Abstract
‘In many women, the circulating levels of estrogen in the perimenopause may be higher than at any other time of her reproductive life when she was not pregnant.’
The perimenopause is usually defined as the time between when a woman notices a change in her menstrual function and the time at which she is amenorrheic for a year, hence, menopausal [1]. The usual beginning of the perimenopause is in the mid-to-late forties, at an age when reproductive function has usually ended. It is a relatively rare event that a woman in her mid-forties will become pregnant with her own gametes even with medical assistance [2]. Starting in her thirties, fertility decreases as a woman ages. This decrease in fecundity is not accompanied by any physical alteration that a woman could recognize. While there is a slight but measurable fall in fecundity, starting from age 30 relative to younger ages, at age 38 a steeper fecundity decline begins [3]. This decline increases through the early forties. By age 45, naturally occurring pregnancy is a rare event. By the early-to-mid forties the spontaneous miscarriage rate increases to over 50% [4]. There is a concomitant increase in karyotypic abnormalities. To view the changes a different way, a couple in their twenties having regular coitus has a greater than 20% chance of pregnancy in a given cycle. At age 40, this chance decreases to less than 5%.
The most common perimenopausal event that women notice is a change in menstrual cycle frequency, which includes somewhat shorter cycles or occasionally missed or longer cycles. The changes in cycle length are the defining events identifying the start of the perimenopause. There may also be a change in the character of the bleeding. In addition, many women specifically complain of breast discomfort or even enlargement during this time period.
It is well known that a woman is born with all the oocytes she will ever have. From birth through menopause these eggs are gradually lost, mostly through atresia, until several years after menopause, when there are no eggs left. At birth, there may be as many as a million eggs in the ovary. There are only roughly 400,000 eggs left at the time of puberty. From puberty to menopause, a woman ovulates approximately 400 times. However, by the perimenopause there may be only several thousand to hundreds or even only dozens of oocytes. There is a fairly constant loss of oocytes through a woman's reproductive years, even in the absence of ovulation. In other words, a woman who at age 20 begins taking oral contraceptive pills and stops them at age 50 will go through a natural menopause at a time not much different from a woman who ovulated monthly, even though she has not ovulated for 30 years.
Running out of eggs is clearly not the whole story in the causation of decreased fecundity and symptoms. There are many dynamic changes and alterations that are now just being discovered. One of the earliest changes that occur as one approaches late reproductive years is an increase in follicle-stimulating hormone (FSH). Unlike luteinizing hormone (LH), which appears to be controlled only by estrogen levels, FSH is controlled by many substances, including estrogen, the inhibins, activin, follistatin and progesterone. The inhibins decrease FSH secretion and activins stimulate FSH secretion [5]. Follistatin is an activin-binding protein, which affects functioning levels of activin. It is likely that the reason for the rise in FSH in the late reproductive years is a decrease in ovarian inhibin production. In contrast to FSH, which rises early in the late-reproductive years, LH changes little until the late perimenopause, when it also rises. Estrogen levels are low in the late perimenopause, in the year after the final menstrual period. For much of the menopausal transition, estrogen levels are the same as they were when the woman was younger. On many occasions, they may even be higher than observed previously. In many women, the circulating levels of estrogen in the perimenopause may be higher than at any other time of her reproductive life when she was not pregnant. The observation of higher circulating levels of estrogen during the late forties may have significant implication for a woman's health, because many of the problems that occur in women in their forties are estrogen related [6].
During the perimenopause, luteal phases are frequently shorter than in younger women. There are also lower levels of progesterone secretion during this timeframe; hence, the peri-menopause may be a time of luteal phase insufficiency [6].
Since study of hormonal mechanisms and changes in the perimenopause is a recent event, it should not be surprising that there is very little understanding of why these changes occur. While it is clear that throughout a woman's reproductive life there is a loss of follicles, this would not necessarily be an explanation for the changes that occur. In any given cycle, most of the steroid hormone production is by the single ovulatory follicle. It is by no means clear why this hormone productivity is affected by the remaining follicles in the ovary. What is clear is that in the very late perimenopause and menopause, there is a lack of responsive follicles. This results in an inability to stimulate follicular development or hormone production and results in a decrease in circulating ovarian hormones. This opens the feedback loops with the pituitary, causing dramatic rises of LH and FSH because of a relative absence of estrogen secretion [7].
For many women, the menopausal hot flash is a hallmark of the menopausal transition. It is responsible for facial flushing, sweats and social embarrassment. Hot flashes originate in the CNS and are mediated through the sympathetic nervous system. The flashes are associated with peripheral vasodilation, increased peripheral temperature and pulses of LH [8]. The causes of the flashes are unknown. It is unlikely that a lack of estrogen is the cause, since many women have high circulating estrogen levels at the time when they are having flashes. In many women, hot flashes occur well before the final menstrual period and the menopause. In some they occur even in the presence of hyperestrogenemia or normal estrogen levels. It is possible that the flash may be due to a relative insensitivity of the hypothalamus to the effects of estrogen. Even though estrogen is an effective therapy for hot flashes, a lack of estrogen is not the cause of a hot flash. Hot flashes occur when women are cycling normally and when their circulating levels of estrogen are equal to or higher than when they were younger and did not have these symptoms. By way of analogy, aspirin is therapeutic for headaches, but headaches are not due to aspirin deficiency. A greater likelihood is that the symptoms of perimenopause are due to hypothalamic dysfunction. This may, in turn, be due to abnormal secretion of second messengers or other factors that control hypothalamic function. What is needed is a clear understanding of the brain areas that result in the menopausal symptoms and enough of a physiologic understanding to develop specific therapies to deal with the symptoms without major effects in other areas, which is a problem with estrogen therapy.
Another major symptom of the menopause that can occur before the final menstrual period is sleep disturbance, characterized by a greater length of time to the onset of sleep and decreased rapid eye movement (REM) sleep. This is also a CNS event. Preliminary studies seem to indicate that the CNS in the perimenopause is less responsive to estrogen compared with responses in younger women. Hence, the conclusion is inescapable that not only are there ovarian changes associated with the perimenopausal transition, but that there are also nervous system changes or aging phenomena associated with the menopause. In other words, menopause is more than ovarian aging. There is also a CNS component that may be responsible for many of the symptoms that occur at this time of a woman's life. A greater understanding of neural control is needed, and this will be a rich area for future research.
‘The conclusion is inescapable that not only are there ovarian changes associated with the perimenopausal transition, but that there are also nervous system changes or aging phenomena associated with the menopause.’
Since the Women's Health Initiative data were released, it is clear that estrogen therapy is unacceptable to many women [9]. It is likely that therapies to deal with hot flashes and sleep disturbances will be directly targeted to CNS changes. Obviously, this will include agents such as the selective serotonin reuptake inhibitors (SSRIs) for hot flashes and benzodiazipines for sleep. New agents have side effects that make them unacceptable to many women. The challenge for the next 10 years is utilizing appropriate animal models, PET scanning and other technologies to develop a better understanding of brain and, specifically, hypothalamic function, and an understanding of the controlling factors responsible for the symptoms of the menopause and perimenopause. This understanding will clearly point to targeted and specific therapies without the general side effects and risks in currently available agents. Safe and specific agents are clearly the target for translational research in the next decade.
‘Most of these hysterectomies are performed for problems that arise secondary to hormonal changes, which occur in the years of the perimenopause.’
A major symptom of the perimenopause is recurrent abnormal bleeding or staining. Abnormal and acyclic bleeding has many causes. Abnormal bleeding may be increased and confounded by the high levels of estrogen that may be present or by the greater estrogen:progesterone ratio seen in the luteal phase. Greater acyclicity and higher estrogen levels may well be associated with endometrial hyperplasia, which occurs at this time in a woman's reproductive life, as well as its ultimate later sequella, endometrial adenocarcinoma [10]. Additional causes of bleeding at this time include fibroids and endometrial polyps. Endometrial polyps are hormonally sensitive and frequently result from acyclic hyperestrogenemia [11]. Thus, abnormal uterine bleeding is usually due to the hormonal function of the perimenopause. Not only are these symptoms unpleasant and inconvenient, but increased estrogen and anovulation can result in endometrial hyperplasia and, ultimately, atypical hyperplasia and endometrial cancer. The greatest risk for this complication is in women who have been chronically exposed to high steroid levels or unopposed estrogens [12], such as women with polycystic ovarian syndrome and obese women. The usual management of abnormal uterine bleeding in the perimenopause requires anatomic evaluation of the endometrial cavity. A hydrosonogram may help identify polyps or other endometrial lesions. Hysteroscopy and curettage are definitive diagnostic procedures. It is likely that the bleeding symptoms can be prevented and/or controlled by controlling the hormonal milieu if no malignant or anatomic lesion is present. Unfortunately, there have been few studies to define the best method of action to deal with this dysfunctional bleeding. Hormone therapy is an unlikely candidate, since its risks may be dramatically increased when given to women who already have normal or even higher levels of circulating estrogen. To produce a progestin-dominant environment and decrease the potential risks, many individuals have chosen to use cyclic low-dose oral contraceptives. This has proven to be helpful in some, but by no means all patients. The risks of this therapy in this age group have not been well defined. It is clear that it should not be available to women who smoke, because of the high cardiovascular risk. Physicians have frequently used cyclic progestin therapy on the theory that, if the woman is nonovulatory, this would be adequate to slough the endometrial lining and decrease the risk of potential problems of unopposed estrogen secretion. This may be helpful for some, but certainly not all women. The best clinical paradigms for management of abnormal bleeding in this reproductive timeframe have yet to be developed.
After cesarean section, hysterectomy is the most common major operation performed on women in the USA. Almost 600,000 are performed annually [13]. The decade of the menopausal transition is also the decade in which most hysterectomies are performed. It is likely that most of these hysterectomies are performed for problems that arise secondary to hormonal changes, which occur in the years of the peri-menopause.
Uterine fibroids are the most common benign tumors in women, affecting approximately 40% of women. A third of hysterectomies are performed for problems related to fibroids. Fibroids are hormonally responsive. They are stimulated by estrogens. They cease to develop and frequently resolve at menopause or after ovariectomy. While estrogen is a major growth factor for fibroids, there may be other hormones and growth factors. The presence of fibroids alone, unless they are very large, is not an adequate reason for surgical intervention. However, many symptoms require surgical therapy. These include interference with bladder or bowel function, persistent menorraghia, pain from bleeding into a fibroid or infarction or other unpleasant symptoms, including the symptoms which may be related to rapid growth or enlargement [14]. The alterations that occur in fibroids may be related to the estrogen alteration in the perimenopausal period. This is a recurrent theme. Control of the hormonal environment may well allow prevention of the pathological occurrences of the menopausal transition.
Almost 20% of all hysterectomies are due to endometriosis. It is well known that endometriomas are also not only steroid responsive, but responsive to other growth factors, which may be altered during the perimenopausal transition [15]. Hence, better understanding and control of the perimenopausal events may decrease endometriosis-related problems.
The perimenopausal period is a dynamic and frequently unpredictable phase in a woman's life. Changes occur and new symptoms may develop. These changes are due both to alterations in ovarian function and the responsiveness of the CNS to reproductive hormones. A clearer understanding of the physiologic changes in this phase of a woman's life may lead to more specific and preferable therapies to deal with the pathologies that may occur at this time. Development of endometrial abnormalities, such as hyperplasia and polyps, are due to unopposed and high levels of estrogen. The growth of fibroids and other conditions stimulated by the opposed estrogen clearly require further study. The development of uterus-specific selective estrogen-receptor modulators (SERMs), which decrease the adverse uterine effects of high levels of menopausal estrogens, would be a boon to perimenopausal women and dramatically decrease the hysterectomy rate as well as the rates of endometrial cancer. Another approach may be the development of safe methods to decrease estrogen secretion. These are long-term goals but well worth achieving since appropriate medical therapies would eliminate the need for the majority of hysterectomies, as well as decrease the unpleasant symptomatology and abnormal bleeding related to abnormalities of perimenopausal hormone secretion.
In summary, it is clear that the perimenopause is that phase in the reproductive spectrum of women most likely to be associated with unpleasant symptoms and unpredictable events. Cycle irregularities are the hallmark of the perimenopause. Hot flashes are more common in the perimenopause then in the post-menopausal period. Anovulation is associated with hyperestrogenemia and may result in fibroid growth, the growth of polyps and endometrial hyperplasia. Estrogen and hormone therapy have not been reasonably studied in the perimenopausal phase. There is at least the potential danger of adding estrogens to a hyperestrogenic state, which will further aggravate many of the problems of the perimenopause. There are, as yet, no specific therapies to deal with the CNS and, specifically, hypothalamic problems that may be the root cause of many of the symptoms. Therapy is therefore directed at symptoms rather than causes. What is needed is further research to reasonably design substances to deal with the causes of perimenopausal symptomatology. Success in these studies would be a true boon for women.