AUTHOR=Yang Fan , Sang Weihua , Liu Yongqing , Wang Jun 

TITLE=The C-reactive protein-to-albumin ratio as a diagnostic biomarker for rheumatoid arthritis: a cross-sectional NHANES analysis

JOURNAL=Frontiers in Medicine

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1624527

DOI=10.3389/fmed.2025.1624527

ISSN=2296-858X

ABSTRACT=BackgroundRheumatoid arthritis (RA) is a chronic inflammatory disorder that leads to joint damage, cartilage and bone destruction, and functional disability. The C-Reactive Protein to Albumin Ratio (CAR), an emerging biomarker reflecting systemic inflammation and nutritional status, has demonstrated prognostic value in various diseases. However, its utility in predicting clinical outcomes in RA patients remains underexplored, warranting further investigation to assess its potential role in disease management and prognosis. This cross-sectional study investigates the potential relationship between CAR and RA in United States adults, develops a clinical prediction model, and validates its effectiveness.ObjectiveTo investigate the association between the CAR and RA using data from the National Health and Nutrition Examination Survey (NHANES).MethodsThis large-scale, cross-sectional study analyzed data from the NHANES database between 1999 and 2018 (excluding 2011–2014). The CAR was calculated as the ratio of C-reactive protein (CRP) to albumin (ALB) levels. RA status was identified through self-reported questionnaire data. Weighted multivariate regression models and subgroup analyses were used to examine the association between CAR and RA. Restricted cubic splines (RCS) were employed to evaluate potential non-linear relationships, and sensitivity analyses were conducted to assess the robustness of the results. Least absolute shrinkage and selection operator (LASSO) were utilized for variable selection in the prediction model. Decision curve analysis (DCA) and receiver operating characteristic (ROC) curve analysis were applied to assess the predictive performance of the models.ResultsThis study included a total of 20,733 patients, among whom 1,744 individuals (4.95%) were diagnosed with RA. After controlling for all covariates, the results of multivariate logistic regression analysis indicated a statistically significant correlation between higher Ln(CAR) levels and the increased incidence of RA (OR:1.77 (95% CI, 1.39–2.25); p < 0.001). The interaction test results showed that there was no statistically significant influence in this specific association. RCS regression modeling demonstrated a linear relationship between Ln-CAR and RA risk. After variable screening, we constructed an RA prediction model incorporating CAR, and the results were visualized using a nomogram. The area under the curve (AUC) was 0.749 (95% CI, 0.738–0.760), and DCA indicated that the model holds clinical significance.ConclusionThese findings suggest that CAR may serve as a promising inflammatory biomarker for predicting the presence of RA. In the RA prediction model incorporating CAR, we validated the effectiveness and clinical utility of this model, providing evidence that CAR can serve as a biomarker for RA risk prediction. Further prospective studies are warranted to validate its clinical utility in RA risk stratification and management.