Secondary Logo

Fitbit And Actigraph: A Comparison Of Physical Activity And Sedentary Time In Overweight/obese Adults1353 Board #28 June 1 900 AM - 1030 AM

Szabo-Reed, Amanda N.1; Willis, Erik2; Ptomey, Lauren1; Washburn, Richard FACSM1; Donnelly, Joseph FACSM1

Medicine & Science in Sports & Exercise: May 2017 - Volume 49 - Issue 5S - p 366
doi: 10.1249/01.mss.0000517883.22655.db
C-31 Free Communication/Poster - Activity Trackers and Smartwatches Thursday, June 1, 2017, 7:30 AM - 12:30 PM Room: Hall F

1Kansas University Medical Center, Kansas City, KS. 2National Institute of Cancer, NIH, MD. (Sponsor: Joseph Donnelly, FACSM)


(No relationships reported)

There is limited information regarding the validity of consumer targeted wearable physical activity (PA) trackers, such as the Fitbit, to assess PA in free-living adults.

PURPOSE: To compare daily minutes of sedentary time, and moderate-to-vigorous (MVPA) colleted using two types of activity trackers, in a sample of overweight and obese adults (n = 12, BMI= 37±4 kg/m2, age = 48±12 yrs.) who completed a 6-mo. weight loss intervention (diet + self-directed PA) delivered using Facebook.

METHODS: Participants wore a Fitbit FlexTM on their non-dominant wrist continuously over 6 mos. They were also asked to wear an ActiGraph GTX1TM on their non-dominant hip for 7 consecutive days at both baseline and 6 mos. Fitbit and ActiGraph data for a minimum of one 10-hr. day at baseline and 6 mos. was required for inclusion in this analysis. Categorization of sedentary time and MVPA was derived from the Fitbit algorithm or using NHANES Actigraph cut-points (Troiano, 2008).

RESULTS: Sedentary time was significantly lower and MVPA was significantly higher when assessed with the Fitbit compared with the ActiGraph at both baseline (sedentary time: Fitbit = 427 ±168, ActiGraph = 639 ±132 min/d, p< 0.001; MPVA: Fitbit= 128 ±49, ActiGraph = 18 ±19 min/d, p <0.0001) or 6 mos. (sedentary time: Fitbit = 459 ±168, ActiGraph = 613 ±103 min/d, p= 0.001; MPVA: Fitbit= 123±53, ActiGraph= 14 ±10 min/d, p<0.0001). Change in both sedentary time and MVPA from baseline to 6 mos. was not statistically significant measured either with the Fitbit (sedentary time = +63 ±46, p =0.16; MVPA = -8±13 min/d, p = 0.72) or the ActiGraph (sedentary time = - 49±52, p =0.39; MVPA = -2±13 min/d, p = 0.48).

CONCLUSIONS: The Fitbit significantly underestimated sedentary time and overestimated MVPA when compared with the ActiGraph in a small sample. Changes in sedentary time and MVPA over 6 mos. were non-significant when assessed by either the Fitbit or ActiGraph. These observations suggest that the Fitbit, which is relatively inexpensive when compared with the ActiGraph, may be useful for assessing changes in sedentary time and MVPA in response to an intervention. However, the absolute values for sedentary time and MVPA assessed by the Fitbit are questionable, and worthy of additional investigation in larger samples of free-living adults.Funded by Kansas City Life Sciences Institute

© 2017 American College of Sports Medicine