Graves’ disease is the most frequent cause of persistent hyperthyroidism due to the overproduction of thyroid hormones (1–3). Graves’ disease is an organ-specific autoimmune thyroid disorder that is characterized by breaking immune tolerance against thyroid antigens resulting in thyroid autoimmunity (4). The primary clinical symptoms or signs of Graves’ disease are diffuse thyroid enlargement and overaction, ophthalmopathy, pretibial myxedema, palpitations, weight loss, and others (5). The incidence of Graves’ disease is 20-40 cases per 100,000 person-years worldwide (6), and is more common in women than men, which approximately affects 0.2% of men and 2% of women (7, 8). Graves’ disease is commonly observed in women aged 20-50 years, whilst all ages could be affected (6). Although the autoimmune mechanism of Graves’ disease has been proposed for several decades, its etiology has not been well accepted and needs to be further elucidated.

Telomeres, the terminal nucleoprotein-DNA complexes of eukaryotic chromosomes, play crucial and pervasive roles in capping and protecting chromosome DNA ends (9, 10). Telomeres, to a certain extent, can maintain the integrity of the replicative capacity of the chromosomes after each cell cycle (11). Telomeres progressively shorten with each cellular division cycle until eventually reaching a critically short length or sufficiently damaged, which triggers subsequently cellular senescence or apoptosis (12). Consequently, telomere length (TL) has been proposed as an underlying biomarker for cellular aging and mortality (13–15). Furthermore, telomere attrition might give rise to potential cell changes, block cellular division, and interfere with the normal function of tissues (12). Phenome-wide analyses provided evidence for association of shorter leukocyte TL with increased risk of thyroid disorders in more than 350,000 UK participants (16). However, whether there is a potential causal effect on leukocyte TL and Graves’ disease has not yet been investigated.

Mendelian randomization (MR) study, an epidemiological method, could minimize potential bias due to confounding and reverse causation (17). MR provided a novel opportunity to explore the potential causal association between an exposure of interest and outcome in an observational study (18). It relies on three core assumptions ( Figure 1 ): (I) the instrumental variables are associated with the exposure of interest (e.g. leukocyte TL); (II) the instrumental variables are not associated with confounders; (III) the effects of instrumental variables on the outcome (e.g. Graves’ disease) are only through the exposure of interest. To the best of our knowledge, there is no study examining leukocyte TL and Graves’ disease using the MR approach.

In the present study, we identified nine SNPs to be of genome-wide significance for leukocyte TL in a large-scale genome-wide association study (GWAS) from the Singapore Chinese Health Study (SCHS) cohort and examined its association with Graves’ disease in Biobank Japan.

The present study is the first of this kind to examine the causal association between leukocyte TL and Graves’ disease using a two-sample MR design. By leveraging several MR estimation approaches, we found that shorter TL was associated with an increased risk of Graves’ disease. Our study, in line with previous observational studies, provides evidence to support a causal role of shorter TL for Graves’ disease.

A number of studies about telomere length and lifetime disease risks have been extensively conducted, including autoimmune disorders (35, 36) and age-related diseases (37). Autoimmune diseases are characterized by spontaneous hyperactivity of the immune system comprising autoantibody production. Inflammation and oxidative stress are recognized as parts of the process for most of these diseases, but their molecular mechanisms have not fully been understood, Intriguingly, escalating telomere shorten was primarily caused by environmental factors concerning inflammation and oxidative stress, probably indicating an association between the telomere system and autoimmune or inflammatory diseases (38, 39). Particularly, for the past decade, data emerging from studies in oncology on the telomere system and its underlying therapeutic implications, stimulated studies estimating the role of the telomere system in the biological mechanisms of autoimmune diseases. Indeed, many MR studies utilized telomere length-associated genetic variants to predict the risk of disease based on the genomic architectures of the individuals (40, 41).

To date, no studies, to our knowledge, have been conducted to investigate the role of TL in Graves’ disease. A recent study showed that the relationship between Graves’ disease and rheumatoid arthritis, implying that these two autoimmune diseases were likely to share similar underlying mechanisms (42). Inflammation and oxidative stress are recognized as parts of the process for most of these diseases, but their molecular mechanisms have not fully been understood. Intriguingly, escalating telomere shortening was primarily caused by environmental factors, such as inflammation and oxidative stress, probably indicating an association between the telomeres and autoimmune or inflammatory diseases. Of note, a previous MR study indicated that genetically predicted TL was associated with a lower risk of rheumatoid arthritis, providing support for an inverse causal association (43). A possible biological explanation was that telomerase mutation carriers with short TL may develop T-cell aging phenotypes and immunodeficiency, and T cells with shorter telomeres could induce DNA damage and upregulate intrinsic apoptosis pathways (44). Another explanation was that telomere dysfunction could activate the production and secretion of inflammatory factors, such as IL-6 and TNF-α, which led to cellular senescence (45). Although these explanations are biologically plausible, further studies are required to clarify underlying mechanisms for the role of TL in the risk of Graves’ disease.

Our study has some strengths. First, the MR method tends to be less biased than conventional observational studies in the presence of unmeasured confounding and reverse causation, therefore it can offer a more robust estimate of a causal association. Indeed, this study obtained consistent estimates by using several MR approaches. Second, this study made the best use of publicly available data associated with the GWAS for leukocyte TL (23,096 individuals) and Graves’ disease (2,176 cases and 210,277 controls) with large sample sizes, which rendered us to gain more precise estimates and greater statistical power. Furthermore, the two-sample MR method was taken advantage of both GWAS summary levels with leukocyte TL and Graves’ disease that were derived from two independent populations (46).

Despite the advantages of the MR design, several limitations of the study should be acknowledged. Firstly, the instrumental variables were derived from blood TL and not TL in thyroid tissues in the present study. However, previous studies have shown that leukocyte TL was highly correlated with TL in other tissues (14, 47). Secondly, we cannot prove that the selected SNPs satisfy the assumption of exclusion restriction. However, MR-Egger regression and MR-PRESSO in this study were applied to assess the extent to which genetic pleiotropy may bias the results. The intercept of the MR-Egger regression analysis was close to zero, implying no strong evidence for supporting directional pleiotropy. Similarly, MR-PRESSO and the leave-one-out analysis did not detect significant outliers. Thirdly, it is assumed that the two samples were from the same underlying populations in a two-sample MR study. Lastly, our findings rested on data from GWAS that was only performed in individuals of Asian ancestry, whereas majority of genetic studies were dominated by European-descent samples, which made it difficult to generalize to other ethnic populations. Because of the differences in genetic population structure, the transferability of genetic results across populations is relatively limited. The validity of our analysis relies on the MR assumption that SNPs are not associated with confounders that are related to the occurrence of Graves’ disease. To minimize bias due to violations of this assumption, we reviewed previous articles regarding telomere length and Graves’ disease and screened risk factors related to both as potential confounders, which included smoking status, vitamin D deficiency, and immune status. We found that the instrumental variables were not associated with these potential confounders. Additionally, we screened these instrumental variables through GWAS catalog to try to identify potential genome-wide significant associations with other traits in Asian population, but we did not find any such traits.

In summary, in the largest study until now, we provided novel evidence to support a causal association of genetically predicted shorter leukocyte TL with an increased risk of Graves’ disease. Further studies are warranted to clarify a biological mechanism of telomeres in the disease onset and progression of Graves’ disease in both the Asian and other ethnic populations.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

YZ designed this study and collected data, and MY and YW performed the analysis and wrote the manuscript. All authors approved the final version of the manuscript.

This work was supported by Sun Yat-Sen University.

The funder had no role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

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.