GENOME-WIDE ASSOCIATION STUDY OF ADOLESCENT-ONSET DEPRESSION

Abstract

Depression is a complex trait disease which emerges most often and most severely during adolescence. Early-onset depression correlates with late-onset depression and has a three-fold higher single nucleotide polymorphism (SNP)-heritability. Early or adolescent depression is also well-predicted by polygenic risk scores (PRS) of adult major depressive disorder (MDD). Though genetic signal likely exists, attempts to determine variants associated with adolescent depression have been unfruitful due to phenotype heterogeneity and the lack of power available in prospective adolescent cohort sample sizes. To overcome the power problem whilst maintaining a stable phenotype, our study proposes to, i) leverage genetic data from both prospective and retrospective cohorts and, ii) restrict the phenotype to self-report symptoms only.

We perform a Genome-Wide Association Study (GWAS) of adolescent-onset depression leveraging approximately 180,000 individuals from over 20 cohorts in the Psychiatric Genomics Consortium (PGC) and Early Genetics and Lifecourse Epidemiology (EAGLE) consortium. Cohorts span 10 countries with diverse ancestries (including African, American-admixed, East Asian, European, Middle Eastern, and South Asian). We first analyse prospective and retrospective cohorts independently before combining all cohorts in a meta-analysis. We perform tissue expression analysis and compare results with findings in the GWAS catalogue.

Current results from 107,721 individuals (14 891 cases and 92 830 controls) have revealed 7 novel independent genome-wide significant SNPs in European ancestry. We determined a SNP-heritability (SE) of 0.053 (0.006). Genetic correlation (SE) between meta-analysed cohorts with the latest MDD GWAS summary statistics was 0.79 (0.04) and between prospective and retrospective cohorts was 0.52 (0.05). Gene-expression analysis determined tissue enrichment in the cortex, cerebellum and hippocampus. Leading SNPs of the association with adolescent-onset depression overlapped with results from previous GWASs of depression, neuroticism and wellbeing in adults.

Association analyses from the remaining contributing cohorts are ongoing. Current results already provide novel genetic associations, and we expect the addition of approximately 70,000 more individuals to further increase power and discovery across ancestries. Once all samples are received, we will derive PRS for out-of-sample prediction across ancestries, test genetic correlation with other traits, use genomic structural equation modelling to investigate shared architecture, run colocalization for shared trait variants and perform Mendelian randomisation to identify causality. We expect that biological insights, potentially hidden within the larger and more heterogeneous adult population, could be uncovered in the more genetically heritable adolescent-onset group. Investigating the genetic architecture of the adolescent-onset depression subtype could facilitate early intervention, risk prediction and stratified treatment.