Elliptical Machine is Lying to You

The Elliptical Machine Is Lying to You

Should you trust the stats from the treadmill? What about your Fitbit? Here's what actually works when it comes to tracking calories.

Getty
 
By Tim Newcomb

Your feet pound the treadmill. Your shirt soaks up the sweat. You can feel the heat—you're working hard, burning those calories and you rack up the steps and the miles. But do you really know how many calories you're burning?

Gym cardio machines such as ellipticals, stair steppers, and stationary bikes offer a reassuring calorie count on their LED screens—a little numerical reminder that you're really doing something. If you're wearing a smartwatch or a Fitbit, though, it might tally a totally different calorie number for the exact same workout. Same for a fitness-tracking app that monitors your steps or tracks your movement by GPS if you're biking or running around the neighborhood.

So what gives? Here's why they're different, and how to know which numbers you can trust.

WHY COUNTING CALORIES IS SO HARD

Daryl Waggott is a senior research scientist at Stanford University, and the co-author of a recent Stanford School of Medicine study that tried to answer this calorie-counting question. Waggott and colleagues tried to find out how accurately an array of devices could measure heart rate and energy expenditure (calories).

"The take-home headline for us was the heart rate was surprisingly good," Waggott says about the study. "Energy expenditure, on the other hand, was pretty shitty—it was really bad."

The Stanford team studied seven devices, all of which combined wrist-detected heart rate with energy expenditure. They all found heart rate data within 5 percent of the true range. But the energy expenditure data was grossly miscalculated, with the most accurate off by as much as 27 percent and the worst machine missing the mark by 93 percent. In the case of fitness trackers, the Stanford study found they tended to underestimate the effects of low-impact movements such as movement while sitting, but tended to overestimate the effects of intense exercises like running.

"ENERGY EXPENDITURE, ON THE OTHER HAND, WAS PRETTY SHITTY—IT WAS REALLY BAD."

It makes sense, really. Using LED lights to monitor blood movement, trackers can gauge heart rate, a relatively simple measurement. But caloric burn offers a more complex metric, which takes into account everything from body composition to age and fitness level to exertion. Finding an all-encompassing algorithm for such a complex metric doesn't come easy.

That's why you shouldn't trust your elliptical machine. While Waggott and his colleagues didn't study exercise equipment in gyms, he says he expects them to be even worse than the wrist-based equipment, simply because the best way to get the best stats is to have as much data about a person and an activity as possible—and a shared treadmill or other piece of gym equipment only knows what you tell it. If you don't tell it anything, then that calorie estimate will be for a fictional default person, and may be way off from what your body has actually burned.

These machines can get more accurate if they use sensors to track stride distance, for example, and pair that with heart-rate data and body calculations. But rarely do people input all that information.

EXERCISE MECHANICS

Even if you feed your weight and age into the machine and hold the heart rate sensors the entire time, that's no guarantee of an accurate calorie count. Consider heart rate data, which, as noted, is (usually) a relatively simple thing to measure. Waggott says this that can differ based on the device. Wrist devices read heart rate by shining LED lights against the skin, but because of battery life concerns, the machines read that data sporadically, which may not provide real-time information on energy expenditure. And with blood flow a key metric in capturing heart rate via the wrist, unnatural wrist positions can change the flow of the blood and, therefore, the data.

Each fitness company, whether offering a wrist-based device or a piece of exercise equipment, has its own proprietary algorithm for figuring out how to give out energy expenditure figures, all based on scientific equations. Another problem, Waggott says, is that those equations don't take into account the variability of people. Skin tones can change a LED light's ability to accurately read heart rates. Body hair, tattoos, and body types will all completely alter a device's ability to read information or best calculate it. The simple many shapes and sizes of people—not to mention the different ways they walk and run and burn energy—makes the device's job all that more difficult.

"If you are walking or running, you need an accurate stride, but everybody has different (body) economics," Waggott says. "Ten thousand steps could be 400 or 800 calories, based purely on mechanics."

Garmin, one of the leaders in heart-rate-based activity tracking, tells PM in a statement that heart rate is a good source of information about the amount of energy the body spends performing any activity. "However, it is very difficult to correctly predict energy expenditure based on heart rate alone, because there are huge difference in heart rate behavior between individuals, such as different maximum heart rates, resting heart rates and fitness levels." Garmin fitness products use a proprietary algorithm from Firstbeat, a provider of physiological analytics, taking into account those differences, as well as other factors, to estimate energy expenditure from heart rate measurements.

"TEN THOUSAND STEPS COULD BE 400 OR 800 CALORIES, BASED PURELY ON MECHANICS."

Fitbit tells PopMech.com in a statement that it is confident in the performance of its trackers based on its own research and development, as well as outside reports. Its trackers show an estimated total number of calories burned based on a user's basal metabolic rate and activity energy expenditure. Using height, weight, age and gender information, plus heart-rate tracking when available, Fitbit says it can provide a more accurate measure than a step-only estimate, especially helpful for non-step workouts like cycling, elliptical or exercise classes.

JUST KEEP GOING

Waggott is a cardiologist by trade, and he says he worries that too many people rely on data-tracking devices to tell them how many calories they burned and how much work they've done—and that many people stop working out too soon because a device overestimates the workload.

When it comes time to choose the most accurate device, start with a few basics. Pick equipment that tracks heart rate. You'll need that. You also want devices that at least attempt to measure you, as simple data from height, weight and age can help streamline the algorithm. For hard-core runners, devices that measure stride rates and cadence will give a fuller account of your workload. Cyclists can find power meters to help define their physical output.

Everyone can agree that activity proves better than no activity, but exactly how to measure that activity remains a work in progress. Just know the more data you put into your device, the more accurate the data you get out.