The Many Faces of Stress: Preliminary Validation of a Remote Photoplethysmography-Based Tool for Psychophysiological Stress and Emotional Distress Monitoring

Background: Chronic stress contributes to mental and physical disorders, including burnout, anxiety, and depression. While self-report assessments remain valuable, they are inherently subjective and may be insensitive to short-term psychophysiological fluctuations. Remote photoplethysmography (rPPG) enables non-contact extraction of cardiovascular signals from facial videos and has increasingly been explored for stress-related monitoring through heart rate and heart rate variability features. Objective: This preliminary study aimed to assess the feasibility, usability, and preliminary construct validity of a mobile rPPG-based application for psychophysiological stress monitoring in daily life by examining usability, stress index distributions, and associations with self-reported psychological distress. Methods: A total of 252 participants from the general population and university students completed standardized facial video acquisition using a smartphone-based rPPG application and self-report questionnaires. The app extracted pulse wave signals, computed cardiovascular features related to heart rate and pulse rate variability, and integrated them into three indices: Stress Level, Stress Recovery, and Stress Response. Correlation and regression analyses examined associations with psychological distress. Results: The three indices showed substantial inter-individual variability. Stress Level was significantly associated with anxiety (r = 0.13, p = 0.036), depressive symptoms (r = 0.13, p = 0.047), and General Emotional Distress (r = 0.17, p = 0.006). In regression analysis, Stress Level emerged as the only significant independent correlate of General Emotional Distress (β = 0.21, p = 0.017). Younger participants and women showed higher Stress Level scores. Conclusions: The present findings should therefore be interpreted as preliminary and exploratory evidence of construct validity, suggesting that the app-derived indices may capture individual differences in stress-related physiological activation in everyday contexts. Currently, the observed associations were weak, the model explained limited variance, and the results do not demonstrate clinical validity, diagnostic utility, or predictive accuracy. Looking ahead, further longitudinal studies, repeated rPPG assessments, correction-aware analyses, and validation against reference physiological measures are needed before these indices can be considered suitable for clinical or preventive use.