AB0933 Development and Validation of a Nomogram for Prediction of Nonalcoholic Fatty Liver Disease in Psoriatic arthritis

Abstract

BackgroundPsoriatic arthritis (PsA) has been linked to an increased risk of metabolic syndrome (MetS). Non-alcoholic fatty liver disease (NAFLD), the hepatic manifestation of MetS, is now the commonest liver disease worldwide. About 65% of PsA patients suffer from NAFLD, and chronic systemic inflammation may be an important predisposing factor.ObjectivesThe purpose of this study was to establish and validate a diagnostic model nomogram for predicting the occurrence of NAFLD in patients with PsA.MethodsA total of 127 PsA patients (46 had NAFLD and 81 had no NAFLD) were enrolled in this study. Retrospectively collected clinical and serological parameters of these patients. The percentage and absolute number of lymphocytes and CD4+T cells were determined by Flow cytometry. The independent risk factors for NAFLD were screened in the PsA patients using univariate and multivariate binary logistic regression analyses and were used for construction of the nomogram prediction model. The AUROC (C index) was used to verify the model discrimination; the calibration curve and Hosmer-Lemeshow test were used to verify the model calibration; and the DCA curve was used to verify the clinical validity of the model.ResultsUnivariate and multivariate logistic regression analyses showed that Body Mass Index (BMI) (OR=1.25, P=0.001), serum triglyceride (TG) (OR=3.51，P=0.015) and peripheral blood Th1 cell percentage (OR=1.12, P < 0.001) is an independent risk factor for NAFLD in PsA patients, and an individualized nomogram prediction model was successfully established. The prediction model had a good discrimination power with AUROC (C-index) of 0.83 (95% CI: 0.76-0.90); the P value in the Hosmer-Lemeshow test was 0.683, suggesting a high reliability of the predicted probability by the model; the DCA curve indicating a good clinical efficiency of the model.ConclusionOur study shows that the establishment of a nomogram prediction model of PsA complicated with NAFLD patients is helpful for early clinical screening and identification of such high-risk patients.Figure 1.A. Example of prediction nomogram for risk of PsA complicated with NAFLD patients; B. The ROC curve of the prediction model; C. The calibration curve of the prediction model; D. The decision curve analysis of the prediction model.Disclosure of InterestsNone declared