Causal Associations Between Chronic Spontaneous Urticaria and Thyroid Function Indicators on Different Ethnic Groups: A Bidirectional Two-Sample Mendelian Randomization Analysis

Some studies across Europe and Asia have shown a significantly increased risk of autoimmune thyroid disease (AITD) in patients with chronic spontaneous urticaria (CSU), with a particularly strong link to Hashimoto's disease, others find no or even an inverse relationship. TPO-anti-TPO IgE interaction in CSU can activate cells releasing inflammatory mediators [1, 2], but antithyroid antibodies alone are not sufficient to trigger mast cell activation. More importantly, the causal relationship between CSU and thyroid function remains unclear.

Traditional studies cannot reliably link CSU and AITDs because of bias and reverse causation. Mendelian randomization (MR), using randomly inherited genetic variants, overcomes these issues. We used a two-sample MR analysis with European and East Asian populations to probe the causality between thyroid function and CSU, and potential reverse causality.

Our current analysis covered genetic variants associated with thyroid conditions (Table S1). European GWAS data came from Open GWAS, FinnGen R10 (L12_URTICA_IDIOPAT), and ThyroidOmics Consortium [3-5]. East Asian AITD and CSU data were from BioBank Japan [4] and Open GWAS [6]. The genetic instrument selection for analysis involved identifying significant SNPs associated with thyroid traits and CSU (p < 1 × 10⁻⁵), excluding those with high linkage disequilibrium (r² < 0.001) or missing outcome data, and ensuring strong F-statistics (F > 10).

Europeans' IVW analysis linked genetic predispositions to Graves' disease (OR = 1.24, p < 0.001), hyperthyroidism (OR = 1.22, p < 0.001), hypothyroidism (OR range: 1.34–22.18, p < 0.001) and Hashimoto's (OR = 1.19, p = 0.0026) with higher CSU risk (Figure 1 and Table S2). No thyroid function indicators significantly affected CSU risk. Reverse MR analysis showed no significant CSU effects on most thyroid functions (p > 0.05), except for hyperthyroidism (OR = 0.92, p = 0.0051) and the lnFT3FT4 ratio (OR = 1.00, p = 0.0127), where an OR close to 1.00 indicates no practical effect. After Bonferroni correction (adjusted p value of 0.05/21), the forward results held, with no significant reverse findings.

In East Asians, no causal relationships were found between thyroid disorders, including Graves' disease, hyperthyroidism, hypothyroidism, and Hashimoto's thyroiditis, and CSU in either direction (Figure 2 and Table S3). This could be attributed to our smaller sample size or regional genetic and environmental factors. Sensitivity analyses (Tables S4 and S5), including MR-PRESSO, revealed no significant heterogeneity or pleiotropy.

Both Graves' disease and CSU are characterized by high Th2 cytokines. Graves' susceptibility gene PTPN22 and its variant alleles correlate with increased ASTT in Poland and CSU susceptibility in Iran. Current studies show a significant overlap in the pathophysiology between CSU and HT, affecting up to 30% of patients, and include the impact of IL-6 and Tregs.

There are two CSU subgroups: autoimmune type I (aaCSU) and type IIb (aiCSU), differing in pathogenesis and markers [2]. Elevated IgE-anti-TPO levels are found in aaCSU, whereas IgG-anti-TPO antibodies mark aiCSU, and patients with type IIb CSU are at risk for autoimmune diseases like Hashimoto's and vitiligo. Patients with CSU show a correlation between IgE-anti-TPO and IgG-anti-TPO levels [2]. However, this is not a consistent feature in all patients [2]. High serum Anti-TPO IgE and IgG levels alone do not predict or prognosticate disease outcomes. The access of thyroid antigen to the skin remains to be demonstrated.

In brief, our bidirectional MR analysis reveals a genetic susceptibility to thyroid dysfunction associated with CSU in Europeans, but not in Asians. No reverse causation was evident, and thyroid function indicators showed no direct link to CSU. These findings suggest new avenues for exploring the relationship between thyroid autoimmunity and CSU. Future GWAS with larger, more diverse cohorts, including stratification by CSU subtypes, may uncover additional genetic variants, thereby refining our understanding of these mechanisms.

Lijun Deng and Ping Xia were involved in drafting the manuscript. Yuxu Yao has contributed substantially to the conception and design in the present study. Tingting Yu, Hui Shen, and Zhenzhong Lu were involved in data acquisition, analysis, and interpretation. Jiang Ji coordinated the study and was involved in data analysis. Qingqing Jiao was the overall study coordinator and was involved in proofreading the manuscript.

The authors declare no conflicts of interest.