Background: Accumulating evidence has suggested that patients with depression have distinct gut microbiome profiles, and antidepressant use may alter the microbial composition. However, large population-based data on how gut microbiota can, in turn, influence the efficacy of antidepressants is lacking.

Methods: To address this research gap, we leveraged stool metagenomic and metabolomic data from the ASPREE-XT Microbiome Sub-Study (n=2,158, age 81.6±3.2 years), an extension observational study after four years of the low-dose aspirin trial (ASPREE). We identified antidepressant use based on ATC code N06 and assessed depression symptoms using the Center of Epidemiologic Studies Depression Scale (CESD)-10. Depression was defined as CESD-10³8, or admission to hospital for greater than 24 hours, where depression was one of the primary reasons for admission.

Results: The overall taxonomic composition moderately varied according to antidepressant use and depression status (PERMANOVA R2=0.3% P=0.001). Among antidepressant users, a lower relative abundance of Faecalibacterium sp CLA AA H233 and GGB3653 SGB 4964 was associated with worse depressive symptoms, that is, worse efficacy of the antidepressants. In contrast, compared to those who were responsive to antidepressants (i.e., no depression), increased relative abundance of 3-dehydroquinate biosynthesis II and chorismite biosynthesis II pathways were associated with less responsiveness to antidepressants (i.e., had depression vs. no depression). 3-dehydroquinate and chorismate are key metabolites in the shikimate pathway involved in the synthesis of folate and aromatic amino acids, which are precursors for neurotransmitters crucial for mood regulation (e.g., serotonin and dopamine). In line with our metagenomic results, our metabolomics analysis revealed interactions between antidepressant use and the metabolomics profiles; the positive correlation between antidepressants and CESD-10 score was attenuated among participants with a metabolomic profile enriched in amino acids. Additionally, the prevalence of anti-inflammatory microbial metabolites that have been implicated in brain function and neuroprotection – N-Stearoyl sphingosine and ectoine – was reduced among participants who were less responsive to antidepressants than those who were responsive.

Conclusions: Our study provides valuable evidence on the role of gut microbiota in individual responsiveness to antidepressants, shedding light on personalized microbiota-targeted therapies to optimize treatment efficacy. Future longitudinal studies are warranted to understand the causal mechanisms.