The Interplay of Personality Functioning and Affect-Event Dynamics in Predicting Future Impairment and Depression – A Large Mobile Mental Health Ambulatory Assessment Study

Since the introduction of the dimensional assessment of personality functioning (PF) in DSM-5 and ICD-11, impairments in PF have consistently been related to transdiagnostic risk-factors for mental health. However, ecological momentary assessment investigating PF in relation to affect event dynamics and other psychopathology are scarce. Leveraging ecological momentary assessment (EMA) data from 16,038 mental health app users, this study examines how within-person affect-event dynamics, baseline impairments in PF, and depression predict longitudinal outcomes in PF impairment and depression severity. We hypothesized a central role of PF in these relationships. Within the first month of usage, on average, 69.3 [range 31-142] mood assessments and 63.3 [range 25 - 133] event assessments were available per user. Dynamic structural equation models showed substantial associations of baseline depression with weakened concurrent, lagged and cross-lagged links between events and mood, especially for positive events, including reduced likelihood of positive events. In contrast, baseline PF impairment was associated with persistence, emotional impact and volatility of interpersonal conflict events as well as with greater mood instability. Over a one-year follow-up (N=1,464; ø 1235,69 assessments per user), dynamic structural equation models identified distinct patterns: Lower concurrent mood-positive event associations, higher negative and lower positive event unpredictability, and baseline average mood/events were predictive of future PF impairment. Among affect-event dynamics, stronger cross-lagged effects of interpersonal conflicts on mood explained highest unique variance of future PF impairment (ΔR2 5.9%). Notably, baseline PF exhibited the strongest predictive utility (ΔR2 = 19.5%), followed by baseline depression symptoms (ΔR2 = 11.9%) with PF showing higher variance explanation in future depression severity than vice versa. Total average variance explanation was 71.2% for future depression severity and 75.2% for future PF impairment. Results offer specific targets for interventions: assessment of PF could therefore help to inform decisions regarding duration, goals, and intervention strategies also beyond treatment of PD. We contextualize the results within the limitations of the study, notably the use of single method self-reports.