Deriving a Cortisol-Related Stress Indicator From Wearable Skin Conductance Measurements: Quantitative Model & Experimental Validation

Abstract

In order to find a stress indicator that can be used to monitor stress with wearables, we compare the almost instantaneous effects of psychological stress on skin conductance, with the effects on the stress hormone cortisol, peaking about 20-30 min later. We modeled this relation deploying a convolution of the height of the skin conductance peaks with the cortisol stress response curve, and used it to determine a skin conductance-derived estimate of stress-induced cortisol. We then conducted a first experiment to validate this model, comparing the stress-induced cortisol estimates with cortisol as measured in saliva samples. Participants (N = 46) completed stressful, boring, and performance tasks in a controlled laboratory setting. Salivary cortisol samples were taken at regular moments. Based upon the pattern of measured salivary cortisol before and after the performed stressful task we divided subjects in high-cortisol responders and low-cortisol responders. For both groups, we found substantial correlations between the skin conductance-based stress-induced cortisol estimates and the measured salivary cortisol. In addition, the (Fisher-corrected) mean within-participant correlation between these variables was found to be 0.48, which proved to be significantly different from zero. These findings support the use of the skin conductance-based stress-induced cortisol estimates as a stress indicator reflecting in-body cortisol changes.

A Computational Approach to Examining Team Coordination Breakdowns During Crisis Situations

“…Pre-Processing. EDA data were preprocessed similarly to an approach by Westerink et al (2020), using the PyPhysio package (Bizzego et al, 2019). The EDA signals were resampled at 10 Hz, and a 10 Hz Butterworth filter was applied to remove wideband noise.…”

Machine Learning-Based Routine Laboratory Tests Predict One-Year Cognitive and Functional Decline in a Population Aged 75+ Years

“…Therefore, the results suggest that at the beginning of a cognitive and functional decline, cortisol levels are not elevated; however, they may increase as the disease progresses. The majority of studies, including ML, evaluated cortisol levels as an outcome, not a feature [ 54 ]. However, few studies included cortisol levels in ML algorithms able to predict different diseases, such as mental disorders [ 55 ], stress and dietary intake [ 56 ], and quality of life [ 57 ].…”

Towards Supporting Emotion Awareness in Retrospective Meetings

“…Among other affective states, stress was recently investigated by Westerink et al [17]. In their study, they demonstrated a correlation between peaks of cortisol, which is the hormone associated with the experience of psychological stress, and peaks of skin conductance, thus suggesting that peaks in the EDA signal can be successfully used to identify stress episodes.…”

“…We compute EDA peaks using the same tool, which implements a method capitalizing on the 1st derivative of the signal curve. The choice to include identification of signal peaks is informed by findings by Westerink et al [17], demonstrating that stressful events are associated to EDA peaks.…”