Smartphone-Based Markers of Social Activity in Schizophrenia and Bipolar Disorder

Social impairment and isolation are indicators of progression and predictors of relapse in psychiatric disorders. Little research has been done to measure these constructs using digital technologies in this population. We investigate the feasibility of collecting active and passive information about the social activity phenotype, in the long-term, using the Beiwe smartphone application.

Our study enrolled 47 patients with schizophrenia and bipolar disorder. Subjects completed daily surveys on digital and in-person social activity, and feeling outgoing or lonely. We described trends and variability of the mobile communication (passive) and survey (active) data streams. We conducted cluster analysis using dynamic time warping distance to explore the similarity of active and passive features of social activity.