Global Burden of Tuberculosis and Drug-Resistant Subtypes Among Women of Childbearing Age from 1990 to 2021, with Projections to 2030 Authors: Jiawang Chen, Xiaozhen Peng, Yixiang Zheng, Ruochan Chen Published: 2026 ABSTRACT AbstractBackgroundTuberculosis (TB) poses a major threat to women of childbearing age (WCBA, 15–49 years), yet comprehensive global burden analyses remain limited. This study provides a comprehensive global spatiotemporal analysis of TB and drug-resistant subtypes among WCBA.MethodsUsing Global Burden of Disease Study 2021 data, we assessed the global incidence and disability-adjusted life years (DALYs) of TB and drug-resistant subtypes among WCBA from 1990 to 2021, and projected trends to 2030 using Bayesian age-period-cohort models.ResultsIn 2021, an estimated 1.99 million WCBA developed TB globally. The age-standardized incidence rate (ASIR) and age-standardized DALY rate (ASDR) declined for overall TB and drug-susceptible TB (DS-TB), but increased for multidrug-resistant TB (MDR-TB) and extensively drug-resistant TB (XDR-TB). The large relative increases in XDR-TB should be interpreted cautiously because baseline estimates were very small and case definitions, testing capacity, and surveillance changed over time. The TB burden showed substantial geographic heterogeneity, with the highest age-standardized rates concentrated in Sub-Saharan Africa. Age-specific patterns showed higher ASIRs in younger populations, whereas ASDRs increased with age. Among the four GBD-attributable exposures included in the comparative analysis, high body-mass index had the largest modeled TB-related DALY rate in 2021. This restricted comparison should not be interpreted as ranking all established TB determinants. Projections suggest that overall TB incidence will decline, whereas the modeled XDR-TB ASIR may continue to rise through 2030 globally.ConclusionReducing TB among WCBA requires improved access to rapid diagnosis, drug susceptibility testing, effective DR-TB treatment, and risk-based TB screening within reproductive and antenatal care, particularly in low-SDI and high-burden settings. Pregnancy- and postpartum-specific surveillance should also be strengthened, as these populations are not separately identifiable in GBD 2021.Keywords: tuberculosis, MDR-TB, XDR-TB, women of childbearing age, global burden of disease, Bayesian age-period-cohort Introduction Tuberculosis (TB), a preventable and curable chronic infectious disease, remains a primary contributor to global infectious disease mortality.1,2 In 2023, approximately 3.6 million women aged 15 years developed TB, with women of childbearing age (WCBA, 15 to 49 years) representing the highest proportion.2 Despite this substantial burden, comprehensive global analyses quantifying TB epidemiology specifically among WCBA remain limited. This demographic encompasses pregnant women, who face unique vulnerabilities: TB can seriously threaten pregnancy, contributing to 6–15% of global maternal mortality and increasing the risks of severe pregnancy complications, adverse fetal outcomes, and vertical transmission.2,3 The risk of active TB (ATB) rises substantially during pregnancy and postpartum,4,5 influenced by disease characteristics, treatment response, gestational age at diagnosis, and human immunodeficiency virus (HIV) co-infection.6 Drug-resistant TB (DR-TB), particularly multidrug-resistant TB (MDR-TB) and extensively drug-resistant TB (XDR-TB), poses a serious public health threat. Pregnant women represent a special sub-population of concern in DR-TB management, yet the global burden in this group remains poorly quantified.7 Clinical management of DR-TB is further complicated by the teratogenicity of second-line anti-TB medications, creating unique therapeutic challenges.8 Moreover, limited pregnancy-specific pharmacokinetic and safety data on anti-TB medications, with pregnant and lactating women remaining underrepresented in clinical trials, underscore the need for their timely and ethically appropriate inclusion in clinical research to improve evidence-based DR-TB treatment recommendations during pregnancy.9 Preventing and controlling TB among WCBA is an important strategy for contributing to Targets of Sustainable Development Goal 3, which aims to end the TB epidemic and reduce the maternal mortality ratio to below 70 per 100,000 live births by 2030.10 Previous GBD studies have examined TB burden in the general population and selected demographic groups.1,11 However, global age-specific and geographic patterns of overall TB and mutually exclusive drug-resistance categories among women aged 15–49 years have not been comprehensively characterized. Such evidence may inform resource allocation, risk-based TB screening within maternal and reproductive health services, and DR-TB care for pregnant and lactating women.12,13 Because GBD 2021 does not identify pregnancy or postpartum status and our analysis was not stratified by HIV status, the findings provide population-level rather than pregnancy-specific or HIV-specific evidence. We therefore characterized temporal, geographic, age, and socioeconomic patterns of TB and its drug-resistant subtypes among WCBA, examined selected GBD-attributable exposures, and projected disease trajectories globally and in eight high-burden countries. Methods Study Population The study focused on WCBA, defined by the WHO as females aged 15 to 49 years, a period characterized by peak reproductive potential and cyclical fluctuations in sex hormones.14 For age-specific analysis, we divided WCBA into seven age groups: 15–19, 20–24, 25–29, 30–34, 35–39, 40–44, and 45–49 years. Data Source The Global Burden of Disease Study (GBD) 2021 employs advanced statistical methods, such as the Bayesian meta-regression modeling tool (DisMod-MR 2.1), to comprehensively assess health loss from 371 diseases/injuries and 88 attributable risk factors across 204 countries and territories.15,16 The analysis provides age-, sex-, year-, and region-specific estimates between 1990 and 2021, integrating data from civil registration systems, mortality interviews, censuses, household surveys, condition-specific registries, health service records, and additional sources.15 All data were extracted from the GBD 2021 Results Tool, accessed through the Global Health Data Exchange (GHDx) platform (https://ghdx.healthdata.org). Prespecified extraction fields included measure, metric, cause, location, age, sex, and year. The detailed search criteria and applied filters are provided in Appendix 1. For projection analyses, age- and sex-specific population projections for 2022–2030 were obtained from the IHME Global Population Forecasts 2017–2100 dataset (https://ghdx.healthdata.org/record/ihme-data/global-population-forecasts-2017-2100). In this study, we quantified TB burden among WCBA using incidence and disability-adjusted life years (DALYs). Here, TB denotes tuberculosis overall. In the GBD cause hierarchy used for this analysis, DS-TB, MDR-TB excluding XDR-TB, and XDR-TB were treated as mutually exclusive categories that together comprise overall TB. This distinction should be retained when comparing subtype estimates. Data for women aged 15 to 49 years (from 1990 to 2021) were extracted across geographic hierarchies: global, five socio-demographic index (SDI) regions, 21 GBD regions (grouped by geographic/epidemiologic homogeneity),15 and 204 countries/territories. The modeled XDR-TB estimate for 1990 was zero; therefore, log-linear trend analyses for XDR-TB began in 1991. Historical XDR-TB estimates require caution because definitions, drug susceptibility testing coverage, and surveillance changed substantially during the study period.11 Rates were calculated per 100,000 population, and the 95% uncertainty intervals (UIs) were defined by the 2.5th and 97.5th percentiles of 1000 posterior draws.17 Socio-Demographic Index The SDI is an integrated metric reflecting the health-associated socioeconomic development level of a nation or region. It is derived as the geometric mean of three parameters: lag-distributed income per capita, average educational attainment in individuals aged 15 years or older, and the fertility rate among females under 25.15,16 SDI scores range from 0 to 1, with higher values denoting greater advancement. Based on GBD 2021 estimates,18 204 countries and territories were classified into five SDI tiers: low (<0.46), low-middle (0.46–0.60), middle (0.61–0.69), high-middle (0.70–0.81), and high (>0.81).15 These SDI tiers were used for stratified analyses to examine how development disparities influence TB burden among WCBA, whereas the underlying SDI values were treated as continuous variables in analyses examining the association between SDI and TB burden. Case Definition TB is a potentially fatal infectious disease which results from infection with Mycobacterium tuberculosis, manifests as pulmonary, extrapulmonary, or other forms confirmed clinically or bacteriologically, with DS-TB demonstrating sensitivity to all first-line anti-TB drugs (eg, isoniazid, rifampicin, pyrazinamide, and ethambutol), while MDR-TB exhibits resistance to both isoniazid and rifampicin, and XDR-TB further resists fluoroquinolones plus either linezolid or bedaquiline.19 Statistical Analysis In order to eliminate confounding effects from variations in age structure among WCBA populations on the incidence and DALY rates and to enable comparable assessments of TB burden across temporal and geographic dimensions, we calculated age-standardized incidence rates (ASIRs) and age-standardized DALY rates (ASDRs) per 100,000 population based on age-specific estimates obtained from the GBD 2021 database. Specifically, age-specific estimates were extracted for the seven predefined WCBA age groups. ASIRs and ASDRs were calculated by applying the corresponding weights from the GBD standard population to the age-specific rates using the following formula: refers to the age-specific rate within the i-th age group of WCBA and represents the corresponding age group’s weight in the GBD 2021 global standard population (derived from aggregated demographic data spanning 1950–2021).17 N represents the total number of age groups. The Joinpoint regression model was applied to estimate average annual percentage changes (AAPCs) with 95% confidence intervals (CIs) to quantify temporal trends in TB burden among WCBA. Analyses were performed using the Joinpoint Regression Program, version 5.2.0.0 (April 2024), developed by the Statistical Research and Applications Branch, National Cancer Institute. The model employed a grid search algorithm for optimal joinpoint identification and weighted least-squares regression for segment-specific slope estimation.20 The optimal number of joinpoints was selected using permutation tests.21 For the 1990–2021 study period, a maximum of six joinpoints was allowed, with a minimum of two observations required between adjacent joinpoints and at least two observations required from each joinpoint to either end of the study period. The AAPC was calculated as follows: Where refers to the slope coefficient for segment i and refers to duration (years) of segment i. The calculated annual ASRs were entered into the Joinpoint Regression Program without additional manual log transformation. Because zero values cannot be included in log-linear models, the XDR-TB estimate for 1990 (ASR = 0) was excluded, and no imputation or replacement was performed. Trend interpretation was based on the AAPC and its 95% CI. An upward (or downward) trend was concluded if both the AAPC estimate and 95% CI were > 0 (or < 0). If the 95% CI encompassed 0, the temporal trend was regarded as non-significant.22 The smoothing spline model was applied to evaluate the nonlinear associations between age-standardized rates (ASRs) for TB, DS-TB, MDR-TB, and XDR-TB and SDI across regions and nations. For these analyses, year-specific SDI values were paired with annual ASRs for each region, rather than using SDI values averaged over the study period. Statistical significance was assessed using Spearman’s rank correlation coefficients (r values) and corresponding P values, and the associations were visualized using locally weighted scatterplot smoothing curves.23 GBD 2021 comprehensively assesses the burden of specific health outcomes attributable to 88 risk factors.15,16 Aligning with the five key TB risk factors (nutritional status, HIV infection, alcohol use disorders, smoking, and diabetes) highlighted in the WHO Global Tuberculosis Report 2024,2 this study selected four risk factors with available comparable estimates within the GBD 2021 risk factor framework for analysis. The selection was further informed by TB risk factor rankings visualized in the GBD Compare tool (https://vizhub.healthdata.org/gbd-compare/) (Figure S1), resulting in the inclusion of high body-mass index (BMI), high fasting plasma glucose (FPG), high alcohol use, and smoking. Although undernutrition is a well-established risk factor for TB, corresponding WCBA-specific attributable estimates were unavailable within the GBD 2021 risk factor framework. Therefore, high BMI was selected as the available metabolic risk factor for comparative risk assessment in this population. For the risk-attributable analyses, DALYs attributable to these four risk factors were extracted for females aged 15–49 years, corresponding to the WCBA population defined in this study. Because age-specific estimates were not consistently available across the seven predefined WCBA age groups for all four risk factors, age-standardized DALY rates could not be calculated. Consequently, crude DALY rates were used to quantify the TB burden attributable to each selected risk factor. Finally, future TB trends (including ASIR and ASDR) at the global level and in the eight high-burden countries (India, Indonesia, China, the Philippines, Pakistan, Nigeria, Bangladesh, and the Democratic Republic of the Congo) from 2022 to 2030 were projected using the Bayesian age-period-cohort (BAPC) model implemented in the BAPC R package (version 0.0.36), with model estimation performed using the integrated nested Laplace approximation (INLA) framework.24,25 Age, period, and cohort effects were modeled using second-order random-walk (RW2) structures with log-gamma priors, according to the default model specification of the BAPC package. An independent and identically distributed overdispersion component was also included, and a Poisson likelihood was specified, with age- and sex-specific projected population-years incorporated as the exposure term in the model. Additionally, separate BAPC models were fitted to project global ASIRs for DS-TB, MDR-TB, and XDR-TB. All analyses and data visualizations were executed in R software (version 4.4.2), with a two-sided P-value < 0.05 denoting statistical significance. Results Regional Disparities in the Overall Burden of TB Among WCBA In 2021, an estimated 1.99 million incident TB cases and 7.83 million DALYs occurred globally among WCBA, corresponding to an ASIR of 102.95 (95% UI: 63.36 to 152.84) and an ASDR of 399.31 (352.31 to 454.75) per 100,000, respectively. From 1990 to 2021, both the ASIR (AAPC: −1.49, 95% CI: −1.51 to −1.48) and ASDR (AAPC: −3.08, 95% CI: −3.14 to −3.03) showed obvious downward trends (Table 1 and Figure 1). At the regional level, Southern Sub-Saharan Africa had the highest ASIR (700.29 per 100,000, 95% UI: 397.61 to 1,057.04), while Central Sub-Saharan Africa had the highest ASDR (2,049.79, 1,435.55 to 2,856.03) in 2021. The lowest rates were observed in High-income North America (ASIR: 2.27, 1.33 to 3.53) and Australasia (ASDR: 2.35, 1.70 to 3.32) (Table S1, Figure 2A and B). All regions experienced declining ASRs, with the most pronounced reductions in Andean Latin America for ASIR (AAPC: −4.09, 95% CI: −4.16 to −4.03) and High-income Asia Pacific for ASDR (−7.76, −7.92 to −7.61) (Table S1, Figure 2C and D). At the national level, ASIRs ranged from 1.95 (the United States) to 845.54 (South Africa), and ASDRs ranged from 0.58 (Andorra) to 4,653.42 (the Central African Republic). Only five countries showed rising ASIRs, led by the Philippines (0.72, 95% CI: 0.70 to 0.74), while only three countries experienced rising ASDRs, led by Lesotho (1.91, 1.73 to 2.08). Conversely, the most pronounced declines were observed in Ecuador for ASIR (−5.06, −5.35 to −4.79) and the Maldives for ASDR (−9.51, −9.62 to −9.40) (Table S1 and Figure S2). Table 1Age-Standardized Rates (per 100,000 Population) and Trends of TB and Drug-Resistant Subtypes Among Women of Childbearing Age Globally, 1990–2021aCharacteristicsNumber (95% UI), 1990ASR (95% UI), 1990Number (95% UI), 2021ASR (95% UI), 2021AAPC of ASR(95% CI), 1990–2021Incidence TB2,219,580.15 (1,812,690.50 to 2,766,148.94)164.32 (100.10 to 243.78)1,991,502.29 (1,663,899.03 to 2,426,135.45)102.95 (63.36 to 152.84)−1.49 (−1.51 to −1.48) DS-TB2,208,364.85 (1,804,105.90 to 2,752,231.47)163.49 (99.66 to 242.59)1,881,245.38 (1,555,568.35 to 2,306,786.12)97.27 (59.62 to 145.18)−1.66 (−1.67 to −1.64) MDR-TB11,215.30 (4,969.88 to 25,807.20)0.83 (0.32 to 1.99)105,646.07 (56,996.80 to 192,081.49)5.45 (2.49 to 11.01)6.08 (5.84 to 6.36) XDR-TBb52.34 (13.47 to 192.33)0.0038 (0.0009 to 0.0140)4,610.84 (3,035.95 to 7,017.01)0.24 (0.13 to 0.40)14.40 (14.06 to 14.94)DALYs TB13,661,194.84 (12,435,634.36 to 15,087,845.61)1,046.98 (941.26 to 1,166.73)7,827,468.54 (7,070,000.72 to 8,730,745.88)399.31 (352.31 to 454.75)−3.08 (−3.14 to −3.03) DS-TB13,563,182.21 (12,319,532.11 to 15,008,050.41)1,039.41 (932.86 to 1,158.05)7,074,205.60 (6,060,682.90 to 7,955,922.41)360.95 (303.86 to 416.70)−3.37 (−3.43 to −3.33) MDR-TB98,012.63 (34,991.81 to 238,319.25)7.56 (2.68 to 18.31)708,278.86 (277,116.17 to 1,412,277.21)36.08 (14.15 to 73.52)5.09 (4.95 to 5.25) XDR-TBb22.47 (5.26 to 94.31)0.0017 (0.0004 to 0.0069)44,984.07 (18,395.28 to 86,034.85)2.28 (0.94 to 4.42)30.86 (28.37 to 35.24)Notes: bFor XDR-TB, ASR in 1990 was zero and was excluded from log-linear Joinpoint analysis. The XDR-TB AAPC therefore covers 1991–2021.Abbreviations: aAAPC, average annual percentage change; ASR, age-standardized rate; CI, confidence interval; DALYs, disability-adjusted life years; DS-TB, drug-susceptible TB; MDR-TB, multidrug-resistant TB; TB, tuberculosis; UI, uncertainty interval; XDR-TB, extensively drug-resistant TB. Age-Standardized Rates (per 100,000 Population) and Trends of TB and Drug-Resistant Subtypes Among Women of Childbearing Age Globally, 1990–2021a Notes: bFor XDR-TB, ASR in 1990 was zero and was excluded from log-linear Joinpoint analysis. The XDR-TB AAPC therefore covers 1991–2021. Abbreviations: aAAPC, average annual percentage change; ASR, age-standardized rate; CI, confidence interval; DALYs, disability-adjusted life years; DS-TB, drug-susceptible TB; MDR-TB, multidrug-resistant TB; TB, tuberculosis; UI, uncertainty interval; XDR-TB, extensively drug-resistant TB. Temporal trends in ASIR and ASDR of TB and its drug-resistant subtypes among women of childbearing age, globally and by SDI region. Here, TB refers to tuberculosis overall. Trends are shown for the period from 1990 to 2021. Data for XDR-TB are available from 1991 onward. Global and regional (21 GBD regions) burden and trends of TB and drug-resistant subtypes among women of childbearing age. (A and B) ASIRs and ASDRs in 2021. (C and D) AAPCs of ASIR and ASDR from 1990 to 2021. For XDR-TB, AAPCs cover 1991–2021. Regional Disparities in the Burden of Drug-Resistant TB Subtypes In 2021, the ASIR and ASDR for DS-TB were 97.27 (95% UI: 59.62 to 145.18) and 360.95 (303.86 to 416.70) per 100,000 population, respectively, reflecting declining trends consistent with overall TB patterns (Table S2, Figures 1, 2 and S3). In contrast, MDR-TB and XDR-TB exhibited distinct patterns. MDR-TB accounted for 105,646 new cases (95% UI: 56,996 to 192,081) and 708,279 DALYs (277,116 to 1,412,277) globally in 2021 (ASIR: 5.45, 2.49 to 11.01; ASDR: 36.08, 14.15 to 73.52). Its ASIR peaked around 2005, then declined gradually before rising again post-2015, while ASDR peaked around 2002 and subsequently declined until 2021. Despite these fluctuations, net increases were observed over the entire period [AAPCs: 6.08 (95% CI: 5.84 to 6.36) for ASIR and 5.09 (4.95 to 5.25) for ASDR] (Table 1 and Figure 1). Geographically, MDR-TB burden was concentrated in Southern Sub-Saharan Africa, which recorded the highest ASIR (22.77 per 100,000) in 2021, followed by Eastern Europe (14.08), South Asia (13.30), and Central Asia (10.30). Sub-Saharan Africa and South Asia also carried the highest ASDR (Table S3, Figure 2A and B). 17 regions had increasing ASIRs and 13 had increasing ASDRs, with the largest increases in Central Asia and Oceania. High-income North America had the largest decreases in both measures (Table S3, Figure 2C and D). The burden of XDR-TB remained comparatively lower, with 4,611 new cases and 44,984 DALYs globally in 2021, corresponding to an ASIR of 0.24 and an ASDR of 2.28 per 100,000 population. However, XDR-TB demonstrated the most aggressive growth over the entire period, with ASIR showing a consistent upward trend since 1991 (AAPC: 14.40, 95% CI: 14.06 to 14.94) and ASDR exhibiting an overall increase (AAPC: 30.86, 28.37 to 35.24) despite a slow and fluctuating downward trend since 2005 (Table 1 and Figure 1). At the regional level, Eastern Europe and Central Asia had markedly higher ASRs than other regions, and all GBD regions experienced substantial increases from 1991 to 2021, with the sharpest rise observed in Oceania [AAPC: 26.60 (25.98 to 27.40) for ASIR and 42.48 (40.18 to 45.21) for ASDR] (Table S4 and Figure 2). These trends should be interpreted with caution, as improvements in laboratory testing capacity, evolving case definitions, and strengthened surveillance systems over the study period may have contributed to the observed increases. At the national level, for MDR-TB, 80.9% of countries exhibited increasing ASIR trends, with the largest increase in Kyrgyzstan (AAPC: 19.71). In contrast, 19.1% of countries showed declining trends, with the largest decrease in Slovenia (−9.08). Similarly, the ASDR increased in the majority of countries (61.8%), with the greatest rise occurring in Lesotho (16.59) and the largest decline in Slovenia (−13.34) (Table S3 and Figure S4). For XDR-TB, Kyrgyzstan (4.16 per 100,000) and the Republic of Moldova (3.62) reported the highest ASIRs in 2021, while Tajikistan recorded the highest ASDR (30.67). From 1991 to 2021, nearly all countries showed an increasing trend in estimated XDR-TB burden, though the magnitude of these increases likely reflects both true epidemiological changes and surveillance improvements. The steepest rises were observed in Lesotho (AAPC: 30.44 for ASIR and 51.65 for ASDR), whereas the smallest increases were in Slovenia (2.25, 8.77) (Table S4 and Figure S5). Age and Socio-Demographic Patterns in TB Burden Among WCBA Globally, both ASIR and ASDR declined across all age groups, with the greatest reduction in women aged 45–49 years and the smallest in those aged 20–24. In 2021, the incidence number and ASIR were highest at 20–24 years, closely followed by the 15–19 group. This indicates a concentration of incident TB in the younger segments. In contrast, ASDR rose steadily with age, reaching a peak at 45–49 group, reflecting the severity and fatal consequences of TB accumulating in older WCBA (Table S5 and Figure 3). Age-specific patterns for DS-TB and MDR-TB resembled overall TB, though DS-TB ASR declined more sharply. Conversely, MDR-TB and XDR-TB burden showed substantial increases across all age groups since 1990. Notably, the ASIR of XDR-TB was highest at 35–39 years (Table S5 and Figure S6–S8). Age-specific disparities in incidence and DALYs of TB among women of childbearing age globally and by SDI region, 1990 and 2021. Incidence in 1990 (A) and 2021 (B) and DALYs in 1990 (C) and 2021 (D). Regionally, low and low-middle SDI regions showed the highest ASIRs for TB and its subtypes in either the 20–24 or 45–49 age groups, with 20–24 years having the most cases and 45–49 years the fewest, reflecting a younger WCBA population in these regions. The global age pattern was most closely mirrored in middle SDI regions. In high-middle SDI regions, the ASIRs of overall TB and DS-TB peaked in the 20–24 group, while ASIRs of MDR-TB and XDR-TB formed an inverted U-shape, peaking in the 35–39 group. In high SDI regions, ASIR showed minimal variation across age groups for all TB types. ASDR trends revealed that age-related variation diminished with higher SDI: lower SDI regions showed steep increases in ASDR with age, whereas higher SDI regions exhibited much less variation. An exception was XDR-TB in high-middle SDI regions, where ASDR rose sharply with age (Figure 3 and Figure S6–S8). The Association Between ASR and SDI From 1990 to 2021, the ASIR and ASDR of TB among WCBA declined across all SDI regions. Although the declines were most substantial in high-middle and high SDI regions, the disparity in TB burden across all quintiles narrowed markedly, as the absolute improvements in lower-SDI regions contributed to closing the gap. Trends for DS-TB closely mirrored those of overall TB (Table S1 and Figure 1). For MDR-TB, the highest ASIR was recorded in low-middle SDI regions (9.04 per 100,000, 95% UI: 3.19 to 21.21), while the highest ASDR were seen in low SDI regions (74.73, 31.60 to 151.44). For XDR-TB, the highest ASIR and ASDR were in high-middle (0.56 per 100,000, 0.29 to 0.96) and low-middle SDI (4.26, 1.33 to 9.66) regions, respectively. Between 1990 and 2021, only high SDI regions showed declining MDR-TB trends. The other four SDI quintiles saw increases, most noticeably in low-middle SDI regions. In contrast, all SDI regions exhibited pronounced upward trends in XDR-TB ASRs (AAPC > 10), with the largest increases in low-SDI regions (Table S3, S4 and Figure 1). Spearman’s rank correlation analysis revealed strong inverse correlations between SDI and both the ASIR (r = −0.78, P < 0.001) and ASDR (r = −0.85, P < 0.001) of TB globally. From 1990 to 2021, rising SDI was consistently associated with declining TB ASRs across all GBD regions. Most regions with SDI between 0.5 and 0.7 exhibited lower burdens, with notable exceptions being Southern/Central Sub-Saharan Africa and Southeast Asia, which consistently had higher-than-expected ASRs (Figure 4A and B). DS-TB followed a similar pattern to overall TB (Figure 4C and D). The correlation for MDR-TB with SDI was also negative but weaker, with most regions exhibiting an initial rapid rise peaking (most notably in Southern Sub-Saharan Africa, Central Asia, Eastern Europe, and South Asia), followed by a slower decline (Figure 4E and F). In contrast, XDR-TB burden showed no significant linear correlation with SDI. Instead, a non-linear pattern emerged: ASRs remained relatively stable at lower SDI levels, began to rise around SDI 0.5, peaked near 0.7, and then declined – a trend prominently observed in Eastern Europe and Central Asia, where ASRs showed rapid increases followed by sharp declines after peaking (Figure 4G and H). Association between SDI and ASIR/ASDR of TB and its drug-resistant subtypes among women of childbearing age globally and across 21 GBD regions from 1990 to 2021. Panels show the relationship for TB (A and B), DS-TB (C and D), MDR-TB (E and F), and XDR-TB (G and H). The solid line and shaded area represent the expected values with 95% CIs, based on SDI and disease rates in all GBD regions. For each region, 32 annual observations are plotted, with each point representing the corresponding year-specific SDI and ASIR/ASDR from 1990 to 2021 (XDR-TB: 1991–2021). Points above the solid line represent a higher-than-expected burden, and those below the line show a lower-than-expected burden. At the national level in 2021, ASRs of TB, DS-TB, MDR-TB, and XDR-TB among WCBA demonstrated a general negative correlation with SDI. However, some higher-SDI countries exhibited higher burdens than certain lower-SDI countries, emphasizing the necessity for TB prevention and control strategies to be tailored to each country’s specific epidemiological and socioeconomic context (Figure S9–S12). Risk Factors Between 1990 and 2021, the global TB DALY rate linked to high BMI rose, ultimately replacing high alcohol use as the primary risk factor for TB among WCBA by 2021. In contrast, rates attributable to high alcohol use, smoking, and high FPG declined globally. This general declining trend was observed across most SDI quintiles and GBD regions. However, Eastern Europe exhibited substantial increases in DALY rates from high BMI, smoking, and high FPG. Rises attributable to high BMI were also noted in low-middle SDI regions, the Caribbean, and Central Sub-Saharan Africa, while Central Asia saw an increase linked to high FPG (Table S6 and Figure 5). Regional rankings of four TB risk factors (high BMI, high FPG, smoking, high alcohol use) by DALYs crude rates in 1990 (A) and 2021 (B). Numbers 1–4 within tiles denote the rank (1 = highest) of each factor per region. In 2021, the leading risk factors varied considerably across regions. High alcohol use remained the primary driver in low SDI regions and much of Sub-Saharan Africa. Despite its diminished role elsewhere, smoking remained the primary risk factor in Central Europe and Oceania. The burden attributable to high FPG was consistently lower in most regions, except in South Asia (Figure 5). Projection of TB Burden Among WCBA to 2030 BAPC analysis projects that by 2030, the global ASIR of TB will decrease by approximately 7.3% from 2021 levels, while the ASDR will decline by about 17.7% (Table S7). Although the global incident cases are projected to remain relatively steady, DALYs number is expected to decline substantially (Figure 6A and B). In contrast, the global ASIR of DR-TB, particularly XDR-TB, is projected to continue rising, although XDR-TB estimates remain subject to surveillance limitations (Table S8). Without sustained investment in both TB control programs and surveillance systems, the DR-TB burden may continue to grow. Projected incidence and DALYs for TB among women of childbearing age (2022–2030) based on BAPC modeling. Forecasts for global level (A and B), India (C and D), and China (E and F). Error bars and shaded area denote 95% UI. Figure 6C–6F present the projections for the two most populous countries, India and China. The projections for the other six high-TB-burden countries are provided in Figure S13. While both India and the Democratic Republic of the Congo are expected to see a rise in TB case numbers, their ASIRs are projected to decline. The Philippines, however, is projected to experience substantial rises in both incidence number and ASIR. The remaining five high-burden countries are projected to show substantial declines in TB burden. Examining longer-term trends (1990–2030), China and Bangladesh demonstrate steady declines in TB incidence, indicating sustained effective control. Conversely, the Philippines and the Democratic Republic of the Congo exhibit persistent upward trends, underscoring an urgent need for intensified interventions. India continues to report an exceptionally high number of new TB cases, far exceeding other countries. This extreme burden is likely associated with rapid population growth, high population density, and healthcare system challenges.26 Discussion While global TB incidence and DALYs declined from 1990 to 2021, driven by improved diagnostics, expanded Bacillus Calmette-Guerin (BCG) vaccination, and socioeconomic advancements, the burden of MDR-TB and XDR-TB showed concerning upward trends. This divergence underscores an urgent need for targeted interventions addressing the dual challenges of declining DS-TB and increasing drug resistance. This study presents a spatiotemporal analysis of TB and its drug-resistant subtypes among WCBA over three decades. Below, we contextualize these trends, propose actionable solutions, and highlight critical areas for policy reform. Our study was not designed to stratify by pregnancy status, as GBD data do not include this information. Nevertheless, our findings provide crucial macro-level evidence for strengthening TB control among WCBA. Within this population, pregnant women may face a unique and heightened vulnerability that warrants particular clinical attention. Pregnancy-induced immune modulation, particularly the suppression of T-helper cell (Th1) pro-inflammatory responses essential for fetal tolerance, can mask TB symptoms, increase susceptibility to new infections, and trigger TB reactivation during postpartum immune reconstitution.27,28 These conditions markedly elevate the risk of mother-to-child transmission.27,28 Pregnant individuals with TB carry a markedly greater risk of adverse maternal and fetal outcomes relative to those without TB.29 These clinical observations provide context for, but should not be inferred directly from, our population-level WCBA estimates. Therefore, targeted latent TB infection (LTBI) screening and monitoring are strongly recommended for high-risk groups, including pregnant women, particularly in TB high burden settings or when combined with additional risk factors for progression to ATB (eg, HIV infection) Close follow-up after diagnosis is also essential.30 Additionally, novel rapid molecular diagnostics with drug susceptibility testing (DST) enable early and accurate TB diagnosis. The prompt initiation of appropriate first-line anti-TB treatment during pregnancy is crucial for effectively preventing severe maternal and perinatal complications.31 Furthermore, comprehensive family planning services are vital for improving pregnancy outcomes among WCBA with TB. These services should include counseling on the risks of pregnancy during TB treatment, provision of safe and preferred contraceptive methods, and appropriate care pathways for women who become pregnant during treatment.32 Our findings showed a strong inverse association between the ASRs of TB and its three subtypes among WCBA and SDI. This demonstrates that countries with lower socioeconomic development bear a higher TB burden, which is fundamentally driven by economic barriers that restrict access to essential healthcare and early TB diagnosis and treatment. Furthermore, low- and middle-income countries (LMICs) face compounding and economically-rooted challenges that exacerbate the TB epidemic among WCBA: high HIV co-infection rates often linked to healthcare system weaknesses, insufficient public health funding and education, widespread malnutrition, overcrowded living conditions associated with poverty, and under-resourced TB control programs with poor policy implementation.33,34 These factors sustain TB as a persistent public health threat, particularly in high-burden regions such as Sub-Saharan Africa, South Asia, and Southeast Asia. Based on the pronounced regional disparities in TB burden, there is a need to tailor context-specific health policies to enhance early detection and targeted management within each country, and address the underlying socioeconomic determinants through concerted efforts to combat poverty and inequality, improve governance standards, foster inclusive growth and equitable distribution, and expand effective health education.35 The growing burden of DR-TB among WCBA poses a major challenge to global TB control. Key drivers include delayed or inaccurate diagnosis and DST, improper anti-TB drug use (incorrect regimens, inadequate dosing, and poor adherence), long treatment duration, comorbidities (eg, HIV, chronic obstructive pulmonary disease [COPD], and diabetes mellitus), and ongoing transmission of resistant strains.36,37 While antibiotic combination therapy aims to prevent resistance, this approach increases adverse drug reaction risks, potentially contributing to treatment failure.38 Currently, DR-TB remains a global challenge, which is exacerbated in resource-limited settings (eg, South Asia and Sub-Saharan Africa) by inadequate diagnostics, limited access to effective regimens, and poor treatment outcomes.39 Addressing this challenge requires a multi-pronged approach. Firstly, rapid molecular DST (eg, Xpert MTB/RIF Ultra) should be expanded to guide first-line therapy and reduce empirical errors.40 Secondly, shorter and all-oral regimens (eg, bedaquiline, pretomanid, linezolid, and moxifloxacin [BPaLM] for MDR-TB) can improve adherence in resource-limited settings.41 Thirdly, patient-centered support through integrating digital adherence tools (eg, video-observed therapy) and community health workers is essential for high-risk groups such as pregnant women.42 Finally, strict airborne precautions in maternal healthcare facilities are needed to prevent nosocomial transmission of resistant strains. Despite a relatively low overall TB burden among WCBA in Eastern Europe and Central Asia, the incidence of DR-TB, particularly XDR-TB, remains substantially higher than that in other GBD regions. Previous TB treatment failure has been identified as an important risk factor for DR-TB in Europe.43,44 Historical factors may have also contributed to the persistence of DR-TB in these regions. The disruption of TB control systems during the health-system transition following the dissolution of the Soviet Union, together with treatment-related factors, may have contributed to delayed diagnosis, treatment interruption, and the emergence of drug-resistant strains.45 These historical challenges have been further compounded by persistent healthcare system limitations, including uneven resource distribution and suboptimal access to TB diagnosis and treatment services.43 Additionally, migration from TB-endemic areas and rising HIV co-infection rates, which are closely linked to inadequate drug control, are key drivers of DR-TB transmission in these two regions.46,47 Therefore, addressing these risk factors calls for targeted strategies: implementing cross-border surveillance by establishing binational TB registries to track migrant populations and ensure continuity of care; integrating HIV-TB services by scaling up “test-and-treat” programs that combine antiretroviral therapy (ART) with TB preventive treatment for WCBA living with HIV; and prioritizing targeted screening for LTBI using interferon-gamma release assays (IGRAs) among high-risk WCBA, including pregnant women, in high-HIV-burden settings.48 Globally, adolescents and young women (15–24 years) experienced the highest TB incidence in our analysis. This higher burden may be associated with HIV co-infection, which may be influenced by social vulnerabilities such as early pregnancy, sexual violence, injecting drug use, and limited healthcare access in high-burden regions.49 Middle-aged women (40+ years) experienced rising TB risk due to immune senescence, caregiving burdens, and comorbidities, particularly in resource-limited settings. Notably, women aged 35–39 showed the highest incidence of XDR-TB in high-middle SDI regions, underscoring unmet needs in early screening and risk factor management in this age group. These results call for the development of age-specific interventions: prioritizing age-stratified TB/HIV screening for young women, addressing social determinants (eg, healthcare access for older WCBA), and tailoring DR-TB prevention to high-risk age groups Consistent with these approaches, international expert recommendations have emphasized the need for adolescent- and young adult-specific TB services, including age-disaggregated surveillance, targeted case finding, and prevention interventions for individuals aged 10–24 years. Implementing such tailored approaches could help reduce the TB burden among WCBA. High BMI has surpassed alcohol use and smoking as the leading contributor to TB-related DALYs among WCBA, marking a major shift from 1990 to 2021. This trend parallels the global increase in obesity,50 which is fueled by urbanization, sedentary lifestyles, and dietary shifts.51 The link between high BMI and increased TB burden presents a complex paradox. Potential explanations include obesity-related immune dysfunction such as intestinal flora imbalance and abnormal cytokine responses (eg, over-expression of interferon-γ [IFN-γ]),52,53 as well as the compounding effects of diabetes, which is both a consequence of obesity and a known amplifier of TB susceptibility and severity.50,54 This risk is particularly acute in regions like Sub-Saharan Africa and Asia, where the dual burden of undernutrition and obesity persists and may compound TB vulnerability.55 Pregnancy further complicates this relationship, as metabolic and immune changes may negate theoretical protective effects of high BMI.56 To address this, integrating metabolic screening (eg, for gestational diabetes and obesity) into prenatal care, promoting community-based nutrition programs, and prioritizing research on immune-metabolic interactions could help mitigate the growing TB burden linked to high BMI. Research shows that smoking and alcohol use increase TB susceptibility and promote LTBI reactivation by impairing immune function and lung tissue.57,58 While tobacco control measures and improved health awareness have reduced the impact of smoking on TB among WCBA, female smoking rates remain largely unchanged or have even increased in many countries.59 Notably, women’s smoking rates are much higher than the global average in Europe and Oceania, yet some developed countries in these regions report relatively lower TB burdens linked to smoking.60 Alcohol use persists as the leading risk factor in high-endemic regions like Sub-Saharan Africa, which may be associated with prevalent drinking patterns.61 These findings underscore the necessity of targeted interventions addressing regional risk factor profiles to strengthen TB prevention. The projections in this study are based on pre-2022 data and may not fully account for recent post-pandemic shifts in TB epidemiology. Nonetheless, the overall trends identified, particularly the divergent trajectories of DS-TB and DR-TB, remain informative for guiding long-term strategic planning. Consistent with these trends, notable advancements have been made in TB prevention and treatment among WCBA globally, and the overall decline in incidence is expected to continue through 2030, though the pace may vary by region. However, both the Philippines and the Democratic Republic of the Congo show persistent increases in case numbers, presenting major challenges for disease management. This situation necessitates focused attention on high-risk groups such as pregnant women, along with precise identification of TB risk factors to develop targeted interventions in these countries. India’s annual TB incidence among WCBA remains substantially higher than other countries, indicating a persistent challenge in disease control. This is closely associated with India’s large population base and nutritional deficiencies, particularly affecting adolescent girls, pregnant and lactating women, and children.62 Furthermore, the general stigma associated with TB in India’s WCBA further exacerbates the disease burden: unmarried women often conceal their condition due to concerns about marriage prospects and social status, while married women may face family isolation or discrimination.63 These factors contribute to their delayed medical care, postponed diagnosis, and poor treatment adherence, ultimately fueling TB transmission. This study aimed to characterize the global patterns, long-term trends, and spatial disparities in incidence and DALYs of TB among WCBA using GBD 2021 data, filling a critical evidence gap. However, several limitations should be noted. First, TB burden may be underestimated in settings with limited surveillance systems, particularly in LMICs. Second, subnational disparities in high-burden countries (eg, China, India) could not be fully characterized due to limited granular data Third, the lack of pregnancy- and postpartum-specific TB data in GBD 2021 precluded stratified analysis of TB burden in this high-risk population. Fourth, although HIV infection is an important risk factor for TB, it was not included in our comparative risk factor analysis because HIV-attributable risk estimates were unavailable within the GBD 2021 risk factor framework. Fifth, this study used GBD 2021 data to project trends to 2030, yet actual data for 2022 and 2023 are now available. These projections may not fully capture post-pandemic changes in TB epidemiology, which have been driven by factors such as healthcare service recovery, improved diagnostic capacity, and adjusted prevention strategies. Sixth, XDR-TB burden estimates are particularly sensitive to changes in laboratory testing capacity, evolving case definitions, and surveillance system strengthening over time. As the analysis period (1991–2021) coincides with substantial global expansion of drug susceptibility testing, the observed upward trends may partly reflect improved detection rather than true increases in transmission. Future studies incorporating the latest GBD data, pregnancy-stratified records, and sensitivity analyses for XDR-TB surveillance bias are needed to validate and update these findings. Conclusion Although the global burden of TB among WCBA has declined over the past three decades, the increasing burden of MDR-TB and XDR-TB remains a major challenge for global TB control. The substantial heterogeneity across regions and age groups highlights the importance of tailored strategies for populations at higher risk, particularly in low-SDI settings and areas with a high burden of DR-TB. Strengthening early diagnosis and rapid drug susceptibility testing, improving access to effective DR-TB treatment, and integrating TB prevention and screening into reproductive healthcare services may help further reduce the TB burden among WCBA worldwide. Future efforts should also prioritize the collection of high-quality TB surveillance data among pregnant and postpartum women, as these population-specific data remain insufficiently captured in current global burden estimates and many national surveillance systems. Acknowledgments We appreciate the works by GBD 2021 collaborators. We also thank Ju Zou for her constructive opinions and technical guidance. Funding Statement This work was supported by grants from the National Natural Sciences Foundation of China (82070613 to RC and 82370638 to RC), Key Project of Teaching Reform Research for Regular Undergraduate Universities in Hunan Province (202502000155 to R.C.), The funders of this study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. Declaration of Generative AI and AI-Assisted Technologies During the preparation of this work, the authors used DeepSeek-V3.2 in order to assist with translation and sentence refinement, improve language fluency and grammatical accuracy. After using this tool, the authors reviewed, edited and re-wrote the content as needed and take full responsibility for the content of the published article. Abbreviations ASIR, age-standardized incidence rate; ASDR, age-standardized DALY rate; ASRs, age-standardized rates; AAPC, average annual percentage change; ATB, active TB; ART, antiretroviral therapy; BAPC, Bayesian age-period-cohort; BMI, body-mass index; BCG, Bacillus Calmette-Guerin; BPaLM, bedaquiline, pretomanid, linezolid, and moxifloxacin; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DALYs, disability-adjusted life years; DS-TB, drug-susceptible TB; DR-TB, drug-resistant TB; DST, drug susceptibility testing; FPG, fasting plasma glucose; GBD, Global Burden of Disease study; HIV, human immunodeficiency virus; IFN-γ, interferon-γ; IGRAs, interferon-gamma release assays; INLA, integrated nested Laplace approximations; LMICs, low- and middle-income countries; LTBI, latent TB infection; MDR-TB, multidrug-resistant TB; SDI, socio-demographic index; TB, tuberculosis; UI, uncertainty interval; WCBA, women of childbearing age; WHO, World Health Organization; XDR-TB, extensively drug-resistant TB. Data Sharing Statement The datasets analyzed during the current study are derived from GBD 2021, available from the Institute for Health Metrics and Evaluation (IHME) at http://ghdx.healthdata.org/gbd-2021. However, data specifically generated or re-analyzed to support the findings of this study are included in this published article and its supplementary information files. Ethics Approval and Informed Consent This study utilized exclusively de-identified and publicly available data from the Global Burden of Disease (GBD) 2021 database. The study involved secondary analysis of aggregated data without access to individual-level identifiable information or interaction with human participants. Based on these characteristics, this study met the exemption criteria described in Article 32 (Items 1 and 2) of the Measures for Ethical Review of Life Sciences and Medical Research Involving Human Subjects (2023), China, which allow exemption from ethical review for research involving legally obtained publicly available data or anonymized information data without involving identifiable personal information. Therefore, institutional ethical approval and individual informed consent were not required. Consent for publication was not applicable. The use of GBD data complied with the database access and use policies. Consent for Publication As this study involves only the analysis of de-identified, publicly available data with no identifiable individual information (such as images, videos, or personal medical records), consent for publication is not applicable. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure All authors declare no competing interests. References Associated Data Data Availability Statement The datasets analyzed during the current study are derived from GBD 2021, available from the Institute for Health Metrics and Evaluation (IHME) at http://ghdx.healthdata.org/gbd-2021. However, data specifically generated or re-analyzed to support the findings of this study are included in this published article and its supplementary information files.