This is published in the Journal of General Internal Medicine, January 2010.

Reuters and others carried the layperson writeup.

This was in older people.  Researchers found that for the highest cohort of narcotics users, the risk of fractures was doubled.  1/3 of these were hip and pelvis, though this means that 2/3 were of another type.  The idea was that patients on narcotics were more likely to get dizzy and fall, though I wonder whether the reason is actual acceleration of osteoporosis in the older population.  The abstract doesn't make it easy to discover whether they looked at this.

The researchers presumably studied narcotics use at Kaiser, and it seems like they did a database review of 2,500 patients.

This is an unexpected association, and it's good every now and then that these types of results come out.  Docs need to be reminded that narcotics can be dangerous, and to think about this before whipping out that prescription pad.

I love this item from the Reuters article, purportedly citing a Cochrane review of the subject: "the risks of long-term opioids may outweigh the modest pain relief."  As if anyone would be o.k. with not getting a prescription for narcotics if there was significant pain.  Would you be a good doctor if your patient told you, "Doc, I have a severe pain in my back and I can't sleep, I've tried Motrin, Tylenol doesn't work, and acupuncture, surgery, physical therapy, and chiropractic are no longer options."  Are we really supposed to say, "Well, you know, the Cochrane Review says..."

I try not to be the one to start the patient on long-term narcotics for non-cancer pain.  I make an effort to identify etiology.  I make the effort to consolidate pain medication prescriptions with one doctor.  And I make them come in for regular visits to make sure we have a handle on the pain and that we are trying to wean off medicines.  Research like this validates the effort to use fewer narcotics, though it remains a reality of practice.