To identify risk factors for osteoporosis and develop a predictive model incorporating conventional clinical data and traditional Chinese medicine (TCM) syndrome patterns. From December 2019 to January 2022, a multi-stage sampling approach was used to recruit adults aged 30–82 years from 12 community-level districts or rural towns in Shanghai, Jilin Province, and Jiangsu Province. Univariate analysis and multivariable logistic regression were employed to examine risk factors and construct osteoporosis prediction models separately for women and men. Model performance was assessed using the receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow goodness-of-fit test. The study enrolled 3,000 participants, comprising 2,243 women (75%) and 757 men (25%). The logistic regression model for osteoporosis in women was: Logit(P) = −2.946 + 0.960 (age ≥50 years) + 0.633 (BMI ≥24 kg/m²) − 0.545 (daily sunlight exposure >30 min) + 0.519 (no dairy product intake) + 0.827 (coronary heart disease) + 0.383 (lumbar disc herniation) + 0.654 (no calcium/vitamin D supplementation) − 0.509 (insomnia) + 0.580 (flushed face and red eyes) + 1.194 (thready and rapid pulse) + 1.309 (sunken and slow pulse). The model for men was: Logit(P) = −1.152 − 0.644 (daily sunlight exposure >30 min) + 0.975 (no calcium/vitamin D supplementation) − 0.488 (insomnia). The area under the ROC curve was 0.743 for the female model and 0.679 for the male model. Hosmer-Lemeshow tests indicated good calibration (p > 0.5 for both models). Risk factors for osteoporosis differ notably between women and men. TCM syndrome elements are significantly associated with osteoporosis risk. Prediction models that integrate routine clinical variables with TCM syndromes demonstrate acceptable discriminative ability and calibration for assessing osteoporosis risk.
