Rheumatoid arthritis (RA) is a chronic inflammatory disease of unknown etiology, characterized by its autoimmune nature and the ability to affect multiple systems. It is distinguished by its symmetrical involvement of peripheral joints and has a global prevalence of approximately 0.46%.OP is one of the most common bone and joint diseases in RA patients, characterized by reduced bone mass and compromised bone microarchitecture.This study aimed to develop a nomogram based on electronic medical records (EMR) to estimate the probability of OP in RA patients.

This retrospective study enrolled 2,101 RA patients from a single center. Patients were randomly split into training (70%) and validation (30%) sets. Least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression were used to identify independent predictors from 29 potential clinical and laboratory variables. A nomogram was constructed based on the final predictors. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

Five independent predictors were identified (Tables 1 and 2): age (OR = 1.068), female gender (OR = 2.931), chronic gastritis (OR = 3.838), platelet count (OR = 1.002), and creatinine (OR = 0.985). The nomogram demonstrated good discrimination, with the area under ROC curve (AUC) values of 0.767 (95% CI: 0.742-0.793) in the training set and 0.785 (95% CI: 0.747-0.822) in the validation set (Tables 3 and Figure 1). Calibration curves showed good agreement between predicted and observed probabilities. DCA confirmed the model's clinical utility across a wide range of risk thresholds.

This study developed a practical nomogram incorporating five readily available predictors to individually estimate the probability of OP in RA patients. In this study, we utilized EMR data to conduct an in-depth analysis of the factors associated with OP in patients with RA and constructed a diagnostic model based on these factors. Our findings indicated that age, gender, CG, platelet count, and Crea are independent risk factors for OP in RA patients. The predictive model, built upon these factors, demonstrated robust performance in both the training and validation sets.The model shows robust predictive performance and has the potential to facilitate early screening and identification for OP in clinical practice.