OBJECTIVES: Fracture risk is higher in older adults with type 2 diabetes mellitus (T2DM). Anti-diabetic drugs have different effects on bone metabolism. We performed a pairwise and network meta-analysis (NMA) including randomized controlled trials (RCTs) to study risk of bone fractures associated with thiazolidinediones (TZDs), compared to placebo or active drugs and rank anti-diabetic drugs for fracture risk in T2DM patients.

METHODS: We searched PubMed/Medline and clinical trial registration websites from inception to May 2, 2017, for published or unpublished RCTs comparing the effects of TZDs with comparators on fracture risk in T2DM patients. Pairwise meta-analyses within DerSimonian-Laird random effects model and NMA within a Bayesian framework were performed to calculate odds ratios (OR) to compare effect of anti-diabetic drugs on fracture risk. Treatment ranking was estimated using surface under the cumulative ranking curve (SUCRA) statistic for the outcome.

RESULTS: We identified