What have we learned from clinical studies? Fractures and the interactions of bone mass and remodeling.

The knowledge that we have gained about risk assessment and fracture prediction from the tools we have available at present has provided considerable information allowing the clinician some improved capability to determine those most at risk for fracture. But the tools are imperfect, and consequently, the physician's capability to estimate fracture risk for individual patients is not precise. To put it into statistical terms, while a point estimate of absolute risk can be developed for individual patients the confidence intervals around that estimate are wide. To a degree, the same is true about treatment effects, and the degree of fracture protection afforded any individual patient may be quite variable. The features that cause the variability in risk prediction and fracture protection are the subject of this symposium. It can be suggested that the imprecision is related to the inexact tool of bone densitometry, which misses many of the features of the skeleton that might contribute to risk. Analogies abound in medicine, including the relationship between lipid status and risk of myocardial infarction, and must be expected among disease outcomes that have multiple pathophysiological mechanisms. Here we attempt to assemble the knowledge that we currently have about the mechanisms involved in fracture events.