Digital biomarkers of avoidance and their relationship with depression and anxiety symptoms

Anxiety and depression are highly comorbid, warranting a need to understand transdiagnostic mechanisms contributing to the high rates of comorbidity. Prior work has implicated avoidance as a transdiagnostic mechanism; however, this has primarily been examined with cross-sectional designs and self-report measures, limiting our ability to fully understand the role of avoidance in daily life. In the current study, depressed persons (N = 268) completed EMAs 3x/day while their phone usage, texts, and GPS were passively collected for 90 days. Using multilevel vector autoregressive modeling, we investigated the temporal associations between individual symptoms and passive sensing biomarkers of avoidance, as well as how these associations differ across samples with different presentations of comorbidity. Results indicate that temporal associations among symptoms and avoidance biomarkers differ considerably across persons with different presentations, providing insight into how the comorbidity of anxiety and depression may be maintained by daily avoidance behaviors.