Fracture Risk Calculator

Fracture probability is estimated rather than measured because bone strength alone does not determine whether a bone breaks. The FRAX approach that this page approximates outputs two ten-year probabilities: hip fracture, and major osteoporotic fracture, meaning the combined chance of a clinical fracture at the hip, spine, distal forearm or proximal humerus. Inputs are age between 40 and 90, sex, body mass index, prior fragility fracture, parental hip fracture, current smoking, glucocorticoid exposure at prednisolone 5 mg daily or equivalent for three months or more, rheumatoid arthritis, secondary osteoporosis and alcohol at three or more units per day, with an optional femoral neck T-score. The T-score expresses bone mineral density in standard deviations below the young adult mean, and the 1994 World Health Organization cut-offs classify values at or above -1.0 as normal, -1.0 to -2.5 as osteopenia and -2.5 or below as osteoporosis. Risk factors act as multipliers on an age-specific baseline hazard, and the model runs against the competing risk of death, which is why probability does not simply rise without limit at extreme ages. In the United States the National Osteoporosis Foundation intervention thresholds of 3 percent ten-year hip risk or 20 percent major osteoporotic risk are widely used, while the United Kingdom applies age-dependent thresholds instead. The important limitations are that the clinical inputs are yes-or-no and so ignore dose-response for glucocorticoids and for multiple prior fractures, that falls history is not included at all despite being the proximate cause of most hip fractures, and that only femoral neck density is accepted, excluding lumbar spine density and trabecular bone score. The figures produced here are a simplified teaching approximation, not the country-calibrated official model.

Modern fracture risk assessment grew out of the 1994 World Health Organization study group report that first defined osteoporosis by a bone density T-score of -2.5 or below. That definition made large-scale screening possible but quickly exposed its own weakness: most fractures occur in people whose density falls in the osteopenic range, because age, prior fracture and other clinical factors carry risk that densitometry does not capture. Through the late 1990s and early 2000s John Kanis and colleagues at the World Health Organization Collaborating Centre for Metabolic Bone Diseases at the University of Sheffield published a series of meta-analyses quantifying each clinical risk factor as an independent hazard ratio, using pooled primary data from prospective population-based cohorts on several continents. Those meta-analyses were assembled into a single probability model and released as FRAX in 2008, alongside country-specific calibration using national hip fracture incidence and mortality statistics, since a given risk profile carries very different absolute probability in Sweden than in Japan. Guideline bodies moved to absolute risk within a few years, and treatment recommendations shifted from treating a density number to treating a probability. Later refinements have added adjustments for trabecular bone score, glucocorticoid dose, recency of prior fracture and hip axis length, reflecting the original criticism that binary inputs discard information.