What’s your number [of steps]? Systematic review evaluating the accuracy of Fitbit activity trackers

Whether you received the Fitbit as a Christmas gift, bought it in the new year hoping “this time would be different,” or it was doctor-recommended, many people rely on this small wrist-worn device for information about their health and daily habits.

The Fitbit activity tracker uses a microelectronic triaxial accelerometer [1], which measures vibration in three axes, allowing it to measure body movement in a three-dimensional space. Types of measurements include daily steps taken, energy expenditure, sleep patterns, distance covered, and time spent in various activities [1].

In a recent article published in JMIR mHealth and uHealth [1], researchers from British Columbia, Canada, performed a systematic review of 67 studies, assessing the measurement accuracy of Fitbits in both controlled and real-world settings.

Accuracy is affected by a variety of external factors

Authors found that there are a variety of factors that influence the accuracy of step counting in a controlled setting. The speed of someone’s motion and where the device is placed on their body, in addition to variations in the extent of body movements (exaggerated, constrained, or inconsistent), were found to affect accuracy of measurement.

In this study, in a controlled setting, Fitbit trackers were found to accurately measure steps 50% of the time. Acceptable accuracy was found 60% of the time when worn on the torso for slow, self-paced, and normal speeds; 70% of the time for ankle placement during slow or very slow walking speeds; and 90% of the time for wrist placement for jogging speeds. However, in all instances in a controlled setting, Fitbit tended to underestimate the step count.

In real-life setting as well, Fitbit trackers were likely to accurately measure steps 50% of the time in adults with no mobility issues, when worn on the torso or wrist, with a tendency to overestimate steps.

Fitbit devices were unlikely to provide accurate measures of energy expenditure and there is evidence that Fitbit overestimates time spent in moderate to vigorous physical activity in a free-living setting.

An important limitation to note is that the studies included in this review used various Fitbit models. Other studies [2,3] suggest that the latest models of activity trackers, including the Fitbit and other activity trackers, are more accurate than older models.

Conclusions important for consumers

Different models of the Fitbit which can be worn on different parts of the body will provide more accurate measurements depending on the type of activity one is hoping to use it for.

The relative accuracy of the Fitbit compared to reference device (research-grade device) does not discount the Fitbit’s intended purpose as a tool for daily activity and fitness self-monitoring.

Due to the inaccuracy of the Fitbit’s step count in people with limited mobility and at various speeds of movement, health care providers, and researchers wishing to use the Fitbit to measure outcomes in patients and participants respectively, should be aware of the mobility patterns of the users and the desired outcome measurement.

References

[1] Feehan LM, Geldman J, Sayre EC, Park C, Ezzat AM, Yoo JY, Hamilton CB, Li LC. Accuracy of Fitbit Devices: Systematic Review and Narrative Syntheses of Quantitative Data. JMIR Mhealth Uhealth 2018;6(8):e10527. http://mhealth.jmir.org/2018/8/e10527. [doi: 10.2196/10527]

[2]  Calabró MA, Stewart JM, Welk GJ. Validation of pattern-recognition monitors in children using doubly labeled water. Med Sci Sports Exerc 2013 Jul;45(7):1313-1322. [doi: 10.1249/MSS.0b013e31828579c3]

[3] Xie J, Wen D, Liang L, Jia Y, Gao L, Lei J. Evaluating the Validity of Current Mainstream Wearable Devices in Fitness Tracking Under Various Physical Activities: Comparative Study. JMIR Mhealth Uhealth 2018;6(4):e94. http://mhealth.jmir.org/2018/4/e94/. [doi: 10.2196/mhealth.9754]

Leave a Reply

Your email address will not be published. Required fields are marked *