Impact of Type 2 Diabetes on Lung Damage and Outcomes in treatment-naïve pulmonary tuberculosis

Objective: To assess the impact of type 2 diabetes mellitus (T2DM) on risk and severity of structural lung disease (SLD) in treatment-naïve pulmonary tuberculosis (TB) patients, identify early predictors, and develop a prediction model for timely clinical intervention. Methods: We retrospectively studied 296 newly diagnosed pulmonary TB patients (201 males and 95 females; mean age 53.8 ± 18.0 years) between December 2021 and December 2024. Patients were categorized into SLD and non-SLD groups based on chest CT findings assessed using integrated visual CT scoring and quantitative image analysis. Baseline demographics, symptoms, and laboratory parameters were compared between groups. Multivariate logistic regression identified independent predictors of SLD; receiver operating characteristic (ROC) analysis evaluated the diagnostic performance of the clinical prediction model. Results: SLD patients were older, predominantly male, and had higher T2DM prevalence (P<0.05). A combined five-parameter model incorporating sex, T2DM, C-reactive protein (CRP), fasting plasma glucose, and erythrocyte sedimentation rate (ESR) demonstrated excellent diagnostic performance with an AUC of 0.881 (95% CI: 0.855–0.927), sensitivity of 91.2%, and specificity of 72.5%. Internal Bootstrap validation (1000 iterations) confirmed model robustness with optimal calibration (Brier score: 0.1352). Among SLD patients, T2DM was independently associated with higher BMI, increased symptom burden, elevated inflammatory markers, and more severe CT abnormalities (P<0.05). Critically, new-onset SLD during 6-month therapy occurred in 63.0% of T2DM patients versus 21.4% without T2DM (P<0.001). Conclusion: T2DM significantly increases structural lung disease risk, severity, and progression in treatment-naïve TB patients. A clinical prediction model incorporating sex, T2DM, C-reactive protein, fasting plasma glucose, and erythrocyte sedimentation rate demonstrates good diagnostic performance (AUC 0.881) for early SLD identification. T2DM was independently associated with higher inflammatory markers and more severe radiological abnormalities. These findings highlight the importance of intensive monitoring, optimal glycemic control, and early diabetes screening to prevent SLD progression and improve TB treatment outcomes in this high-risk population.