FITNESS TRACKER INACCURACY: IMPROVEMENTS ON THE WAY

 FITNESS TRACKER INACCURACY: IMPROVEMENTS ON THE WAY

Fitness trackers have become daily companions for people trying to keep tabs on their activity and calories burned. But for those with obesity, these devices often get it wrong, underestimating energy use and offering up health data that’s more discouraging than helpful. A frustrating moment in an exercise class — where a scientist noticed his mother-in-law’s hard work barely registered on the class leaderboard — sparked a shift in approach. Now, a Northwestern University team has developed an open-source smartwatch algorithm that can accurately track energy expenditure for people with obesity, matching the accuracy of lab-grade equipment and making health tracking more inclusive.

A New Approach to Fitness Tracking

Many people rely on fitness trackers to measure calories burned, but most devices have been built and calibrated for individuals without obesity. People with higher body weight often move differently—gait, speed, and energy output vary. These differences mean that standard algorithms, especially those in hip-worn devices, often miscalculate energy burn because they don’t account for these variations. Wrist-worn trackers are a bit better, offering more comfort and potentially better accuracy, but until now, no one has rigorously adapted these devices for people with obesity.

Nabil Alshurafa, associate professor at Northwestern University Feinberg School of Medicine, led the team that developed and tested this new algorithm. Created in the university’s HABits Lab, their open-source, dominant-wrist algorithm is transparent, testable, and ready for other researchers to build on. The team plans to launch an activity-monitoring app for both iOS and Android later this year.

“People with obesity could benefit tremendously from accurate fitness tracking, but most devices just aren’t built for them,” Alshurafa said.

Why Accuracy Matters

Without an algorithm designed for wrist devices and tailored to this population, it’s difficult to know how much activity and energy people with obesity burn daily. This lack of precise data makes it harder to develop effective health interventions. In their study, the Northwestern researchers tested their algorithm against 11 existing algorithms using research-grade devices, and they used wearable cameras to verify when calorie counts were off.

The new algorithm delivered over 95% accuracy in measuring calories burned for people with obesity, a significant improvement over existing consumer trackers.

What Sparked the Research

Alshurafa’s motivation came from personal experience: after seeing his mother-in-law’s effort go unrecognized in a group fitness class leaderboard, he realized that fitness shouldn’t be a losing game for the people who need it most.

How the Study Was Done

Participants in the study wore a fitness tracker and a metabolic cart — a mask that measures oxygen intake and carbon dioxide output to calculate energy burn and resting metabolic rate. They completed various physical activities, and the researchers compared the tracker’s results to the metabolic cart’s gold-standard measurements.

A second group of participants wore a fitness tracker and a body camera as they went about their daily lives. The camera lets researchers check when the algorithm over- or underestimates calorie burn.

Alshurafa challenged participants to try as many pushups as possible in five minutes. While many couldn’t do floor pushups, they powered through wall pushups with real effort—a reminder that traditional workout standards often leave people out. He emphasized the need to rethink how gyms, trackers, and exercise programs define success so that everyone’s hard work is recognized.

The study, “Developing and comparing a new BMI inclusive energy burn algorithm on wrist-worn wearables,” will be published in Nature Scientific Reports on June 19.

The research team included lead author Boyang Wei, Christopher Romano, Bonnie Nolan, Mahdi Pedram, and Whitney A. Morelli. The work was funded by the National Institute of Diabetes and Digestive and Kidney Diseases, the National Science Foundation, the National Institute of Biomedical Imaging and Bioengineering, and the NIH’s National Center for Advancing Translational Sciences.