Approved Research

Predicting the risk of Osteoporosis or Osteoporotic fracture using novel female-specific data fields.

Lay summary

Aims:

The first aim of our research is to assess whether currently used risk factors for osteoporosis and osteoporotic fracture risk show any bias in prediction performance between men and women. Our second aim is to investigate whether data points specific to women, such as number of pregnancies, time since perimenopause started amongst others, influence the risk of osteoporosis and associated fractures. Finally, we will assess whether including data points specific to women improves the accuracy of models predicting osteoporosis and osteoporotic fracture in women.

Project duration:

The duration of our project will be approximately 3 years from data acquisition to obtaining results.

Rationale:

Osteoporosis, which refers to low bone mineral density, is the most common bone disease in humans and one that disproportionately affects women. Osteoporosis leads to an increased risk of fractures, which leads to an increased morbidity and mortality. As effective interventions exist, preventative treatment, when applied to those at the highest risk of developing the disease, can significantly reduce the incidence of the disease and lower fracture rates. However, as osteoporosis is an 'invisible' disease that is largely symptomless pre-fracture, identification of those high-risk individuals that might benefit from treatment poses a great difficulty. Clinical tools that incorporate risk factors for osteoporosis, such as age, BMI, smoking status etc, estimate an individual's risk of fracture within a specified timeframe. Despite approximately 90% of osteoporosis sufferers being women, it is not known whether current risk predictor tools are equally accurate for predicting fractures in men and women, or whether female specific data fields improve the quality of risk prediction for women.

Impact:

Osteoporotic fractures have severe health implications for their sufferers and a large economic cost. For instance, hip fractures correspond to a 15-20% increased mortality rate within a year and up to 50% of fractures require long-term nursing care and suffer from decreased quality of life, social isolation, depression, and loss of self-esteem. More accurately predicting risk and developing new metrics will allow for preventative measures to be better targeted to the appropriate high-risk individuals and reduce the risk of life-changing fractures. Further, aspects of women's health have been underacknowledged in the history of scientific research. We hope that this study will help to close this gender data gap by exploring the impact of women's health features on an increasingly prevalent disease.