The latent tuberculosis infection survey using two interferon γ release assay tests among the elderly in a well-confined rural county in Eastern China

Abstract

Background

Tuberculosis (TB) latent infection is a crucial precursor to the development of active pulmonary tuberculosis while currently no gold standard for detecting latent infection. This study aims to investigate the prevalence of latent infection among the elderly in high-incidence areas in Eastern China. Additionally, it compares the agreement of two interferon-γ release assay (IGRA) reagents in detecting latent TB infection in this elderly population.

Methods

A random sampling method was adopted to select two townships from Lanxi City, China. Based on the health check-ups for the elderly under the National Basic Public Health Service Project, a questionnaire survey and latent infection testing were conducted among the recruited elderly population aged 65 and above. A multi-factor analysis was used to identify the influencing factors for latent infections in the elderly population. Meanwhile, the optimal cutoff value was determined using the ROC curve.

Results

This study enrolled a total of 1,583 elderly individuals aged 65 years and above for latent tuberculosis infection (LTBI) testing. Among them, 424 tested positive with an infection rate of 26.78%. Multivariable analysis revealed that smoking (Adjusted OR = 1.810, 95% CI 1.315–2.491) and engaging in physical exercise at least once per week (Adjusted OR = 2.179, 95% CI 1.021–4.051) were significantly positively associated with LTBI. The comparative study of latent infection using two different reagents was conducted among 108 elderly participants, showing an extremely high level of agreement with a Kappa value of 0.812, a correlation coefficient (R²) of 0.713 (P < 0.001).Using the traditional IGRA as a reference, we found that in the elderly population, increasing the cutoff value of the domestic AIMTB reagent to 24.02 could achieve the optimal cutoff value with the area under the ROC curve 0.952, with a sensitivity of 83.3% and a specificity of 97.4%.

Conclusion

The prevalence of LTBI among individuals aged 65 and above in the relatively confined rural areas of Zhejiang Province was as high as 26.78%, which required focused attention with the intensification of population aging. Male sex, smoking duration, and exercise frequency were all associated with LTBI, and the underlying mechanisms needed to be explored in the future. The domestic reagent AIMTB of China showed good agreement with IGRA reagents in old adults, which offered a model for tailoring cutoff thresholds to specific populations, helping reduce LTBI screening costs in low- and middle-income countries. Besides, the need to appropriately raise diagnostic standards should be further verified through large-scale population studies in the future.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-025-06739-2.

Introduction

Tuberculosis (TB) is a chronic infectious disease caused by the infection of Mycobacterium tuberculosis (MTB). Among them, pulmonary tuberculosis (PTB) is the most common type, accounting for more than 80% of all types of TB [1]. This condition, characterized by pulmonary infection due to the invasion of MTB, has imposed a significant burden on low- and middle-income countries [2]. According to the Global Tuberculosis Report 2024, an estimated 10.8 million new cases of TB were reported globally in 2023, marking an increase compared to 2022 [2, 3]. China remains one of the high TB burden countries, with an estimated 741,000 new cases in 2023, ranking third globally [3]. Globally, about 25% of people are infected with MTB but show no TB symptoms or active disease evidence, qualifying as latent tuberculosis infection (LTBI) [1]. Previous research indicated that about 5% of healthy adults infected with MTB would rapidly progress to active TB within two years. Although this risk diminishes over time, MTB could still reactivate after a latent period of several years to decades [4, 5]. Additionally, studies have shown that in rural areas, the burden of LTBI is significantly higher among the elderly than among younger individuals [6]. Therefore, early identification and timely intervention of LTBI in the elderly are crucial for controlling local TB outbreaks.

Although various detection methods for LTBI are currently available, the lack of a definitive gold standard for diagnosis continues to pose a significant challenge. The most common methods are the tuberculin skin test (TST), the recombinant mycobacterium tuberculosis fusion protein (EC), and interferon-γ release assay (IGRA) [7, 8]. Both TST and EC relied on the principle of utilizing the delayed-type hypersensitivity (DTH) reaction elicited by MTB antigens for indirect detection [9, 10]. However, the TST was limited by the interference from Bacillus Calmette-Guérin vaccination, which was prevalent in high-incidence countries like China [11]. Meanwhile, EC also had limitations, as it was not suitable for children under six months of age and older adults over 65 years old [9]. For both, a 48- to 72-hour interval was required to measure the size of the induration or erythema, which could lead to follow-up loss, especially among populations with poor follow-up rates, such as the elderly. The IGRA stimulated specific T-cell immune responses with MTB antigens in vitro, leading to the secretion of interferon-γ. The presence of LTBI was determined by quantitatively detecting the level of released interferon-γ or counting the number of responsive T cells. This specific method had demonstrated high sensitivity and specificity across the entire population [12]. Therefore, in the screening of latent infections in the elderly, particularly in high-incidence areas, IGRA were prioritized over TST and EC.

Currently, two internationally recognized and mature industrialized reagents are widely recommended for identifying interferon-γ. The first is the QuantiFERON-TB Gold In-Tube test (QFT-IT), developed by Cellestis in Australia, which employed enzyme-linked immunosorbent assay (ELISA) to directly measure the concentration of interferon-γ in whole blood. The other was the T-cell spot test for TB (T-SPOT.TB), produced by Oxford Immunotec in the UK utilizing enzyme-linked immunospot assay (ELISPOT) to quantify the number of spots representing MTB-specific effector T cells [13]. Besides, in recent years, an increasing number of domestic reagents in China had been introduced into the field of LTBI screening such as AIMTB Rapid Test Assay [14, 15]. However, it was regrettable that agreement evaluations comparing these domestic reagents with internationally established standards, particularly in the context of elderly populations, was still limited.

The Lanxi City located in the middle of Zhejiang, China, presents a high-incidence for PTB among the elderly. In recent years, with the implementation of the “Tuberculosis-Free Community” project, active screening had been carried out in this specific population to reduce the prevalence of active PTB and identify the LTBI levels in the targeted population [16]. Thus, the aim of this study was to explore the influencing factors associated with LTBI in the elderly population, and evaluate the agreement between emerging domestic LTBI test and conventional detection methods, which would be helpful to provide scientific data and policy support for future large-scale latent infection population study in developing countries with high TB burden.

Methods

Location

This study was conducted in Lanxi City, which is situated in the central region of Zhejiang Province, in the Jinhua area. It has a subtropical monsoon climate, with an annual average temperature of 17.7℃ and precipitation of 1,439 millimeters. Besides, Lanxi City had a comparatively high risk of PTB (70/100,000–80/100,000), while the notification rate and prevalence of the elderly more than 60 had reached 161.6/100,000 and 233.1/100,000, respectively. The location details were described in our previous study [17].

Sample size calculation and implementation process

We used the cross - sectional design in this study, and its sample size was calculated using the formula from Daniel et al. [18]:

Based on previous study, p for LTBI rate was set as 34.9% in people greater than 65 years old, with Z value of 1.96 and an allowable error of 3% [17].The calculation yielded a required number of 970 elderly subjects for the investigation. Considering the cooperation level and the completeness of variable collection among the elderly population, the sample size was further increased by 30%. Therefore, it was planned to include no fewer than 1,261 individuals for LTBI.

In this study, a random sampling method was used to select two sub-districts from the 16 sub-districts of Lanxi City. Participants were recruited on a daily basis from the two selected sub-districts. All elderly individuals aged 65 and above who attended the health check-ups and provided informed consent on the day when the target number was reached were included in the study. Also, inclusion criteria contained that participants were required to be either local residents or permanent residents with a minimum of six months of continuous residence in the preceding year. Exclusion criteria included that participants were in the treatment period for TB, or had a history of TB, as well as those unwilling to participate in this study.

A total of 1,583 elderly individuals from Lubu subdistrict and Youbu subdistrict in Lanxi City were ultimately included in this study. Then, a questionnaire survey and LTBI testing were performed for all participants. The questionnaire included information on participants’ sex, age, body mass index (BMI), smoking status, alcohol consumption, exercise habits, and whether they had diabetes. The definition of some variables had been introduced in our previous study [19]. Additionally, smoking status included never smoker, prior smoker, and current smoker; prior smoker defined as individuals with a history of smoking who had completely quit; current smokers were defined as individuals who reported smoking any tobacco products within the past 30 days, with lifetime consumption of ≥ 100 cigarettes. Besides, exercise status was categorized by participation frequency (excluding sessions lasting less than 10 min) into four groups: Never (no exercise participation), Occasionally (less than once per week), At least once per week (1–6 times weekly), and Every day (daily participation). For the LTBI testing, two types of enzyme-linked immunosorbent assay IGRA reagents were randomly selected: QFT-Plus (Qiagen, Hilden, Germany), and the AIMTB Rapid Test Assay (Leide Biosciences, Guangdong, China). For people who agreed to participate in the agreement study, a second informed consent was required before proceeding with the LTBI testing using the other reagent. A total of 108 individuals agreed to participate in the subsequent research.

Detection methods and interpretation criteria

QFT-Plus Detection was performed by local CDC staff. The details were listed as follows: The venous whole blood was thoroughly mixed, and 1 mL was added to each of the following tubes: the negative control tube (N tube), the mitogen positive control tube (M tube), the TB1 antigen tube, and the TB2 antigen tube. Then, the blood culture tubes were shaken to ensure thorough mixing of the antigens with the blood. Subsequently, the tubes were incubated in a constant temperature incubator at 37 °C (± 0.5) for 16–24 h. After incubation, the samples were centrifuged to separate the supernatant. A one-step ELISA method was employed to detect the level of interferon-γ in the supernatant. The positive or negative results were interpreted according to the manufacturer’s instructions. Meanwhile, the AIMTB was carried out. The venous whole blood was thoroughly mixed, and 0.6 mL was aliquoted into each of the following tubes: the N tube, the TB-specific antigen tube (T tube), and the positive control tube (P tube). The blood culture tubes were shaken to ensure thorough mixing of the antigens with the blood. Subsequently, the tubes were incubated in a constant temperature incubator at 37 °C (± 0.5) for 20 ± 2 h. After incubation, the samples were centrifuged to separate the plasma. A one-step ELISA method was employed to detect the level of interferon-γ in the supernatant.

The positive or negative results were interpreted according to the manufacturer’s instructions. The detailed results were interpreted as follows: (1) The QFT-Plus Test: A positive result was defined as an N tube concentration of ≤ 8.0 IU/ml, with either the TB1-Nil or TB2-Nil concentration being ≥ 0.35 IU/ml and ≥ 25% of the N tube value. (2) the AIMTB Test: A positive result was determined when the T tube concentration minus the N tube concentration was ≥ 20 pg/mL, and the T tube concentration was ≥ 25% of the N tube value.

Statistical analysis

Descriptive statistics and all analyses were used to demonstrate participant characteristics. Categorical variables were presented using frequencies/percentages, and the Chi-square (χ²) test was used to detect the differences between subgroups. Binary logistic regression identified risk factors for latent infection, calculating ORs and 95% CIs, with P < 0.05 indicating significance. Kappa statistic was used to assess the agreement between kits: <0.00 (poor), 0.00–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial), ≥ 0.81 (almost perfect). Simple linear regression and Bland-Altman plots assessed group differences, with P < 0.05 indicating significance. Additionally, the positive agreement rate, the negative agreement rate, and overall agreement rate were used to evaluate the concordance between the AIMTB and QFT-Plus test. The receiver operating characteristic (ROC) curve were used to find cutoff value of the AIMTB test using the QFT-Plus as reference [20]. The optimal cutoff value was determined by the Youden index. Data processing, visualization, and analysis were performed using R 4.5.0, including the ggplot2 package (Wickham, 2016), the pROC package (Robin et al., 2011), and the forestplot package (Gordon& Lumley, 2022).

Results

Demographic characteristics of the participants

A total of 1,583 individuals aged over 65 years were enrolled in this study to investigate latent infection level. Among them, 424 individuals were identified as positive for latent infection, yielding a positive LTBI rate of 26.78%. The male-to-female ratio was 0.95:1. The majority of the participants were aged 70 years or older (approximately 70%). 55.15% individuals had the BMI within the range of 18.5–23.9. The proportions of participants who never drank alcohol and never smoked accounted for 69.17% and 93.49%, respectively. Additionally, 35.94% of the participants had diabetes. For detailed demographic characteristics, see Table 1.

Table 1.

Demographic characteristics of all participants for LTBI test

Characteristic	Number(%)	Prevalence of LTBI(%)
Total	1583(100.00)	26.78
Sex
Male	775(48.96)	32.52
Female	808(51.04)	21.29
Age(years)
65–69	471(29.75)	25.9
70–74	556(35.12)	26.44
75–79	369(23.31)	39.77
80–84	128(8.09)	23.44
≥ 85	59(3.37)	33.9
BMI (kg/m²)
<18.5	90(5.59)	24.44
18.5–23.9	873(55.15)	24.4
≥ 24	620(39.17)	30.48
Smoking status
Non-smoker	1264(79.85)	24.6
Former smoker	83(5.24)	31.53
Current smoker	236(14.91)	36.86
Drinking status
Never	1095(69.17)	25.66
Occasionally	133(8.4)	24.06
Frequently	209(13.2)	26.79
Every day	146(9.22)	37.67
Exercise status
Never	1480(93.49)	26.35
Occasionally	18(1.14)	27.78
At least once a week	29(1.83)	44.83
Every day	56(3.54)	28.57
Diabetes
Yes	569(35.94)	27.42
No	1014(64.06)	26.43

LTBI Latent tuberculosis infection, BMI Body mass index

The influencing factors for the LTBI using multivariable analysis

Further multivariable analysis revealed that male sex (Adjusted OR = 1.64, 95% CI 1.25–2.15), former smoking status (Adjusted OR = 1.42, 95% CI 1.01–2.01), and engaging in exercise at least once per week (Adjusted OR = 2.21, 95% CI 1.03–4.75) were positively associated with LTBI in the elderly. The results were shown in Fig. 1.

Fig. 1.

Multivariable analysis of factors influencing LTBI in the elderly

Agreement analysis of two reagents

A total of 108 participants aged 65 years and above were included in the agreement analysis study. The sex ratio was 1:1. The age group of 65–69 years accounted for the largest proportion (53.70%). The majority of participants had a BMI of 18.5–23.9 (57.41%). Most participants had never smoked (75.00%) and had never consumed alcohol (55.56%). The proportion of participants who did not engage in physical exercise was 93.51%. The prevalence of diabetes in this specific group was low, with 87.96% of participants not having diabetes. The overall agreement of two reagents was 0.812. More details were provided in Supplement 1.

In this study, AIMTB was evaluated against QFT-Plus as the reference. The results indicated that the positive agreement rate of AIMTB was 89.29% (95% CI 73.93%–96.91%), and the negative agreement rate was 93.75% (95% CI 87.67%–97.16%). The overall agreement rate was 92.59% (95% CI 86.78%–96.17%). Further quantitative analysis revealed a good correlation between the two assays in the same samples, with a correlation coefficient R² = 0.713 (P < 0.05). The mean difference in interferon-γ levels between the two systems was − 3.51 IU/ml, with 95% Limits of Agreement (LoA) ranging from − 16.64 to 9.63 IU/ml. Notably, 5.37% (16/298) of the data points fell outside the 95% LoA. The results were shown in Fig. 2(A)–(B).

Fig. 2.

Agreement analysis of two reagents. (A) Correlation analysis of IFN-γ values detected by QFT-Plus and AIMTB using scatter plot; (B) Bland-Altman plot for the agreement analysis of IFN-γ values detected by QFT-Plus and AIMTB; (C) The ROC curve of AIMTB using QFT-Plus as the reference

With QFT-Plus results as the reference, an ROC curve was constructed to identify the optimal cutoff value for IFN-γ levels detected by the AIMTB system in diagnosing latent infection. The results showed that when the cutoff value was set at 24.02 pg/ml, the AUC for AIMTB in diagnosing latent infection was 0.952, with a specificity of 97.4%. The corresponding ROC curve was displayed in Fig. 2 (C).

Discussion

With the intensification of population aging, China, as one of the 30 high-burden countries for TB, is confronted with the dual challenge of increasing the detection of PTB and reducing the activation of LTBI among the elderly population [3, 21, 22]. In the pilot area of Lanxi City, we previously implemented community-based chest radiography screening to enhance the detection of active PTB in the elderly. The study demonstrated that active case finding through chest radiography could significantly increase the detection rate of asymptomatic active tuberculosis in the elderly population and improve the equity of TB medical services for marginalized and vulnerable groups [19].This study aimed to explore the situation of LTBI in the region from the perspective of latent infection, identify the risk factors associated with LTBI in the elderly population, and assess the agreement of currently used LTBI diagnostic reagents in the area. It was important to note that this study would provide insights into establishing a unified standard for LTBI levels among the elderly population, not only in China but also globally, in the context of the diverse range of diagnostic reagents available on the market.

Previous studies have shown that in 2021, the screening incidence rate of PTB among individuals aged 60 and above in Lanxi City was as high as 233 per 10,000, making it a high-burden area for elderly TB in rural Zhejiang [19]. In this study, through a sampling survey, we found that in 2023, the rate of LTBI among individuals aged 65 and above in Lanxi City was 26.78%, while the reported PTB rates for the entire population and for individuals aged 65 and above in the region were 73.13 per 100,000 and 208.62 per 100,000, respectively. Meanwhile, in a survey conducted in eight villages of Jiangsu Province, the latent infection rate detected by TB-IGRA among old individuals ranged from 31.29% to 33.33%, with the reported PTB rates for the entire population and for the old people being 51.05 per 100,000 and 85.59 per 100,000, respectively [23]. In Deqing, a low-incidence area in Zhejiang Province, the LTBI rate using IGRA test of AIMTB among targeted elderly group was only 7.24%, with a reported incidence rate of 59.59 per 100,000 during the same period [14]. Additionally, using the B-SHADE model to assess the latent infection status of TB in China in 2013, we found that the infection rate among individuals aged 60 and above was 38.36%, with a 95% confidence interval of 31.25%−45.46%, while the national reported incidence rate of TB for individuals aged 60 and above was 142.72 per 100,000 [24]. Therefore, the prevalence of LTBI among the elderly population varies significantly across different regions in China. This is not only affected by the risk of active PTB in the region but might also be associated with other external factors such as environment and diet [25, 26]. Future studies needed to incorporate potential external variables to further explore the relationship between these factors and the LTBI rate. Moreover, as age increased, the immune status of the elderly deteriorated, and they often suffered from comorbid chronic diseases, used related medications, and might experience malnutrition. The complex interplay of these factors made the elderly population with LTBI more susceptible to developing active PTB [27]. Therefore, future precision prevention and control strategies should prioritize identifying high-risk target groups among the elderly population. Based on a comprehensive assessment of individual health status, standardized preventive anti-tuberculosis treatment or nutritional intervention should be provided to reduce the risk of transmission within families and communities caused by the activation of latent infections in this population [28–31].

This study identified several key risk factors for LTBI in the elderly individuals through multivariable analysis, including male sex, smoking, and engaging in exercise at least once per week. The association between male and a higher prevalence of LTBI might be attributed to the relatively more active social lives of men, which increased their exposure to the illness [32]. Interestingly, our study found a correlation between smoking cessation and LTBI. Upon further investigation of smoking duration, we discovered that individuals who had quit smoking had an average smoking history of 51.38 years, compared to 45.32 years for current smokers. This suggested that those who quit smoking had been exposed to tobacco for a longer period. We hypothesized that the long-term damage to the respiratory tract caused by smoking, causing the interference of structure and function of the respiratory tract, such as damaging the airway epithelial barrier, thereby increasing the risk of MTB infection in this group [33, 34]. Future research might consider replacing the smoking status variable with smoking duration to achieve more accurate assessment results.

Moreover, we found a correlation between engaging in exercise at least once per week and LTBI. This might be because, compared to elderly individuals who exercise daily, those who exercised only once a week might already be in a suboptimal health state, with certain deficiencies in their physical condition and immune function, making them more susceptible to PTB. Additionally, compared to those who did not exercise at all, individuals who exercised once a week, while having some physical activity, might still face increased exposure risks. Their exercise frequency and intensity might not be sufficient to maintain a robust immune function to counteract the risks associated with exposure. Therefore, future research needed to further explore the detailed mechanisms and critical nodes between exercise frequency, exercise intensity, and LTBI to better understand this phenomenon.

In this study, we compared the agreement of the domestic AIMTB reagent in China with the QFT-Plus reagent from Qiagen, Germany, using the latter as a reference. The results showed an extremely high agreement between the two reagents, with a Kappa value of 0.812 (P < 0.001). This finding was similar to a previous study conducted in Deqing involving 200 participants, which further confirmed the reliability of China’s domestic reagent in detecting LTBI in the elderly population [14]. Unlike simple correlation comparisons, our study also explored the optimal cutoff value for the domestic AIMTB reagent. The results indicated that raising the standard of the value of T tube (TB-specific antigen) concentration minus the N tube concentration from 20 to 24 could achieve higher agreement in the elderly group. Considering that this adjustment was related to the characteristics of TB strains in different regions and the immune status of the elderly population, further agreement evaluations through large-scale, multicenter cohort studies are still needed in the future.

Limitations

This study conducted latent infection screening in two streets of Lanxi City using a sampling method rather than in all streets, which might introduce some bias in representativeness. Additionally, although the sample size met the statistical requirements, the limited number of participants to some extent restricted the statistical power of variables, especially in agreement evaluation. Future studies should consider conducting large-scale agreement evaluation studies in multiple centers.

Conclusions

The prevalence of LTBI among individuals aged 65 and above in the relatively confined rural areas of Zhejiang Province was as high as 26.78%, which required focused attention with the intensification of population aging. Male sex, smoking duration, and exercise frequency were all associated with LTBI, and the underlying mechanisms needed to be explored in the future. The domestic reagent AIMTB of China showed good agreement with IGRA reagents in the old adults, which offered a model for tailoring cutoff thresholds to specific populations, helping reduce LTBI screening costs in low- and middle-income countries. Besides, the need to appropriately raise diagnostic standards should be further verified through large-scale population studies in the future.

Abbreviations

TB

Tuberculosis

PTB

Pulmonary tuberculosis

LTBI

Latent tuberculosis infection

IGRA

Interferon-γ release assay

MTB

Mycobacterium tuberculosis

TST

Tuberculin skin test

EC

Recombinant mycobacterium tuberculosis fusion protein

DTH

Delayed-type hypersensitivity

T-SPOT.TB

T-cell spot test for TB

QFT-IT

QuantiFERON-TB Gold In-Tube test

ELISA

Enzyme-linked immunosorbent assay

BMI

Body mass index

CI

Confidence interval

OR

Odds ratio

LoA

Limits of Agreement

AUC

Area Under the Curve

QFT-Plus

QuantiFERON^®-TB Gold Plus

Authors’ contributions

KL and JJ concepted and designed this study; ZY and YF acquired the data; BC, WW, GJ and LH involved in the analysis and interpretation of the data; ZY and XJ wrote the manuscript; WT, QC, MZ, and YZ revised it critically and finished the final approval; and all authors agree to be accountable for all aspects of the work.

Funding

This study was supported by National Key Research and Development Program of China (Grant No.2024YFC2311202), the “Pioneer” and “Leading Goose” R&D Program of Zhejiang (Project No.2025C01134), and Zhejiang Provincial Medical and Health Project (2025KY774).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study was approved by Zhejiang Provincial Center for Disease Control and Prevention (ZJCDC). All participants provided informed consent to finish the study, and this study was conducted in accordance with the Declaration of Helsinki and the legal requirements of China.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.