Wearable technology continuously monitors heart-rate recovery to predict risk

CHAMPAIGN, Ill. —  The time it takes the heart to return to its baseline rhythm after exercise can predict a host of cardiovascular or metabolic disorders. In a new study, scientists at the University of Illinois Urbana-Champaign used a “smart shirt” equipped with an electrocardiogram to track participants’ heart-rate recovery after exercise and developed a tool for analyzing the data to predict those at higher or lower risk of heart-related ailments.

They report their findings in the IEEE Journal of Health Informatics.

“The heart’s response to exercise provides us with an early indicator of changes in health, in particular cardiovascular function and mortality,” said Manuel Hernandez, a professor of biomedical and translational sciences at the Carle Illinois College of Medicine who led the research with industrial and enterprise systems engineering professor Richard Sowers and graduate student Ayse Dogan.

“Extensive research has shown [an] association between abnormal heart-rate recovery and various cardiovascular diseases, including heart failure, coronary artery disease, diabetes mellitus, hypertension and sudden cardiac death,” the researchers wrote. But measuring HRR has traditionally been an involved process requiring a cardiologist, a treadmill and other costly equipment and personnel.

The team wanted to develop a more accessible approach to assessing and predicting cardiovascular risk. If a wearable device could capture relevant data as a person goes through their daily routines, shipping off that data to a laboratory or doctor’s office for analysis, it could make early diagnosis of potential problems available to many more people.

To achieve this, the researchers made use of a smart shirt developed by the Quebec-based company Carre Technologies. The shirt provides sensors to capture continuous measures of heart performance, including tracking electrical activity and heart-rate variability. In total, 38 participants ranging in age from 20 to 76 walked on a treadmill at varying speeds and inclines while wearing the device. The study was conducted in 2021 in Illinois during the COVID-19 lockdown.

The team used machine-learning and other techniques to extract the most meaningful signals of cardiac health from the data, designing a system for predicting those at highest risk of cardiovascular maladies.

“We chose the median heart rate recovery value of 28 beats per minute as the threshold to separate participants into high-risk and low-risk groups,” the researchers wrote. They used other statistical measures to cross-check the findings.

“We are finding consistent values with the different traditional classifiers and cross-validation approaches,” Dogan said. That means that the algorithm developed for the study yielded reasonably accurate results despite the small sample size, she said.

The study is a first step toward using wearables to help people more readily assess their risk of heart-related problems, perhaps catching worrisome trends before they develop into full-fledged disorders or cause sudden death, Hernandez said.

“We want to use it to provide us with some greater insight in terms of our underlying cardiovascular function,” he said. “And we want to make something that’s clinically actionable.”

“One would like to have a whole bunch of data from wearables, and then that data is transmitted to a doctor’s office, and the doctor can interpret it,” Sowers said. This would be especially useful for those living in rural communities or other areas with poor access to advanced medical facilities.

Future studies of the use of wearable technology to predict cardiovascular risk should increase the number of people studied, follow participants over time and compare their heart activity during exercise and at rest, the researchers said. Further studies should also focus on integrating the technology into standard healthcare practices.

Hernandez also is an affiliate of the Beckman Institute for Advanced Science and Technology, the departments of bioengineering and health and kinesiology, and of the Center for Social and Behavioral Science at the U. of I. Sowers also is a professor of mathematics and in the National Center for Supercomputing Applications at Illinois.

Editor’s note:  

To reach Manuel Hernandez, email mhernand@illinois.edu  

To reach Richard Sowers, email r-sowers@illinois.edu

To reach Ayse Dogan, email adogan2@illinois.edu
 

The paper “Continuous heart rate recovery monitoring with ECG signals from wearables: identifying risk groups in the general population” is available online.

DOI: 10.1109/JBHI.2025.3550092