Several risk factors are important

EDITOR—Wilkin argues for broadening the indication for treatment of osteoporosis to “infirm older people.”1 There are several problems with the specific case he makes but evidence nevertheless to support a move in this direction. He deduces that frequency of impact is the main risk factor for fracture Falling is indeed a recognised risk factor, but the evidence suggests that it is just one among several predictors.2

Surprisingly, few researchers have attempted to combine their results to produce a risk score. We have identified one study that produced a score with a sensitivity of 70% and specificity of 98% on the basis of three factors—bone mineral density, body sway, and muscle strength.3 This prediction is for fractures occurring within a fairly short follow up time and probably overestimates the potential for effective intervention. Our preliminary results from a similar cohort provide support for a risk score approach over a longer period.4

We suggest that there is a case for focusing on elderly people but that the approach should be based on firm evidence and involve a specific risk score rather than a nebulous concept of infirmity. Hormone replacement therapy and bisphosphonates are costly and have important side effects and should not be used indiscriminately in any age group.

^a Competing interests: None declared.

Replacing bone mineral density with bone turnover is not a solution

EDITOR—Low bone mineral density is a disease surrogate; as Eastell points out in his commentary on Wilkin's article on osteoporosis, it is unfortunate that this particular surrogate has been given the status of a true diagnosis.1 But Wilkin's suggestion of replacing bone mineral density with another surrogate, bone turnover, is not a solution to the deficiencies of bone mineral density.

An important use for a surrogate is to identify groups of high risk patients so that treatment is directed most economically at those …