Predicting fractures in an international cohort using risk factor algorithms without BMD.

Area of Science:

• Gerontology
• Epidemiology
• Clinical Medicine

Background:

• Postmenopausal osteoporosis poses a significant fracture risk.
• Accurate fracture risk prediction is crucial for timely intervention.
• Bone mineral density (BMD) is a key factor, but clinical risk factors also play a role.

Purpose of the Study:

• To compare the predictive performance of different models for incident fractures in postmenopausal women using only self-reported clinical risk factors.
• To evaluate the WHO Fracture Risk (FRAX) and Garvan Fracture Risk Calculator (FRC) against a simple age and prior fracture model.

Main Methods:

• The GLOW study enrolled 19,586 postmenopausal women aged 60+ from primary care in 10 countries.
• Data on clinical risk factors and previous fractures were collected via questionnaires over a 2-year follow-up.
• Model performance was assessed using the C-index for predicting hip, major, and osteoporotic fractures.

Main Results:

• FRAX (C-index 0.78) and FRC (C-index 0.76) demonstrated moderate accuracy for hip fracture prediction.
• Prediction for major (C-index 0.61) and osteoporotic (C-index 0.64) fractures showed lower accuracy.
• A simple model using age and prior fracture history performed comparably to FRAX and FRC for hip fracture prediction (C-index 0.78).

Conclusions:

• Clinical risk factors alone are sufficient for estimating fracture risk in postmenopausal women without requiring BMD measurements.
• More complex models incorporating multiple clinical factors, including falls, did not outperform simpler models in this population.
• Parsimonious models based on age and fracture history are effective for predicting future fractures.