Association Between Systemic Inflammation and Radiographic Severity in Knee Osteoarthritis: A Trial Baseline Cohort Analysis

Background Knee osteoarthritis (OA) is a common cause of pain and disability. Blood-based inflammatory markers may provide additional information on disease severity, but their clinical value is uncertain. Objectives To evaluate the association between inflammatory blood markers and radiographic severity of knee OA, and to assess their diagnostic performance. Methods In this cross-sectional study, 112 adults with radiographically confirmed knee OA [Kellgren-Lawrence (KL) grades I-III] were assessed. Demographic, hematological, and biochemical parameters were recorded. The Kruskal-Wallis test was used for between-group comparisons, Spearman's rank for correlations, ordinal logistic regression for predictors, and receiver operating characteristic (ROC) curves for diagnostic accuracy. Results Significant differences across KL grades were observed for monocyte count, neutrophil count, neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), erythrocyte sedimentation rate (ESR), and vitamin D (all p < 0.05). However, correlation analysis revealed only a weakly significant association between KL grade and ESR (ρ = 0.241, p = 0.010), while other markers showed no significant correlation. Logistic regression identified ESR as the only independent predictor [odds ratio (OR) 1.046, 95% confidence interval (CI): 1.013-1.083, p = 0.008]. ROC analysis showed fair discrimination for ESR [area under the curve (AUC) = 0.646], C-reactive protein (CRP) (AUC = 0.662), and CRP-to-albumin ratio (CAR) (AUC = 0.685), whereas NLR, MLR, and platelet-to-lymphocyte ratio (PLR) were not significant. Conclusions ESR was the only inflammatory marker consistently associated with radiographic severity of knee OA. Other indices demonstrated limited diagnostic value, indicating that evaluation of OA severity should not rely solely on hematological ratios.