Evaluating the causal effects of circulating metabolites on major depression disorder

Abnormal circulating metabolite profiles have been reported in major depressive disorder (MDD), but whether specific metabolites contribute to MDD or reflect consequences of this disorder remains uncertain.

We used two-sample Mendelian randomization (MR) to evaluate 233 genetically proxied circulating metabolites in relation to MDD (294,322 cases and 741,438 controls). We additionally conducted reverse MR with MDD as the exposure and validated the set of 56 discovery-stage forward associations using an independent FinnGen R12 dataset with MDD as an outcome (N = 494,164). Random-effects inverse-variance weighted (IVW) estimates were primary, with weighted-median, MR-Egger, heterogeneity, and directional-pleiotropy analyses used for sensitivity assessment.

In the discovery analysis, a total of 25 metabolites were inversely associated with MDD (odds ratio [OR], 0.93-0.98) and further 31 were positively associated with MDD (OR, 1.03-1.07). In FinnGen, all 56 IVW point estimates were directionally concordant; a total of 41 associations had P < 0.05 and 35 of them remained significant after false discovery rate (FDR) correction. Among the 35 FDR-significant associations, weighted-median and MR-Egger estimates were directionally concordant with IVW. Reverse MR identified two FDR-surviving associations: genetic liability to MDD was associated with a lower omega-6 fatty acids-to-total fatty acids ratio (OR, 0.93; 95% confidence interval [CI], 0.89-0.96; FDR = 0.011) and higher glycoprotein acetylation (OR, 1.10; 95% CI 1.05-1.15; FDR = 0.011).

Genetically proxied circulating metabolites showed reproducible associations with MDD, while reverse effects were supported by only two metabolic traits. Those findings may help prioritization of candidate biomarkers for risk stratification in patient cohorts and generate hypotheses for mechanistic research. Clinical utility requires confirmation in well-designed randomized trials.