Causal association between body mass index and autoimmune thyroiditis: evidence from Mendelian randomization

Autoimmune thyroiditis (AIT), particularly Hashimoto's disease, is the most common autoimmune thyroid disease and is estimated to affect about 1–5% of the general population [1]. As the most common autoimmune inflammatory disease of the thyroid gland, AIT is also the main cause of autoimmune hypothyroidism [2]. Importantly, AIT-induced hypothyroidism typically persists permanently and usually necessitates lifelong thyroid hormone replacement therapy [2, 3]. In addition, as thyroid tissue is gradually destroyed and production of thyroid hormones decreases, it can cause various symptoms such as fatigue, weight gain, constipation, feeling cold, dry skin, depression, muscle pain, and reduced tolerance to exercise [4, 5].

The treatment for AIT depends on the severity of the condition and the symptoms experienced by the patient. Clinical evidence has shown that myo-inositol and selenium can be effective in treating AIT [6‐8]. A systematic review and meta-analysis have found that vitamin D supplementation may have a beneficial effect on thyroid autoantibody levels in the treatment of AIT [9]. In addition, Glycyrrhizin, a direct HMGB1 antagonist, has been shown to ameliorate inflammatory infiltration in a murine model of AIT via inhibition of TLR2-HMGB1 signaling [10]. These factors may influence the development of AIT. Although research has implicated oxidative stress and psychiatric disorders as potential risk factors for developing AIT [11, 12], the exact mechanisms underlying this relationship are not well understood.

Adipocytes in adipose tissue produce adipokines, which can have either pro-inflammatory or anti-inflammatory effects in various tissues and organs, including the thyroid gland [13‐15]. Adipocytes can also recruit and activate immune cells, promoting the release of pro-inflammatory mediators, and contributing to the development of chronic inflammatory conditions [16, 17]. Recent studies have reported that adipocytes possess the ability to engage in direct interactions with immune cells, either through cell-to-cell contact or via the release of extracellular vesicles containing microRNAs and other molecules. These interactions have been found to modulate the activity of immune cells and subsequently impact the magnitude of the immune response [18‐21].

Body mass index (BMI) is a commonly used measure of body fatness and is calculated by dividing an individual's weight in kilograms by the square of their height in meters. It serves as a proxy for overall adiposity and is widely used in epidemiological studies to assess the relationship between obesity and various health outcomes [22, 23]. Consequently, it is hypothesized, based on the aforementioned research, that individuals with elevated BMI may be at an increased susceptibility for the development of AIT. This supposition finds support in observational studies [24, 25]. However, these types of studies can be influenced by biases such as reverse causation and residual confounding, which may limit our understanding of the impact of BMI on AIT.

Mendelian randomization (MR) is a statistical technique that uses genetic variants as instrumental variables (IVs) to evaluate the causal relationship between an exposure factor and an outcome [26]. This method can provide stronger evidence of causality than observational studies since it can minimize confounding and reverse causation bias [26]. Several MR studies have demonstrated the causal relationship between BMI(as the exposure) and numerous immune-related diseases [27, 28]. However, to our knowledge, no relevant studies have reported the causal effect of BMI on the risk of AIT.

In this study, we performed a two-sample summary data MR analysis to comprehensively investigate the causal relationship between BMI (as the exposure) and AIT (as the outcome).

This study utilized the MR approach to assess the relationship between BMI and AIT. For our primary analysis, we employed the IVW method, which has been shown to have high sensitivity in detecting causal effects. Our IVW results provide evidence supporting a potential causal association between BMI and AIT. We also conducted additional analyses, including the weighted median, MR-Egger, and weighted mode methods, in the two-sample MR framework. These analyses revealed a trend towards causality between BMI and AIT, although statistical significance was not observed. Nonetheless, the consistency of our results across various methods strengthens the robustness of our findings.

Previous studies have suggested a potential link between BMI and AIT, where higher BMI levels have been associated with an increased risk of developing AIT [24, 25]. However, these studies were observational, meaning that they examined the relationship between BMI and AIT in real-world settings without interventions or control groups. Note that observational studies have some limitations, such as the presence of confounding factors and the possibility of reverse causation. Confounding factors are variables that can independently influence both BMI and AIT, which makes it challenging to determine if BMI directly causes AIT. On the other hand, reverse causation suggests that AIT could actually lead to weight gain and higher BMI, rather than higher BMI causing AIT. While these observational studies offer valuable insights, they cannot definitively establish a causal link between BMI and AIT. By utilizing the Mendelian randomization approach, we were able to overcome these limitations and provide more robust evidence supporting a causal link between BMI and AIT.

The mechanisms linking BMI and AIT are unclear. Research suggests that obesity can lead to gut dysbiosis, an alteration in the gut microbiota composition, which can trigger autoimmune diseases such as AIT [38, 39]. Dysbiosis can activate immune cells and produce pro-inflammatory cytokines as the gut microbiota plays a crucial role in regulating the immune system [39, 40]. Obesity can also hinder human self-tolerance mechanisms by promoting pro-inflammatory processes and reducing the number of Bregs and Tregs, leading to increased Th17 and Th1 cells that create an environment favorable for autoimmune disorders [14]. In Hashimoto's thyroiditis, adipose tissue enlargement in patients with high BMI results in the release of monocyte chemoattractant protein (MCP) from podocytes and monocytes that play a feedback role based on thyroid hormones [24]. Adipokines, which are biologically active substances derived from adipocytes, are involved in regulating metabolism, hormonal balance, and immune response [41, 42]. Obesity exacerbates the severity of certain autoimmune diseases and reduces treatment response. Moreover, excess body fat may promote the progression of autoimmune diseases [24]. Therefore, we hypothesize that AIT in high BMI patients might develop via similar mechanisms. Obesity could contribute to AIT development, which in turn may result in autoimmune disease progression. The thyroid and adipose tissues may interact bilaterally since adipocytes express functioning thyroid-stimulating hormone (TSH) receptors, and TSH binding to its receptor on brown adipocytes can stimulate thermogenesis. A schematic diagram to illustrate this concept is shown in Fig. 6.

Our study identified multiple SNPs associated with both BMI and immune function and inflammatory signaling, suggesting that immune function and signaling pathways may mediate the relationship between BMI and AIT. These results support our proposed hypothesis. However, the question of causality between BMI and AIT remains unresolved since not all patients with AIT are obese.

Previous studies have indicated that obesity is positively correlated with AIT, with obese individuals having a higher prevalence of AIT compared to non-obese individuals. The mechanism behind this correlation is not fully understood, but it is hypothesized that increased leptin production in obese individuals may promote Th17 cell differentiation, thus contributing to the development of AIT. Using MR, our findings provided evidence supporting a causal relationship between BMI and AIT. This provides stronger support than the mere correlation observed in observational studies and meta-analyses [43‐45]. Our results are consistent with these findings, indicating that a high BMI may increase the risk of AIT. However, MR studies have limitations since they assume that instrumental variables are only related to the outcome through the risk factor being studied and require robust genetic instrumentation and large sample sizes [46].

In this study, we conducted heterogeneity and sensitivity analyses to assess the reliability and validity of our findings. Firstly, we assessed the level of heterogeneity among SNPs by employing the Q-statistic based on the IVW and MR-Egger methods. Our results indicated that there was no significant heterogeneity among SNPs, which strengthens the reliability of our findings. This was also confirmed by funnel plots, which did not show trends towards bias. Secondly, we performed a leave-one-out sensitivity analysis to explore if any single SNP was responsible for influencing the MR findings. Our results showed that the links were not reliant on a particular SNP. Thirdly, we employed the MR-Egger regression approach to detect directional pleiotropy, which can negatively impact MR techniques. We used the intercept of MR-Egger regression to provide free-of-bias estimates of causality. The results indicated that there was no evidence of genetic pleiotropy affecting our outcomes.

However, it should be acknowledged that our MR analysis may have certain limitations. Firstly, the accuracy of genetic instruments is essential for the validity of the MR approach. Despite using a significant number of independent SNPs as IVs, there remains the possibility of weak instrument bias. Secondly, one limitation of this study is the relatively small number of cases of AIT in the GWAS summary dataset. This may limit the statistical power to detect significant associations in some of the analyses. Additionally, it is worth noting that only the IVW method showed significant results, while other MR methods did not provide consistent findings. Therefore, caution should be exercised when interpreting the results as the evidence is not entirely convincing across all MR methods used in this study. Lastly, population stratification may influence the findings of MR analysis, potentially causing spurious associations. Although the MR approach minimizes the risk of confounding, it cannot fully eliminate the possibility of residual confounding, which might impact the results. Therefore, the interpretation of the results of this study should be cautious, and further research is necessary to confirm the causal relationship between BMI and AIT.

Our study underscores the significance of sample size selection in GWAS and how utilizing the European population for both exposure and outcome groups can address some of the limitations in MR. By using the European population, we mitigated the impact of genetic heterogeneity and population structure on the validity of our study. Our analysis involved a substantial sample size, using GWAS summary statistics data from 339,224 individuals of European ancestry as the exposure factor and 187,928 individuals of European descent from the autoimmune thyroiditis GWAS summary statistics as the outcome factor, which significantly increased the study's power.

The results of this study have clinical significance in terms of identifying a possible causal link between BMI and AIT. If confirmed through further investigation, healthcare providers may need to consider BMI as a risk factor for AIT and screen patients with high BMI for AIT. From a public policy perspective, our study may provide additional evidence for promoting healthy weight management practices to prevent the development of AIT. However, it should be noted that the effect size of possible interventions cannot be determined from this study alone, highlighting the need for further research to investigate the impact of interventions such as weight loss on the risk of developing AIT.

In summary, our MR analysis suggests that there may be a causal link between BMI and an increased risk of AIT, indicating that obesity could contribute to the onset of AIT. Our results provide a basis for future research into the association between BMI and AIT.