Affect Dynamics as Predictors of Symptom Severity and Treatment Response in Mood and Anxiety Disorders: Evidence for Specificity

Studies of affect dynamics in psychopathology often focus on the prediction of broad constructs like subjective well-being and psychological health. Less is known about how fluctuation in affect over time relates to specific symptom measures (e.g., anxiety or depression), or how these domains change in treatment. A clinical sample of 32 adults with mood and anxiety disorders (13 generalized anxiety, 5 major depression, 14 comorbid) completed four daily assessments of positive (PA) and negative affect (NA) for 30 days prior to receiving cognitive behavioral treatment. Anxiety and depression symptom severity were assessed pre- and post-treatment. We calculated three metrics of affect dynamics for each person’s PA and NA time series: (1) variability (experiencing emotional extremes, the standard deviation of a person’s PA or NA vector); (2) instability (magnitude of point-to-point change in emotion, the vector’s mean squared successive difference); and (3) inertia (the extent to which emotions self-perpetuate over time, the lag-1 autocorrelation of the vector). Multiple regression models were run to test dynamics of positive and negative affect as between-subjects predictors of symptom severity and pre-to-posttreatment change in symptoms. Findings suggest NA dynamics are unrelated to depression symptom severity or treatment response, but we observed a specific effect of NA instability (MSSD) on both severity and response of anxiety symptoms. All PA dynamics were unrelated to anxiety or depression symptom severity. However, variability, instability, and inertia of PA were all found to relate to treatment response for both anxiety and depression symptoms. Taken together, our results suggest that affect dynamics have some specificity in their relationship to clinically relevant phenomena such as symptom severity and treatment outcomes at the between-subjects level of analysis.