Polygenic scores and antidepressant treatment outcomes in major depression: a critical integrative review.

Major depressive disorder (MDD) shows a large heterogeneity in antidepressant treatment outcome, a variability explained in part by common-variant liability captured by polygenic scores (PGSs). We performed a review of PubMed- and Google Scholar-indexed studies that related any PGS to treatment response, remission or resistance in adult MDD. Thirty-nine investigations met inclusion criteria. MDD-PGS emerged as the most consistently replicated predictor, each standard deviation increase raising non-remission odds by ∼10-14 % across six European trials and two biobank analyses. Schizophrenia genetic liability (SCZ-PGS) was uniformly detrimental under monoaminergic monotherapy (OR∼2.2 for top quintile) while possibly associated with improvement to electroconvulsive therapy (ECT) or lithium augmentation. Bipolar disorder PGS predicted faster lithium response (HR∼1.5), while attention-deficit/hyperactivity-disorder PGS was reliably associated with treatment-resistant depression (TRD) in multiple studies. PGSs for coronary-artery disease and stroke similarly reduced antidepressant efficacy and doubled TRD odds. Across traits, individual PGSs generally explained <1 % of outcome variance, but multi-omic models integrating pharmacokinetic alleles, inflammatory markers and clinical covariates reached 8-12 %. Methodological barriers include ancestry bias (>85 % European), heterogeneous phenotyping and small discovery cohorts. Even so, PGSs may support pathway-specific treatment modifiers and justify prospective trials of genotype-guided lithium, ECT or augmentation with drugs modulating inflammation-metabolism. Scalable precision psychiatry will require multi-ancestry discovery, harmonised longitudinal outcomes and equitable implementation frameworks, but current evidence signals a realistic path from genomic insight to stratified care.