Abstract and Introduction
Abstract
Objective: The impact of selenium (Se) on human thyroid function remains unclear, with inconsistent results from recent epidemiological studies. Moreover, the observed associations are prone to bias due to potential confounding and reverse causation. Mendelian randomization (MR) analysis facilitates the large minimization of biases produced by environmental and lifestyle influences, providing unconfounded estimates of causal effects using instrumental variables. We aim to examine the association between Se concentrations and human thyroid function using a two-sample MR analysis.
Design and Methods: Genetic instruments for Se concentrations, including toenail and blood (TAB) and blood Se concentrations, were identified from a genome-wide association study (GWAS) of blood Se (n = 5477) and toenail Se levels (n = 4162). GWAS summary statistics on thyroid phenotypes were downloaded from the ThyroidOmics consortium, including thyroid-stimulating hormone (TSH) (n = 54,288), free thyroxin (FT4) (n = 49,269), hypo (n = 53,423), and hyperthyroidism (n = 51,823). The MR study was conducted using the inverse-variance weighted (IVW) method, supplemented with the weighted median and the mode-based method.
Results: Genetically determined TAB Se was negatively associated with FT4 (β = −.067; 95% confidence interval [CI] = −0.106, −0.028; p = 0.001) using the IVW analyses, as well in the additional analyses using the weighted median and weighted-mode methods. No evidence in heterogeneity, pleiotropy or outlier single-nucleotide polymorphisms was detected (all p > 0.05). Suggestive casual association between increased genetically determined TAB Se concentrations and decreased hypothyroidism risk was found by the IVW method (odds ratio [OR] = 0.847; 95% CI = 0.728, 0.985; p = 0.031). The causal effect of TAB Se on FT4 was observed in women (β = −.076; 95% CI = −0.129, −0.024; p = 0.004). However, the influence of genetically determined higher Se concentrations on TSH levels and hyperthyroidism revealed insignificance in the primary and sensitivity analyses.
Conclusions: The present MR study indicated that high Se concentration enable the decreasing of FT4 levels, and the effects of Se concentrations on FT4 remain sex-specific.
Introduction
Selenium (Se) is an essential trace element that maintains multiple normal physiological functions in human bodies.[1] It is obtained from a wide variety of food sources, including grains, cereals, seafood organ meats, dairy products, and vegetables. Toenail and blood (TAB) levels of Se exposure as biomarkers are influenced not only from dietary Se intake, but also by broader socioeconomic, geographic and physiological factors such as age, gender, geographic location and smoking status.[2]
The thyroid gland contains the highest tissue level of Se and expresses a variety of selenoproteins (e.g., thioredoxin reductases, deiodinases and glutathione peroxidases), involving protection against oxidative damage and metabolism of thyroid hormones.[3] Hence, an appropriate Se intake is required to protect thyroid health and reduce the risk of thyroid diseases. Despite the expected link between Se and thyroid function, previous epidemiological studies on the associations between Se and thyroid hormones were inconclusive. Some studies found that low plasma levels of Se affect glutathione peroxidase activity, total triiodothyronine (T3), free thyroxin (FT4) and thyroid-stimulating hormone (TSH) levels in humans living,[4,5] while others did not find significant correlations between Se status and thyroid metabolism.[6,7] Several studies also examined the effect of Se supplementation on thyroid hormones, but not all trials have reported positive health benefits.[2,8,9]
Iodothyronine deiodinases are selenoproteins that catalyze the deiodination of prohormone thyroxine (T4) to the biologically active triiodothyronine by peripheral conversion. The excess thyroid hormones can lead to hyperthyroidism, whereas deficiency may cause hypothyroidism. Andrade et al.[10] evaluated the dietary Se intake using a previously validated food frequency questionnaire, and found that lower Se intake was inversely associated with subclinical hypothyroidism. A meta-analysis revealed that serum/plasma Se levels in subjects with hypothyroidism were statistically lower compared with those of healthy controls.[11] Randomized controlled trials (RCTs) of Se supplementation and thyroid diseases are currently limited and showed conflicting results.[12–14] In this case, it is uncertain whether the association is causal or biased by reverse causality or residual confounding.
Nowadays, the Mendelian randomization (MR) approach has been widely used in analyses to provide more robust causal evidence on exposures to clinical outcomes.[15] Based on the principle of independent assortment during meiosis, the MR approach is a type of instrumental variable (IV) analysis that uses genetic variants, such as single-nucleotide polymorphisms (SNPs), to infer the causality of the exposure for an outcome. The calculated correlation of a genetically predicted exposure with an outcome minimizes the bias produced by environmental influences since alleles are randomly allocated at conception. Unlike RCTs, where periods of exposure are typically brief, MR studies give the effects of lifetime exposures, and can therefore show longer-term effects.
As the role of Se in thyroid function using MR-based studies is largely unknown, we analyzed the genome-wide association studies (GWAS) data set using a two-sample MR approach to explore the association between the predicted Se concentrations and thyroid-related phenotypes.
Clin Endocrinol. 2023;98(6):813-822. © 2023 Blackwell Publishing
