Related Experiment Video
Updated: Jul 15, 2026
An Application for Pairing with Wearable Devices to Monitor Personal Health Status
Published on: February 3, 2022
Investigating behavioral inertia in passively sensed smartphone parameters to differentiate affective episodes in
E M Langner1, C Bittendorf2, E Mühlbauer1
1Department of Psychiatry and Psychotherapy, Carl Gustav Carus University Hospital, Faculty of Medicine, Dresden University of Technology, Dresden, Germany.
Abstract:
Bipolar disorders are severe, recurrent mental illnesses characterized by distinct mood states associated with different behavioral patterns. While (emotional) inertia, the tendency of a system to persist or change, has been linked to depressive episodes, research on its role in hypomanic or manic states remains sparse. Moreover, behavioral inertia, assessed via passive sensing using smartphone parameters is equally underexplored. This study investigated the temporal dynamics during (hypo-)manic and depressive episodes in passively sensed data in patients with bipolar disorder, complementing previous research on depression and using autocorrelation (AR) and moving averages (Mov.AVG) as dynamic measures. Data were drawn from the BipoSense study (N = 29), with 10 587 observed patient days including 20 (hypo)manic and 30 depressive episodes. Multilevel logit models were used to analyse AR and Mov.AVG across several parameters of activity, communication, sleep and phone use. During depressive episodes, AR was significantly increased in activity and communication domains, particularly in later depressive weeks, reflecting heightened behavioural inertia. In contrast, Mov.AVG showed consistent decreases across activity, communication and sleep domains. For (hypo-)manic episodes, AR patterns were more heterogenous (both, elevated and decreased values), suggesting inconsistent temporal dynamics. However, Mov.AVG revealed a coherent pattern of increased activity, communication and phone use and reduced sleep, aligning with the clinical picture of (hypo-)mania. Concomitant analyses revealed that AR and Mov.AVG seem to explain different variance in the dynamic of patients experiencing bipolar episodes. This study provides the first insight of behavioural inertia in patients with BD using passive sensing smartphone parameters. It underscores the importance of future research investigating behavioral inertia and especially the dynamics during (hypo-)manic episodes on larger cohorts and longer observation periods.
