Effect of metabolites levels on type 2 diabetes mellitus and glycemic traits: a Mendelian randomization study.

OBJECTIVE To assess the causal associations of plasma levels of metabolites with type 2 diabetes mellitus (T2DM) and glycemic traits. METHODS Two-sample mendelian randomization (MR) was conducted to assess the causal associations. Genetic variants strongly associated with metabolites at genome-wide significance level (P < 5 × 10 -8) were selected from public GWAS, and SNPs of Outcomes were obtained from the DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) consortium for T2DM and from the Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC) for the fasting glucose, insulin and HbA1c. The Wald ratio and inverse-variance weighted methods were used for analyses, and MR-Egger was used for sensitivity analysis. RESULTS The β estimates per 1 SD increasement of arachidonic acid (AA) level was 0.16 (95% CI: 0.078, 0.242; P<0.001). Genetic predisposition to higher plasma AA levels were associated with higher FG levels (β 0.10 [95%CI: 0.064, 0.134], P<0.001), higher HbA1c levels (β 0.04 [95%CI: 0.027, 0.061]) and lower FI levels (β -0.025 [95%CI: -0.047, -0.002], P=0.033). Besides, 2-hydroxybutyric acid (2-HBA) might have positive causal effect on glycemic traits. CONCLUSIONS Our findings suggest that AA and 2-HBA may have the causal associations on T2DM and glycemic traits. It is beneficial for clarifying the pathogenesis of T2DM, which would be valuable for early identification and prevention for T2DM.