Objective To explore the risk factors for elderly patients receiving knee arthroplasty for 1 year and concomitant periprosthetic osteolysis, and to construct a risk prediction model. Methods A total of 250 elderly patients receiving knee arthroplasty were selected as the research subjects, and they were divided into modeling group (n=175) or validation group (n=75). Patients were followed up for 1 year after surgery, and the modeling group was further assigned to concomitant group or non-concomitant group according to concurrent status of patients with periprosthetic osteolysis. Univariate analysis and multivariate Logistic regression model were used to analyze the risk factors for concomitant periprosthetic osteolysis after knee arthroplasty, and a risk prediction model was constructed using Nomogram based on the risk factors. The Hosmer⁃Lemeshow test was employed to validate the goodness of fit of the Logistic regression model, and the receiver operating characteristic (ROC) curve, calibration curve and decision curve were drawn to evaluate the discrimination, calibration and clinical applicability of the risk prediction model, respectively. Results A total of 116 cases (46.40%) of elderly patients receiving knee arthroplasty and concomitant periprosthetic osteolysis postoperatively. Compared with the non⁃concomitant group, the concomitant group exhibited statistically significant differences in the Hospital for Special Surgery (HSS) knee score, preoperative macrophage levels, fixation method, postoperative posterior condylar offset (PCO), postoperative posterior tibial slope (PTS), and postoperative joint line (P<0.05). Multivariate Logistic regression analysis revealed that a low HSS knee score, preoperative abnormal macrophage levels, biological fixation method, postoperative PCO≥2 mm, postoperative PTS<5°, and postoperative joint line≥4 mm were the risk factors for concomitant periprosthetic osteolysis in elderly patients after knee arthroplasty (P<0.05). Hosmer⁃Lemeshow test indicated favorable goodness of fit of Logistic regression model (χ²=6.423, P=0.600). The areas under the ROC curve of the risk prediction model for the modeling and validation groups were 0.891 and 0.885, respectively. The calibration curve demonstrated close agreement between predicted and actual probabilities. Decision curve analysis interpreted high net benefit values in both internal and external validation. Conclusion The rate of concomitant periprosthetic osteolysis in elderly patients undergoing knee arthroplasty is relatively high 1 year after surgery. A low HSS score for knee, abnormal macrophage levels, biological fixation method postoperative PCO≥2 mm, postoperative PTS<5 º, postoperative joint line≥4 mm are independent risk factors for concomitant periprosthetic osteolysis in elderly patients after knee arthroplasty. The risk prediction model constructed based on the aforementioned factors can help clinical medical and nursing personnel assess the risk of concomitant periprosthetic osteolysis in these patients after knee arthroplasty.

无