Unveiling risk factors and predicting osteoporosis through bone density based aging model: a community-based cohort...

Osteoporosis is a progressive skeletal disorder influenced by multiple clinical and lifestyle factors. Early identification of individuals at high risk is essential for prevention and personalized management. This study aimed to identify key determinants of osteoporosis and to establish a bone density based aging model to evaluate deviations in bone health.

This retrospective study included 4,275 adults from Shunde District, Guangdong, China (mean age 60.62 years, 67.36% female). Bone mineral density (BMD) was measured at the lumbar spine (L2-L4) and proximal femur sites (femoral neck and trochanter) using dual-energy X-ray absorptiometry (DXA). Participants were classified as normal, osteopenic, or osteoporotic according to WHO and Chinese guidelines. Univariable and multivariable regression models were applied to assess the associations between clinical and lifestyle factors and osteoporosis risk. A bone density aging model was developed using support vector regression to estimate bone density age, and bone density age acceleration (BDAA) was calculated as a marker of deviation from age-matched bone health.

Risk analyses identified age, sex, body mass index, exercise frequency, OSTA score, and post-menopausal years (in women) as significant risk factors for osteoporosis. The bone density aging model showed good predictive performance (mean absolute error = 5.716 years, R² = 0.145). Higher BDAA was positively correlated with osteoporosis risk across bone health categories, including individuals without clinical diagnosis. In addition, sex-specific analyses showed that BDAA was elevated in male with insufficient exercise or smoking, and in female with osteopenia and osteoporosis, with the increase being more pronounced in osteoporosis patients, particularly under the influence of diet and alcohol consumption.

Our findings indicate that the bone density based biological aging models can effectively capture deviations in bone health, improve early identification and personalized risk stratification of osteoporosis. This approach may facilitate and support the development of precision medicine strategies in osteoporosis prevention and management.

Not applicable.