Examining the clinical utility of using affective dynamics to characterize and predict depression symptoms within a comorbid sample

Background: Dynamic indices calculated from ecological momentary assessment (EMA) data can model individual-level symptom fluctuations over time. However, the clinical utility for using these complex metrics to characterize comorbid symptomatology and to predict future symptom changes has not been adequately examined. Methods: Women (N = 35) with social anxiety disorder and a history of major depression completed a clinical diagnostic interview and self-report measures at baseline and approximately two months later. In between these assessments, participants completed EMA surveys on mood and anxiety symptoms five times a day for approximately 30 days (T = 5,250). Symptom severity and dynamic indices (i.e., variance, inertia) during EMA were calculated. The relative predictive value of these indices for characterizing baseline symptomatology, as well as the relative prospective predictive value of dynamic indices predicting future depressive symptomatology was assessed. Results: Baseline depressive symptoms were associated with mean levels of anxiety (b = .38, p = .001); whereas, baseline social anxiety symptoms were associated with mean levels of depression during EMA (b = -.38, p = .013). Moment-to-moment variance in anxiety during EMA (b = .22, p = .020) and baseline self-report measures of depression (b = .50, p < .001) predicted future depressive symptoms. Conclusions: Moment-to-moment fluctuations in anxiety may constitute a unique early warning sign of future increases in depression for individuals with mood and anxiety comorbidity.