Assessing the Scalability & Feasibility of Digitally Phenotyping Stress

Northeastern University, FY23 Tier 1 Grant

07/2022 — 06/2024

Aarti Sathyanarayana, Robert Volpe, Varun Mishra (Northeastern University)

This project aims to develop scalable tools to identify moments of heightened stress in undergraduate students by combining physiological signals, behavioral data from their phones, along with academic deadlines and performance. This effort addresses the growing prevalence of stress and anxiety, especially among incoming students navigating critical life transitions. The study employs a phased approach:

1. Calibration Phase: In-clinic assessments, including demographic data collection, psychological evaluations, and stress responses measured through physiological markers.
2. Passive Data Collection: Continuous monitoring over one month using smartphone applications and wearable devices (e.g., Empatica E4 and Garmin wearables).
3. Evaluation Phase: Focus groups assessing the scalability, comfort, and feasibility of the tools.

The outcomes include creating a pilot dataset, identifying optimal tools for passive data collection, and developing preliminary digital phenotypes of stress. These results will provide a foundation for larger-scale studies and interventions, contributing to advancements in stress detection and just-in-time intervention delivery. The project integrates expertise in psychology, digital health, and signal processing.

This study addresses a critical public health challenge by focusing on early detection and prevention of stress among college students, a population at significant risk for long-term mental health issues. By developing and evaluating digital tools to phenotype stress, the project seeks to enable scalable, passive monitoring that minimizes user burden while maximizing data quality. The findings will pave the way for innovative stress-reduction interventions using just-in-time adaptive strategies, potentially transforming mental health support for students. The methodologies and insights gained will be transferable to other high-stress populations, promoting broader societal impact.