Abstract

Background

   There is still controversy regarding the relationship between
   hypothyroidism and rheumatoid arthritis (RA), and there has been a
   dearth of studies on this association. The purpose of our study was to
   explore the shared genetic architecture between hypothyroidism and RA.

Methods

   Using public genome-wide association studies summary statistics of
   hypothyroidism and RA, we explored shared genetics between
   hypothyroidism and RA using linkage disequilibrium score regression,
   ρ-HESS, Pleiotropic analysis under a composite null hypothesis (PLACO),
   colocalization analysis, Multi-Trait Analysis of GWAS (MTAG), and
   transcriptome-wide association study (TWAS), and investigated causal
   associations using Mendelian randomization (MR).

Results

   We found a positive genetic association between hypothyroidism and RA,
   particularly in local genomic regions. Mendelian randomization analysis
   suggested a potential causal association of hypothyroidism with RA.
   Incorporating gene expression data, we observed that the genetic
   associations between hypothyroidism and RA were enriched in various
   tissues, including the spleen, lung, small intestine, adipose visceral,
   and blood. A comprehensive approach integrating PLACO, Bayesian
   colocalization analysis, MTAG, and TWAS, we successfully identified
   TYK2, IL2RA, and IRF5 as shared risk genes for both hypothyroidism and
   RA.

Conclusions

   Our investigation unveiled a shared genetic architecture between these
   two diseases, providing novel insights into the underlying biological
   mechanisms and establishing a foundation for more effective
   interventions.

   Keywords: hypothyroidism, rheumatoid arthritis, genome-wide association
   studies, shared genetic architecture, shared risk genes

1. Introduction

   The condition of hypothyroidism is characterized by low thyroid hormone
   levels and high thyroid-stimulating hormone (TSH) levels. Approximately
   4% to 10% of the population is affected by hypothyroidism, and
   subclinical hypothyroidism has been reported in up to 10% of
   individuals ([37]1), with a higher prevalence among women and the
   elderly ([38]2). Hypothyroidism is associated with cardiac dysfunction,
   atherosclerosis, hypertension, and coagulation disorders, potentially
   reducing patients’ lifespans ([39]3).

   Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune and
   autoimmune disease that affects 0.5% to 1% of the world’s population
   ([40]4). It leads to inflammation and structure damage of joints,
   reduced mobility, and increased disability ([41]5). Even though their
   precise etiology is not fully understood, these diseases are believed
   to be associated with an interplay between genetic predisposition and
   environmental influences ([42]6–[43]9). Despite notable advancements in
   RA treatment, patients experience less favorable outcomes concerning
   quality of life, morbidity, and mortality than the general population
   ([44]10).

   Previous studies have demonstrated an elevated rate of hypothyroidism
   in RA patients compared to controls ([45]11). Recently, Gao Y et al.
   provided compelling evidence supporting a causal association between
   hypothyroidism and increased risk of RA ([46]12). In a cohort study,
   there was no significant difference in the prevalence of hypothyroidism
   between RA patients and healthy adults ([47]13). However, RA and
   hypothyroidism have not always been associated consistently in the
   literature, and there has been a dearth of studies on this association.

   There is an elusive relationship between hypothyroidism and RA. Our
   study aims to clarify the shared genetic architecture and molecular
   pathways of hypothyroidism and RA, providing new information on their
   underlying biological mechanisms and paving the way for more effective
   interventions. [48]Figure 1 illustrates the overall study design.

Figure 1.

   [49]Figure 1
   [50]Open in a new tab

   Overall study design.

2. Methods

2.1. Data source

   We used genome-wide association studies (GWAS) summary statistics for
   hypothyroidism involving 462,933 individuals (22,687 cases and 440,246
   controls) of European ancestry from the GWAS Catalog
   ([51]https://www.ebi.ac.uk/gwas/). For RA, we applied summary
   statistics for the discovery dataset from GWAS with 253,417 individuals
   (12,555 cases and 240,862 controls) from the FinnGen Consortium
   ([52]https://r9.finngen.fi/) ([53]14). To corroborate the findings from
   the initial discovery dataset, a replication GWAS of RA was performed.
   This replication involved 58,284 individuals, comprising 14,361 RA
   cases and 43,923 controls ([54]15). [55]Supplementary Table S1 presents
   a comprehensive description of the characteristics of each dataset
   utilized in our study. To mitigate the potential bias of ethnic
   diversity, we focused our investigation on individuals of European
   ancestry. Ethical clearance was not required, as the study relied on
   publicly available data.

2.2. Linkage disequilibrium score regression

   Heritability estimates for hypothyroidism and RA were derived using
   LDSC in Python 2.7 ([56]16). Additionally, we calculated the genetic
   correlation (r[g]) between the two diseases, quantifying the shared
   genetic variance relative to the square root of their respective Single
   Nucleotide Polymorphism (SNP) heritability estimates. The analysis
   utilized the 1000 Genomes European Reference dataset to convert GWAS
   summary statistics into precalculated linkage disequilibrium (LD)
   scores. Sensitivity analysis was conducted with a single-trait
   heritability intercept constraint.

2.3. ρ-HESS

   We used the ρ-HESS approach to investigate the local genetic
   correlations between hypothyroidism and RA ([57]17), which estimates
   the local heritability of SNPs and genetic covariance based on genomic
   references constructed from genomes. We could estimate local genetic