Combined neutrophil-to-lymphocyte, monocyte-to-lymphocyte, and platelet- to-lymphocyte ratios as inflammatory biomarkers in tuberculosis: A retrospective study from Lampang, Thailand

Background: Tuberculosis (TB) remains one of the major infectious diseases causing global morbidity and mortality. Systemic inflammation plays a central role in TB pathogenesis. Hematological indices derived from complete blood counts (CBC) including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and platelet-to-lymphocyte ratio (PLR) have been proposed as surrogate inflammatory biomarkers. Objectives: To investigate the role of CBC-derived ratios (NLR, MLR, and PLR) as inflammatory biomarkers in TB patients compared with healthy controls, and to assess their association with clinical characteristics and laboratory findings. Materials and methods: This retrospective study was conducted at Lampang Hospital, Thailand. Data from 213 TB patients and 170 healthy controls were collected from medical records and hematology reports. CBC-derived ratios were calculated and analyzed using Mann-Whitney U test, multivariate logistic regression adjusted for age, gender, and body mass index (BMI), and receiver operating characteristic (ROC) curve analysis. Results: Median NLR, MLR, and PLR values were significantly higher in TB patients compared with controls (NLR 4.21 vs 1.58; MLR 0.47 vs 0.16; PLR 223.0vs119.9; all p<0.001). In multivariate logistic regression analysis, only MLR (OR = 511.567; 95% CI: 7.384-35,439.660; p=0.004) remained independently associated with TB after adjustment, whereas NLR and PLR did not retain statistical significance. ROC analysis indicated good discrimination for NLR (AUC=0.852), MLR (AUC=0.893), and PLR (AUC=0.812). The combined model (NLR+MLR+PLR) achieved the highest accuracy (AUC=0.903; sensitivity=79.3%; specificity=92.4%). Conclusion: Elevated NLR, MLR, and PLR values reflect systemic inflammation in TB patients. Among these markers, MLR showed the strongest independent association with TB. Although their specificity is limited, combining these ratios enhances their ability to represent the host inflammatory response and may provide additional information for clinical assessment and disease monitoring.