Welcome to Dear Julia, a weekly column where readers can submit everyday health questions. Which over-the-counter painkillers work best? Is it better to run or walk for exercise? How much harm does frequent flying do to your body? Julia Belluz will sift through the research and consult with experts in the field to figure out how science can help us live happier and healthier lives.
Dear Julia: Is there any way to reliably know how many calories you burn during exercise? I'm skeptical of what my fitness tracker says.
Dear reader: You should be skeptical. Estimates of how many calories you burn for a given activity can vary widely. Two different fitness trackers can show very different results for the exact same exercise. These numbers are best viewed as rough guesses, not precise scientific calculations.
I reached out to a few experts; they all said measuring energy expenditure during exercise with precision is extremely difficult, and requires the tools of a physiology lab. One scientist even laughed at the notion that people would take their calorie burn estimates seriously. So — assuming you don't want to spend $30,000 on a wearable calorimeter, or millions on a metabolic chamber — here are some ideas about how to navigate the more accessible tools out there.
Free online trackers are the least trustworthy
The free online trackers — the ones that let you put in a few details like age or weight, then estimate that you'll sweat off 700 calories in a hot yoga class — are probably the least trustworthy. That's because these aren't very individualized. They tend to make all kinds of assumptions about things like your resting metabolic rate and muscle mass based on broader averages. And because online trackers don't measure movement, they can't make subtle distinctions between, say, a 30-minute walk on a flat street and a 30-minute walk up a hill.
More fundamentally, online trackers can't measure what's actually going on inside your body during exercise, like your heart rate or breathing. That's a major limitation, since two people will expend energy differently doing the exact same exercise based on their existing levels of fitness. Usually, picking up on these variations requires more detailed heart-rate data, for starters.
"A highly trained runner who is 120 pounds likely burns less calories than a newbie who is 120 pounds and running the same distance and pace," explains Sherry Pagoto, co-founder of the UMass Center for mHealth and Social Media. "This will bear out in their heart rate data — the more conditioned one gets, the lower the heart rate during the same activity, and the fewer calories burned."
By the way, the same limitations apply to machines at the gym. Even when you’ve entered a few variables — sex, weight, age — they're not always calibrated correctly and they often lack accurate heart-rate monitors. So they'll end up spitting out an estimate for the "average person" — which is not likely to be you. At best, you’ll get a broad guess from these machines.
Wearables are a little more accurate — but still far from perfect
The more sophisticated wearable devices — JawBones, Fitbits — fare a bit better than online trackers, since they can measure things like movement and heart rate.
Even so, the calorie estimates on these trackers are far from perfect, and can vary pretty significantly from device to device. It's telling that if you strap on two kinds of fitness trackers and do an exercise, you're likely to get two different calorie estimates. (See this excellent Wired article for a demonstration of that.)
Each of these devices take some data from your body and feeds it through a proprietary algorithm particular to that device. "Step counters multiply distance covered by body weight to get work performed," explained sports nutritionist Matt Fitzgerald. "Bicycle power meters use a similar method, substituting power for distance covered. Heart rate monitors use heart rate to estimate the amount of oxygen consumed during exercise."
But these algorithms necessarily have to make assumptions and over-simplifications. And there's no way to tell how well they apply to your particular situation. "All of these devices have formulas that work better for some people and not so well for others," said Dan Heil, an exercise physiology professor at Montana State University. "You simply never know which group you fall into."
With more than 165,000 mobile health apps on the market, only a tiny fraction have been studied scientifically — that is, measured against more sophisticated calorie counters in laboratories. Worse still, most of that research is a few years out of date because of lag time in publishing. But researchers have found that tracking devices can have a pretty big margin of error, underestimating or overestimating calories used by as much as 20 percent. A related study, led by researchers at Stanford, compared seven popular wrist-worn fitness trackers and found no device had an error rate that was less than 20 percent (and the worst performing device was wrong 93 percent of the time).
Fitness trackers may not be accurate — but they're not totally useless
Operating in this imperfect world, the researchers I spoke with suggested using a few different tracking tools and averaging out their counts to see where you stand. This sounds super-tedious, but if you do the same exercises every week — and you’re a stickler for accuracy — it may be worth your time.
"Its almost impossible to tell which [tracking tools] are better than others," Heil advised. "But if there are two to three different devices that are giving similar predictions, then there is a good chance that these devices are giving reasonable calorie calculations for you...but not necessarily for everyone else."
For people who are trying to lose weight, the investment in cross-checking is even more worthwhile. Pagoto suggested starting by assuming a device is overestimating calorie burn — or, if you’ve cross-referenced, going with the most conservative number. "Even though the lowest estimate may not necessarily be the most accurate, the direction of the error will be such that you will lose weight more quickly [than you expect to] rather than more slowly," she said.
If all else fails, the most out-of-date tracking device you're likely to own — your scale — won’t lie. "If your device is telling you that you are burning thousands of calories and the food you enter is under your calorie goal," Pagoto said, "yet you are still not losing weight, the information is not likely accurate."
Whatever your goal, managing expectations is probably important, says Kevin Hall, the National Institute of Health researcher who devised this science-based body-weight planner. Since there's no way to ensure accuracy, Hall suggested using the devices as a way to measure your levels of physical activity over a long period: whether you're doing more or less exercise from one month to the next.
"I laugh when I see people taking a careful look at the calorie counts from those devices because they’re really not intended for that," he said. He puts his study participants in a multi-million-dollar metabolic chamber to see how much energy they burn while exercising. Anything less, he suggested, might give you an accurate count — if, that is, you add or subtract 20 percent for a "conservative" margin of error.