Activity Monitor Accuracy in Assessing Caloric Expenditure in Obese Adults

Thursday, March 19, 2015
Exhibit Hall Poster Area 1 (Convention Center)
Justin Menickelli, Maridy Troy, Tom Watterson, Chris Cooper and Dan Grube, Western Carolina University, Cullowhee, NC

Background/Purpose:

One challenge for weight loss intervention programs is to accurately assess caloric expenditure.  Waist-worn, piezoelectric pedometers are accurate for measuring steps, but much less accurate for assessing caloric expenditure (Crouter et al, 2003).  Newly designed, wrist-worn triaxial accelerometers claim to accurately assess caloric expenditure.  This study examined the accuracy of the Nike FuelBand (NFB) and the Fitbit Force (FBF) in assessing caloric expenditure during motor-driven treadmill walking and running in obese adults (M BMI= 35.43, SD= 3.22). 

Method:

Nineteen participants (M age= 21.16 years, SD= 3.59) were tested on one day in three consecutive but separate sessions: 1) resting supine for 10 minutes, 2) walking on a treadmill at self-selected speed for 10 minutes, and 3) running on a treadmill at self-selected speed for 10 minutes.  During treadmill sessions, participants wore a NFB and FBF on the non-dominant wrist.  Estimated caloric expenditure from the NFB and FBF was recorded for the last five minutes of each testing session.  Metabolic data was measured using a TrueOne 2400 Metabolic Measurement System and energy expenditure in kcals was calculated based on the amount of oxygen consumed during the last five minutes of steady-state walking and running.    

Analysis/Results:

The mean self-selected speed was 80.36 m/min (SD= 5.33) for walking and 136.21 m/min (SD= 26.11) for running. 

Correlation coefficients between actual and estimated caloric expenditure for walking were -0.11 (NFB) and 0.28 (FBF); and for running were 0.64 (NFB) and 0.62 (FBF).

A repeated measures ANOVA determined caloric expenditure while walking differed significantly (F(2, 36)= 4.08, p= .025).  Post hoc tests using Bonferroni correction revealed that actual caloric expenditure did not differ significantly (p= 1.0) from estimated caloric expenditure for the NFB (31.84 ± 14.51 vs. 32.0 ± 9.92) but did differ significantly (p= .048) for the FBF (31.84 ± 14.51 vs. 41.32 ± 10.99).

A repeated measures ANOVA determined caloric expenditure while running did not differ significantly (F(2, 36)= .018, p= .982).  Post hoc tests using the Bonferroni correction revealed that actual caloric expenditure did not differ significantly from estimated caloric expenditure for either the NFB (73.53 ± 16.87 vs. 73.42 ± 12.65) or the FBF (73.53 ± 16.87 vs. 73.00 ± 10.66).

Conclusions:

Both the NFB and FBF accurately assessed caloric expenditure in obese adults during the running session. The results lend support to findings (Dondzila et al, 2012; Swartz et al, 2009) in which speed of locomotion was an impactful variable.