With an increasingly ageing population, osteoporosis and osteoporosis-related fractures is fast becoming an important public health problem placing a considerable economic burden on health service resources. This does not account for the substantial pain, disability and indeed mortality incurred after a fracture, particularly a hip fracture. Osteoporosis is a systemic skeletal disorder which results from an imbalance in bone remodeling. This leads to a reduction in bone strength and increased susceptibility to fracture. It affects up to 1 in 2 women and 1 in 5 men. In the past 2 decades, there have been significant advances in bone biology which have helped in the understanding of the pathogenesis of osteoporosis and have led to improved therapies. In developing strategies for fracture prevention, it is important to identify those individuals with the highest fracture risk who will require pharmacological intervention. Treatment is aimed at fracture prevention and includes modification of general lifestyle factors which have been linked to fractures in epidemiological studies and ensuring optimum calcium and vitamin D intake as adjunct to active anti-fracture therapy. A number of drugs are now approved for the treatment of osteoporosis. This review article will describe the pathogenesis of osteoporosis and focus on the methods currently in use for the identification of patients at high fracture risk and will highlight their usefulness and limitations. The existing anti-fracture pharmacotherapies and those in development will be reviewed. Assessment of their effectiveness including the use of biochemical markers of bone turnover in this clinical context will be reviewed.

In this analysis, although both approaches were reasonably accurate in women, FRAX discriminated fracture risk poorly in men. These data support the concept that all algorithms need external validation before clinical implementation.

We thank Drs. Pluskiewicz and Drozdzowska for their interest in our work [1] and their thoughtful remarks [2]. We would like to comment on their remarks as follows:1. We agree that hip fracture is important, and ideally any validation study should consider a separate analysis for hip fracture. However the number of hip fractures was small (n=20) in our study and thus was not sufficient for a full stratified analysis. Despite these small numbers, the concordance for hip fracture prediction between FRAX and Garvan nomogram predicted risk of hip fracture was 0.73 for women and 0.29 for men. 2. The concordance between FRAX and Garvan nomogram predicted probabilities of fracture was generally higher in women than in men. For instance, for any osteoporotic fracture, the correlation between the two algorithms was 0.82 in women but only 0.20 for men. 3. Determining an appropriate risk threshold for treatment depends, among other things, on the effectiveness of treatment and risk-benefit considerations. The latter are, in turn, dependent on the wealth and healthcare system of a country. The National Osteoporosis Foundation recommended thresholds [3] of 20% for any fracture and 3% for hip fracture are relevant to the US setting but not necessarily to non-US populations.We consider that treatment thresholds need further country-specific studies.The assessment of fracture risk has entered a new era with individualized or absolute risk being the preferred approach. However, the predictive performance of any algorithm should be thoroughly evaluated in relevant cohorts before it is implemented in clinical settings. We thank our colleagues for their thoughtful contribution to the on-going discussion on fracture risk assessment.