T32 - Investigation Of Familial Co-Aggregation And Genetic Correlation Of Age At Onset And Episodicity With Personality Dimensions In Major Depressive Disorder

Abstract

Background Personality dimensions have been shown to be heritable and are linked to liability of psychiatric disorders. Neuroticism, in particular, has a pervasive association with major depressive disorder (MDD) partially explained by shared genetic effects. Neuroticism predicts the clinical presentation and comorbidity of MDD, the onset of new episodes and affects its course and response to treatment. Age at onset (AAO) and episodicity, widely considered as proxies of disease severity, are both familial and heritable (Ferentinos et al. 2015) hence providing clues to dissecting MDD heterogeneity. A recent meta-analysis of GWAS of neuroticism by the Genetics of Personality Consortium (GPC; de Moor et al. 2015) pinpointed a genome-wide significant SNP in MAGI1 previously associated with episodicity in MDD (Ferentinos et al. 2014). The aims of this study were: first, to investigate the familiality of personality dimensions and their co-aggregation with AAO and episodicity within families of MDD siblings; second, to assess the SNP-heritability of personality dimensions and their genetic correlations with AAO and episodicity in unrelated genotyped MDD subjects. Methods Personality dimensions (neuroticism-N, extraversion-E, psychoticism-P) were extracted from the Eysenck Personality Questionnaire (EPQ). For our first aim, we used 1498 subjects with recurrent MDD from the DeNt affected siblings study (691 families with 2-5 affected full siblings). Familiality of EPQ dimensions was investigated with linear mixed models (LMMs). We then created a dataset of all possible sibpairs and calculated cross-trait within-subject and cross-trait cross-siblings associations of AAO and episodicity with each EPQ dimension. Analyses were performed with LMMs or negative binomial generalized linear mixed models (GLMMs) for AAO or episode frequency, respectively and applying a family size weight to adjust for non-independence of sibpairs (Suarez & Van Eerdewegh, 1984). For our second aim, we used 2695 unrelated MDD cases from the RADIANT studies. AAO, episodicity and EPQ scores were similarly analyzed in LMMs or GLMMs as appropriate. Derived residuals were then used to estimate SNP-heritabilities of EPQ scores or their pairwise genetic correlations with AAO and episodicity in GREML bivariate analyses with GCTA software. Results In the DeNt dataset, all personality dimensions were familial, with intraclass correlation coefficients (ICCs) of 0.21 (SE 0.04), 0.11 (SE 0.03) and 0.17 (SE 0.04) for N, E and P scores, respectively. AAO was negatively associated with neuroticism (p Discussion Neuroticism and episodicity co-aggregate within families of siblings with MDD. Larger samples are required to test whether this familial covariation is due to shared additive genetic effects. As a next step, the predictive accuracy of GPC personality polygenic scores on AAO and episodicity in the RADIANT study is being tested.