View article

Rumination, characterized by repetitive negative thoughts about one's emotions and experiences, is a transdiagnostic process involved in various emotional disorders. While statistical tools from affective dynamics have been largely applied to elucidate the temporal dynamics of affect, it has seldom been used in rumination research. In this study, we aimed to examine the predictive value of moment-to-moment variability and inertia of rumination features, comparing them against person-level means, standard deviations, and trait rumination. Forty participants answered five rumination items four times daily over two weeks, with self-report questionnaires at the end. From these intensive time-series data, we computed inertia (using a multilevel vector autoregressive model) and moment-to-moment variability per participant and then analyzed the data using regression models to predict depression and anxiety symptomatology. Rumination metrics derived from the time-series data better predicted depression and anxiety than trait rumination alone. Yet, the best-performing models included both these temporal measures and trait rumination, accounting for 67.2% of depression variance and 51.1% of anxiety. Only the inertia of one specific feature of rumination, that is, the negativity of thoughts, had added predictive value for depression, after controlling for person-level means, standard deviations, and trait rumination. Importantly, the moment-to-moment variability of rumination features did not add any predictive value, corroborating previous research on affective dynamics. In addition to inertia of negativity, the person-level means of content features of …