A prospective cohort study in patients with type 2 diabetes found that treatment with a thiazolidinedione (glitazone) increased fracture risk in both men and women; pioglitazone appeared to be associated with a greater risk than rosiglitazone.
Observational studies and analysis of clinical trial data indicate that treatment with a glitazone increase the risk of bone fractures in women, however the evidence is still uncertain especially in relation to effects on risk in men. The authors of this study used healthcare data on a large population to study the association between bone fractures and treatment with a glitazone or a sulphonylurea. Their study population was derived from all residents of British Columbia at any time between January 1998 and December 2007 who were registered for healthcare services in the Province. Using prescription data, they identified all users of either a glitazone or a sulphonylurea between January 1996 and December 2007: sulphonylureas were chosen as the comparator because they were, like glitazones, more likely to be used second line. They then obtained data on peripheral and all fractures, and estimated adjusted hazard ratios for fracture in the two user groups.
After general exclusions, the source population was about 4.2 million people; of these, 127,581 began treatment with one of the study drugs during the relevant period, and 84,339 were eligible after exclusion (mainly because of treatment with a drug from the other study group). Average age of the study cohort was 59 years, and 43% were women.
There were 2,214 fractures in the study patients, most (76%) peripheral. Average time to fracture was 1.71, 1.66, and 1.44 patient-years in the sulfonylurea, rosiglitazone, and pioglitazone cohorts respectively.
Compared to those treated with sulphonylureas, patients receiving a glitazone had an increased risk of peripheral fracture (adjusted hazard ratio [HR], 1.28; 95% CI, 1.10 to 1.48) across the group. When the glitazones were examined separately in comparison to sulphonylureas, the risk with rosiglitazone for women was not significant (HR, 1.17; 95% CI, 0.91 to 1.50) whereas that for pioglitazone was (HR, 1.77; 95% CI, 1.32 to 2.38).
Overall fracture risk for men was not significantly different in the glitazone group (HR, 1.20; 95% CI, 0.96 to 1.50), however analysis by individual drug found no significant risk with rosiglitazone (HR, 1.00; 95% CI, 0.75 to 1.34) but a significant increase for pioglitazone (HR, 1.61; 95% CI, 1.18 to 2.20).
The authors conclude that their analysis further supports the association of glitazones with increased risk of fractures in women, and indicates a possible association between pioglitazone use and increased risk in men. They caution, however, that the 95% CI in the subgroup analysis overlap and the association should thus be regarded as a basis for further research rather than a definitive result. Overall, they conclude that glitazone treatment increases fracture risk in both men and women compared to sulphonylureas, and that the effect may be stronger with pioglitazone than with rosiglitazone. Further research is needed to gain greater certainty.
Arch Intern Med 2009; 169: 1395-402 (link to abstract)