| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
University of Michigan (J.F.R., S.D.H., R.L., B.N., M.Y.), Ann Arbor, Michigan 48109; University of Massachusetts (S.C.), Worcester, Massachusetts 01655; University of Melbourne (L.D.), Victoria 3050, Australia; University of Washington (K.C., E.S.M.), Seattle, Washington 98195; and Monash University (J.T.), Victoria 3168, Australia
Address all correspondence and requests for reprints to: John F. Randolph, Jr., M.D., L4228 Women’s Hospital, University of Michigan Health System, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109-0276. E-mail: jfrandol{at}med.umich.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: This paper investigates the longitudinal relationship of annual serum FSH levels with four proposed bleeding criteria for the late menopausal transition in two cohort studies. The goal is to provide empirically based guidance regarding application of hormonal criteria that may be optimal for widespread application in clinical and research settings for assessing menopausal stage.
Design/Setting: Prospective menstrual calendar and annual serum FSH data were collected from two population-based cohort studies: the Melbourne Women’s Midlife Health Project and the Study of Women’s Health Across the Nation.
Participants: Participants in the study were 193 Melbourne Women’s Midlife Health Project and 2223 Study of Women’s Health Across the Nation women aged 42–57 yr at baseline who contributed 10 or more menstrual cycles and at least one annual serum FSH value.
Main Outcome Measure(s): Association between bleeding criteria for the late menopausal transition and FSH was a main outcome measure. Associations of bleeding criteria, FSH, and hot flashes with the final menstrual period were also measured.
Results: A single FSH measure is an independent marker of the late menopausal transition, but FSH concentrations are less predictive of menopausal stage than any of four proposed bleeding criteria. Criterion FSH values for the late transition are similar across both studies. Experience of hot flashes adds no information in the presence of hormonal and bleeding criteria.
Conclusions: An annual serum FSH concentration of 40 IU/liter could be incorporated, in conjunction with bleeding markers, into the Stages of Reproductive Aging Workshop paradigm for markers of the late menopausal transition.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The need for a rigorous characterization of the menopausal transition both to standardize terminology and criteria for investigational purposes and to inform women and their care providers prompted the Stages of Reproductive Aging Workshop (STRAW) in July, 2001 (9), which resulted in a seven-stage classification of female reproductive life based on the best available evidence. The proposed STRAW paradigm included bleeding, symptom, and hormonal criteria, but operational definitions of hormone parameters were not included because insufficient information was available to establish definitive criteria. Consensus was reached that, of the candidate biomarkers of the transition, FSH was the best available, and STRAW recommended that specific criteria be developed for incorporating FSH into the evolving system for staging reproductive aging (9). Vasomotor symptoms, including hot flashes, although characteristic of the menopausal transition, are not experienced by all women and are not considered essential criteria within the staging system.
The ReSTAGE collaboration is conducting an empirical evaluation of proposed criteria for staging reproductive aging in the context of the natural history of the menopausal transition. In a previous paper (10), we assessed proposed bleeding criteria for the late menopausal transition using prospectively collected menstrual calendar data from four large cohort studies: TREMIN (11), the Melbourne Women’s Midlife Health Project (MWMHP) (1), the Seattle Midlife Women’s Health Study (SMWHS) (3), and the multisite, multiethnic, Study of Women’s Health Across the Nation (SWAN) (4). Although our findings supported the recommendation from STRAW that amenorrhea of 60 d be used as the bleeding criterion for the late transition, a correlation between this criterion and underlying hormonal changes and symptoms has yet to be demonstrated. Annual serum FSH levels are available from MWMHP and SWAN, and the assays have been cross-validated to allow comparability. This paper investigates the longitudinal relationship of annual serum FSH with four proposed bleeding markers of the late menopausal transition. Our underlying goal is to address the question: does information on serum FSH and hot flashes improve the determination of a woman’s entry into the late menopausal transition compared with the use of bleeding criteria for the late menopausal transition alone?
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This study has been approved by Institutional Review Boards at the Universities of Michigan, Washington, Massachusetts, and Melbourne. For this paper, we conducted secondary analyses using menstrual calendar information, annual serum FSH concentrations, and self report of hot flashes accumulated in two longitudinal cohort studies of the menopausal transition.
The MWMHP began in 1991 with a cross-sectional survey of 2001 mid-aged Australian-born women of Anglo-European heritage identified by random telephone digital dialing (response rate = 71%) (12). From these women, the MWMHP enrolled 438 women, aged 45–55 yr, who had menstruated in the prior 3 months and were not using hormone therapy into a longitudinal study (response rate = 56%). This analysis includes records from the 193 women who maintained menstrual records through at least 10 untreated bleeding segments and provided at least one untreated serum sample for FSH measurement. (The term bleeding segment is analogous to menstrual cycle but acknowledges that menstrual calendar data cannot distinguish menstrual from nonmenstrual bleeds.)
The SWAN is an ongoing multiethnic, multisite longitudinal community-based study of mid-aged women begun in 1995 (4). A cross-sectional survey randomly selected women aged 40–55 yr from a variety of lists including a large managed health care plan, community census, utility households, and registered voters, or by random digit dialing. Each site then recruited about 450 women including Caucasian women and women from one minority group (African-Americans at four sites; and Japanese, Chinese, and Hispanic women at one site each). Eligibility requirements for the longitudinal cohort included age 42–52 yr, an intact uterus, no reproductive hormone use at time of enrollment, and at least one menstrual period in the previous 3 months. A total of 3302 women were enrolled. This analysis includes records of the 2223 women who maintained menstrual records through at least 10 untreated segments and provided at least one untreated serum sample for FSH measurement.
Both studies included a menstrual calendar and asked women to record menstrual bleeding on each day bleeding occurred. Although slightly different in format, calendars identified the date of bleeding onset and days of bleeding in a comparable manner. Ascertainments of the onset of the late transition and of the FMP are based on the menstrual calendar data.
In MWMHP, each menstrual calendar card covered 1 yr. Participants were followed up to 10 yr with annual assessments which included symptoms, measured body mass index (BMI), smoking history, hormone use, and blood sampling to measure FSH. Experience of hot flashes was obtained from the North American symptom checklist (13). Fasting morning blood samples for hormone assays were taken between d 4 and 8 of the menstrual cycle for those still cycling, or after 3 months of amenorrhea as previously described (14). FSH was measured initially by RIA in yr 1 (7). Two subsequent changes in method were used, the automated Microparticulate Enzyme Immunoassay (IMX Analyzer; Abbott Diagnostics, Chicago, IL) (yr 2 and 3), and for all samples from yr 4 onward, the TOSOH AIA1200 automated Enzyme Immunoassay. Correlation coefficients were: FSH (IMX) and RIA, 0.98, and FSH (TOSOH) with IMX 0.99. Interassay and intraassay coefficients of variation were 5.4 and 4.0%, respectively.
In SWAN, women were provided monthly menstrual calendars. Assessments at baseline and annual follow-up visits included symptoms, measured BMI, smoking history, hormone use, and blood sampling to measure FSH. Blood was drawn between d 2 and 5 of the follicular phase of the menstrual cycle in regularly cycling women. Serum was frozen at –80 C and sent on dry-ice to the SWAN Endocrine Laboratory at the University of Michigan. Hormone assays were conducted using an ACS-180 automated analyzer (Bayer Diagnostics, Norwood, MA). Serum FSH concentrations were measured with a two-site chemiluminometric immunoassay that uses constant amounts of two monoclonal antibodies (provided by Bayer Diagnostics). Each antibody is directed to different regions on the ß-subunit (one coupled to paramagnetic particles and the other labeled with dimethylacridinium ester) with specificity for intact FSH. Interassay and intraassay coefficients of variation were 12.0 and 6.0%, respectively.
Cross-validation of serum assays used in the Melbourne and SWAN Studies was conducted. Laboratories for each study received and analyzed samples from the other study. As neither measure can be defined as the gold standard, we created standardized values using an equating procedure (15) such that the standard scores (z scores) of the log FSH distribution from each assay are equated to each other. All analyses used these standardized FSH values.
Using definitions recommended by the World Health Organization (16) and modified by ReSTAGE, a bleeding episode is a period of consecutive bleeding days; a bleeding-free interval is a period of consecutive bleeding-free days; and a bleeding segment is a bleeding episode and the subsequent bleeding-free interval. In ReSTAGE a single day of bleeding, as well as consecutive days of spotting/bleeding, were coded as bleeding episodes. Bleeding-free intervals had to consist of at least 3 d. One or two bleed-free days between two bleed days were considered part of the bleeding episode. Pregnancies and the first three segments after birth/abortion are considered nonmenstrual intervals and are excluded from analyses (17). For each bleeding segment, we indicate whether a woman was using hormone therapy or oral contraceptives, fertility medications, selective estrogen receptor modulators, injectable contraceptives, contraceptive implants, or a short course of hormones to treat a menstrual disorder. Gaps in the menstrual record are coded as missing.
This paper compares three recently proposed bleeding criteria (3, 18, 19, 20) for onset of the late menopausal transition with the current standard of 90 d of amenorrhea (5). We refer to the bleeding episode that marks when each criterion is met as the "marker event", with age at occurrence defined as age on the first day of that bleeding episode. We calculated the following marker events: the first observed 1) segment of at least 90 d (3); 2) segment of at least 60 d (20); 3) skipped segment, defined as a segment that exceeded twice the median segment length of the previous 10 segments (3); and 4) running range over a 10-segment sequence (18). For the running range, we identify the first day of the bleeding segment when the running range exceeds 42 d. Menopause is defined as the date of the FMP recorded in the menstrual calendars established retrospectively after 12 months of amenorrhea were observed.
We created a dichotomous variable based on whether the woman reported experiencing any hot flashes (yes/no). In MWMHP, women were asked whether they had been bothered by hot flashes in the last 2 wk. Originally scored on a four-point scale (0, not present; 1, causes minor irritation; 2, interferes with normal life; 3, debilitating), we defined scores of 1–3 as having hot flashes. SWAN inquired about the presence of hot flashes in the past 2 wk, with frequency originally indicated on a five-point scale (0, not at all, to 4, everyday). We defined scores of 1–4 as having hot flashes.
Smoking status (current, former, never) is based on annual questionnaires. BMI is based on a linear interpolation of annual height and weight measures, and categorized as less than 25, 25–29.9, 30–35, and more than 35.
Analysis
Women were censored at hysterectomy, bilateral oophrectomy, initiation of hormonal birth control, hormone therapy, or chemotherapy (Table 1
). Segments during which a short-course of hormonal medication, including fertility medication, was used to treat a menstrual disorder were treated as gaps in the menstrual record.
|
We examined the association of FSH with each bleeding marker by estimating a binomial logistic regression model for each marker (has occurred, yes/no) as a function of log-transformed FSH, before and after adjusting for concurrent chronologic age, baseline BMI, and baseline smoking status, and ethnicity in SWAN. Within-woman correlation from multiple observations per woman was accounted for by using generalized estimating equations (21). We tested interactions between log FSH and each covariate to determine whether associations between FSH and bleeding markers varied by subgroup. From these models, we estimated the probability that a marker had occurred for various values of FSH. In addition, we computed age-adjusted odds ratios for marker occurrence in terms of categorized serum FSH, where categories were defined based on our unadjusted associations and on published standards (19, 22). Estimated probabilities of marker occurrence, done separately for each study, reflect not only the role of FSH but also the age distribution of each study. In contrast, the odds ratios indicate the contribution of FSH apart from age.
We used a Cox model to assess how well bleeding markers, serum FSH concentrations, and hot flash experience predict FMP. We started with a varying coefficient Cox model (28) where the marker effect on FMP varies with age at the marker event (19). Estimation proceeded using the regression spline method. Because the log hazard as a function of age at the marker event is approximately linear, the final regressions model the log hazard of the marker to FMP as a linear function of age at marker. These models were compared with corresponding models excluding FSH or hot flashes to assess the additional contributions of FSH and hot flashes in predicting the FMP. Potential confounding and interactions by smoking status, BMI and, within SWAN, by ethnicity, were examined.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. Reflecting eligibility criteria, MWMHP participants were older than SWAN participants. SWAN participants were heavier, and MWMHP participants were less likely to have smoked.
|
). The steepest increase in the marker probability occurred with an increase in FSH from 10–20 IU/liter. The curves for the 10-segment running range and a 60-d segment were similar to each other, particularly in the MWMHP dataset. At any given FSH value, estimated probabilities for a skipped segment and a 90-d segment were lower. Upon adjustment for age, the increase in the marker probability with an increase in FSH was less steep, but FSH remained significantly positively associated with each marker. Further adjustment for covariates resulted in little additional change in the FSH associations with markers, and interactions of FSH with these covariates were not statistically significant (data not shown). Associations were similar for SWAN and MWMHP.
|
, an increase in FSH was more strongly related to the 10-segment running range and the 60-d segment than to the skipped segment, as seen in the larger odds ratios (Table 3). The odds ratio was larger still for the 90-d segment. For all markers, odds ratios were highest for FSH equal to or more than 40 IU/liter (vs. <10 IU/liter). Adjustment for age reduced the magnitude of the FSH odds ratios relative to unadjusted odds ratios (data not shown), but patterns of association were similar before and after adjustment for age. Results were similar for SWAN and MWMHP.
|
presents the relative hazard of FMP for serum FSH concentrations and hot flashes. Cutpoints of 20–39 IU/liter and equal to or more than 40 IU/liter were selected based on Fig. 1 and the published standards (19, 22). The likelihood of reaching the FMP increased with age, thus, the log-relative hazard of menopause for those women who have experienced a bleeding marker as compared with those who have not declined with age (19, 20). When considered alone, serum FSH concentrations of 20–39 IU/liter were associated with a 2- to 4-fold increase in the log relative hazard of FMP, whereas values of 40 IU/liter and above were associated with a 6- to 12-fold increase, although the confidence intervals are wide for both cutpoints and in both cohorts. However, when serum FSH concentrations were considered jointly with bleeding markers, the effect of FSH was markedly attenuated, reflecting the associations of FSH with the bleeding markers seen in Fig. 1 and Table 3. In the MWMHP, the relative hazard of FMP was increased 4-fold when serum FSH concentrations were 20 IU/liter or higher. In SWAN, which includes younger women and a shorter period of follow-up, a similar magnitude of effect was seen when serum FSH concentrations were 40 IU/liter or higher. The adjusted association between FSH and FMP varied little across models including different bleeding markers of the late transition. Annual serum FSH concentration provided additional information about which women were closest to menopause.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Single annual serum FSH concentrations have been associated with the menopausal transition (24) as an indirect measure of the change in ovarian function that underpins reproductive aging, but they have not been rigorously related to well-defined bleeding markers of menopausal stage (5). FSH levels were qualitatively included in the STRAW model for reproductive aging (9), but specific cutpoints were not specified given the absence of a well-defined relationship between FSH concentrations and onset of the transition. A recent analysis of randomly drawn serum FSH levels in 576 women ages 35–60 participating in the National Health and Nutrition Examination Survey in 1999–2000 (25) concluded that FSH alone "has limited utility in distinguishing among women in different reproductive stages", but was constrained by cross-sectional design, a classification of menopausal transition as newly irregular menses in the last 12 months attributed to "going through the menopause", and a population skewed to premenopausal women.
Specification of cutpoints for annual serum FSH levels that might serve as criteria for onset of the menopausal transition has been problematic due to the variable pattern of FSH secretion, the variation in available assays (McConnell, D. S., unpublished data), and the independent effects of age, body size, and other variables (26). Cutpoints in this analysis reflect clinically used values but should not be construed as having specific intrinsic biological implications. We found that an annual serum FSH concentration of 40 IU/liter is informative across both studies and could clearly be incorporated in conjunction with bleeding markers into the STRAW paradigm. In MWMHP there was no difference between the cutpoints 20 and 40 IU/liter. In SWAN, 20 IU/liter was less predictive than 40 IU/liter. SWAN includes younger women and women who have transitioned earlier, suggesting that the higher FSH values may be most predictive of women who transition faster or at a younger age. MWMHP includes older women who have all transitioned, suggesting that FSH over 20 IU/liter may be more informative for older women, whereas an FSH level over 40 IU/liter may be more informative for younger women.
Bleeding markers are most informative of a woman being in the late menopausal transition (19, 20, 22), but several competing criterion are currently proposed (19, 20). FSH levels are less informative, but do add to the predictability of time to FMP. The additional information provided by serum FSH concentrations is similar regardless of the choice of bleeding criteria, suggesting that the ovarian biology underlying these changes in menstrual bleeding is the same for all of the proposed bleeding criteria for the late menopausal transition. Thus, selection of an optimal bleeding marker should be based on considerations such as frequency of occurrence, association with time to FMP, and ease of measurement.
FSH stimulates folliculogenesis and estradiol (E2) production in the ovary, and is regulated primarily in a classic negative feedback loop by the inhibins A and B secreted by the corpus luteum and follicle, respectively (14, 24, 27, 28), and to a lesser extent by E2. As oocyte competence and the total number of ovarian follicles declines with age, negative feedback decreases as reflected in a progressive rise in FSH initially in the early phase of the menstrual cycle (24, 29). In both cross-sectional (7, 25, 26) and longitudinal (24, 30, 31) studies of the menopausal transition, single measures of these biomarkers have substantial variability, but decreases in inhibin B and increases in FSH are observed before decreases in E2 (14, 24, 30). Widespread use of the inhibin B measure has been precluded by cost and sensitivity of the assay (32). Assays of FSH vary by as much as 2-fold (McConnell, D. S., unpublished data), making comparisons between studies problematic and the establishment of criteria concentrations as reliable markers of transition stages difficult. These difficulties have been overcome in the current analysis by the similarities of the study design and subject populations, and by cross-validation of the FSH assay in the respective laboratories (15).
The association of FSH with the bleeding markers is expected because bleeding is the clinical reflection of the underlying utero-ovarian biology. Given the variability of annual serum measures of FSH, bleeding markers may be most informative in the late menopausal transition as they reflect the cumulative impact of a changing hormonal environment including the most biologically significant changes in FSH. As such, they remain the most clinically significant markers of the late transition. Serum FSH may be a more important marker for the early menopausal transition when bleeding markers are less predictive of FMP, and future analyses of the ReSTAGE collaboration will address this question (33).
Self report of hot flashes has also been associated with the menopausal transition (1, 5, 34) and serum FSH levels (35), and is independently informative in the absence of any other marker of the late transition, albeit only modestly. However, when either FSH level or any bleeding marker is available, reported hot flashes do not add additional information regarding the likelihood of being in the late transition.
This analysis has some limitations. Both studies set a lower age limit for enrollment and excluded women who were not still menstruating, so we cannot evaluate whether FSH concentrations are more or less relevant for women with an earlier onset of the late menopausal transition. Moreover, SWAN is still in progress and not all women have completed their reproductive transition. Nonetheless, the results are notably robust, and support recommendations regarding the modification of the STRAW paradigm to include an annual serum FSH concentration of equal to or more than 40 IU/liter as an additional criterion for the late menopausal transition.
In summary, we have shown that a single annual early follicular serum FSH level is positively associated with four bleeding markers of the late menopausal transition and is an independent but less robust marker of the late transition. Hot flashes similarly are associated positively with the late transition, but are much less robust and are no longer significant when considered together with FSH and bleeding markers. Thus, serum FSH levels should be incorporated into classification criteria for the late menopausal transition.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
ReSTAGE is supported by Grant AG 021543 (Siobán Harlow, PI) from the National Institute of Aging. Data collection for the MWMHP was supported by the Victorian Health Promotion Foundation and the National Health and Medical Research Council of Australia (Lorraine Dennerstein, PI). The SWAN has grant support from the National Institutes of Health, Department of Health and Human Services, through the National Institute on Aging, the National Institute of Nursing Research, and the National Institutes of Health Office of Research on Women’s Health (Grants NR004061, AG012505, AG012535, AG012531, AG012539, AG012546, AG012553, AG012554, AG012495). The Seattle Midlife Women’s Health Study is supported by the National Institute of Nursing Research (Grants NR004141 and NR04001).
First Published Online May 23, 2006
1 The ReSTAGE Collaboration. Authors are listed in alphabetical order. 
Abbreviations: BMI, Body mass index; E2, estradiol; FMP, final menstrual period; MWMHP, Melbourne Women’s Midlife Health Project; STRAW, Stages of Reproductive Aging Workshop; SWAN, Study of Women’s Health Across the Nation.
Received February 2, 2006.
Accepted May 11, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. G. Hudson, G. L. Gierach, F. Modugno, J. Simpson, J. W. Wilson, R. W. Evans, V. G. Vogel, and J. L. Weissfeld Nonsteroidal Anti-inflammatory Drug Use and Serum Total Estradiol in Postmenopausal Women Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 680 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Hale, X. Zhao, C. L. Hughes, H. G. Burger, D. M. Robertson, and I. S. Fraser Endocrine Features of Menstrual Cycles in Middle and Late Reproductive Age and the Menopausal Transition Classified According to the Staging of Reproductive Aging Workshop (STRAW) Staging System J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3060 - 3067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D. Harlow, K. Cain, S. Crawford, L. Dennerstein, R. Little, E. S. Mitchell, B. Nan, J. F. Randolph Jr., J. Taffe, and M. Yosef Evaluation of Four Proposed Bleeding Criteria for the Onset of Late Menopausal Transition J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3432 - 3438. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
