Drugs for pre-osteoporosis: prevention or disease mongering?

We have read with interest the letters from Dr. Compston et al on
behalf of the National Osteoporosis Guideline Group, Dr. Seeman and Dr.
Rizzoli.

Both Dr. Compston and Dr. Seeman note that most fractures occur in
women who do not have osteoporosis. Dr. Seeman further notes "Reducing
this disease burden among the many at moderate risk has been eloquently
written about by Professor G Rose many years ago". Dr. Rizzoli states that
the goal of management is to “to identify patients at increased risk of
fracture” and “to to allow them to benefit from efficacious available
treatments” without specifying the absolute risk of fracture to which he
is referring. Such statements are easily interpreted as support for
widespread treatment of very large numbers of woman at "moderate risk" -
women who are at very low risk in absolute terms. Dr. Compston's, Seeman's
and Rizzoli’s letters provide safeguards against such interpretation; the
material we reviewed did not

Dr. Compston accuses us of ignorance of risk models of osteoporosis.
When she subsequently acknowledges that we are in fact quite aware of risk
assessment and the desirability of treating only those at highest absolute
risk she accuses us of failing to provide a revelation. We plead guilty of
failure to provide a revelation, but not to the charge of ignorance.

Dr. Compston accuses us of scare-mongering by mentioning the
possibility of osteonecrosis of the jaw. This rare but serious side-effect
is acknowledged as causally linked to alendronate in the web of the
National Osteoporosis Foundation that Dr. Compston represents (1). Is she
suggesting that occurrence of rare and serious side effects (even if
absolute proof of causation is unavailable) should not be included in the
balance when considering desirable and undesirable consequences of
treating women "at risk"? We hope not.

Our intention is not to trivialise the problem of fragility fractures
but to point out that the population at risk of being medicalised by being
labelled “osteopenic” is growing. The problems with the reanalyses we
examined include increasing the likelihood of women, without their prior
consent, receiving first a label of “at risk”, and subsequently of
receiving a treatment that, given full information delivered in an
unbiased way, they would not choose.

We agree with Dr. Seeman that the absolute risk algorithm developed
by the WHO will, if adhered to, avoid the exposure of low risk women to
unnecessary interventions. Nevertheless, even “high risk” women are at low
absolute risk. There is a psychological cost to labelling a woman as “high
risk”. Indeed, many women are likel to be disinclined to take a
medication the appropriate duration of which has not been established,
associated with cost, inconvenience and side effects (some potentially
serious) for a small absolute risk reduction. The women will nevertheless
be left carrying the label of “at risk”.

If the industry and osteoporosis researchers funded by the industry
are scrupulous in recommending that only women with clinical risk factors
undergo bone mineral density testing, that only those at highest risk
following testing be labelled "at risk", and that those highest risk women
be presented with a balanced case on the desirable and undesirable
consequences of treatment (with, for instance, a well-constructed decision
aid) we would have no problem. In the articles and advertisements that we
addressed, we have seen no evidence of such scrupulousness.

Pablo Alonso-Coello, Gordon Guyatt, Alberto Lopez

Iberoamerican Cochrane Center-Department of Clinical Epidemiology and
Public Health. CIBER de Epidemiología y Salud Pública (Hospital de Sant
Pau) 08041, Barcelona, Spain

Competing interests:
None declared