Assessing the overlap of personality traits and internalizing psychopathology using multi-informant data: Two sides of the same coin?

Abstract

Personality and psychopathology share a hierarchical dimensional structure, developmental trajectories and correlations with varied outcomes. However, quantifying the extent and details of their direct empirical overlap has been hindered by over-reliance on self-reports and broad construct domains. Using multi-method data, we estimated the Big Five personality domains’ and nuances’ “true” correlations (rtrue) with, and true predictive accuracy (rtruePRED) for, various psychopathology state domains, free of single-method and occasion-specific biases, random error, and direct content overlap. Our sample included Estonian Biobank participants (N = 16 226) who completed comprehensive personality and psychopathology questionnaires, and whose personality traits were also rated by close informants. Personality nuances out-predicted the Big Five domains for psychopathology, with rtruePRED = 0.31…0.58 for specific psychopathology domains of distress, fear, inattention, hyperactivity, insomnia and fatigue, and rtruePRED = 0.52 for the p-factor. For the Big Five, neuroticism was the strongest correlate of distress, fear and fatigue (rtrue = 0.08…0.29), while inattention (rtrue = –0.56) and insomnia (rtrue = 0.12) correlated mostly with conscientiousness, and hyperactivity with agreeableness (rtrue = –0.20). Associations based on self-reports alone were weaker. These findings highlight the need for multi-rater assessments and finer-grained personality and psychopathology dimensions to reveal the full extent and details of personality-psychopathology overlap, which is likely stronger than self-report data suggest. Yet psychopathology is far from being empirically redundant with personality traits.