Original Article The Long-Term Effectiveness of Preventive Strategies for Osteoporosis in Postmenopausal Women: A Modeling Approach C. Le Pen 1,2 , F. Maurel 2 , G. Breart 3 , P. Lopes 4 , P.-F. Plouin 5 , M.-P. Allicar 6 and C. Roux 7 1 University Paris-Dauphine, Paris; 2 CLP Sante, Paris; 3 Ho Ãpital du Port Royal, Unite  INSERM 149, Paris; 4 CHU de Nantes, Nantes; 5 Ho Ãpital Broussais, Paris; 6 Lilly France, Saint-Cloud; and 7 Center of Bone Diseases Assessment, Ho Ãpital Cochin, Paris, France Abstract. Based on data from the literature, we have developed a computer-based simulation model to compare the long-term effectiveness of different pre- ventive strategies of osteoporotic fractures. The Markov model comprises 25 states, including states which describe women distributed according to three levels of fracture risk, fractures states, post-fracture states and a death state. We chose eight standard preventive strategies, which we compare with the `No Treatment' reference strategy. The ®rst two strategies consist in treating all 50-year-old women for 5 or 10 years with hormone replacement therapy HRT). Strategies 3 and 4 aim at assessing a 5-year course of treatment with bisphosphonates in osteopenic and osteoporotic 65- or 75-year-old women. Strategies 5 and 6 combine 5 years of HRT in all 50-year-old women with 5 years of bisphosphonates in osteopenic and osteoporotic women at 65 or 75 years. The last two strategies simulate 10 years of HRT in all 50-year-old women, followed by strategy 3 or strategy 4. Simulated life expectancy and mean ages of fracture occurrence ®t well with the observed data. All the preventive strategies tested reduced the number of fractures. Early 10-year HRT in all women, plus 5 years of bisphosphonates in women at risk of fractures at 65 or 75 years, are the most effective strategies, with an 18.4±19.0% reduction in all fractures, and a 25.6±26.1% reduction in the number of hip fractures. Strategy 2 has a similar outcome, thus demonstrating the value of treatment started early and sustained over a long period. The strategies implemented later, S3 and S4, only concern women at risk i.e., osteopenic or osteoporotic), and are less effective, with a 1.5±2.1% decrease in all fractures. The combined strategies, S5 and S6, produce intermediate results: a 12.9±13.5% reduction in the number of all fractures and a 17.5±17.9% reduction in hip fractures. Keywords: Fractures; Modeling; Osteoporosis; Post- menopausal Introduction Consensus conferences and a World Health Organization WHO) working party have de®ned osteoporosis as a systemic skeletal disease characterized by both low bone mass and microarchitectural deterioration of the bone [1]. The clinical signi®cance of osteoporosis lies in the fractures it occasions, mainly of the hip, the spine and the wrist, although many other bones may be affected [2]. Several prospective studies have shown that there is a gradient of risk between a decrease in bone mineral density BMD) and an increase in the incidence of fractures. In 50-year-old Caucasian women, the lifetime risk is estimated at about 15% for hip fracture and about 45% for all osteoporotic fractures [3]. Incidence rates of osteoporotic fractures are expected to rise dramatically over the decades to come, mainly because of demo- graphic changes [4]. Given the cost of osteoporotic fractures especially hip fractures), and other conse- quences in terms of mortality, disability, loss of autonomy and quality of life, preventive strategies are clearly required. Hormone replacement therapy HRT) [5,6,] and bisphosphonates [7,8] have, in observational Osteoporos Int 2000) 11:524±532 ß 2000 International Osteoporosis Foundation and National Osteoporosis Foundation Osteoporosis International Correspondence and offprint requests to: Professor Claude Le Pen, CLP Sante Â, 20, rue de Boulainvilliers, F-75016 Paris, France. Tel: +33 1 53922400. Fax: +33 1 53922404. e-mail: lepen@clp-sante.fr