Jacobson N, Weingarden H, Wilhelm S. (2019). Digital biomarkers of mood disorders and symptom change. npj Digital Medicine. 2: 1-3. doi: 10.1038/s41746-019-0078-0
Researchers analyzed actigraphy data from a previous 55-participant study to identify digital biomarkers of mood disorders and predict symptom change over a 2-week period. Digital biomarkers are objective, quantifiable physiological and behavioral data that a digital device (i.e. actigraph) collects and measures. Researchers chose actigraphy for its ability to capture the increases and decreases in movement, energy level, sleep disruptions, and goal-directed behavior that characterize bipolar and major depressive disorder (MDD), respectively. Actigraphy data came from a study of mood disorder patients (MDD: n = 15, bipolar I and II: n = 8) and a control (n = 32). All participants wore an actiwatch (an actigraph watch) on the right wrist at all times (excluding bathing) for up to 2 weeks. Participant actiwatches collected and measured gross motor activity and rest. To measure change in symptom severity, the mood disorder group completed assessments of depressive symptoms at baseline and after the 2-week study period. Researchers extracted features from actigraphy data to form a set of 4 generalizable digital biomarkers: movement intensity distribution, movement intensity variability, autoregressive lags of movement intensity, and oscillatory pattern. In a machine-learning algorithm, actigraphy biomarkers accurately predicted participant diagnostic status (mood disorder or control) 89% of the time. Researchers also found a significant correlation between change in participant depression severity predicted by actigraphy biomarkers and actual change in symptom severity. Actigraphy may offer an affordable, low-burden supplement to conventional, resource-intensive diagnostic assessments for earlier and broader detection of bipolar disorder and MDD.