[Frontiers in Bioscience S3, 474-486, January 1, 2011]

The endocrinology of perimenopause: need for a paradigm shift

Jerilynn C Prior, Christine L Hitchcock

Endocrinology and Metabolism, Department of Medicine, and Centre for Menstrual Cycle and Ovulation Research (CeMCOR), University of British Columbia, and Vancouver Coastal Health Research Institute, 2775 Laurel St, 4th floor, Vancouver, British Columbia, Canada V5Z 1M9

TABLE OF CONTENTS

1. Abstract
2. Purpose, perspective and definitions
3.Ovarian changes of perimenopause
3.1. Estrogen changes-the paradox of perimenopause
3.2. Progesterone, ovulation and luteal phase changes in perimenopause
4. Hypothalamic-pituitary-ovarian feedback changes of perimenopause
4.1. Inhibin and control of perimenopausal follicle stimulating hormone (FSH)
4.2. Feedback loops, ovulation and the luteinizing hormone (LH) peak
5. Additional perimenopause hormonal changes in cortisol and catecholamines
6. Perimenopausal hormonal origins for common experience changes:
6.1. Heavy menstrual flow
6.2. Increased premenstrual symptoms
6.3. Weight gain
6.4. Night sweats/hot flushes
7. Perimenopausal hormonal changes and future health risks:
7.1. Cardiovascular disease
7.2. Breast and endometrial cancer
7.3. Osteoporosis and fracture risks
8. Summary and perspective
9. References

1. ABSTRACT

Perimenopause, rather than a time of declining estrogen, is characterized by three major hormonal changes that may begin in regularly menstruating women in their mid-thirties: erratically higher estradiol levels, decreased progesterone levels (in normally ovulatory, short luteal phase or anovulatory cycles), and disturbed ovarian-pituitary-hypothalamic feedback relationships. Recent data show that approximately a third of all perimenopausal cycles have a major surge in estradiol occurring de novo during the luteal phase. This phenomenon, named "luteal out of phase (LOOP)" event, may explain a large proportion of symptoms and signs for symptomatic perimenopausal women. Large urinary hormone data-sets from women studied yearly over a number of years in the Study of Women Across the Nation (SWAN) and in the Tremin data will eventually provide a more clear prospective understanding of within-woman hormonal changes. Predicting menopause proximity with FSH or Inhibin B levels is documented to be ineffective. Anti-Mullerian hormone levels may prove predictive. Finally, there is an urgent need to change perimenopause understandings, language and therapies used for midlife women's symptoms to reflect these hormonal changes.

2. PURPOSE, PERSPECTIVE AND DEFINITIONS

Current care of midlife women is often informed by outdated knowledge about the nature of perimenopause. Recent scientific reviews (1-4) have concluded that perimenopause is characterized by intermittently and sometimes dramatically high levels of estrogen, lower progesterone levels and disturbed hormonal feedback loops. This evidence is in contrast with assumptions of dropping or lower estrogen levels, and the casual use of the term "estrogen deficiency" as a synonym for perimenopause. Although we can easily link perimenopausal heavy flow with both high estrogen levels and inadequate progesterone-related endometrial effects (5), today this menorrhagia is commonly treated with oral contraceptives that make it worse (6), or with ablative surgeries of the endometrium or the uterus rather than with effective high dose long-cycle progesterone/progestins (7). Thus, although the major purpose of this review is to highlight current new knowledge of the changing endocrinology of perimenopause, we also want to contribute to improved clinical care, education and research.

Our perspective is a complex integration of several points of view-that of scientists studying the confusing endocrinology of perimenopause and women's changing daily experiences over time, as investigators committed to knowledge translation from research into clinical care, and as educators trying to change concepts that may be prejudicial for women. One of us has also spent 40 years as a physician attempting to make the data gathered in a clinical setting speak a scientific language (8).

Taxonomy, or the naming of things, is fundamental for any field of science (9). We understand that definitions not only arise from but also create concepts that may prevent scientific advancement-for example, with the diagnosis of hysteria, which was formerly thought caused by the uterus flying throughout the body. There is a danger in prematurely naming incompletely understood phenomena, but it is a necessary part of the scientific process. However, clear definitions are essential for this paper and for our purpose. Therefore, building on the work performed by numerous investigators and the Stages of Reproductive Aging Workshop (STRAW; 10; Figure 1), we will create and justify definitions that fit with what we now know (see Table 1).

For the purposes of this review we define perimenopause as the entire transitional period from women's mature reproductive to a non-reproductive state. Thus, as shown in Figure 2, perimenopause begins when the estrogen levels have already begun to rise and progesterone levels have decreased but cycles continue to be regular (11). We have postulated a clinical definition of changes in women's experiences any three of which will make the diagnosis of perimenopause. It is likely that perimenstrual cyclic night sweats without daytime vasomotor symptoms is sufficient unique that it, alone, may be diagnostic. This clinical diagnostic tool (Table 2) needs to be validated against the final menstruation prospectively in a population-based sample. This diagnosis of perimenopause in women with regular cycles is controversial because all of the current taxonomies define the onset of women's midlife transition by decreases in the regularity or frequency of menstruation (10; 12; 13) yet by the onset of irregular cycles, women have usually experienced several years of hormonal changes typical of perimenopause and their associated symptoms.

Perimenopause ends, as adopted by STRAW, one year beyond the final menstrual flow (10). Thus, perimenopause as defined here encompasses part of the STRAW "late reproductive stage," the early and late menopause transition stages and one year beyond the final menstrual flow (Figure 2).

The end of perimenopause also defines the onset of what we prefer to call menopause (rather than postmenopause, which relies on the use of the word "menopause" to mean the retrospectively defined literal final menstrual period). This taxonomy avoids the inappropriate dual naming of the year following the last menstrual flow as both "postmenopause" and "perimenopause" that both the WHO and STRAW schema preserved (10; 14). Menopause, in our understanding, although it is also a hormonally complex and an experientially changing life phase, is the state within which women achieving it then permanently reside.

3. CHANGES IN OVARIAN STEROIDS DURING PERIMENOPAUSE

Before discussing the hormonal changes of perimenopause it is important to state clearly that our normal referent is the endocrinology of the premenopausal normally ovulatory cycle (15; 16). Ideally, all studies of perimenopausal hormonal changes would have a within-center premenopausal control group that is similar in body mass index, racial mixture and dietary and activity variables.

3.1. Estradiol changes-the paradox of perimenopause

The word "menopause" is commonly used as inclusive of both women's midlife transitional phase and the stable non-reproductive phase. Because estradiol is truly low beginning about a year after the final menstrual flow, it is assumed that during perimenopause levels must be dropping. Teleologically, the functional requirement in perimenopause is to rid the ovary of hormonally responsive follicles that could be stimulated by FSH later in life (17). Anatomical studies have found an increased rate of recruitment of small follicles that undergo atresia-this starts about age 37 (17). At the same time estradiol levels become, on average, higher. Many of the symptoms of perimenopause are also associated with unusually high levels of estradiol. Thus perimenopause is a time during which estradiol levels become erratic and often high before eventually becoming lower than in reproductive-aged women.

Santoro and colleagues were the first to state the then-radical notion that estradiol levels were higher in perimenopause (18). This observation was based on a cross-sectional single cycle, daily urinary hormone study published in 1996 showing significantly higher levels especially during the luteal phase (Figure 3). This observation was confirmed by a meta-analysis comparing samples from premenopausal women to perimenopausal women within the same research center; average estradiol levels were statistically higher in perimenopausal women (4). Not one of the primary studies from which these data were extracted had the power to see significant differences or even noted these higher levels (4). Most remarkably, the cross-sectional analysis of data from an excellently designed, population-based study of perimenopausal women aged 45-55 was summarized as a "decrease in E2. . .levels" (19), despite abnormally high and hugely variable follicular phase serum estradiol levels (Figure 4). This inability to "see" data other than what is expected (20) appears to persist, especially about the higher estradiol levels of perimenopause.

The meta-analysis, previously mentioned, of follicular and premenstrual estradiol levels from studies published before 1998 showed that perimenopausal mean serum estradiol levels are 29% higher in the follicular phase and 22% higher in the premenstruum than in premenopausal women (4). We now know that estradiol levels are dramatically higher in some cycles and for some of the time (21). Although it is hard, yet, to put into a population perspective, in their prospective observational study, Hale and colleagues estimate that about a third of all menopausal transition cycles show evidence of "luteal out of phase" (LOOP) events, meaning, as shown in Figure 5, that a second, and usually higher estradiol peak appears after the normal midcycle estradiol peak (21). It is not yet known whether LOOP events are the primary explanation for the erratically higher estradiol levels of perimenopause.

3.2. Progesterone, ovulation and luteal phase changes in perimenopause

By contrast with the higher levels of estradiol, lower levels of progesterone characterize perimenopause. These lower levels arise through three mechanisms: 1) decreased progesterone production within normal-length ovulatory cycles; 2) shortened luteal phase lengths within ovulatory cycles; and 3) more frequently anovulatory cycles. The evidence about perimenopausal changes in ovulation and luteal phase lengths are sparser than the now extensive literature on the higher and erratic estradiol levels of perimenopause, and none are from population-based samples (22). Early observers such as Metcalf using once/week urinary pregnanediol excretions (PdG) documented that only 37% of irregular perimenopausal cycles were ovulatory (23). Likewise using classical (non-quantitative) basal body temperature, Doring showed that ovulatory cycles only occurred in 50% of cycles in women ages 46-50 (24), and Vollman by quantitative basal temperature (QBT) (25) noted ovulatory disturbances (meaning short luteal phases and/or anovulatory cycles) were significantly increased with increasing gynecological age over 35 years. Illustrative prospective QBT data that are valid against PdG, in one woman with regular cycles documented over more than 10 years showed that, in cycles that averaged 29± 2 days long, only a single cycle out of a series of 14 had a short luteal phase length (8%) with no anovulatory cycles when she was in her mid-30s. However, 10 years later her regular cycles were shorter (27± 2 d) but now 86% of cycles showed short luteal phases and 14% were anovulatory (22).

Several longitudinal studies have monitored ovulatory characteristics in perimenopause. Hale and colleagues from Australia (26) used thrice-weekly serum samples over a single cycle in a convenience sample comparing four groups: mid reproductive (controls), late reproductive, early menopausal transition, and late menopausal transition (26). They found lower progesterone levels in perimenopausal ovulatory cycles as well as higher estradiol levels (26). Other studies have used analysis of urinary hormone metabolites, which are more easily obtained, but absolute levels must be interpreted with caution, because of genetic variability in steroid metabolism (33).

Miro and colleagues presented the results of the British FREEDOM study (Fertility Recognition Enabling Early Detection Of Menopause), a convenience sample using daily first morning urines provided by 103 women prospectively studied over six to 18 months (27). Results showed that some women with increased urinary FSH levels but still regular cycles had higher PdG excretions than did women who were premenopausal, however, luteal phase lengths are not described (27). In general normal-length ovulatory cycles had higher PdG levels than "delayed" ovulatory cycles in what would be the early menopause transition by STRAW. In subsequent reproductive stages PdG decreased as estrogen excretions (E1G) increased (27).

The Study of Women Across the Nation (SWAN) prospectively followed a multiethnic cohort of over 3,000 women ages 42-52 from seven centers in the USA. Some were invited to join a daily hormone sub-study; of these, 840 collected daily first morning urines for 50 days/y. The majority of cycles (81%) met criteria for being ovulatory (28; 30). Women with ovulatory cycles were further separated into those who were still menstruating regularly (late reproductive age) and those who had skipped a menstrual period. Within ovulatory cycles, women in late menopausal transition had higher excretions of gonadotrophins, lower PdG and no difference in E1G (30). Anovulatory cycles were separated into three forms: normal midcycle peak of estrogen followed by an LH peak but not by PdG, a normal midcycle estrogen peak not followed by an LH peak, or no evidence of usual midcycle events suggesting that anovulation and low progesterone production were related to various kinds of hypothalamic-pituitary ovarian feedback disturbances (28). Of women starting with regular ovulatory cycles, having oligomenorrhea and anovulation (in data truncated at 50 days) was a strong predictor of being in the late perimenopause stage. In regular cycles, anovulation continued at about 10% of all cycles. However, within ovulatory cycles PdG levels decreased by about seven percent per year (29; Figure 6). Obesity and greater education were associated with lower PdG levels (29).

Finally, the Biodemographic Model of Reproductive Aging (BIMORA) study of the Tremin longitudinal menstrual cycle data collection in college-educated Caucasian women asked 156 women to collect daily first morning urines over six months (January to July) yearly for five years (31). Assessment of gonadal steroids primarily focused on reporting estrogen excretions that were not counterbalanced by PdG-this "unopposed" estrogen exposure increased during the menopausal transition because the 95th percentile for estrogen remained the same while ovulatory cycles and the amount of PdG were decreasing (32). This perimenopausal increased estradiol to progesterone ratio is similar to the observations made earlier by Metcalf and MacKenzie (34).

In summary, in perimenopause both older studies employing QBT and more recent studies monitoring serum and urinary hormones have found decreasing levels of progesterone within ovulatory cycles and a rising incidence of ovulation disturbances. Ovulatory cycle progesterone production decreases gradually but continuously with progress toward menopause. Estrogen production unopposed by appropriate progesterone production has been documented in one of the earliest (34) as well as one of the most recent studies of the endocrinology of perimenopause (32). In contrast to menopause, when both estrogen and progesterone are low, in perimenopause the two primary ovarian steroids are changing in opposite directions-estradiol levels higher and progesterone levels lower. These changed ovarian steroid hormone levels and ratio would be expected to increase women's risk for both endometrial (35) and breast cancer (36). It remains to be seen what contribution these changing steroid ratios make to women's midlife experiences and risk factors for subsequent menopausal health.

4. HYPOTHALAMIC-PITUITARY-OVARIAN FEEDBACK CHANGES OF PERIMENOPAUSE

Recent research has clarified that the primary hormonal changes of perimenopause result not only from the "aging ovary" but also from disruption of the usual positive and negative hormonal and paracrine feedback networks controlling the normal ovulatory menstrual cycle. The evidence for these changes in control mechanisms will now be briefly discussed.

4.1. Inhibin and control of perimenopausal follicle stimulating hormone (FSH)

We initially postulated that the higher estradiol levels in midlife women represented "perimenopausal endogenous ovarian hyperstimulation" (4). The term "endogenous" indicates that these changes occur from within a woman's reproductive system. The term "ovarian hyperstimulation" is used by analogy with that of ovulation induction for in vitro fertilization that can create extreme estradiol levels and medical emergencies. Higher estradiol levels normally act to suppress rising FSH levels; in perimenopause this feedback fails, particularly at the follicular-luteal transition.

The prime mover in the feedback disruptions that result in the hormonal changes of perimenopause is now confirmed to be Inhibin B. At the point at which each ovary contains fewer than 100 follicles (2), Inhibin B levels, made by small antral follicles, decline and no longer hold early cycle FSH level in check. This, in turn, leads to increased recruitment of follicles, each of which contributes to the increasing estradiol levels. Important prospective studies in a few women show decreases in Inhibins A and B and progesterone levels but maintained estradiol productions (37). That observation has been confirmed with sophisticated multiple linear regression studies of FSH and luteinizing hormone (LH) and their relationships with Inhibins A and B and their complex relationships with estradiol and progesterone (37; 38). These studies confirm the role of declining Inhibin B levels in allowing FSH to rise in the follicular phase. Elevated FSH, in turn, appears to stimulate the second estradiol peak called LOOP during the luteal phase (38). Because normally ovulatory progesterone levels may prevent a further cohort of follicles from becoming stimulated enough to cause a second mid-cycle like estradiol peak, this is further evidence of hypothalamic-pituitary-ovarian feedback disruption.

Although there are yet no clear feedback roles for anti-Mullerian hormone (AMH) in the feedback and control of perimenopausal changes, several studies from the same Australian group highlight the dropping levels of AMH across the perimenopausal transition (26; 38). Anti-Mullerian hormone is produced by granulosa cells from small follicles and levels are parallel with the number of remaining ovarian follicles as measured by antral follicle counts (on transvaginal ultrasound). Research suggests that AMH is highly reproducible at the same cycle phase (39), and is basically stable across the menstrual cycle (40). In addition, AMH levels decline over time in within-woman studies at ages 36 and 40 (41). It appears clear that lower AMH levels provide a more reliable biomarker of approaching menopause than any other known measure (42). However, at the present moment whether or not AMH plays a role in the disturbed hormonal feedback loops in perimenopause is unknown (38).

4.2. Feedback loops, ovulation and the luteinizing hormone (LH) peak

The just discussed FSH changes relate to the control of estradiol in both the follicular phase, and also during the luteal phase in the form of LOOP events. However, there are additional changes in the control of ovulation and progesterone production that appear to relate more to the hypothalamic-pituitary-axis than to ovarian aging. An elegant description of SWAN daily urinary hormone data in cycles without evidence of luteal activity (anovulatory) shows that both estradiol and LH peaks can be normal but ovulation still not occur (28). One might speculate, given the strong negative associations between FSH (controlled for LH) and progesterone shown in the Australian analysis, that higher FSH levels might interfere with release of an egg and progesterone production (38). That phenomenon also can occur in premenopausal cycles but appears to be more common in perimenopause.

Additional perimenopausal feedback imbalances leading to anovulation involve disturbances of estradiol's positive feedback on LH such that a normal estradiol peak occurs but an LH peak does not follow (28). Finally, cycles may have no evidence of either an estradiol or an LH peak despite high follicular phase estradiol levels-there is also no ovulation (28). Thus there can be no doubt that the hypothalamus or pituitary can become insensitive to estradiol feedback resulting in anovulation.

5. ADDITIONAL PERIMENOPAUSE HORMONAL CHANGES-CORTISOL AND CATECHOLAMINES

To our knowledge there are no prospective studies of the autonomic and sympathetic system and cortisol stress hormone changes across the perimenopause. It is accepted that perimenopausal women are more likely than premenopausal women to report "mood swings" and various symptoms (tachycardia, chest pain, dry mouth, tingling, panic attacks) that are related to a stress response-further research should investigate whether associated hormones such as cortisol or Adrenocorticotrophic Hormone (ACTH) and/or sympathetic nervous system responses (such as norepinephrine) also show increases during perimenopause.

Moreover, the most symptomatic women in perimenopause even more commonly describe stress-related symptoms. Given this, it would be surprising if cortisol and catecholamine levels were not higher in symptomatic perimenopause, and higher levels of estrogen may be responsible. In an experimental study, healthy young men wore either a transdermal estradiol patch or placebo and 24-48 hours later were subjected to the Trier Social Stress Test (43). Those men wearing the estradiol patch experienced greater increases than controls in cortisol, ACTH and norepinephrine to the social stressors of the testing (43). Elevated endogenous estradiol is a potential mechanism to explain the experience women have of being less able to cope with stress than they were earlier in their lives. However, further changes in hypothalamic-pituitary responses may also play a role. Certainly women with disturbed sleep and vasomotor symptoms may experience greater stress hormone responses (as discussed below).

6. PERIMENOPAUSAL HORMONAL ORIGINS FOR COMMON EXPERIENCE CHANGES

There is a shared hope among physicians and midlife women that "a test" will tell them whether or not they are in perimenopause, if they are now menopausal (if they have had a hysterectomy, for example), or how close they are to becoming menopausal. Some physicians use the level of serum FSH level early in the cycle (say cycle day 3) as a test for perimenopause. For an individual woman, however, FSH is neither sensitive nor specific (44). There are hopes that the AMH will prove to be useful for deciding about proximity to menopause but further validation is needed.

The hypothesis that elevated perimenopausal estradiol levels were behind perimenopausal experiences was based on clinical observations of estrogen-associated experiences (increasingly heavy flow, increased premenstrual symptoms, mastalgia, fluid retention, weight gain) in cycles documented with the Daily Perimenopause Diary (45) and QBT (46). Cycle lengths tend to shorten in older, regularly menstruating women and the follicular phase becomes shorter (47). Shortened follicular phase lengths are associated with higher early cycle serum estradiol levels (16) and with higher urinary FSH levels in both follicular and late luteal phases (48).

6.1. Heavy menstrual flow

The best evidence linking symptoms to high estrogen comes from the analysis of very heavy flow (menorrhagia) (5). In a clinical study, 28 women over age 40 with heavy and/or heavy irregular flow were age-matched (± 2 years) with 28 women with regular cyclical bleeding. Estradiol and FSH levels were measured in a standardized way in all women showing that, although FSH levels were not different, estradiol levels were almost doubled in those with heavy flow (5). Twenty of the women with heavy flow had endometrial biopsies and 50% of these showed endometrial hyperplasia, also suggesting progesterone deficiency may have played a role in the heavy flow (5).

A recent quantitative study of the amount of flow during two consecutive menstruations in the Australian study by Hale and colleagues shows that flow increases in absolute amount and in variability from mid- to late-reproductive and across until the late menopausal transition (49). Flow was over 250 ml/menstrual period in women in the late perimenopause who experienced an ovulatory cycle with very high estradiol levels or the unique LOOP phenomenon (49). Thus two studies strongly associate higher estradiol levels with clinically and absolutely heavy menstrual flow (5; 49).

Although heavy flow in perimenopause is commonly attributed to fibroids, it is rare for this benign tumor, originating in the uterine muscle, to impinge on the endometrium. Rather heavy flow and fibroids co-occur because both are associated with higher estradiol and lower progesterone levels.

6.2 Increased premenstrual symptoms

We have previously suggested that the midlife increase in premenstrual symptoms is a result of the higher estradiol levels of regularly cycling women in very early perimenopause (50). Increased premenstrual symptoms also predict night sweats or hot flushes (51; 52). In healthy ovulatory young women (mean age 33), we found no hormonal relationship, however, with premenstrual mood symptoms (53). Although the exact combined hormonal and sociocultural etiologies of premenstrual symptoms are unclear, prospective within-woman data relate them to cycles with higher estradiol/lower progesterone levels (54) that are characteristic of the ovarian hormone changes in perimenopause. However, studies are inconsistent and contradictory, and the relationship between mood and hormones is not clear.

6.3. Weight gain

Whether or not the metabolic changes (weight, body mass index (BMI), waist circumference, blood pressure and cardiovascular adverse lipid changes) that increase in midlife women bear any relationship to the hormonal changes of perimenopause is currently unclear. The prospective daily urinary hormone study in SWAN found, however, that estradiol levels were not different in those with and without a diagnosis of the metabolic syndrome (29). What is quite clear, however, is that the BMI increases that occur in the population across ages 25-50 differ in men and women (55; Figure 7). Using prospective data from the randomly selected men and women in the 9243-strong Canadian Multicentre Osteoporosis Study, men have a gradual BMI increase after age 25 whereas women's BMI remains stable until it increases dramatically between ages 35-44 and ages 45-54. The majority of women ages 45-54 in the population will be in perimenopause, given that menopause occurs on average at age 52 in women without hysterectomy or ovariectomy (55). This is a strong suggestion that there are gender differences in the metabolic changes that occur during the 40s and 50s.

6.4. Night sweats/hot flushes

Although night sweats are often labeled as "estrogen deficiency symptoms", they are present in midlife women who continue to have regular menstrual cycles (and usually called "premenopausal" in surveys) (45). The initial large survey of hot flushes in women noted that seven women volunteered that their night sweats were cyclic and occurred around flow (56). In an observational study of Daily Perimenopause Diaryã data from our laboratory, we observed not only that night sweats predominated over daytime hot flushes/flashes (vasomotor symptoms, VMS) but that they had a cyclic character with increased prevalence before or during menstruation (45). Whether night sweats occur in cycles with greater downward estradiol swings and higher average estradiol levels, as we postulate, is not currently known-we are currently performing a daily urine study to explore hormonal relationships with night sweats in very early perimenopausal women with regular cycles.

7. PERIMENOPAUSAL HORMONAL CHANGES AND FUTURE HEALTH RISKS

7.1. Cardiovascular disease

Heart diseases and stroke (together called cardiovascular disease or CVD) are considered by health experts to be the leading killer of women in North America although they don't enter most perimenopausal women's consciousness and predominantly occur for much older women. However, it is important to ask the question whether the hormonal changes of perimenopause increase or decrease the risks for menopausal CVD. If the notion were correct that estrogen treatment earlier rather than later in menopause is protective against CVD, one could postulate that those with the highest perimenopausal endogenous estradiol levels would have the least subsequent CVD. Unfortunately, to date, no studies that have carefully characterized endogenous estradiol levels or tracked estradiol/estrogen treatment in perimenopause have subsequently observed the cohort for the development of CVD.

However, there are new data suggesting that lower perimenopausal progesterone levels may be related to subsequent CVD. Before sharing data from that study it is useful to remember there are mechanisms through which progesterone may decrease CVD. Intra-arterial progesterone compared with intra-arterial estradiol has been shown to equally improve endothelial function in a random-ordered vehicle-controlled experiment in 28 healthy women who were recently menopausal (57). Oral micronized progesterone also lowers blood pressure (58) and appears lipid neutral when given with estradiol (59). The strongest data relating progesterone to subsequent CVD are from a nested case control study from the Netherlands in which menstruating women of mean age 47 collected first morning urines for hormone levels on cycle day 22 of three consecutive cycles in 1982-86. All new myocardial infarctions among women in the same community were observed through 1991 to determine perimenopausal CVD risks (60). Forty-five women had acute myocardial infarction and in these versus controls, estrogen and testosterone excretions and other reproductive variables did not predict heart attacks over approximately eight years (60). However, those with lower progesterone levels and with probable anovulatory cycles tended to be more likely to have a heart attack during follow-up (smoking-adjusted odds-ratio: 1.3, 95% CI 0.3-3.7). With longer follow-up and an increased incidence of heart attack these data are likely to become stronger.

7.2. Breast and endometrial cancer

In contrast to CVD, midlife women are very concerned about breast cancer because most women have known someone who developed it or died from it. Given that two randomized controlled trials show that breast topical estradiol treatment stimulates and progesterone treatment inhibits human breast cell proliferation and cell division (61; 62), the increased frequency of perimenopausal days of estradiol unopposed by progesterone (32; 34) should increase breast cancer risk. Consistent with those data, higher perimenopausal progesterone levels in a case-control study were associated with a lower risk for breast cancer in the European Prospective Investigation into Nutrition and Cancer study (odds-ratio: 0.61, 95% CI 0.38, 0.98) (63). Finally, women in their 40s (mostly perimenopausal) are more likely than women in their 50s (mostly menopausal) to develop interval breast cancer between screenings (64; British Columbia Cancer Agency data). The strongest data suggesting that the lower progesterone levels of perimenopause may pose a breast cancer risk are from the large E3N cohort in which menopausal women treated with estradiol without progesterone had an increased breast cancer risk that did not occur when estradiol plus progesterone was the treatment (36).

Perimenopausal women appear little concerned about endometrial cancer, perhaps because it is less prevalent than breast cancer. Unopposed estrogen as a pharmacological therapy is a well recognized risk factor for endometrial cancer. Metcalf and MacKenzie (34) and O'Connor and colleagues (32) found that endogenous estrogen unopposed by progesterone was common in perimenopause. In addition, perimenopause has been identified as a "window of risk" for endometrial cancer (35). Further research is needed into whether perimenopausal exposure to high estradiol and low progesterone levels predicts subsequent risk for endometrial cancer.

7.3. Osteoporosis and fracture risks

Although the common belief has been, and still may be, that bone loss and fractures begin in women with estrogen deficiency who are menopausal, there is incontrovertible evidence that rapid bone loss begins in perimenopause (65-68). This is a paradox because of the higher estradiol levels earlier outlined. However, these observations fit perfectly with newer understandings of bone physiology-downward swinging estradiol levels release cytokines and increase RANKL that causes increased bone resorption. In a meta-analysis of spinal bone change observed in perimenopausal versus women early in menopause we documented a mean rate of bone loss of -1.8% per year in perimenopause that exceeded the rate of -1.2 % in early menopausal women (4).

Although it is predominantly bone resorption that determines bone loss and fracture (69), it has becoming increasingly clear that endogenous progesterone levels increase bone formation and contribute with estradiol to positive premenopausal bone change (70-72). An ongoing prospective study in Munich is documenting QCT spinal bone change prospectively (67). This group is now observing the prevalence of ovulatory cycles in relationship to bone change and showing greater bone loss in those with fewer normally ovulatory cycles (73). These data strongly suggest that the perimenopausal years-with swinging estradiol levels leading to increased bone resorption and lower progesterone levels causing less formation-set the stage for menopausal osteoporosis and fracture.

8. SUMMARY AND PERSPECTIVE

This review has presented evidence that demands a shift in the paradigm about perimenopausal changes. Perimenopausal women are not estrogen deficient; rather they have physiologically high and erratic levels of estrogen. Many of the clinical presentations of perimenopause are related to these high levels of estrogen, particularly heavy menstrual bleeding, breast tenderness and an increased response to psychological stressors. Moreover, perimenopausal estrogen is also poorly suppressible, because of impaired feedback mechanisms. The second ovarian hormone, progesterone, is largely ignored in the literature, but is clearly low compared with what was normal earlier in life. Lower progesterone levels or the abnormal ratio of estrogen and progesterone also likely causes many perimenopausal symptoms.

In the absence of relevant perimenopausal clinical trials, clinicians have extrapolated from studies of menopausal women to menstruating perimenopausal women whose hormonal circumstances are very different. The practice of prescribing estrogen (as menopausal hormone therapy, or as oral contraceptives) to perimenopausal women makes little physiological sense, and may cause harm. A few clinicians (including the first author) have instead prescribed progesterone as a monotherapy in perimenopause. Clinical trials have demonstrated that progesterone and progestins are effective therapies for heavy flow and progesterone improves sleep fragmentation. In clinical practice progesterone has also proved effective for many other issues in perimenopause, especially hot flushes. Clinical trials are needed to address other therapeutic uses of progesterone as a monotherapy in perimenopause.

The hormonal changes of perimenopause plus the onset of the reduced social status associated with aging in western culture make perimenopause a complex transition. However, it can also be a time of increased self-awareness and adaptation that can bring women feelings of improved self-worth. Should the perimenopausal changes that women notice not be explained and appropriately named, this can interfere with potential adaptation and resilience (74). There is urgent need for controlled trial research into therapies for the consequences of perimenopausal endocrine changes. A number of the current medical responses to perimenopausal symptoms and signs are no longer appropriate. For example, when a perimenopausal woman is told that she is "too young" to be having night sweats, or that she must be imagining things if she becomes symptomatic while she is still regularly menstruating, when her heavy flow is treated with hysterectomy rather than proven non-surgical strategies (7), when she is given therapies such as combined hormonal contraceptives with additional estrogen that add to her already erratically high endogenous estrogen, there may well be harm (6). Even within a purely clinical framework, the use of ineffective therapies or ones that make women feel worse have the potential to damage the clinician-patient relationship going into midlife and thus to decrease the potential for healthy aging. Within the broader context of women's lives and of recent perimenopausal endocrine research findings, there needs to be a paradigm shift.

9. REFERENCES

1. H.G. Burger, G.E. Hale, D.M. Robertson, L. A. Dennerstein: Review of hormonal changes during the menopausal transition: focus on findings from the Melbourne Women's Midlife Health Project. Hum Reprod Update 13(6), 559-65 (2007)
doi:10.1093/humupd/dmm020
PMid:17630397

2. HG Burger, GE Hale, L Dennerstein, DM Robertson: Cycle and hormone changes during perimenopause: the key role of ovarian function. Menopause 15(4 Pt 1), 603-12 (2008)
doi:10.1097/gme.0b013e318174ea4d
PMid:18574431

3. J.C. Prior: Ovarian aging and the perimenopausal transition: the paradox of endogenous ovarian hyperstimulation. Endocrine 26(3), 297-300 (2005)
doi:10.1385/ENDO:26:3:297

4. J.C. Prior: Perimenopause: The complex endocrinology of the menopausal transition. Endocr Rev 19, 397-428 (1998)
doi:10.1210/er.19.4.397
PMid:9715373

5. M.H. Moen, H. Kahn, K.S. Bjerve, T.B. Halvorsen: Menometrorrhagia in the perimenopause is associated with increased serum estradiol. Maturitas 47(2), 151-5 (2004)
doi:10.1016/S0378-5122(03)00250-0

6. R.F. Casper, S. Dodin, R.L. Reid, Study Investigators: The effect of 20 �g ethinyl estradiol/1 mg norethindrone acetate (MinestrinTM), a low-dose oral contraceptive, on vaginal bleeding patterns, hot flashes, and quality of life in symptomatic perimenopausal women. Menopause 4, 139-47 (1997)
doi:10.1097/00042192-199704030-00004

7. G.A. Irvine, M.B. Campbell-Brown, M.A. Lumsden, A. Heikkila, J.J. Walker, I.T. Cameron: Randomised comparative trial of the levonorgestrel intrauterine system and norethisterone for treatment of idiopathic menorrhagia. Br J Obstet Gynaecol 105(6), 592-8 (1998)
PMid:9647148

8. AR Feinstein AR. Scientific methodology in clinical medicine. II. Classification of human disease by clinical behavior. Ann Intern Med 61, 757-81 (1964)
PMid:14217139

9. A.R. Feinstein, Clinical judgment. Baltimore: The Williams & Wilkins Company, 1967.

10. M.R. Soules, S. Sherman, E. Parrott, R. Rebar, N. Santoro, W. Utian, N. Woods: Executive summary: stages of reproductive aging workshop (STRAW). Fertil Steril 76, 874-8 (2001)
doi:10.1016/S0015-0282(01)02909-0

11. J.C. Prior: Clearing confusion about perimenopause. B C Med J 47(10), 534-8 (2005)

12. D.J. Brambilla, S.M. McKinlay, C.B. Johannes: Defining the perimenopause for application in epidemiologic investigations. Am J Epidemiol 140:12, 1091-5 (1994)
PMid:7998591

13. S.D. Harlow, E.S. Mitchell, S. Crawford, B. Nan, R. Little, J. Taffe: The ReSTAGE Collaboration: defining optimal bleeding criteria for onset of early menopausal transition. Fertil Steril 89(1), 129-40 (2008)
doi:10.1016/j.fertnstert.2007.02.015
PMid:17681300    PMCid:2225986

14. WHO Technical Report Series. Research on the menopause in the 1990's. A report of the WHO Scientific Group. 866, 1-107. 1996. World Health Organization, Geneva, Switzerland, World Health Organization.

15. H.K. Nielsen, K. Brixen, R. Bouillon, L. Mosekilde: Changes in biochemical markers of osteoblastic activity during the menstrual cycle. J Clin Endocrinol Metab 70, 1431-7 (1990)
PMid:2110577

16. B.H. Landgren, A.L. Unden, E. Diczfalusy: Hormonal profile of the cycle in 68 normally menstruating women. Acta Endocrinol (Copenh) 94, 89-98 (1980)

17. S.J. Richardson, V. Senikas, J.F. Nelson: Follicular depletion during the menopausal transition: evidence for accelerated loss and ultimate exhaustion. J Clin Endocrinol Metab 65:1231 (1987)
doi:10.1210/jcem-65-6-1231

18. N. Santoro, J. Rosenberg, T. Adel, J.H. Skurnick: Characterization of reproductive hormonal dynamics in the perimenopause. J Clin Endocrinol Metab 81:4, 1495-501 (1996)
doi:10.1210/jc.81.4.1495

19. H.G. Burger, E.C. Dudley, J.L. Hopper, J.M. Shelley, A. Green, A. Smith, L. Dennerstein, C. Morse: The endocrinology of the menopausal transition: a cross-sectional study of a population-based sample. J Clin Endocrinol Metab 80, 3537-45 (1995)
doi:10.1210/jc.80.12.3537

20. J.C. Prior, S.I. Barr, Y.M. Vigna: The controversial endocrinology of the menopausal transition (letter). J Clin Endocrinol Metab 81, 3127-8 (1996)
doi:10.1210/jc.81.8.3127

21. G.E. Hale, C.L. Hughes, H.G. Burger, D.M. Robertson, I.S. Fraser: Atypical estradiol secretion and ovulation patterns caused by luteal out-of-phase (LOOP) events underlying irregular ovulatory menstrual cycles in the menopausal transition. Menopause 16(1), 50-9 (2009)
doi:10.1097/gme.0b013e31817ee0c2
PMid:18978637

22. J.C. Prior. The ageing female reproductive axis II: ovulatory changes with perimenopause. In: D.J. Chadwick, J.A. Goode, editors. Endocrine Facets of Ageing. Chichester, UK: John Wiley and Sons Ltd p. 172-86 (2002)
doi:10.1002/0470846542.ch11

23. M.G. Metcalf: Incidence of ovulatory cycles in women approaching the menopause. J Biosoc Sci 11, 39-48 (1979)
doi:10.1017/S0021932000012037
PMid:422596

24. G.K. Doring: The incidence of anovular cycles in women. J Reprod Fertil (Suppl 6), 77-81 (1969)

25. R.F. Vollman. The menstrual cycle. In: E.A. Friedman, editor. Major Problems in Obstetrics and Gynecology, Vol 7. 1 ed. Toronto: W.B. Saunders Company, 1977. p. 11-193.

26. G.E. Hale, X. Zhao, C.L. Hughes, H.G. Burger, D.M. Robertson, 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 92(8), 3060-7 (2007)
doi:10.1210/jc.2007-0066

27. F. Miro, S.W. Parker, L.J. Aspinall, J. Coley, P.W. Perry, J.E. Ellis: Sequential classification of endocrine stages during reproductive aging in women: the FREEDOM study. Menopause 12(3), 281-90 (2005)
doi:10.1097/01.GME.0000147018.30796.25
PMid:15879917

28. G. Weiss, J.H. Skurnick, L.T. Goldsmith, N.F. Santoro, S.J. Park: Menopause and hypothalamic-pituitary sensitivity to estrogen. JAMA 292(24), 2991-6 (2004)
doi:10.1001/jama.292.24.2991
PMid:15613667

29. N. Santoro, S.L. Crawford, W.L. Lasley, J.L. Luborsky, K.A. Matthews, D. McConnell, J.F. Randolph, Jr, E.B. Gold, G.A. Greendale, S.G. Korenman, L. Powell, M.F. Sowers, G. Weiss: Factors related to declining luteal function in women during the menopausal transition. J Clin Endocrinol Metab 93(5), 1711-21 (2008)
doi:10.1210/jc.2007-2165

30. N. Santoro, B. Lasley, D. McConnell, J. Allsworth, S. Crawford, E.B. Gold, J.S. Finkelstein, G.A. Greendale, J. Kelsey, S. Korenman, J.L. Luborsky, K. Matthews, R. Midgley, L. Powell, J. Sabatine, M. Schocken, M.F. Sowers, G. Weiss: Body size and ethnicity are associated with menstrual cycle alterations in women in the early menopausal transition: The Study of Women's Health across the Nation (SWAN) Daily Hormone Study. J Clin Endocrinol Metab 89(6), 2622-31 (2004)
doi:10.1210/jc.2003-031578

31. R.J. Ferrell, K.A. O'Connor, D.J. Holman, E. Brindle, R.C. Miller, G. Rodriguez, J.A. Simon, P.K. Mansfield, J.W. Wood, M. Weinstein: Monitoring reproductive aging in a 5-year prospective study: aggregate and individual changes in luteinizing hormone and follicle-stimulating hormone with age. Menopause. 14(1), 29-37 (2007)
doi:10.1097/01.gme.0000227859.50473.20
PMid:17019379

32. K.A. O'Connor, R.J. Ferrell, E Brindle, J. Shofer, D.J. Holman, R.C. Miller, D.E. Schechter, B. Singer, M. Weinstein: Total and unopposed estrogen exposure across stages of the transition to menopause. Cancer Epidemiol Biomarkers Prev 18(3), 828-36 (2009)
doi:10.1158/1055-9965.EPI-08-0996
PMid:17416772

33. Y. Lin, G.D. Anderson, E. Kantor, L.M. Ojemann, A.J. Wilensky: Differences in the urinary excretion of 6-b -hydroxycortisol/cortisol between Asian and Caucasian women. J Clin Pharmacol 39(6), 578-82 (1999)
doi:10.1177/00912709922008182

34. M.G. Metcalf, J.A. MacKenzie: Menstrual cycle and exposure to estrogens unopposed by progesterone: relevance to studies on breast cancer incidence. J Endocrinol 104, 137-41 (1985)
doi:10.1677/joe.0.1040137
PMid:3968497

35. G.E. Hale, C.L. Hughes, J.M. Cline: Endometrial cancer: hormonal factors, the perimenopausal "window of risk", and isoflavones. J Clin Endocrinol Metab 87, 3-15 (2002)
doi:10.1210/jc.87.1.3

36. A. Fournier, F. Berrino, F. Clavel-Chapelon: Unequal risks for breast cancer associated with different hormone replacement therapies: results from the E3N cohort study. Breast Cancer Res Treat 107(1), 103-11 (2008)
doi:10.1007/s10549-007-9523-x
PMid:17333341    PMCid:2211383

37. C.K. Welt, D.J. McNicholl, A.E. Taylor, J.E. Hall: Female reproductive aging is marked by decreased secretion of dimeric inhibin. J Clin Endocrinol Metab 84, 105-11 (1999)
doi:10.1210/jc.84.1.105

38. D.M. Robertson, G.E. Hale, D. Jolley, I.S. Fraser, C.L. Hughes, H.G. Burger: Interrelationships between ovarian and pituitary hormones in ovulatory menstrual cycles across reproductive age. J Clin Endocrinol Metab 94(1), 138-44 (2009)
doi:10.1210/jc.2008-1684

39. R. Fanchin, J. Taieb, D.H. Lozano, B. Ducot, R. Frydman, J. Bouyer: High reproducibility of serum anti-Mullerian hormone measurements suggests a multi-staged follicular secretion and strengthens its role in the assessment of ovarian follicular status. Hum Reprod 20(4), 923-7 (2005)
doi:10.1093/humrep/deh688
PMid:15640257

40. W.J. Hehenkamp, C.W. Looman, A.P. Themmen, F.H. de Jong, E.R. te Velde, FJ Broekmans: Anti-Mullerian hormone levels in the spontaneous menstrual cycle do not show substantial fluctuation. J Clin Endocrinol Metab 91(10), 4057-63 (2006)
doi:10.1210/jc.2006-0331

41. IA van Rooij, FJ Broekmans, GJ Scheffer, CW Looman, JD Habbema, FH de Jong, B.J. Fauser, A.P. Themmen, E.R. te Velde: Serum anti-Mullerian hormone levels best reflect the reproductive decline with age in normal women with proven fertility: a longitudinal study. Fertil Steril 83(4), 979-87 (2005)
doi:10.1016/j.fertnstert.2004.11.029
PMid:15820810

42. M.R .Sowers, A.D .Eyvazzadeh, D. McConnell, M. Yosef, M.L .Jannausch, D. Zhang, S. Harlow, J.F. Randolph, Jr.: Anti-mullerian hormone and inhibin B in the definition of ovarian aging and the menopause transition. J Clin Endocrinol Metab 93(9), 3478-83 (2008)
doi:10.1210/jc.2008-0567

43. C. Kirschbaum, N. Schommer, I. Federenko, J. Gaab, O. Neumann, M. Oellers, N. Rohleder, A. Untiedt, J. Hanker, K.M. Pirke, D.H. Hellhammer: Short-term estradiol treatment enhances pituitary-adrenal axis and sympathetic responses to psyhosocial stress in healthy young men. J Clin Endocrinol Metab 81, 3639-43 (1996)
doi:10.1210/jc.81.10.3639

44. H.G. Burger: Diagnostic role of follicle-stimulating hormone (FSH) measurements during the menopausal transition--an analysis of FSH, oestradiol and inhibin. Eur J Endocrinol 130, 38-42 (1994)
doi:10.1530/eje.0.1300038
PMid:8124478

45. G.E. Hale, C.L. Hitchcock, L.A. Williams, Y.M. Vigna, J.C. Prior: Cyclicity of breast tenderness and night-time vasomotor symptoms in mid-life women: information collected using the Daily Perimenopause Diary. Climacteric 6(2), 128-39 (2003)
PMid:12841883

46. J.L. Bedford, J.C. Prior, C.L. Hitchcock, S.I. Barr: Detecting evidence of luteal activity by least-squares quantitative basal temperature analysis against urinary progesterone metabolites and the effect of wake-time variability. Eur J Obstet Gynecol Reprod Biol 146(1):76-80 (2009)
doi:10.1016/j.ejogrb.2009.05.001

47. E.A. Lenton, B.H. Landgren, L. Sexton, R. Harper: Normal variation in the length of the follicular phase of the menstrual cycle: effect of chronological age. Br J Obstet Gynaecol 91, 681-4 (1984)
PMid:6743609

48. F. Miro, S.W. Parker, L.J. Aspinall, J. Coley, P.W. Perry, J.E. Ellis: Relationship between follicle-stimulating hormone levels at the beginning of the human menstrual cycle, length of the follicular phase and excreted estrogens: the FREEDOM study. J Clin Endocrinol Metab 89(7), 3270-5 (2004)
doi:10.1210/jc.2003-031732

49. G.E. Hale, F. Manconi, G. Luscombe, I.S. Fraser: Quantitative measurements of menstrual blood loss in ovulatory and anovulatory cycles in middle- and late-reproductive age and the menopausal transition. Obstet Gynecol 115(2 Pt 1), 249-56 (2010)
doi:10.1097/AOG.0b013e3181ca4b3a
PMid:20093896

50. J.C. Prior. Premenstrual symptoms and signs. In: R.E. Rabel, E.T. Bope, editors. Conn's Current Therapy 2002. New York: W.B. Saunders Company, 2002. p. 1078-80.

51. J.R. Guthrie, L. Dennerstein, J.L. Hopper, H.G. Burger: Hot flushes, menstrual status, and hormone levels in a population-based sample of midlife women. Obstet Gynecol 88(3), 437-42 (1996)
doi:10.1016/0029-7844(96)00196-2

52. E.W. Freeman, M.D. Sammel, P.J. Rinaudo, L. Sheng: Premenstrual syndrome as a predictor of menopausal symptoms. Obstet Gynecol 103(5 Pt 1), 960-6 (2004)
PMid:15121571

53. A.T. Harvey, C.L. Hitchcock, J.C. Prior: Ovulation disturbances and mood across the menstrual cycles of healthy women. J Psychosom Obstet Gynaecol 30(4), 207-14 (2008)

54. M. Wang, L. Seippel, R.H. Purdy, T. Backstrom: Relationship between symptom severity and steroid variation in women with premenstrual syndrome: Study on serum pregnenolone, prenenolone sulfate, 5α-Pregnane-3,20-Dione, and 3α-Hydroxy-5α-Pregnan-20-one. J Clin Endocrinol Metab 81, 1076-82 (1996)
doi:10.1210/jc.81.3.1076

55. W.M. Hopman, C. Leroux, C. Berger, L. Joseph, S.I. Barr, J.C. Prior, M. Harrison, S. Poliquin, T. Towheed, T. Anastassiades, D. Goltzman, CaMos Research Group: Changes in body mass index in Canadians over a five-year period: results of a prospective, population-based study. BMC Public Health 7, 150 (2007)
doi:10.1186/1471-2458-7-150
PMid:17620129    PMCid:1947966

56. F. Kronenberg: Hot flashes: epidemiology and physiology. Ann.N.Y.Acad.Sci. 592, 52-86 (1990)
doi:10.1111/j.1749-6632.1990.tb30316.x
PMid:2197954

57. K.J. Mather, E.G. Norman, J.C. Prior, T.G. Elliott: Preserved forearm endothelial responses with acute exposure to progesterone: a randomized cross-over trial of 17-b estradiol, progesterone, and 17-b estradiol with progesterone in healthy menopausal women. J Clin Endocrinol Metab 85, 4644-9 (2000)
doi:10.1210/jc.85.12.4644

58. P.B. Rylance, M. Brincat, K. Lafferty, J.C. De Trafford, S. Brincat, V. Parsons, J.W. Studd: Natural progesterone and antihypertensive action. Br Med J 290, 13-4 (1985)
doi:10.1136/bmj.290.6461.13
PMid:3917316    PMCid:1415335

59. A.P. Cheung: Acute effects of estradiol and progesterone on insulin, lipids and lipoproteins in postmenopausal women: a pilot study. Maturitas 35, 45-50 (2000)
doi:10.1016/S0378-5122(00)00091-8

60. W.J. Gorgels, Y. Graaf, M.A. Blankenstein, H.J. Collette, D.W. Erkelens, J.D. Banga: Urinary sex hormone excretions in premenopausal women and coronary heart disease risk: a nested case-referent study in the DOM-cohort. J Clin Epidemiol 50(3), 275-81 (1997)
doi:10.1016/S0895-4356(96)00367-8

61. K.J. Chang, T.T.Y. Lee, G. Linares-Cruz, S. Fournier, B. de Lignieres: Influence of percutaneous administration of estradiol and progesterone on human breast epithelial cell cycle in vivo. Fertil Steril 63, 785-91 (1995)
PMid:7890063

62. J. Foidart, C. Collin, X. Denoo, J. Desreux, A. Belliard, S. Fournier, B. de Lignières: Estradiol and progesterone regulate the proliferation of human breast epithelial cells. Fertil Steril 5, 963-9 (1998)
doi:10.1016/S0015-0282(98)00042-9

63. R. Kaaks, F. Berrino, T. Key, S. Rinaldi, L. Dossus, C. Biessy, G. Secreto, P. Amiano, S. Bingham, H. Boeing, H.B. Bueno de Mesquita, J. Chang-Claude, F. Clavel-Chapelon, A. Fournier, C.H. van Gils, C.A. Gonzalez, A.B. Gurrea, E. Critselis, K.T. Khaw, V. Krogh, P.H. Lahmann, G. Nagel, A. Olsen, N.C. Onland-Moret, K. Overvad, D. Palli, S. Panico, P. Peeters, J.R. Quirós, A. Roddam, A. Thiebaut, A. Tjønneland, M.D. Chirlaque, A. Trichopoulou, D. Trichopoulos, R. Tumino, P. Vineis, T. Norat, P. Ferrari, N. Slimani, E. Riboli: Serum sex steroids in premenopausal women and breast cancer risk within the European Prospective Investigation into Cancer and Nutrition (EPIC). J Natl Cancer Inst 97(10), 755-65 (2005)
PMid:15900045

64. L. J. Warren Burhenne, L. Ken, and I. A. Olivotto: A new analysis of interval cancer in teh screening mammography program of British Columbia (SMPBC) based on five year age groupings. Radiology 217, 353. (2000.)

65. B.L. Riggs, L.J. Melton, R.A. Robb, J.J. Camp, E.J. Atkinson, L. McDaniel, S. Amin, P.A. Rouleau, S. Khosla: A population-based assessment of rates of bone loss at multiple skeletal sites: evidence for substantial trabecular bone loss in young adult women and men. J Bone Miner Res 23(2), 205-14 (2008)
doi:10.1359/jbmr.071020
PMid:17937534    PMCid:2665699

66. R. Recker, J. Lappe, K. Davies, R. Heaney: Characterization of perimenopausal bone loss: a prospective study. J Bone Min Res 15(10), 1965-73 (2000)
doi:10.1359/jbmr.2000.15.10.1965
PMid:11028449

67. V. Seifert-Klauss, T.M. Link, C. Heumann, P. Luppa, M. Haseitl, J. Rattenhuber, M. Kiechle: Influence of pattern on menopausal transition on the amount of trabecular bone loss. Results from a 6-year prospective longitudinal study. Maturitas 55, 317-24 (2006)
doi:10.1016/j.maturitas.2006.04.024
PMid:16781095

68. C. Berger, L. Langsetmo, L. Joseph, D.A. Hanley, S. Davison, R. Josse, N. Kreiger, A. Tenenhouse, D. Goltzman: Change in bone mineral density as a function of age in women and men and association with the use of antiresorptive agents. CMAJ 178, 1660-8 (2008)

69. C. Berger, L. Langsetmo, L. Joseph, D.A. Hanley, S. Davison, R.G. Josse, J.C. Prior, N. Kreiger, A. Tenenhouse, D. Goltzman, CaMos Research Group: Association between change in bone mineral density (BMD) and fragility fracture in women and men. J Bone Miner Res 2008.

70. J.C. Prior, Y.M. Vigna, M.T. Schechter, A.E. Burgess: Spinal bone loss and ovulatory disturbances. N Engl J Med 323, 1221-7 (1990)
PMid:2215605

71. E.J. Waugh, J. Polivy, R. Ridout, G.A. Hawker: A prospective investigation of the relations among cognitive dietary restraint, subclinical ovulatory disturbances, physical activity, and bone mass in healthy young women. Am J Clin Nutr 86(6), 1791-801 (2007)
PMid:18065600

72. J.L. Bedford, J.C. Prior, S.I. Barr: A prospective exploration of cognitive dietary restraint, subclinical ovulatory disturbances, cortisol and change in bone density over two years in healthy young women. J Clin Endocrinol Metab (2010)

73. M.E. Schmidmayr, A. Ehle, P. Luppa, M. Kiechle, V.R. Seifert-Klauss: Cycle characteristics in perimenopausal women with normal and osteopenic bone density. First results from the PeKnO (Perimenopausale Knochendichte Und Ovulation) study. Endocrine Society Abstract

74. J.C. Prior: Perimenopause lost - reframing the end of menstruation. J Reprod Infant Psychol 24(4), 323-35 (2006)
doi:10.1080/02646830600974071

Abbreviations: LH, luteinizing hormone; FSH, follicle stimulating hormone; QBT, quantitative basal temperature; QCT, quantitative computed tomography (for bone density); AMH, anti-Mullerian hormone; ACTH, adrenocorticotrophic hormone; PdG, pregnanediol glucuronide; E1G, estrone glucuronide; STRAW, Stages of Reproductive Aging Workshop; FREEDOM, Fertility Recognition Enabling Early Detection Of Menopause; BIMORA, Biodemographic Observations in Reproductive Aging; LOOP, luteal out of phase; VMS, vasomotor symptoms (hot flushes/flashes and night sweats); CaMOS, Canadian Multicentre Osteoporosis Study; BMI, body mass index = kg/m2

Key Words: Perimenopause, Estradiol, Progesterone, Inhibin B, LH, FSH, AMH, Menorrhagia, Cyclic VMS, LOOP, diagnosis, CVD, Breast Cancer, Endothelial Cancer, Osteoporosis, Menopause, Review

Send correspondence to: Jerilynn C Prior, Endocrinology and Metabolism, Department of Medicine, and Centre for Menstrual Cycle and Ovulation Research (CeMCOR), University of British Columbia, and Vancouver Coastal Health Research Institute, 2775 Laurel St, 4th floor, Vancouver, British Columbia, Canada V5Z 1M9, Tel: 604-875-5927, Fax, 604-875-5915, E-mail:jerilynn.prior@vch.ca