![[Visit Client Website]](/img/acr/banner_2009.gif)
Search
Print Page
Browse by Day: 
Purpose: Several observational studies have demonstrated a relationship between compliance to osteoporosis medications and fracture. While medication compliance would be expected to reduce risk, compliance may also be a surrogate for other healthy behaviors that affect fracture risk. Little is known about the size of this potential effect; thus we evaluated the relationship between compliance to placebo and clinical fractures among women participating in the Fracture Intervention Trial (FIT).
Method: Compliance to placebo among participants in the placebo arm of the FIT was evaluated using pill counts. Women were defined as having high compliance if they took ≥ 80% of dispensed placebo. Cox proportional hazards models analyzed the relationship between placebo compliance and various types of clinical fractures. Factors that could affect fracture outcomes were included in multivariable models and included age, baseline bone mineral density, height, body mass index, general health, smoking status, calcium intake, prior clinical fracture, and baseline radiographic vertebral fracture.
Results: Among 3169 women randomized to placebo, 81.8% had high compliance. Among placebo treated women, there were 45 hip, 111 wrist, 75 clinical vertebral and 489 total clinical fractures over a median follow-up period of 4.03 years. For wrist, hip or spine fractures combined, high compliance was associated with a statistically significant 30% lower fracture incidence after adjustment, with similar trends for individual fracture types (Table).
Conclusion: Women participating in a clinical trial
of alendronate that were highly compliant to placebo had significantly reduced
risk for wrist, hip or clinical spine fractures, suggesting that adherence may
be a proxy for other behavior that confers benefit independent of the effect of
the medication. Previous observational studies that have reported benefit of
bisphosphonates may in part be confounded by healthy behaviors associated with
medication adherence.
Table: Relationship between compliance* and fracture | ||||
Crude fracture rate | Hazard Ratio unadjusted | Hazard Ratio adjusted‡ | ||
| Compliant | Non-compliant |
|
|
Hip fracture | 0.35 | 0.52 | 0.65 (0.34, 1.31) | 0.82 (0.40, 1.69) |
Wrist fracture | 0.90 | 1.10 | 0.82 (0.52, 1.29) | 0.75 (0.47, 1.19) |
Spine fracture | 0.55 | 1.00 | 0.55 (0.33, 0.91) | 0.66 (0.93, 1.11) |
Wrist, hip or spine fracture | 1.68 | 2.51 | 0.67 (0.49, 0.92) | 0.70 (0.51, 0.96) |
Any clinical fracture | 4.31 | 4.90 | 0.88 (0.71, 1.10) | 0.90 (0.72, 1.13) |
‡ adjusted for age, baseline BMD, height, BMI, general health, smoking status, baseline calcium intake, baseline vertebral fracture, other clinical fracture after age 45
* compliance measured using ≥ 80% threshold