Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study.
F, S., Y, Z., H, M., M, Y., & Y, Y. (2026). Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study.. Frontiers in cellular and infection microbiology. https://doi.org/10.3389/fcimb.2026.1836144
F S, Y Z, H M, M Y, Y Y. Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study.. Frontiers in cellular and infection microbiology. 2026; doi: 10.3389/fcimb.2026.1836144
F S, Y Z, H M, et al. Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study.[J]. Frontiers in cellular and infection microbiology. 2026. DOI: 10.3389/fcimb.2026.1836144.
@article{f2026,
author = {Sun F and Zhou Y and Ma H and Ye M and Yang Y},
title = {Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study.},
journal = {Frontiers in cellular and infection microbiology},
year = {2026},
doi = {10.3389/fcimb.2026.1836144},
note = {PMID: 42666300},
}
TY - JOUR AU - Sun F AU - Zhou Y AU - Ma H AU - Ye M AU - Yang Y TI - Red blood cell distribution width is associated with coronary artery disease in latent tuberculosis infection: a machine learning-based study. T2 - Frontiers in cellular and infection microbiology PY - 2026 DO - 10.3389/fcimb.2026.1836144 AN - PMID:42666300 ER -
BACKGROUND: Latent tuberculosis infection (LTBI) is associated with an increased risk of coronary artery disease (CAD), potentially mediated by systemic inflammation. Traditional lipid-based risk models may have limited predictive performance in tuberculosis-affected populations. OBJECTIVES: This study aimed to identify CAD risk factors in individuals with LTBI and to develop and externally validate a predictive model using a Boruta-LASSO machine learning approach. METHODS: In this dual-center retrospective case-control study, patients who underwent both coronary CT angiography (CCTA) and interferon-γ release assay (IGRA) were enrolled. Cases were patients with CAD diagnosed by CCTA; controls were those without CAD. Candidate predictors were selected using Boruta and LASSO regression, then incorporated into a multivariable logistic regression model to construct a nomogram, which was externally validated. RESULTS: Among 862 screened records, 616 participants were enrolled (derivation cohort, n = 453; validation cohort, n = 163). Age, red blood cell distribution width (RDW), and hemoglobin A1c (HbA1c) were identified as independent predictors and incorporated into the nomogram. The model achieved an area under the receiver operating characteristic curve (AUC) of 0.783 (MAE 0.285) in the derivation cohort and 0.738 (MAE 0.303) in the external validation cohort, with favorable net clinical benefit. A significant RDW-LTBI interaction was associated with a 1.737-fold (95% CI 1.092-2.877, P = 0.025) increase in CAD risk. After Benjamini-Hochberg correction (FDR < 0.05), RDW was positively correlated with interleukin-6 (IL-6; r = 0.345), C-reactive protein (CRP; r = 0.251), serum amyloid A (SAA; r = 0.238), and monocyte-to-lymphocyte ratio (MLR; r = 0.214). CONCLUSION: A nomogram for predicting CAD risk in LTBI-positive individuals was developed and externally validated using three routine clinical variables. A significant multiplicative interaction between RDW and LTBI was identified, and positive correlations between RDW and pro-inflammatory markers were observed specifically in LTBI-positive patients. These findings are consistent with the hypothesis that LTBI-associated chronic inflammation may contribute to CAD risk and that this low-cost tool may facilitate early risk stratification in resource-limited settings.