New algorithm makes fitness trackers more accurate for people with obesity

For many, fitness trackers have become indispensable tools for monitoring how many calories they've burned in a day. But for those living with obesity, who are known to exhibit differences in walking gait, speed, energy burned and more, these devices often inaccurately measure activity - until now.

Scientists at Northwestern University have developed a new algorithm that enables smartwatches to more accurately monitor the calories burned by people with obesity during various physical activities.

The technology bridges a critical gap in fitness technology, said Nabil Alshurafa, whose Northwestern lab, HABits Lab, created and tested the open-source, dominant-wrist algorithm specifically tuned for people with obesity. It is transparent, rigorously testable and ready for other researchers to build upon. Their next step is to deploy an activity-monitoring app later this year that will be available for both iOS and Android use.

People with obesity could gain major health insights from activity trackers, but most current devices miss the mark."

Nabil Alshurafa, associate professor of behavioral medicine at Northwestern University Feinberg School of Medicine

Current activity-monitoring algorithms that fitness trackers use were built for people without obesity. Hip-worn trackers often misread energy burn because of gait changes and device tilt in people with higher body weight, Alshurafa said. And lastly, wrist-worn models promise better comfort, adherence and accuracy across body types, but no one has rigorously tested or calibrated them for this group, he said. 

"Without a validated algorithm for wrist devices, we're still in the dark about exactly how much activity and energy people with obesity really get each day - slowing our ability to tailor interventions and improve health outcomes," said Alshurafa, whose team tested his lab's algorithm against 11 state-of-the-art algorithms designed by researchers using research-grade devices and used wearable cameras to catch every moment when wrist sensors missed the mark on calorie burn.

The findings will be published June 19 in Nature Scientific Reports. 

The exercise class that motivated the research

Alshurafa was motivated to create the algorithm after attending an exercise class with his mother-in-law who has obesity. 

"She worked harder than anyone else, yet when we glanced at the leaderboard, her numbers barely registered," Alshurafa said. "That moment hit me: fitness shouldn't feel like a trap for the people who need it most."

Algorithm rivals gold-standard methods

By using data from commercial fitness trackers, the new model rivals gold-standard methods of measuring energy burn and can estimate how much energy someone with obesity is using every minute, achieving over 95% accuracy in real-world situations. This advancement makes it easier for more people with obesity to track their daily activities and energy use, Alshurafa said.

How the study measured energy burn

In one group, 27 study participants wore a fitness tracker and metabolic cart - a mask that measures the volume of oxygen the wearer inhales and the volume of carbon dioxide the wearer exhales to calculate their energy burn (in kilocalories/kCals) and resting metabolic rate. The study participants went through a set of physical activities to measure their energy burn during each task. The scientists then looked at the fitness tracker results to see how they compared to the metabolic cart results.

In another group, 25 study participants wore a fitness tracker and body camera while just living their lives. The body camera allowed the scientists to visually confirm when the algorithm over- or under-estimated kCals.

At times, Alshurafa said he would challenge study participants to do as many pushups as they could in five minutes. 

"Many couldn't drop to the floor, but each one crushed wall-pushups, their arms shaking with effort," he said, "We celebrate 'standard' workouts as the ultimate test, but those standards leave out so many people. These experiences showed me we must rethink how gyms, trackers and exercise programs measure success - so no one's hard work goes unseen."

The study is titled, "Developing and comparing a new BMI inclusive energy burn algorithm on wrist-worn wearables."

Other Northwestern authors include lead author Boyang Wei, and Christopher Romano and Bonnie Nolan. This work also was done in collaboration with Mahdi Pedram and Whitney A. Morelli, formerly of Northwestern.