ObjectivePreclinical studies, observational studies, and clinical trials suggest that thiazolidinediones (TZDs) reduce bone mineral density (BMD) and increase fracture risk. Most of the evidence on the skeletal effects of TZDs is from studies of rosiglitazone. We set out to investigate the magnitude and etiology of the adverse skeletal effects of pioglitazone.DesignDouble-blind, randomized controlled trial.Trial registrationAustralia New Zealand Clinical Trials Registry, actr.org.au Identifier: ACTRN12607000610437, date of registration 28/11/07.MethodsA total of 86 people with type 2 diabetes mellitus (T2DM) or impaired glucose tolerance (IGT), median age 64 years, were randomized to receive either pioglitazone 30 mg/day or placebo for 1 year, in addition to their usual diabetes treatments. The primary outcome was change in lumbar spine BMD; secondary outcomes included changes in BMD at other sites and in biochemical markers of bone turnover.ResultsChange in spine BMD was not altered by treatment with pioglitazone (Ptreatment×time=0.5). After 1 year, the mean (95% CI) between-groups difference in lumbar spine BMD was −0.7% (−2.1, 0.7). Pioglitazone increased bone loss at the proximal femur (Ptreatment×time=0.03). After 12 months, the between-groups difference in total hip BMD was −1.2% (−2.1, 0.2). Pioglitazone did not alter change in BMD at other skeletal sites, nor did it affect changes in the levels of either of the biochemical markers of bone turnover, procollagen type 1 N-terminal propeptide, or β-C-terminal telopeptide of type 1 collagen.ConclusionsOver 1 year, treatment with pioglitazone 30 mg/day did not produce consistent effects on either BMD or bone turnover in people with T2DM or IGT. The mechanism(s) by which pioglitazone increases fracture risk in T2DM is unclear.
Compromised cortical bone compartment in type 2 diabetes mellitus patients with microvascular disease
