Chapter 16: Menopause and Postmenopausal Hormone Therapy

Menopause is the permanent cessation of menstruation due to loss of ovarian follicular function. It is diagnosed retrospectively after 12 months of amenorrhea. The average age at menopause is 51 years among U.S. women. Perimenopause refers to the time period preceding menopause, when fertility wanes and menstrual cycle irregularity increases, until the first year after cessation of menses. The onset of perimenopause precedes the final menses by 2–8 years, with a mean duration of 4 years. Smoking accelerates the menopausal transition by 2 years.

Although the peri- and postmenopausal transitions share many symptoms, the physiology and clinical management of the two differ. Low-dose oral contraceptives have become a therapeutic mainstay in perimenopause, whereas postmenopausal hormone therapy (HT) has been a common method of symptom alleviation after menstruation ceases.

PHYSIOLOGY

Ovarian mass and fertility decline sharply after age 35 and even more precipitously during perimenopause; depletion of primary follicles, a process that begins before birth, occurs steadily until menopause (Chap. 13). In perimenopause, intermenstrual intervals shorten significantly (typically by 3 days) as a result of an accelerated follicular phase. Follicle-stimulating hormone (FSH) levels rise because of altered folliculogenesis and reduced inhibin secretion. In contrast to the consistently high FSH and low estradiol levels seen in menopause, perimenopause is characterized by “irregularly irregular” hormone levels. The propensity for anovulatory cycles can produce a hyperestrogenic, hypoprogestagenic environment that may account for the increased incidence of endometrial hyperplasia or carcinoma, uterine polyps, and leiomyoma observed among women of perimenopausal age. Mean serum levels of selected ovarian and pituitary hormones during the menopausal transition are shown in Fig. 16-1. With transition into menopause, estradiol levels fall markedly, whereas estrone levels are relatively preserved, a pattern reflecting peripheral aromatization of adrenal and ovarian androgens. Levels of FSH increase more than those of luteinizing hormone, presumably because of the loss of inhibin as well as estrogen feedback.

DIAGNOSTIC TESTS

The Stages of Reproductive Aging Workshop +10 (STRAW+10) classification provides a comprehensive framework for the clinical assessment of ovarian aging. As shown in Fig. 16-2, menstrual cycle characteristics are the principal criteria for characterizing the menopausal transition, with biomarker measures as supportive criteria. Because of their extreme intraindividual variability, FSH and estradiol levels are imperfect diagnostic indicators of perimenopause in menstruating women. However, a consistently low FSH level in the early follicular phase (days 2–5) of the menstrual cycle does not support a diagnosis of perimenopause, while levels >25 IU/L in a random blood sample are characteristic of the late menopause transition. FSH measurement can also aid in assessing fertility; levels of <20 IU/L, 20 to <30 IU/L, and ≥30 IU/L measured on day 3 of the cycle indicate a good, fair, and poor likelihood of achieving pregnancy, respectively. Antimüllerian hormone and inhibin B may also be useful for assessing reproductive aging.

SYMPTOMS

Determining whether symptoms that develop in midlife are due to ovarian senescence or to other age-related changes is difficult. There is strong evidence that the menopausal transition can cause hot flashes, night sweats, irregular bleeding, and vaginal dryness, and there is moderate evidence that it can cause sleep disturbances in some women. There is inconclusive or insufficient evidence that ovarian aging is a major cause of mood swings, depression, impaired memory or concentration, somatic symptoms, urinary incontinence, or sexual dysfunction. In one U.S. study, nearly 60% of women reported hot flashes in the 2 years before their final menses. Symptom intensity, duration, frequency, and effects on quality of life are highly variable.

TREATMENT

One of the most complex health care decisions facing women is whether to use postmenopausal HT. Once prescribed primarily to relieve vasomotor symptoms, HT has been promoted as a strategy to forestall various disorders that accelerate after menopause, including osteoporosis and cardiovascular disease. In 2000, nearly 40% of postmenopausal women age 50–74 in the United States had used HT. This widespread use occurred despite the paucity of conclusive data, until recently, on the health consequences of such therapy. Although many women rely on their health care providers for a definitive answer to the question of whether to use postmenopausal hormones, balancing the benefits and risks for an individual patient is challenging.

Although observational studies suggest that HT prevents cardiovascular and other chronic diseases, the apparent benefits may result at least in part from differences between women who opt to take postmenopausal hormones and women who do not. Those choosing HT tend to be healthier, have greater access to medical care, are more compliant with prescribed treatments, and maintain a more health-promoting lifestyle. Randomized trials, which eliminate these confounding factors, have not consistently confirmed the benefits found in observational studies. Indeed, the largest HT trial to date, the Women’s Health Initiative (WHI), which examined more than 27,000 postmenopausal women age 50–79 (mean age, 63) for an average of 5–7 years, was stopped early because of an overall unfavorable benefit-risk ratio in the estrogen-progestin arm and an excess risk of stroke that was not offset by a reduced risk of coronary heart disease (CHD) in the estrogen-only arm.

The following summary offers a decision-making guide based on a synthesis of currently available evidence. Prevention of cardiovascular disease is eliminated from the equation due to lack of evidence for such benefits in recent randomized clinical trials.

BENEFITS AND RISKS OF POSTMENOPAUSAL HORMONE THERAPY

See Table 16-1.

Definite benefits

Symptoms of menopause

Compelling evidence, including data from randomized clinical trials, indicates that estrogen therapy is highly effective for controlling vasomotor and genitourinary symptoms. Alternative approaches, including the use of antidepressants (such as paroxetine, 7.5 mg/d; or venlafaxine, 75–150 mg/d), gabapentin (300–900 mg/d), clonidine (0.1–0.2 mg/d), or vitamin E (400–800 IU/d), or the consumption of soy-based products or other phytoestrogens, may also alleviate vasomotor symptoms, although they are less effective than HT. Paroxetine is the only nonhormonal drug approved by the U.S. Food and Drug Administration for treatment of vasomotor symptoms. Bazedoxifene, an estrogen agonist/antagonist, in combination with conjugated estrogens has also received approval for vasomotor symptom management. For genitourinary symptoms, the efficacy of vaginal estrogen is similar to that of oral or transdermal estrogen; oral ospemifene is an additional option.

Osteoporosis

(See also Chap. 35)

Bone density

By reducing bone turnover and resorption rates, estrogen slows the aging-related bone loss experienced by most postmenopausal women. More than 50 randomized trials have demonstrated that postmenopausal estrogen therapy, with or without a progestogen, rapidly increases bone mineral density at the spine by 4–6% and at the hip by 2–3% and that those increases are maintained during treatment.

Fractures

Data from observational studies indicate a 50–80% lower risk of vertebral fracture and a 25–30% lower risk of hip, wrist, and other peripheral fractures among current estrogen users; addition of a progestogen does not appear to modify this benefit. In the WHI, 5–7 years of either combined estrogen-progestin or estrogen-only therapy was associated with a 33% reduction in hip fractures and 25–30% fewer total fractures among a population unselected for osteoporosis. Bisphosphonates (such as alendronate, 10 mg/d or 70 mg once per week; risedronate, 5 mg/d or 35 mg once per week; or ibandronate, 2.5 mg/d or 150 mg once per month or 3 mg every 3 months IV) and raloxifene (60 mg/d), a selective estrogen receptor modulator (SERM), have been shown in randomized trials to increase bone mass density and decrease fracture rates. Other options for treatment of osteo­porosis are bazedoxifene in combination with conjugated estrogens and parathyroid hormone (teriparatide, 20 μg/d SC). These agents, unlike estrogen, do not appear to have adverse effects on the endometrium or breast. Increased physical activity, adequate calcium intake (1000–1200 mg/d through diet or supplements in two or three divided doses), and adequate vitamin D intake (600–1000 IU/d) may also reduce the risk of osteoporosis-related fractures. According to the Institute of Medicine’s 2011 report, 25-hydroxyvitamin D blood levels of ≥50 nmol/L are sufficient for bone-density maintenance and fracture prevention. The Fracture Risk Assessment (FRAX^¯) score, an algorithm that combines an individual’s bone-density score with age and other risk factors to predict her 10-year risk of hip and major osteoporotic fracture, may be of use in guiding decisions about ­pharmacologic treatment (see www.shef.ac.uk/FRAX/).

Definite risks

Endometrial cancer (with estrogen alone)

A combined analysis of 30 observational studies found a tripling of endometrial cancer risk among short-term users (1–5 years) of unopposed estrogen and a nearly tenfold increased risk among long-term users (≥10 years). These findings are supported by results from the randomized Postmenopausal Estrogen/Progestin Interventions (PEPI) trial, in which 24% of women assigned to unopposed estrogen for 3 years developed atypical endometrial hyperplasia—a premalignant lesion—as opposed to only 1% of women assigned to placebo. Use of a progestogen, which opposes the effects of estrogen on the endometrium, eliminates these risks and may even reduce risk (see later).

Venous thromboembolism

A meta-analysis of observational studies found that current oral estrogen use was associated with a 2.5-fold increase in risk of venous thromboembolism in postmenopausal women. A meta-analysis of randomized trials, including the WHI, found a 2.1-fold increase in risk. Results from the WHI indicate a nearly twofold increase in risk of pulmonary embolism and deep vein thrombosis with estrogen-progestin and a 35–50% increase in these risks with estrogen-only therapy. Transdermal estrogen, taken alone or with certain progestogens (micronized progesterone or pregnane derivatives), appears to be a safer alternative with respect to thrombotic risk.

Breast cancer (with estrogen-progestin)

An increased risk of breast cancer has been found among current or recent estrogen users in observational studies; this risk is directly related to duration of use. In a meta-analysis of 51 case-control and cohort studies, short-term use (<5 years) of postmenopausal HT did not appreciably elevate breast cancer incidence, whereas long-term use (≥5 years) was associated with a 35% increase in risk. In contrast to findings for endometrial cancer, combined estrogen-progestin regimens appear to increase breast cancer risk more than estrogen alone. Data from randomized trials also indicate that estrogen-progestin raises breast cancer risk. In the WHI, women assigned to receive combination hormones for an average of 5.6 years were 24% more likely to develop breast cancer than women assigned to placebo, but 7.1 years of estrogen-only therapy did not increase risk. Indeed, the WHI showed a trend toward a reduction in breast cancer risk with estrogen alone, although it is unclear whether this finding would pertain to formulations of estrogen other than conjugated equine estrogens or to treatment durations of >7 years. In the Heart and Estrogen/Progestin Replacement Study (HERS), 4 years of combination therapy was associated with a 27% increase in breast cancer risk. Although the latter finding was not statistically significant, the totality of evidence strongly implicates estrogen-progestin therapy in breast carcinogenesis.

Some observational data suggest that the length of the interval between menopause onset and HT initiation may influence the association between such therapy and breast cancer risk, with a “gap time” of <3–5 years conferring a higher HT-associated breast cancer risk. (This pattern of findings contrasts with that for CHD, as discussed later in this chapter.) However, this association remains inconclusive and may be a spurious finding attributable to higher rates of screening mammography and thus earlier cancer detection in HT users than in nonusers, especially in early menopause. Indeed, in the WHI trial, hazard ratios for HT and breast cancer risk did not differ among women 50–59, those 60–69, and those 70–79 years of age at trial entry. (There was insufficient power to examine finer age categories.) Additional research is needed to clarify the issue.

Gallbladder disease

Large observational studies report a two- to threefold increased risk of gallstones or cholecystectomy among postmenopausal women taking oral estrogen. In the WHI, women randomized to estrogen-progestin or estrogen alone were ∼55% more likely to develop gallbladder disease than those assigned to placebo. Risks were also increased in HERS. Transdermal HT might be a safer alternative, but further research is needed.

Probable or uncertain risks and benefits

Coronary heart disease/stroke

Until recently, HT had been enthusiastically recommended as a possible cardioprotective agent. In the past three decades, multiple observational studies suggested, in the aggregate, that estrogen use leads to a 35–50% reduction in CHD incidence among postmenopausal women. The biologic plausibility of such an association is supported by data from randomized trials demonstrating that exogenous estrogen lowers plasma low-density lipoprotein (LDL) cholesterol levels and raises high-density lipoprotein (HDL) cholesterol levels by 10–15%. Administration of estrogen also favorably affects lipoprotein(a) levels, LDL oxidation, endothelial vascular function, fibrinogen, and plasminogen activator inhibitor 1. However, estrogen therapy has unfavorable effects on other biomarkers of cardiovascular risk: it boosts triglyceride levels; promotes coagulation via factor VII, prothrombin fragments 1 and 2, and fibrinopeptide A elevations; and raises levels of the inflammatory marker C-reactive protein.

Randomized trials of estrogen or combined estrogen-progestin in women with preexisting cardiovascular disease have not confirmed the benefits reported in observational studies. In HERS (a secondary-prevention trial designed to test the efficacy and safety of estrogen-progestin therapy with regard to clinical cardiovascular outcomes), the 4-year incidence of coronary death and nonfatal myocardial infarction was similar in the active-treatment and placebo groups, and a 50% increase in risk of coronary events was noted during the first year among participants assigned to the active-treatment group. Although it is possible that progestin may mitigate estrogen’s benefits, the Estrogen Replacement and Atherosclerosis (ERA) trial indicated that angiographically determined progression of coronary atherosclerosis was unaffected by either opposed or unopposed estrogen treatment. Moreover, no cardiovascular benefit was found in the Papworth Hormone Replacement Therapy Atherosclerosis Study, a trial of transdermal estradiol with and without norethindrone; the Women’s Estrogen for Stroke Trial (WEST), a trial of oral 17β-estradiol; or the Estrogen in the Prevention of Reinfarction Trial (ESPRIT), a trial of oral estradiol valerate. Thus, in clinical trials, HT has not proved effective for the secondary prevention of cardiovascular disease in postmenopausal women.

Primary-prevention trials also suggest an early increase in cardiovascular risk and an absence of cardioprotection with postmenopausal HT. In the WHI, women assigned to 5.6 years of estrogen-progestin therapy were 18% more likely to develop CHD (defined in primary analyses as nonfatal myocardial infarction or coronary death) than those assigned to placebo, although this risk elevation was not statistically significant. However, during the trial’s first year, there was a significant 80% increase in risk, which diminished in subsequent years (p for trend by time = .03). In the estrogen-only arm of the WHI, no overall effect on CHD was observed during the 7.1 years of the trial or in any specific year of follow-up. This pattern of results was similar to that for the outcome of total myocardial infarction.

However, a closer look at available data suggests that timing of initiation of HT may critically influence the association between such therapy and CHD. Estrogen may slow early stages of atherosclerosis but have adverse effects on advanced atherosclerotic lesions. It has been hypothesized that the prothrombotic and proinflammatory effects of estrogen manifest themselves predominantly among women with subclinical lesions who initiate HT well after the menopausal transition, whereas women with less arterial damage who start HT early in menopause may derive cardiovascular benefit because they have not yet developed advanced lesions. Nonhuman primate data support this concept. Conjugated estrogens had no effect on the extent of coronary artery plaque in cynomolgus monkeys assigned to receive estrogen alone or combined with progestin starting 2 years (∼6 years in human terms) after oophorectomy and well after the establishment of atherosclerosis. However, administration of exogenous hormones immediately after oophorectomy, during the early stages of atherosclerosis, reduced the extent of plaque by 70%.

Lending further credence to this hypothesis are results of subgroup analyses of observational and clinical trial data. For example, among women who entered the WHI trial with a relatively favorable cholesterol profile, estrogen with or without progestin led to a 40% lower risk of incident CHD. Among women who entered with a worse cholesterol profile, therapy resulted in a 73% higher risk (p for interaction = .02). The presence or absence of the metabolic syndrome (Chap. 22) also strongly influenced the relation between HT and incident CHD. Among women with the metabolic syndrome, HT more than doubled CHD risk, whereas no association was observed among women without the syndrome. Moreover, although there was no association between estrogen-only therapy and CHD in the WHI trial cohort as a whole, such therapy was associated with a CHD risk reduction of 40% among participants age 50–59; in contrast, a risk reduction of only 5% was observed among those age 60–69, and a risk increase of 9% was found among those age 70–79 (p for trend by age = .08). For the outcome of total myocardial infarction, estrogen alone was associated with a borderline-significant 45% reduction and a nonsignificant 24% increase in risk among the youngest and oldest women, respectively (p for trend by age = .02). Estrogen was also associated with lower levels of coronary artery calcified plaque in the younger age group. Although age did not have a similar effect in the estrogen-progestin arm of the WHI, CHD risks increased with years since menopause (p for trend = .08), with a significantly elevated risk among women who were ≥20 years past menopause. For the outcome of total myocardial infarction, estrogen-progestin was associated with a 9% risk reduction among women <10 years past menopause as opposed to a 16% increase in risk among women 10–19 years past menopause and a twofold increase in risk among women >20 years past menopause (p for trend = .01). In the large observational Nurses’ Health Study, women who chose to start HT within 4 years of menopause experienced a lower risk of CHD than did nonusers, whereas those who began therapy ≥10 years after menopause appeared to receive little coronary benefit. Observational studies include a high proportion of women who begin HT within 3–4 years of menopause, whereas clinical trials include a high proportion of women ≥12 years past menopause; this difference helps to reconcile some of the apparent discrepancies between the two types of studies.

For the outcome of stroke, WHI participants assigned to estrogen-progestin or estrogen alone were ∼35% more likely to suffer a stroke than those assigned to placebo. Whether or not age at initiation of HT influences stroke risk is not well understood. In the WHI and the Nurses’ Health Study, HT was associated with an excess risk of stroke in all age groups. Further research is needed on age, time since menopause, and other individual characteristics (including biomarkers) that predict increases or decreases in cardiovascular risk associated with exogenous HT. Furthermore, it remains uncertain whether different doses, formulations, or routes of administration of HT will produce different cardiovascular effects.

Colorectal cancer

Observational studies have suggested that HT reduces risks of colon and rectal cancer, although the estimated magnitudes of the relative benefits have ranged from 8% to 34% in various meta-analyses. In the WHI (the sole trial to examine the issue), estrogen-progestin was associated with a significant 38% reduction in colorectal cancer over a 5.6-year period, although no benefit was seen with 7 years of estrogen-only therapy. However, a modifying effect of age was observed, with a doubling of risk with HT in women age 70–79 but no risk elevation in younger women (p for trend by age = .02).

Cognitive decline and dementia

A meta-analysis of 10 case-control and two cohort studies suggested that postmenopausal HT is associated with a 34% decreased risk of dementia. Subsequent randomized trials (including the WHI), however, have failed to demonstrate any benefit of estrogen or estrogen-progestin therapy on the progression of mild to moderate Alzheimer’s disease and/or have indicated a potential adverse effect of HT on the incidence of dementia, at least in women ≥65 years of age. Among women randomized to HT (as opposed to placebo) at age 50–55 in the WHI, no effect on cognition was observed during the postintervention phase. Determining whether timing of initiation of HT influences cognitive outcomes will require further study.

Ovarian cancer and other disorders

On the basis of limited observational and randomized data, it has been hypothesized that HT increases the risk of ovarian cancer and reduces the risk of type 2 diabetes mellitus. Results from the WHI support these hypotheses. The WHI also found that HT use was associated with an increased risk of urinary incontinence and that estrogen-progestin was associated with increased rates of lung cancer mortality.

Endometrial cancer (with estrogen-progestin)

In the WHI, use of estrogen-progestin was associated with a nonsignificant 17% reduction in risk of endometrial cancer. A significant reduction in risk emerged during the postintervention period (see later).

All-cause mortality

In the overall WHI cohort, estrogen with or without progestin was not associated with all-cause mortality. However, there was a trend toward reduced mortality in younger women, particularly with estrogen alone. For women 50–59, 60–69, and 70–79 years of age, relative risks (RRs) associated with estrogen-only therapy were 0.70, 1.01, and 1.21, respectively (p for trend = .04).

Overall benefit-risk profile

Estrogen-progestin was associated with an unfavorable benefit-risk profile (excluding relief from menopausal symptoms) as measured by a “global index”—a composite outcome including CHD, stroke, pulmonary embolism, breast cancer, colorectal cancer, endometrial cancer, hip fracture, and death (Table 16-1)—in the WHI cohort as a whole, and this association did not vary by 10-year age group. Estrogen-only therapy was associated with a neutral benefit-risk profile in the WHI cohort as a whole. However, there was a significant trend toward a more favorable benefit-risk profile among younger women and a less favorable profile among older women, with RRs of 0.84, 0.99, and 1.17 for women 50–59, 60–69, and 70–79 years of age, respectively (p for trend by age = .02).

Changes in health status after discontinuation of hormone therapy

In the WHI, many but not all risks and benefits associated with active use of HT dissipated within 5–7 years after discontinuation of therapy. For estrogen-progestin, an elevated risk of breast cancer persisted (RR = 1.28 [95% confidence interval, 1.11–1.48]) during a cumulative 12-year follow-up period (5.6 years of treatment plus 6.8 years of postintervention observation), but most cardiovascular disease risks became neutral. A reduction in hip fracture risk persisted (RR = 0.81 [0.68–0.97]), and a significant reduction in endometrial cancer risk emerged (RR = 0.67 [0.49–0.91]). For estrogen alone, the reduction in breast cancer risk became statistically significant (RR = 0.79 [0.65–0.97]) during a cumulative 12-year follow-up period (6.8 years of treatment plus 5.1 years of postintervention observation), and significant differences by age group persisted for total myocardial infarction and the global index, with more favorable results for younger women.

APPROACH TO PATIENT