Are Fitness Trackers Fit to Track Your Fitness? – Part I: Activity/Step Counts

Fitness trackers have taken the world by storm, led by the popularity of FitBit and the Apple Watch. There are, literally, hundreds of fitness trackers available worldwide from all sorts of companies, from hard-core fitness/sport companies to digital technology companies to companies focused on narrow health areas, such as pediatric obesity.

FitBit was one of the first to market the new generation of “wearable” fitness trackers, but it was a surprise hit! At a TechCrunch 50 conference in September 2008, the founders of FitBit hoped that they could get 50 pre-orders of their new activity tracker. (In reality, they thought they would only get 5 orders.) Instead, in just one day, they got 2000 pre-orders. Now, just 9 years later, FitBit had over $2 Billion of revenue in 2016.

FitBit’s success and that of subsequent devices like the Apple Watch begs the obvious question: “Are these fitness trackers accurate?” or as the title of my blog asks, “Are Fitness Trackers Fit to Track Your Fitness?”

Over this and the next two blogs, we will examine the scientific data on the accuracy of such trackers for the three most popular uses: activity/step count tracking, heart rate tracking, and sleep tracking.

Today we examine step count tracking. Unlike typical medical devices, fitness trackers are usually not regulated, nor are they required to provide any proof or validation beyond what the market and competitors demand. Nonetheless, the growth in fitness trackers in general, and FitBit in particular, has been driven first and foremost by the hope/belief/thought that if one monitors one’s steps and activity, one will take action to be more active and, therefore, healthier.

Although we will not examine all the assumptions in that last statement today, the concept that steps lead to more activity and healthier lifestyle has been critical in the adoption of FitBit by large corporate wellness departments for their employees. Many of these large corporations are “self-insured,” meaning that they pay for all the cost and risk associated with the health insurance of their employees. Although employees may think of their employer’s designated insurance company (e.g. Aetna, Cigna, Blue Cross, etc) as the one that pays for their health care, these costs are all passed back to these self-funded employers via insurance premiums. The employer takes the risk of employee health care costs, not the insurance company. Therefore, these employers, through their corporate wellness programs, want employees to be healthier so that health care costs will be lower. Enter the FitBit revolution!

So how accurate are fitness trackers at counting steps?

Many quality medical studies have looked at this question. One of the best was published by Case et al. in 2015 in the Journal of the American Medical Association (JAMA). These authors recruited 14 healthy adults and had them undergo a very specific treadmill walking protocol of exactly 500 or 1500 steps using both smartphones in pants pockets and three different wearables (including a couple of traditional pedometers), simultaneously. A total of four smartphone-based tracking apps and 6 wearables were tested at both the step protocols. Overall, the researchers found that all devices performed very well in accurately measuring the total number of steps. There was more deviation in step counts among wearables than among the smartphone-based apps. One device that had a particularly mixed result was the Nike Fuelband (step count on average 20% lower than actual). Since 2014, the Fuelband is no longer on the market. Excluding this device, both the wearables and smartphone apps performed similarly, with less than 10% average difference from the actual step count in most cases. This study demonstrated that, with one exception, popular activity trackers seem to accurately track the number of steps.

The next relevant question is whether such tracking results in greater activity or a healthier lifestyle?

An interesting analysis from a group of researchers in Spain (Bort-Roig, et al.) in 2014, published in the journal Sports Medicine, reviewed 26 of the best research articles (culled from a list of over 1,500 publications) that studied the role of activity tracking in health and disease. Most studies reported average-to-excellent measurement accuracy.

Regarding the effect of activity tracking on behavior and other health outcomes, very few studies and very small numbers of subjects were scientifically assessed for outcomes. With regards to step counts, three studies specifically looking for changes in step counts since the start of tracking did find statistically significant changes, but these changes were reported as “small” and “observed over short intervention and measurement periods.” The authors concluded that more research and better designed studies are needed to understand the full intervention effect of activity tracking.

In other words, the jury is still out on whether activity tracking leads to more activity…

In our next blog, we will examine the accuracy of fitness trackers on heart rate measurement.