Deciding whether and how to intervene involves assessment of short-term fracture risk and other non-skeletal health issues.
Although physical activity in youth is positively associated with peak bone mass, exercise in later life has probably only a modest effect on slowing bone loss. The benefits of resistance or weight-bearing physical activity are shown in clinical trials which include striking improvements in muscle strength and function and a reduced risk of falling (by about 25%) even in very old age.
Calcium intake must equal obligatory calcium loss to maintain calcium homeostasis and prevent bone loss. Without dairy products, the diet rarely provides more than 300 mg of calcium per day, less than the obligatory calcium loss from skin, urine, and stool. About 1200 mg of calcium per day is recommended for postmenopausal women without hormone replacement therapy. This requires increased consumption of milk products (each dairy portion contains approximately 300 mg), or other calcium-supplemented foods (such as calcium-supplemented orange juice, which is particularly useful for women with lactose intolerance), or calcium supplements. More than 20 clinical trials have shown that calcium supplements improve bone mineral density, especially in elderly women. Calcium alone is not sufficient to prevent fractures, however.
Vitamin D, primarily produced in the skin after exposure to sunlight, is necessary for optimal calcium absorption. Low Vitamin D levels are relatively common in winter, and year round in the elderly who no longer go outdoors. Physiological supplements of calciferol (400 – 800 mg/day) increase calcium absorption, prevent secondary hyperparathyroidism, and increase bone density at the femoral neck in the elderly. Vitamin D supplementation of 600 – 800 IU/day is recommended for housebound women and for those 65 and older.
Evidence that vitamin D prevents fracture is mixed. Two large trials of vitamin D (plus adequate calcium) for fracture prevention in the elderly had contradictory results. One found no difference in fracture rates in older Dutch men and women assigned to calciferol 400 IU/day or placebo. In England, a trial of 100 000 IU of oral vitamin D given every four months for two years reduced fracture risk by 22%.
Many controlled clinical trials have shown that postmenopausal oestrogen therapy increases bone mass moderately, whether begun at the time of the menopause or in old age. Oestrogen has several bone-sparing effects, increasing calcium absorption from the gut, reducing renal calcium loss, and directly reducing bone resorption. The greatest effects are seen at the spine. Bone density increases for the first 3 years of hormone replacement therapy, with no evidence of significant further gain. Bone loss resumes when oestrogen is discontinued.
Transdermal and oral oestrogens have similar effects on bone density. Although bone augmentation is better at higher doses, a majority of healthy women maintain bone with as little as 0.3 mg of conjugated equine oestrogen or equivalent (half the previously recommended 0.625 mg/day) when oestrogen is taken with adequate calcium. Studies consistently show better bone preservation in women who take both oestrogen and calcium. Oestrogen combined with cyclical or continuous medroxyprogesterone acetate or with cyclical progesterone does not improve bone density more than treatment with oestrogen alone, but the addition of a more androgenic progestogen or testosterone may provide additional bone benefit.
Although oestrogen unquestionably preserves bone, until recently the strongest evidence that hormone replacement therapy prevents fractures was based on a systematic review of 22 trials of hormone replacement therapy, which found a 27% reduction in non-vertebral fracture risk Women in these trials did not have a diagnosis of osteoporosis based on low bone mineral density. In the large HERS trial, hormone replacement therapy did not reduce fracture risk in older women who had heart disease but not osteoporosis.
The recently reported Women’s Health Initiative is the first large trial to unequivocally show that hormone replacement therapy prevents fractures. In this trial, 16 608 women were assigned to placebo or conjugated equine oestrogen plus medroxyprogesterone acetate (Prempro); after 5 years hormone replacement therapy reduced the risk of all fractures by 24% and reduced the risk of hip fractures by 33% compared to placebo (Writing Group for the Women’s Health Initiative Investigators 2002).
Unfortunately, the Women’s Health Initiative reported an early excess of cardiovascular disease and a delayed excess of breast cancer, greatly reducing the potential use of hormone replacement therapy for fracture prevention because continued hormone replacement therapy is required for bone maintenance while hormone replacement therapy for 4 or more years promotes breast cancer. Consensus is building that hormone replacement therapy (oestrogen plus a progestogen) should be used only by women with menopause symptoms and ideally for no more than 4 years after age 50. This conclusion does not necessarily apply to unopposed oestrogen. Another part of Women’s Health Initiative is examining the risk – benefit ratio of unopposed oestrogen in 10 739 women; results are not yet available.
In the absence of contraindications, it is usual to initiate hormone replacement therapy after a premature menopause, to be continued until about age 50 or 55, in order to mimic the oestrogen-replete lifespan. This treatment is reasonable, based on oestrogen’s proven ability to delay postmenopausal bone loss, although not tested in clinical trials.
Non-oestrogen medication for osteoporosis
Since 1990 the bisphosphonates, alendronate and risedronate, and raloxifene, a SERM (Selective oEstrogen Receptor Modulator) have been shown in large clinical trials to prevent spine fractures in women at high risk of fracture (women who were on average 15 years postmenopause and had low bone density or prior spine fracture). There are no head-to-head comparisons of these therapies with fracture outcomes, but each suppresses turnover and reduces fracture risk by 40 – 50% with overlapping 95% confidence intervals. These treatments do differ with regard to bone mineral density increments, with greater increases for alendronate and risedronate than raloxifene. Because only small improvements in bone density are necessary to reduce fracture risk, the choice of medication should be based on considerations other than the change in bone density.
Parathyroid hormone stimulates bone formation and increases bone mass more dramatically than any other medication. In a clinical trial, a 20 Вµg dose of Parathyroid hormone reduced the risk of new vertebral fractures by 65% over a period of less than 2 years; the number of women with new moderate or severe fractures was reduced by 90%. This drug is expensive and must be taken by daily injection, but may be life-saving for the small subset of women with severe osteoporosis who do not respond to or tolerate other treatment. Present recommendations are for one or two years of use, after which another antiresorptive medication can be initiated.
Tibolone in doses of 1.25 mg/day significantly improves bone mineral density of both spine and hip in early postmenopausal women. No large clinical trial with fracture outcomes has been reported.
Treatment choice by age
Oestrogen or tibolone are most often prescribed to preserve bone in the early postmenopausal years. Oestrogen is effective when begun later, but the resultant bleeding and breast pain are often unacceptable for older women. Women 5 – 15 years postmenopause without severe osteoporosis may prefer raloxifene for its possible reduced risk of breast cancer and heart disease. Older women with osteoporotic T-scores on bone mineral density or prior fracture(s), may prefer alendronate or risedronate for their rapid-acting bone-specific effects. Alendronate and risedronate but not raloxifene have been shown to reduce the risk of non-spine fracture by 20 – 25%.
There are practical reasons to focus on older women. Given a 3-year treatment that reduces fracture risk by 50%, in women aged 65 and older it would be necessary to treat 13 who have had a fracture and 63 without a prior fracture in order to prevent one new fracture. To prevent one fracture in 50-year-old women, it would be necessary to treat 50 who have had a fracture and 250 without prior fracture. In addition to the smaller number needed to treat, older women are often more willing to take and continue osteoporosis treatment because their risk of fracture is more proximate than the risk for younger women, making the risk – benefit ratio for medication more favourable. On the other hand, women who have already sustained a fracture and are treated with the strongest available bone-specific osteoporosis drug will still have a higher fracture risk than similarly treated women without prior fracture.
For women with very advanced osteoporosis who do not respond to other therapy, Parathyroid hormone can be given in combination with antiresorptive agents such as oestrogen or a bisphosphate, with further increase in bone formation. Other than this, combined use of two osteoporosis therapies is rarely necessary, incurs greater costs, increases side effects, and only increases bone density slightly above that achieved by single-drug therapy. No clinical trials have shown an additive effect of dual therapies on fracture protection.
Because all treatments have side effects, only women at high risk of fracture in the near future should be prescribed aggressive pharmacotherapy. The most serious common side effect of hormone replacement therapy, raloxifene, and tibolone is venous thromboembolic disease including pulmonary embolus. The risk, about two to four times that in non-users, decreases but does not disappear with continued therapy (Writing Group for the Women’s Health Initiative Investigators 2002). Given the poor risk – benefit ratio for combined oestrogen plus progestogen therapy, the initiation of combined hormone replacement therapy for fracture prevention must be seriously questioned, particularly now that alternative treatment options are available. The risk – benefit ratio for unopposed oestrogen and fracture risk reduction is still being studied in Women’s Health Initiative.
Raloxifene increases hot flushes, more often in younger women. In clinical trials, approximately 20% of women younger than 60 years of age and 10% of older women developed hot flushes, but symptoms were usually mild and rarely led to discontinuation of therapy. Unlike oral oestrogen, raloxifene does not increase the risk of gallbladder disease or endometrial hyperplasia.
Alendronate is most often stopped because of upper gastrointestinal symptoms, particularly in women aged 70 and older and those with active gastro-oesophageal reflux disease. Women who experience upper gastrointestinal symptoms on daily alendronate usually can tolerate 70 mg once a week, an interval that seems to allow superficial gastric erosions to heal. Trials of intermittent dosing with alendronate and risedronate have shown bone density changes equal to those achieved with daily dosing, and it is likely that most women will prefer weekly therapy.
Women who have already suffered osteoporotic fractures have a very high risk of new fractures in the future, and require a bone-specific medication such as alendronate or risedronate, each of which can reduce the risk of fracture by half within a year or two. It is unknown how many years to continue bisphosphonate therapy.
Statin use has been reported to reduce risk of bone loss or fracture risk in some observational studies, but this was not confirmed in the largest observational study from the Women’s Health Initiative. No clinical trial data have been reported.
Thiazide diuretics may reduce bone loss and fracture rates, as suggested by observational studies and one clinical trial.
Soya food (soya protein isolate at 20 – 25 g soya protein/day) has little or no effect on bone density, but ipriflavone, a synthetic isoflavone, reduces bone loss. No fracture trials have been reported.
Observational studies suggest vitamin C is associated with higher levels of bone mineral density. No trials have been reported.