Abstract
It is a commonplace of both clinical and investigative experience that things vary. No two individuals have quite the same bone size or density. Neither do bone mineral content (BMC) and remodeling activity respond to an intervention equally in every individual treated with a given agent. Moreover, the patients enrolled in an investigation of a new treatment for osteoporosis may differ, not only from one another, but from the larger group of patients the drug is intended to treat. Finally, those patients selected by their physicians to receive a particular treatment (e.g., hormone replacement therapy) may be different from those not selected. (Indeed, it would be hoped that the treatment choices by physicians and patients had a rational basis and were not decided by flips of a coin.).