Abstract Zoledronate is most effective when bone turnover is elevated, as in postmenopausal women. Individuals with type 2 diabetes mellitus show lesser bone mineral density gains compared to non-diabetes mellitus individuals with zoledronate, and fractures often occur despite preserved bone mineral density. It is unclear whether zoledronate reduces fracture risks in type 2 diabetes mellitus. This study aimed to evaluate skeletal outcomes, including fracture incidence, over 5 years in postmenopausal osteoporotic women with and without type 2 diabetes mellitus. This prospective cohort included postmenopausal women with bone mineral density T-scores of≤–2.5 at the lumbar spine, femoral neck, or hip. Participants were classified as type 2 diabetes mellitus or non-diabetes mellitus. All received annual zoledronate (4 mg), daily calcium (1000 mg), and cholecalciferol (500 IU), with standardized fall-prevention measures. Fracture history was recorded at follow-up visits, and annual spine radiographs were performed to detect morphometric vertebral fractures. Women completing≥5 years of follow up were included in the final analysis. The primary end point was fracture incidence; secondary end points were changes in bone mineral density and bone turnover markers. Of 221 women enrolled, 150 completed 5 years (63 type 2 diabetes mellitus and 87 non-diabetic mellitus; median age 59 y). The baseline bone mineral density was similar, but bone turnover markers were lower in type 2 diabetes mellitus. At a minimum of 60 months follow up, 20 new fractures occurred in 15 women: 8 (12.7%) with type 2 diabetes mellitus and 7 (8.0%) without. The relative risk (RR) was 1.58 (95% confidence interval: 0.60–4.13; p=0.2). Both groups showed comparable bone mineral density improvements. Despite differences in baseline bone turnover, fracture incidence did not differ significantly between postmenopausal women with and without type 2 diabetes mellitus treated with annual zoledronate over a 5-year follow-up. However, the low number of fracture events and wide confidence intervals limit definitive inference.
Giri et al. (Wed,) studied this question.