Non-invasive recording of physiological variables under stress conditions and aerobic and anaerobic physical activity

Abstract

Understanding how acute stress and exercise affect glycemic control in type 1 diabetes patients is of great interest for inclusion in an Artificial Pancreas (AP) control algorithm. As a first step towards integrating stress and exercise in an AP, this study presents a database with recordings from 18 healthy volunteers during stress and exercise conditions. Blood volume pulse, temperature, electrodermal activity and heart rate were non-invasively acquired with the Empatica E4 device. Participants accomplished well defined stress, aerobic and anaerobic exercise protocols. The procedure designed to induce stress showed efficiency generating significant changes (p < 0.05) for each test with respect its previous rest period in the recorded signals, primarily for the opinion and arithmetical tasks. Heart Rate (HR) and Electrodermal Activity (EDA) changes followed the same performance that the self-perception stress level of the volunteers through stress protocol. HR variation registered along physical activity revealed the potential to discriminate between aerobic and anaerobic exercise, showing a different behaviour between both conditions. The recorded physiological changes presented in this work provide a valuable data source for generating models for stress and exercise for future research.