Researchers from Canada’s University of Waterloo have discovered that the Apple Watch’s ECG sensor data could be used to develop a stress prediction tool. The ECG app and electrical heart rate sensors were introduced with the Apple Watch Series 4 and have subsequently been part of Series 5, Series 6, Series 7, Series 8, or Ultra. The tool can be used to track and record the wearer’s heartbeat and rhythm, and then check the recording for atrial fibrillation (AFib), a form of irregular rhythm.
The study has been published in Frontiers Digital Health and explores the possibility of using ECG sensor data to deduce a stress score. Previous studies cited by Apple have also shown similar results in classifying the rhythm of the Apple Watch ECG compared to standard 12-lead ECGs, a clinical trial of 600 participants revealed that the ECG sensor had over 99% specificity when identifying sinus rhythm. For sensitivity in atrial fibrillation, there is an over 98% specificity.
Participants in the study were instructed to collect data 6 times a day within a 3-hour interval with the help of an iPhone 7 with iOS 15.0 and an Apple Watch Series 6 containing an installed Apple Watch ECG app for two weeks. Before recording data on the app, participants had to complete a stress questionnaire on the iPhone and the relevant data was then collected using HealthKit and was eventually loaded into Kubios to determine heart rate variability (HRV).
In the study researchers also considered the impacts of other variables on stress prediction models. These included age, gender, profession, and socioeconomic status.
The researchers found that “In general, the “stress” models had a high level of precision but lower recall. The “no stress” models performed generally well with a recall typically above 60%. Considering the ultra-short duration of the ECG measurements performed here compared to the standard and the nature of real-life measurements, the results presented were quite promising.”
Apart from usual frequency domain features such as SDNN, the model included valuable HRV features like heart acceleration (AC) and deceleration capacity (DC). Researchers also shared that the predictive power of the current model can be improved if ECG data is used with other stress-related variables such as sleep and physical activity, both of which are already recorded by the Apple Watch.
The benefit of having a wearable device that can monitor stress in real-time is that it would allow people to respond early to changes in their mental health and at a large scale data collected through this method can be used to inform public health initiatives and policies.
While Apple currently does not have a stress score feature, other wearable devices such as Fitbit and Garmin do offer such a feature. In the case of Fitbit the HRV is used to determine the stress score with the assistance of EDA sensors and skin temperature tracking that look out for physical indications of stress. In comparison, Garmin employs a simplistic HRV derivation model to decipher users’ stress levels, the watch determines the interval between each heartbeat, and less variability between beats equates to higher stress levels, whereas an increase in variability represents less stress.