This Article Full Text Full Text (PDF) Supplemental Data Submit a related Letter to the Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Google Scholar Articles by Solomon, D. H. Articles by Stürmer, T. PubMed PubMed Citation Articles by Solomon, D. H. Articles by Stürmer, T. Related Collections Calcium and Bone Metabolism Diabetes and Insulin

A Cohort Study of Thiazolidinediones and Fractures in Older Adults with Diabetes

Divisions of Pharmacoepidemiology (D.H.S., S.M.C., N.K.C., C.C., R.L., T.S.) and Rheumatology (D.H.S.), Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115; and Leslie Dan Faculty of Pharmacy (S.M.C.), University of Toronto, Toronto, Canada M5S 1A1

Address all correspondence and requests for reprints to: Daniel H. Solomon, M.D., M.P.H., Division of Rheumatology, Brigham and Women’s Hospital, 75 Francis Street, Boston Massachusetts 02115. E-mail: dsolomon{at}partners.org.

Context: Thiazolidenediones (TZDs) are selective ligands of peroxisome-proliferator-activated receptor- and have been shown to reduce bone mineral density. Recent results from several randomized controlled trials find an increased risk of fracture with TZDs compared with other oral antidiabetic agents.

Objective: The aim of the study was to determine the association between TZD use and fracture risk among older adults with diabetes.

Design: We conducted a cohort study.

Participants: Medicare beneficiaries with at least one diagnosis of diabetes initiating monotherapy for an oral hypoglycemic agent participated in the study.

Main Outcome: We measured the incidence of fracture within the cohort.

Results: Among the 20,964 patients with diabetes eligible for this study, 686 (3.3%) experienced a fracture during the median follow-up of approximately 10 months. Although not statistically significant, patients using only a TZD were more likely to experience a fracture than those using metformin (adjusted relative risk, 1.31; 95% confidence interval, 0.98–1.77; P = 0.071) or a sulfonylurea (adjusted relative risk, 1.21; 95% confidence interval, 0.94–1.55; P = 0.12). Each individual TZD was associated with an increased risk, with confidence intervals overlapping unity, compared with both metformin and sulfonylureas. The adjusted risk of any fracture associated with TZD use compared with metformin was elevated for non-insulin-using patients, women and men. If TZD use is associated with fractures, the number needed for one excess fracture when comparing TZD users to sulfonylurea users was 200, and the number was 111 when comparing TZDs with metformin.

Conclusions: As has been found with other analyses, our data suggest that TZDs may be associated with an increased risk of fractures compared with oral sulfonylureas and metformin.