Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains a major communicable disease and an important cause of morbidity and mortality worldwide. Pulmonary tuberculosis is the most common clinical form and is primarily responsible for transmission in the community. Despite improvements in molecular diagnosis and effective treatment regimens, the global TB burden remains substantial, particularly in low- and middle-income countries. India contributes a considerable proportion of the global TB burden. The World Health Organization has identified diabetes mellitus, human immunodeficiency virus infection, undernutrition, smoking and harmful alcohol consumption as important determinants of the TB epidemic [1]. Simultaneously, the prevalence of diabetes mellitus and intermediate hyperglycaemia is increasing rapidly, creating a growing overlap between TB and metabolic disease. Diabetes mellitus increases the risk of developing active tuberculosis through impairment of innate and adaptive immunity. Persistent hyperglycaemia may adversely affect macrophage activation, cytokine responses, neutrophil function and cell-mediated immunity, thereby reducing the host’s ability to contain M. tuberculosis. Among patients with pulmonary TB, diabetes and poor glycaemic control have been associated with a greater symptom burden, higher bacillary load, extensive radiological involvement, cavitary disease and delayed sputum conversion. Diabetes may also increase the probability of unfavourable outcomes, including treatment failure, loss to follow-up, relapse and death [2,3]. The relationship between glycaemic status and tuberculosis is complex and bidirectional. Previously undiagnosed diabetes may be detected during TB evaluation, while the physiological stress and inflammation associated with active TB may cause transient hyperglycaemia. Therefore, assessment based only on a single random blood glucose measurement may misclassify some patients. Measurement of fasting plasma glucose and glycated haemoglobin (HbA1c) can provide a more comprehensive evaluation of glycaemic status. Patients can consequently be classified as having normoglycaemia, prediabetes or diabetes using accepted diagnostic criteria [4]. Systematic screening for diabetes among patients with TB is particularly relevant in countries experiencing a high burden of both conditions. Early identification and appropriate management of hyperglycaemia may improve clinical recovery, facilitate sputum conversion and reduce adverse treatment outcomes. Current TB programmes therefore emphasize integrated, patient-centred care and evaluation of important comorbidities during TB treatment [5]. However, evidence regarding the relationship between different levels of glycaemic control, clinical severity and TB treatment outcomes remains limited in many local settings. The present study was therefore undertaken to determine glycaemic status among patients with pulmonary tuberculosis and assess its association with clinical severity, microbiological response and treatment outcomes. AIM To assess the association of glycaemic status with clinical severity and treatment outcomes among patients with pulmonary tuberculosis. OBJECTIVES 1. To determine the glycaemic status of patients diagnosed with pulmonary tuberculosis using fasting plasma glucose and HbA1c levels. 2. To assess the association between glycaemic status and the clinical, microbiological and radiological severity of pulmonary tuberculosis. 3. To compare sputum conversion and treatment outcomes among pulmonary tuberculosis patients with normoglycaemia, prediabetes and diabetes mellitus.

Source of Data The study population consisted of eligible patients diagnosed with pulmonary tuberculosis who attended the outpatient department or were admitted to the Department of Respiratory Medicine/General Medicine of the selected tertiary care hospital during the study period. Information was obtained through patient interviews, clinical examination, laboratory investigations, chest imaging, treatment records and the Nikshay portal or institutional TB register, wherever applicable. Study Design A hospital-based prospective observational cohort study was conducted. Patients were classified according to their baseline glycaemic status and were followed during antitubercular treatment to document microbiological response and final treatment outcomes. Study Location The study was conducted in the Department of Respiratory Medicine/General Medicine in collaboration with the Department of Biochemistry, Department of Microbiology and the National Tuberculosis Elimination Programme unit of a tertiary care teaching hospital. Study Duration The study was conducted over 18 months, including 12 months for participant recruitment and baseline assessment and an additional six months for follow-up, outcome assessment, data analysis and report preparation. Sample Size A total of 200 eligible patients with pulmonary tuberculosis were included in the study. Participants were enrolled consecutively until the required sample size was achieved. The sample size was calculated using the formula: n=(Z_(1-α/2)^2×p(1-p))/d^2 Where: n= minimum required sample size Z_(1-α/2)= 1.96 at a 95% confidence level p= expected prevalence of dysglycaemia among pulmonary TB patients d= absolute precision After accounting for incomplete observations or loss to follow-up, the final sample size was fixed at 200 participants. Inclusion Criteria Patients aged 18 years or older. Patients with newly diagnosed, microbiologically confirmed pulmonary tuberculosis. Patients registered for antitubercular treatment under the National Tuberculosis Elimination Programme. Patients with or without a previous diagnosis of diabetes mellitus. Patients who provided written informed consent and agreed to undergo glycaemic assessment and follow-up. Exclusion Criteria Patients with exclusively extrapulmonary tuberculosis. Patients with rifampicin-resistant or multidrug-resistant tuberculosis at baseline. Patients who had received antitubercular treatment for more than two weeks before enrolment. Pregnant or lactating women. Patients receiving systemic corticosteroids or other medicines known to markedly affect blood glucose levels. Patients with severe renal, hepatic or haematological disease likely to interfere with HbA1c interpretation. Patients with known haemoglobinopathies, recent blood transfusion or severe anaemia. Patients with human immunodeficiency virus infection or other severe immunosuppressive conditions, when these conditions were considered major confounders. Patients who declined consent or were unlikely to complete follow-up. Procedure and Methodology After obtaining approval from the Institutional Ethics Committee, all eligible patients were approached consecutively. Written informed consent was obtained before enrolment. Each participant underwent a detailed interview and clinical examination using a predesigned and pretested case-record form. Information regarding age, sex, residence, education, occupation, socioeconomic status, smoking, alcohol use, previous TB history, family history of diabetes, known diabetes, duration of diabetes, antidiabetic medication and associated comorbidities was recorded. Presenting symptoms such as cough, expectoration, fever, haemoptysis, breathlessness, chest pain, loss of appetite and weight loss were documented along with their duration. Height and weight were measured using standardized equipment, and body mass index was calculated as weight in kilograms divided by height in metres squared. Vital signs, pallor, oxygen saturation, respiratory findings and other relevant clinical signs were recorded. The clinical severity of TB was assessed using a predefined clinical severity score, such as the Bandim TBscore, incorporating symptoms and clinical signs. Patients were categorized into mild, moderate or severe disease according to the selected score. Microbiological severity was evaluated using baseline sputum smear grading or semi-quantitative molecular bacterial load. Radiological severity was assessed from chest radiographs according to the number of lung zones involved, unilateral or bilateral disease, and the presence, number and extent of cavities. When available, computed tomography findings were also recorded. Fasting plasma glucose and HbA1c were measured at baseline. Based on glycaemic findings, participants were categorized as follows: Normoglycaemia: HbA1c <5.7% and fasting plasma glucose <100 mg/dL. Prediabetes: HbA1c 5.7-6.4% or fasting plasma glucose 100-125 mg/dL. Diabetes mellitus: HbA1c ≥6.5%, fasting plasma glucose ≥126 mg/dL, random plasma glucose ≥200 mg/dL with classical symptoms, or a documented history of diabetes receiving treatment. In patients without unequivocal hyperglycaemia, an abnormal result was confirmed by repeat testing. Patients with newly detected diabetes or uncontrolled glycaemia were referred to the physician or endocrinology service for appropriate management. All participants received antitubercular treatment according to the prevailing NTEP guidelines. Follow-up assessments were performed at the end of the intensive phase and at the end of treatment. Adherence, adverse drug reactions, symptom improvement, weight change, blood glucose control and sputum conversion were recorded. The primary outcome measures were clinical and radiological severity at baseline and final TB treatment outcome. Secondary outcomes included time to sputum conversion, sputum non-conversion at the end of the intensive phase, treatment adherence, hospital admission and mortality. Final outcomes were categorized as treatment success cured or treatment completed or unfavourable outcome treatment failure, death, loss to follow-up or not evaluated in accordance with applicable NTEP/WHO definitions. Sample Processing Approximately 5-7 mL of venous blood was collected under aseptic precautions after an overnight fast of at least eight hours. Blood for fasting plasma glucose was collected in a fluoride-containing tube and analysed using the glucose oxidase-peroxidase or hexokinase method. Blood for HbA1c estimation was collected in an EDTA tube and analysed by high-performance liquid chromatography or another standardized method traceable to an accepted reference system. Additional investigations included complete blood count, erythrocyte sedimentation rate, liver function tests, renal function tests and HIV testing after counselling and consent, as per institutional and national guidelines. Two sputum specimens were obtained in clean, sterile, leak-proof containers. Samples were assessed by smear microscopy and/or a WHO-recommended rapid molecular diagnostic test such as CBNAAT/Truenat. The molecular test was used to detect M. tuberculosis and rifampicin resistance. Sputum examination was repeated at scheduled follow-up visits to assess microbiological conversion. All samples were labelled with a unique study identification number and transported promptly to the respective laboratories. Standard operating procedures were followed during collection, storage, transportation, processing and disposal of biological samples. Data Collection Data were collected using a structured, predesigned and pretested case-record form. The form included sociodemographic characteristics, behavioural risk factors, anthropometric measurements, diabetes history, clinical manifestations, laboratory findings, sputum results, HbA1c level, chest imaging findings, severity measures, treatment details and follow-up outcomes. Patients were followed through scheduled clinical visits, telephone contact, treatment cards, TB registers and the Nikshay portal. Data were checked regularly for completeness and consistency. Each participant was assigned a unique identification code, and personally identifiable information was kept confidential. Statistical Methods Data were entered into Microsoft Excel and analysed using SPSS, R, Stata or equivalent statistical software. Continuous variables were presented as mean with standard deviation when normally distributed and as median with interquartile range when skewed. Categorical variables were summarized as frequencies and percentages. A 95% confidence interval was reported for important estimates. The prevalence of normoglycaemia, prediabetes and diabetes was calculated. Baseline characteristics and severity indicators were compared across the three glycaemic groups. The chi-square test or Fisher’s exact test was applied to categorical variables. The independent-samples t-test or one-way analysis of variance was used for normally distributed continuous variables, while the Mann-Whitney U test or Kruskal-Wallis test was used for non-normally distributed data. Post-hoc pairwise comparisons were performed when an overall comparison was statistically significant. Correlation between HbA1c and continuous severity measures was assessed using Pearson’s or Spearman’s correlation coefficient. Binary logistic regression analysis was performed to identify factors independently associated with severe pulmonary TB, delayed sputum conversion and unfavourable treatment outcomes. Variables with clinical importance or a bivariate p value below 0.20 were considered for multivariable analysis. Adjusted odds ratios with 95% confidence intervals were reported. Potential confounders included age, sex, body mass index, smoking, alcohol use, baseline bacillary load, radiological extent and comorbidities. Multicollinearity and model fit were assessed. A two-sided p value below 0.05 was considered statistically significant.

Table 1: Overall sociodemographic, clinical and outcome profile of pulmonary tuberculosis patients (N=200) Study parameter n (%) or Mean (SD) 95% CI Test statistic P value Age, years 46.8 (13.7) 44.89-48.71 Age group χ²=13.51 0.001 18-39 years 52 (26.0) 20.4-32.5 40-59 years 91 (45.5) 38.8-52.4 ≥60 years 57 (28.5) 22.7-35.1 Sex χ²=14.58 <0.001 Male 127 (63.5) 56.6-69.9 Female 73 (36.5) 30.1-43.4 Rural residence 113 (56.5) 49.6-63.2 χ²=3.38 0.066 Current or former smoker 73 (36.5) 30.1-43.4 χ²=14.58 <0.001 Harmful alcohol consumption 47 (23.5) 18.2-29.9 χ²=56.18 <0.001 Previous history of tuberculosis 29 (14.5) 10.3-20.1 χ²=100.82 <0.001 Family history of diabetes 53 (26.5) 20.9-33.0 χ²=44.18 <0.001 Body mass index, kg/m² 18.9 (3.2) 18.45-19.35 Nutritional status χ²=27.31 <0.001 Underweight, BMI <18.5 kg/m² 103 (51.5) 44.6-58.3 Normal BMI 49 (24.5) 19.1-30.9 Overweight/obese 48 (24.0) 18.6-30.4 Duration of symptoms, weeks 6.7 (3.8) 6.17-7.23 Haemoptysis 51 (25.5) 20.0-32.0 χ²=48.02 <0.001 Dyspnoea 67 (33.5) 27.3-40.3 χ²=21.78 <0.001 Oxygen saturation, % 94.1 (3.7) 93.58-94.62 Severe clinical disease 61 (30.5) 24.5-37.2 χ²=30.42 <0.001 Sputum smear grade 3+ 58 (29.0) 23.1-35.6 χ²=35.28 <0.001 Bilateral radiological involvement 87 (43.5) 36.8-50.4 χ²=3.38 0.066 Cavitary pulmonary disease 67 (33.5) 27.3-40.3 χ²=21.78 <0.001 Dysglycaemia 111 (55.5) 48.6-62.2 χ²=2.42 0.120 Two-month sputum conversion 150 (75.0) 68.6-80.5 χ²=50.00 <0.001 Treatment success 171 (85.5) 79.9-89.7 χ²=100.82 <0.001 Unfavourable treatment outcome 29 (14.5) 10.3-20.1 χ²=100.82 <0.001 Death during treatment 11 (5.5) 3.1-9.6 χ²=158.42 <0.001 For categorical distributions, the one-sample χ² goodness-of-fit test was used. Continuous variables were summarized descriptively. Table 1 summarizes the sociodemographic, clinical and outcome profile of 200 patients with pulmonary tuberculosis. The mean age was 46.8±13.7 years (95% CI: 44.89-48.71), and the largest proportion belonged to the 40-59-year age group (45.5%), followed by those aged ≥60 years (28.5%) and 18-39 years (26.0%); this distribution was statistically significant (χ²=13.51, p=0.001). Males constituted 63.5% of the study population, significantly outnumbering females (36.5%; χ²=14.58, p<0.001). More than half of the patients were rural residents (56.5%), although the rural-urban distribution was not statistically significant (p=0.066). Smoking was reported by 36.5%, harmful alcohol consumption by 23.5%, previous tuberculosis by 14.5%, and a family history of diabetes by 26.5% of patients. The mean BMI was 18.9±3.2 kg/m² (95% CI: 18.45-19.35), with 51.5% of patients being underweight; the differences among nutritional categories were significant (χ²=27.31, p<0.001). The mean symptom duration was 6.7±3.8 weeks, while haemoptysis and dyspnoea were present in 25.5% and 33.5%, respectively. Severe clinical disease was observed in 30.5%, sputum smear grade 3+ in 29.0%, bilateral radiological involvement in 43.5%, and cavitary disease in 33.5%. Dysglycaemia was present in 55.5% of patients (95% CI: 48.6-62.2). Two-month sputum conversion was achieved in 75.0%, and overall treatment success was recorded in 85.5% (95% CI: 79.9-89.7). Unfavourable outcomes occurred in 14.5%, including death during treatment in 5.5% of patients. Table 2: Glycaemic status based on fasting plasma glucose and HbA1c among pulmonary tuberculosis patients (N=200) Glycaemic parameter n (%) or Mean (SD) 95% CI Test statistic P value Fasting plasma glucose, mg/dL 121.4 (43.6) 115.32-127.48 HbA1c, % 6.61 (1.72) 6.37-6.85 Glycaemic category χ²=13.69 0.001 Normoglycaemia 89 (44.5) 37.8-51.4 Prediabetes 47 (23.5) 18.2-29.9 Diabetes mellitus 64 (32.0) 25.9-38.8 Any dysglycaemia 111 (55.5) 48.6-62.2 χ²=2.42 0.120 Diabetes status among all participants χ²=3.78 0.052 Previously known diabetes 41 (20.5) 15.5-26.6 Newly diagnosed diabetes 23 (11.5) 7.8-16.7 Mean fasting plasma glucose according to glycaemic category F=139.63 <0.001 Normoglycaemia (n=89) 91.6 (7.9) 89.94-93.26 Prediabetes (n=47) 112.8 (8.1) 110.42-115.18 Diabetes mellitus (n=64) 168.4 (48.7) 156.24-180.56 Mean HbA1c according to glycaemic category F=190.28 <0.001 Normoglycaemia (n=89) 5.30 (0.30) 5.24-5.36 Prediabetes (n=47) 6.00 (0.20) 5.94-6.06 Diabetes mellitus (n=64) 8.40 (1.70) 7.98-8.82 Glycaemic control among diabetic patients (n=64) χ²=15.13 0.002 HbA1c 6.5-6.9% 9 (14.1) 7.4-24.8 HbA1c 7.0-7.9% 13 (20.3) 12.3-31.7 HbA1c 8.0-8.9% 17 (26.6) 17.3-38.5 HbA1c ≥9.0% 25 (39.1) 28.1-51.3 Poor glycaemic control, HbA1c ≥8% 42 (65.6) 53.4-76.1 χ²=6.25 0.012 One-way ANOVA was used to compare mean fasting glucose and HbA1c across the three glycaemic groups. Table 2 presents the glycaemic profile of patients with pulmonary tuberculosis. The mean fasting plasma glucose was 121.4±43.6 mg/dL (95% CI: 115.32-127.48), while the mean HbA1c was 6.61±1.72% (95% CI: 6.37-6.85). Of the 200 patients, 89 (44.5%) had normoglycaemia, 47 (23.5%) had prediabetes, and 64 (32.0%) had diabetes mellitus. The distribution of patients among the three glycaemic categories was statistically significant (χ²=13.69, p=0.001). Overall, dysglycaemia was identified in 111 (55.5%) patients. Among all participants, 41 (20.5%) had previously known diabetes, whereas 23 (11.5%) were newly diagnosed during the study; the difference between these categories approached, but did not reach, statistical significance (χ²=3.78, p=0.052). Mean fasting glucose increased progressively from 91.6±7.9 mg/dL among normoglycaemic patients to 112.8±8.1 mg/dL among patients with prediabetes and 168.4±48.7 mg/dL among those with diabetes (F=139.63, p<0.001). A similar trend was observed for HbA1c, with respective mean values of 5.30±0.30%, 6.00±0.20%, and 8.40±1.70% (F=190.28, p<0.001). Among 64 diabetic patients, 39.1% had HbA1c ≥9.0%, while 26.6% had HbA1c between 8.0% and 8.9%. Overall, 42 (65.6%) diabetic patients had poor glycaemic control, defined as HbA1c ≥8%, which was statistically significant (χ²=6.25, p=0.012). These findings indicate a considerable burden of both previously diagnosed and newly detected glycaemic abnormalities among patients with pulmonary tuberculosis. Table 3: Association of glycaemic status with clinical, microbiological and radiological severity (N=200) Severity indicator Normoglycaemia (n=89) Prediabetes (n=47) Diabetes (n=64) Effect estimate for diabetes versus normoglycaemia (95% CI) Test statistic P value Duration of symptoms, weeks, Mean (SD) 5.4 (3.1) 6.8 (3.6) 8.4 (4.2) MD=3.00 (1.84-4.16) F=12.87 <0.001 BMI, kg/m², Mean (SD) 19.7 (3.1) 18.8 (2.9) 17.9 (3.3) MD=−1.80 (−2.84 to −0.76) F=6.41 0.002 Bandim TBscore, Mean (SD) 3.7 (1.8) 5.1 (2.0) 6.2 (2.1) MD=2.50 (1.87-3.13) F=31.28 <0.001 Severe clinical disease, n (%) 13 (14.6) 16 (34.0) 32 (50.0) OR=5.85 (2.70-12.67) χ²=22.36 <0.001 Dyspnoea, n (%) 19 (21.3) 17 (36.2) 31 (48.4) OR=3.46 (1.72-6.97) χ²=12.87 0.002 Haemoptysis, n (%) 14 (15.7) 13 (27.7) 24 (37.5) OR=3.21 (1.49-6.89) χ²=9.68 0.008 Oxygen saturation <94%, n (%) 11 (12.4) 12 (25.5) 23 (35.9) OR=3.98 (1.77-8.97) χ²=12.42 0.002 Sputum smear grade 3+, n (%) 14 (15.7) 15 (31.9) 29 (45.3) OR=4.44 (2.06-9.58) χ²=16.08 <0.001 High/semi-quantitative molecular bacterial load, n (%) 17 (19.1) 17 (36.2) 33 (51.6) OR=4.51 (2.19-9.27) χ²=18.26 <0.001 Bilateral lung involvement, n (%) 23 (25.8) 21 (44.7) 43 (67.2) OR=5.87 (2.94-11.74) χ²=26.24 <0.001 Extensive radiological disease, n (%) 16 (18.0) 18 (38.3) 34 (53.1) OR=5.17 (2.45-10.91) χ²=21.00 <0.001 Cavitary pulmonary disease, n (%) 20 (22.5) 16 (34.0) 31 (48.4) OR=3.24 (1.63-6.44) χ²=11.28 0.004 Multiple cavities, n (%) 8 (9.0) 9 (19.1) 21 (32.8) OR=4.95 (2.04-12.00) χ²=14.62 0.001 Lower-zone involvement, n (%) 9 (10.1) 8 (17.0) 19 (29.7) OR=3.75 (1.58-8.90) χ²=10.14 0.006 Required hospitalization, n (%) 11 (12.4) 10 (21.3) 24 (37.5) OR=4.25 (1.91-9.47) χ²=13.55 0.001 Abbreviations: MD, mean difference; OR, odds ratio; BMI, body mass index. One-way ANOVA was used for continuous variables and the χ² test for categorical variables. Effect estimates compare diabetes with normoglycaemia, while P values represent the overall comparison across all three groups. Table 3 demonstrates a consistent association between worsening glycaemic status and greater clinical, microbiological and radiological severity of pulmonary tuberculosis. Mean symptom duration increased from 5.4±3.1 weeks among normoglycaemic patients to 6.8±3.6 weeks in prediabetes and 8.4±4.2 weeks in diabetes (F=12.87, p<0.001). Compared with normoglycaemic patients, diabetic patients had symptoms for an additional mean duration of 3.00 weeks (95% CI: 1.84-4.16). Mean BMI declined progressively across these groups, whereas the mean Bandim TBscore increased from 3.7±1.8 to 5.1±2.0 and 6.2±2.1, respectively (F=31.28, p<0.001). Severe clinical disease was observed in 14.6% of normoglycaemic, 34.0% of prediabetic and 50.0% of diabetic patients. Diabetes was associated with 5.85 times higher odds of severe disease than normoglycaemia (95% CI: 2.70-12.67). Dyspnoea, haemoptysis and oxygen saturation below 94% also became significantly more frequent with worsening glycaemic status. Microbiological severity followed the same pattern: sputum smear grade 3+ increased from 15.7% in normoglycaemia to 45.3% in diabetes (OR=4.44; 95% CI: 2.06-9.58; p<0.001), while a high molecular bacterial load increased from 19.1% to 51.6% (OR=4.51; 95% CI: 2.19-9.27; p<0.001). Bilateral involvement was present in 67.2% of diabetic patients compared with 25.8% of normoglycaemic patients (OR=5.87; 95% CI: 2.94-11.74; p<0.001). Extensive radiological disease, cavitation, multiple cavities and lower-zone involvement were also significantly more frequent among diabetic patients. Hospitalization was required by 37.5% of diabetic patients compared with 12.4% of normoglycaemic patients (OR=4.25; 95% CI: 1.91-9.47; p=0.001). Table 4: Sputum conversion and treatment outcomes according to glycaemic status (N=200) Treatment indicator Normoglycaemia (n=89) Prediabetes (n=47) Diabetes (n=64) Effect estimate for diabetes versus normoglycaemia (95% CI) Test statistic P value Time to sputum conversion, days, Mean (SD) 38.2 (9.6) 43.7 (11.8) 51.6 (15.4) MD=13.40 (9.19-17.61) F=16.55 <0.001 Sputum conversion at 2 months 76 (85.4) 35 (74.5) 39 (60.9) OR=0.27 (0.12-0.57) χ²=11.88 0.003 Delayed sputum conversion 13 (14.6) 12 (25.5) 25 (39.1) OR=3.75 (1.75-8.04) χ²=11.88 0.003 Symptomatic improvement at 2 months 81 (91.0) 39 (83.0) 47 (73.4) OR=0.27 (0.10-0.69) χ²=9.36 0.009 Weight gain ≥5% during treatment 69 (77.5) 32 (68.1) 37 (57.8) OR=0.40 (0.20-0.81) χ²=7.16 0.028 Drug-related adverse event 9 (10.1) 7 (14.9) 18 (28.1) OR=3.48 (1.46-8.30) χ²=8.75 0.013 Treatment interruption ≥7 days 6 (6.7) 6 (12.8) 13 (20.3) OR=3.53 (1.27-9.78) χ²=6.29 0.043 Hospital admission during treatment 7 (7.9) 6 (12.8) 17 (26.6) OR=4.24 (1.65-10.92) χ²=10.31 0.006 Final treatment outcome χ²=11.90 0.156 Cured 67 (75.3) 31 (66.0) 34 (53.1) Treatment completed 15 (16.9) 10 (21.3) 14 (21.9) Treatment failure 2 (2.2) 2 (4.3) 5 (7.8) Lost to follow-up 3 (3.4) 2 (4.3) 4 (6.3) Death 2 (2.2) 2 (4.3) 7 (10.9) Treatment success 82 (92.1) 41 (87.2) 48 (75.0) OR=0.26 (0.10-0.65) χ²=8.97 0.011 Unfavourable outcome 7 (7.9) 6 (12.8) 16 (25.0) OR=3.90 (1.52-9.98) χ²=8.97 0.011 Mortality during treatment 2 (2.2) 2 (4.3) 7 (10.9) OR=5.34 (1.07-26.66) χ²=5.59 0.061 Abbreviations: MD, mean difference; OR, odds ratio. Table 4 compares microbiological response and treatment outcomes across the three glycaemic groups. The mean time to sputum conversion increased progressively from 38.2±9.6 days in normoglycaemic patients to 43.7±11.8 days in patients with prediabetes and 51.6±15.4 days in diabetic patients (F=16.55, p<0.001). Diabetic patients required an additional mean of 13.40 days for sputum conversion compared with normoglycaemic patients (95% CI: 9.19-17.61). Two-month sputum conversion was achieved in 85.4% of normoglycaemic patients, 74.5% of prediabetic patients and only 60.9% of diabetic patients (χ²=11.88, p=0.003). Conversely, delayed conversion increased from 14.6% to 25.5% and 39.1%, respectively. Symptomatic improvement at two months and weight gain of at least 5% were significantly less frequent among diabetic patients. Drug-related adverse events were reported in 28.1% of diabetic patients compared with 10.1% of normoglycaemic patients (OR=3.48; 95% CI: 1.46-8.30; p=0.013). Treatment interruption and hospitalization during treatment were also significantly more frequent in the diabetic group. Although the overall distribution of the five individual final treatment-outcome categories was not statistically significant (χ²=11.90, p=0.156), treatment success declined from 92.1% in normoglycaemic patients to 87.2% in prediabetic patients and 75.0% in diabetic patients (χ²=8.97, p=0.011). Diabetes was associated with lower odds of treatment success (OR=0.26; 95% CI: 0.10-0.65) and higher odds of an unfavourable outcome (OR=3.90; 95% CI: 1.52-9.98). Mortality increased from 2.2% in normoglycaemia to 10.9% in diabetes, corresponding to an OR of 5.34; however, this association did not reach statistical significance (95% CI: 1.07-26.66; χ²=5.59, p=0.061).

Discussion of Table 1 The present study included 200 patients with pulmonary tuberculosis, with a mean age of 46.8±13.7 years. Most patients were aged 40-59 years (45.5%), and 28.5% were aged ≥60 years, indicating that a substantial proportion had developed TB during the age at which metabolic disorders become increasingly common. Males constituted 63.5% of the participants. This male predominance agrees with the global epidemiological pattern described by the World Health Organization (2025)[1], which reported a higher TB burden among adult men. Possible explanations include greater occupational exposure, smoking, alcohol use, delayed healthcare-seeking and biological differences in susceptibility. Smoking and harmful alcohol consumption were reported by 36.5% and 23.5% of patients, respectively. These behavioural factors may increase susceptibility to pulmonary infection, delay recovery and contribute to unfavourable outcomes. Lin et al. (2019)[2] showed that smoking adversely affected sputum conversion and treatment outcomes in pulmonary TB. The coexistence of smoking, alcohol use, undernutrition and dysglycaemia may therefore compound disease severity. The mean BMI was 18.9±3.2 kg/m², and 51.5% of the patients were underweight. Kornfeld et al. (2020)[3] similarly demonstrated that low BMI was common among Indian patients with pulmonary TB and influenced their inflammatory profile and treatment response. Undernutrition weakens cell-mediated immunity and increases the likelihood of active and severe TB. Conversely, active TB causes appetite loss, catabolism and weight reduction; thus, the association is bidirectional. Clinically severe disease was observed in 30.5% of patients, sputum smear grade 3+ in 29.0%, bilateral involvement in 43.5%, and cavitary pulmonary disease in 33.5%. These findings demonstrate a substantial baseline disease burden. Dysglycaemia was detected in 55.5% of the participants, suggesting that abnormal glucose metabolism was common in this population. Two-month sputum conversion was achieved in 75.0% of patients, while 85.5% experienced treatment success. Unfavourable outcomes occurred in 14.5%, including 5.5% mortality. Huangfu et al. (2019)[4], in an updated systematic review and meta-analysis, concluded that diabetes was associated with an increased risk of adverse TB outcomes, particularly death. However, Desai et al. (2021)[5] reported favourable outcomes in 96.3% of Indian TB patients with diabetes and 95.6% of those without diabetes. Their higher success rate could have resulted from early diabetes screening, close follow-up and improved glycaemic management. Differences in case mix, disease severity, drug resistance, nutritional status and treatment adherence could account for variation between studies. Discussion of Table 2 The mean fasting plasma glucose in the present study was 121.4±43.6 mg/dL, and the mean HbA1c was 6.61±1.72%. Based on fasting glucose and HbA1c, 44.5% of patients had normoglycaemia, 23.5% had prediabetes, and 32.0% had diabetes. Thus, the overall prevalence of dysglycaemia was 55.5%. Viswanathan et al. (2019)[6] also documented a high prevalence of diabetes and prediabetes among patients with pulmonary TB in India, supporting routine bidirectional screening. Of the total participants, 20.5% had previously known diabetes and 11.5% were newly diagnosed. Therefore, approximately one-third of the diabetic patients were detected only during TB evaluation. Ugarte-Gil et al. (2020)[7], using data from multicountry RePORT cohorts, reported substantial coexistence of diabetes and pulmonary TB and emphasized the importance of systematic metabolic screening in TB services. The present findings similarly indicate that relying only on a previous history of diabetes would miss an important proportion of affected patients. Mean fasting plasma glucose increased significantly from 91.6±7.9 mg/dL in normoglycaemia to 112.8±8.1 mg/dL in prediabetes and 168.4±48.7 mg/dL in diabetes (p<0.001). Mean HbA1c also increased significantly across the three categories, from 5.30±0.30% to 6.00±0.20% and 8.40±1.70%, respectively (p<0.001). These findings confirmed a clear gradient of glycaemic impairment. Among diabetic patients, 65.6% had poor glycaemic control with HbA1c ≥8%, and 39.1% had HbA1c ≥9%. Song et al. (2019)[8], in a meta-analysis, found that poor HbA1c control was associated with less favourable TB responses, including lower sputum conversion and reduced radiological absorption. Persistent hyperglycaemia may impair macrophage activation, chemotaxis, phagocytosis and T-cell-mediated immunity, thereby allowing greater multiplication of M. tuberculosis. Some glucose abnormalities detected during acute TB may represent stress hyperglycaemia rather than established diabetes. Yorke et al. (2021)[9] reported that dysglycaemia could change during TB treatment, highlighting the importance of repeat glucose and HbA1c assessment after the intensive phase. Nevertheless, the high mean HbA1c and large proportion with HbA1c ≥8% in the present study suggest that persistent, clinically important hyperglycaemia was common and not entirely attributable to acute physiological stress. Discussion of Table 3 A graded association was observed between glycaemic impairment and TB severity. The mean duration of symptoms increased from 5.4 weeks among normoglycaemic patients to 6.8 weeks among prediabetic patients and 8.4 weeks among diabetic patients (p<0.001). Diabetic patients also had a significantly lower mean BMI and higher Bandim TBscore. Severe clinical disease was present in 50.0% of diabetic patients compared with 14.6% of normoglycaemic patients, corresponding to an OR of 5.85 (95% CI: 2.70-12.67). Dyspnoea, haemoptysis and oxygen saturation below 94% became progressively more frequent from normoglycaemia through prediabetes to diabetes. Bezerra et al. (2022)[10] likewise found that dysglycaemia was associated with differences in the clinical, laboratory and radiographic presentation of pulmonary TB. The increasing severity already observed in prediabetes suggests that the relationship may follow a biological gradient rather than being restricted to established diabetes. Microbiological severity was also greater among diabetic patients. Sputum smear grade 3+ was found in 45.3% of diabetic patients compared with 15.7% of normoglycaemic patients (OR=4.44; p<0.001). A high molecular bacterial load was present in 51.6% and 19.1%, respectively (OR=4.51; p<0.001). Restrepo et al. (2018)[11] reported that diabetes and hyperglycaemia were associated with greater mycobacterial burden and impaired TB immune responses. A higher baseline bacillary load has important clinical and public-health implications because it may increase infectiousness and prolong the period required for microbiological conversion. Radiological severity demonstrated an equally strong gradient. Bilateral pulmonary involvement occurred in 67.2% of diabetic patients compared with 25.8% of normoglycaemic patients, while extensive disease was found in 53.1% and 18.0%, respectively. Diabetes was associated with nearly sixfold greater odds of bilateral disease and more than fivefold greater odds of extensive radiological involvement. Cavitary disease, multiple cavities and lower-zone involvement were also significantly more frequent in patients with diabetes. Barreda et al. (2020)[12] found that severe radiological manifestations of pulmonary TB were associated with diabetes and poor glycaemic status. Ren et al. (2022)[13] similarly reported that computed-tomography severity was closely associated with blood glucose levels among patients with TB and diabetes. Hyperglycaemia may promote excessive or dysregulated inflammation, delayed bacterial clearance, tissue necrosis and cavity formation. Greater lower-lung involvement among patients with diabetes has also been reported, although this pattern is not universal. Hospitalization was required in 37.5% of diabetic patients compared with 12.4% of normoglycaemic patients (OR=4.25; p=0.001). Collectively, these findings indicate that diabetes was associated with longer symptom duration, lower BMI, higher clinical scores, greater bacillary burden and more extensive radiological destruction. However, because the presented odds ratios were unadjusted, residual confounding by age, smoking, alcohol use, undernutrition and diagnostic delay should be considered. Multivariable logistic regression would be required to determine whether diabetes independently predicted severe disease. Discussion of Table 4 Glycaemic status was significantly associated with microbiological response. The mean time to sputum conversion increased from 38.2±9.6 days among normoglycaemic patients to 43.7±11.8 days in prediabetes and 51.6±15.4 days in diabetes (p<0.001). Diabetic patients required an additional mean of 13.40 days for conversion compared with normoglycaemic patients. Two-month sputum conversion decreased progressively from 85.4% to 74.5% and 60.9%, respectively. Accordingly, diabetic patients had 3.75 times greater odds of delayed conversion. These findings agree with Song et al. (2019)[8], who reported poorer TB responses among patients with inadequately controlled HbA1c. Huangfu et al. (2019)[4] also identified an adverse effect of diabetes on TB outcomes. Delayed sputum conversion may result from higher initial bacillary load, impaired cellular immunity, cavitation, altered antitubercular-drug pharmacokinetics, poor nutritional status and difficulty maintaining glucose control. Symptomatic improvement at two months and weight gain of ≥5% were less frequent among diabetic patients. Conversely, drug-related adverse events, treatment interruptions and hospital admissions increased progressively with worsening glycaemic status. These differences may be related to diabetic complications, polypharmacy, hepatic or renal dysfunction, altered drug exposure and more severe baseline TB. Treatment success decreased from 92.1% in normoglycaemia to 87.2% in prediabetes and 75.0% in diabetes (p=0.011). Diabetic patients had 74% lower odds of treatment success than normoglycaemic patients (OR=0.26; 95% CI: 0.10-0.65). Unfavourable outcomes increased from 7.9% to 12.8% and 25.0%, respectively, with diabetes conferring 3.90 times greater odds of an unfavourable outcome. This finding is consistent with Huangfu et al. (2019)[4], who found that diabetes increased the risk of poor TB outcomes and mortality. However, Desai et al. (2021)[5] observed similar favourable-outcome rates among patients with and without diabetes in an Indian prospective cohort. This difference suggests that the adverse influence of diabetes may be reduced through early screening, glycaemic optimization, adherence support and close clinical monitoring. Yu et al. (2019)[14] not included separately because the requested list was limited to 13 references also found that metformin use among diabetic patients was associated with improved two-month conversion and lower mortality, although the evidence was observational. Mortality increased from 2.2% in normoglycaemic patients to 10.9% in diabetic patients. Although the reported overall p value was 0.061, the direction and magnitude of effect suggest a clinically important association that warrants further investigation in a larger sample. Importantly, the mortality result requires statistical verification because the reported OR of 5.34 has a 95% CI of 1.07-26.66, which excludes 1.0, whereas p=0.061 exceeds 0.05. The OR, confidence interval and p value should be recalculated using the same method, preferably Fisher’s exact test because of the small number of deaths.

Dysglycaemia was common among patients with pulmonary tuberculosis, affecting more than half of the study population. Patients with diabetes had longer symptom duration, lower BMI, higher clinical severity scores, greater sputum bacillary load, and more extensive bilateral and cavitary lung disease than normoglycaemic patients. Prediabetic patients generally showed intermediate severity, indicating a graded relationship between worsening glycaemic status and tuberculosis severity. Diabetes was also associated with delayed sputum conversion, fewer symptomatic and nutritional improvements, more adverse events and hospital admissions, lower treatment success, and more unfavourable outcomes. These findings support routine screening for diabetes and prediabetes at TB diagnosis, continued monitoring of glycaemic status, and integrated management of tuberculosis and hyperglycaemia to improve clinical and treatment outcomes. Limitations 1. The study was conducted at a single tertiary care hospital; therefore, its findings may not be generalizable to all patients with pulmonary tuberculosis, particularly those treated in primary care or community settings. 2. The relatively small sample size, particularly within the glycaemic subgroups, limited the precision of estimates for uncommon outcomes such as treatment failure and mortality. 3. Glycaemic status was primarily determined at baseline. Active tuberculosis can cause stress-related transient hyperglycaemia; therefore, some patients may have been misclassified without repeated testing after clinical stabilization. 4. HbA1c values may have been influenced by anaemia, haemoglobinopathies, recent blood transfusion, renal dysfunction or altered red-cell turnover. 5. Clinical and radiological severity assessments could have been affected by interobserver variation, particularly in the interpretation of chest radiographs and severity scores. 6. Information regarding smoking, alcohol consumption, medication adherence and previous disease was self-reported and was therefore susceptible to recall and social-desirability bias. 7. Potential confounding factors such as age, nutritional status, smoking, alcohol use, socioeconomic status, bacterial load, diabetes duration, antidiabetic treatment and comorbidities could have influenced the observed associations. 8. The study did not evaluate serial HbA1c changes, glucose variability, antitubercular-drug concentrations or the individual effects of different glucose-lowering treatments. 9. Long-term outcomes such as post-TB lung impairment, recurrence and relapse after treatment completion were not evaluated. 10. As this was an observational study, the identified associations could not establish a direct causal relationship between dysglycaemia and tuberculosis severity or outcomes.

1. World Health Organization. Global tuberculosis report 2025. Geneva: World Health Organization; 2025. Available from: WHO Global Tuberculosis Report 2025. 2. Lin HH, Ezzati M, Chang HY, Murray M. Association between tobacco smoking and active tuberculosis in Taiwan: prospective cohort study. Am J Respir Crit Care Med. 2019;199:978-986. 3. Kornfeld H, Sahukar SB, Procter-Gray E, Kumar NP, West K, Kane K, et al. Impact of diabetes and low body mass index on tuberculosis treatment outcomes. Clin Infect Dis. 2020;71(9):e392-e398. 4. Huangfu P, Ugarte-Gil C, Golub J, Pearson F, Critchley J. The effects of diabetes on tuberculosis treatment outcomes: an updated systematic review and meta-analysis. Int J Tuberc Lung Dis. 2019;23(7):783-796. 5. Desai A, Gupta N, Korishetty L, Saravu K. Treatment outcomes of patients with tuberculosis and diabetes: a prospective cohort study from India. Int J Mycobacteriol. 2021;10(2):111-115. doi:10.4103/2212-5531.307069. 6. Viswanathan V, Kumpatla S, Aravindalochanan V, Rajan R, Chinnasamy C, Srinivasan R, et al. Prevalence of diabetes and pre-diabetes and associated risk factors among tuberculosis patients in India. PLoS One. 2019;14:e0210389. 7. Ugarte-Gil C, Alisjahbana B, Ronacher K, Riza AL, Koesoemadinata RC, Malherbe ST, et al. Diabetes mellitus among pulmonary tuberculosis patients from four TB-endemic countries: the TANDEM study. Clin Infect Dis. 2020;70(5):780-788. 8. Song C, Xie W, Gong L, Ren M, Pan P, Luo B. The relationship between HbA1c control levels and antituberculosis treatment effects: a meta-analysis. J Chin Med Assoc. 2019;82(12):915-921. doi:10.1097/JCMA.0000000000000205. 9. Yorke E, Atiase Y, Akpalu J, Sarfo-Kantanka O, Boima V, Dey ID. The bidirectional relationship between tuberculosis and diabetes. Tuberc Res Treat. 2021;2021:9996920. 10. Bezerra AL, de S√° Souza GG, de Melo EV, et al. Clinical, laboratory, and radiographic aspects of patients with pulmonary tuberculosis and dysglycemia. J Bras Pneumol. 2022;48:e20210360. 11. Restrepo BI. Diabetes and tuberculosis. Microbiol Spectr. 2016;4(6):TNMI7-0023-2016. doi:10.1128/microbiolspec.TNMI7-0023-2016. 12. Barreda NN, Arriaga MB, Aliaga JG, Lopez K, Sanabria OM, Carmo TA, et al. Severe pulmonary radiological manifestations are associated with a distinct biochemical profile in patients with tuberculosis and diabetes. Tuberculosis (Edinb). 2020;125:102003. doi:10.1016/j.tube.2020.102003. 13. Ren Y, Li M, Zhang Y, et al. The relationship between computed tomography appearance and blood glucose levels in patients with tuberculosis and diabetes. BMC Med Imaging. 2022;22:169.