Validity and reliability of Fitbit Flex for step count, moderate to vigorous physical activity and activity energy expenditure

Sushames, Ashleigh, Edwards, Andrew, Thompson, Fintan, McDermott, Robyn, and Gebel, Klaus (2016) Validity and reliability of Fitbit Flex for step count, moderate to vigorous physical activity and activity energy expenditure. PLoS ONE, 11 (9). e0161224. pp. 1-14.

[img]
Preview
PDF (Published Version) - Published Version
Available under License Creative Commons Attribution.

Download (625kB) | Preview
View at Publisher Website: http://dx.doi.org/10.1371/journal.pone.0...
 
107
1139


Abstract

Objectives: To examine the validity and reliability of the Fitbit Flex against direct observation for measuring steps in the laboratory and against the Actigraph for step counts in free-living conditions and for moderate-to-vigorous physical activity (MVPA) and activity energy expenditure (AEE) overall.

Methods: Twenty-five adults (12 females, 13 males) wore a Fitbit Flex and an Actigraph GT3X+ during a laboratory based protocol (including walking, incline walking, running and stepping) and free-living conditions during a single day period to examine measurement of steps, AEE and MVPA. Twenty-four of the participants attended a second session using the same protocol.

Results: Intraclass correlations (ICC) for test-retest reliability of the Fitbit Flex were strong for walking (ICC = 0.57), moderate for stair stepping (ICC = 0.34), and weak for incline walking (ICC = 0.22) and jogging (ICC = 0.26). The Fitbit significantly undercounted walking steps in the laboratory (absolute proportional difference: 21.2%, 95%CI 13.0–29.4%), but it was more accurate, despite slightly over counting, for both jogging (6.4%, 95%CI 3.7–9.0%) and stair stepping (15.5%, 95%CI 10.1–20.9%). The Fitbit had higher coefficients of variation (Cᵥ) for step counts compared to direct observation and the Actigraph. In free-living conditions, the average MVPA minutes were lower in the Fitbit (35.4 minutes) compared to the Actigraph (54.6 minutes), but AEE was greater from the Fitbit (808.1 calories) versus the Actigraph (538.9 calories). The coefficients of variation were similar for AEE for the Actigraph (Cᵥ = 36.0) and Fitbit (Cᵥ = 35.0), but lower in the Actigraph (Cᵥ = 25.5) for MVPA against the Fitbit (Cᵥ = 32.7).

Conclusion: The Fitbit Flex has moderate validity for measuring physical activity relative to direct observation and the Actigraph. Test-rest reliability of the Fitbit was dependant on activity type and had greater variation between sessions compared to the Actigraph. Physical activity surveillance studies using the Fitbit Flex should consider the potential effect of measurement reactivity and undercounting of steps.

Item ID: 46130
Item Type: Article (Research - C1)
ISSN: 1932-6203
Additional Information:

© 2016 Sushames et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funders: James Cook University (JCU)
Projects and Grants: JCU Research Infrastructure Block Grant
Date Deposited: 19 Oct 2016 02:45
FoR Codes: 42 HEALTH SCIENCES > 4207 Sports science and exercise > 420799 Sports science and exercise not elsewhere classified @ 34%
42 HEALTH SCIENCES > 4207 Sports science and exercise > 420701 Biomechanics @ 33%
42 HEALTH SCIENCES > 4206 Public health > 420699 Public health not elsewhere classified @ 33%
SEO Codes: 92 HEALTH > 9204 Public Health (excl. Specific Population Health) > 920499 Public Health (excl. Specific Population Health) not elsewhere classified @ 100%
Downloads: Total: 1139
Last 12 Months: 97
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page