Tuberculosis (TB) is a severe communicable disease that is among the top 10 causes of death worldwide. According to the World Health Organization, ~10 million new cases and 1.3 million deaths were reported in 2020. Of all TB cases in 2020, the proportion of adult men, adult women, and children is 56, 33, and 11%, respectively (1). Despite achieving a microbiological cure, many survivors face post-TB sequelae, thus increasing the overall disease burden (2).

Taiwan became an aged society in 2018. There were 16,472 and 7,823 new TB cases in 2005 and 2020, respectively, and the incidence rate decreased from 72.5 to 33.2 per 100,000 population during this period. More than half of the new TB cases occur in the elderly population (age ≥ 65 years) since 2005. Besides, there were 460 TB-related deaths in 2020 and the cumulative reduction between 2005 and 2020 was 53.5 %. Since 2006, the coverage rate of directly observed treatment, short course has reached 100%, and ~70% of patients with bacteriologically-positive TB were treated successfully in 2018 (3). Between 2006 and 2013, hypertension (HTN), diabetes mellitus (DM), chronic obstructive pulmonary disease (COPD), cardiovascular disease, and cancer were identified as common comorbidities in patients with TB in Taiwan. The risk of death in patients with TB and any comorbidity was 2.12 times higher than that in those without comorbidities (4). The Taiwan Centers of Disease Control highlights preventive treatment of high-risk individuals and early diagnosis of TB, and the elimination of TB is expected to be completed gradually by 2035, echoing the World Health Organization End TB Strategy (1).

Hypothyroidism is characterized by decreased thyroid hormone production in the thyroid gland. It can be classified as primary (due to thyroid hormone deficiency) or secondary/tertiary (due to abnormal hypothalamic-pituitary function). The diagnosis of hypothyroidism depends heavily on laboratory testing because of variable non-specific clinical manifestations (5–7). Women are more likely to have hypothyroidism than men (8). More than 95% cases of hypothyroidism are of the primary type (9, 10). Overt hypothyroidism and subclinical hypothyroidism are two degrees of primary hypothyroidism, and the prevalence of these disorders ranges from 0.1 to 3% and 1.6 to 15% (5, 8, 11–15), respectively. Annually, ~2–5% of cases of subclinical hypothyroidism may progress to overt hypothyroidism (6). Although chronic autoimmune thyroiditis is the main cause of primary hypothyroidism (5), other possible causes include iatrogenic diseases, deficiency or excessive consumption of iodine, drugs, and infiltrative diseases (5, 16–18). The common mechanism of drug-induced hypothyroidism includes inhibition of thyroid hormone synthesis, decreased absorption of T4, increased T4 clearance, increased type 3 deiodination, destructive thyroiditis, suppression of TSH, and immune dysregulation. Common drugs can cause hypothyroidism including lithium, amiodarone, omeprazole, lansoprazole, interferon alfa, interleukin-2, tyrosine kinase inhibitors (sunitinib, sorafenib, imatinib, etc.), and checkpoint inhibitors (ipilimumab, pembrolizumab, nivolumab, etc.). Patients taking these drugs should monitor possible hypothyroid symptoms and serum TSH should be measured at least every 6 to 12 months (17).

Thyroid hormones modulate various immune system functions, including chemotaxis, phagocytosis, production of reactive oxygen species, and the release of cytokine release (19). Hypothyroidism may have a detrimental effect on the immune system and subsequently make patients vulnerable to infection. Recent studies have shown that thyroid hormone signaling plays a vital role in optimal immune response during Mycobacterium tuberculosis (Mtb) infection (20), which could be related to infection or drug-related hypothyroidism. Thyroid tuberculosis, which has a frequency of 0.1–0.4%, may cause hypothyroidism owing to extensive glandular destruction due to caseous necrosis (21). Infectious agents have been reported to trigger autoimmune thyroid diseases by possible mechanisms including molecular mimicry theory and bystander activation theory (22). Rifampin increases T4 clearance, possibly because of enhanced hepatic T4 metabolism and biliary excretion of iodothyronine conjugates. Vaidya (16) and Montanelli (17) reported that rifampin causes primary hypothyroidism and there were few case reports of rifampin-induced hypothyroidism, and most of them had underlying Hashimoto's Thyroiditis, and some of them without underlying thyroid disease. Prior systemic review and meta-analysis showed that ethionamide and para-aminosalicylic acid were the most frequently reported drugs associated with the occurrence of hypothyroidism (23) and these drugs can cause hypothyroidism by inhibiting thyroid hormone synthesis through a mechanism of iodine organification inhibition (24, 25).

These studies may partially explain the causal relationship between TB and hypothyroidism; however, evidence from longitudinal analysis is lacking. Therefore, we conducted a bidirectional, nationwide, population-based cohort study to investigate the association between TB and hypothyroidism.

To the best of our knowledge, this is the first longitudinal, nationwide, population-based cohort study to investigate the bidirectional relationship between hypothyroidism and TB. We found that patients with hypothyroidism have a 2.91 times higher risk of developing TB than those without hypothyroidism, and that patients with TB have a 2.01 times higher risk of developing hypothyroidism than that those without. Although the pathophysiological association between hypothyroidism and tuberculosis remains unclear, several hypotheses have been proposed.

In our data, hypothyroidism was predominated in the female (64.2%). But male with hypothyroidism had a higher risk of TB development, especially in the age ≥45 years. Besides, the higher urbanization level 1 and 2 also had an impact on TB development. We proposed some explanations for these interesting findings. First, previous studies revealed that male and old age were the risk factors of TB development (35, 36). Testosterone can impair macrophage activation and decrease the production of proinflammatory cytokines, which may increase the susceptibility to Mtb infection. In contrast, estrogen can enhance macrophage activation and induce proinflammatory cytokines, which provide protection against Mtb infection (37). Second, a modeling study in Taiwan showed that immune senescence played an important role in the age disparity between children and elders, which means patients with latent TB infection have an increased risk of developing TB as they age due to a weakening immune system with older age (38). Third, TB is a disease whose transmission is favored by the density of urban populations (39). Taiwan has seen rapid growth in living standards and nationwide coverage of high-quality, publicly funded healthcare services (28). Due to the patient preference, better service quality, increased geographic accessibility, better community awareness, and access to better diagnostics and treatment, the TB case notification rates are higher in urban areas than in rural areas (39). Our findings are compatible with the above explanations.

In Taiwan, more than half of the new TB cases occur in the elderly population (age ≥ 65 years) since 2005 (3) and it can be seen in our TB cohort. In general concept, hypothyroidism development is much more common in female than male, and the incidence rate increase with age (40). However, these findings were not seen in our study. We think there may be other possible causes, with drug-induced hypothyroidism being one of the most likely, and additional study is needed in the future. Besides, this study also showed that patients with hypothyroidism have an interaction effect with comorbidities including DM and COPD toward the risk of developing TB. Individual links between TB and these two comorbidities among the general population are well established among papers (41). This demonstrates a relationship between hypothyroidism and DM and COPD and how all these interrelations remain unexplored.

Previous studies have demonstrated that serum triiodothyronine (T3) and thyroxine (T4) modulate specific immune responses, including innate and adaptive immune responses, which change significantly during aging and in cases of hypothyroidism. The dysregulation of innate and adaptive immune responses may subsequently increase susceptibility to infection (42–44). A recent population-based cohort study showed that patients with hypothyroidism were associated with a risk of pneumonia. In addition, the use of TRT (>30 days) can attenuate the hypothyroidism-linked pneumonia risk (45). However, there was no evidence or mechanism to explain the TB-linked hypothyroidism. In a rabbit Mtb infection model, defective production of thyroid hormones increased susceptibility to Mtb infection (46). Furthermore, T3 and T4 restricted Mtb growth in human monocyte-derived macrophages through interleukin-1α production; therefore, young, healthy household contacts of patients with TB with a decreased production of T4 at baseline have an increased risk of developing active TB (20). Our findings strengthen the evidence that patients with hypothyroidism have an increased risk of developing TB. The non-TRT group showed a significantly higher risk of developing TB, especially pulmonary TB, than the non-hypothyroidism cohort. Additionally, we found that long-term TRT (>1 year) may decrease the risk of developing TB, which supports the findings of a previous study (20). Further prospective cohort studies to evaluate the association between TRT and the risk of TB are warranted.

In our study, patients with TB, pulmonary TB, or extra-pulmonary TB had a significantly higher risk of developing hypothyroidism than those without TB. However, the results of previous, similar studies (20, 47) have been controversial. Kleynhans investigated the changes in the immune system during treatment for TB in patients who were cured and in those in whom it failed. They found that T3 concentrations increase during TB infection, whereas T4 concentrations remain unchanged in the failed group. T4 concentrations were lower in the cured group after 6 months of TB treatment than in the failed group. T4 plays a critical role in Mtb infection and may be a potential biomarker to differentiate treatment outcomes (47). Recently, T3 and T4 concentrations were reported to be elevated in TB progressors (healthy household contacts developing TB during the follow-up period) after 6 months of TB treatment, although T4 concentrations decreased in treated TB progressors 3 years after TB treatment (20). Although previous studies have demonstrated an association between TB and thyroid hormone levels, our findings reinforce that TB is a risk factor for hypothyroidism.

Moreover, we observed that the TB and hypothyroidism cohorts have a higher risk for hypothyroidism and TB, respectively during a follow-up period of <5 years; the highest risks were noted in the 3–6 month follow-up. To our knowledge, this is the first study to reveal this correlation. We proposed possible explanations for the short-term and long-term risks of this observation. First, rifampin enhances T4 clearance and may cause hypothyroidism (16, 17). Without treatment, ~4–15% of patients with latent TB infection develop TB within 1–5 years after getting infected (48–50). Devalraju et al. found that 17 of 688 healthy household contacts developed active TB, and that 12 of 17 progressors received a diagnosis of TB within the first year (20). These may be the reasons for the short-term risk of both diseases. Second, Mtb infection may trigger autoimmune reactions by molecular mimicry of Mtb antigens with human proteins (51) and infection may subsequently result in autoimmune thyroid diseases (22). Third, asymptomatic populations of hypothyroidism and TB may result in delayed diagnosis. And finally, genetic polymorphism plays a major role in the progression to active TB (52) and two single nucleotide polymorphisms in immune-and inflammation-related genes (interleukin-6 rs2066992 and rs1524107) increased the risk of active TB (53). In addition, interleukin-6 can inhibit thyroid function in vivo (54, 55). The relationship between the long-term risk of TB and hypothyroidism may involve complex interactions between genes, cytokines, and the immune system, and additional studies are warranted to elucidate these mechanisms.

Our study had some limitations. First, we were unable to analyze the type or disease severity of hypothyroidism and TB in both cohorts because of the lack of symptoms and the absence of associated laboratory data, such as concentrations of thyroid-stimulating hormone, free T4, T3, or anti-thyroid peroxidase antibodies, in the NHIRD. Second, similar to issues with all electronic health databases, there can be problems related to coding errors and deliberate upcoding. Although previous studies have demonstrated high accuracy and validity for ICD-9-CM-based LHID disease identification (27, 29–31), we sought to reduce misclassification bias with a stricter definition of hypothyroidism and TB (29). On the other hand, we may underestimate the populations of asymptomatic subclinical hypothyroidism. Third, clinicians did not routinely examine chest radiography scans and thyroid functions in either cohort, which might have resulted in underdiagnosis. Fourth, although we adjusted for the latent variables in the table, there might have been unknown or unmeasured confounding biases. Fifth, although rifampin, ethionamide and para-aminosalicylic acid were the possible cause of hypothyroidism (16, 17, 23), we did not perform an analysis of these drugs. Additional research should be performed in the future.

In conclusion, the current study revealed a bidirectional relationship between TB and hypothyroidism. Patients with hypothyroidism showed a 2.91-fold higher risk of developing TB than the general population, especially those combined with DM or COPD. The hypothyroidism-linked TB risk may be attenuated by long-term TRT. Physicians should be aware of the risk of developing hypothyroidism when treating patients with TB. We recommend that public health authorities should conduct surveillance for both diseases to reduce the associated disease burden.

The data analyzed in this study is subject to the following licenses/restrictions: The data on the study population that were obtained from the NHIRD are maintained in the NHIRD. The National Health Research Institutes (NHRI) is a non-profit foundation established by the government. Only citizens of Taiwan who fulfill the requirements of conducting research projects are eligible to apply for access to the NHIRD. The use of the NHIRD is limited to research purposes only. Applicants must follow the Computer-Processed Personal Data Protection Law and the related regulations of the National Health Insurance Administration and NHRI, and an agreement must be signed by the applicant and their supervisor upon application submission. All applications are reviewed for approval of data release. Requests to access these datasets should be directed to https://nhird.nhri.org.tw/.

The studies involving human participants were reviewed and approved by Institutional Review Board of the Tri-Service General Hospital (approval number: B-111-01). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

L-TC, C-CS, S-HW, and S-ET contributed to conception and design of the study. C-HC and W-CC organized the database and performed the statistical analysis. L-TC wrote the first draft of the manuscript. C-KP, S-HW, and S-ET wrote sections of the manuscript. L-TC, S-YW, G-JW, C-HT, C-AS, and S-ET were responsible for the critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.

This study was supported by the Tri-Service General Hospital Research Foundation (TSGH-E-111230), the sponsor had no role in the study design, data collection, analysis, decision to publish, or preparation of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.