The most recent issue of the Lancet Diabetes & Endocrinology published results of a randomized clinical trial evaluating the impact of fitness tracker usage on consumers’ activity levels and health benefits. The trial, taking place in Singapore and involving 800 participants over a 12-month period from 2013 to 2014, found that fitness trackers, such as Fitbit, do not raise activity levels enough to improve health. Even when employers offer financial rewards, the outcome remains disappointing.

The results could potentially put a damper on likely buyers’ interest and hurt fitness tracker manufacturers’ ambition to target the corporate wellness market. Fitbit, in response to this study’s results, gave a very standard rebuttal by stating that “Numerous published studies, along with internal Fitbit data, continue to demonstrate the health benefits of using a fitness tracker combined with a mobile app to support health and fitness goals."

To me, this trial seemed well designed and run. The design has its limitations but the authors recognize them and their scientific approach gives this study creditability. It is one of the largest randomized trials that I have seen evaluating a fitness tracking device’s impact on fitness and health and it also lasted 12 months, a decent period to evaluate trial outcomes. So overall, I believe this study is creditable, well designed, well run, and results are trustworthy.

However, the study was done in 2013-2014 and used an older model (Fitbit Zip). The study did not track whether and how participants used any Fitbit app features to increase usage. At the time of this trial, Fitbit’s hardware and app were not as sophisticated as its current technology and features are today, so the results are not conclusive.

This is important because the industry already recognizes that just counting steps as a measure of physical activities is not adequate, as the study shows and validates. The intensity of activities, measured more accurately through heart rate monitoring in the latest fitness tracking devices, is a better metric and could lead to better outcomes on usage and health anecdotally. Of course, this hypothesis needs further proof through clinical trials like this one. User engagement level can also be improved through social and gamification features now common in many fitness apps. Digital coaching could also nudges users when such needs are warranted and with user consent. These features were not part of the trial, I believe, because at that point, software was not as advanced and personalized as today’s. Future clinical trials need to factor in those software feature usage.

The use of incentive in an employee wellness program is always a controversial topic, and this study probably helps validate that cash incentive only has a limited impact and may aggravate the adverse impact on usage once incentive stops. I agree with authors’ analysis that how to implement incentive (i.e., tying it to which metrics) can have a huge impact on user behaviors. Forms of incentives are also important. The industry already concludes that cash incentive is largely ineffective for long-term behavior changes. We have seen more adoption of rewards points that need to be earned by consumers. These points are tied to a variety of healthy living decisions that people make because healthy habits are not only just about more physical activities, but also decisions on sleep, stress, diet, and managing relationships. When a program offers a more broad range of rewards/incentive categories, it would be more effective in fostering behavioral changes. That’s the prevailing hypothesis now and I’d like to see some clinical studies targeting these outcomes as end points.

The industry is getting smarter about what technology combined with what types of incentives can have the best impact on what types of users. We are unlikely to see a generic fitness device with programs targeting a general population today. Instead, depending on what attributes and health conditions that specific consumer cohorts have, employers would offer different, but targeted programs with a more appropriate tracking device. The end results may not be simply miles run or monthly calorie count, but rather whether consumers have formed better health habits and gotten rid of bad ones. For instance, for obese consumers, the goal can be a certain activity level measured by weekly exercise minutes at or above an age-appropriate heart rate threshold, accordingly an Apple Watch is offered; but for people who are relatively healthy, the goal maybe is not about steps, but improving their sleep quality and reducing stress. So a sleep tracker could be provided as part of the program. These differentiated programs will be more effective at sub population level when targeted well.

Overall, I think this study confirms three things:

-Tracking fitness through steps alone would not achieve desired outcome;

-Using a device alone without using companion solutions would also fail to achieve desired outcome

-Using incentive has no long-term impact on users’ behaviors.

Believe me, these areas are of keen interest from the fitness technology and healthcare industries, which are looking for more evidence and better approaches. They have become much smarter since the time this study was done. So let’s hope that with better understanding of human behavior changes and better technology designs, we can find a winning formula that can be validated by future large-scale scientific studies.

Fitbit may already be conducting one right now, I hope.