Combining information from multiple bone turnover markers as diagnostic indices for osteoporosis using support vector...

Area of Science:

• Biomedical Science
• Osteoporosis Research
• Diagnostic Tool Development

Background:

• Osteoporosis (OP) is a systemic bone disease.
• Dual-energy X-ray absorptiometry (DXA) is the current gold standard for OP diagnosis.

Purpose of the Study:

• To evaluate the diagnostic potential of combined bone turnover markers (BTMs) for osteoporosis.
• To assess the efficiency of various BTM combinations using machine learning models.

Main Methods:

• Recruited 9053 Chinese postmenopausal women (2464 OP patients, 6589 controls).
• Assayed serum levels of six BTMs: BAP, BSP, CTX, OPG, OST, and sRANKL.
• Constructed support vector machine (SVM) models to analyze BTM combinations for OP diagnosis.

Main Results:

• Increasing the number of BTMs enhanced the predictive power of SVM models.
• A single BTM (BAP) achieved a kappa coefficient of 0.7783 compared to DXA.
• The model using all six BTMs achieved a high kappa coefficient of 0.9786.

Conclusions:

• Single BTMs are insufficient for accurate osteoporosis diagnosis.
• Combinations of multiple BTMs in SVM models demonstrate high diagnostic agreement with DXA.
• This multi-BTM approach shows potential as a clinical diagnostic tool for osteoporosis.