Home / Specialties / Bone & Joint / New Evidence Links Glitazones to Broken Bones

New Evidence Links Glitazones to Broken Bones

Jun 23, 2009
The largest study to date looking at whether the risk of bone fractures is increased in the setting of thiazolidinedione drugs (TZDs) suggests that fracture risk is more than 40% higher in people taking TZDs and that both men and women are vulnerable 
The analysis, presented by Dr. Merri Pendergrass (Harvard University, Boston, MA) during the ADA 2009 Scientific Sessions, looked at almost 70,000 patients taking either rosiglitazone (Avandia, GlaxoSmithKline) or pioglitazone (Actos, Takeda) and, unlike other studies, found no difference in fracture risk between the two TZDs.
Pendergrass stated that, “I think these agents should be avoided in people at high risk for fracture–unless, of course, particular benefits for a particular patient seem high.”  For example, a patient with high hypoglycemia risk who is not a good candidate for other classes of medications. Of note, Pendergrass continued, diabetes itself increases the risk for fracture. “So a postmenopausal woman with diabetes would have a particularly high risk–especially if she smoked, had a family history of fracture, or other additional fracture risk factors.”
Pendergrass and colleagues reviewed the Medco database–more than 13 million people–looking for all patients between the ages of 43 and 63 at study onset with diabetes and a TZD prescription or any diabetic patients within the same age group taking metformin, exenatide, or a sulfonylurea. They then used a linear-regression model to adjust for age, chronic obstructive pulmonary disease (COPD), asthma, osteoporosis, stroke, and prior fracture to compare fracture risks in patients with “glitazone” prescriptions and in patients with no TZD prescription over the study period (January 2006 through June 2008).
They found that fracture rates were higher among all patients taking TZDs, with no difference between those taking pioglitazone versus rosiglitazone. Fracture rate was also higher in both women taking TZDs versus controls and in men taking TZDs versus controls (although the fracture rate was higher in women than men). According to Pendergrass, this is the first study to show an increased fracture rate in men as well as women. Of note, however, an analysis that looked only at recent TZD prescriptions did not find an increased fracture risk in men, suggesting that men may need to be on the drugs for longer than 18 months before developing an increased risk of broken bones.
Both older women and older men (age 50 to 65 at study conclusion) had a significantly increased fracture risk, but in younger subjects (age 43 to 49) only women were at significantly increased risk. Of note, the study could not control for alcohol consumption and smoking, both of which are known risk factors for fractures. Older adults (age 65 and older) could not be included in the analysis, due to a lack of data, the authors note.
Odds Ratios for Fracture Risk 


Odds ratio

95% CI

All patients



All women



All men



New TZD Rx, women*



New TZD Rx, men*



Rosiglitazone vs pioglitazone



*<18 mo 
Pendergrass has previously disclosed research grant support from Novo-Nordisk. 
Aubert RE, Herrera V, Tully L, et al. Thiazolidinedione treatment increases the risk of fracture. American Diabetes Association 2009 Scientific Sessions; June 7, 2009; New Orleans, LA. 601-P. Available here.