To the Editor: The article by Seeman and Eisman on treatment of osteoporosis1 not only neglects the pathogenesis and prevention of this condition, but also fails to appreciate the profound differences between the three main types of fragility fracture (peripheral non-hip, hip and spine).

To say that adults lose bone with age because the “volume of bone resorbed is greater than the volume replaced” is a spectacular but common tautology which simply describes what would be expected from an external negative calcium balance. The common postmenopausal bone loss can generally be accounted for by the rise in calcium requirement due to a fall in calcium absorption and rise in obligatory calcium excretion.2-4 This can be corrected with hormones or compensated for with a calcium supplement, which is known to suppress bone resorption.5 In 20 trials of calcium therapy completed up to 1997, the loss of bone in 855 postmenopausal women treated with calcium was 0.3% per annum, compared with 1.0% per annum in 635 untreated control women (P < 0.001).6

In older women, this relative calcium deficiency is complicated by a decline in vitamin D status caused by reduced exposure to sunlight and progressive thinning of the skin. It has been known for 30 years that vitamin D deficiency is common in patients with hip fracture,7 for 20 years that this was also true in Australia,8 and for 12 years that vitamin D with calcium can reduce the hip fracture rate by 43% in 18 months in women in residential care.9

When it comes to established osteoporosis (the combination of low bone density with fracture), treatment needs to distinguish between the three main types of fracture referred to above. In hip fractures, surely the first priority must be to give adequate vitamin D and calcium. Most non-hip peripheral fractures occur in women with bone densities in the normal range,10 so the need for treatment is debatable unless bone turnover is very high.

Vertebral fractures are a different matter. Unlike peripheral fractures, they do not “heal” in the usual sense — the deformity remains and often causes local pain and mechanical dysfunction. The recurrence rate is very high, because all the vertebrae have much the same bone density, and the osteoporotic collapse of one indicates that the others are ready to follow. This condition is notoriously difficult to manage and should almost be regarded as a medical emergency that requires full investigation — not least to exclude myeloma — and rapid, effective treatment. It is in the secondary prevention of these fractures that the remedies advocated by Seeman and Eisman probably have their main role and are most cost-effective. In fact, although the authors place great emphasis on prevalent fracture as a risk factor for further fracture, the reference they quote deals with prevalent vertebral fracture, not with prevalent non-hip peripheral fracture, where the evidence of benefit from bisphosphonates and raloxifene is very much weaker.

We are arguing for much greater emphasis on the prevention of osteoporosis with adequate calcium after the menopause and adequate vitamin D in the elderly, and the use of appropriate investigation and selective treatment in the management of the condition when it is established.

In reply: Our article concerned a discussion of evidence-based treatment for the prevention of osteoporotic fractures. Prevention of osteoporosis per se is important, but the approaches chosen must be evidence- based. Calcium supplementation may diminish, but not abolish, bone loss by reducing remodelling rate.1 Any association of a lifelong diet high in calcium appears to be with peak bone mass, not rates of bone loss,2 and may be attributable to differences in protein intake or physical activity. Despite opinion, meta-analysis of prospective, randomised, double-blind, placebo-controlled studies does not support a role for calcium supplementation in reducing fractures.3

Later rather than earlier intervention avoids needless exposure to treatment for large numbers of individuals at low absolute risk of fracture (ie, those who are unlikely to sustain a fracture even without treatment).4 Vitamin D is of course indicated in vitamin D deficiency. However, there is no evidence for anti-fracture efficacy of vitamin D in ambulant community dwellers.5