<p><strong>Background and Objectives: </strong>There are methodological challenges with research in osteoporosis. The first is to predict the lifetime risk of hip fracture incorporating trends in the rates of hip fracture and mortality. The second is to identify optimum pharmacotherapy to reduce fractures in the absence of active-comparator trials. A third is to isolate the costs for incident and prevalent fractures. The objective of this thesis is to investigate these issues.</p> <p><strong> </strong><strong>Methods: </strong></p> <p>Project 1: From national administrative data, we estimated the lifetime risk of hip fracture for age 50 years to end of life using life tables.</p> <p>Project 2: A literature review identified randomized placebo-controlled trials with nine drugs for post-menopausal women to estimate odds ratios between drugs for fractures.</p> <p>Project 3: From provincial administrative data from Manitoba excess costs relative to matched controls were estimated for incident fractures, prevalent fractures and non-fracture osteoporosis. .</p> <p><strong>Results and Conclusions:</strong></p> <p>Project 1:<strong> </strong>For women and men, the crude lifetime risks of hip fracture was 12.1% and 4.6% respectively, and lower after incorporating trends, 8.9% and 6.7%. The risk is expected to continue to fall for both women and men.</p> <p>Project 2: Three drugs, zoledronic acid, teriparatide and denosumab, had the highest odds of reducing fractures and the largest effect sizes. Estimates were consistent between Bayesian and classical approaches.</p> <p>Project 3: All incident fracture types and most prevalent fractures had significant excess costs, and the results were robust to assessment of missing variances. Excluding prevalent fractures underestimates the cost of illness of fractures.</p> / Doctor of Philosophy (PhD)