Executive Summary of the Stages of Reproductive
Aging Workshop ⴙ 10: Addressing the Unfinished
Agenda of Staging Reproductive Aging
Sioba´ n D. Harlow, Margery Gass, Janet E. Hall, Roger Lobo, Pauline Maki,
Robert W. Rebar, Sherry Sherman, Patrick M. Sluss, and
Tobie J. de Villiers, for the STRAW ⫹ 10 Collaborative Group
Department of Epidemiology (S.D.H.), University of Michigan, Ann Arbor, Michigan; The North American
Menopause Society (M.G.), Mayfield Heights, Ohio; The Endocrine Society (J.E.H.), Department of
Medicine, Harvard Medical School; Boston, Massachusetts; Department of Obstetrics and Gynecology
(R.L.), Columbia University, New York, New York; Department of Psychiatry and Psychology (P.M.),
University of Illinois, Chicago, Illinois; American Society for Reproductive Medicine (R.W.R.), Birmingham,
Alabama; National Institute of Aging (S.S.), Bethesda, Maryland; Department of Pathology (P.M.S.),
Harvard Medical School, Boston, Massachusetts; and International Menopause Society (T.J.d.V.) Cape
Town, South Africa
Objective: The aim of this article is to summarize the recommended updates to the 2001 Stages of
Reproductive Aging Workshop (STRAW) criteria. The 2011 STRAW ⫹ 10 reviewed advances in
understanding of the critical changes in hypothalamic-pituitary-ovarian function that occur before
and after the final menstrual period.
Methods: Scientists from five countries and multiple disciplines evaluated data from cohort studies
of midlife women and in the context of chronic illness and endocrine disorders on change in
menstrual, endocrine, and ovarian markers of reproductive aging including antimu¨ llerian hor-
mone, inhibin-B, follicle-stimulating hormone, and antral follicle count. Modifications were ad-
opted by consensus.
Results: STRAW ⫹ 10 simplified bleeding criteria for the early and late menopausal transition,
recommended modifications to criteria for the late reproductive stage (Stage -3) and the early
postmenopause stage (Stage ⫹1), provided information on the duration of the late transition
(Stage -1) and early postmenopause (Stage ⫹1), and recommended application regardless of wom-
en’s age, ethnicity, body size, or lifestyle characteristics.
Conclusions: STRAW ⫹ 10 provides a more comprehensive basis for assessing reproductive aging
in research and clinical contexts. Application of the STRAW ⫹ 10 staging system should improve
comparability of studies of midlife women and facilitate clinical decision making. Nonetheless,
important knowledge gaps persist, and seven research priorities are identified. (J Clin Endocrinol
Metab 97: 1159 –1168, 2012)
T
he 2001 Stages of Reproductive Aging Workshop
(STRAW) proposed nomenclature and a staging sys-
tem for ovarian aging including menstrual and qualitative
hormonal criteria to define each stage (1– 4). The STRAW
staging system is widely considered the gold standard for
characterizing reproductive aging through menopause, just
as the Marshall-Tanner Stages characterize pubertal matu-
ration (5). Research conducted during the past 10 years has
advanced knowledge of the critical changes in hypothalamic-
pituitary and ovarian function that occur before and after the
ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2012 by The Endocrine Society
doi: 10.1210/jc.2011-3362 Received December 14, 2011. Accepted January 20, 2012.
First Published Online February 16, 2012
SPECIAL FEATURE
Consensus Statement
J Clin Endocrinol Metab, April 2012, 97(4):1159 –1168 jcem.endojournals.org 1159
final menstrual period. These advances were the topic of a
follow-up workshop “STRAW ⫹ 10: Addressing the Unfin-
ished Agenda of Staging Reproductive Aging” (STRAW ⫹
10). STRAW ⫹ 10, held in Washington, DC, on September
20 and 21, 2011, reviewed these scientific advances and up-
dated the STRAW criteria. The sponsors were the National
Institute on Aging (NIA) and Office of Research on Women’s
Health (ORWH) of the National Institutes of Health, The
North American Menopause Society (NAMS), the American
Society for Reproductive Medicine (ASRM), the Interna-
tional Menopause Society (IMS), and the Endocrine Society.
The STRAW ⫹ 10 workshop achieved the following aims:
1. to reevaluate criteria for the onset of late reproduc-
tive life and early menopausal transition, given new
population-based data relating to follicle-stimulat-
ing hormone (FSH), antral follicle count (AFC), an-
timu¨ llerian hormone (AMH), and inhibin-B;
2. to reevaluate criteria for staging postmenopause,
given new population-based data on changes in FSH
and estradiol concentrations after the final men-
strual period (FMP);
3. to reevaluate applicability to women based on vari-
ations in body size, lifestyle characteristics, and
health status; and
4. to identify remaining gaps in scientific knowledge
and research priorities.
BACKGROUND AND SIGNIFICANCE
The menopausal transition marks a period of physiologic
changes as women approach reproductive senescence. Evi-
dence supports the clinical importance of the transition for
many women as a period of temporal changes in health and
quality of life (ie, vasomotor symptoms, sleep disturbance,
depression) and longer-term changes in several health out-
comes (ie, urogenital symptoms, bone, lipids) (6–15) that
may influence women’s quality of life and the likelihood of
healthy aging. As a standardized staging system for repro-
ductive aging, STRAW made a substantial contribution to
women’s health research by providing consistent classifica-
tion of menopause status for studies of midlife women. Im-
portantly, STRAW facilitated research that aimed to distin-
guish the health effects of ovarian versus somatic aging. The
STRAW staging system also serves as a clinical tool for
women and their healthcare providers to guide the assess-
ment of fertility, contraceptive needs, and healthcare deci-
sion making (16, 17).
Building upon previous consensus meetings of the World
Health Organization and the Council of Affiliated Meno-
pause Societies, (18) STRAW reevaluated nomenclature,
proposed a standardized staging system, and recommended
criteria for defining the onset of each stage. STRAW partic-
ipants evaluated potential criteria including menstrual cy-
cles; endocrinologic parameters including FSH, estradiol,
AMH, and inhibin-B; symptoms; fertility; and ovarian im-
aging including AFC. Of the candidate biomarkers consid-
ered in 2001, only FSH was consistently measurable in a
clinical setting. Data were insufficient to define quantitative
criteria for FSH or to clarify the precise timing of change in
FSH levels. Information on AFC and on the relationship
among AMH, inhibin-B, and the timing of ovarian aging was
limited. Symptoms were considered to be subjective and ac-
knowledged not to be universally experienced. STRAW
therefore restricted staging recommendations to menstrual
cycle bleeding criteria and qualitative FSH criteria.
Seven Stages of the 2001 STRAW Staging System
STRAW divided the adult female life into three broad
phases: reproductive, the menopausal transition, and
postmenopause. These three phases included a total of
seven stages centered on the FMP (Stage 0) (1– 4). The
reproductive phase was divided into Stages -5, -4, and -3
corresponding to early, peak, and late, respectively. The
menopausal transition phase consisted of Stage -2 (early)
and Stage -1 (late), and the postmenopause phase con-
tained Stages ⫹1 (early) and ⫹2 (late). Stage -3 was char-
acterized by regular menstrual cycles and increasing levels
of FSH. Stage -2 was characterized by variability in men-
strual cycle length and increased levels of FSH. Stage -1
was characterized by onset of skipped cycles or amenor-
rhea of at least 60 days and continued elevation of FSH.
The ReSTAGE Collaboration
The ReSTAGE Collaboration subsequently conducted
empirical analyses to assess the validity and reliability of the
2001 STRAW’s original menstrual cycle criteria in four co-
hort studies—the TREMIN study, the Melbourne Women’s
Midlife Health Project, the Seattle Midlife Women’s Health
Study, and the Study of Women’s Health Across the Nation
(SWAN). Findings supported STRAW’s recommendations,
provided more precise specification of menstrual criteria for
early and late transition, and recommended a quantitative
cutpoint for FSH levels characteristic of the late transition
(19–22).
Generalizability
A limitation of the original STRAW was its recommen-
dation, based on the available evidence, that the staging sys-
tem only be applied to healthy women. STRAW explicitly
recommended against applying the criteria to seven sub-
groups of women, (1– 4) including smokers (19% of US
women aged 45–64 y
23
), women with a body mass index
(BMI) greater than 30 kg/m
2
(38% of US women (24)), and
1160 Harlow et al. STRAW ⫹ 10 Staging Reproductive Aging J Clin Endocrinol Metab, April 2012, 97(4):1159–1168
women who had undergone hysterectomy (35% of US
women (25)). Women engaged in heavy aerobic exercise and
women with chronic menstrual cycle irregularities, uterine
abnormalities or ovarian abnormalities, or significant illness
such as cancer were also excluded. Another limitation of the
2001 STRAW was the lack of insight regarding the applica-
bility of the staging system in diverse populations. In 2001,
few data were available from studies of multiethnic or diverse
socioeconomic populations. Recent data from multiethnic
cohorts now permit the assessment of generalizability, (17,
22, 26–33) although data from low-resource countries re-
main quite limited (34, 35).
STRAW had a sustained influence on research in the field,
prompting the assessment of trajectories of change in endo-
crine levels and biomarkers of ovarian senescence as well as
the evaluation of how these trajectories vary by body size,
smoking, ethnicity, and other factors (26–33, 36–55). Ten
years later, the understanding of ovarian aging and its endo-
crine and clinical correlates has advanced considerably,
providing a more nuanced and comprehensive under-
standing of the critical junctures that occur during repro-
ductive aging before and after the FMP. For example, Fig.
1 illustrates changes in mean FSH and estradiol concen-
trations in relation to time before and after the FMP in the
SWAN cohort (22). The role of AMH and inhibin-B as
markers of declining fertility and ovarian aging is more
clearly understood, as are the relationships among pat-
terns of decline in AMH, inhibin-B, AFC, and primordial
follicle counts (26,37,38,42,43,45–48,51,53,56–58).
The goal of the STRAW ⫹ 10 was to review significant
advances in the field and develop recommendations for
updating the original STRAW criteria.
METHODS
STRAW ⫹ 10 involved a 2-day in-person meeting hosted
at the 2011 Annual Meeting of NAMS. On the first day,
international experts gave oral presentations reviewing
recent data bearing on the goals, as part of a public sym-
posium, followed by comments and discussion from the
audience. The first two sessions focused on data from pro-
spective cohort studies of midlife women; clinical findings
related to changes in patterns and levels of menstrual, en-
docrine, and ovarian markers of reproductive aging; and
data relevant to how these trajectories vary by ethnicity,
body size, and smoking status. A particular focus was on
patterns of change in AMH, inhibin-B, FSH, estradiol, and
AFC and their interrelationships. A third session focused
on emerging evidence related to staging reproductive ag-
ing in the context of cancer treatment, chronic illness in-
cluding cancer and human immunodeficiency virus
(HIV)–acquired immunodeficiency syndrome (AIDS),
and endocrine disorders including polycystic ovary syn-
drome (PCOS) and primary ovarian insufficiency (POI,
otherwise known as premature ovarian failure). At the end
of day 1, a panel reviewed and the participants discussed
modifications that had been proposed by symposium
speakers. STRAW ⫹ 10 explicitly considered the feasibil-
ity of applying criteria in low-resource countries.
Subsequently, 41 scientists convened to develop con-
sensus and propose modifications to the STRAW staging
system. These participants had clinical and/or research
experience in female reproductive aging and included sci-
entists from several key research groups in the United
States, Canada, Australia, the Netherlands, and South Af-
rica, as well as representatives from National Institutes of
Health–funded cohort studies of midlife women that have
biologic samples (59) including the SWAN, Michigan
Bone Health and Metabolism Study, Seattle Midlife Wom-
en’s Health Study, Biodemographic Models of Reproduc-
tive Aging, and the Penn Ovarian Aging Study as well as
the Australian Melbourne Women’s Midlife Health Proj-
ect and junior investigators who submitted qualifying
posters.
Three breakout groups were formed based on scientific
expertise and interest. Group 1 reviewed criteria for
STRAW Stages -4 to -2. Group 2 reviewed criteria for
STRAW Stages -1 to ⫹2. Each of these two groups was
subdivided into two subgroups and assigned a rapporteur.
Each subgroup proposed modifications to the STRAW
paradigm separately, considering criteria for the relevant
stages in healthy women and the weight of evidence con-
cerning the appropriateness of applying these criteria to
smokers and women regardless of body size. Each sub-
group then reviewed the recommendations of the other
FIG. 1. Adjusted population means (95% CI) for segmented mean
profiles of follicle-stimulating hormone and estradiol across the final
menstrual period in the Study of Women’s Health Across the Nation
(N ⫽ 1,215). *The y axis is unitless. The units of hormone are marked
in the corresponding curves. Reproduced with permission from
Randolph et al, J Clin Endocrinol Metab 2011;96:746 –754.
J Clin Endocrinol Metab, April 2012, 97(4):1159–1168 jcem.endojournals.org 1161
subgroup and discussed points of disagreement until con-
sensus was reached. Group 3 discussed staging in the con-
text of endocrine disorders and chronic illness and pro-
posed modifications.
On the second day, the scientists reconvened to review
and discuss proposed modifications to the STRAW stag-
ing system. First, Group 1 and Group 2 reviewed the other
group’s recommendations. In this way, all groups re-
viewed all stages under consideration (Stages -4 to ⫹2).
Afterward, the group-at-large met to discuss each pro-
posal, and final recommendations were adopted by con-
sensus. Points of disagreement were discussed until con-
sensus was achieved on common principles. In general,
disagreements reflected points for which data was not yet
adequate to make a recommendation. Preliminary recom-
mendations of the STRAW ⫹ 10 were presented at the
NAMS annual meeting on September 22, with comments
and requests for clarification considered by the STRAW ⫹
10 program committee.
RESULTS
STRAW ⫹ 10 retained the criteria for an ideal staging
system used by the 2001 Workshop. Therefore, a staging
system should
1. rely primarily on objective data;
2. use widely available, reliable, noninvasive, and in-
expensive tests;
3. allow for prospective classification of women; and
4. permit unambiguous classification of women into a
unique stage.
In addition, it was concluded that the modified staging
system should
5. retain the same widely accepted nomenclature;
6. consider menstrual cycle criteria to remain the most
important criteria given the continuing lack of in-
ternational standardization of biomarker assays as
well as their cost and/or invasiveness, particularly in
the context of resource-poor countries;
7. consider biomarker criteria as supportive criteria
given the lack of assay standardization (supportive
criteria are to be used only as necessary and should
not be interpreted as required for diagnosis); and
8. use criteria that are independent of age, symptoms,
and pathology (because no universal menopausal
syndrome has been established across ethnic groups,
(60) two key symptoms are incorporated only as de-
scriptive additional information that may support
other criteria in assessing stage (61)).
The revised STRAW ⫹ 10 Staging System is presented in
Fig. 2. STRAW ⫹ 10 recommended the acceptance of the
ReSTAGE Collaboration’s more precise and simplified
specification of the menstrual cycle criteria for the early
and late menopausal transition and concurred with Re-
STAGE recommendations that the quantification of the
FSH criteria in Stage -1 is possible given the improved
standardization of this assay and additional population-
based data. In addition, STRAW ⫹ 10 recommended
modifications to the criteria for the late reproductive stage
(Stage -3) as well as the early postmenopause stage (Stage
⫹1) and provided information on the duration of the late
transition (Stage -1) and early postmenopause (Stage ⫹1)
stages. Although additional biomarkers, especially AMH
and AFC, have considerable promise, the lack of stan-
dardized assays and data from noninfertility populations
remain important limitations to their incorporation into
the STRAW staging system and their utility as clinical
tools for staging reproductive aging. Nonetheless, the re-
vised STRAW ⫹ 10 Staging System includes qualitative
criteria for these biomarkers during the late reproductive
life when relative changes in these parameters have im-
portant consequences for fertility potential.
Definition and Rationale for Key Revisions to the
Staging Criteria
Late reproductive stage (Stage -3). The late reproductive
stage marks the time when fecundability begins to decline
and during which a woman may begin to notice changes in
her menstrual cycles. Given that critical endocrine parame-
ters begin to change before overt changes in menstrual cy-
clicity and that these endocrine changes are important to
fertility assessments, STRAW ⫹ 10 recommended that the
late reproductive stage be subdivided into two substages (-3b
and -3a). In Stage -3b, menstrual cycles remain regular with-
out change in length or early follicular phase FSH levels;
however, AMH and antral follicle counts are low. Most (53,
62) but not all (63) studies suggest that inhibin-B is also low.
In Stage -3a, subtle changes in menstrual cycle characteris-
tics, specifically shorter cycles, (64 – 66) begin. Early follic-
ular phase (cycle days 2–5) FSH increases and becomes more
variable, with the other three markers of ovarian aging being
low. The lack of standardized AMH assays prevented the
development of quantitative recommendations for this
biomarker.
Early menopausal transition (Stage -2). Early meno-
pausal transition is marked by increased variability in
menstrual cycle length, defined as a persistent difference of
7 days or more in the length of consecutive cycles. Persis-
tence is defined as recurrence within 10 cycles of the first
variable length cycle. Cycles in the early menopausal tran-
1162 Harlow et al. STRAW ⫹ 10 Staging Reproductive Aging J Clin Endocrinol Metab, April 2012, 97(4):1159–1168
sition are also characterized by elevated but variable early
follicular phase FSH levels and low AMH levels and AFC.
Late menopausal transition (Stage -1). The late meno-
pausal transition is marked by the occurrence of amenor-
rhea of 60 days or longer. Menstrual cycles in the late
menopausal transition are characterized by increased vari-
ability in cycle length, extreme fluctuations in hormonal
levels, and increased prevalence of anovulation. In this
stage, FSH levels are sometimes elevated into the meno-
pausal range and sometimes within the range character-
istic of the earlier reproductive years, particularly in as-
sociation with high estradiol levels. The development of
international standards and the availability of substantive
population-based data now permit the definition of quan-
titative FSH criteria, with levels greater than 25 IU/L in a
random blood draw characteristic of being in late transi-
tion, based on current international pituitary standards
that approximate more than 40 IU/L in the previously used
urine-based gonadotropin standards (67–69). Empirical
analyses should be undertaken to confirm this recommen-
dation, and researchers and clinicians should carefully
evaluate the appropriate FSH value, depending on the as-
say they use. Based on studies of menstrual calendars and
on changes in FSH and estradiol, this stage is estimated to
last, on average, 1 to 3 years. Symptoms, most notably
vasomotor symptoms, are likely to occur during this stage.
Early postmenopause (Stage ⫹1a, ⫹1b, ⫹1c). New data
on the trajectories of change in mean levels of FSH and es-
tradiol (22, 30, 32, 40, 41, 52, 54, 55) indicate that FSH
continues to increase and that estradiol continues to decrease
until approximately 2 years after the FMP, after which the
levels of each of these hormones stabilize. Therefore,
STRAW ⫹ 10 recommended that early postmenopause be
subdivided into three substages (⫹1a, ⫹1b, and ⫹1c).
Stages ⫹1a and ⫹1b each last 1 year and end at the time
point at which FSH and estradiol levels stabilize. Stage
⫹1a marks the end of the 12-month period of amenorrhea
required to define that the FMP has occurred. It corre-
sponds to the end of “perimenopause,” a term still in com-
mon usage that means the time around menopause and
begins at Stage -2 and ends 12 months after the FMP. Stage
⫹1b includes the remainder of the period of rapid changes
in mean FSH and estradiol levels. Based on studies of hor-
monal changes, Stages ⫹1a and ⫹1b together are esti-
mated to last, on average, 2 years. Symptoms, most no-
tably vasomotor symptoms, are most likely to occur
during this stage.
Stage ⫹1c represents the period of stabilization of high
FSH levels and low estradiol values that is estimated to last
3 to 6 years; therefore, the entire early postmenopause
lasts approximately 5 to 8 years. Further specification of
this stage will require additional studies of trajectories of
change in FSH and estradiol from the FMP through the
late postmenopause.
Late postmenopause (Stage ⫹2). Stage ⫹2 represents
the period in which further changes in reproductive en-
FIG. 2. The Stages of Reproductive Aging Workshop ⫹ 10 staging system for reproductive aging in women.
J Clin Endocrinol Metab, April 2012, 97(4):1159–1168 jcem.endojournals.org 1163
