Depression genetic risk score is associated with anhedonia-related markers across units of analysis

Abstract

Investigations of pathophysiological mechanisms implicated in vulnerability to depression have been negatively impacted by the significant heterogeneity characteristic of psychiatric syndromes. Such challenges are also reflected in numerous null findings emerging from genome-wide association studies (GWAS) of depression. Bolstered by increasing sample sizes, recent GWAS studies have identified genetics variants linked to MDD. Among them, Okbay and colleagues (Nat. Genet. 2016 Jun;48(6):624-33) identified genetic variants associated with three well-validated depression-related phenotypes: subjective well-being, depressive symptoms, and neuroticism. Despite this progress, little is known about psychopathological and neurobiological mechanisms underlying such risk. To fill this gap, a genetic risk score (GRS) was computed from the Okbay's study for a sample of 88 psychiatrically healthy females. Across two sessions, participants underwent two well-validated psychosocial stressors, and performed two separate tasks probing reward learning both before and after stress. Analyses tested whether GRS scores predicted anhedonia-related phenotypes across three units of analyses: self-report (Snaith Hamilton Pleasure Scale), behavior (stress-induced changes in reward learning), and circuits (stress-induced changes in striatal reward prediction error; striatal volume). GRS scores were negatively associated with anhedonia-related phenotypes across all units of analyses but only circuit-level variables were significant. In addition, the amount of explained variance was systematically larger as variables were putatively closer to the effects of genes (self-report < behavior < neural circuitry). Collectively, findings implicate anhedonia-related phenotypes and neurobiological mechanisms in increased depression vulnerability, and highlight the value of focusing on fundamental dimensions of functioning across different units of analyses.

Fig. 1. Coronal slices showing the functional and structural regions-of-interest (ROIs) considered for the analyses.

a Functional (bilateral) nucleus accumbens ROI (yellow color) previously found to be associated with stress-induced RPE reductions. b Functional (bilateral) putamen ROI (red color) emerging from a meta-analysis of RPE studies in healthy controls. c Structural (bilateral) nucleus accumbens (yellow color) and putamen (red color) ROI. For analyses, structural ROIs were extracted for each participant individually and entered into analyses. For panel a, circles highlight the nucleus accumbens ROI that was considered for analyses

Fig. 2

Histogram of genetic risk scores across 83 female participants

Fig. 3

Scatterplots of associations between depression-related genetic risk scores and a bilateral nucleus accumbens stress-induced RPE changes (ΔR² = 0.065, p = 0.033); b bilateral putamen stress-induced RPE changes (ΔR² = 0.074, p = 0.027); c bilateral nucleus accumbens volume (ΔR² = 0.095, p = 0.028); and d bilateral putamen volume (ΔR² = 0.095, p = 0.009). For the anhedonia markers, residualized values are plotted (removing variance associated with ancestry-related variables (i.e., PC1 and PC2 scores)). For the volumetric data, intracranial volume was calculated to correct for interindividual differences in total brain size

Fig. 4

Amount of variance explained by the genetic risk score for anhedonia-related markers across units of analysis (self-report → behavior → brain circuits)