"I think there's enough data to put into question whether to use Avandia," Hammes said. "Most physicians are going to shy away from it and, in our clinic, we shy away it and if you're going to put someone on a TZD, you're going to put them on Actos."
These researchers analyzed data from an earlier observational study of diabetes patients, some of whom were taking TZDs.
Women over 50 who had broken bones were 71 percent more likely to have been prescribed a TZD.
In men, the increased risk (more than triple) was seen among those taking both TZDs and loop diuretics, but not in just one or the other.
And in both genders, the risk went up the longer a person was on the medication, which, according to Banerji, poses a particular problem.
"People have diabetes for many years not a short period of time. You can't put patients on these for a long period of time," she said. "We have to be very careful about prescribing TZDs at all and need to wait for the new class of TZDs that understands the [disease] process enough to hit the right target and not all these unanticipated targets."
But Hammes did point out that the study, which will appear in the October issue of the Journal of Clinical Endocrinology & Metabolism, did have some limitations. It was not, for instance, a randomized controlled study nor was the original study the data came from designed to look at this question.
The authors, from the University of Michigan and other institutions, called for larger, randomized trials.
The U.S. National Library of Medicine has more on thiazolidinediones.
SOURCES: Mary Ann Banerji, M.D., professor, medicine, and director, diabetes treatment center, State University of New York Downstate Medical Center, New York City; Steve Hammes, M.D., Ph.D., chi
All rights reserved