The most effective treatment is estrogen, which may be given alone to women without a uterus or combined with a progestogen in women with a uterus. As with any medicine, these hormone treatments are not trouble-free, but alternatives such as herbal remedies, natural foods, and neuroactive drugs are less effective. Given that estrogen with or without progestogen is the best therapy, the challenge lies in finding ways to optimize the benefits and minimize the risks of hormone treatment.
Optimizing the Benefits—When and What to Prescribe
The most effective treatment is estrogen, which may be given alone to women without a uterus or combined with a progestogen in women with a uterus. As with any medicine, these hormone treatments are not trouble-free, but alternatives such as herbal remedies, natural foods, and neuroactive drugs are less effective. Given that estrogen with or without progestogen is the best therapy, the challenge lies in finding ways to optimize the benefits and minimize the risks of hormone treatment.
Optimizing the Benefits—When and What to Prescribe
Hormones are usually prescribed when symptoms are disturbing enough to justify the bother and the cost. Hormone treatment reduces the severity of hot flashes by nearly 80%2 and their frequency by about 75% (which is equal to about 18 fewer flashes per week with hormones than with placebo).3 Hormones may improve secondary symptoms, such as forgetfulness and difficulty concentrating, but these effects are usually noticed only in women who have hot flashes to begin with.2
When to Prescribe
The majority of women will begin taking hormones soon after the menopause, which occurs on average at around 51 years of age. Hormones should be used only for as long as necessary, which is usually two to three years. As some rare adverse risks increase after five years of usage, the majority of women who take hormones will be below 60 years of age.
What to Prescribe
Hormone prescriptions should specify the lowest effective dose. Whether the estrogen is synthetic or natural does not influence the effect on hot flashes and sleeplessness.4,5 The effect on symptoms is not influenced by adding a progestogen, whether it is taken every day or in a cyclic pattern.1 Hormones can be taken orally or absorbed through the skin from a patch without altering the effect on symptoms.4,5
Vaginal Dryness
Estrogen is an effective treatment for this symptom of estrogen deficiency and for the sexual discomfort that may occur.6 For this problem, applying creams or gels directly to the vulva and vagina is more effective than pills or skin patches. Minimizing Side Effects
Irregular bleeding occurs in about 15% of cycles, and may be a reason for women to stop using hormone treatment.7,8 Bleeding is less likely with daily than with cyclic progestogen, and the chance of bleeding decreases during the first year of treatment. As bleeding may be distressing for a post-menopausal woman, it is helpful to make new users aware of the possibility that it may occur.
Breast tenderness is more likely with combined estrogen–progestogen than with estrogen alone.9 For every 21 women treated with progestin formulations for three years, one will have breast symptoms attributable to the estrogen–progestogen hormone.1
Muscular aches and pains, joint pain, muscle stiffness, and headaches are common symptoms with or without hormone treatment. All but headache may be improved during treatment with estrogens and progestogens.2 Women are also concerned about weight gain during use of hormone treatment, but weight gain is similar among women who are not taking hormones.2
The figure shows estrogen (left side) and estrogen–progestogen (right side) risks of myocardial infarction or coronary-artery-related death according to age at treatment (top) and number of years after menopause that treatment was started (bottom). Relative hazard (RH) corresponds to the relative rate of coronary artery events in hormone groups compared with placebo groups. The circles are proportional to the number of events and size of the subgroup. Horizontal bars are 95% confidence intervals (CIs). CHD = coronary heart disease
Potential Long-term Benefits
Epidemiological studies published before the Women’s Health Initiative (WHI) trials suggested that hormone treatment during the menopause reduced the risk for osteoporosis, senile dementia, coronary heart disease (CHD), and colon cancer. However, epidemiological studies are susceptible to bias, while the WHI trials involved random allocation, which minimizes bias arising from baseline differences between the hormone and placebo groups. Evidence from randomized controlled trials (RCTs) is consistent with the epidemiological evidence only for osteoporosis and colon cancer. Prevention of Osteoporosis and Fractures
Osteoporosis is associated with 1.5 million fractures per year in the US. Hip fracture is the most severe fracture, but is uncommon in women below 60 years of age, who suffer less than one hip fracture per 1,000 women per year.10 Hormone treatment can prevent osteoporosis, but most fractures are due to trauma from a fall rather than osteoporosis.11 Estrogen with or without progestogen reduces the likelihood of hip fracture by about one-quarter.12,13 The WHI trials were the first to show a significant overall reduction in fractures with any kind of treatment in women who were not known to be at high risk for osteoporotic fracture. Nevertheless, the absolute effect of estrogen or estrogen–progestogen on hip fracture incidence is small, involving less than one hip fracture per 1,000 women per year. Hormone treatment for the prevention of osteoporosis is not warranted in women who are typically low-risk and who are below 60 years of age. Non-prescription alternatives including exercise and vitamin D have a better risk–benefit ratio.
Memory Impairment and Senile Dementia
More than 33% of women above 65 years of age will develop dementia during their lifetime.14 In a summary of epidemiological studies, hormone treatment seemed to reduce the risk for dementia by 34%.15 However, these studies may be affected by bias, as women with dementia may not recall whether they took hormones. The WHI Memory Study (WHIMS) involved only women above 65 years of age, and neither estrogen alone nor estrogen–progestogen prevented the mild impairment of memory.16–18 In the estrogen trial, 28 estrogen patients and 19 placebo patients developed dementia, a 1.5-fold increase in risk for dementia (95% confidence interval [CI] 0.8–2.7).17 With estrogen–progestin, the dementia risk was 2.1-fold higher (95% CI 1.21–3.48) than with placebo.16 Thus, the best available evidence does not support a role for hormone treatment in the prevention of memory impairment or dementia.
Coronary Heart Disease
Cardiovascular disease is the leading cause of death in post-menopausal women. CHD rates are lower in pre-menopausal women than in men of a comparable age, but the incidence rises after the menopause to approximately three to five cases per 1,000 per year in low-risk women.12,13 One mechanism for the pre-menopausal protective effect in women is the estrogen-mediated prevention of lipid deposition in the walls of the coronary arteries. Epidemiological studies suggested that hormone treatment reduced the risk for myocardial infarction (MI) by 44% (relative rate 0.56, 95% CI 0.51–0.61) compared with no use.19,20
Primary prevention is the treatment of women who do not have heart disease with the aim of diminishing the risk for developing heart disease. The primary prevention trials did not show the benefits that were expected from the epidemiological study results for either estrogen alone or estrogen–progestogen.13,21 In the WHI estrogen-alone trial, the likelihood of CHD (relative rate 0.91, 95% CI 0.75–1.12) was unaffected.13 In the WHI estrogen–progestin trial, there was a small but significant increase in CHD risk (relative rate 1.24, 95% CI 1–1.54).21 The increased CHD risk occurred despite favorable effects on lipids.
While estrogen in younger women appears to protect against the development of lipid plaques on the coronary artery walls, once plaques have developed after the menopause estrogen reduces the integrity of the capsule, making rupture of the plaque more likely. Estrogen also increases the chance that a clot will form when rupture occurs, and the clot may block the lumen of the coronary artery, causing an MI.22 Thus, in women who have developed lipid plaques many years after the menopause, estrogen treatment may increase their chance of an MI. Theoretically, starting to take hormones soon after the menopause and before lipid plaques form is more beneficial than beginning treatment 10 years later.The WHI addressed the issue of when to start hormone treatment in an analysis of both hormone trials.22 CHD risks with either estrogen or estrogen–progestogen were lower in younger women and those who began treatment less than 10 years after the menopause, although the trends were only significant in the combined data and for the timing of hormone treatment after the menopause, but not for age.22 However, a significantly lower risk in younger women was shown in a study that combined data from 30 trials involving a total of 26,708 women.23 In the trials of women with a mean age of <60 years, there were 14 deaths per 10,000 women, and in the trials on women with a mean age of >60 years, there were 175 deaths per 10,000 women. Hormone treatment reduced total mortality in trials among younger women (relative rate 0.61, CI 0.3–0.95), but not in those among older women (relative rate 1.03, CI 0.90–1.18). As the likelihood of having a MI is lower among younger women, the number of deaths that can be prevented is small: the data from the 30 trials indicate that approximately five deaths would be prevented by giving estrogen to 10,000 younger women (calculated as 1 minus the 0.61 relative rate = 0.39, and 0.39 x 14 deaths = 5.46).
Secondary Prevention
Among women with a history of MI or clinically significant coronary artery disease, the secondary prevention of CHD is an important goal because clinical events are more frequent in women with heart disease. The Heart and Estrogen/Progestin Replacement (HERS) secondary prevention trial involved 2,763 post-menopausal women aged 55 and 80 years of age who had coronary artery disease and an intact uterus.24 Treatment over an average of four years with oral estrogen and progestin had no effect on the likelihood of another MI (relative rate 0.99, 95% CI 0.80–1.22) compared with placebo. The Estrogen in the Prevention of Re-infarction Trial (ESPRIT) included 1,017 post-menopausal women between 50 and 69 years of age who had a recent first MI. Here, hormone treatment also had no effect on the frequency of another MI compared with placebo (relative rate 0.99, 95% CI 0.70–1.41).25 These results show that hormones are not indicated for the secondary prevention of MI in women with CHD.
Summary of Coronary Heart Disease Prevention Studies
Reliable evidence indicates that hormones are not indicated for either the primary or secondary prevention of CHD events. Alternative strategies to maintain a healthy weight and fitness should be considered. The WHI and other trials indicate that CHD risks are small among women within 10 years of the menopause or women below 60 years of age. There is little evidence that MI risk depends on doses of hormones, different formulations, or routes of hormone administration in a review of products used in recent studies. Colon Cancer
The incidence of colorectal cancer in post-menopausal women is approximately 1.6 cases per year per 1,000 women.12,13 Age, family history, and diet are risk factors, and the use of oral contraception and hormones has been associated with reduced risks.26 Data from two of three trials are consistent with a reduced risk for colon cancer in users of hormones. In the HERS follow-up study, the risk was reduced by about 20% (relative rate 0.81, 95% CI 0.46–1.45).27 In the WHI estrogen–progestogen trial, colon cancer was more than 40% less likely in the hormone group compared with the placebo group (relative rate 0.63, 95% CI 0.43–0.92).12 However, in the WHI estrogen-only trial, the risk was not reduced (relative rate 1.08; 95% CI 0.63–1.86).13 However, a reduced risk for colon cancer would not constitute an indication to prescribe hormones.
Disorders That May Be More Frequent with Hormone Treatment
As with other medicinal drugs, hormone use has the potential to cause adverse effects, although the short-term use that is sufficient for most patients is safe. The main focus of studies has been on cardiovascular diseases such as stroke and venous thromboembolism, and cancers of the breast, endometrium, and ovary.
Stroke
The incidence of stroke among post-menopausal women aged between 50 and 59 years is approximately eight per 1,000 per year,22 and approximately 75% of strokes are ischemic.12,13,28,29 No consistent hormone effect on stroke risk was noted in 29 different epidemiological studies using different stroke end-points and hormone definitions.30 Stroke risk with estrogen alone was significantly increased in one of three relevant trials. In the ESPRIT study, among women who had a history of MI stroke risk with estrogen alone was not significant (relative rate 1.64, 95% CI 0.60–4.47) compared with placebo.25 In the Women’s Estrogen for Stroke Trial (WEST), estrogen alone did not affect the risk for stroke over 2.8 years of follow-up (relative rate 1.1, 95% CI 0.8–1.6).31 The WHI estrogen-alone trial reported a small increase in stroke risk (relative rate 1.33; adjusted 95% CI 1.05–1.68).22 With respect to estrogen– progestogen use and stroke risk, a small increase in the risk for ischemic stroke with estrogen–progestogen was not significant in the HERS trial (relative rate 1.18, 95% CI 0.84–1.43) compared with placebo. The WHI estrogen–progestogen trial reported a modest significant increase in risk for any stroke (relative rate 1.31, 95% CI 1.08–1.68).22 In the combined WHI analysis of both hormone trials, the number of strokes due to hormone use over and above the number that would be experienced by non-users was 9.3 cases (95% CI 3.4–15.13).22 Venous Thromboembolism
Venous thromboembolism is a term that includes thrombophlebitis in deep veins and pulmonary embolism. The incidence among post-menopausal women is about one per 1,000 women per year.12,13 Data from epidemiological studies and RCTs consistently demonstrate an increased risk for thromboembolism associated with post-menopausal hormone use.32,33 The increased risk declines from approximately four-fold in the first year of use to less than two-fold after the third year of use.32,33 The WHI studies confirmed th
venous thromboembolism risk estimates from the epidemiological studies. In the estrogen-alone trial, the risk increase was not significant (relative rate 1.32, 95% CI 0.99–1.75), and it was highest in the first two years.32 There were two more cases of venous thrombosis per 1,000 women per year in the estrogen group than in the placebo group. The WHI estrogen–progestogen trial reported a two-fold increase in venous thromboembolism risk (relative rate 2.11, 95% CI 1.26–3.55).12
Breast Cancer
Breast cancer incidence in post-menopausal women is approximately three cases per 1,000 women per year.12 An association between breast cancer and hormone use is plausible because breast cancer incidence is increased by hormonal factors such as early menarche and late menopause.34 The combined results of 51 epidemiological studies showed that breast cancer risk increased by 2.3% per year of hormone use. This was similar to the increased risk of 2.8% per year of natural delay in the onset of the menopause.35 Data from four RCTs involving 12,643 women indicate that the use of estrogen alone may not affect breast cancer risk.25,36–38 The average risk for invasive breast cancer in the trials was 0.81 (95% CI 0.63–1.03).39 In a further four randomized trials involving 19,756 women who used estrogen–progestogen, the average breast cancer risk was 1.24 (95% CI 1.03–1.50).39 In 19 recent epidemiological studies, the hormone-associated breast cancer risk diminished soon after discontinuing hormones, and normalized within five years.39 Breast cancer risk does not vary significantly with different types of estrogen or progestogen preparations, with use of lower dosages, or with different routes of administration.39 Relative risks of less than two-fold are classified as moderate breast cancer risk factors. Other risk factors in this range include alcohol consumption, obesity, and nulliparity.40
Endometrial Cancer
The incidence of endometrial cancer in post-menopausal women is less than one case per 1,000 women per year.12 Epidemiological studies since 1975 have consistently shown that unopposed estrogen use in women with a uterus increases the risk for endometrial cancer by approximately 2.8-fold (95% CI 2.6–3) compared with those who have never taken estrogen.41 Combining estrogen with progestogen appears to minimize the effect of unopposed estrogen on endometrial cancer risk.42 In the WHI estrogen–progestogen trial, the risk was similar to that seen in the placebo group (relative rate 0.83, 95% CI 0.29–2.32).12
Epithelial Ovarian Cancer
Invasive ovarian cancer incidence in post-menopausal women is approximately 0.3 cases per 1,000 women per year.9,43 The use of hormones during the menopause has been associated with higher ovarian cancer incidence, but social class is a predisposing factor for ovarian cancer and women in the highest social classes are most able to pay for hormones.44 Combined data from five epidemiological studies involving 2,501 women showed that the ovarian cancer risk was 1.28-fold higher (95% CI 1.05–1.56) for hormone users compared with non-users.45 In the WHI trial of estrogen–progestogen the ovarian cancer risk was not significantly increased (relative rate 1.58, 95% CI 0.77–3.24).9
Conclusions
Hot flashes occur in over 50% of women entering the menopause and their frequency declines to 30% after three years. Estrogen with or without progestogen is the most effective treatment for hot flashes and for symptoms due to vulvar and vaginal atrophy. The average patient is a woman aged between 45 and 60 years who takes hormones for less than three years. Long-term hormone use is not indicated to prevent CHD, hip fractures, or senile dementia. The impact of side effects can be diminished by appropriate counseling. Venous thromboembolism may occur during the first year of treatment, but adverse effects on cancer and stroke incidence are mainly associated with using hormones for more than three years.
