Bisphosphonate therapy was effective for reducing fracture incidence among postmenopausal women with diabetes, a post hoc analysis of randomized trial data found.

Among women with diabetes, the relative risk of nonvertebral fracture was 0.52 (95% CI 0.33-0.80) and the relative risk of morphometric vertebral fracture was 0.34 (95% CI 0.18-0.67), reported Ann Schwartz, PhD, of the University of California San Francisco, at theAmerican Society for Bone and Mineral Research meeting.

For nondiabetic women, the relative risks of nonvertebral and vertebral fractures were 0.83 and 0.39, respectively.

“Clearly, the bisphosphonate treatment was protective in both diabetic and nondiabetic women,” she said.

Type 2 diabetes is associated with a higher fracture risk at a given bone density, but the reasons for this greater bone fragility are unclear. However, there is evidence that diabetic bone is characterized by reduced bone formation and higher levels of advanced glycation end products.

“One hypothesis is that this accumulation of advanced glycation end products contributes to the bone fragility in diabetic bone. This has led to a theoretical concern that bisphosphonates, which have an antiresorptive mechanism of action, might not be an effective treatment for fracture prevention in this setting,” she said.

To test this hypothesis, she and her colleagues analyzed data from two randomized trials, the Fracture Intervention Trial (FIT) of alendronate (Fosamax) and the Health Outcomes and Reduced Incidence with Zoledronic Acid Once Yearly (HORIZON) pivotal fracture trial.

FIT included just under 6,500 postmenopausal women whose femoral neck T-score was -1.6 or lower and who received alendronate, 5 to 10 mg/day or placebo for up to 4 years. Their mean age was 68 and mean body mass index (BMI) was 25.1 kg/m².

HORIZON enrolled more than 7,700 women with osteoporosis based on their femoral neck T-score and who were given a once-yearly infusion of 5 mg zoledronic acid for up to 3 years. Their mean age was 73 and mean BMI was 25.3 kg/m².

Diabetes was ascertained by self-report or by self-report of use of an antidiabetic medication. Fasting glucose levels were available for the women in HORIZON and for 20% of the women in FIT, and for others nonfasting glucose levels were used.

In FIT, 4.2% of participants had baseline diabetes, as did 7.7% in HORIZON. Most of the women were probably type 2 diabetics, although this was not specifically ascertained, Schwartz noted.

For the analyses of fractures, the two studies were combined.

Overall, 87 of 869 women with diabetes experienced one or more nonvertebral fracture, as did 1,411 of 13,288 women without diabetes.

Morphometric vertebral fractures were reported in 44 of 769 diabetic women and in 786 of 12,312 nondiabetic women.

In both FIT and HORIZON, bone turnover markers also were suppressed in both diabetic and nondiabetic women, supporting the finding that the anti-fracture efficacy of bisphosphonate therapy was not inferior in diabetic women, Schwartz concluded.

A limitation of the study was its post hoc design, in which assessment for diabetes was not part of the original protocol.