[Skip to Navigation]
Editor's Correspondence
May 8, 2006

Thiazolidinediones and the Risk of Nontraumatic Fractures in Patients With Diabetes: In Reply

Arch Intern Med. 2006;166(9):1043. doi:10.1001/archinte.166.9.1043

In their letter published in the January 9, 2006, issue of the ARCHIVES, Mascitelli and Pezzetta1 state that thiazolidinedione use has been shown to decrease hip and femoral neck bone mineral density (BMD)2 and suggest that this is an explanation for our report that adults with type 2 diabetes mellitus are at higher fracture risk compared with adults without diabetes in the Health, Aging, and Body Composition (Health ABC) Study.3 While we agree that thiazolidinedione use may be associated with bone loss, the prevalence of thiazolidinedione use was too low to account for the increased fracture risk associated with type 2 diabetes mellitus that we reported.3 The reference that Mascitelli and Pezzetta1 cite is also from the Health ABC Study,2 and only 11 (2%) of 566 participants with diabetes in our report used a thiazolidinedione at the start of the study in 1997-1998. In the first 4 years of the study during which our fracture outcomes were collected, only 69 (12%) of 566 participants with diabetes used a thiazolidinedione.4 Our analyses of diabetes and fracture risk considered baseline BMD to adjust for differences among thiazolidinedione users.3 In addition, although not included our report, we found a similar prevalence of thiazolidinedione use among participants with diabetes who did and did not experience a fracture. In a recently published abstract,4 we noted that there were modest losses in total hip BMD associated with longer duration of thiazolidinedione use during the first 4 years of the study, although it is an unlikely explanation for our findings given the low prevalence of thiazolidinedione users over this time. Furthermore, several cohort studies conducted prior to the introduction of thiazolidinedione use have found an association between type 2 diabetes mellitus and incident fracture risk,5,6 suggesting that other aspects of diabetes are risk factors for fracture. As more participants with diabetes are exposed to thiazolidinedione use over time, it may be possible to determine if there is a specific effect on fracture that is independent of diabetes.