Background: Antidepressants are considered one of the first-line intervention for major depressive disorder (MDD), but treatment responses vary and are often unsatisfactory. It was reported that only 35% MDD achieve remission with initial antidepressant treatment and therefore more rigorous intervention, or different therapeutic approaches are needed to achieve better outcomes. Recent research has demonstrated that that differential antidepressant response is partly attributable to genetic polymorphism and associated with genetic liability to other mental health conditions. Methods: The current study aimed to investigate associations between multiple polygenic scores (PGS) and the early treatment response to antidepressants using data from the International SSRI Pharmacogenomics Consortium (ISPC). Specifically, we selected traits with publicly available GWAS which have been reported to predict treatment outcomes of MDD and traits that have implications in antidepressant metabolism. Because the ISPC cohorts include independent participants of European and East Asian ancestries, we applied deep learning algorithms to derive multivariable polygenic models separately. We then compare model performance and predictor differences between the ancestral groups. Results: Among all the preselected PGS, PGS for schizophrenia, diastolic blood pressure, systolic blood pressure, BMI and sleep duration were significantly associated with treatment response in participants of Asian ancestry. More PGSs significantly associated with treatment response were found participants of European ancestry. Discussion: Most of the available GWAS are derived from participants of European ancestry and therefore differences in polygenic score prediction may merely reflect different genetic architectures rather than shared disease liability. More data from different ancestral groups are needed to better understand potential mechanism underlying antidepressant responses. Findings are invaluable for treatment planning to improve patients' outcomes.
Date:
2021-10
Relation:
European Neuropsychopharmacology. 2021 Oct;51:E159.
