There’s mounting evidence that everyday fitness wearables can serve as tools to predict heart disease and mortality risk. And can do so by helping to predict the wearer’s VO2 max, the highest volume of oxygen the body can use during maximum exercise. Conventional wisdom holds that a higher level of fitness — measured through VO2 max — leads to better heart health. Research has shown that an increase in V02 max is directly linked to a decrease in the risk of heart disease and death due to cardiac-related issues. Measuring VO2 max has long been confined to clinical settings and high-tech sports labs. But findings in a study out of the University of Cambridge could change that.
Cambridge researchers turned to machine learning to help them predict VO2 max without using the expensive lab equipment that’s traditionally needed. In their study, more than 11,000 participants wore consumer-grade fitness trackers and smartwatches continuously for 6 days, engaging only in everyday activities. The sensors of these wearables recorded biometric information, including heart rate and accelerometer data, gathering 60 values per second. Researchers took this trove of raw sensor data and used an AI-based deep neural network to predict V02 max. They then compared results from two groups using wearables to a third group of study participants whose VO2 max was measured in a lab. The results were striking.
In the initial comparison, there was 82% agreement between the VO2 max scores of the wearables group and the lab-tested group. Follow-up testing showed 72% agreement. The study’s co-lead author, Dr Dimitris Spathis, says, “We had to design an algorithm pipeline and appropriate models that could compress this huge amount of data and use it to make an accurate prediction. The free-living nature of the data makes this prediction challenging because we’re trying to predict a high-level outcome with noisy low-level data [from wearable sensors].” The study’s senior author, Professor Cecilia Mascolo, adds, “Cardio-fitness is such an important health marker, but until now we did not have the means to measure it at scale.”
The ability of wearables to offer insights into cardiac-related health issues can’t be overstated; according to the Centers for Disease Control and Prevention (CDC), heart disease is the leading cause of death for people in the United States. Someone dies from cardiovascular disease about every 30 seconds in the U.S., with about 1 on 5 of all deaths attributed to heart disease. The CDC estimates that the cost of heart disease — in healthcare and productivity loss — is over $220 billion each year. And now it seems that simple, low-cost fitness wearables could be used to help lessen that impact.
