Nils Haller

Background: We examine the potential of behavioral activation by wearable devices in inpatients with late-life depression (LLD). Methods: In randomized order, 17 inpatients with LLD wore a GPS watch that either allowed monitoring their activity level or had a disabled activity display. The patients answered questionnaires repeatedly in addition to completing an ambulatory assessment of biopsychological stress. Results: Acceptance and satisfaction were high. We found short-term differential effects of the randomization order, with patients first in the control group being more active but also more stressed. A high number of missing values limits the informative value of ambulatory assessment. Conclusions: The intervention was feasible and well accepted. However, the effect of continuous feedback on activity levels in a crossover design needs to be critically reassessed.