Automatically generated smartphone data and subjective stress in healthy individuals - a pilot study.

Research paper by Maria M Faurholt-Jepsen, Helga H Þórarinsdóttir, Maj M Vinberg, Henrik H Ullum, Mads M Frost, Jakob J Bardram, Lars Vedel LV Kessing

Indexed on: 29 Dec '19Published on: 28 Dec '19Published in: Nordic journal of psychiatry


Most people will also experience symptoms of stress at some point. Smartphone use has increased during the last decade and may be a new way of monitoring stress. Thus, it is of interest to investigate whether automatically generated smartphone data reflecting smartphone use is associated with subjective stress in healthy individuals. to investigate whether automatically generated smartphone data (e.g. the number of outgoing sms/day) was associated with (1) smartphone-based subjectively reported perceived stress, (2) perceived stress (Cohen's Perceived Stress Scale (PSS)) (3) functioning (Functioning Assessment Short Test (FAST)) and (4) non-clinical depressive symptoms (Hamilton Depression Rating Scale 17-items (HDRS)). A cohort of 40 healthy blood donors used an app for daily self-assessment of stress for 16 weeks. At baseline participants filled out the PSS and were clinically evaluated using the FAST and the HDRS. The PSS assessment was repeated at the end of the study. Associations were estimated with linear mixed effect regression and linear regression models. There were no statistically significant associations between automatically generated smartphone data and perceived stress, functioning or severity of depressive symptoms, respectively (e.g. the number of outgoing text messages/day and self-assessed stress ( 0.30, : -0.40; 0.99,  .40). Participants presented with low levels of stress during the study. Automatically generated smartphone data was not able to catch potential subjective stress among healthy individuals in the present study. Due to the small sample and low levels of stress the results should be interpreted with caution.