When joy feels wrong: Identifying key dampening features predicting depressive symptomatology using machine learning and network analysis

Dampening of positivity is implicated in depression risk, yet prior research traditionally overlooks dampening’s multifaceted nature, depression’s heterogeneity, and differential relations between dampening features and depressive symptoms. This preregistered study pooled data from 13 studies, yielding four cross-sectional (N = 4015; 13-86 years) and four longitudinal (N = 1457; 14-86 years) datasets based on depressive symptom measures. Random forest (RF) and network analyses examined the predictive utility of individual dampening features for specific depressive symptoms. Across analyses, dampening features most strongly predicted core cognitive-affective symptoms such as negative self-perceptions, pessimism, pervasive negative emotions, and, to a lesser extent, fearful feelings. The features “I don’t deserve this,” “These feelings won’t last,” and “My streak of luck is going to end soon” showed consistently high concurrent predictive utility, with the latter two emerging as longitudinal predictors in RF analyses. Findings refine understanding of dampening’s relation to depressive symptoms and highlight intervention targets.