The relationship between bisphosphonate adherence and fracture: is it the behavior or the medication? Results from the placebo arm of the fracture intervention trial.

Medication compliance may be a surrogate for factors that improve health outcomes such as fractures. Little is known about the size of this potential "healthy adherer" effect. We evaluated the hypothesis that compliance with placebo is associated inversely with bone loss and fractures among women participating in the Fracture Intervention Trial (FIT). Compliance with placebo and alendronate was evaluated using daily medication diaries. Women were defined as having high compliance if they took 80% or more of dispensed study medication. Change in bone mineral density (BMD) was assessed using mixed models comparing women with high versus lower compliance with placebo. Cox proportional-hazards models analyzed the association between placebo compliance and various types of fractures. Among 3169 women randomized to placebo, 82% had high compliance. Compared with women with lower placebo compliance, bone loss at the total hip was lower in compliant placebo-treated women (-0.43%/year versus -0.58%/year, p = .04). Among placebo-treated women, there were 46 hip, 110 wrist, 77 clinical vertebral, and 492 total clinical fractures. Compared with women with lower placebo compliance, women with high placebo compliance had a nonsignificant reduced risk for hip fracture [adjusted hazard ratio (HR) = 0.67, 95% confidence interval (CI) 0.30-1.45]. This trend was not observed for other fractures. Medication compliance may be a proxy for factors that confers benefit on reducing hip fracture (but not other types of fractures) independent of the effect of the medication itself. Nonrandomized studies of interventions designed to maintain or improve bone density and/or hip fracture may need to consider medication compliance as a confounder to better estimate true intervention effects.