There is a lot of interest in using passively captured data to gain insights into an individuals health. A new study by researchers at the University of Washington, Oregon Health and Science University, and the Veterans Affairs health system gives us insights into how unobtrusive sensors can be used to detect low mood and depression symptoms in older adults.

The study, published in the Journal of the American Geriatrics Society, followed 157 adults for 3.7 years on average, with an average age of 84, and tracked their mood on a weekly basis and correlated it with the amount of movement they made within their home as detected by sensors embedded in the ceiling. They also correlated each participant’s mood with the amount of time that the participants spent on the compute as well as tracked input like mouse clicks and keyboard presses.

Two of the variables they found — time out of residence and amount of time for computer use — appeared to be markers for low mood. Low mood, as they measured it, was answered by a yes/no question, “During the last week, have you felt downhearted or blue for more than three days?” rather than established depression questionnaires such as the Patient Health Questionnaire (PHQ-9) or Geriatric Depression Scale (GDS).

They found no association between low mood and movement, which is surprising given that sluggish movement or hyperactive movement — also known as psychomotor retardation and agitation — is part of the depressive episode criteria in the Diagnostic & Statistics Manual (DSM). However, the authors note that this criterion was studied in inpatient psychiatric units. They did not correct for baseline depression and noted their sample of adults lived independently and had no significant depression at baseline.

The study’s findings could bolster use cases for mood detection within smartphone apps and wearable technologies. Detecting an excessive amount of time a user spends at home or decreased engagement with desktop or mobile devices could be used for interventions in depression, which is often a co-morbidity with other medical conditions.