Genetically predicted selenium is negatively associated with serum TC, LDL-C and positively associated with HbA1C levels

Pervious epidemiological evidence on the associations of selenium, zinc with lipid profile and glycemic indices was contradictory. The aim of this study was to investigate whether selenium and zinc were casually associated with lipid profile and glycemic indices using mendelian randomization (MR) analysis. A two-sample MR was used to evaluate the causal-effect estimations. Summary statistics for selenium, zinc, lipids and glycemic indices were retrieved from previous large-scale genome-wide association study (GWAS). Single nucleotide polymorphisms (SNPs) that independently and strongly associated with the selenium and zinc were selected as the instrumental variables. The casual estimates were calculated using inverse variance weighted method (IVW), with weighted median, MR-Egger, and MR-PRESSO test as sensitivity analysis, respectively. In the standard IVW analysis, per SD increment in selenium was associated with an 0.077 mmol/L decrease of TC (95 %CI: −0.102,−0.052) and 0.074 mmol/L of LDL-C (95 %CI: −0.1,−0.048). Suggestive casual associations were found between selenium and insulin or HbA1c. With IVW method, per SD increase in selenium was associated with an 0.023 mmol/L increase of insulin (95 %CI: 0.001,0.045), and an 0.013 mmol/L increase of HbA1c (95 %CI: 0.003,0.023). The results were robust in the sensitivity analysis. Zinc was not casually associated with any of lipid and glycemic markers. Our MR analysis provides evidence of the potential causal effect of Se on beneficial lipid profile, including decreased TC and LDL-C. Furthermore, suggestive casual evidence was suggested between Se and increased serum HbA1c levels. Careful consideration is required for the protective effects of Se supplementation. No casual-effect association was found between Zn and any indices of the lipid and glucose parameters.