Association between Tuberculosis, Statin Use, and Diabetes [TBN 100]

A nationwide retrospective cohort study examined whether statin use was associated with a lower risk of tuberculosis (TB), including whether the association differed according to diabetes status. The investigators conducted a propensity score-matched analysis using South Korea’s National Health Insurance Service databases from 2003 through 2013. These databases represent people of all ages and regions across South Korea and contain sociodemographic information, diagnoses, deaths, prescriptions, and insured medical service use.

Among 1,025,340 people in the National Health Insurance cohort, the researchers identified 125,841 new statin users with a first-ever statin prescription between 2003 and 2013. After excluding 189 people younger than 18 years and 2,184 who died or emigrated within 7 days of cohort entry, 123,468 statin users remained. Potential controls were drawn from 569,999 participants in the medical-checkup cohort. After excluding previous statin users, people younger than 18 years, and those who died or emigrated within 7 days, 439,546 never-users were eligible. Propensity scores were estimated at statin initiation for users and at the first medical checkup in the corresponding year for nonusers. Matching produced 28,018 statin users and 28,018 nonusers, with standardized differences below 10% for all adjusted variables. Analyses considered 22 covariates, including age, sex, socioeconomic characteristics, diabetes, hypertension, dyslipidemia, renal and pulmonary disease, rheumatoid disease, malignancy, and cardiovascular disease. The definition used to identify incident TB was not specified in the supplied text.

In the unmatched population, crude TB incidence was similar between groups: 168 cases during 158,491 person-years among statin users, or 1.06 per 1,000 person-years (95% CI, 0.90 to 1.22), compared with 2,845 cases during 3,026,596 person-years among nonusers, or 0.94 per 1,000 person-years (95% CI, 0.91 to 0.98). The unadjusted hazard ratio (HR) was 1.05 (95% CI, 0.89 to 1.23). After adjustment for 22 covariates, statin use was associated with a lower TB risk (HR, 0.60; 95% CI, 0.49 to 0.74). In the propensity score-matched analysis, 30 TB cases occurred among statin users during 30,303 person-years, compared with 235 cases among nonusers during 167,857 person-years. Incidence was 0.99 versus 1.40 per 1,000 person-years, corresponding to an HR of 0.67 (95% CI, 0.46 to 0.98). The association weakened after statin discontinuation. In the matched analysis, the HR was 0.73 within 1 month after discontinuation (95% CI, 0.51 to 1.05) and 1.00 within 6 months (95% CI, 0.72 to 1.38). Diabetes modified the association in the matched analysis. Among participants without diabetes, statin use was associated with substantially lower TB risk, with 7 versus 147 cases and an HR of 0.28 (95% CI, 0.13 to 0.60). Among those with diabetes, there was no evidence of reduced risk, with 23 versus 88 cases and an HR of 1.05 (95% CI, 0.66 to 1.67). The interaction was statistically significant (P for interaction = 0.003). Statin use was also associated with lower TB risk among patients with cardiovascular disease (HR, 0.51; 95% CI, 0.28 to 0.92), but not clearly among those without cardiovascular disease (HR, 0.74; 95% CI, 0.44 to 1.23), although the interaction was not significant.

In conclusion, statin use was associated with a modestly lower incidence of TB after extensive adjustment and propensity score matching, but this association was concentrated among people without diabetes and appeared to diminish after statin discontinuation. Because this was an observational claims-based study, it cannot establish that statins prevent TB. Residual confounding, differences between the insurance and medical-checkup cohorts, exposure misclassification, and the small number of TB events in several matched subgroups may limit precision and causal interpretation. 

Source: Kim MC, Yun SC, Lee SO, Choi SH, Kim YS, Woo JH, Kim SH. Association between Tuberculosis, Statin Use, and Diabetes: A Propensity Score-Matched Analysis. The American Journal of Tropical Medicine and Hygiene. 2019 Aug 1;101(2):350-356.

Statin Use Is Associated With a Lower Risk of TB [TBN 099]

A nationwide, population-based retrospective cohort study evaluated whether statin use was associated with a lower risk of incident tuberculosis (TB) disease. The study used Taiwan’s National Health Insurance Research Database from January 1, 2000, through December 31, 2013, which covers more than 99% of the Taiwanese population.

Adults aged 20 years or older with cumulative statin prescriptions for at least 30 days were included as statin users. Statins assessed included simvastatin, lovastatin, pravastatin, fluvastatin, atorvastatin, cerivastatin, rosuvastatin, and pitavastatin. Patients with TB before cohort entry, missing data, or no eligible matched control were excluded. Each new statin user was matched by age and sex to a nonuser using incidence-density sampling and the same index date. Statin exposure was quantified using cumulative defined daily doses (cDDDs), categorized as less than 180, 180 to 365, or more than 365 cDDDs. Incident TB required an ICD-9-CM code of 010 to 018 plus prescriptions for at least two anti-TB drugs for at least 4 weeks within 180 days of diagnosis. Participants were followed until TB diagnosis, death, withdrawal from insurance, or December 31, 2013. Analyses adjusted for age, sex, comorbidities, urbanization, and annual outpatient visits. This observational cohort provides moderate-level evidence for association, but not proof of causality.

The final analysis included 102,424 statin users and 202,718 nonusers, contributing 571,568 and 1,027,385 person-years of observation, respectively. Statin users had substantially more cardiometabolic and chronic diseases than nonusers, including diabetes (35.0% versus 7.1%) and coronary heart disease (18.3% versus 4.6%). After adjustment, statin use was associated with a 47% lower hazard of incident TB (adjusted hazard ratio [HR], 0.53; 95% CI, 0.47 to 0.61). The association showed a dose-response pattern: fewer than 180 cDDDs were not associated with reduced risk (HR, 1.06; 95% CI, 0.91 to 1.24), whereas 180 to 365 cDDDs were associated with lower risk (HR, 0.57; 95% CI, 0.45 to 0.72), and more than 365 cDDDs with the lowest risk (HR, 0.27; 95% CI, 0.22 to 0.33). A second exposure categorization reported in the supplied text produced similar estimates, including HRs of 0.65 for the intermediate category and 0.29 for the highest category. Independent predictors of higher TB risk included age 65 years or older (HR, 1.60; 95% CI, 1.42 to 1.80), male sex (HR, 3.68; 95% CI, 3.24 to 4.18), diabetes (HR, 1.22; 95% CI, 1.06 to 1.40), heart failure (HR, 1.51; 95% CI, 1.08 to 2.11), COPD (HR, 1.72; 95% CI, 1.42 to 2.10), asthma (HR, 1.51; 95% CI, 1.17 to 1.94), and systemic glucocorticoid use (HR, 1.47; 95% CI, 1.26 to 1.72).

In conclusion, statin use was associated with a substantially lower incidence of TB in Taiwan, particularly at higher cumulative exposure. Because the study was observational and based on administrative claims, residual confounding, exposure misclassification, healthy-user bias, and differences in underlying health status may remain. Generalizability outside Taiwan or to populations with different TB epidemiology is uncertain. Randomized or prospective studies would be needed before statins could be considered for TB prevention. 

Source: Su VY, Su WJ, Yen YF, Pan SW, Chuang PH, Feng JY, Chou KT, Yang KY, Lee YC, Chen TJ. Statin Use Is Associated With a Lower Risk of TB. Chest. 2017 Sep;152(3):598-606.

Statin treatment is associated with a decreased risk of active TB [TBN 098]

A study investigated whether statin use, which inhibits cholesterol biosynthesis, is associated with a lower risk of developing active tuberculosis (TB). This was a population-based cohort study using Taiwan's National Health Insurance Research Database (NHIRD). The cohort included all adults aged 18 years or older who were followed longitudinally from January 1999 through December 2011. Drug exposure was assessed during 1999, and participants were followed from January 1, 2000 until the first occurrence of active TB diagnosis, termination of insurance coverage, death, or the end of the study.

The study included all eligible adults in the NHIRD, with 8,098 patients who developed active TB and 809,800 controls. The mean follow-up period was 9.8 years. Statin users were defined as individuals with at least 7 days of statin prescriptions. Statins included simvastatin, lovastatin, pravastatin, fluvastatin, atorvastatin, cerivastatin, rosuvastatin, and pitavastatin. Exposure was categorized as current use (prescription within 30 days before the index date), recent use (31 to 90 days), past use (91 days to 1 year), and chronic use (more than 90 cumulative prescription days within the year of TB diagnosis). The analysis adjusted for multiple confounders, including disease risk score (DRS) adjustment, and also evaluated subgroup effects and duration-response relationships.

All categories of statin use, including current, recent, past, and chronic use, were associated with a reduced risk of active TB compared with non-users. Chronic statin use showed the strongest association, with an unadjusted relative risk (RR) of 0.74 (95% CI, 0.63 to 0.87), an adjusted RR of 0.66 (95% CI, 0.56 to 0.78), and a DRS-adjusted RR of 0.62 (95% CI, 0.53 to 0.72). Current statin use demonstrated a similar protective association. Subgroup analyses consistently showed reduced TB risk across all predefined groups, although none of the interaction tests reached statistical significance. A duration-response analysis found progressively lower TB incidence with longer statin use. When cumulative statin exposure was analyzed as a continuous variable, each additional day of statin therapy was associated with a 0.002% reduction in the risk of active TB (p < 0.010). Baseline comparisons showed that patients with TB generally had greater comorbidity, more TB risk factors, more outpatient visits, and higher cardiovascular medication use than controls.

In this large nationwide cohort from Taiwan, statin use, particularly chronic use, was associated with a significantly lower risk of developing active TB, with evidence of a duration-response relationship. As an observational cohort study, the findings support an association rather than causation. The excerpt does not specify the study's funding source, conflicts of interest, inclusion and exclusion criteria beyond age eligibility, or detailed limitations, so these cannot be assessed from the available text. The level of evidence is moderate for an observational population-based cohort study.

Source: Lai CC, Lee MT, Lee SH, Hsu WT, Chang SS, Chen SC, Lee CC. Statin treatment is associated with a decreased risk of active tuberculosis: an analysis of a nationally representative cohort. Thorax. 2016 Jul 1;71(7):646-51.

TB and increased incidence of cardiovascular disease [TBN 097]

A study investigated whether tuberculosis (TB) increases the risk of incident cardiovascular disease (CVD), while accounting for differences in CVD risk that existed before TB diagnosis. This was a retrospective matched cohort study conducted using large electronic health record databases from the United States and the United Kingdom. The study examined CVD incidence during the two years before and two years after TB diagnosis, allowing replication of findings across two independent healthcare systems.

The analysis included 17,941 people with TB (2,121 in the United States and 15,820 in the United Kingdom) and 130,494 matched individuals without TB. Most participants with TB had at least four matched controls. The median age at TB diagnosis was 59 years in the United States and 44 years in the United Kingdom. Covariates measured two years before TB diagnosis included smoking status, body mass index (BMI), comorbidities, and prescriptions for statins or antihypertensive medications. Incident CVD events were identified from primary care and hospital records using ICD-9 and ICD-10 diagnostic codes in the United States and primary care diagnostic codes together with ICD-10 codes in the United Kingdom. Incidence rates and incidence rate ratios (IRRs) were estimated, with adjustment for demographic and clinical covariates, followed by additional adjustment for baseline differences in CVD incidence before TB diagnosis.

Over the four-year observation period, 462 CVD events occurred among participants with TB in the United States and 622 in the United Kingdom, corresponding to incidence rates of 65 and 11 per 1,000 person-years, respectively. Among matched individuals without TB, CVD incidence rates were lower at 31 and 6 per 1,000 person-years. People with TB already had higher CVD incidence during the two years before diagnosis, but the greatest increase occurred during the acute period surrounding TB diagnosis. During this acute period, CVD incidence reached 127 versus 30 per 1,000 person-years in the United States and 16 versus 6 per 1,000 person-years in the United Kingdom for participants with and without TB, respectively. After adjustment for demographic and clinical factors, TB was associated with a significantly increased risk of CVD during the acute period, with adjusted IRRs of 3.5 (95% CI, 2.7 to 4.4) in the United States and 2.7 (95% CI, 2.2 to 3.3) in the United Kingdom. After further accounting for preexisting differences in CVD incidence before TB diagnosis, the association remained significant but was attenuated, with adjusted relative risks of 3.2 (95% CI, 2.2 to 4.4) in the United States and 1.6 (95% CI, 1.2 to 2.1) in the United Kingdom. Stratified analyses showed similar relative risks between men and women in the United Kingdom, with no consistent differences according to age or race/ethnicity.

Active TB was associated with a substantially increased risk of incident cardiovascular disease, particularly during the period surrounding TB diagnosis, even after accounting for preexisting cardiovascular risk. These findings were consistent across independent cohorts from the United States and the United Kingdom, supporting the robustness of the association. As a retrospective observational study, causal inference is limited, and residual confounding remains possible despite extensive adjustment. The findings suggest that cardiovascular risk assessment and monitoring may be important during and shortly after TB diagnosis.

Source: Critchley JA, Limb ES, Khakharia A, Carey IM, Auld SC, De Wilde S, Harris T, Phillips LS, Cook DG, Rhee MK, Chaudhry UA. Tuberculosis and increased incidence of cardiovascular disease: cohort study using United States and United Kingdom health records. Clinical Infectious Diseases. 2025 Feb 15;80(2):271-9.

BMI trajectories and association with TB risk in Southern Africa [TBN 096]

A study investigated the nutritional status of tuberculosis (TB)-affected households and evaluated the association between baseline body mass index (BMI), changes in BMI over time, and TB risk, while describing longitudinal BMI trajectories. This was a prospective, noninterventional observational household contact cohort study (ERASE-TB) conducted in Zimbabwe, Mozambique, and Tanzania. Recruitment began between March and September 2021, enrolling household contacts aged 10 years or older of individuals with microbiologically confirmed pulmonary TB. Participants were followed every six months for up to 24 months.

The study included 2,107 household contacts from 822 households (699 in Zimbabwe, 710 in Mozambique, and 698 in Tanzania), with a median follow-up of 23.8 months (IQR, 21.8 to 26.3). At enrollment and follow-up visits, investigators collected sociodemographic data, medical and TB history, and anthropometric measurements including height, weight, and mid-upper arm circumference (MUAC). Participants also underwent blood pressure, hemoglobin, and optional HIV testing. TB screening consisted of the World Health Organization symptom questionnaire and chest radiography, followed by Xpert MTB/RIF Ultra and mycobacterial culture when indicated. An independent endpoint review committee classified TB cases, with confirmed or likely TB used as study outcomes. Prevalent TB was defined as diagnosis at baseline, while incident TB was diagnosed more than 30 days after enrollment. Associations between BMI and TB were assessed using adjusted logistic regression, Cox proportional hazards models, restricted cubic spline analyses, Poisson regression with time-varying BMI changes, and growth mixture modeling to identify latent BMI trajectories.

Among participants, 62.2% were female, the median age was 27 years (IQR, 16 to 42), and 29.5% were adolescents. Underweight was common among adolescents (61.8%) but uncommon among adults (9.2%), whereas 36.6% of adults were overweight or obese. A household-level dual burden of malnutrition, with both underweight and overweight individuals, was observed in 14% to 19% of households across the three countries. Twenty-one participants (1.0%) had prevalent TB and 41 (1.9%) developed incident TB, corresponding to an incidence rate of 13.2 per 1,000 person-years (95% CI, 9.5 to 17.9). Baseline underweight alone was not associated with prevalent TB (adjusted odds ratio [aOR], 0.94; 95% CI, 0.28 to 2.73), but underweight combined with anemia was associated with a substantially higher odds of prevalent TB (aOR, 4.83; 95% CI, 1.03 to 16.8). Similarly, baseline underweight alone was not significantly associated with incident TB (adjusted hazard ratio [aHR], 1.06; 95% CI, 0.49 to 2.26), whereas overweight or obesity showed a nonsignificant trend toward lower risk (aHR, 0.42; 95% CI, 0.15 to 1.16). Underweight combined with anemia was associated with a markedly higher hazard of incident TB (aHR, 3.77; 95% CI, 1.50 to 9.51). Restricted cubic spline analysis demonstrated a nonlinear inverse relationship between BMI and TB risk, with sharply increasing risk below a BMI Z-score of 0. The population attributable fraction for underweight was estimated at 17.3% of incident TB. Among participants who developed incident TB and had repeated anthropometric measurements, 64.7% lost weight during follow-up, and 31.8% of these lost more than 10% of baseline BMI. Adults who developed TB generally started with higher BMI and lost weight before diagnosis, whereas adolescents were often underweight from baseline. Time-varying BMI loss of at least 10% was not significantly associated with incident TB (adjusted incidence rate ratio, 2.27; 95% CI, 0.22 to 22.9), likely reflecting limited statistical power. Growth mixture modeling identified four BMI trajectory groups: decreasing, low stable, high stable, and increasing BMI. Participants in the decreasing BMI trajectory had the highest TB incidence (5.8%) compared with the low stable (1.3%), high stable (0.8%), and increasing (0.0%) groups (P = .005). The decreasing BMI group also had the highest baseline median BMI (30.6 kg/m²).

In this East and Southern African household contact cohort, baseline underweight alone was not independently associated with prevalent or incident TB, but underweight combined with anemia identified individuals at substantially higher risk. Declining BMI over time, particularly among adults who were initially overweight or obese, was associated with subsequent TB development, highlighting the value of longitudinal nutritional monitoring. The observational design limits causal inference, and the relatively small number of TB events reduced statistical precision for some analyses. The findings support integrating repeated nutritional assessment, particularly BMI and anemia evaluation, into TB household contact follow-up programs.

Source: Larsson L, Calderwood CJ, Marambire ET, Held K, Banze D, Mfinanga A, Madziva K, Walsh P, Jacob J, Fernandez FT, Lungu P. Body mass index trajectories and association with tuberculosis risk in a cohort of household contacts in Southern Africa. Clinical Infectious Diseases. 2025 Dec 15;81(6):e600-11.

Prognostic value of CRP in adults with TB meningitis [TBN 095]

A prospective cohort study investigated whether baseline serum C-reactive protein (CRP), an inexpensive marker of systemic inflammation, predicts poor clinical outcomes in adults with tuberculous meningitis (TBM), particularly among people living with HIV. Participants were enrolled between March 2021 and November 2023 at Mulago and Kiruddu National Referral Hospitals in Kampala, Uganda. The study aimed to evaluate the prognostic value of CRP for disability or death after TBM treatment.

Adults aged 18 years or older with definite, probable, or possible TBM according to the uniform case definition were included. Microbiological confirmation was based on positive cerebrospinal fluid (CSF) Xpert MTB/RIF Ultra or mycobacterial culture. Participants presenting within 3 months of antiretroviral therapy initiation were classified as having unmasking immune reconstitution inflammatory syndrome (IRIS). All participants were enrolled through the HARVEST phase 3 randomized clinical trial, which compared high-dose rifampicin (35 mg/kg/day) with standard-dose rifampicin (10 mg/kg/day) for 8 weeks. All patients received dexamethasone starting at 0.4 mg/kg/day with tapering over 6 weeks. Baseline serum CRP was measured, typically within 24 hours of diagnostic lumbar puncture, using a Cobas c311 clinical chemistry analyzer. The primary outcome was poor neurological outcome, defined as a modified Rankin Scale (mRS) score of 4 or greater at 8 weeks.

Among 178 enrolled adults with TBM, 135 (75.8%) had both baseline CRP measurements and 8-week outcome data available for analysis. The median age was 36 years (IQR, 30 to 42), 46% were female, and 83% were living with HIV, with a median CD4 count of 79 cells/μL (IQR, 38 to 177). Altered mental status at presentation was common, occurring in 76% of participants. As a continuous predictor, baseline CRP demonstrated moderate discrimination for poor outcome, with an area under the receiver operating characteristic curve (AUC) of 0.70 (95% CI, 0.61 to 0.79), improving to 0.76 (95% CI, 0.68 to 0.84) after adjustment for rifampicin dose, TBM diagnostic certainty, Medical Research Council (MRC) severity grade, and HIV status. A CRP threshold of 40 mg/L provided the best balance between sensitivity (66%) and specificity (64%) for predicting mRS ≥4 at 8 weeks. Severe TBM (MRC grade 3) was more common among participants with CRP ≥40 mg/L than among those with lower CRP levels (33% vs 9%). Among participants with poor outcomes (mRS ≥4), 61.2% had baseline CRP ≥40 mg/L compared with 30.9% among those with better outcomes (mRS ≤3). Baseline CRP ≥40 mg/L was associated with a more than threefold increase in the odds of disability or death at 8 weeks (unadjusted OR, 3.53; 95% CI, 1.73 to 7.19; P < .001). This association remained significant after adjustment for potential confounders (adjusted OR, 2.78; 95% CI, 1.28 to 6.04; P = .010). The association did not differ significantly according to TBM diagnostic certainty or suspected unmasking TBM-IRIS status.

Elevated baseline serum CRP was independently associated with a substantially increased risk of disability or death at 8 weeks among adults with TBM, supporting its potential role as a low-cost prognostic biomarker. As an observational cohort analysis nested within a randomized clinical trial, the study provides moderate-level evidence for prognostic utility but does not establish causality. Important limitations include the relatively small sample size, missing CRP or outcome data in a subset of enrolled participants, and restriction to Ugandan referral hospitals, which may limit generalizability to other settings.

Source: Tugume L, Cresswell FV, Engen NW, Tukundane A, Kimuda S, Mugabi T, Namombwe S, Kagimu E, Kabahubya M, Ellis J, Bahr NC. Prognostic value of C-reactive protein in adults with tuberculous meningitis: a prospective cohort study. Clinical Infectious Diseases. 2025 Nov 15;81(5):e410-3.

Stool processing methods for Xpert ultra testing in childhood TB [TBN 094]

A prospective multicountry diagnostic accuracy study evaluated and compared three stool processing methods for Xpert MTB/RIF Ultra (Xpert Ultra) testing in children with presumed pulmonary tuberculosis (TB): the Simple Processing Kit (SPK), the Stool Optimization and Standardization (SOS) method, and the Optimized Sucrose Flotation (OSF) method. Children younger than 15 years were consecutively enrolled between June 2019 and March 2021 from tertiary hospitals and referral networks in India, South Africa, and Uganda. The study also assessed laboratory staff perceptions of the acceptability and usability of these stool processing approaches.

A total of 607 children were included. Eligible participants had microbiologically confirmed TB or at least one symptom suggestive of pulmonary TB, including prolonged cough, fever, failure to thrive, weight loss, or a chest radiograph consistent with TB. Children who had received anti-TB treatment for more than 72 hours were excluded. All participants underwent standardized TB evaluation including clinical assessment, chest radiography, and respiratory specimen collection. Respiratory samples were tested with Xpert Ultra and mycobacterial culture. Stool specimens were collected, processed within 72 hours, and tested using one or more of the three stool processing methods. Confirmed TB was defined by positive sputum Xpert Ultra or culture. Stool Xpert Ultra results were excluded from case classification. Laboratory personnel processing stool samples completed anonymous surveys evaluating usability and acceptability.

Among the 607 children, 61.1% were enrolled in Uganda, 60.5% were younger than 5 years, 50.7% were underweight, and 15.5% were living with HIV. Confirmed TB was diagnosed in 147 children (24.2%), and 92.5% of these cases were sputum Xpert Ultra positive. The proportion of valid stool test results was similar across methods, ranging from 87.4% for OSF to 90.3% for SPK. Positive stool Xpert Ultra results ranged from 9.3% (OSF) to 11.2% (SOS). Against the microbiological reference standard (MRS), all methods showed high specificity (97.1% to 98.2%) but modest sensitivity: 36.9% for SPK, 38.6% for SOS, and 31.3% for OSF. Sensitivities were substantially higher among children with higher sputum bacillary burden, reaching 95.8% to 100% when sputum Xpert Ultra semiquantitative results were low or higher, compared with only 16.7% to 22.6% among children with trace or very low sputum results (P < .001). 

Among the 179 children tested with all three methods, diagnostic performance was similar, with overlapping 95% confidence intervals. Compared with sputum culture, sputum Xpert Ultra sensitivity was 82.4% (95% CI, 56.6% to 96.2%), higher than stool Xpert Ultra, although the difference was not statistically significant. Combining stool and sputum testing increased sensitivity by 17.6% to 23.5% compared with stool testing alone, while adding stool testing to sputum Xpert Ultra increased sensitivity by up to 11.8% at the expense of a small reduction in specificity (2.1% to 2.9%). Survey responses from 17 laboratory staff indicated that all methods were acceptable and perceived as beneficial. SOS was viewed as the least time-consuming method, and 75% of respondents believed it could be performed by nonlaboratory staff in peripheral facilities.

The three stool processing methods demonstrated similar diagnostic accuracy, characterized by high specificity but limited sensitivity for childhood pulmonary TB. The SOS method appeared most favorable operationally because of its ease of use and lower time requirements. Key limitations include incomplete testing of all children with all three methods, variation in method implementation over time, and reliance on a microbiological reference standard that may miss pediatric TB cases. As a prospective multicountry diagnostic accuracy study, the evidence supports stool Xpert Ultra as a useful noninvasive adjunct to respiratory testing, particularly in settings where sputum collection is challenging.

Source: Jaganath D, Nabeta P, Nicol MP, Castro R, Wambi P, Zar HJ, Workman L, Lodha R, Singh UB, Bavdekar A, Sanghavi S. Stool processing methods for Xpert ultra testing in childhood tuberculosis: a prospective, multicountry accuracy study. Clinical Infectious Diseases. 2026 Mar 15;82(3):526-34.

Long-term mortality trends among individuals with TB in Brazil [TBN 093]

A study examined excess mortality among people diagnosed with tuberculosis (TB) in Brazil and assessed how mortality during the TB episode and the post-TB period contributed to long-term mortality. Using a retrospective population-based cohort design, investigators linked TB notifications from Brazil’s National Notifiable Disease Information System (SINAN) with mortality records from the Mortality Information System (SIM) for 2007–2016. The objective was to compare mortality among individuals with TB against the general population and evaluate how causes of death evolved over time after TB diagnosis.

The analysis included 834,594 individuals with TB after excluding records with postmortem diagnoses, revised non-TB diagnoses, missing demographic information, and individuals aged 90 years or older. Participants contributed 4,089,883 person-years of follow-up, with a mean follow-up of 4.9 years. Deaths were identified through linked mortality records and categorized into 11 cause-of-death groups based on ICD-10 codes. The primary outcome was mortality rate ratios (MRRs) comparing the TB cohort with the general population by year since diagnosis. The TB episode was defined as the first 12 months after diagnosis, while follow-up beyond 12 months was considered the post-TB period. Analyses were stratified by age, sex, and region, and adjusted MRRs (aMRRs) were estimated for risk factors associated with mortality.

Among 834,594 individuals, 120,330 deaths occurred during follow-up (14.4%). Mortality was markedly elevated during the first year after TB diagnosis (MRR 11.28, 95% CI 11.18–11.37) and declined over time but remained above that of the general population even 10 years later (MRR 1.46, 95% CI 1.34–1.59). During the TB episode, mortality was highest among individuals aged 30–44 years (MRR 24.97, 95% CI 24.59–25.33) and lowest among those aged 75–89 years (MRR 4.04, 95% CI 3.94–4.15). HIV infection was the strongest predictor of mortality during the TB episode (aMRR 5.16, 95% CI 5.05–5.27). Other factors associated with higher mortality included female sex (aMRR 1.67, 95% CI 1.64–1.70), combined pulmonary and extrapulmonary TB (aMRR 1.57, 95% CI 1.52–1.62), and abnormal chest X-ray findings (aMRR 1.77, 95% CI 1.65–1.89). During the post-TB period, elevated mortality remained associated with HIV infection (aMRR 2.80, 95% CI 2.73–2.87), treatment reinitiation after abandonment (aMRR 1.78, 95% CI 1.74–1.84), alcohol use disorder (aMRR 1.52, 95% CI 1.49–1.55), female sex (aMRR 1.41, 95% CI 1.38–1.43), treatment loss to follow-up, and drug-resistant TB requiring regimen changes (aMRR 1.65, 95% CI 1.53–1.77). Excess mortality accumulated substantially over time. At 1 year, cumulative mortality was 6.82% in the TB cohort versus 0.70% in the general population, yielding 6.12% excess mortality. By year 10, cumulative mortality reached 17.88% versus 7.97%, corresponding to 9.90% excess mortality, of which 5.22% occurred during the post-TB period. Causes of death shifted over time. In the first year, TB (31.3%) and HIV (23.9%) were the leading causes of death. By year 10, respiratory disease (16.3%), cardiovascular disease (16.1%), and cancer (13.3%) became the predominant causes.

The findings indicate that TB is associated with substantial excess mortality that persists for at least a decade after diagnosis, with more than half of total excess mortality occurring after completion of the initial TB episode. This large national cohort provides strong observational evidence, although causal inference is limited by its retrospective design and reliance on routinely collected administrative data. The results suggest that long-term follow-up and management of chronic comorbidities may be important components of care for TB survivors.

Source: Kim S, Pelissari DM, Harada LO, Sanchez M, Oliveira PB, Johansen FD, Maciel EL, Cohen T, Castro MC, Menzies NA. Long-term mortality trends among individuals with tuberculosis: a retrospective cohort study of individuals diagnosed with tuberculosis in Brazil. Clinical Infectious Diseases. 2026 Jan 15;82(1):e1-8.

A point-of-care prediction tool for recurrent tuberculosis [TBN 092]

A study aimed to develop and validate a simple prediction model for recurrent tuberculosis (TB) that could be used at treatment completion by outreach workers to identify TB survivors at highest risk for recurrence in resource-limited, high-burden settings. The analysis used data from the TB Aftermath noninferiority trial conducted in Maharashtra, India. Participants were enrolled between January 2021 and October 2023, and follow-up data through October 2024 were analyzed.

The TB Aftermath trial enrolled 1,076 adults (≥18 years) who had completed or been cured of TB according to India's National TB Elimination Programme (NTEP) definitions. Individuals with pulmonary and/or extrapulmonary TB were eligible regardless of treatment regimen. Participants entered the study within 60 days of treatment completion and were randomized to phone-based or home-based symptom screening at 6 and 12 months after treatment. All participants also received an 18-month home visit. For this analysis, investigators included participants with at least 12 months of follow-up or those who experienced an outcome. Individuals lost to follow-up within 12 months and those who died without documented TB recurrence were excluded. The primary outcome was recurrent TB diagnosed within 18 months of treatment completion, including both microbiologically confirmed and clinically confirmed recurrences. Candidate predictors were evaluated using exploratory analyses and LASSO regression, followed by model development and validation using separate training (60%) and validation (40%) datasets.

Among 1,033 eligible TB survivors, 85 (8.2%) experienced recurrent TB within 18 months. Of these recurrences, 52 (61%) were microbiologically confirmed and 64 (75%) involved pulmonary disease. The median participant age was 35 years (IQR 27-48), 52% were male, and the median BMI was 19.9 kg/m² (IQR 17.6-23.2). Nine predictors were shortlisted: sex, household income, biomass fuel exposure, BMI, unhealthy alcohol use (AUDIT), smoking history, peak expiratory flow (PEF), disease site, and history of more than one TB episode. The highest-performing and most practical models contained five variables. Several five-item models demonstrated moderate discrimination, all including BMI and PEF. The selected model consisted of sex, household income, BMI, PEF, and history of more than one TB episode, achieving a cross-validated c-statistic of 0.69 (95% CI 0.56-0.83). Model calibration in the validation set was acceptable (Hosmer-Lemeshow P=.053; calibration intercept 0.03, 95% CI -0.03 to 0.09; slope 0.66, 95% CI 0.08-1.24). Performance did not differ significantly between early and late recurrence. Sensitivity analyses, including complete-case analyses and analyses restricted to microbiologically confirmed recurrence, identified the same five predictors and supported the robustness of the model.

The study concluded that a parsimonious five-item model using routinely obtainable characteristics can moderately predict TB recurrence after treatment completion among TB survivors in India. As a prediction-model development and validation study, the level of evidence is moderate. Limitations include modest discriminatory performance, relatively low recurrence event numbers, missing and poor-quality PEF measurements, and potential limitations in generalizability beyond similar high-burden public-sector settings. 

Source: Cox SR, Moe AH, Gupte AN, Kadam A, Valawalkar S, Gupte N, Lele G, Kendall EA, Baillie C, Barthwal MS, Kakrani A. A point-of-care prediction tool for recurrent tuberculosis. Clinical Infectious Diseases. 2025 Dec 15;81(6):e612-22.

BMI and incident tuberculosis in close TB contacts [TBN 091]

A longitudinal observational study developed and internally validated a prediction model for progression to active tuberculosis (TB) among close contacts of culture-confirmed pulmonary TB patients in the RePORT-Brazil cohort. The study enrolled participants between June 2015 and June 2019 across four cities in three Brazilian states (Rio de Janeiro, Bahia, and Amazonas) and followed contacts for 24 months. The objective was to identify predictors of progression from exposure or latent infection to active TB disease.

The study included 1,846 close contacts of 619 pulmonary TB index patients, with 86% completing at least 1.5 years of follow-up. Close contacts were defined as individuals exposed to a culture-positive pulmonary TB patient for at least 4 hours per week during the 6 months preceding diagnosis. At baseline, contacts underwent clinical evaluation, chest radiography, interferon-gamma release assay (IGRA) testing, HIV testing, and collection of sociodemographic and clinical data. Contacts were eligible only if they were asymptomatic for TB at enrollment. Follow-up assessments occurred at 6 months in person and every 6 months thereafter by telephone, with clinical evaluation if symptoms developed. Progression to TB was defined by microbiological confirmation or clinical diagnosis. Thirty baseline variables were initially considered. Predictor selection involved empirical review followed by least absolute shrinkage and selection operator (LASSO) regression with 5-fold cross-validation. Cox proportional hazards models were used after variable selection.

Among the 1,846 contacts, 25 (1.4%) developed TB within 24 months. Of these, 13 cases were microbiologically confirmed, 2 had histopathological evidence suggestive of TB, and 10 were clinically diagnosed; 75% had pulmonary TB. Baseline IGRA positivity was more common among progressors than non-progressors (75% vs 36.3%). The final prediction model identified three predictors of progression: tuberculosis preventive therapy (TPT), smoking status, and baseline IGRA positivity. 

Internal validation demonstrated good discrimination with an optimism-corrected AUC of 0.81 (95% CI 0.74-0.88). Contacts who never initiated TPT had substantially higher risk of developing TB than those who completed TPT (adjusted hazard ratio [aHR] 16.55, 95% CI 2.20-124.43). Among IGRA-positive contacts, TPT remained the only retained predictor, with an optimism-corrected AUC of 0.73 (95% CI 0.64-0.79); failure to start TPT was associated with increased TB risk (aHR 14.75, 95% CI 1.95-111.68). In the subgroup of IGRA-positive contacts who did not receive TPT or received it for less than 30 days, lower body mass index (BMI) emerged as an additional predictor. Each 1 kg/m² increase in BMI was associated with a 13% reduction in TB risk (aHR 0.89, 95% CI 0.80-0.99). A BMI threshold of approximately 25 kg/m² was identified, with BMI <25 kg/m² associated with higher risk of progression (aHR 4.14, 95% CI 1.17-14.67). In this high-risk subgroup, TB incidence was 8.4% among those with BMI <25 kg/m² compared with 2.1% among those with BMI ≥25 kg/m².

The study concluded that lack of TPT, baseline IGRA positivity, smoking status, and lower BMI are important predictors of progression to active TB among close contacts of pulmonary TB patients. Completion of TPT appeared to be the strongest protective factor. Major limitations include the small number of TB progressors (25 events), which may have reduced statistical precision and increased uncertainty around effect estimates. Internal validation showed good model performance, but external validation in other populations is needed before broader implementation. 

Source: Arriaga MB, Amorim G, Figueiredo MC, Staats C, Kritski AL, Cordeiro-Santo M, Rolla VC, Rebeiro PF, Andrade BB, Sterling TR. Body mass index and incident tuberculosis in close tuberculosis contacts. Clinical Infectious Diseases. 2026 Jan 15;82(1):e100-9.