Addressing Postmenopausal Estrogen Deficiency: A Position Paper of the American Council on Science and Health

Sander Shapiro, MD

Abstract

During the past century, Medicine has made immense strides toward improving the general health, so much so that it now claims dominion over many events that were once perceived as immutably natural. In the case of postmenopausal estrogen replacement therapy (ERT), it is not clear that the public has entirely accepted Medicine's mandate. Despite cogent arguments for adopting a long-term regimen of prophylaxis for estrogen deficiency-related problems, relatively few women have received sustained treatment. Medical providers will be more effective advocates of long-term estrogen therapy by becoming more conversant with its effects and proficient in its application. This review aims to provide readers with a foundation in estrogen physiology and pharmacology that will enhance present therapeutic efforts and facilitate a critical appraisal of new regimens as they become available.

Etiology of Menopause

Natural menopause occurs when essentially all follicles are gone from a woman's ovaries. This eliminates the body's most productive estrogen source and results in a substantial decrease in circulating levels of the hormone. Without an adequate estrogen stimulus, endometrial proliferation then ceases and there is insufficient tissue to produce a slough from the uterus. In this way, the number of follicles that are available within the ovaries and the mechanism by which they are lost determine when menstrual activity ends. Clinically, amenorrhea of 6-12 months' duration during the fourth or fifth decade of life suggests the onset of menopause.

Every follicle that will ever appear in the ovary has formed by the midpoint of fetal life. Cells ordained to become oocytes (germ cells) have their origin in the early embryonic yolk sac. About 1000 of these cells migrate into the gonadal ridge tissue that is destined to form the ovary. There they undergo rapid mitoses reaching a maximum of 5-7 million by the fifth gestational month. Multiplication within this cohort of cells then stops, never to resume. Most oocytes acquire a halo of support cells capable of estrogen synthesis (granulosa cells) and form individual follicles. Those that do not organize into follicles undergo rapid atresia. Thereafter, a steady loss of primordial follicles occurs so that by birth the ovaries contain considerably fewer than were present at midgestation: about 1-2 million.[1] From birth onwards, this number continues to decrease independent of any hormonal influence (Figure 1). Throughout infancy and childhood, there is a steady loss of primordial follicles. This occurs through what currently appears to be a random, spontaneous initiation of folliculogenesis by individual primordial follicles followed by early atresia. The absence of circulating gonadotropins is probably what prevents maturation beyond the antral stage, dooming these follicles to an atretic course. Similarly, during the reproductive years, an unfavorable milieu sustains only limited maturation for most maturing follicles. Since only 400-500 follicles actually reach the point of ovulation over a reproductive lifetime, and there are essentially no follicles left at menopause, 99.9% undergo degenerative changes as a normal course.[2] This steady loss of ovarian follicles is not influenced by ordinary physiologic events.

The relationships between age and primordial follicle numbering using data from three studies as described by Richardson[4]: with permission.

During the interval between midgestation and puberty, primordial follicles continually enter the maturational process and are lost through early atresia. Once the hypothalamic-pituitary-ovarian axis becomes completely functional in late puberty, there is opportunity for a fraction of these developing follicles to become gonadotropin sensitive and, for a very few, to achieve ovulatory competence.[3] In adulthood, one follicle from each advanced cohort that becomes gonadotropin sensitive is recruited toward full maturity during a menstrual cycle. By about the seventh day of the cycle, it outstrips the rest of its mates, becomes more sensitive to follicle-stimulating hormone (FSH), grows larger by granulosa cell mitosis and antral fluid accumulation, and produces increasingly large amounts of 17beta-estradiol. The rise in circulating estrogen that results triggers pituitary release of a luteinizing hormone (LH) bolus which, in turn, induces ovulation and luteinization of the granulosa cell population in the mature follicle. Progesterone as well as estradiol is secreted by the corpus luteum. When it involutes, which it is destined to do if not rescued by pregnancy, levels of both steroid hormones fall abruptly (Figure 2). This concentration change induces menstruation and also provokes an increase in the release of FSH from the pituitary. As serum FSH rises, a new cycle of follicular recruitment is initiated. During the 30- or 40-year interval in which follicles mature, ovulate, and luteinize, relatively high levels of circulating estrogen are maintained. When, toward the end of the fifth decade of life, follicular maturation ceases and serum estrogen levels fall below those experienced during follicular cycles, menstruation ends. These low estrogen levels also evoke hot flashes, irritability, and genital atrophy -- signs and symptoms that have been recognized throughout history as being associated with the menopause. An appreciation that there are chronic biomolecular events that occur within various organ systems (central nervous, cardiovascular, skeletal) as a result of chronic estrogen deprivation is a much more recent development.

Estradiol and FSH levels in the serum of a woman during a normal ovulatory cycle (top panel) and over a lifetime (bottom panel).

In the last decade of menstrual activity, there is an accelerated rate of follicular depletion.[4] Ovulatory cycles become less frequent, menstrual shedding occurs after luteal phases in which only limited amounts of progesterone are produced, and intervals of early follicular development transpire without development of a mature follicle or ovulation. These cycles are accompanied by a gradual rise in serum FSH levels[5] (see Figure 2). Once menopause has occurred, FSH levels reach a plateau that is 10-20 times higher than that found during the early follicular phase (5-10 vs 50-150 mIU per milliliter). Serum estrogen levels slowly decrease during this interval, as do those of another granulosa cell product, inhibin.[6,7] The fall in estrogen and inhibin levels results from a decrease in total granulosa cell numbers. The rise in FSH is caused by absence of the inhibitory influences that these two agents exerted upon its production by the anterior pituitary.[8,9] This perimenopausal period lasts for 2-10 years and, in its later portion, is usually associated with a less regular pattern of menses. There is no clear endocrine marker of the final cycle. When 6-12 months elapse without flow, menopause has, by definition, occurred.

Testosterone and androstenedione are both produced by the ovary during the reproductive years. Conversion of androstenedione by peripheral tissue accounts for about half of the testosterone circulating at this time; the other half comes directly from the ovary and adrenal gland. When menopause occurs, there is a 50% reduction in serum androstenedione levels, hence a significant drop in the main precursor of peripherally produced testosterone. At the same time, there is actually a slight increase in ovarian testosterone production. This is probably due to high levels of gonadotropin, which stimulate ovarian stromal steroidogenesis. The overall result is that, after menopause, circulating testosterone levels are about two thirds of their premenopausal concentration.[10]

The median age of menopause in the United States today is 51.3 years, with a normal range of 47-55.[11] Women who are exceptionally thin, smoke cigarettes, experience prolonged malnutrition, or come from non-European ancestry tend to cease menstruation at a slightly earlier age; 3-4 years earlier at most.[12-14] Early menopause has also been observed among women who have had a unilateral oophorectomy. Parity and a history of irregular menstrual cycles are associated with a later age of menopause.[15] No other condition or treatment, including the use of oral contraceptive pills, has been consistently shown to foster a similar extension.[16] An individual's age at menarche is not predictive of her age at menopause.[15] Unlike the age of menarche, there is considerable question as to whether a change in the average age of menopause has occurred over historic intervals. The ancient Egyptians and Greeks recorded menopause at between 40 and 50 years, while records from medieval Europe most frequently cited 50 as the average age.[17,18] Several reports from mid-19th century industrial Europe give 45 as the most frequent menopausal age.[19] Thus, while there is no consensus, it would seem that over the last century, there has been a trend toward a later age for menopause.[20] More importantly, this century has seen a dramatic increase in the number and proportion of women living long beyond the age of menopause.

Medicalization of Menopause

Menopause, the natural, age-related cessation of cyclic menstruation, has been an acknowledged milestone in women's lives throughout recorded history.[21] In most times and cultures, it was viewed simply as a sign that a woman's term of reproductive competence had ended; in a few it had specific social/cultural significance.[22] Attitudes within the upper classes of industrializing countries began to change during the 19th century as certain behavioral characteristics became temporally associated with the cessation of menses.[23] What had once appeared as a singular event became conceptualized as a condition, a season, a state of mind and being.[24] Much of what was included in this state was considered as unwelcome and in need of remediation. Medicalization of the menopause was under way.[18] However, since almost nothing was known about the biochemical events underlying the normal menstrual cycle, little could be done for the menopausal woman beyond cataloguing the associated complaints and addressing them directly.[25] Purges, diet, and limited exercise were the clinicians' prescription in mid-19th century Europe and America.[25] The objective was to get each individual complainant through the "crisis" and into a more tranquil state.[26]

Simultaneous with medicalization of the menopause, the concept that "internal secretions" contribute to an individual's physiologic condition was gaining credence.[27] First the testes and then the ovaries were accorded a secretory function. Hormones that were their proposed products were assumed to have rejuvenating potential -- a theory consistent with the more magical attributes accorded to the gonads by various societies throughout history. By mid-19th century, this concept had taken sufficient hold that clinics dispensing gonadal tissues or their extracts appeared in Berlin and other European centers. The extent to which women attained relief from menopausal symptoms after imbibing these potions remains unclear. Some substantive success must have been realized, because the dispensing clinics flourished and commercial compounding firms worked toward producing more potent products.[28] At the beginning of the 20th century, extracts of ovary and corpus luteum, as well as dried tissue powders, had become part of the contemporary physician's pharmacopoeia. Therapy was directed toward perceived psychosocial problems as frequently as to the address of physical complaints.

During the first decades of the 20th century, physiologic modeling techniques (biologic assays) and pharmacologic extraction methods provided a means for quantifying the effects of ovarian products.[29] Nonetheless, substantial variability in the potency of many commercial ovarian preparations was being noted and decried as late as 1921.[30] In 1930, estrone became the first active ovarian product to be isolated in pure form and have its structure determined. Five years later, the more potent estrogen, 17beta-estradiol, was identified. These chemical advances, together with new insights into ovarian function, provided a framework upon which chemically observed paramenopausal events could be evaluated. Over the next 20 years, a substantial body of information accumulated that sustained the notion that many events associated with the onset of menopause were directly related to a decrease in circulating estrogens. Administering the newly available estrogens, clinicians began to identify and reverse many of the alterations brought about by the loss of ovarian secretory output. It then took only a slight extrapolation to conclude that the postmenopausal period was one of physiologic hormone deficiency that could benefit from replacement therapy. Medicalization of the menopause was now firmly grounded in scientific observation, and routine therapy seemed justified.[31] A growing awareness of the clinical benefits provided by estrogen replacement began to crystallize during the decade following midcentury. Yet, standard medical texts of the time suggested only limited-term, complaint-directed, therapy, an approach that had been advocated as early as 1897.[24,32] Only a relatively few physician investigators were recommending prophylactic, long-term estrogen therapy.

The book Feminine Forever became, in 1966, a watershed event in the public's perception of the menopause.[33] In it, Dr. Robert Wilson, a New York gynecologist, called attention to cardiovascular and bone mineral data that suggested long-term physiologic benefits could result from estrogen therapy. Seductively, he also suggested that a youthful appearance and attitude would result from estrogen therapy. These latter goals caught the public's attention, attracted readers, and propelled the book into multiple printings. An interest in estrogen replacement therapy (ERT) was now planted in the general public's mind, as well as that of the medical profession. Since the mid-1960s, there has been a gradual, albeit fitful, rise in the frequency with which long-term estrogen therapy is prescribed. Public and professional enthusiasm have been tempered by both experience and theory. The frequency with which women have encountered uterine bleeding while taking estrogens as replacement has led many to cease therapy. Fears that estrogens might provoke breast, uterine, or other cancers alarmed some physicians and many potential users, thereby dampening enthusiasm for the practice of ERT among large segments of the menopausal population. Additionally, concerns about the therapy's cost-effectiveness, when monitoring its use and potential adverse complications were taken into account, inhibited acceptance within the public health community. Over this same interval, a substantial change in the medical perception of menopause occurred. Whereas in the past it had been noted as a discrete, temporal event, it has increasingly come to be recognized as having a prelude and a prolonged epilogue. Recognition of both has been made possible by development of sensitive methods for hormone measurement and a more sophisticated understanding of the physiology associated with estrogen deprivation. Now it is recognized that the events that culminate in menopause have their beginnings at an earlier time, and that their influence extends through the remainder of each woman's life: hence, the growing medical appreciation of menopause as a heraldic event in the aging process.[34] Actual community medical practices with regard to menopause are frequently substantially discordant from those advocated in the recent medical literature. This difference will be narrowed as the public acquires a better understanding of menopausal physiology and the prophylactic benefits of therapy. Such knowledge can only provide succor to this population, which seems destined to extend its life span even further than the present mean of almost 80 years.

Physiology of Reduced Estrogen Production

The most common perceivable and physiologically detectable events associated with onset of menopause are the occurrence of hot flashes and body sweats. These vasomotor complaints may begin several years before menopause and usually last for 2-5 years beyond that landmark event. Infrequently, they are experienced by women over 60 years of age.[35] As many as 70% of American women experience these symptoms, to some degree, while only 20% of Japanese women report them.[11,36,37] Most European surveys on incidence produce rates similar to those reported in the United States.[38,39] The rather substantial difference between surveys from Eastern and Western countries remains unexplained. Cultural influences on perceptions and attitudes as well as biologic factors such as diet have been offered as possible reasons for the discrepancy.[40]

The characteristics of a hot flash have been described heterogeneously and are quite variable from individual to individual.[36] Generally, flashes are experienced as transient warm feelings that start on the upper body or face, spread centripetally, and last less than 2 minutes. They are accompanied by tachycardia and may be immediately followed by perspiration that is sometimes profuse. Episodes are usually more frequent during waking hours but can awaken a person multiple times each night. A pituitary luteinizing hormone pulse accompanies each flash.[41] Objective findings during each episode include a drop in skin electrical resistance, an increase in peripheral blood flow, and a rise in skin temperature. Peripheral vasodilatation and sweating often lead to a fall in core temperature with subsequent vasoconstriction, chills, and shivering.[42] These events have caused physiologists to suspect that both labile hypothalamic thermoregulatory and peripheral vascular response mechanisms are at the etiologic center of this phenomenon.[43,44] The precipitating event that affects the hypothalamus is estrogen withdrawal. Appearance and extent of symptoms is directly related to the serum concentration of postmenopausal estrogen.[45] Age of the individual does not relate to incidence of this type of vasomotor instability, as young women who are oophorectomized experience these symptoms at about the same rate as older women.[46]

Recognition of the temporal relationship between menopause and ovarian atresia is of long standing. It formed the basis for the early 20th century hypothesis that gradually falling levels of an ovarian product might be the primary initiator of vasomotor flushing. Supporting this belief was the observation that an increase in frequency and intensity of hot flashes resulted from an abrupt loss of the ovarian agent, as happens following oophorectomy.[47] This hypothesis was finally sustained when pure samples of various estrogens became available and were shown to abolish menopausal vasomotor symptoms. Additional support for the central role of estrogen withdrawal came from the observation that prepubertal girls and adult women with ovarian agenesis, both of whom have low serum estrogen levels, are unencumbered by hot flashes. Further evidence is provided by the observation that once exposed to estrogen, women with ovarian agenesis experienced hot flashes following cessation of administration.[48]

More recently, the mechanistic role of LH pulses as a requirement for initiation of each hot flash has been pondered. The finding that flashes continue to occur in women whose capacity to secrete LH has been inhibited by long-acting gonadotropin agonist administration has established that this pituitary agent does not trigger flashes. Similar findings have been noted in hypophysectomized women. Mechanistic studies have also substantially ruled out menopausally elevated levels of FSH and the absence of inhibin as primary initiators of menopausal vasomotor instability.[49] Hence, the cellular site and underlying mechanisms that cause hot flashes remain undetermined.

Prospective studies of perimenopausal women do not provide data to demonstrate that they experience an increase in psychological or physical problems other than vasomotor complaints and diminished vaginal lubrication.[11,50] Convincing evidence has not been developed to support claims for the occurrence of more frequent or intense psychiatric problems during this period.[51] The oft enunciated 19th century view that menopause is associated with nervous irritability and hysterical states does not bear up to scrutiny through comparative testing.[19,52] However, when vasomotor complaints become severe, there can be sufficient sleep deprivation to induce irritability, tiredness, and difficulty with concentration.[53] Also, test instruments designed to detect clinical depression have been found to provide mean numerical scores among perimenopausal women that, while not consistent with overt depression, do decrease significantly after estrogen treatment.[54]

Estrogen replacement has been demonstrated to be an effective treatment for hot flashes.[36,55] The reduction in frequency and severity is time dependent, requiring from 4-6 weeks for maximum effect. After this length of therapy, alterations in quantifiable sleep parameters can also be expected. The oral dose of conjugated estrogen that is most frequently effective is 0.625 mg/day. Higher doses are sometimes required. However, when 1.25 mg is inadequate, the initial diagnosis must be questioned. There is rarely a substantive difference in response when other routes (transdermal, intramuscular) or estrogen formulations are prescribed.

Almost contemporaneous with the onset of flushing, some women begin to experience dyspareunia and vaginitis. These signs are thought to result from a decrease in vaginal lubrication and a change in pH.[7] Later events include restricted blood flow, epithelial atrophy, and loss of connective tissue elasticity. Since sexual interest can remain high (over 50% of peri- and postmenopausal women) these physiologic changes are unwelcome.[56] Preventing their appearance or inducing their reversal can usually be accomplished by administration of the same systemic estrogen doses as are given for vasomotor symptoms.[57] When direct vaginal application of estrogen cream is undertaken, therapeutic serum levels of the hormone are usually attained. Local estrogen delivery systems have recently become commercially available for the treatment of vaginal atrophy. They provide sustained hormone release without causing an appreciable rise in serum estrogen concentration or the induction of systemic events.

Estrogens contribute to the maintenance of many tissues and their metabolic activities. When estrogen levels fall toward the end of the perimenopause, two distinguishable kinds of change take place. One includes a group of rather abrupt and directly observable events, including vasomotor flushing. The other comprises a larger, diverse group of more slowly developing processes whose observable consequences reach threshold status only years after ovarian estrogen output has fallen. Outwardly, the late changes seem indistinguishable from other degenerative events related to aging. In the past, there was nothing that directly connected them to the temporally distant menopause. However, by mid-20th century, their collective relationship to an absence of estrogen stimulation had become evident. Simultaneously, the sequelae of these physiologic degenerative processes began to draw clinical attention. This conjunction fostered the concept that postmenopausal women were in a state of chronic estrogen deficiency which, in turn, provided the rationale for long-term ERT. Those organ systems that are considered to be most profoundly affected by limited postmenopausal estrogen levels are the cardiovascular, skeletal, genitourinary, and central nervous system (CNS).

Cardiovascular Disease. Cardiovascular disease is the most common cause of death among women in the United States.[58] Deaths due to coronary heart disease are less frequent than in men at all ages and are extremely infrequent before the advent of menopause.[59] Postmenopausal coronary heart disease rises in women, paralleling the increase seen in men as they age. Women who undergo premature menopause are at greater coronary risk than menstruating women of similar age.[60] These observations suggest that estrogens play a significant role in delaying the onset of cardiovascular disease in women. They also support the hypothesis that by replacing estrogens after menopause, substantial reductions in the incidence of cardiovascular disease can be obtained. A number of specific direct and indirect estrogen effects are thought to potentially contribute to this favorable outcome.[61]

Estrogens exert a major influence on lipoprotein metabolism. Hepatocytes produce more high-density lipoprotein (HDL) and catabolize more low-density lipoprotein (LDL) when under estrogen stimulation. This induces a serum lipid profile that is less favorable to atheroma formation. When estrogen levels fall the ratio of HDL to LDL changes, adversely resulting in an increased rate of atherosclerosis production.[62] Estrogen administration after menopause will maintain a favorable lipoprotein profile.[63] ERT can alter hepatic protein production profiles that may then influence cardiovascular risk by changing intravascular coagulation characteristics. In addition, estrogens have several direct effects on blood vessel metabolism that are favorable.[61,63,64] These include their antioxidant activity, which protects endothelial cells from injury, and their capacity to inhibit myointimal proliferation after vascular injury.[65] Numerous other estrogen-induced effects, some that involve nonclassical (nongenomic) steroid mechanisms, have also been proposed.[62,66] All are potential inhibitors of cardiovascular disease. However, the fractional contribution of these mechanisms to the overall result brought on by estrogen is unknown.[67] At present, conjecture places one third of an estrogen's favorable overall effect on cardiovascular disease in its influence on lipoprotein profiles.[60,68,69]

Despite evidence that estrogens have beneficial effects on arterial health and the wide use of these steroids to inhibit the development of coronary heart disease, therapeutic efficacy remains in question. Almost every observational study published to date has found a lower risk of coronary heart disease in postmenopausal women who were taking ERT as compared with those who were not.[70] In aggregate, these findings suggest that estrogen replacement is responsible for a reduction in the relative risk of death from coronary heart disease to 0.70, with some individual studies doing considerably better.[68] Studies that included a progestin in the therapeutic regimen produced similar reductions. The large number and variable design used in these studies make it unlikely that these results were produced by some unidentified confounding factor.[71] Nevertheless, controversy remains primarily because of the limited amount of information available from randomized, prospective studies.

Recently, skepticism has been heightened by findings from the only large-scale, blinded, placebo-controlled trial yet to be published.[72] In that study, women with preexisting acute coronary disease were put on an estrogen-progestin or placebo regimen for 4 years. Overall, the number of coronary events in the two study limbs was the same. However, there was an increase in the number of clinical adverse events among estrogen users during the trial's first year. Furthermore, a prospective, randomized trial that employed sequential coronary angiography to evaluate women with documented coronary artery disease failed to demonstrate any substantive differences between those who were or were not on an HRT regimen.[73] Thus, while these results seem contrary to those of several nonrandomized trials and may not be indicative of future results, they do provide a note of caution.[58,74]

Osteoporosis. Osteoporosis is defined as a loss of normal bone micro-architecture and density that leads to structural fragility. The cellular processes that cause this deterioration are intrinsic to the general process of aging. In women, a more rapid diminution of bone mineral density is superimposed upon this slow loss during the 5-10 years immediately following menopause.[75] Among persons of limited peak bone density, age-related bone loss results in a greater risk of fracture -- common sites being vertebral bodies, wrists, and hips. Clinical measurements of bone density have established a threshold for substantially increased fracture risk at 2.5 standard deviations below the mean density found in young adults.[76] Women are more likely to cross this threshold then men, accounting for 80% of all hip fractures. Women who are big boned, darkly pigmented, or have higher postmenopausal estrogen levels are less likely to experience fractures.[77] Cigarette smokers and those who have had a limited calcium intake during their reproductive years are more susceptible.

Osteoporosis plays a significant role in most of the 1.5 million annual fractures that occur in the United States.[78] It was the first chronic medical problem recognized to result from long-term estrogen deficiency.[79] In recent years, the identification of estrogen receptors within bone cells and the realization that estrogens play an influential role in cellular bone remodeling activity have begun to provide a basic understanding of the mechanisms by which postmenopausal women rapidly lose bone density.[80]

Chronic estrogen administration during and following menopause has been shown to substantially reduce the rate of osteoporotic bone loss and decrease the incidence of both hip and vertebral fractures by about 50%.[81] ERT can also improve bone density measurements in osteoporotic women.[82] However, once estrogen administration ceases, the rate of bone mineral density decline returns to that of untreated women during the immediate postmenopausal period. Treatment intervals, routes of administration, and dosages have varied among the many favorably reported studies. On average, it takes from 5-10 years of ERT before a significant difference in fracture incidence can be demonstrated.[83] An improvement in bone density can be demonstrated after as short a therapy interval as 6 months. The effective dose of conjugated estrogens has been found to be as low as 0.3 mg/day.[84] However, most studies have found a dose of 0.625 mg to be more widely effective. Other forms of oral estrogen seem to be equally efficacious and, because the action is directly upon the target tissue, route of administration does not seem to matter.[85] Since little, if any, protection from fracture is gained with short-term estrogen replacement and because rapid bone deterioration resumes after estrogen withdrawal, therapy should be planned as a long-term activity. Moreover, once substantial loss of trabecular micro-architecture has occurred, estrogen-induced stabilization and/or increased bone density are thought less likely to diminish fracture risk.[86,87]

Genitourinary Disease. Hypoestrogenism during the perimenopause can result in the rather abrupt onset of dyspareunia. This is thought to be primarily the consequence of decreased vaginal secretions and the resultant lack of coital lubrication. Over longer intervals, more profound alterations in genitourinary anatomy and physiology occur. Studies in laboratory animals have demonstrated genitourinary sensitivity to estrogens. Estrogen receptors have been identified in the human urinary tract as well as in the various tissues of the internal genitalia.[88] After menopause, atrophic changes gradually occur in the urethra, periurethral connective tissue, and genital mucosa. A correlation has been shown between atrophic urogenital conditions and decreased sexual well being.[89] These changes are associated with urinary incontinence in more than 15% of the non-nursing home population who are over 60 years of age.[90] At present, however, the degree to which estrogen deficiency changes contribute to either urge or stress incontinence is unknown. Clearly, estrogens have a direct effect on urogenital tissues and are capable of inducing measurable improvements in several urodynamic parameters. Definitive studies that might establish the relative importance of estrogen deficiency among the several known causative factors for incontinence are lacking. Only the positive experiences from estrogen replacement studies are available to support the contention that a hormone deficiency contributes to age-related genitourinary symptomatology.

Meta-analysis of the available studies addressing estrogen therapy (as of 1994) for stress incontinence shows that while significant subjective improvement was found, a decrease in the quantity of urine loss was not demonstrable.[91] More recently, a single, large, controlled study of stress incontinence (inadvertent urine loss following an abrupt increase in intra-abdominal pressure) failed to obtain positive results from estrogen therapy.[92] On the other hand, studies evaluating estrogen therapy for urge incontinence (inadvertent urine loss associated with a sensation of voiding urgency) have been slightly more positive.[93] This has led one observer to suggest that the benefits of estrogen in some stress incontinence trials have actually been due to the presence of a mixed-type incontinence among members of the study groups.[94] A more substantive positive claim can be made for inhibition of recurrent cystitis and improved sexual function in postmenopausal women. Here, evidence strongly supports the use of either local or systemic estrogen therapy.[95]

CNS Deterioration. Among the faculties affected by aging, mentation stands out. Only minor quantifiable deterioration is experienced during middle age, while more severe decrements frequently occur in those who are over age 65.[96] Memory and learning capacity normally degenerate modestly in old age. When the decline is severe enough to interfere with daily affairs it is called dementia. This form of incapacity becomes increasingly prevalent after 70, and by age 80, new cases appear in the population with the same frequency as myocardial infarction.[97,98] Alzheimer's disease accounts for two thirds of all dementia, occurring in about 40% of the population that is over 80 years of age.[97] An association between estrogens, menopause, and cognition has long been postulated.[99] However, relatively little attention was directed toward the influence of estrogens on CNS activity until the last 2 decades. During this interval, persuasive evidence has been developed suggesting that estrogens promote neural cell growth, alter CNS function, and protect the system against toxins.

Work on ovariectomized rodents has demonstrated that estrogens enhance synapse formation within the cerebellum, hippocampus, and other CNS structures.[100,101] These synapses have been shown to actively transmit neural signals.[102] In the forebrain, estrogens stimulate production of choline acetyltransferase, which promotes production of the neurotransmitter acetylcholine.[103] A deficit in acetylcholine reduces memory, while estrogens can raise levels of the transmitter and prevent cognitive loss. Estrogens also act as an antioxidant, reducing the toxicity of several free radical producers; one of which, beta amyloid, accumulates in the brain of Alzheimer's patients.[104,105] In addition, estrogens have been shown to enhance synaptic sprouting through an apolipoprotein-dependent mechanism.[106]

These biochemical/molecular findings have provided a biological rationale upon which to explore the possible relationship between postmenopausal estrogen therapy and cognition. Two provocative questions have come under clinical investigation: Can ERT influence the modest general decline in age-related mental function? and Can ERT inhibit the onset and/or the progression of Alzheimer's dementia? Definitive answers are not currently available, but initial observations have shown sufficient promise to encourage more rigorous, ongoing study.

If estrogen replacement has any effect on the modest, age-related declination in women's cognition, it is likely to be in specific functional areas. This is because some components of cognition, such as primary and remote memory, do not show any change with age, while others, such as new information recall, do decline.[107,108] From the limited studies carried out to date, there is only modest evidence to suggest ERT improves specific types of cognitive performance; principally verbal learning.[109] However, methodologic problems and a diversity of findings among the available studies have led to the recommendation that more rigorous investigations be carried out.[110,111]

A small group of studies that addressed cognitive function in women already affected by Alzheimer's disease showed consistent positive evidence for an estrogen effect. These efforts suggested that estrogen replacement in standard therapeutic doses can improve verbal IQ, comprehension, memory, and some but not all other measures of dementia.[110,112] A recent, larger study, however, failed to support the hypothesis that 1 year of estrogen administration is effective in limiting the progression of mild to moderate Alzheimer's disease.[113]

More substantive evidence can be mustered for estrogens' capacity to inhibit the onset of Alzheimer's disease.[110] In both case control and cohort studies, the relative risk of developing Alzheimer's disease was between 0.3 and 0.7, suggesting that a reduced incidence of 50% might be obtained with long-term ERT.[110,114] In one of these studies, evidence was accumulated that showed a greater effect when estrogen dosage was in the high rather than low therapeutic range.[115] The efficacy of progestin or androgen addition to estrogen therapy has not been determined for either age-related cognition or Alzheimer's dementia.

Sexual Drive. The relative extent to which cultural and biological factors influence sexual activity is not well established. Libido, a significant component of sexual activity, is thought to be primarily a cognitive phenomenon. In women it begins to wane during the early perimenopause, decreasing more rapidly some years after menopause. Nevertheless, most healthy 70-year-olds engage in sexual intercourse, and a substantial proportion of postmenopausal women express regret at their loss of libido.[116,117] Studies of postmenopausal women have shown that estrogen-reversible, atrophic genital tract changes substantially influence sexual activity but that ERT does not alter libido. Premenopausal women lose libido rapidly after oophorectomy. Reversal does not come from estrogen replacement but does when testosterone is provided.[118] In these patients, androgen administration produced an increase in sexual fantasies, sexual desire, and arousal during intercourse but did not alter the number of orgasms or the frequency of intercourse.[118,119] Thus, by extrapolation one might suppose that androgens play a major role in postmenopausal loss of libido. This hypothesis remains in question, however, because ovarian testosterone production and serum testosterone concentrations do not fall substantially in the immediate postmenopausal period.[120] Only a small number of postmenopausal women with intact ovaries have been evaluated for an androgen effect on libido.[121] These studies have not, unambiguously, supported the many positive anecdotal experiences of clinicians in the field.[117] This lack of support for androgen therapy is important, as the commonly used therapeutic doses (1.25-2.5 mg/day) are capable of adversely altering estrogen-induced lipid profiles during ERT.

Uterus. Estrogens induce hypertrophy of the myometrium and proliferation of the endometrium. With chronic exposure, the endometrium becomes prone to carcinomatous degeneration.[122] Reduction of endometrial mitotic activity occurs during the luteal phase of each ovulatory cycle as a result of progesterone exposure. This progesterone effect is the mechanism by which the estrogen-driven propensity toward neoplasia is held in check during the reproductive years. Recognition, during the early 1970s, that a substantial increase in endometrial cancer accrued (relative risk, 2-8) to women either currently or previously on estrogen-only replacement regimens dampened enthusiasm for chronic postmenopausal ERT.[123] The observation that iatrogenic estrogen-induced malignancy was more likely to be identified at an early stage failed to mollify.[124] Addition of a progestin to ERT regimens became almost mandatory for prophylaxis in those women possessing a uterus. This scheme proved effective in limiting the incidence of endometrial cancer and may actually have reduced its rate below that found in non-estrogen users.[125] Addition of a progestin (usually medroxyprogesterone acetate, 10 mg/day) for 12 or more days of each cycle effectively lowered the epithelial proliferation rate to that of the midluteal phase, but it also caused cyclic bleeding in at least 50% of those taking ERT.[126,127] Since this was an inconvenience and a major reason that many women abandoned therapy, new protocols incorporating progestins came into vogue (see Figure 3). Today, almost all postmenopausal women taking estrogen also take a progestin; ERT has become HRT except for those women who are ex-uteri.

Protocols for administering estrogen and progestin during HRT: the light boxes represent days of estrogen intake; the dark boxes represent days of progestin intake.

Venous System. Ever since epidemiologic studies suggested a relationship between the amount of estrogen in oral contraceptive pills and an increase in the rate of venous thrombosis, there has been concern that a similar effect might result from postmenopausal ERT. The mechanisms by which oral contraceptives increase venous thrombosis are not entirely clear, but it is generally believed that the effect is mediated by estrogen-induced changes in the metabolism of multiple plasma coagulation factors. Despite substantially lower estrogen doses during postmenopausal therapy than are used for contraception, an increase in the rate of venous thrombosis does appear to exist.[128,129] The relative risk has been found to increase with the amount of orally administered hormone (using conjugated estrogen: 0.3 mg, RR = 2.1; 0.625 mg, RR = 3.3; 1.25 mg, RR = 6.9).[128] Nevertheless, postmenopausal baseline rates of venous thrombosis are so low (1.3 per 100,000 women years) that these occurrences have little impact on overall mortality rates.

Because estrogens' overall effect on coagulation is most likely the cumulative result of many metabolic alterations in hepatic protein metabolism, use of transdermal rather than oral hormone has been suggested as a means of limiting direct liver stimulation and thereby minimizing coagulation problems. However, when the "first pass" effect of orally administered hormones on hepatic coagulation factors has been evaluated and compared with routes that initially bypass the liver, uniform alterations have not been found[130-132]; nor has data accumulated to show that avoidance of an oral dose, first-pass effect can prevent the increase in venous thrombosis during ERT. Further study is required in order to ascertain whether avoiding first-pass hepatic exposure can limit venous thrombosis. Coincidently, the degree to which absence of a first-pass effect might diminish those benefits derived from altered HDL/LDL concentrations will need to be determined. Furthermore, it is not clear whether the addition of progestins to estrogen replacement regimens influences venous thrombosis rates.

Colon. There have been a number of observational studies suggesting that estrogen inhibits the development of colon cancer among postmenopausal women.[133,134] The relative risk for estrogen users, as compared with nonusers, in these studies varies from 0.5 to 0.7 for women of the same age. For past users, rates returned to those of nonusers 5 years after discontinuing estrogen intake.[135] Because the large bowel is not ordinarily thought of as an estrogen target, it has been suggested that altered hepatic bile output might be the mechanism by which cancer induction is inhibited.[136] However, there is growing evidence that large-bowel mucosa actually is an estrogen-sensitive tissue. Steroid hormone receptors have been found in normal colonic mucosa as well as in colonic adenocarcinoma, and some physiologic activities of the colon are known to be modified during the menstrual cycle.[137,138] With women accounting for almost half of the 100,000 new colon cancers appearing in the United States each year, a potential reduction in incidence of 50% would have a considerable impact on the overall mortality statistics of postmenopausal women. But, because the vitality of this estrogen action diminishes rapidly with cessation of therapy, chronic administration would be required in order to obtain full benefit. Prospective, randomized trials are needed to confirm the observational evidence of this estrogen-produced benefit and to establish its effect on mortality rates. Trials that include nonoral routes of administration will also be needed if the impact of first-pass hepatic influences is to be determined.

Breast. Biologists became aware that breast tissue was sensitive to estrogen almost simultaneously with their recognition that estrogens influence genital tract tissues. Shortly thereafter, estrogens' tumorigenic influence on breast tissue was established and its role in maintaining neoplastic growth determined. This information induced oncologists to attempt therapeutic manipulations of systemic estrogen levels in women with breast cancer. It also stimulated epidemiologists to investigate the influence that estrogen exposure might have on breast cancer incidence. Resulting investigations showed that women exposed to intrinsic estrogens for longer time periods, as is the case with early onset of menarche or late onset of menopause, experienced higher rates of breast cancer.[139,140] Additionally, it was noted that there was a decrease in the breast cancer incidence rate following menopause, no matter what the age of onset, and suggested that this decrease could be a result of the coincident fall in serum estrogen concentrations.[141] All of these findings suggest that an excess of exposure to estrogens during the woman's ovulatory years is associated with an increased risk of breast cancer.

Despite these and many other observations linking estrogen to breast metabolism and neoplasia, the relationship between estrogen use in postmenopausal therapy and breast cancer remains unclear. To date there are only limited, conflicting data sets concerning whether short-term ERT (< 5 years) alters the incidence of breast cancer. For longer therapeutic intervals, the study data, which are derived almost exclusively from observational investigations, are less ambiguous. The results of these studies lack consistency, with some showing no difference in incidence while others demonstrate increases in relative risk and a few contrarily suggest the existence of a protective effect. However, analysis of 51 epidemiologic studies involving 52,000 women with and 110,000 women without breast cancer sheds substantial light on the issue.[142] The evaluators, while recognizing the potential for selection bias in these studies, concluded that there was an increase in relative risk when women between 50 and 70 years of age took ERT for more than 5 years. They estimated that after 15 years of therapy, the risk was about 1.25 times that for nonusers: an increase that would add 12 cases per 1000 women to the base rate of 45 cases per thousand over the same time interval.

Some evidence suggests that the addition of a progestin to estrogen may confer a greater increase in breast cancer risk than the use of estrogen alone.[142,143] In a recent case-control study, combination HRT was associated with a 24% increase in breast cancer risk over nonusers (statistically significant), whereas unopposed estrogen was associated with a 6% increase that was not statistically significant.[144]

This increase may, however, not be of a comparable type disease to that found in women who are not taking estrogens.[145] For example, breast cancer with a favorable histology has been found to be 2.65 times more likely among HRT users than nonusers.[146] Fairly consistent evidence has been developed showing that the prognosis for neoplastic breast disease contracted while on ERT is better than that for disease contracted without estrogen use.[146-148] Consistent data exploring whether route, amount, and type of estrogen administered alter these incidence figures have not been developed. There is also a dearth of information to support the contention that women with a family history of breast cancer experience a greater change in relative risk by taking estrogen than do women without a positive family history.

Once a woman stops ERT, her relative risk appears to return to that of a nonuser within 2-5 years.[142] More conclusive information about the impact of ERT on breast cancer incidence presently does not exist. When, in about 5 years, data from randomized trials become available, a clearer picture of this critical aspect of ERT may emerge. At present, most authoritative analysts favor the concept that chronic estrogen therapy probably does result in a slight increase in the relative risk of breast cancer.[149-151] Their individual interpretations of the significance that this increase imparts to the therapy's overall clinical value vary immensely. Because breast cancer affects a large fraction of the female population, small changes in its incidence will produce a large change in the number contracting the disease. Thus, an individual's (investigator, clinician, patient) view of breast cancer's medical impact will substantially influence their appreciation for the overall value that may be gained from ERT.[152]

Survivors of breast cancer experience the same estrogen deficiency problems as other postmenopausal women. Their rising numbers make the question of efficacy for postcancer ERT extremely cogent. Unfortunately, convincing study data that either support or denigrate the use of estrogen in the post-breast cancer patient are not available. Nor are there any studies that demonstrate a worse prognosis for women who take estrogens after breast cancer.[153]

Is Estrogen Replacement Efficacious?

At midcentury, postmenopausal estrogen therapy seemed to promise great medical benefits. Fifty years later, enthusiasm for short-term estrogen therapy remains, while that for chronic replacement has become somewhat dampened. Epidemiologic data support a consensus view that, on balance, estrogen therapy provides substantial relief to those women who experience mild to severe vasomotor symptomatology. Using modest doses with or without a concomitant progestin can essentially eliminate hot flashes and their sequelae. Subsequently, usually after several years and regardless of the regimen selected, the estrogen can be gradually withdrawn without provoking symptoms of a similar magnitude. The financial and adverse health costs of such a protocol are minimal. For those women with intact uteri, concomitant progestin therapy is advantageous but can cause vaginal bleeding. Otherwise, short-term estrogen does not seem to impose negative consequences other than the very slight risk of venous thrombosis. There is, for instance, essentially no evidence to suggest that when estrogen is taken for less than 5 years, the risk of developing breast cancer is increased. Along this same line, temporally limited estrogen therapy is not associated with any of the recognized benefits bestowed by long-term therapy.

A consensus about the relative merit of chronic estrogen administration has not developed. In part this is because of the multiple target tissues involved, the several organ-specific outcomes to be evaluated, and an absence of available data from controlled, randomized trials. In order to evaluate the aggregate effect of long-term estrogen therapy, analysts have tended to rely on mortality and life expectancy estimates as a means of quantifying therapeutic results. Using relative risk figures for specific diseases and the altered mortality estimates that are implied, various evaluators have made quantitative estimates of the changes in life expectancy that might result from adopting specific hormone replacement regimens.

Available data suggest that estrogen administration reduces the risk of coronary heart disease by 30% to 50% when therapy is carried on for 10 years. The extent of benefit that occurs when a progestin is added to the regimen remains unclear. Most progestins adversely alter serum lipoproteins while some, including medroxyprogesterone acetate (Provera), also influence additional vascular events in a negative manner.[154,155] These negative actions would not occur in posthysterectomy estrogen users but might substantially reduce the benefits for those requiring progestins as a means of limiting the risk of estrogen-induced endometrial hyperplasia and neoplasia. Neither the reduction in hip fractures that result from estrogen intake nor the associated decrease in mortality are altered by the addition of progestins.

ERT also modifies mortality rates by influencing the development of certain cancers: increasing the rate for breast cancer and decreasing the rate for colon cancer. Suspicion also exists that progestins increase the relative risk of cancer but with almost no supporting data.[147] Nevertheless, some analysts have factored this concern into their estimations of breast cancer incidence. The same kind of progestational effect has not been invoked for colon cancer. Combining all of these estimates, epidemiologists find that ERT in posthysterectomy women probably adds 1 or 2 years to life expectancy.[156] For women with intact uteri who must also take a progestin, the estimated extension is zero to half a year.[157,158] These figures have inspired some enthusiasm for ERT but considerable skepticism for combined estrogen and progestin therapy, ie, HRT.[159] Authorities decry the lack of solid data upon which these estimates and the therapeutic recommendations described below are made. Nevertheless, they continue to state their recommendation emphatically. The practitioner and potential user of the therapy are then left to make their own value judgements. Interestingly, as a group, female physicians and especially gynecologists are more apt to personally undertake HRT than is the general public.[160-162]

Quantitative mortality estimates are helpful when considering the value of long-term therapy, but there are other factors that also should be included in such an evaluation. Since a decision to recommend HRT as public policy would have financial consequences, the cost of therapy and surveillance as well as the projected saving from a lowered mortality rate need to be considered. When such an analysis was undertaken in Great Britain, ERT appeared to offer "good value," while that for combined hormone therapy was found more difficult to evaluate and to possess questionable value.[163]

The influence that replacement estrogens exert on the quality of a user's life should be part of a global appraisal. Nonquantifiable benefits, such as improved cognition and delayed onset of Alzheimer's disease, become important considerations for a population that can be expected to live for more then 80 years. The long-term impact of femoral fractures on both mobility and independent living is also of considerable consequence. Finally, because so much of the negative impact on mortality estimates is influenced by assumptions about progestin effects, thought must be given to the use of progestins that do not inhibit the positive cardiovascular effects of estrogen. Such a progestin might be progesterone itself. While not yet substantiated by clinical outcome studies, there is reason to assume that progesterone will not evoke the hypothetical negative cardiovascular and breast changes attributed to Provera and the 19-nortestosterone compounds. Should this turn out to be the case, a major potential liability of long-term therapy would have been removed. As our society comes to understand the implications inherent in having large segments of the population live a full life span, interest in the nonquantifiable benefits of chronic estrogen therapy should increase.[164] More emphasis will then be placed on extending the interval that a person experiences good health by postponing the onset of age-related diseases.[165]

Bimolecular Action of Estrogens and Progesterones

Ever since the structural identification of estrone and 17beta-estradiol in the 1930s, estrogens have been thought of structurally as a group of 18 carbon-containing steroids. However, as the name implies, any chemical agent that induces estrous in a female mammal may be called an estrogen. These ligands need not resemble the 4 adjacent carbon rings of a steroid; much simpler molecules such as the dye phenol red can act as estrogens. Yet no matter what their structure, they all are thought to work by triggering the same general sequence of molecular events.

Estrogens passively enter cells by simple diffusion. In those target tissues that are capable of responding, the estrogen encounters a nuclear protein to which it binds. The binding is highly specific and quite firm. In the absence of this receptor protein, a cell cannot respond as an estrogen target in a genomic fashion. During binding, the receptor protein alters its shape (configuration). Then, the steroid receptor complex binds, selectively, to a short nucleotide sequence on the DNA chain, called an estrogen response element (ERE).[166] This sequence is usually situated upstream from each of those genes that are responsive to the hormone. The bound steroid-receptor complex acts as a transcription factor, modulating the synthesis of that gene's messenger RNA product. The altered mRNA production rate is mirrored by a change in synthesis of that gene's coded protein. Changes in protein synthesis are then responsible for the final metabolic, structural, and mitotic effects of the hormone.[167] The exact number of genes influenced by estrogens is unknown but is thought to be greater than 100. Each gene codes for a single protein, but in some situations the observable estrogenic event is a result of changes in the transcription of several sensitive genes.[168]

This classical theory of estrogen action has been studied and refined during the years since the estrogen receptor was first identified.[169] The theory accounts for almost all of the known actions attributable to estrogens. The recent discovery of a less ubiquitous second estrogen receptor (designated beta) has not substantially influenced this theory. Moreover, the identification of a cell surface estrogen receptor has given support to the classical theory by providing a possible mechanistic explanation for those nongenomic events attributed to estrogens.[167] The extent to which beta-receptor genomic and cell surface receptor nongenomic mechanisms play a role in the overall influence of estrogen on organizational events has not been fully defined.[170] Their number and breadth is thought to be small and unlikely to be an appreciable part of the overall clinical response seen during ERT.

Simply put, application of classic theory would seem to suggest that, paraphrasing the words of a lapsed medical student, "an estrogen is an estrogen is an estrogen."[171] Experience tells us something else. Estrogens of similar molecular weight can have substantially different potencies on a weight-to-weight basis, and their potency ratios may vary considerably depending on the chosen route of administration.[172,173] Moreover, two estrogens may have a potency ratio for one quantifiable effect that is reversed when measuring another outcome.[174] And, finally, some ligands can act as estrogens in one cell type or for one genomic event while behaving as antagonists in another estrogenic system.[175] These seeming dichotomies have driven research about estrogen action and stimulated discoveries that have provoked modification of the classic theory. They have, in turn, fostered appreciation that differences in the route of administration and molecular configuration can potentially induce differences in the outcome of estrogen therapy. Conceptually, advances in estrogens' mechanism of action are getting to a point that specific molecular constructs will become available for stimulating one or another estrogen effect while inhibiting others. As the relative clinical value of ERT in each tissue and anatomic system becomes clear, the ability to target specific tissue sites will become desirable.[176] In order to do this, physicians are going to need to develop a more comprehensive appreciation of the mechanistic pharmacology of estrogens.

Differences in biological outcome among individual estrogenic agents are the result of operational alterations at any one of several sites in the pathway for estrogen action. Agents that are delivered orally will, themselves, undergo a first-pass effect as they are metabolized in the liver while simultaneously inducing a first-pass effect on the liver. In other words, estrogen absorbed from the gastrointestinal tract will be chemically altered during passage through the liver and the liver will, as an estrogen-sensitive tissue, be modified by its exposure to a relatively high concentration of the agent. Metabolic transformation and conjugation limit the amount of estrogen entering the peripheral circulation. As a result, those estrogens that are relatively resistant to hepatic degradation appear to be more potent.[177] When delivery is by transvaginal or transdermal routes, smaller loads will usually produce serum concentrations equal to those obtained with larger amounts of orally administered drug, and the estrogenic stimulus to the liver will be correspondingly less. Selecting between routes for delivery is, therefore, unlikely to produce substantive differences in outcome unless alterations in lipoprotein or coagulation profiles are of paramount importance.

Once in a target cell, estrogens show considerable variation in their binding affinity with the estrogen receptor. Those agents that have short residence time on the receptor produce smaller effects unless administered repeatedly. However, since clinical applications usually involve chronic drug administration, differences in binding affinity are of little therapeutic consequence. The conformational differences that various estrogens evoke in the estrogen receptor influence the way that the estrogen receptor complex binds to each gene's estrogen response element. This can cause certain estrogens to produce quantitatively different stimuli for transcription at one or another estrogen-sensitive gene.[178] These substantive differences in estrogen action have been demonstrated in isolated laboratory systems and conform to classical theories of estrogen action. However, they have thus far been found to make very little clinical impact. Among commonly used estrogens, routes of administration, variations in potency, and variations in genomic stimulatory capabilities have not translated into substantial differences in outcome during clinical ERT trials.[142] This generalization holds for both desirable and undesirable actions of the steroidal estrogens. Thus, it would appear that for the clinical objectives of estrogen replacement, an estrogen is an estrogen is an estrogen -- as long as the dose is adjusted to provide an optimal plasma level of the hormone.

For those women with an intact uterus who embark upon ERT, the addition of a progestin is the accepted standard.[179] The progestin most frequently used by both clinical investigators and prescribing physicians in the United States has been medroxyprogesterone acetate, a synthetic C-21 steroid.[180] The addition of this progestin to ERT has a number of undesirable effects on cardiovascular metabolism that partially nullify the beneficial alterations provided by estrogen. It reduces the vascular dilatory actions of estrogen, minimizes the rise in circulating HDL, and accelerates uptake of LDL into vascular plaque.[173] These unfavorable effects have led epidemiologists to forecast a much lower positive influence on life expectancy from HRT than from ERT and to discount the overall value of replacement therapy in women with intact uteri.[159] Similar liabilities are assumed to result from the use of 19-nortestosterone progestins (norethindrone, etc). The development of an orally effective delivery system for progesterone has, therefore, been eagerly awaited -- the expectation being that this ovarian hormone will provide the same endometrial proliferative inhibitions as synthetic progestins without invoking any of their deleterious cardiovascular actions. The recent availability of micronized, crystallin progesterone has allowed these theoretical notions to be field tested. Orally delivered, micronized progesterone provides a predictable serum concentration.[181] Metabolism during the first hepatic pass is substantial but not sufficient to prevent establishment of therapeutic blood levels. The effect of these serum levels is a consistent inhibition to both endometrial proliferation and the development of atypical adenomatous hyperplasia. Moreover, there does not appear to be a significant modification of the advantageous lipid profile produced by estrogen administration nor an induction of the other deleterious cardiovascular laboratory events brought on by synthetic progestins.[182]

Selective Estrogen Response Modulators

The effects of antiestrogens or partial estrogens provide a contrasting story to that of the progestins. Though the earliest representatives of this genre (eg, clomiphene citrate) have been known for 40 years, it is only in the last decade that their peculiar characteristics have been aggressively studied and clinically exploited. These synthetic agents competitively bind to the estrogen receptor but do not evoke the full panoply of estrogen responses. Some are complete antagonists, some are partial antagonists, and still others possess both agonist and antagonist characteristics.[183] The exact mechanisms by which they produce these actions are not fully understood. Findings to date suggest that a new layer of complexity will need to be added to the classic theory of estrogen action in order to integrate their actions. These agents seem to cause a distinct conformational change in the estrogen receptor molecule that selectively alters the estrogen-receptor complexes' ability to influence transcription.[184] As a result, in some estrogen target cells they prevent estrogen-induced events while in other cells they provoke the same changes as would a standard estrogen.[185,186] Because they possess these agonist/antagonist capabilities in varying combination, they have collectively been called selective estrogen response modulators (SERMs). Their potential lies in the possibility that by choosing judiciously, SERMs possessing only favorable estrogenic activity can be applied as alternatives to classic estrogens in the treatment of postmenopausal estrogen deficiency.

At present there are two clinically available SERMs; tamoxifen and raloxifene. Both are estrogen agonists in their action on bone and cholesterol metabolism but act as antagonists in the breast. Tamoxifen induces estrogenic events in endometrium, including proliferation, while raloxifene behaves as a complete antagonist in that tissue.[187] The extent to which either drug affects the cardiovascular system beyond changing lipid profiles has not yet been determined. Unfortunately, a similar lack of information exists for both agents with regard to cognitive events. Additionally, raloxifene does not ameliorate vasomotor symptoms.[188] Thus, it appears that raloxifene might serve advantageously as a replacement for estrogen in long-term ERT regimens but not in short-term regimens aimed solely at combating vasomotor symptoms. Women ingesting it would not experience vaginal spotting or bleeding and would not need to take a progestin to protect against endometrial hyperplasia. Users would, however, fail to derive relief from vasomotor symptoms and would, for the present, be unaware of the degree to which they were attaining the cognitive or cardiovascular benefits expected from classic estrogens.[176] Obviously, the results of long-term, randomized clinical trials are needed before a wholesale switch to this type of designer estrogen can be contemplated.

Ironically, there are societies that may have, albeit unknowingly, been taking advantage of a SERM-like effect for centuries.[189] Some East Asian cultures ingest diets containing large amounts of phytoestrogens (plant-derived compounds that have agonist/antagonist actions). Evidence is accumulating to suggest an estrogen antagonistic effect on breast tissue, and a substantial agonist effect on serum lipoprotein profiles is obtained from these plant-derived agents.[189,190] In large quantities, the phytoestrogens from soybeans are also agonists on the vasomotor flush apparatus.[191] At present, neither the full spectrum nor the extent and specificity of the alterations induced by these estrogen-like compounds are known; nor is it established that phytoestrogens actually act through the same classical mechanisms as do raloxifene and the other synthetic partial estrogens.[192]

Hormone Replacement Regimens

For most women, the treatment of postmenopausal estrogen deficiency will involve administration of both an estrogen and progestin. This can be accomplished using any number of different regimens (see Figure 3). Most are based, for convenience, on the calendar month. None are peculiar to a specific type of estrogen, dosage, or route of administration. Estrogen is given for either 25 days or throughout the whole month. Progestin is provided daily during each day of estrogen administration or for 12-14 days out of each cycle. When given in a discontinuous fashion, the progestin is administered for an interval of more than 10 days in order to ensure adequate suppression of endometrial proliferative activity.[180,193] Higher doses of progestin are required for effective intermittent therapy than for continuous therapy. Stopping therapy for 5 days out of each month has the advantage of limiting chronic breast symptoms, for some. This may not be as important to those women who, because they do not have a uterus, are only taking an estrogen.[194] Women seldom complain of vasomotor problems during this interval of abstinence. And, when the 25-day-per-month regimen includes daily intake of both hormones, withdrawal uterine bleeding is seldom experienced. On the other hand, when the progestin is limited to a portion of the estrogen cycle, there is a much greater likelihood that cyclic withdrawal bleeding will occur (about 60%), whether the estrogen is administered for 25 days or the entire month.

Because of this propensity to bleed with intermittent progestin regimens, most therapists favor continuous progestin protocols. However, some women (10% to 20%) experience intermittent spotting on continuous combined estrogen/progestin regimens and will, therefore, favor intermittent progestin with its more predictable blood flow.[193] Clinical experience has shown that, no matter what regimen is followed, it will take at least 6 months before amenorrhea or a predictable pattern of uterine bleeding is established.

A variant of the intermittent progestin regimen that aims at diminishing the frequency of uterine bleeding is one that provides progestin for 14 days at 3-month intervals. This protocol seems to offer adequate protection from endometrial neoplasia.[195] However, large-scale studies of sufficient length have not yet been reported so that its efficacy must remain in question. The regimen also provides advantage to those who experience depression or other progestin-induced side effects.

While little work has been done to document differences in long-range outcome or measurable intermediate parameters between these various regimens, the increased incidence of bleeding experienced with intermittent progestin therapy may dissuade some women from continuing chronic therapy. One study has found that between 10% and 50% of women who stop taking HRT cite bleeding as their reason for doing so.[196] Continuous combined therapy is therefore favored as a first-try regimen.

Before beginning HRT, there are certain particulars that need to be documented in the medical record, including: the candidate's estrogen status, the content of her HRT counseling, findings of a general physical exam, cervical cytology, and mammographic results. Interval evaluations of women on HRT should include description of the hormonal regimen and their endometrial status as ascertained by bleeding history and other appropriate monitors.[197]

Postmenopausal estrogen replacement is given either to suppress vasomotor symptoms, a short-term aim, or to diminish the estrogen deficiency sequelae of old age. In the former situation, individual, objective changes in frequency and severity of symptoms provide information with which to establish an appropriate hormone dose. For the latter, there are no directly measurable parameters upon which to judge the eventual impact of a particular hormone dose in a particular individual. A very substantial experience with these HRT regimens gives assurance that, with adequate progestin exposure, endometrial hyperplasia is a rare event. Biopsy of the uterine lining is, therefore, seldom necessary. Sampling ought to be done when uterine bleeding occurs at times other than late in the progestin portion of an intermittent regimen or during withdrawal. In continuous combination, hormone cycle changes in the established spotting pattern or heavy/prolonged bleeding suggest the need for biopsy. And, among those few women with intact uteri who elect to abstain from progestin, sampling should be considered an annual event unless assurance is gained from ultrasound that the endometrium remains hypoplastic.

The dose of estrogen required to provide relief from vasomotor symptoms is, not surprisingly, about the same as the dose required to raise serum estrone and estradiol concentrations to their early or midfollicular phase levels: about 100 pg/mL for estrone and 50 pg/mL for estradiol. During a woman's reproductive years, estrone and estradiol concentrations cycle, reaching their lowest level with the onset of menses (see Figure 2). Doses of estrogen that raise serum concentrations appreciably beyond these levels will frequently cause side effects such as breast engorgement and may, over prolonged periods, increase the risk of breast cancer.[198] When vasomotor symptoms are not relieved by these doses, poor uptake or increased hepatic metabolism is likely. When adequate serum estrogen concentrations are present and vasomotor complaints persist, another cause for the problem should be sought. Even if only a limited treatment interval is anticipated, a progestin must nevertheless be included from the onset. Unopposed estrogen therapy has an effect on endometrial proliferation that goes beyond the period of treatment, thereby increasing the relative risk of endometrial neoplasia even when estrogen treatment lasts for less than 5 years.

Determining the estrogen dose needed for attainment of long-term ER goals is problematic because reliable biofeedback parameters are limited. The extent to which circulating FSH levels are lowered during estrogen replacement does not correlate with the degree of estrogen response induced in other sensitive tissues. This is because inhibin, a formidable suppressor of pituitary FSH production, is all but gone from the circulation after menopause.[199] Similarly, bio-indicators of bone metabolism and altered profiles for serum lipoproteins provide only limited predictive information about the overall, long-term benefits that are expected from ERT.[200] At present, the best information available concerning estrogen dosage for long-term therapy comes from two sources. First, favorable cardiovascular outcome trials have, for the most part, not shown a dose-dependent outcome when estrogen dose was above threshold. Second, the lowest dosage of conjugated equine estrogen (Premarin) to show general effectiveness in preventing osteoporosis has been 0.625 mg per day.[201] It therefore appears that the lowest estrogen dose capable of producing a generally favorable long-term outcome for most women is, like that for the treatment of vasomotor instability, one that produces serum hormone levels comparable to those experienced during early follicular development. Laboratory testing aimed at individualizing the optimum estrogen dose for long-term ERT is not generally performed or productive.

As long as effective circulating levels of estrogen are achieved, the route of administration is of little concern. While there are theoretical reasons to expect that beneficial cardiovascular actions will occur with lower circulating levels of estrogen delivered via the enterohepatic route, clinical data are not available to support this notion. Similarly, convincing data that demonstrate a more favorable clinical outcome when using one or another estrogenic compound have not, as yet, been published. Therefore, it seems prudent to prescribe 0.625 mg of conjugated equine estrogen, 0.625 mg of esterified estrogen, 2 mg of micronized estradiol, or their daily equivalent for long-term therapy (see Table 1). The daily dose of progestin that effectively accompanies these estrogens is 2.5-5.0 mg of medroxyprogesterone acetate or 1-5 mg norethindrone acetate. Both of these agents induce detrimental effects on cardiovascular function, making the new, as yet clinically unevaluated, micronized progesterone products potentially superior for long-term HRT. One hundred mg once each day or 200 mg each day for 12 days of the cycle may eventually be proven to be as effective as the standard doses of synthetic progestins for endometrial control and superior for cardiovascular outcomes (Figure 3).

Hormone replacement is most frequently begun when menopause is first recognized. However, some women will experience significant vasomotor episodes a year or more before menstruation ceases. When night sweats cause sleep deprivation or are otherwise disconcerting, HRT ought to be considered despite ongoing menstruation. Usually this therapy will override the preexisting spontaneous bleeding episodes.

Because very large decrements in bone mineral density occur in the first few years after menopause, it was once thought that late initiation of estrogen therapy was of little therapeutic value. More recently it has been recognized that the continued, albeit slower bone loss of the later postmenopause can be arrested and bone density even increased somewhat when the initiation of therapy is delayed. In addition, there is evidence that substantial cardiovascular benefits can accrue even when therapy begins 10 or more years after the onset of menopause. It is, therefore, recommended that HRT be initiated whenever the postmenopausal candidate is ready to undertake therapy.[202]

When the aim of ERT is solely the relief of vasomotor symptoms, successful therapy can be gradually withdrawn after an interval of a few years. An abrupt cessation of estrogen intake seems to induce the resumption of symptoms more frequently than does a stepwise reduction, though experimental data are not available in support of this contention. Similarly, there is an absence of information from which to conclude the appropriate time for a trial withdrawal. Long-term replacement therapy, on the other hand, should be considered a lifetime endeavor. There is ample evidence to show that major treatment benefits rapidly wane once hormone administration ceases.

While the addition of progesterone to an estrogen replacement regimen is made necessary by the presence of the uterine endometrium, inclusion of an androgen must be considered as optional. Serum testosterone levels decline somewhat around the time of menopause but then remain relatively stable in subsequent years.[10] Dehydroepiandrosterone, a very weak androgen, declines steadily from young adult levels throughout the remainder of life. Premenopausal women who are oophorectomized experience a much more profound drop in circulating androgens, due almost entirely to testosterone. Most of these women note a substantial change in libido. However, sexual desire does not change as predictably after menopause, and the libidinal effect of exogenous androgen remains to be thoroughly evaluated.

Efforts to determine the frequency and extent of libidinal decline in postmenopausal women have been of limited scope. It is estimated that about 50% of women experience some decrease in sexual interest around the time of menopause and that 20% consider this to be unwelcome.[117] There is persuasive evidence that when parenterally administered testosterone attains supraphysiologic levels, it will evoke higher rates of sexual desire, arousal, coitus, and orgasm than when estrogen is the only replacement steroid.[118] Neither large observational nor controlled/prospective studies evaluating the use of orally administered, physiologic doses of testosterone for libidinal enhancement have been reported. Anecdotal experience suggests that 2.5 mg of daily oral methyltestosterone will provide relief for most women who complain of a postmenopausal libidinal defect. Higher doses of testosterone have occasionally been recommended but are more likely to induce unwanted side effects such as facial hair growth and acne.[203,204] Unfortunately, even the lower doses of testosterone are known to reduce HDL levels and, thereby, enhance some potentially detrimental events affecting cardiovascular disease development.[205,206] It is primarily for this reason that the addition of testosterone to hormone replacement regimens should be individualized and that testosterone need not be continued for the full interval of long-term estrogen therapy.

Contraindications to Replacement Therapy

There are only a few instances in which a therapist must consider withholding postmenopausal estrogen replacement. A recent episode of deep vein thrombophlebitis may be aggravated by the alterations in coagulation factors that are brought on by standard replacement regimens. However, there do not seem to be any studies that address the question of whether ERT increases the risk of recurrence. At present, the post-thrombosis time interval before therapy can be safely begun or resumed is unknown. The fact that there is a slight increase in thrombosis risk among women on HRT has led some to consider a positive thrombosis history an absolute contraindication.[128] Probably the most prudent course is to withhold estrogen in those with a hypercoagulable state and to begin therapy in those without known predisposing factors after a wait of 1-2 years.[207] Undiagnosed uterine bleeding is considered a reason to withhold ERT. But, once the reason for bleeding is found to be other than endometrial neoplasia and has been treated, therapy can be undertaken. A few women with stage I endometrial cancer have received HRT after appropriate cancer therapy without recurrence.[208]

The question of whether to administer estrogen therapy when there is a history of breast cancer is problematic. At present, the suggestive evidence for an increased risk of breast cancer after 10 or more years of postmenopausal ERT has not been accompanied by evidence that those with a strong family history are any more prone to acquire the disease while on estrogens. Women who have had breast cancer have been given ERT after a 2-year or longer disease-free interval without significant cancer recurrence.[145,209] The lack of a large-scale, randomized study in support of these findings leaves the practitioner without substantial guidance.[153] Each potential candidate must therefore be evaluated as an individual and the decision to treat made in concert after considering her particular indications and risk tolerance.[210] As the general actions of SERMs such as tamoxifen and raloxifene are better understood, these agents may become effective substitutes for estrogens in postmenopausal women who have been treated for estrogen-sensitive neoplasia. In addition to breast cancer, melanotic adenocarcinoma of the uterus falls into this category. Other forms of melanoma, though potentially estrogen dependent, may actually respond favorably to estrogens and are therefore not considered an absolute contraindication to estrogen replacement.[211]

Severe acute and chronic liver disease is usually considered an absolute contraindication to HRT. However, there is little information available about the effects of estrogen, at the dosages most frequently used for postmenopausal therapy, on liver function. Liver disease might better be considered a relative contraindication necessitating a higher degree of clinical observation during HRT. This would include periodic measurement of serum estrogen levels, since its metabolism may be compromised because of the underlying disease.[212] Hepatic injury may also present a situation for which a nonoral route of estrogen administration is better suited. As time and experience have accrued, several additional conditions that were once considered to be contraindications to ERT have been reevaluated and then recategorized; among them are diabetes mellitus, endometriosis, hypertension, and otosclerosis.[211]

Sustaining HRT

Despite the strong case that can be made for long-term hormone replacement, relatively few women undertake therapy and even fewer continue to use it for more than 5 years. In 1990, it was estimated that less than 20% of postmenopausal women in the United States had received replacement estrogens, and that use declined dramatically after age 65.[213] With the advantages of HRT becoming more apparent, it is likely that greater numbers of women will embark upon long-term use.[214] Individual decisions to use HRT are influenced by at least three major determinants: a candidate's perception of the health problem, her provider's attitudes about therapy, and her personal experience with hormone replacement.[215] Many postmenopausal women and some of their healthcare providers are unfavorably disposed toward HRT. Some of these people are ill informed, their biases based on misconception. Others have belief systems that are opposed to interfering with what they consider a natural process. A few are well informed but, by choosing to assign different relative values to the benefits and complications of HRT, conclude that the undertaking is not worthwhile. Finally, there are those who think the available data insufficient to sustain a rational determination.[159] No matter what a candidate's decision, conscientious healthcare workers will want to assure that it is based on logically evaluated, accurate information.

The public receives more of their information about HRT from the mass media than from healthcare providers.[202] Often that information is incomplete, distorted, or of a sensational nature. Determining the extent and adequacy of a candidate's information can effectively direct the educational efforts of her healthcare provider.[216] For example, the dangers of breast cancer are of paramount concern to peri- and postmenopausal women. Opinion surveys have found that women believe that breast cancer accounts for 40% of women's deaths, while cardiovascular disease accounts for 19%.[217] Thus, when they hear that chronic estrogen therapy may induce as much as a 40% increase in breast cancer incidence over a lifetime, their concerns for the risks of therapy are appreciably heightened. If, on the other hand, candidates were given to understand that women with a history of postmenopausal estrogen use actually have a lower mortality rate from their breast cancer than those women who are not exposed to the hormone, and if they further understood that breast cancer accounted for 4% of women's deaths (not 40%), they might be less fearful of undertaking hormone replacement. Similarly, understanding that cardiovascular events account for 45% of postmenopausal women's deaths and that estrogen therapy reduces that rate by as much as 50% might induce candidates to reconsider the priorities they use in their evaluation of HRT.

Another area where there is considerable misunderstanding with regard to HRT is the type of formulation that is to be used. Many women will refuse hormones that are not "natural." This may mean a hormone that has or has not been synthesized from precursors originating in plants. Or, it may mean that the formula ought to be identical to that of estrogens produced by the human ovary. For others, the concern is that the source of the drug not be a chemical company that synthesized its product from elemental agents or a pharmaceutical house that obtains its base from the urine of pregnant mammals. The public is, in general, quite naive about the structure of estrogens, the differences in structure, and the relationship of structure to biological function. Frequently, deficits in understanding the chemistry of estrogens lie at the foundation of their misconceptions. A few women will actually be aware of the differences in metabolic effect between the contents of birth control pills (alkylestrogens) and those given in HRT. Whatever the cause of concern, an effective strategy for prescribing once that concern is discerned is to initiate a thorough discussion of the topic. Most women will welcome this and acquiesce to the use of the provider's choice formulation. When the candidate remains unconvinced, it is usually possible, having recognized her concerns, to prescribe another effective agent. As was proposed earlier, for the purpose of inducing multiple estrogen responses during long-term postmenopausal therapy, an estrogen is an estrogen is an estrogen.

The public's increasing interest in alternative medicine has induced many women to seek postmenopausal medicaments from their health-food store. Some will ask their primary care provider about the efficacy of specific agents or regimens, phytoestrogens among them. In most cases, these products will cause no immediate harm to the individual. However, when they are used as substitutes for adequate doses of effective estrogen, they produce long-term harm by depriving the individual of the known benefits that come from HRT.[191] Women considering using such agents as their major means of dealing with menopausal problems should be made aware that studies to demonstrate efficacy are seldom available, that the amounts of active constituents in these preparations usually vary from batch to batch, and that they have not gone through the rigorous US Food and Drug Administration approval process that is required of prescription pharmaceuticals.

Summary

During the past quarter century, observational trials have begun to give the medical community an appreciation for the systemic alterations that can be induced by postmenopausal estrogen replacement. Most of the benefits envisioned by midcentury advocates have been demonstrated in these studies. Some additional benefits and at least two notable drawbacks (malignancy, thrombosis), not foreseen 50 years ago, have become apparent. These findings have significantly altered standard therapeutic protocols as well as public perceptions about the value derived from estrogen replacement. Considerable progress has also been made in understanding the biology of estrogens during this period. The conjunction of knowledge gained from clinical experience and biomedical research now provides us with the capacity to substantially inhibit many of the degenerative processes intrinsic to old age. Controlled, prospective trials are currently under way that will better define the risk/benefit aggregate of our most common HRT regimens.[218] The findings from these studies will hopefully suggest types of treatment modifications that can further enhance clinical outcomes. The use of progesterone in place of synthetic progestins has the potential to diminish the negative cardiovascular effects that result from the synthetics' influence on serum lipid profiles. A better understanding of estrogens' effect on breast cancer rates, gained from currently ongoing studies, will provide a base upon which to evaluate the role of SERMs in postmenopausal therapy. Additional information about estrogens' capacity to inhibit deterioration in CNS function will enlarge our appreciation for the nonquantifiable benefits that public health officials often ignore when they consider the merits of HRT. All of these changes are likely to provide support for a wider use of postmenopausal hormone replacement. If clinicians are to play a significant role in advancing public understanding of the value to be gained from replacement therapy, they will need to become more conversant with the biomolecular actions of estrogens as well as the concerns and issues that women ponder as they face the last third of their lives.[160]

References

  1. Baker TG. A quantitative and cytological study of germ cells in human ovaries. Proc R Soc Lond B Biol Sci. 1963;158:417-433.

  2. Richardson SJ, Nelson FJ. Follicular depletion during the menopausal transition. Ann NY Acad Sci. 1990;592:13-20.

  3. Gougeon A. Origin and growth of the preovulatory follicle in spontaneous and stimulated cycles. In: Testart J, Frydman R, eds. Human In vitro Fertilization: Actual Problems and Perspectives. New York, NY: Elsevier Science; 1985.

  4. Richardson SJ. The biological basis of the menopause. Baillieres Clin Endocrinol Metab. 1993;7:1-16.

  5. Klein N, Battaglia D, Fujimoto V, et al. Reproductive aging: accelerated ovarian follicular development associated with a monotropic follicle-stimulating hormone rise in normal older women. J Clin Endocrinol Metab. 1996;81:1038-1045.

  6. MacNaughton J, Bangah M, McCloud P, et al. Age related changes in follicle stimulating hormone, luteinizing hormone, oestradiol and immunoreactive inhibin in women of reproductive age. Clin Endocrinol (Oxf). 1992;36:339-345.

  7. Moslers WH, Johnson VE. Human Sexual Response. Boston: Little Brown and Co.; 1966.

  8. Klein N, Illingworth P, Groome N, et al. Decreased inhibin B secretion is associated with the monotropic FSH rise in older, ovulatory women: a study of serum and follicular fluid levels of dimeric inhibin A and B in spontaneous menstrual cycles. J Clin Endocrinol Metab. 1996;81:2742-2745.

  9. Santoro N, Adel T, Skurnick J. Decreased inhibin tone and increased activin A secretion characterize reproductive aging in women. Fertil Steril. 1999;71:658-662.

  10. Judd H, Fournet N. Changes of ovarian hormonal function with aging. Exp Gerontol. 1994;29:285-298.

  11. McKinlay SM, Brambilla DJ, Posner JG. The normal menopause transition. Maturitas. 1992;14:103-116.

  12. Frere G. Mean age at menopause and menarche in South Africa. S Afr J Med Sci. 1971;36:21-24.

  13. Scrogg RF. Menopause and reproductive span in rural Niuginsi. Proc 9th Ann Symp Papua New Guinea Med Soc. 1973:126-131.

  14. Adena MA, Gallagher HG. Cigarette smoking and the age at menopause. Ann Hum Biol. 1982;9:121-130.

  15. Stanford JL, Hartge P, Brinton LA, et al. Factors influencing the age at natural menopause. J Chronic Dis. 1987;40:995-1002.

  16. Van Keep P, Brand P, Lehert P. Factors affecting the age at menopause. J Biosoc Sci. 1979;6(suppl):37-55.

  17. Amundsen DW, Diers CJ. The age of menopause in medieval Europe. Hum Biol. 1973;45:605-612.

  18. Wilbush J. Historical perspective: Tilt, E.J. and the change of Life (1857) -- the only work on the subject in the English language. Maturitas. 1980;2:259-267.

  19. Tilt EJ. The Change of Life in Health and Disease. London: Churchill; 1857..

  20. Frommer DJ. Changing age of the menopause. Br Med J. 1964;2:349-351.

  21. Ricci SV. Aetros of Amida; a translation. Philadelphia, Pa: Blakeston; 1950.

  22. Barbre JW. From "Goodwives" to Menoboomers: Reinventing Menopause in American History. Minneapolis/St. Paul, Minn: University of Minnesota Press; 1994.

  23. Wilbush J. La Menespausie -- the birth of a syndrome. Maturitas. 1979;1:145-151.

  24. Curries AF. The Menopause. New York, NY: Appleton; 1897.

  25. Leake S. Chronic or slow disease peculiar to women. London: Baldwin; 1977.

  26. Smith-Rosenberg C. Puberty to menopause: the cycle of femininity in nineteenth century America. Feminist Studies. 1973;1:58-72.

  27. Kerr JM, Johnstone RW, Phillips MH. Historical Review of Obstetrics and Gynecology, 1800-1950. London: Livingstone; 1954.

  28. Novak E. Ovarian therapy. JAMA. 1924;83:2016-2018.

  29. Marshal FH, Jolly WA. Contributions to the physiology of mammalian reproduction: Part II. The ovary as an organ of internal secretion. Philos Trans R Soc Lond B Biol Sci. 1906;198:99-102.

  30. Novak E. Menstruation and Its Disorders. New York, NY: Appleton; 1921.

  31. Bell S. Changing ideas: the medicalization of menopause. Soc Sci Med. 1989;24:535-542.

  32. Brewer JI. Textbook of Gynecology. New York, NY: Williams and Wilkins; 1961: 112.

  33. Wilson RA. Feminine Forever. New York, NY: M. Evans; 1966.

  34. Lamberts SW, Van den Beld AW, Van der Lely A. The endocrinology of aging. Science. 1997;278:419-424.

  35. Berg G, Gottwall T, Hammar M, et al. Characteristic symptoms among women aged 60-62 in Lurkoping Sweden in 1986. Maturitas. 1988;10:193-199.

  36. Kronenberg F. Hot flashes: epidemiology and physiology. Ann NY Acad Sci. 1990;592:52-86.

  37. Avis NE, Kaufert PA, Lack M, et al. The evolution of menopausal symptoms. Baillieres Clin Endocrinol Metab. 1993;7:17-32.

  38. Hawkinson LF. The menopausal syndrome. JAMA. 1938;111:390-393.

  39. Barrett L, Cullis W, Fairfield L, et al. An investigation of the menopause in one thousand women. Lancet. 1933;1:106-108.

  40. Lock M. Japanese experience and perceptions of menopause. Cult Med Psychiatry. 1986;10:23-46.

  41. Tataryn IV, Melabrun DR, Lu KH, et al. FSH and skin temperature during menopausal hot flash. Clin Endocrinol Metab. 1979;49:152-154.

  42. Sturdee DW, Wilson KA, Pipili E, et al. Physiological aspects of menopausal hot flashes. Br Med J. 1978;2:79-80.

  43. Altura BM. Sex and estrogens and responsiveness of terminal arterioles to neurohypophyseal hormones and catecholamines. J Pharmacol Exp Ther. 1975;193:403-412.

  44. Tataryn IV, Lomax P, Bajorek JG, et al. Postmenopausal hot flashes: a disorder of thermoregulation. Maturitas. 1980;2:101-107.

  45. Longcope C, Crawford S, McKinlay S. Endogenous estrogens: relationship between estrone, estradiol, non-protein bound estradiol. and hot flashes and lipids. Menopause. 1996;3:77-84.

  46. Aksel S, Schomberg D, Tyrey L, et al. Vasomotor symptoms, serum estrogens and gonadotropin levels in surgical menopause. Am J Obstet Gynecol. 1976;126:165-169.

  47. Utian WH. The true clinical features of postmenopausal oophorectomy, and their response to oestrogen therapy. S Afr Med J. 1972;46:732-737.

  48. Casper RF, Yen SS, Wilkes MM. Menopausal flashes: a neuroendocrine link with pulsatile luteinizing hormone secretion. Science. 1979;205:823-825.

  49. Burger HG, Dudley EC, Hopper JL. The endocrinology of the menopausal transition: a cross sectional study of a population based sample. J Clin Endocrinol Metab. 1995;80:3537-3545.

  50. Hunter MS. Emotional well-being, sexual behavior and hormone replacement therapy. Maturitas. 1990;12:299-314.

  51. Schmidt PJ, Rubinow DR. Menopause-related affective disorders: a justification for further study. Am J Psych. 1991;148:844-850.

  52. Ballinger CB. Psychiatric aspects of the menopause. Br J Psychiatry. 1990;156:773-787.

  53. Stone AB, Pearlstein TB. Evaluation and treatment of changes in mood, sleep and sexual functioning associated with menopause. Obstet Gynecol Clin North Am. 1994;21:391-403.

  54. Ditkoff EC, Gary WG, Cresto M, et al. Estrogen improves psychological function in asymptomatic postmenopausal women. Obstet Gynecol. 1991;78:991-995.

  55. Haas S, Walsh B, Evans S, et al. The effect of transdermal estradiol on hormone and metabolic dynamics over a six-week period. Obstet Gynecol. 1988;71:671-676.

  56. Semmens JP, Tsai CC, Semmens EC, Loadholt CB. Effects of estrogen on vaginal physiology during menopause. Obstet Gynecol. 1985;66:15-18.

  57. Pfeiffer E, Verwoerdt A, Davis GC. Sexual behavior in middle life. Am J Psychiatry. 1972;128:1262-1267.

  58. Bush TL, Barrett-Connor E, Cowan LD, et al. Cardiovascular mortality and non contraceptive use of estrogen in women: results of Lipid Research Clinics Program follow-up study. Circulation. 1987;75:1102-1109.

  59. Tunstall-Pedoe H. Myth and paradox of coronary risk and the menopause. Lancet. 1998;351:1425-1428.

  60. Vendola AR, Simon JA. Nonlipid effects of estrogen replacement on the cardiovascular system. Infertil Repro Med Clin North Am. 1995;6:829-840.

  61. Mendelsohn ME, Karas RH. The protective effects of estrogen on the cardiovascular system. N Engl J Med. 1999;340:1801-1811.

  62. Wagner JD, Clarkson TB, St. Clair RW, et al. Estrogen and progesterone replacement therapy reduces low density lipoprotein accumulation in the coronary arteries of surgically menopausal cynomolgus monkeys. J Clin Invest. 1991; 88:1995-2002.

  63. Wild RA. Estrogen: effects on the cardiovascular tree. Obstet Gynecol. 1996;87:27s-35s.

  64. Stevenson JC. Mechanisms whereby oestrogens influence arterial health. Eur J Obstet Gynecol. 1996;65:39-42.

  65. Chen S, Li H, Durand J, et al. Estrogen reduces myointimal proliferation after balloon injury of rat carotid artery. Circulation. 1996;93:577-584.

  66. Iafrati MD, Karas RH, Aronovitz M, et al. Estrogen inhibits the vascular injury response in estrogen receptor a-deficient mice. Nat Med. 1997;3:545-548.

  67. Williams JK, Adams M, Klopfenstein H. Estrogen modulates responses of atherosclerotic coronary arteries. Circulation. 1990;81:1680-1687.

  68. Bush TL. Estrogen replacement and coronary disease: the evidence for primary and secondary prevention. J Clin Endocrinol Metab. 1996;81:3830-3832.

  69. The Writing Group. Effects of estrogen and estrogen/progestin regimens on heart disease risk factors in post menopausal women: the post menopausal estrogen/progestin interventions (PEPI) trial. JAMA. 1995;273:199-208.

  70. Barrett-Connor E, Grady D. Hormone replacement therapy, heart disease and other considerations. Annu Rev Public Health. 1998;19:55-72.

  71. Rodstrom K, Bengtsson C, Lissner L, et al. Pre-existing risk factor profiles in users and non-users of hormone replacement therapy: prospective cohort study in Gottenburg, Sweden. BMJ. 1999;319:890-893.

  72. Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. JAMA. 1998;280:605-613.

  73. Herrington D. Effects of estrogen replacement on the progression of coronary artery atherosclerosis. N Engl J Med. 2000;343:522-529.

  74. Sullivan JM, Vander Zwaag R, Hughes JP, et al. Estrogen replacement and coronary artery disease: effect on survival in postmenopausal women. Arch Intern Med. 1990;150:2557-2562.

  75. Ettinger B. Optimal use of postmenopausal hormone replacement. Obstet Gynecol. 1988;72:31s-36s.

  76. Hui SL, Slemenda CW, Johnston CC. Baseline measurement of bone mass predicts fracture in white women. Ann Intern Med. 1989;111:355-361.

  77. Cummings SR, Browner WS, Bauer D, et al. Endogenous hormones and the risk of hip and vertebral fractures among older women. N Engl J Med. 1998;339:733-738.

  78. Melton LJ, Lane AW, Cooper C, et al. Prevalence and incidence of vertebral deformities. Osteoporos Int. 1993;3:113-119.

  79. Albright F, Smith PH, Richardsen AM. Postmenopausal osteoporosis. JAMA. 1941;116:2465-2474.

  80. Turner RT, Riggs BL, Spelsberg TC. Skeletal effects of estrogens. Endocrinol Rev. 1994;15:275-300.

  81. Ettinger B, Genant HK, Cann CE. Late-term estrogen replacement therapy prevents bone loss and fractures. Ann Intern Med. 1985;102:319-324.

  82. Lindsay R, Tohme JF. Estrogen treatment of patients with established postmenopausal osteoporosis. Obstet Gynecol. 1990;76:290-295.

  83. Felson DT, Zhang Y, Hannan MT, et al. The effect of postmenopausal estrogen therapy on bone density in elderly women. N Engl J Med, 1993;329:1141-1146.

  84. Genant HK, Lucas J, Weiss S, et al. Low-dose esterified estrogen therapy: effective on bone, plasma estradiol concentrations, endometrium and lipid levels. Arch Intern Med. 1997;157:2609-2615.

  85. Lindsay R. Long-term estrogen therapy for treatment and prevention of osteoporosis. J Clin Endocrinol Metab. 1996;81:3829-3820.

  86. Greendale GA, Judd HL. The menopause: health implications and clinical management. J Am Geriatr Soc. 1993;41:426-436.

  87. Schneider DL, Barrett-Connor EL, Morton DJ. Timing of postmenopausal estrogen for optimal bone mineral density: the Rancho Bernardo study. JAMA. 1997;277:543-547.

  88. Iosif CS, Batra S, Ek A, et al. Estrogen receptors in the human female lower urinary tract. Am J Obstet Gynecol. 1981;141:817-820.

  89. Lindgren R, Berg G, Hammar M, et al. Hormonal replacement therapy and sexuality in a population of Swedish postmenopausal women. Acta Obstet Gynecol Scand. 1993;72:292-297.

  90. Diokno AC, Brock BH, Brown MB, et al. Prevalence of urinary incontinence and other urological symptoms in the non-institutionalized elderly. J Urol. 1986;136:1022-1025.

  91. Fantl JA, Cardozo L, McClish DK. Estrogen therapy in the management of urinary incontinence in postmenopausal women: a meta-analysis. First report of the Hormones and Urogenital Therapy Committee. 1994;83:12-18.

  92. Fantl JA, Bump RC, Robinson D, et al. Efficacy of estrogen supplementation in the treatment of urinary incontinence. Obstet Gynecol. 1996;88:745-749.

  93. Ouslander JG, Bruskewitz R. Disorders of micturition in the aging patient. Arch Intern Med. 1989;34:165-190.

  94. Speroff L. Advances in transdermal postmenopausal hormone therapy. Contemp Ob/Gyn. 1999(suppl 1):23.

  95. Raz R, Stamm WE. A controlled trial of intravaginal estriol in postmenopausal women with recurrent urinary tract infections. N Engl J Med. 1993;329:753-756.

  96. Botwinick J. Intellectual abilities. In: Birren JE, Schaie KW, eds. Handbook of the Psychology of Aging. New York, NY: Van Nestrand Reinhold; 1977: 580-605.

  97. Jorm AF, Korten AE, Henderson AS. The prevalence of dementia: a quantitative integration of the literature. Acta Psychiatr Scand. 1987;76:465-479.

  98. Katzman R. Alzheimer's disease. N Engl J Med. 1986;314:964-973.

  99. Kopera H. Estrogens and psychic functions. Front Horm Res. 1973;2:118-133.

  100. Toran-Allerand CD. Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: II. Morphological correlates and hormone specificity. Brain Res. 1980;189:413-427.

  101. Singh M, Meyer EM, Simpkins JW. The effect of ovariectomy and estradiol replacement on brain-derived neurotrophic factor messenger ribonucleic acid expression in cortical and hippocampal brain region of female Sprague-Dawley rats. Endocrinology. 1995;136:2320-2324.

  102. Morrison JH, Hof PR. Life and death of neurons in the aging brain. Science. 1997;278:412-419.

  103. Luine VN. Estradiol increases choline acetyltransferase activity in specific basal forebrain nuclei and projection areas of female rats. Eur Neurol. 1985;89:484-490.

  104. Behl C, Skutella T, Lezoualch F, et al. Neuroprotection against oxidative stress by estrogens: structure activity relationship. Mol Pharmacol. 1997;51:535-541.

  105. Jaffe AB, Toran-Allerand CD, Greengard P, et al. Estrogen regulates metabolism of Alzheimer amyloid beta precursor protein. J Biol Chem. 1994;269:13065-13068.

  106. Stone DJ, Rozovsky Morgan TE, et al. Increased synaptic sprouting in response to estrogen via an apolipoprotein E-dependent mechanism: implications for Alzheimer's disease. J Neurosci. 1998;18:3180-3185.

  107. Craik FM, Yaragihara T, Peterson RC. Memory functions in normal aging. In: Peterson T (ed). Memory Disorders: Research and Clinical Practice. New York, NY: Dekker; 1991: 347-367.

  108. Drachman DA. Memory and cognitive function in normal aging. Dev Neuropsychol. 1986;2:277-285.

  109. Sherwin BB. Estrogen and cognitive functioning in women. Proc Soc Exp Biol Med. 1998;217:17-21.

  110. Yaffe K, Sewaya G, Lieberberg I, et al. Estrogen therapy in postmenopausal women. JAMA. 1998;279:688-695.

  111. Shaywitz SE, Shaywitz BA, Pugh KR, et al. Effects of estrogen on brain activation patterns in postmenopausal women during working memory tasks. JAMA. 1999;281:1197-1202.

  112. Grundman M, Corey-Bloom J, Thal LJ. Perspectives in clinical Alzheimer's disease research and the development of antidementia drugs. J Neural Transm. 1998;53:255-275.

  113. Mulnard R, Cotman C, Kawas C, et al. Estrogen replacement therapy for treatment of mild to moderate Alzheimer disease: a randomized controlled trial. JAMA. 2000;283:1007-1015.

  114. Henderson VW. Estrogen, cognition and a woman's risk of Alzheimer's disease. Am J Med. 1997;103:115-185.

  115. Paganini-Hill A, Henderson VW. Estrogen replacement therapy and risk of Alzheimer's disease. Arch Intern Med. 1996;156:2213-2217.

  116. Kaiser FE. Sexuality in the elderly. Urol Clin North Am. 1996;23:99-109.

  117. Bachmann GA, Leiblum SR, Sandler B, et al. Correlates of sexual desire in postmenopausal women. Maturitas. 1985;7:211-216.

  118. Sherwin BB, Gelfand MM. The role of androgen in the maintenance of sexual function in oophorectomized women. Psychosom Med. 1987;49:397-409.

  119. Shifren J, Braunstein G, Sunon J, et al. Transdermal testosterone treatment in women with impaired sexual function after oophorectomy. N Engl J Med. 2000;343:682-688.

  120. Plouffe L Jr. Ovaries, androgens and the menopause: practical applications. Semin Reprod Endocrinol. 1998;16:117-120.

  121. Rosenberg M, King TD, Timmons MC. Estrogen-androgen for hormone replacement: a review. J Reprod Med, 1997;42:394-404.

  122. Smith DC, Prentice R, Thompson DJ, et al. Association of exogenous estrogen and endometrial carcinoma. N Engl J Med. 1975;293:1164-1167.

  123. Grady D, Rubin SM, Petitti DB, et al. Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann Intern Med. 1992;117:1016-1037.

  124. Chu J, Schweid AI, Weiss NS. Survival among women with endometrial cancer: a comparison of estrogen users and non-users. Am J Obstet Gynecol. 1982;143:569-573.

  125. Hammond CB, Jelovsck FR, Lee KL, et al. Effects of long-term estrogen replacement therapy. II. Neoplasia. Am J Obstet Gynecol. 1979;133:537-547.

  126. Whitehead MI, Townsend PT, Pryse-Davis J, et al. Effects of estrogens and progestins on the biochemistry and morphology of the menopausal endometrium. N Engl J Med. 1981;305:1599-1505.

  127. Archer DF, Picker JH, Bottiglioni F. Bleeding patterns in postmenopausal women taking continuous combined or sequential regimens of conjugated estrogens with medroxyprogesterone acetate. Obstet Gynecol. 1994;83:686-692.

  128. Jick H, Derby LE, Meyers WM, et al. Risk of hospital admission for idiopathic venous thromboembolism among users of postmenopausal oestrogens. Lancet. 1996;348:981-983.

  129. Perez Gutthann S, Garcia Rodriguez LA, Castellsaque J, et al. Hormone replacement therapy and risk of venous thromboembolism: population based case-control study. BMJ, 1997. 314: p. 796-800.

  130. Kroon UB, Silfverstolpe G, Tengborn L. The effects of transdermal estradiol and oral conjugated estrogens on haemostasis variables. Thromb Haemost. 1994;71:420-423.

  131. Pinto S, Bruni V, Rosati D, et al. Effects of estrogen replacement therapy on thrombin generation. Thromb Res. 1997;85:185-193.

  132. Boschetti C, Cortellaro M, Nencioni T, et al. Short- and long-term effects of hormone replacement therapy on blood coagulation factors in menopausal women. Thromb Res. 1991;62:1-8.

  133. Newcomb PA, Storer BE. Postmenopausal hormone use and risk of large-bowel cancer. J Natl Cancer Inst. 1995;87:1067-1071.

  134. Calle EE. Hormone replacement therapy and colorectal cancer: interpreting the evidence. Cancer Causes Control. 1997;8:127-129.

  135. Calle EE, Miracle-McMahill HL, Thun MJ, et al. Estrogen replacement therapy and risk of fatal colon cancer in a prospective cohort of postmenopausal women. J Natl Cancer Inst. 1995;87:517-523.

  136. McMichael AJ, Potter JD. Reproduction, endogenous and exogenous sex hormones and colon cancer: a review and hypothesis. J Natl Cancer Inst. 1980;65:1201-1207.

  137. Wald A, Van Thiel D, Hoechstetter L, et al. Gastrointestinal transit: the effect of the menstrual cycle. Gastroenterology. 1981;80:1497-1500.

  138. Meggouh F, Lointier P, Pezet D, et al. Status of sex steroid hormone receptors in large bowel cancer. Cancer. 1991;67:1964-1970.

  139. Trichopoulos D, MacMalon B. Menopause and breast cancer risk. J Natl Cancer Inst. 1972;48:605-613.

  140. Frisch RE. The right weight: body fat, menarche and fertility. Proc Nutr Soc. 1994;53:113-129.

  141. Clemmesen J. Statistical studies in malignant neoplasms. II. Basic Tables, Denmark 1943-1957. Acta Pathol Microbiol Scand. 1965;174:2-3.

  142. Collaborative Group on Hormone Factors in Breast Cancer. Breast cancer and hormone replacements: collaborative reanalysis of the data from 51 epidemiological studies of 52,705 women with breast cancer and 108, 411 women without breast cancer. Lancet. 1997;350:1047-1059.

  143. Schairer C, Rubin J, Troisi R, et al. Menopausal estrogen and estrogen-progestin replacement therapy and breast cancer risk. JAMA. 2000;283:485-491.

  144. Ross R, Paganini-Hill A, Wan P. Effect of hormone replacement therapy on breast cancer risk: estrogen versus estrogen plus progestin. J Natl Cancer Inst. 2000;92:328-332.

  145. Willis DB, Callee EE, Miracle-McMahill H, et al. Estrogen replacement therapy and risk of fatal breast cancer in a prospective cohort of postmenopausal women in the United States. Cancer Causes Control. 1996;7:449-457.

  146. Gapstur SM, Marrow M, Sellers TA. Hormone replacement therapy and risk of breast cancer with favorable histology. JAMA. 1999;281:2091-2097.

  147. Bergvist L, Adami HO, Persson I, et al. Prognosis after breast cancer diagnosis in women exposed to estrogen and estrogen-progesterone replacement therapy. Am J Epidemiol. 1989;130:221-228.

  148. Bonnier P, Romain S, Giacalone PL, et al. Clinical and biological prognostic factors in breast cancer diagnosed during postmenopausal hormone replacement therapy. Obstet Gynecol. 1995;85:11-17.

  149. Faiz O, Fentiman IS. Hormone replacement therapy and breast cancer. Int J Clin Pract. 1998;52:98-101.

  150. Colditz GA. Relationship between estrogen levels, use of hormone replacement therapy and breast cancer. J Natl Cancer Inst. 1998;90:814-823.

  151. Witt DM, Lousberg TR. Controversies surrounding estrogen use in postmenopausal women. Ann Pharmocother. 1997;31:745-755.

  152. Burger HG. The weighty issues of perimenopausal and menopausal hormone therapy. Med J Aust, 1998;168:206-207.

  153. Roy JA, Sawka CA, Pritchard KI. Hormone replacement therapy in women with breast cancer. Do the risks outweigh the benefits? J Clin Oncol. 1996;14:997-1006.

  154. Clarkson TB. Progestins and cardiovascular disease: a critical review. J Reprod Med. 1999;44:180-184.

  155. Haarbo J, Hassager C, Jensen SB, et al. Serum lipids, lipoproteins and apolipoproteins during postmenopausal estrogen replacement therapy combined with either 19-nortestosterone derivatives or 17-hydroxyprogesterone derivatives. Am J Med. 1991;90:584-589.

  156. Schairer C, Adani H, Hoover R, et al. Cause-specific mortality in women receiving hormone replacement therapy. Epidemiology. 1997;8:59-65.

  157. Grady D, Rubin SM, Petitti DB, et al. Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann Intern Med. 1992;117:1016-1037.

  158. Barrett-Connor E. Risk and benefit of replacement estrogen. Annu Rev Med. 1992;43:239-251.

  159. Hannaford PC. Is there sufficient evidence for us to encourage the widespread use of hormone replacement therapy to prevent disease? Br J Gen Pract. 1998;48:951-952.

  160. Dunne FP, Sadler GJ, Crampton JA. Hormone replacement therapy: why are general practitioners not prescribing more? Int J Clin Pract. 1998;52:285-288.

  161. Andersson K, Pedersen AT, Mattsson L. Swedish gynecologists' and general practitioners' views on the climacteric period: knowledge, attitudes and management strategies. Acta Obstet Gynecol Scand. 1998;77:909-916.

  162. Isaacs AJ, Britton AR, McPherson K. Utilization of hormone replacement therapy by women doctors. BMJ. 1995;311:1399-1401.

  163. Daly E, Vessey MP, Barlow D, et al. Hormone replacement therapy in a risk-benefit perspective. Maturitas. 1996;23:247-259.

  164. Daly E, Gray A, Barlow D, et al. Measuring the impact of menopausal symptoms on quality of life. BMJ. 1993;307:836-840.

  165. Hammond C. Menopause and hormone replacement therapy: an overview. Obstet Gynecol. 1996;87:2s-15s.

  166. O'Malley BW, Tsai SY, Bagchi M, et al. Molecular mechanisms of a steroid hormone receptor. Recent Prog Horm Res. 1991;47:1-26.

  167. Revelli A, Massobrio M, Tesarik J. Non-genomic actions of steroid hormones in reproductive tissues. Endocr Rev. 1998;19:3-17.

  168. Roper RJ, Griffith JS, Lyttle CR, et al. Interacting quantitative trait loci control phenotypic variation in murine estradiol-regulated responses. Endocrinology. 1999;140:556-561.

  169. Gorski J, Toft DO, Shyamala G, et al. Hormone receptors: studies on the interaction of estrogens with the uterus. Recent Prog Horm Res. 1968;24:45-80.

  170. Kuiper G, Carlsson B, Grandien K, et al. Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology. 1997;138:863-870.

  171. Stein G. Sacred Emily, In: Familiar Quotations. Boston, Mass: Little, Brown & Company; 1913.

  172. Bhavnani BR. Pharmocokinetics and pharmacodynamics of conjugated equine estrogens: chemistry and metabolism. Proc Soc Exp Biol Med. 1998;217:6-16.

  173. Mashchak CA, Lobo RA, Dozono-Takano R, et al. Comparison of pharmacodynamic properties of various estrogen formulations. Am J Obstet Gynecol. 1982;144:511-518.

  174. Wren BG, Brown LB, Routledge DA. Differential clinical response to oestrogens after menopause. Med J Aust. 1982;2:329-332.

  175. Heaney RP, Draper MW. Raloxifene and estrogen: comparative bone-remodeling kinetics. J Clin Endocrinol Metab. 1997;82:3425-3429.

  176. Fuleihan GE. Tissue-specific estrogens - the promise for the future. N Engl J Med. 1997;337:1648-1649.

  177. O'Connell MB. Pharmacokinetic and pharmacologic variation between different estrogen products. J Clin Pharm. 1995;35:18s-24s.

  178. Hyder SM, Shipley GL, Stancel GM. Estrogen action in target cells: selective requirements for activation of different hormone response elements. Mol Cell Endocrinol. 1995;112:35-43.

  179. Gordon J, Reagan JW, Finkle WD, et al. Estrogen and endometrial carcinoma. An independent pathology review supporting original risk estimate. N Engl J Med. 1977;297:507-511.

  180. Weiderpass E, Adami H, Baron J, et al. Risk of endometrial cancer following estrogen replacement with and without progestins. J Natl Cancer Inst. 1999;91:1131-1137.

  181. Simon JA, Robinson DE, Andrews MC, et al. The absorption of oral micronized progesterone: the effect of food, dose proportionality, and comparison with intramuscular progesterone. Fertil Steril. 1993;60:26-33.

  182. The Writing Group for the PEPI Trial. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA. 1995;273:199-208.

  183. McDonnell DP, Clemm DL, Herman T, et al. Analysis of estrogen receptor function in vitro reveals three distinct classes of antiestrogens. Mol Endocrinol. 1995;9:659-669.

  184. Yang NN, Venugopalan M, Hardikar S, et al. Identification of an estrogen response element activated by metabolites of 17ß-estradiol and raloxifene. Science. 1996;273:1222-1225.

  185. Willson TM, Norris JD, Wagner BL, et al. Dissection of the molecular mechanism of action of GW5638, a novel estrogen receptor ligand, provides insight into the role of estrogen receptor in bone. Endocrinology. 1997;138:3901-3911.

  186. Sato M, Rippy M, Bryant H. Raloxifene, tamoxifen, nafoxidine, or estrogen effects on reproductive and nonreproductive tissues in ovariectomized rats. FASEB J. 1996;10:905-912.

  187. Love RR, Mazess RB, Borden HS, et al. Effects of tamoxifen on bone mineral density in postmenopausal women with breast cancer. N Engl J Med. 1992;326:852-856.

  188. Delmas PD, Bjarnson NH, Mitlack BH, et al. Effects of raloxifene on bone mineral density, serum cholesterol concentrations and uterine endometrium in postmenopausal women. N Engl J Med. 1997;337:1641-1647.

  189. Clarkson TB. Soy phytoestrogens: what will be their role in postmenopausal hormone replacement therapy? Menopause 2000;7:71-75.

  190. Barns S. Phytoestrogens and breast cancer. Bailliere's Clin Endocrinol Metab. 1998;12:559-579.

  191. Albertazze P, Pansini F, Bonaccorsi G, et al. The effect of dietary soy supplementation on hot flashes. Obstet Gynecol. 1998;91:6-11.

  192. Anderson J, Garner S. Phytoestrogens and bone. Bailliere's Clin Endocrinol Metab. 1996;12:543-557.

  193. Whitehead MI, Hillard TC, Crook D. The role and use of progestogens. Obstet Gynecol. 1990;75(4 suppl):S59-S76.

  194. Greendale GA, Reboussin BA, Hogan P, et al. Symptom relief and side effects of postmenopausal hormones: results from the Postmenopausal Estrogen/Progestin Interventions Trial. Obstet Gynecol. 1998;92:982-988.

  195. Ettinger B, Selby J, Citron JT, et al. Cyclic hormone replacement therapy using quarterly progestin. Obstet Gynecol. 1994;83:693-700.

  196. Nachtigall LE. Enhancing patient compliance with hormone replacement therapy at menopause. Obstet Gynecol. 1990;75:77S-80S.

  197. Committee on Quality Assessment. Ambulatory care criteria set: hormone replacement therapy. Int J Gynecol Obstet. 1997;57:331-332.

  198. Dupont WD, Page DL. Menopausal estrogen and prevention bias. Ann Intern Med. 1991;115:455-456.

  199. Varma T, Everard D, Hole D. Effect of natural estrogen on the serum level of follicle-stimulating hormone (FSH), estradiol and estrone in post-menopausal women and its effect on endometrium. Acta Obstet Gynecol Scand. 1985;64:105-109.

  200. Mazess RB. Biochemical markers do not predict bone changes in individual patients in response to estrogen. Am J Med. 1998;104:80-81.

  201. Lindsay R, Hart DM, Clark DM. The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstet Gynecol. 1984;63:759-763.

  202. Rozenberg S, Vasquez JB, Vandromme J, et al. Educating patients about the benefits and drawbacks of hormone replacement therapy. Drugs Aging. 1998;13:33-41.

  203. Beard MK, Curtis LR. Libido, menopause, and estrogen replacement therapy. Postgrad Med. 1989;86:225-228.

  204. Barrett-Conner E, Timmons C, Young R, et al. Interim safety analysis of a two-year study comparing oral estrogen-androgen and conjugated estrogens in surgically menopausal women. J Womens Health. 1996;5:593-601.

  205. Sarrel PM. Cardiovascular aspects of androgens in women. Semin Reprod Endocrinol. 1998;16:121-128.

  206. Hickok LR, Toomey C, Speroff L. A comparison of esterified estrogens with and without methyltestosterone: effects on endometrial histology and serum lipoproteins in postmenopausal women. Obstet Gynecol. 1993;82:919-924.

  207. McKinney KA, Thompson W. A practical guide to prescribing hormone replacement therapy. Drugs. 1998;56:49-57.

  208. Baker DP. Estrogen-replacement therapy in patients with previous endometrial carcinoma. Compr Ther. 1990;16:28-35.

  209. DiSaia P, Grosen E, Kurosaki T, et al. Hormone replacement therapy in breast cancer survivors: a cohort study. Am J Obstet Gynecol. 1996;174:1494-1498.

  210. Braendle W. Hormone replacement therapy in women with breast cancer. Anticancer Res. 1998;18:2253-2256.

  211. Hartmann BW, Huber JC. The mythology of hormone replacement therapy. Br J Obstet Gynaecol. 1997;104:163-168.

  212. Berga SL. Hormonal management of the sick menopausal woman. Obstet Gynecol Clin North Am. 1994;21:231-244.

  213. Cauley JA, Cummings SR, Black DM, et al. Prevalence and determinants of estrogen replacement therapy in elderly women. Am J Obstet Gynecol. 1990;163:1438-1444.

  214. Moorhead T, Hannaford P, Warskyj M. Prevalence and characteristics associated with use of hormone replacement therapy in Britain. Br J Obstet Gynaecol, 1997;104:290-297.

  215. Marmoreo J, Brown JB, Batty HR, Cummings S, Powell M. Hormone replacement therapy: determinants of women's decisions. Patient Educ Couns. 1998;33:289-298.

  216. Kaufert P, Boggs P, Ettinger B, et al. Women and menopause: beliefs, attitudes, and behaviors. The North American Menopause Society 1997 Menopause Survey. Menopause. 1998;5:197-202.

  217. Gallup Survey. 1995. Unpublished

  218. Wren BG. Megatrials of hormone replacement therapy. Drugs & Aging. 1998;12:343-348.

  219. Lobo RA. Absorption and metabolic effects of different types of estrogens and progestogens. Obstet Gynecol Clin North Am. 1987;14:143-167.