Is food logging a better bet for weight loss than activity tracking?

By Jonah Comstock
Calorie Counter and Food Diary by MyNetDiary MyNetDiary's popular food tracking app

When it comes to weight loss, a simple food tracking application like MyNetDiary or MyFitnessPal might be a much better investment than $100 Fitbit or Jawbone device, at least according to newly released data from the Activity Exchange, a startup that aggregates health tracking data for payers, employers, and providers.

"We had sort of a more holistic picture of a person," Jialu Chen, an associate at Asset Management Ventures who is helping the startup with marketing, told MobiHealthNews. "So a lot of our users will connect, not only with their Fitbit but also their MyFitnessPal, or they might also be tracking their weight on Withings. So suddenly we have three data points on a person as opposed to one. Or maybe they stopped using Fitbit and started using Jawbone, but they’re still running and so we have this sort of continuous data, whereas other app or device companies might only have data from a user using their app or device."

The Activity Exchange has 185,000 users on various apps or devices, according to the company's data scientist Naveed Ahmad. The startup works with health plans and self-insured employers, collecting data from their apps and devices and aggregating it, then using that data combined with claims data to help their clients communicate with their members.

To get robust data for this study, the company looked at 10,000 users who tracked their health in a number of different ways, including tracking weight either through a connected scale or manually. They found that users who logged their food three times a day lost an average of 20 pounds over the course of a year. To achieve the same results with exercise, users had to exercise an average of five days a week. They found that on average, if users logged their food for even two to four weeks, they would start to see results. 

"Something we see across a number of different analyses we’ve done is, if you look at someone’s weight gain or weight loss, it’s driven a lot more by what they’re eating rather than how many calories they’re burning," Ahmad told MobiHealthNews. "We see this with Fitbit as well, where people who are more active aren’t actually losing too much weight because they tend to kind of compensate for their activity by eating more. They don’t have a good understanding of what a scoop of ice cream is vs 1,000 steps. And the reality is what really tends to drive weight loss is reducing the amount that you eat [more so than] exercising."

Chen also pointed out that the figure of 20 pounds a year should serve to adjust people's expectations about how quickly digital health interventions can drive weight loss.

"It’s a little reassuring that the weight loss happens, but it’s slow and steady," she said. "I think a lot of people start exercising and they expect to lose 10 pounds in a month, and that’s just not happening."

Activity Exchange also looked at social media usage and sleep as possible predictors of weight loss. They found that Twitter use -- even tweeting a lot about health or about working out -- didn't make anyone any more likely to lose weight. Sleep, on the other hand, did have an effect: People with consistent, night to night sleep schedules tended to lose more weight than people with chaotic or variable sleep patterns.

The data is real-world data, so it's neither randomized and controlled not peer-reviewed like a published study would be. Activity Exchange plans to continue tracking these and other variables, and feeding the results of that data back into how they interact with their health plan customers.