Common and specific determinants of 9-year depression and anxiety course-trajectories: A machine-learning investigation in the Netherlands Study of Depression and Anxiety (NESDA).

Background: Given the strong relationship between depression and anxiety, there is an urge to investigate their shared specific long-term course determinants. The current study aimed identify compare main determinants of 9-year trajectories combined pure anxiety symptom severity. Methods: Respondents with a 6-month and/or diagnosis (n=1,701) provided baseline data on 152 sociodemographic, clinical biological variables. Depression severity assessed at baseline, 2-, 4-, 6- follow-up, were used data-driven course-trajectory subgroups for general psychological distress, depression, scores. For each outcome (class-probability), Superlearner (SL) algorithm identified optimally weighted (minimum mean squared error) combination machine-learning prediction algorithms. outcome, top in SL by determining variable-importance correlations SL-predicted observed (ρpred) calculated. Results: Low high (ρpred: 0.41-0.91, median=0.73) found. In SL, important distress age, young age onset, respiratory rate, participation disability, somatic disease, low income, minor depressive disorder mastery score. severity, similar Specific included several types healthcare-use, pure-anxiety arousal distress. Limitations: Limited sample size machine learning. Conclusions: depression- anxiety-severity are mostly shared. Domain-specific exceptions healthcare use course.