Personality improves prediction of the onset of common mental disorders

Common mental disorders (CMDs) affect hundreds of millions of people globally yet remain difficult to predict before onset. Personality traits may offer a scalable solution given their temporal stability, assessment efficiency, and conceptual links to psychopathology. Using elastic net regularization with temporally independent validation, we tested whether personality traits prospectively predict new-onset CMDs, using 20 years of registry-verified diagnoses among Estonian Biobank participants (N = 63,672). Among adults disorder-free for 17+ years, specific personality traits (nuances; single ques-tionnaire items) predicted new-onset CMDs up to 3 years later (MAUC = 0.71; range 0.68–0.79), out-performing broader trait domains (MAUC = 0.67; range 0.62–0.76) and demographic variables (MAUC = 0.59; range 0.53–0.62). Predictions generalized across disorders, indicating that traits capture shared transdiagnostic vulnerability rather than diagnosis-specific risks. Sparse models including 15–20 top-ranked nuances achieved near-peak accuracy, supporting the feasibility of brief population-level risk assessments for early CMD detection. These findings position personality traits among the strongest and most practical predictors of new-onset mental health problems.