POLYGENIC RISK SCORES FOR A SPECTRUM OF PSYCHIATRIC OUTCOMES ARE ALSO ASSOCIATED WITH DEPRESSION TRAJECTORIES FROM CHILDHOOD TO EARLY ADULTHOOD: FINDINGS FROM THE AVON LONGITUDINAL STUDY OF PARENTS AND CHILDREN

Abstract

Depression is a complex and multifactorial disorder that has genetic and environmental influences. Genome-wide association studies have shown that common genetic variants are implicated in depression. These common variants, when combined into polygenic risk scores, are associated with depression case status, severity and age of onset. However, less is known about how genetic risk affects change in depression symptoms longitudinally. Furthermore, psychiatric disorders are comorbid and recent studies have shown genetic risk is shared between them, but the association between this shared polygenic risk and how depression manifests and changes over time is not yet understood.

We used data from the Avon Longitudinal Study of Parents and Children (ALSPAC). Self-reported depressive symptoms were assessed on 10 occasions between the ages of 10 and 25 using the 13-item Short Mood and Feelings Questionnaire. Polygenic risk scores (PRS) for major depressive disorder (MDD), anxiety (ANX), neuroticism (NEU), and schizophrenia (SCZ) were computed with PRSice-2 using summary statistics from recent genome-wide associations studies, in which ALSPAC was not included. Additionally, we used genomic structural equation modelling (GSEM) to create a multi-trait PRS of MDD, ANX, NEU, SCZ, bipolar disorder, autism spectrum disorder, and attention deficit hyperactivity disorder, to capture the spectrum of psychopathology and explored how this genetic risk score was associated with depression trajectories.

We modelled depression trajectories using generalised additive models, with age as the time metric, and included sex, age-sex interaction, PRS, age-PRS interaction, and the first 10 principal components as predictors. We ran separate models for each PRS.

Depression trajectories for those in the top and bottom deciles of MDD and NEU PRS start to show divergence around mid- to late-adolescence with higher genetic risk associated with worse trajectories. With the multi-trait PRS, differences emerge as early as childhood, again with higher genetic risk indicative of worse trajectories. In these three models, the separation between the trajectories then continues to increase into adulthood. No clear pattern of separation was observed with the ANX or SCZ PRS.

These findings suggest that psychiatric PRS are associate with (and may influence) the longitudinal course of depressive symptoms from childhood into early adulthood. The multi-trait PRS was superior to PRS of individual psychiatricdisorders in delineating depression trajectories in association with genetic risk. One interpretation is that a spectrum of psychiatric genetic risk could underpin developmental differences in depression trajectories.