hess KR. Graphical methods for assessing violations of the proportional hazards assumption in Cox regression. Stat Med 1995;14:1707–23.
35. Freedman RR. Physiology of hot flashes. Am J Hum Biol 2001;13:453–64.
36. Loprinzi CL, Barton DL. On hot flash mechanism, measurement, and treatment. Menopause 2009;16:621–3.
37. Freeman EW, Sammel MD, Lin H, Gracia CR, Kapoor S, Ferdousi T. The role of anxiety and hormonal changes in menopausal hot flashes. Menopause 2005;12:258–66.
38. Juang KD, Wang SJ, Lu SR, Lee SJ, Fuh JL. Hot flashes are associated with psychological symptoms of anxiety and depression in peri- and post- but not premenopausal women. Maturitas 2005;52:119–26.
39. Simpkins JW, Brown K, Bae S, Ratka A. Role of ethnicity in the expression of features of hot flashes. Maturitas 2009;63:341–6.
40. Appling S, Paez K, Allen J. Ethnicity and vasomotor symptoms in postmenopausal women. J Womens Health (Larchmt) 2007;16:1130–8.
41. Gold EB, Colvin A, Avis N, Bromberger J, Greendale GA, Powell L, et al. Longitudinal analysis of the association between vasomotor symptoms and race/ethnicity across the menopausal transition: study of women's health across the nation. Am J Public Health 2006;96:1226–35.
42. Thurston RC, Bromberger JT, Joffe H, Avis NE, Hess R, Crandall CJ, et al. Beyond frequency: who is most bothered by vasomotor symptoms? Menopause 2008;15:841–7.
43. Freeman EW, Sammel MD, Lin H, Gracia CR, Pien GW, Nelson DB, et al. Symptoms associated with menopausal transition and reproductive hormones in midlife women. Obstet Gynecol 2007;110:230–40.
44. Thurston RC, Sowers MR, Chang Y, Sternfeld B, Gold EB, Johnston JM, et al. Adiposity and reporting of vasomotor symptoms among midlife women: the study of women's health across the nation. Am J Epidemiol 2008;167:78–85.
45. Freeman EW, Sammel MD, Lin H, Gracia CR. Obesity and reproductive hormone levels in the transition to menopause. Menopause 2010;17:718–26.
46. Randolph JF Jr, Sowers M, Bondarenko IV, Harlow SD, Luborsky JL, Little RJ. Change in estradiol and follicle-stimulating hormone across the early menopausal transition: effects of ethnicity and age. J Clin Endocrinol Metab 2004;89:1555–61.
47. Dennerstein L, Smith AM, Morse C, Burger H, Green A, Hopper J, et al. Menopausal symptoms in Australian women. Med J Aust 1993;159:232–6.
48. Sievert LL, Obermeyer CM, Price K. Determinants of hot flashes and night sweats. Ann Hum Biol 2006;33:4–16.
© 2011 by The American College of Obstetricians and Gynecologists.