Population-based study of the effectiveness of bone-specific drugs in reducing the risk of osteoporotic fracture

Aim Evidence supports bone-specific drugs (BSDs) efficacy in the fracture risk reduction. But treatment rates for osteoporosis among high-risk patients are far below the recommended guidelines. A major concern about BSDs is the lack of adherence with treatment. Objective To determine if BSDs decrease fracture risk in high-risk elderly women in real clinical setting. Methods A nested case-control design was used in a cohort of elderly women from the Quebec health databases. Women enter into the cohort if they are 70 years or older between 1995 and 2003. Nested case-controls were designed for women with a diagnosis of osteoporosis (OP) and for those with a prior fracture. All cases of fractures occurring during follow-up were matched with 10 randomly selected controls based on age, time period, bone mass density testing, and having a diagnosis of OP or a prior fracture. Use of BSDs before the index date was categorized as follows: short-term (<= 1 year), intermediate-term (> 1 and <= 3 years), and long-term (>3 years). We used an adjusted conditional logistic regression model to assess BSD effect on fracture. Results Among 3170 women who had a fracture, of these women, 1824 had OP and 1346 had a prior fracture. Only long-term exposure to BSDs among women with OP reduced the fracture risk by 16% (odds ratio: 0.84; 0.73-0.97). Among women with OP, a high number of medical services or use of anticonvulsants or narcotics increased the fracture risk by 12-73%. Among women with a prior fracture, a high number of medical services or risk of fall or use of benzodiazepines, antidepressants, or narcotics increased the fracture risk by 23-77%. Conclusion The incidence of fractures decreased by 16% among women with OP when more than 80% of BSDs was used for at least 3 years. Among women with a prior fracture, fracture risk reduction was not significant, Exposure to BSDs among women with a prior fracture is troubling, given that only approximately 12% of these individuals were being treated, and only 2% was using BSDs for the long term. Copyright (c) 2008 John Wiley & Sons, Ltd.