Natural menopause is the permanent cessation of menstruation resulting from the loss of ovarian follicular activity. It is confirmed after 12 consecutive months of amenorrhoea for which there is no other obvious cause. The average age at natural menopause, about 51 years, has been constant for centuries. A majority of women experience natural menopause between ages 45 and 55 years. Cigarette smoking, a vegetarian diet, and simple hysterectomy are each associated with an earlier (by 1 – 2 years) average age at menopause. Premature menopause (last menstrual period before age 40) can follow bilateral oophorectomy, chemotherapy, or radiation (i.e. induced menopause), or be caused by premature ovarian failure.

The menopause transition (perimenopause) is usually defined as the interval before menopause when hormone levels fluctuate and fall. When the menopause transition begins depends on its definition: reduced fertility typically begins about 10 years before the last menstrual period, hormone changes are apparent about 8 years before the last menstrual period, and menstrual irregularities with or without vasomotor symptoms, disturbed sleep, and mood changes usually begin about 4 years before the last menstrual period.

Postmenopause describes the years after the menopause. Late postmenopause, defined as 10 or more years after the last menstrual period, is more important than chronological age for conditions that are clearly related to low postmenopausal oestrogen levels, such as osteoporosis.

The perimenopause

General overview of hormone replacement therapy

HRT: Alternative treatment for vasomotor symptoms

Quality of life, mood, and depression

Randomized placebo-controlled trials of less than two years duration have reported improved quality of life (estimated by standard quantitative tests) in oestrogen-treated women. This benefit is only clearly documented in women who have vasomotor symptoms. In a recent publication from the HERS trial, women treated with Prempro who had no menopause symptoms at baseline reported less vigor and well being than women on placebo.

The relation of oestrogen to depressed mood during the menopause transition is controversial. A 1995 review of 14 clinical trials concluded that oestrogen therapy improved depressed mood during the menopause transition or after oophorectomy, but most of these trials included women with vasomotor symptoms and impaired sleep. Women with clinical depression do not usually improve on standard oestrogen regimens. However, very high doses of transdermal oestradiol may provide significant benefit, as suggested by small short trials.

Progestogens may worsen perimenopausal depressive symptoms. Women who had premenstrual syndrome before menopause are more likely to have negative mood effects during the progestogen phase of hormone replacement therapy. A trial of progestogen withdrawal should be considered for any woman who becomes depressed after starting hormone replacement therapy.

St Johns Wort has been recommended as an alternative therapy for menopausal depression, based on a meta-analysis of 23 trials in 1757 women with depressive disorders; the authors concluded that St John’s Wort was superior to placebo and comparable to standard antidepressant medication; Linde and Mulrow’s subsequent review of 27 trials in 2291 women reported similar results. By contrast, two other clinical trials in a total of 540 women with major depression found no benefit.


Women may report either increases or decreases in migraine frequency or severity with menopause or hormone therapy. Women who had few or mild headaches premenopause report worsening with hormone replacement therapy. One clinical trial reported a significant increase in the frequency and duration of migraines after oral but not transdermal oestrogen and no effect of hormone replacement therapy on tension headaches.

Weight gain

Weight gain is common during the menopause transition, and has been attributed to both oestrogen deficiency and oestrogen therapy. A quantitative review of 22 randomized controlled trials found no evidence that postmenopausal oestrogen alone or with a progestin promotes or prevents weight gain. A clinical trial has shown that weight gain during the menopause transition can be prevented by diet and exercise: premenopausal women were randomly assigned to an intervention of reduced saturated fat plus increased physical activity or to assessment only; after the menopause transition, weight had decreased 0.2 lb in the intervention group and increased 5.2 lb in the control group, and waist circumference had decreased 2.9 cm compared to 0.5 cm in the control group.

HRT: Sexuality, urogenital symptoms, and libido

Osteoporosis and fractures

Bone loss

Osteoporosis is due to imbalance between bone formation and bone resorption, leading to loss of bone quantity and quality. Throughout life there is constant removal and replacement of bone in myriad small bone remodelling units on bone surfaces. Low bone mineral density (and even osteoporotic fractures) can occur before the menopause in young women with anorexia nervosa or extreme physical activity. In these young women, amenorrhoea is a classical feature of malnutrition-related bone loss, which is usually reversible with adequate nutrition.

The balance between bone formation and bone resorption is disturbed during the menopause transition before menses cease when follicle-stimulating hormone levels are high but oestrogen levels are relatively normal. Bone loss is often accelerated for several years after menopause, averaging 2% per year (range <1% – 5%), after which there is a steady slow bone loss of about 0.5% year; however, some women show little change in bone mineral density after menopause. With advanced age, bone loss accelerates, reflecting immobility, illness, poor nutrition, or secondary hyperparathyroidism due to vitamin D deficiency.

Peak bone mass

The age when a woman has osteoporotic bone levels depends on her peak bone mass and the rate of bone loss after menopause. The age when peak bone mass is achieved is uncertain, but is thought to be between 20 and 30. Peak bone mass is determined in part by small effects of multiple genes, birth weight, childhood growth and development, not smoking, a diet adequate in protein and calcium, and weight-bearing physical activity. Women of African heritage tend to have higher peak bone mass and women of Asian heritage tend to have lower peak bone mass than Caucasian women. Women who have poor diet and sedentary habits in youth and young adulthood never attain their potential peak bone mass.

Fracture risk

Fracture risk increases with age and is also closely related to bone density, bone connectivity, body size, and the propensity to fall. A 50-year-old white woman has a 15% lifetime risk of hip fracture, usually occurring after age 70; a 40% risk of at least one spine fracture, often by age 60; and a 15% risk of Colles fracture, typically before age 60. Wrist, spine, and hip fractures can each cause functional limitations. After a hip fracture, only one-third of women regain their prefracture level of function and independence, and one in five die within the year of fracture.

HRT and Osteoporosis: Clinical diagnosis

HRT: Prevention and treatment of osteoporosis

HRT and coronary heart disease

Hormone replacement therapy and breast cancer

Endometrial cancer

Unopposed oestrogen causes endometrial hyperplasia, a cancer precursor. In the PEPI trial, unopposed oestrogen increased endometrial thickness assessed by transvaginal ultrasound and endometrial hyperplasia by needle biopsy by 10% per year (Writing Group for the PEPI Trial 1996). These changes were not observed in women taking oestrogen plus continuous or cyclical medroxyprogesterone acetate or cyclical micronized progesterone. In another study, treatment with up to 5 years of continuous combined oestradiol (2 mg) plus norethisterone (1 mg) daily was not associated with endometrial hyperplasia; in fact, pre-existing hyperplasia reverted to normal in hormone replacement therapy-treated women.

Five years of unopposed oestrogen increases the endometrial cancer risk four- to five-fold; the risk for 10 years” use is increased approximately 10-fold. A good prognosis is usual when the oestrogen-stimulated cancer causes bleeding leading to early diagnosis and treatment, but the relative risk for invasive cancer is also increased. Women who take 10 – 12 days of a progestogen with their oestrogen are protected from endometrial cancer; the reported exceptions may reflect non-compliance with the progestogen.

Some women who cannot tolerate a progestogen may want to take oestrogen alone. This requires a clear explanation of endometrial cancer risk and annual endometrial biopsy and/or transvaginal ultrasound. Unfortunately, a 5 mm increase in endometrial thickness by ultrasound does not always precede the development of atypical endometrial hyperplasia, a cancer precursor.

Colon cancer

In a systematic review of 21 published observational studies of hormone replacement therapy and colorectal cancer, 9 reported a significant reduction in colorectal cancer, 9 reported no significant protective effect, and 3 reported a small non-significant increased risk. Women’s Health Initiative reported an absolute risk reduction of 6 fewer colorectal cancers per 10 000 person-years (Writing Group for the Women’s Health Initiative Investigators 2002).

Ovarian cancer

Some observational studies have reported an excess risk of ovarian cancer in women taking hormone replacement therapy or unopposed oestrogen but most have found no association. None of these studies is convincing regarding safety or harm.

Memory loss and dementia

LeBlanc et al. (2001) reported a significantly reduced relative risk of dementia in women taking oestrogen but there is no clinical trial evidence that oral oestrogen slows memory loss or the progression of dementia. Only small trials in recently oophorectomized women have shown improved verbal memory. In HERS, approximately 1000 older intact women with heart disease completed six cognitive function tests; after 4 years, women assigned to hormone replacement therapy did not perform better on any test than women assigned to placebo. The Women’s Health Initiative Memory Study (WHIMS), an ancillary 4-year study of 4532 Women’s Health Initiative women aged 65 and older, reported that more women in the estrogen plus progestogen than placebo group had a clinically significant decline in global cognitive funtion (6.7% vs. 4.5%) and twice as many women developed probable dementia. No benefit has been shown in a one-year clinical trial in which 120 women with early Alzheimer’s disease were randomly assigned to placebo or unopposed Premarin (either 0.625 mg or 1.25 mg); cognitive function scores on the clinical dementia rating scale were actually significantly worse in women assigned to oestrogen.

By contrast, raloxifene shows some promise; in a 3-year trial, older women (aged 70+) assigned to raloxifene had less decline on two of six cognitive function tests than women assigned to placebo.

Venous thromboembolic disease

There is no question that oral oestrogen increases the risk of Venous thromboembolic disease and pulmonary emboli. In observational studies of healthy women, the absolute numbers of excess Venous thromboembolic disease are about 1 in 5000 hormone replacement therapy users per year. In HERS women with documented heart disease, the absolute risk was much higher, about 1 per 250 users per year (even though a majority were taking aspirin and statins, medications that seem to reduce the risk of venous thromboembolism). An increased risk of Venous thromboembolic disease persisted throughout the 4-year HERS trial, showing that not all susceptible women who will have Venous thromboembolic disease do so during the first year of treatment. In Women’s Health Initiative, women taking oestrogen plus progestin had twofold greater rates of Venous thromboembolic disease including pulmonary emboli and deep vein thrombosis (Writing Group for the Women’s Health Initiative Investigators 2002). Again, excess risk was greatest in the first year but was still apparent after 4 – 5 years.

Mutations in factor V Leiden or prothrombin are associated with an increased risk of Venous thromboembolic disease, but most women with VTEs do not have either genotype, and screening is not currently recommended. There are insufficient data to assume that transdermal hormone replacement therapy will be safer. A careful history for prior Venous thromboembolic disease should be obtained, and women with a positive history should avoid oestrogen.

Gallbladder disease

The risk of gallbladder disease is approximately doubled by oral oestrogen. In HERS there was about one extra case per 250 hormone replacement therapy users per year. Observational data and laboratory studies suggest that transdermal oestrogen will have little or no effect on gallbladder disease.


Despite the apparent ubiquity of oestrogen receptors in diverse human tissues, and its multiple biological effects, randomized clinical trial evidence of hormone replacement therapy benefit or harm is presently limited to only a few important outcomes: relief of vasomotor and genital symptoms, an increased risk of Venous thromboembolic disease that wanes a little over time, an increased risk of CHD and stroke that does not appear to be limited to the first year of treatment, an increased risk of breast cancer first apparent after 4 – 5 years of hormone replacement therapy, an increased risk of gallbladder disease, and a decreased risk of fracture and colon cancer. Overall, the benefits of the extended use of combined hormone replacement therapy necessary to prevent fractures and colon cancer are clearly exceeded by the risks. The risk – benefit ratio for unopposed oestrogen is not clear. Fortunately, other non-hormone replacement therapy medications have been shown to be effective in reducing women’s risk of cardiovascular disease and fractures.

10-Minute consultation: osteoporosis

Mrs B is a 62-year-old Caucasian woman worried about osteoporosis after reading an advertisement in a magazine. She says she is in excellent health, has no complaints, and has never had a fracture. But her previously healthy mother died 2 years ago at age 89, within 6 months of a hip fracture.

What would you do next?

  • Obtain a menopause and hormone replacement therapy history (oestrogen deficiency is the commonest cause of osteoporosis).
  • Obtain a diet, weight and lifestyle history to elicit risk factors for osteoporosis.

Mrs B was 48 when she had her last menstrual period and began hormone replacement therapy (hormone replacement therapy) on the advice of her physician. She quit hormone replacement therapy 2 years later after reading an article about breast cancer, and she takes no prescription medications now.

Mrs B’s main exercise is walking two blocks to the shop for fresh bread and her daily pack of cigarettes. She drinks one or two glasses of wine when out with her friends. Because she dislikes milk and cheese, she has started taking 1000 mg of supplemental calcium each morning. She worries about wrinkles and skin cancer, and is thus careful to avoid sun exposure.

Mrs B’s physical examination is normal. She is quite pleased that her weight, 8 stone 3 pounds, has not changed since she was 25, but she is surprised at her measured height, which at 5 feet, 5 inches is nearly 2 inches shorter than she remembers.

Is Mrs B at high risk for osteoporosis, and why?

Mrs B has several characteristics that make osteoporosis a real possibility for her. She has lived more than 10 years with postmenopausal oestrogen levels; the short amount of time she used oestrogen would have had no sustained benefit, because estrogen must be continued to preserve bone. And she has several lifestyle risk factors.

Osteoporosis is important only because it is a risk factor for fracture. The strongest risk factor for a fracture is a previous fracture. Mrs. B’s 2-inch height loss is compatible with unrecognized vertebral fractures. The next two strongest risk factors for osteoporosis are years since menopause and low body weight. Mrs. B’s menopause was 2 or 3 years earlier than the average age (51 years); cigarette smoking is associated with an earlier menopause and low bone mineral density. She is also fairly thin. Low body fat and smoking are each associated with lower levels of endogenous oestrogen and with lower circulating levels of oestrogen after hormone replacement therapy. Limited weight-bearing exercise impairs muscle and bone strength and balance, all risk factors for a fragility fracture. She is also likely to be vitamin D deficient. Furthermore, because 1000 mg of calcium is about twice the amount the body can absorb at one time, her effective calcium supplementation is only half the amount she is taking.

On the positive side, she is in your office because she is worried about osteoporosis, which suggests that she may be receptive to bone-sparing recommendations. Also her general health is good, she has not lost weight (weight loss is a risk factor for osteoporosis), and her mother’s first fracture did not occur until she was nearly 90. Although there are few data on this point, it is likely that only a family fracture occurring before age 75 or 80 is a useful predictor of increased fracture risk in another family member. Modest alcohol intake that does not replace a reasonable diet is not bad for bones, and may actually have a small positive effect on bone mineral density.

How can you be sure the patient has osteoporosis?

Expensive tests should not be ordered unless the results will change management and patient compliance. Even without tests, this woman’s concerns open the door at this and future visits to educational opportunities about positive lifestyle changes she can make and treatment options that are available.

In the absence of a confirmed previous low trauma fracture, the only way to be certain a patient has osteoporosis is to obtain a spine film and look for vertebral fractures (two-thirds of which are unrecognized without an x-ray), or to obtain a central or peripheral bone scan to determine whether her bone mineral density is at least 2.5 standard deviations below the young adult mean (i.e. a T-score of -2.5 or lower).

A prior fracture carries a much higher risk of subsequent fractures than an osteoporotic T-score, and in some countries documentation of a prior fracture may be required before prescription of a bone-specific agent is approved. In this case the physician chose to order spine films first, based on their very strong risk prediction, the reported height loss, and because x-rays could be obtained without waiting for a referral appointment.

Which bone scans and when?

There is a difference of opinion about the optimal time to order a bone scan when osteoporosis is suspected but no fracture has occurred. Unless there are several risk factors, it is usually reasonable to wait until the patient is at least 5 years postmenopause, allowing some bone loss to have occurred. But a combination of such risk factors as family history, height loss, tobacco habit, and physical inactivity noted above can prompt much earlier evaluation.

The gold standard for a diagnosis of osteoporosis is dual energy x-ray absorptiometry (DEXA) of the hip. When more than one site is scanned, it is not uncommon to see considerable divergence in the results. In Mrs B, the lumbar spine T-score was -1.5 and total hip T-score was -2.3. Most clinicians would call the hip result close enough to the WHO criteria to warrant a diagnosis of osteoporosis. Although postmenopausal bone loss occurs first at the spine, it is often obscured by osteoarthritis at this site, as was the case here. Low bone density based on bone scans at peripheral sites including the forearm and heel has also been shown to predict fracture risk, and these smaller, less expensive instruments are suitable for the primary care office.

After a diagnosis of osteoporosis, what additional work-up is necessary?

At this point in the work-up there is divergent opinion depending as much on the specialty of the physician as the characteristics of the patient. Many specialists order serum calcium, phosphate, creatinine, alkaline phosphatase, a sensitive TSH test, and vitamin D and PTH levels, and others add 24-hour urine calcium, serum protein electrophoresis (for multiple myeloma), and tests for coeliac disease to diagnose the most common of the secondary causes of osteoporosis. These tests are certainly appropriate for osteoporotic women who are young, or who have unusually severe disease, or who are otherwise ill. For the apparently well postmenopausal patient like Mrs B, these additional tests would not be ordered by most generalists unless the patient fails to respond to therapy.

What treatment would you recommend?

Mrs B should make the following lifestyle changes:

  • diet (specifically, she needs to divide the calcium into a twice-a-day regimen and add vitamin D to her diet);
  • exercise (she should do more weight-bearing exercise plus resistance training); and
  • smoking cessation (she should get help through a smoking-cessation programme).

In addition, the three main regimens for the treatment of osteoporosis in a patient such as Mrs B are oestrogen, raloxifene, or one of the bisphosphonates  –  alendronate or risedronate.


Hormone replacement therapy has been shown in multiple clinical trials to preserve bone, and the prevention of osteoporosis is a desired side effect when oestrogen is used to treat menopausal symptoms in the early postmenopausal years. However, hormone replacement therapy is no longer approved in the USA for the treatment of osteoporosis. This decision was based on the paucity of clinical trial data showing fracture prevention in women with osteoporosis. Hormone replacement therapy does preserve bone when given to older women, but women who have been without oestrogen for more than 5 years tend to have more problems than younger women with breast pain and bleeding. In any event, Mrs B is worried about cancer and has already stopped oestrogen once, so hormone replacement therapy would not be a good choice for her.


This selective oestrogen receptor modulator is approved for prevention and treatment of osteoporosis, based on clinical trials showing a 40% reduced risk of spine (but not hip) fractures in women with osteoporosis with or without prior fractures. Raloxifene does not increase the risk of breast cancer (preliminary data suggest it may reduce the risk), heart disease, or stroke. The only serious side effect reported to date is venous thromboembolic disease, with rates similar to those reported with hormone replacement therapy. Raloxifene is not effective for the treatment of hot flushes, but instead slightly but significantly increases hot flushes. Mrs B did not have bothersome hot flashes and is worried about breast cancer, so this might be a good bone medication for her.


Alendronate and risedronate are oral bisphosphonates that have been shown to reduce spine fractures and hip fractures in women with osteoporosis. Although there are no head-to-head comparisons, the 40 – 50% spine fracture risk reduction is similar for these two bisphosphonates and also comparable to raloxifene (based on overlapping 95% confidence intervals). This is the only regimen shown to prevent hip fracture, and would also be an acceptable choice for this patient. The only common potentially serious side effect is oesophageal erosion and bleeding, so the medication should not be given to patients who report GERD or a hiatus hernia. These upper GI side effects are uncommon when the total weekly dose is given once a week instead of daily (70 mg alendronate or 35 mg risedronate).


Fracture risk reduction is not concordant with the change in bone mineral density. No repeat scan need be obtained for at least 2 or 3 years. Bone markers such as NTX or CTX can be helpful within the first few weeks of treatment, but are not used unless the patient has unusually severe osteoporosis. A new fracture does not necessarily mean treatment failure, in that the patient might have had more fractures or more severe fractures without the medication.

Further encouragement for lifestyle change, especially smoking cessation, is the most important follow-up activity for this patient.

Share →