Abstract
Objective
The microbiome is emerging as an important component of brain–body communication and may influence treatment outcomes in psychiatric disorders. This naturalistic cohort study prospectively evaluated whether baseline microbiome abundance and composition of circulating bacterial extracellular vesicles (EVs) are associated with subsequent response to selective serotonin reuptake inhibitors (SSRIs) in patients with obsessive-compulsive disorder (OCD).
Methods
Thirty-two drug-naïve or medication-free patients with OCD were enrolled and assessed at baseline and at 12 weeks following the initiation of SSRI pharmacotherapy. Serum EV microbiome profiling was performed at baseline, before any pharmacological treatment was commenced, ensuring that all biological measurements reflect a pharmacologically unconfounded pre-treatment state. Good responders were defined as patients with a 35% or greater reduction in the Yale-Brown Obsessive-Compulsive Scale score at 12 weeks, while poor responders were defined as those with less than a 35% reduction. Baseline microbiome abundance and composition of circulating bacterial EVs in serum were compared between these groups.
Results
Of the 32 participants, 29 completed the 12-week study. Based on the treatment response at 12 weeks, the OCD group was classified into 15 good responders (5 females) and 14 poor responders (7 females). The two groups did not differ significantly in age, sex, age of onset, baseline Y-BOCS or MADRS scores, or medication type. No significant differences in alpha or beta diversity were observed between the groups. Taxonomic analysis of pre-treatment serum EV profiles revealed a nominally higher relative abundance of the genus
Acinetobacter
in patients who subsequently showed poor response compared to good responders. However, this association did not remain statistically significant after correction for multiple comparisons, age adjustment, or compositionally-aware reanalysis.
Conclusion
These results provide no statistically defensible evidence of a taxon-level microbiome difference between response groups, and the nominal
Acinetobacter
finding should not be regarded as a validated biomarker. Global diversity measures also did not distinguish the response groups. These findings are hypothesis-generating and support the feasibility of pre-treatment serum EV-associated bacterial profiling in OCD. Larger, adequately powered studies incorporating compositionally appropriate statistical methods and rigorous contamination-control procedures are warranted to determine whether reproducible microbiome–treatment response associations exist.