Causal relationship between hypothyroidism and ulcerative colitis: a bidirectional Mendelian randomization study

Abstract

Objective

Ulcerative colitis (UC) and Hashimoto’s thyroiditis frequently cooccur in patients with multiple autoimmune conditions, but the specific association between UC and hypothyroidism is unknown. We used Mendelian randomization (MR) methods to determine the causal relationship between UC and hypothyroidism.

Methods

We obtained single nucleotide polymorphisms (SNPs) related to ulcerative colitis (UC) and hypothyroidism from genome-wide association studies (GWAS) available in the public database of the Integrated Epidemiology Unit (IEU). To assess the causal relationship between UC and hypothyroidism, we employed MR-Egger, weighted median, inverse variance weighted (IVW), simple mode, and weighted mode methods. Sensitivity analyses were performed using Cochran’s Q test, the horizontal pleiotropy test, and the leave-one-out (LOO) method to assess the reliability of the MR data. The genes corresponding to instrumental variables (IVs) were subjected to Gene Ontology (GO) functional annotation, Kyoto Encyclopedia of the Genome (KEGG) pathway enrichment analysis, and protein-protein interaction (PPI) analysis to explore the mechanisms behind the causal relationships at the gene level.

Results

Forward MR analysis indicated that hypothyroidism was associated with an increased risk of UC (IVW: P = 0.02, OR = 9.71, 95% confidence interval (CI) = 1.36–69.46). In contrast, reverse MR did not demonstrate a causal relationship between UC and hypothyroidism (IVW: P = 0.53). Sensitivity analysis proved the reliability of the results. The PPI network revealed CD247, CD80, and STAT4 as central genes. GO and KEGG analyses revealed significant enrichment of the T cell, gamma interferon (IFN-γ), and programmed cell death-1 (PD-1)/programmed cell death ligand-1 (PD-L1) pathways.

Conclusion

Hypothyroidism was a risk factor for UC. The balance of T-cell differentiation played an important role in the process of hypothyroidism-induced UC, and IL-21 might be the key to finding a cure. Enrichment of PD-1/PD-L1 might attenuate inflammation by suppressing the immune action of T cells.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-024-03461-y.

Introduction

Ulcerative colitis (UC) is a chronic, relapsing inflammatory disease that primarily affects the mucosal lining of the colon and rectum. Its most typical symptom is rectal bleeding accompanied by diarrhea, and the condition is characterized by periodic attacks and remissions of mucosal inflammation [1]. The diagnosis of this disease usually relies on a comprehensive evaluation, including the patient’s clinical manifestations, endoscopy and biopsy results, such as colonoscopy ulcers, mucosal congestion and edema, etc.

Hypothyroidism is characterized by low thyroid function, resulting in insufficient secretion and release of thyroid hormone. The main hormones secreted by the thyroid gland are thyroxine and triiodothyronine, which are essential for regulating human metabolism [2]. Common symptoms include fatigue, weakness, weight gain, dry skin and hair, muscle aches, unbearable colds, and sleep disturbances [3]. Diagnosing hypothyroidism usually requires a blood test to measure the levels of thyroid-stimulating hormone (TSH) and thyroxine (T4). The disease is mainly treated with hormone replacement therapy, such as taking levothyroxine [4].

One study has shown that the therapeutic dose of thyroxine was increased in almost all patients with hypothyroidism combined with UC [4]. UC and Hashimoto’s thyroiditis often cooccur in multiple autoimmune patients [5]. Additionally, Hashimoto’s thyroiditis is one of the most common causes of hypothyroidism [6]. However, the causal relationship between hypothyroidism and UC remains unclear. Validating this causal relationship between hypothyroidism and UC risk in a real-world setting presents challenge due to the long-term nature of cohort studies, the influence of confounding factors, and the low prevalence of UC. Therefore, we suggested Mendelian randomization (MR) analysis as a feasible research strategy to explore the causal relationship between UC hypothyroidism.

MR encompasses epidemiological study design, data analysis, and the formulation of aetiological hypotheses based on Mendel’s law of independent assortment. This principle, rooted in Mendel’s second law, involves the independent and random assortment of alleles during gamete formation as genetic material is passed from both parents to their offspring [7]. This genetic variation is typically free from the influence of confounding factors and remains unaltered by acquired environmental, behavioural, psychological, and socioeconomic factors [8]. MR offered a means to address the limitations of traditional analyses in elucidating the intricate causes of diseases, which were often susceptible to interference from confounding factors and ambiguous causal temporal sequences [9].

In our study, we obtained single nucleotide polymorphism (SNP) data for UC and hypothyroidism based on genome-wide association study (GWAS) data from the public database Integrated Epidemiology Unit (IEU) and investigated the causal relationship between the two by using a bidirectional MR approach. Moreover, Gene Ontology (GO) functional annotation, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, and protein-protein interaction (PPI) analysis were performed on the genes corresponding to the instrumental variables (IVs) to explore the causal relationship between UC and Hypothyroidism, as well as the interactions.

Materials and methods

Data sources and preprocessing

The ebi-a-GCST004133 data (n = 45, 975, SNP = 9, 474, 559) for UC and ukb-b-19,732 data (n = 462, 933, SNP = 9, 851, 867) for hypothyroidism were obtained from the IEU Open GWAS database. In the R language Two Sample MR package (v 0.5.6) [10], i nstrumental variables (IVs) that were significantly correlated (P < 5 × 10 − 8) with exposure factors were firstly extracted using the extract-instruments function. Next, independent and distant SNPs were retained as IVs by setting clump = TRUE, r² = 0.001 and kb = 10,000 to remove the effect of linkage disequilibrium (LD). Then, the bidirectional MR analysis was performed based on the acquired IVs. Three fundamental premises underlie MR studies: [1] a robust and significant correlation exists between IVs and exposure; [2] IVs are unrelated to confounding factors; and [3] IVs can influence outcomes only through exposure and not through other pathways.

Bidirectional MR analysis

The harmonize data function of the Two Sample MR package (v 0.5.6) was employed to standardize effect equivalents and effect sizes. The primary MR methods included MR-Egger [11], weighted median [12], inverse variance weighted (IVW) [13], simple mode [14], and weighted mode [15]. IVW is the most crucial method due to its superior performance in detecting causation. This method yielded a P < 0.05, indicating statistical significance between the exposure factor and the outcome variable. An odds ratio OR > 1 indicated a risk factor, while an OR < 1 indicated a protective factor. Mendelian randomization was conducted utilizing several visual techniques: scatter plots illustrated the correlation between exposure and outcome; forest plots depicted the diagnostic efficacy of the exposed SNP locus for the outcome; and funnel plots assessed potential publication bias and heterogeneity in the results.

Sensitivity analysis

To assess the reliability of the MR analysis, a sensitivity analysis was conducted. First, a heterogeneity test using Cochran’s Q test, with Q_pval > 0.05 indicated the absence of heterogeneity. Subsequently, a horizontal pleiotropy test was carried out via the two-sample MR mr-pleiotropy-test function, with P > 0.05 indicating no horizontal pleiotropy, suggesting the absence of confounding factors in the study. Finally, the leave-one-out (LOO) method was employed by systematically eliminating each SNP. If the effect of the remaining SNPs on the outcome variable did not significantly change, this indicated the reliability of the MR analysis results.

Functional enrichment analysis and PPI

Genes corresponding to the IVs in the forward MR were identified using the Variant Effect Predictor (VEP) tool. To elucidate the functions and pathways of genes regulated by the IVs, GO and KEGG analyses of the genes corresponding to the IVs were conducted using the cluster-Profiler package (v 4.7.1) [16], with a significance threshold of P < 0.05. The genes were then input into the String database (v 11.5) (https://string-db.org) to construct the generic protein‒protein interaction (PPI) network, with a confidence level of 0.700. Subsequently, the core genes were further screened to obtain the most essential PPI network.

Results

Hypothyroidism was a risk factor for UC

The overall design of the study is shown in Fig. 1. All the information was available in the online database. Forward MR analysis was conducted to investigate the impact of hypothyroidism on UC. A total of 104 SNPs associated with hypothyroidism were screened and utilized for forward MR analysis. Detailed information was shown in Table S1. Among the various MR methods, the most significant IVW model indicated a statistically significant association between the exposure factor hypothyroidism and the outcome of UC (P = 0.02, OR = 9.71, 95% confidence interval (CI) = 1.36–69.46). The results for the remaining four methods were MR Egger (P = 0.06, OR = 75.73, 95% CI = 0.87-6593.43), Weighted median (P = 0.07, OR = 4.64, 95% CI = 0.88–24.38), Weighted mode (P = 0.30, OR = 3.11, 95% CI = 0.36–26.62) and Simple median (P = 0.24, OR = 2.55, 95% CI = 0.53–12.21) (Table 1). This suggestted that hypothyroidism was causally linked to UC and is considered a risk factor.

Fig. 1.

Flowchart of the experimental design

Table 1.

Forward MR analyses of hypothyroidism and UC

outcome	exposure	Method	P-value	OR (95%CI)
Ulcerative colitis	hypothyroidism	MR Egger	0.06	75.73 (0.87-6593.43)
		Weighted median	0.07	4.64 (0.88–24.38)
		Inverse variance weighted	0.02	9.71 (1.36–69.46)
		Weighted mode	0.30	3.11 (0.36–26.62)
		Simple median	0.24	2.55 (0.53–12.21)

The MR scatterplot demonstrated a small IVW intercept, signifying the minimal impact of confounding factors on the reliability of the results (Fig. 2A). Furthermore, the positive slope of the IVW indicates that exposure factors act as risk factors, implying that an elevated incidence of hypothyroidism corresponds to an increased risk of UC. Subsequently, a forest plot was utilized to assess the diagnostic efficacy of the predicted exposure factors on the outcome for each SNP locus and was further corroborated by Fig. 2B, revealing that the exposure factor hypothyroidism increased the risk of developing UC under the IVW method. The symmetrical distribution of IV along the IVW line on both sides in the funnel plot suggest that MR conforms to the second law of Mendelian inheritance (Fig. 2C).

Fig. 2.

Hypothyroidism is the cause of ulcerative colitis. (A) Scatterplot. The slope of each line indicates the estimated effect of each MR method. (B) Forest plot. The red dots indicate the combined estimates of all SNPs using the IVW method. The horizontal line represents the 95% confidence interval. (C) Funnel plot. The vertical lines represent the estimates of all SNPs. The symmetry of the funnel plot indicates that there is no significant horizontal fold effect. (D) Leave-one-out analysis. The black dots indicate that a single specific variant was excluded from the analysis using the IVW approach to assess causal effects. Red dots indicate inverse variance weighted estimates using all SNPs. IVW: inverse variance weighted; SNP: single-nucleotide polymorphism

Sensitivity analysis confirmed the reliability of forward MR results

To ascertain the robustness of the MR analyses, sensitivity analyses were conducted on the forward MR assessments. First, the heterogeneity test revealed significant heterogeneity between the two groups (MR Egger P = 1.41 × 10^− 71, IVW P = 2.82 × 10^− 72), and the P value of the random effects IVW model utilized in the analyses was less than 0.05, indicating the reliability of the MR results. Second, the outcomes of the horizontal multivariate test indicated the absence of horizontal multivariate effects (P = 0.32), implying that there were no confounding factors in the MR forward analyses (Table S2). Finally, LOO was performed, which demonstrated minimal bias and confirmed the reliability of the results (Fig. 2D).

UC was not a risk factor for hypothyroidism

A reverse MR analysis was conducted to investigate the impact of UC on hypothyroidism. Screening 50 independent SNPs as IVs for the analysis revealed no causal relationship between UC and hypothyroidism (IVW P = 0.53, OR = 1.0007, 95% CI = 0.9986–1.003; MR Egger P = 0.89, OR = 0.9995, 95% CI = 0.9921–1.007; Weighted median P = 0.43, OR = 0.9993, 95% CI = 0.9976–1.001; Weighted mode P = 0.68, OR = 0.9995, 95% CI = 0.9970–1.002; Simple median P = 0.71, OR = 0.9994, 95% CI = 0.9961–1.003) (Table 2). The results of the reverse MR visualization and sensitivity analysis are included in the supplementary material.

Table 2.

Reverse MR analyses of UC and hypothyroidism

outcome	exposure	Method	P-value	OR (95%CI)
hypothyroidism	Ulcerative colitis	MR Egger	0.89	0.9995 (0.9921–1.007)
		Weighted median	0.43	0.9993 (0.9976–1.001)
		Inverse variance weighted	0.53	1.0007 (0.9986–1.003)
		Weighted mode	0.68	0.9995 (0.9970–1.002)
		Simple median	0.71	0.9994 (0.9961–1.003)

CD247, CD80 and STAT4 were pivotal genes involved in hypothyroidism-induced UC.

Based on the analysis of 104 IVs using forward MR, VEP identified 107 corresponding genes and revealed 32 focus genes (Fig. 3). The PPI network showed complex interactions between proteins, with CD247, CD80 and STAT4 more closely linked to other genes. The ITK, CTLA4, CD69, and CD274, which surrounded CD80 and CD247, were primarily involved in regulating T-cell homeostasis. They inhibited T-cell activation and cytokine production and inactivate cytotoxic T cells. In addition, the CD274 gene encoded an immune-inhibitory receptor ligand. The interaction with its ligand during infection or inflammation of normal tissues could prevent autoimmunity by maintaining the balance of the immune response. Among the genes surrounding the STAT4, PTPN2, IL12RB2, and TERT were involved in cell growth, differentiation, and senescence cycles. NFκB and IL-21 affected the value-added activation of immune cells and played a role in the immune response.

Fig. 3.

PPI network diagram of genes corresponding to independent variables

T-cell differentiation and the PD-L1/PD-1 pathway were essential for the pathogenesis of hypothyroidism-induced UC.

To further understand the potential functions of the 107 genes involved in the development of hypothyroidism-induced UC, we performed GO and KEGG enrichment analyses (Fig. 4). GO analysis revealed that the biological process (BP) terms associated with the genes were mainly associated with T-cell activation, IFN-γ production and the immune response. KEGG signalling pathways included the T-cell receptor signalling pathway, Th1 and Th2 cell differentiation pathway, Th17 cell differentiation pathway, PD-L1 expression pathway and PD-1 checkpoint pathway in cancer.

Fig. 4.

GO analysis and KEGG signal pathway diagram. GO: Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; Count: Number of genes; BP: Biological process;

Discussion

In our study, we first found that hypothyroidism is a risk factor for UC, but UC is not a risk factor for hypothyroidism. Second, T-cell imbalance was involved in the process of thyroid function-induced UC development by triggering inflammation and modulating immunity, and the CD247 gene was also involved. Third, the PD-1/PD-L1 pathway has significance in hypothyroidism-induced UC, but its exact role was unclear. In recent years, studies have shown that UC and thyroid disease frequently occured together in patients with autoimmune disorders and that hypothyroid patients with combined UC received increased thyroxine treatment [4, 5]. Despite this, a causal link between hypothyroidism and UC has not been previously established. We found a significant association between UC and hypothyroidism by meticulous examination of public databases and the application of Bidirectional MR analysis to mitigate confounding variables and other potential effects. Our study suggested that thyroid function should be assessed in the treatment of UC, and determining whether UC was comorbid with thyroid disease was essential in the clinic.

The interrelationship between autoimmune diseases is a major clinical challenge. It has been reported that in patients with gastrointestinal disorders, such as UC, susceptibility to a variety of autoimmune diseases was commonly demonstrated [17]. A common feature of hypothyroidism and UC is an abnormal immune activity against the body’s own tissues, leading to disorders of the immune system [18]. Hypothyroidism was found to be a potential risk factor for UC by MR analysis, providing new insights into the pathogenesis of UC. This association is particularly noteworthy because hypothyroidism can be modified by medical interventions such as hormone replacement therapy. Clinically, this link suggests that enhanced monitoring of gastrointestinal symptoms in patients with hypothyroidism could lead to earlier detection and treatment of UC. There appears to be a complex interplay between thyroid, microbiota, and UC pathogenesis, given that hypothyroidism may alter intestinal motility and microbiota composition, which may in turn affect intestinal immunity and barrier function [19]. In addition, if the specific subgroup of thyroiditis were to be analysed separately from the group of patients suffering from hypothyroidism, it would be effective in reducing the heterogeneity of the study, thus making the study results more accurate and reliable. This is undoubtedly a work of great research potential and value, and is expected to bring new insights to the field of ulcerative colitis. In conclusion, our findings emphasised the critical role of thyroid function assessment in the management of UC. Early intervention of thyroid dysfunction is potentially therapeutically significant in the progression of UC disease and warrants further exploration.

T-cell differentiation plays a crucial role in the development of hypothyroidism-induced UC. Various immune cells, including Th1, Th2, Treg, and Th17 cells, were involved in the pathogenesis of UC [20]. CD247, an essential protein for T cell receptor signaling, is critical for T cell activation [21]. Helper T cells, such as Th1, Th2, and Th17 cells, produce IL-21, a cytokine that influences intestinal inflammation by stimulating IFN-γ production [22, 23]. Elevated levels of IL-21 have been linked to active UC and increased risk of UC recurrence [24, 25]. IL-21 also impacts the JAK-STAT pathway, affecting both innate and adaptive immunity [26]. It inhibits Treg suppression and promotes Th17 responses, contributing to inflammation [27]. The balance of T-cell differentiation, particularly the Th1/Th2 dichotomy and Th17/Treg imbalance, plays a significant role in autoimmune thyroid disease and UC due to hypothyroidism, and IL-21 may be the key to finding a cure.

PD-1 and PD-L1 are a pair of negative costimulatory molecules that function to maintain immunological tolerance to autoantigens, attenuate T-cell activation signals, and mediate immune tolerance or tumour immune escape [28]. The PD-1 and PD-L1 pathways are also significantly involved in hypothyroidism-induced UC, and the elevated levels of PD-1/PD-L1 in mucosal tissues of patients with UC may be a protective feedback mechanism generated by immune cells in an inflammatory state [29]. PD-1 expression on Th cells was found to be positively correlated with disease activity in patients with active UC [30]. PD-L1 can suppress T lymphocyte immune responses in the gut by inducing and maintaining the Treg pathway [31], and activation of the PD-1/PD-L1 pathway in Th1 cells mediates their transformation into Treg cells by reducing STAT activation [32]. PD-1 interacts with PD-L1 to reduce T-cell activation and function, thereby counteracting positive signals from T-cell receptors [33, 34]. Currently, the role of PD-1/PD-L1 in UC is unclear due to the limited number of studies. It may suppress intestinal inflammation by inhibiting the immune response of T cells, or it may be self-protective during chronic inflammation. In addition, CD80 interacts more stronglywith PD-L1 than with PD-1/PD-L1, and the PD-L1/CD80 pathway also plays a key role in inducing and maintaining T-cell tolerance [35]. Moreover, CD80 is specifically overexpressed in the colonic mucosa of patients with UC with allopatric hyperplasia [36]. However, the specific mechanisms of the CD80/PD-L1 pathway involved in UC development still need to be explored. Our analysis indicated that the PD-L1/PD-1 pathway plays a significant role in the development of hypothyroidism-induced UC. By specifically regulating the PD-L1/PD-1 pathway, this study may provide new treatment strategies for alleviating UC symptoms caused by hypothyroidism.

We used large-scale SNP data for bidirectional MR analysis to explore the causal relationship of hypothyroidism on UC, as well as sensitivity analyses of the results to provide strong evidence for a causal relationship between UC and ypothyroidism, and functional pathway analyses for the PPI network and GO and KEGG functional pathway analyses to explore the mechanism of hypothyroidism contributing to the development of UC, which will provide an important reference for further research in this field. In conclusion, our study revealed that hypothyroidism is a significant risk factor for the onset of UC. Dysregulation of T-cell differentiation is a key player in this pathogenesis, with IL-21 potentially serving as a critical mediator in maintaining a controlled immune balance. The involvement of the PD-1/PD-L1 pathway in the development of UC in the context of hypothyroidism is evident. However, the precise mechanism of action remains to be elucidated.

Although the current research results are enlightening, to truly reveal the depth and breadth of these findings, they need to be verified and expanded through a series of more extensive and in-depth studies. This includes but is not limited to expanding the study sample size, conducting multi-center collaborative research, and implementing long-term follow-up to comprehensively examine the role of the PD-L1/PD-1 pathway in the process of hypothyroidism-induced UC and evaluate its practical application. Feasibility and effectiveness in clinical application. In the future, based on this study, we plan to launch a cohort study focusing on the complex association between hypothyroidism and UC. Through this study, we will be able to deeply analyze the subtle differences in blood protein profiles and other related factors between UC patients with hypothyroidism and ordinary UC patients, thereby revealing the potential pathophysiological mechanisms of hypothyroidism-induced UC. Such efforts will not only help us more fully understand the connection between the two diseases, but may also provide new clues and directions for future precision treatments.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

J.M. conceived and designed the project. Y.M.Y. interpreted the data and drafted the manuscript. Y.M.Y, J.J.L, X.W. and J.M. contributed to the editing of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the Natural Science Foundation of China (NSFC) [grant number 82341223], the Youth Science Foundation of NSFC [grant number 82202844] and the Tang Du Hospital Phoenix Project.

Data availability

All data involved in this study were obtained from the public database Integrated Epidemiology Unit. Trait: Ulcerative colitis - IEU OpenGWAS project (mrcieu.ac.uk)，Trait: Non-cancer illness code, self-reported: hypothyroidism/myxoedema - IEU OpenGWAS project (mrcieu.ac.uk).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.