WARNING: This post contains relatively more speculation and less evidence. Toward the end I will make an evidence-based comment.
As I was driving to work this morning, the local radio news broadcast a spot about fears that long-term use of bisphosponates such as alendronate (Fosamax) might increase risks of fractures in certain locations, leading the FDA to consider new warnings. The natural reaction one might have to such news would be to pooh-pooh the idea that a drug that (supposedly) builds stronger bones could be a cause of bone weakness.
To see why this possibility might be more reasonable than would first appear, I'll refer you back to an excellent book that I have not mentiooned here in a while, John Abramson's Overdo$ed America, and its excellent chapter on osteoporosis. Dr. Abramson there reported data (which he admitted were preliminary and not conclusive) on the mechanisms by which bisphosponate drugs work.
He reminded us that there are basically two types of bone. If bones were PVC pipes, hollow, then the main type of bone would be cortical bone. The other type of bone, trabecular bone, is the internal part that provides the cross-bracing within the hollow tube. It does not take much engineering know-how to figure out that the real contributor to stronger, fracture-resistant bones is the trabecular component. It is therefore worrisome that studies suggest that bisphosphonates selectively increase cortical bone and in some circumstances might even disrupt growth of trabecular bone.
Now, if you are trying to make bones stronger, you'd want a drug that increases trabecular bone. But if your goal is to market drugs, you are very happy with a drug whose sole impact is on cortical bone. That's because bone-mineral-density (BMD) machines measure cortical bone more than they do trabecular bone. So if your marketing plan is to sell bisphosponate drugs by encouraging as many women as possible to go in for BMD testing-- see http://brodyhooked.blogspot.com/2009/12/npr-how-osteopenia-became-treatable.html-- then you want a drug that once women start taking it, makes their numbers look better soon.
As Dr. Abramson pointed out, this helps to explain why bisphosponates show very little if any effectiveness in preventing fractures in women with osteoporosis unless they have had a previous fracture, and why there is virtually no evidence that they help women with the less severe stage of bone density loss, osteopenia--and why there is even a chance that bisphosponates might increase fracture risk in some women with osteopenia.
So we don't know the final answer by any means, but as more news comes out about bisphosphonates and fracture risk, we should keep these possible underlying mechanisms in mind.
Evidence-based comment: This is yet another example of the flaws of using surrogate endpoints to determine whether or not to prescribe drugs. So long as we continue to believe that making BMD numbers look better is the same as preventing osteoporotic fractures, or that making blood sugar numbers look better in Type II diabetes is the same as preventing serious diabetic complications, then we are not practicing smart medicine; we're doing pharmaceutical industry marketing.