Geolocation features differentiate healthy from remitted depressed adults.

Area of Science:

• Digital psychiatry
• Computational psychiatry
• Behavioral data science

Background:

• Depression recurrence poses a significant clinical challenge, necessitating tools for early symptom detection.
• Bridging the gap between symptom reemergence and treatment initiation is crucial for managing depressive disorders.

Purpose of the Study:

• To investigate the utility of mobile sensing data, including geolocation and ecological momentary assessment, for detecting early signs of depression recurrence.
• To develop a predictive model for identifying individuals at risk of depression relapse using objective behavioral data.

Main Methods:

• Remitted depressed adults and healthy controls were monitored using smartphone apps for 21 days, collecting geolocation and ecological momentary assessment data.
• Clinical interviews and self-report measures were administered at baseline.
• A least absolute shrinkage and selection operator (LASSO) regression model was employed to identify key predictors of recurrence.

Main Results:

• Remitted depressed adults showed reduced circadian routine regularity and lower average daily distance traveled compared to controls.
• Reduced travel distance was associated with increased negative affect, irrespective of depression severity.
• A model combining circadian routine, travel distance, and baseline depression severity achieved 72% accuracy in classifying remitted depressed individuals.

Conclusions:

• Mobile sensing offers a promising avenue for improving clinical care in depressive disorders through real-time monitoring.
• Objective behavioral data, when integrated with clinical assessments, can enhance the prediction of depression recurrence.
• Technological advancements in mobile sensing can potentially reduce high recurrence rates by facilitating timely treatment.