Nutrition Status and Comorbidities Are Important Factors Associated With Mortality During Anti-Tuberculosis Treatment

Tuberculosis (TB) was the leading cause of death from infectious diseases globally, prior to the coronavirus disease 2019 (COVID-19) pandemic. Unfortunately, COVID-19 has impeded global efforts to eliminate TB,1 causing substantial decline in the number of newly diagnosed TB cases. Between 2019 and 2020, an 18% reduction was observed in the number of newly diagnosed TB cases, from 7.1 million to 5.8 million, indicating underdiagnosed and untreated TB cases because of the pandemic.2 This setback has undone years of progress in reducing TB mortality, with the total number of deaths in 2020 reaching 2017 levels. Mortality, in addition to the incidence and prevalence, is a crucial indicator for assessing the burden of TB from a public health perspective. The primary objective of anti-TB treatment is to prevent TB mortality. Therefore, understanding and assessing the factors associated with TB mortality is essential to achieve the goals outlined in the World Health Organization (WHO) End TB strategy.

According to the global TB report of the WHO, TB incidence among high-income countries was 9.0 per 100,000 individuals in 2022. The Republic of Korea continued to bear a disproportionately higher burden of TB than other high-income countries, with a TB notification rate of 39.8 per 100,000 individuals in 2022,3 which is approximately four times higher than that in high-income countries. The elevated TB incidence and mortality rates among older individuals present significant challenges in controlling TB endemics in Korea.4 Older patients with TB often experience unfavorable treatment outcomes because of coexisting chronic diseases. Undernutrition, a well-known risk factor of TB, is particularly prevalent among older individuals. Studies have revealed that concomitant chronic disease5, 6 and undernutrition7, 8 adversely affect the prognosis of anti-TB treatments, including mortality. As the global population ages, TB in the geriatric population has become a significant challenge to global TB control. Analyzing the unique TB epidemiology in Korea could serve as a case study for developing strategies to reduce TB mortality in countries with aging populations.

In this study, we presented the clinical characteristics of TB mortality cases and analyzed the factors associated with mortality, focusing on nutritional status and comorbidities among patients with pulmonary TB in Korea, a setting characterized by an intermediate TB burden and a high-income status.

Between 2019 and 2021, 1,119 patients with pulmonary TB were registered in the cohort database (Fig. 1). Forty-eight participants with rifampicin-resistant TB or a history of multi-drug-resistant TB treatment were excluded. Among the 1,071 participants with rifampicin-susceptible pulmonary TB, 59 (5.5%) died. A total of 858 participants were included in this study after excluding those with other treatment outcomes. The survival and mortality groups had 799 and 59 participants, respectively; their clinical characteristics are presented in Table 1. The mean age was higher in the mortality group than that in the survival group (74.3 vs. 60.4, P < 0.001). The proportion of positive sputum smear test results was higher in the mortality group than in the survival group (39.0% vs. 20.8%, P = 0.001). Patients in the mortality group experienced alarming symptoms more frequently than those in the survival group (47.5% vs. 27.8%, P = 0.001). The four components of the NRS, comprising BMI (20.4 ± 3.3 vs. 22.0 ± 3.3 kg/m²,P = 0.010), serum albumin levels (3.2 ± 0.7 vs. 4.0 ± 0.6 g/dL, P < 0.001), total cholesterol (133.3 ± 40.3 vs. 161.3 ± 41.7 mg/dL, P < 0.001), and serum lymphocyte count (12.5 ± 9.5% vs. 23.6 ± 10.0%, P < 0.001) were significantly lower in the mortality group than those in the survival group. Approximately one-third of participants in the mortality group had an NRS of ≥ 1. Individuals in the mortality group had a higher NRS (P < 0.001) and CCI scores (P = 0.004) than those in the survival group.

Fig. 1
Flow chart of participant enrollment.
TB = tuberculosis, MDR/RR-TB = multidrug-resistant/rifampicin-resistant tuberculosis.

• Click for larger image
• Download as PowerPoint slide

Table 1
Baseline characteristics of enrolled pulmonary tuberculosis patients with and without mortality during anti-tuberculosis treatment

• Click for larger image
• Click for full table
• Download as Excel file

The results of the univariable and multivariable logistic regression analyses of the variables associated with mortality are presented in Table 2. Among the variables, age ≥ 65 years (odds ratio [OR], 6.23; 95% confidence interval [CI], 3.58–10.86; P < 0.001), positive smear result (OR, 2.44; 95% CI, 1.41–4.22; P = 0.002), alarming symptoms (OR, 2.35; 95% CI, 1.38–4.00; P = 0.002), NRS (OR, 3.04; 95% CI, 2.24–4.13; P < 0.001), CCI score (OR, 1.67; 95% CI, 1.47–1.90; P < 0.001), and HREZ as the initial treatment regimen (OR, 2.42; 95% CI, 1.32–4.43; P = 0.004) were selected and included in the multivariable logistic regression. NRS (adjusted OR, 2.63; 95% CI, 1.76–3.71; P < 0.001) and CCI score (adjusted OR, 1.64; 95% CI, 1.40–1.92; P < 0.001) remained statistically significant in the multivariable analysis.

Table 2
Factors associated with mortality during anti-tuberculosis treatment among patients with pulmonary tuberculosis

• Click for larger image
• Click for full table
• Download as Excel file

Age-based subgroup analysis was conducted, categorizing the patients into two groups based on whether they were < 65 years (Table 3). Among patients aged < 65 years, alarming symptoms (adjusted OR, 3.86; 95% CI, 1.00–14.89; P = 0.049), NRS (adjusted OR, 2.81; 95% CI, 1.38–5.69; P = 0.004), and CCI score (adjusted OR, 2.13; 95% CI, 1.45–3.13; P < 0.001) remained statistically significant in the multivariate analysis. For patients aged ≥ 65 years, NRS (adjusted OR, 2.21; 95% CI, 1.46–3.33; P < 0.001) and CCI score (adjusted OR, 1.47; 95% CI, 1.21–1.78; P < 0.001) also remained statistically significant in the multivariable analysis.

Table 3
Age-based subgroup analysis investigating factors associated with mortality during anti-tuberculosis treatment among patients with pulmonary tuberculosis

• Click for larger image
• Click for full table
• Download as Excel file

In this study, we analyzed the characteristics of patients who died during anti-TB treatment and identified factors associated with mortality. To objectively assess the nutritional status and comorbidities of the participants, we adopted previously validated tools: the NRS and CCI score. Our findings indicated that for each incremental increase in the NRS and CCI score, the odds of mortality were approximately 2.44 and 1.62 times higher, respectively. This trend was consistent with the results of age-stratified subgroup analysis. Accurately assessing the nutritional status and comorbidities of patients at the time of TB diagnosis and identifying reversible factors, such as nutritional status, body weight, and total cholesterol levels, are important to prevent TB mortality.

Risk factors for TB deaths vary according to human immunodeficiency virus (HIV) prevalence and regional conditions. In areas with high HIV rates, factors such as HIV positivity, advanced immunosuppression, and malnutrition increase the risk of TB mortality.11 In regions with a low HIV prevalence, older age, male sex, comorbidities, and substance abuse were significant risk factors.11, 12 Recently, Jeong et al.6 assessed the factors associated with mortality among patients with TB registered in the public-private mix TB control project in Korea. They found that older age, lower BMI, dyspnea, general weakness, and bilateral lesions were common risk factors for both TB- and non-TB-related deaths. TB-related deaths were linked to fever and smear positivity, whereas non-TB deaths were associated with the male sex and comorbidities such as heart or liver disease. Risk factors for TB-related mortality may vary depending on TB epidemiology, population structure, socioeconomic status, and healthcare system. Tailored public health strategies based on regional data are essential to reduce TB mortality.

Our study revealed that NRS had the highest odds of mortality, indicating that undernutrition plays a crucial role. However, assessment of the nutritional status of individuals with TB is not well established. Notably, BMI is a useful tool for evaluating nutritional status. Kim et al.10 developed and validated the NRS, demonstrating that nutritional deficits that are assessed using the NRS are associated with poor outcomes in patients with miliary TB. A comprehensive assessment, incorporating simple physical examinations and laboratory tests, should be emphasized at the time of TB diagnosis. Further research is needed to formulate nutritional supplements based on the initial nutritional assessment. Recent clinical trials have indicated that nutritional support is associated with a substantial reduction in TB incidence among household contacts.13 This concept may also apply to individuals with active TB.

In 2021, the proportion of patients with TB aged ≥ 65 years among all notified cases reached 51.1%, signifying an inevitable increase in TB burden among older individuals in an aging population in Korea.4 Older patients are often frail and have significant comorbidities that contribute to adverse drug reactions and poor medication adherence during anti-TB treatment. The effect of comorbidities on mortality is likely attributable to non-TB-related causes of death. A recent nationwide Korean study revealed that comorbidities, particularly heart, liver, and kidney diseases, were associated with non-TB-related death.6 Lin et al.14 found that a significant proportion of deaths were attributed to underlying comorbidities, particularly liver cirrhosis. Patients with TB and chronic liver disease have a higher risk of drug-induced liver injury.15 Elevation of hepatic enzymes is one of the most common adverse drug reactions during anti-TB treatment with first-line drugs.16, 17 The synergistic effect of aging and underlying liver disease may increase the likelihood of adverse drug reactions, such as drug-induced liver injury, posing a potential threat to treatment outcomes. Metabolic disorders are associated with drug-induced liver injury during anti-TB treatment.18 Untreated and complicated diabetes increases the risk of unfavorable outcome and mortality.19 Understanding the impact of comorbidities on TB and its treatment outcomes can assist physicians in preparing proactive interventions.

Despite a notable proportion of TB-related deaths among older patients in Korea,20 age did not emerge as a significant factor associated with mortality in the multivariate analysis conducted in the current study. Initially, we hypothesized that the effects of nutritional status and comorbidities would vary with age, prompting further subgroup analyses. The impact on mortality was similar between older patients aged ≥ 65 years and those aged < 65 years. A recent cluster analysis21 revealed that older patient comorbidities and who were underweight were more likely to exhibit dyspnea, positive smear results, and bilateral disease on chest radiography among individuals with pulmonary TB, regardless of sex. These clinical characteristics have been recognized as factors associated with mortality. For patients aged < 65 years, alarming symptoms indicative of severe lung parenchymal destruction and heightened disease severity emerged as significant factors associated with mortality in the subgroup analysis. An individualized approach is warranted for younger patients with alarming symptoms to strengthen patient-centered care and improve survival. However, the presence of alarming symptoms was not significant in patients aged ≥ 65 years. This might be ascribed to chronic symptoms similar to those of TB in older patients with multi-morbidities. This indicates that, in older patients, TB severity cannot be determined based on symptoms alone. In addition, distinguishing the causes of symptoms when treating older patients is crucial to prevent delays in TB diagnosis.

TB mortality is a crucial indicator for national TB control programs aligned with the ‘End TB Strategy’ of the WHO, which aims to reduce the number of TB-related deaths by 95% between 2015 and 2035. Despite being a preventable and curable disease, TB was the world’s second leading cause of death from a single infectious agent after COVID-19 in 2022. Globally, an estimated 1.3 million deaths were attributed to TB in 2022.22 Pandemic-related disruptions are projected to have led to nearly half a million excess TB deaths between 2020 and 2022, compared to the anticipated numbers. The global trend showed a decline in the absolute number of TB deaths until 2019, followed by an increase between 2020 and 2021 and then a reversal in 2022. To address potential delays in TB diagnosis and treatment due to the impact of COVID-19, the Korea Disease Control and Prevention Agency plans to strengthen public-private partnerships for TB care and prevention in Korea.23 The agency aims to provide systematic management for patients with TB from diagnosis to completion of anti-TB treatment. In the third National Strategic Plan for TB control announced in 2023, efforts will be intensified across government agencies to reduce mortality rates. This involves improving treatment accessibility and success rates, particularly for patients aged ≥ 65 years and those facing economic vulnerabilities, by integrating various health programs. Our study highlights that controlling comorbidities and improving nutritional status are important for reducing mortality in patients with TB and that public health programs should address these factors through a multidisciplinary approach. For example, we recently found that untreated and complicated diabetes at the time of TB diagnosis increases the risk of TB mortality.19 Policymakers must revise medical policies to screen for diabetes status and complications at baseline and incorporate proactive management.

The strength of our study is that data were extracted from a nationwide multi-center prospective cohort study. However, this study has several limitations. First, a convenience sampling strategy was employed for participant enrollment, which, owing to its non-probabilistic nature, limited the generalizability of the study. Second, there may have been selection bias, as only patients who provided informed consent were enrolled. Individuals with a severe clinical presentation who were unable to provide informed consent were excluded from the study. Third, because our study was conducted from 2019 to 2021, which was during the COVID-19 pandemic, there may be confounding bias due to COVID-19. Fourth, although other well-known covariables, such as positive sputum smear results, presence of alarming symptoms, and initial treatment regimen, were not statistically significant, this does not mean that these were not important factors associated with TB mortality. This could be ascribed to the small sample size of our study. Notably, the NRS and CCI score remained statistically significant despite the small sample size. Fifth, specific comorbidities were not included in this analysis. If each comorbidity was added as a variable in the current statistical analysis, the number of variables required to minimize bias would be exceeded. Therefore, we attempted to avoid this problem. Additional research is needed to determine how chronic diseases, such as hypertension or cardiovascular disease, affect TB treatment prognosis.

In conclusion, our findings highlight a significant association between nutritional status and comorbidities at the time of TB diagnosis and mortality during anti-TB treatment. There is a compelling need for a comprehensive assessment of patients with pulmonary TB to identify and address the reversible factors that ultimately enhance survival of patients with TB.