Fitbit Activity Trackers Enhance The Effectiveness Of Physical Activity Interventions Offering Personalized Advice

Being active is recommended to prevent developing diabetes, cardiovascular disease, cancer, mental health problems, and much more. Unfortunately, most people are not meeting physical activity guidelines. This is costly to society, but more importantly, it reduces our quality of life and increases the disease burden. Because so many people are inactive, we need interventions that can reach large numbers of people at an affordable cost.

In this regard, web-based, personally tailored interventions have demonstrated promising outcomes. Personally tailored interventions mimic consultations with health professionals and provide personally relevant information to support healthy lifestyles.

The personalized physical activity advice is provided after participants complete one or more web-based surveys. Based on participant responses and using IF-THEN algorithms (e.g., IF not meeting activity guideline, THEN provide advice to increase activity levels), relevant feedback is selected from a large database with all possible response options. Although these interventions have shown outcomes that are much better compared to providing generic (one-size-fits-all) and targeted (this-size-fits-specific-group) interventions, they can still be improved and be more effective for more people.

One way to improve personally tailored physical activity interventions is to stop using questionnaires to estimate how active people are. It is well known that many people overestimate how active they are by a large margin. Inaccurate activity estimations will lead to participants being provided with incorrect advice. For example, someone might receive the message that they are meeting the activity guidelines and do not need to become more active when this is actually not the case. When this happens, the intervention is not providing accurate and credible advice to participants and will, therefore, not be as effective as it could be.

The popularity of new activity trackers (e.g., Fitbit, Garmin) offer opportunities to improve personally tailored interventions. These activity trackers provide better estimates of physical activity (compared to survey data), they can easily be worn all day and the data they collect is automatically uploaded to the Internet, and can thus be used in other applications.

Therefore, our study compared the effectiveness of two personally tailored physical activity interventions: one using a traditional questionnaire to measure physical activity, and the other using Fitbit activity trackers instead.

We recruited 243 people from all over Australia to be in the study; all contact with participants was via e-mail or phone. We collected data on their physical activity levels immediately before they received the intervention and again 1 and 3 months later. After collecting baseline data, participants were randomly assigned to one of two groups. Both groups received a comprehensive personally tailored physical activity intervention, that provided them with eight modules of personalized advice over a three-month period. However, one group had to complete a physical activity survey every time before receiving a new module of personalized advice, whereas the other group was provided with a Fitbit Flex at the start of the study and asked to wear it as much as possible. All they had to do was making sure that their Fitbit data was synced with our website before accessing a new module.

Image republished with permission from Elsevier from https://doi.org/10.1016/j.ypmed.2018.09.003 under a CC-BY 4.0 license

The first remarkable result of this study was that many people stopped participating in the study before it was finished. This is probably because we didn’t have any contact with the participants other than by phone or e-mail, so accountability was low. However, more importantly, nearly two times as many of the drop-outs were in the group that didn’t receive the Fitbit. People that received a Fitbit were much more motivated to keep on going, and this is also visible in the number of modules that each group completed, which was much higher in the Fitbit group.

Image republished with permission from Elsevier from https://doi.org/10.1016/j.ypmed.2018.09.003 under a CC-BY 4.0 license

Secondly, those in the Fitbit group really liked using their Fitbit. They reported that they wanted to keep using it, they found it easy use and comfortable to wear. They also thought it improved the value, credibility, user-friendliness and personal relevance of the personalized advice. Most of them wore their Fitbit most of the time over the 3-month period. Consequently, it was no surprise that on average the Fitbit group rated the acceptability of the intervention higher than those who did not receive a Fitbit. For example, the Fitbit group more often thought the personalized physical activity advice was credible, that it helped them to achieve goals, and that they learned something new compared to the group who did not receive the Fitbit. Interestingly, Fitbit group participants even though the website looked better and was easier to use compared to the other group, despite both groups receiving the exact same website.

Finally, physical activity levels both groups went up a lot (this was expected as both groups received a comprehensive intervention), but they went up a lot more in the Fitbit group. In the Fitbit group activity levels increased by 285 minutes per week, whereas they only increased 120 minutes in the other group. More importantly, there were also strong increases in moderate and vigorous intensity physical activity. Physical activity at higher intensities contributes most to health improvements.

Image republished with permission from Elsevier from https://doi.org/10.1016/j.ypmed.2018.09.003 under a CC-BY 4.0 license

Thus, in conclusion, activity trackers really help to improve the effectiveness of personally tailored physical activity interventions. However, some caution is needed when considering these outcomes. While this is an innovative study with strong outcomes, some limitations to the study design need to be mentioned. Firstly, there was no control group, so it’s possible that something other than the intervention received in both groups caused the increases in activity (e.g., a prolonged bout of good weather). Secondly, there was no “tracker-only” group, so it is impossible to tell to what extent the tracker itself increased activity and not so much the personally tailored intervention. Though the high module completion in the Fitbit groups does suggest the tailored intervention did also contribute to increased activity levels. And thirdly, preventing chronic disease by engaging in regular physical activity only makes sense in the long run, thus we must also examine whether the intervention outcomes persist in the long run and follow participants much longer than three months as we did in this study.