Many observational studies have reported an association between thyroid function and type 2 diabetes mellitus (T2DM), but the causal relationship remains unclear. This study aims to examine the association between thyroid function and T2DM by using the National Health and Nutrition Examination Survey (NHANES) database 2007–2012 and Mendelian randomization (MR) analysis.

A cross-sectional analysis was performed utilizing data from the NHANES, and logistic regression was employed to investigate the relationship between thyroid function and T2DM. Genome-wide association studies (GWAS) data of thyroid function and T2DM were obtained from the website of the IEU OpenGWAS project and the Thyroidomics Consortium. Single nucleotide polymorphisms (SNPs) strongly related to thyroid function were used as instrumental variables. Moreover, the inverse-variance weighting MR-Egger regression model and weighting median method were used to analyze the causal effect of thyroid function on T2DM. Cochran’s Q test, the MR-Egger intercept test, and “leave one out” cross-validation were applied to evaluate the pleiotropy and heterogeneity levels of the above included SNPs.

In the NHANES study, no association was identified between thyroid function and the risk of T2DM in a fully adjusted model. A total of 17, 31, 7, and 79 SNPs were identified in the cohorts for free thyroxine (FT4), thyroid-stimulating hormone (TSH), hyperthyroidism, and hypothyroidism, respectively. The ORs with 95% confidence intervals (95% CIs) derived from MR-Egger regression, WME, and inverse variance weighting (IVW) for TSH and T2DM were 0.86 (0.75–0.99), 0.91 (0.85–0.98), and 0.94 (0.89–0.99), respectively. For hypothyroidism and T2DM, the ORs (95% CI) from MR-Egger regression, WME, and IVW were 5.71 (1.76–18.53), 3.33 (1.57–7.04), and 2.60 (1.50–4.49), respectively. Additionally, Cochran’s Q test results demonstrated no heterogeneity among the SNPs associated with TSH and T2DM, while heterogeneity was observed among the SNPs related to hypothyroidism and T2DM (Q = 133.40, P < 0.05), necessitating the use of the IVW random effects model.

The findings of the present study point to a potential causal relationship between TSH, hypothyroidism, and T2DM. Further research is needed to explore the relationship between thyroid function and T2DM.